   The nineteenth century is widely and rightly held to be the century in which
the mathematical revolution in logic achieved its breakthrough. W.V. Quine once
remarked that logic is an ancient discipline, but since 1879 it has been a great one.
Of course, 1879 marks the publication of Gottlob Frege’s Begriffsschrift, and 1870
and 1883 the appearance of Charles Peirce’s “Description of a Notation for the
Logic of Relatives” and “Note B: The Logic of Relatives”. Frege and Peirce are
the independent co-founders of modern quantification theory. Frege (1848–1925)
was a German and Peirce (1839–1914) an American (their contributions are chron-
icled in volume three of this Handbook, The Rise of Mathematical Logic: Leibniz
to Frege). Although Frege’s work was little recognized and little appreciated by
British logicians of the period — Russell was a late exception — important steps
toward the mathematicization of logic were taken in Britain. Augustus De Mor-
gan (1806–1871) made significant contributions to the logic of relatives, of which
Peirce took respectful heed, and also to probability theory, an interest in which
he did much to revive. Until 1847, De Morgan was virtually the lone force in the
algebraicization of logic. Then George Boole (1815–1864) published The Mathe-
matical Analysis of Logic which appeared on the same day as De Morgan’s Formal
Logic and in which he gives to logic a somewhat different algebraic twist.Boole also
did valuable work on probability theory. Even the lesser figure, William Hamilton
(1788–1856), had pertinent things to say about predicate quantifiers. But the fact
remains that the revolution in logic was not put “over the top” by British logicians
of the nineteenth century. Why, then, does the Handbook of the History of Logic
make a place for an entire volume on this subject? The answer is that there is no
better place than Britain to witness the demise of the old logic and the beginnings
of the new.
   Aristotle’s logic — the old logic — was an immense achievement. Aristotle
originated the logic of syllogisms, he gave some expression to a separate logic of
immediate inference, and he made a number of attempts to extend the syllogistic
to modal contexts. The syllogistic was Aristotle’s most complete logic. Indeed the
perfectability proof of the Prior Analytics is an almost sound demonstration of
something like the completeness (in the modern sense) of syllogistic logic. Over the
centuries, the syllogistic has had its critics. The Megarians and Stoics expressed
reservations, as did legions of smart mediaeval logicians. The Renaissance was not
a good time for logic, and Aristotle’s supremacy was tested (or at least questioned)
by the likes of Descartes and the Port Royal logicians, by Locke and Leibniz, by
Kant and Hegel. All the same, the theory of syllogisms was logic’s paradigm
viii                                       Preface
for over two millennia. It provided the essential framework which others would
attempt to reinterpret or supplement with additional insights.
    All this ended in the 19th century. This was the century in which the hegemony
of the syllogistic fell apart like a collapsing empire. It was not, however, a clean
death. It was a lingering demise. Some of the century’s more original logicians
still paid their obeisances to the syllogistic even while they were in fact giving it no
role, or no central role, in their own work. A case in point is J.S. Mill (1806–1873),
who plights his great admiration for the syllogistic in his Autobiography, and who
in A System of Logic, pretty much concedes deductive logic to Richard Whately
(1887–1963), who in turn concedes it to Aristotle. But everything else in that
significant work was directed to issues in which it can only have been obvious that
the syllogistic would have no place. What makes the 19th century interesting is
that while the syllogism is losing its paramouncy, logic launches itself in genuinely
new ways, not all of which flow directly to the waters that create the Peircean-
Fregean tsunami of mathematical logic. An important additional development is
the logic of science.
    Squarely in the idealist tradition was the great Romantic poet, Samuel Taylor
Coleridge (1772–1834). Hegel had taught that the absolute could be fully artic-
ulated, that everything worth knowing is knowable within a dialectical system
in which truth is immanent. Coleridge developed a logic which deviated from
Hegel’s in two principal ways. His logic is foundationalist rather than coherentist;
and the knowing subject is beyond conceptualization, and hence ineffable. Lest it
be thought that Coleridge’s logic is largely an historical curiosity, to say nothing
of its being an eccentricity on its author’s part, idealism was philosophically dom-
inant in 19th century Britain, and idealism retains a broadly Hegelian orientation
not only in philosophy, but in logic as well.
    George Bentham (1800–1884), a nephew of Jeremy, was a botanist of note,
who took an interest in jurisprudence and logic. His Outline of a New System of
Logic (1827) is a work of considerable importance. It expressly formulated for the
first time the idea of predicate-quantification, the priority of which over Hamilton
would in due course be established by Herbert Spencer, and is described by W.S.
Jevons (1835–1882) somewhat breathlessly as the most important discovery in
formal logic since Aristotle.
    Whately’s Elements of Logic (1826) is a solid and formally correct re-telling of
deductive logic in the Aristotelian tradition to which Bentham’s book was intended
to be a critique, and which Whately ignored in subsequent editions. Whately
played a large role in restoring logic to Oxford’s curriculum after a period of
shameful neglect. In his review of Whately, Mill defended the syllogistic against
what he considered encroachments of Scottish philosophers who had proposed its
displacement by inductive logic. Mill took this position notwithstanding that A
System of Logic (1843) itself made a substantial contribution to inductive logic,
and was the originator of the deductive-nomological model of explanation. We
see in this a blend of something old and something new. In matters deductive,
Mill sought no quarrel with the syllogistic. In matters non-deductive, he was
                                       Preface                                     ix
in the descendent class of Francis Bacon, as was Mill’s contemporary William
Whewell (1794–1866), and A System of Logic brims with efforts to get at the
logic of those inductions that underpin the experimental sciences. Despite his
satisfaction with Whately’s treatment of the old logic of deduction, Mill was in
process of empiricizing it. He sought for deduction the only certitude and the only
objectivity that a serious and deep commitment to empiricism could consistently
allow. And, in the spirit of the Stoic skeptics, Mill would allow to syllogisms no
non-circular place in human cogitation. Deductive logic would impose consistency
constraints on consequence-drawing, but it would not give us positive principles
of reasoning.
   Whewell’s approach to the logic of science differed significantly from Mill’s.
Philosophy of the Inductive Sciences, founded upon Their History (1840) expressly
rejects the hypothetico-deductive claim that scientific hypotheses are discovered
by mere guesswork, and argues that hypothetical entities lie properly in the ambit
of induction.
   The new work in deductive logic — the work of Boole, De Morgan, and Hugh
MacColl (1837–1909) — facilitated the drift away from the syllogistic towards
mathematical treatments the newly emerging symbolic logic. We see in this the
convergence of three factors: the erosion of the syllogistic paradigm, the mathe-
maticization of deduction, and the rise of inductive logic, not excluding the theory
of probability.
   Arguably Britain’s leading idealist, F.H. Bradley’s The Principles of Logic ap-
peared in the same year (1883) as Frege’s Grundlagen and Peirce’s “Note B: The
Logic of Relatives”. One might be forgiven for thinking that, next to Frege and
Peirce, Bradley is a backwards looking museum-piece. Bradley, the idealist, like
Mill, the empiricist, has a principled aversion to the formalization of reasoning,
but he also vigorously attacks the traditional syllogistic’s analysis of propositions
as a necessary composition of three elements or ideas — the subject-idea, the at-
tribute, and the joining of these two ideas. In so doing, he anticipates difficulties
Russell was to have in analyzing propositions as unities of mutually independent
constituents, difficulties which crop up in Frege’s attempt to preserve a sharp dis-
tinction between concepts and objects. Bradley also anticipates Russell’s theory
of descriptions, and Quine’s extension of it to names, arguing that logically proper
names are disguised general terms. Russell credits Bradley with the idea of an-
alyzing general propositions as conditionals. Bradley himself thought a further
reason to reject the traditional syllogistic was its failure to accommodate the logic
of relations. It is therefore simply a misconception to think of Bradley’s logic as
hide-bound to the old ways.
   Hugh MacColl was born in 1837, the first year of Victoria’s long reign. In a
number of papers he fostered developments that would take hold in the century
to come, in Boolean algebra, modal, free and paraconsistent logic, as well as di-
alogue logic. Notwithstanding that these papers were vigorously — some would
say unfairly — condemned by Russell, it is clear that MacColl anticipates the
analysis of strict implication, whose precedence C.I. Lewis had the characteristic
x                                       Preface
grace to acknowledge. More generally, it is now clear that MacColl’s modal logic
is equivalent to the system T of Robert Feys (1937) and the equivalent system of
Georg von Wright (1951).
   W.S. Jevons (1835–1882), best known as a pioneer of mathematical economics,
published in 1874 a substantial and important monograph on logic under the title
The Principles of Science. Its main achievement was the extension and clarifi-
cation of the theory of induction developed by Whewell (and criticized by Mill).
It knitted together a general theory of probability with the analysis of induction,
and is arguably the single most important contribution to logical theory in 19th
century Britain.
   Virtually every student of logic is familiar with Venn diagrams. They have
an even wider currency than the De Morgan equivalences of propositional logic.
In this, John Venn (1835–1882), a student of De Morgan’s, was anticipated by
the figures of Euler a century earlier and of Leibniz in the century before that,
and was rivaled by those of his contemporary Lewis Carroll. Of arguably greater
originality is Venn’s The Logic of Chance. Venn was concerned to find a conception
of probability that would leave room for freedom of human action. This was the
frequentist conception which, among other things, exercised a benign influence on
the development of mathematical statistics. His writings on deductive logic are
also important, containing early investigations of mechanized proofs.
   Lewis Carroll (1832–1898), author of the Alice stories, was an accomplished
puzzler and paradoxer. His Achilles paradox convincingly demonstrates the neces-
sity of not requiring principles of inference to appear as premisses in deductions
that they themselves validate. He is also thought by some commentators to have
originated the method of semantic tableaux.
   Before the passing of the century, Bertrand Russell (1872–1970), whose contri-
butions are described in detail in our companion volume, Logic From Russell to
Church, was largely innocent of developments in mathematics that would shortly
shape his and A.N. Whitehead’s (1861–1947) logical programme in the next cen-
tury. It is true that Russell reviewed for Mind Louis Courtarat’s De l’infini
math ́matique (1896), and thus became acquainted with set theory. He also knew
      e
Whitehead’s Universal Algebra (1898), which contains a version of Boole’s logic.
Russell’s conversion to the new logic appears to have occurred one August morn-
ing in 1900 at the First International Congress of Philosophy. Guiseppe Peano
(1858–1932) read a paper on forms of definition in mathematics, and bested Ernst
Schr ̈der in subsequent discussion. For the remainder of 1900, Russell developed
     o
the ideas that would animate Principles of Mathematics (1903).
   It hardly wants saying that the three themes of British logic in the 19th century
– the abandonment of the syllogistic paradigm, the mathematization of deduction,
and the advancement of inductive logic – are not all expressly present in the works
of most of the logicians of that country and time, although the first two are clearly
in Boole. These broad thematic developments were the products of a collective
efforts not always graced by individual clarity. But in their net affect these were
efforts that made the logic of the period something to sit up and take notice of,
                                    Preface                                     xi
both then and now.
   Once again the Editors are deeply and most gratefully in the debt of the vol-
ume’s superb authors. The Editors also warmly thank the following persons:
Professor Margaret Schabas, Head of the Philosophy Department and Professor
Nancy Gallini, Dean of the Faculty of Arts, at the University of British Columbia;
Professor Michael Stingl, Chair of the Philosophy Department and Christopher
Nicol, Dean of the Faculty of Arts and Science, at the University of Lethbridge;
Jane Spurr, Publications Administrator in London; Carol Woods, Production As-
sociate in Vancouver, and our colleagues at Elsevier, Senior Acquisitions Editor,
Lauren Schultz and Mara Vos-Sarmiento, Development Editor. For their excel-
lent advice and support the Editors also owe a special debt of gratitude to Ivor
Grattan-Guiness and Fred Wilson.
                                                                  Dov M. Gabbay
                                                           King’s College London
                                                                       John Woods
                                                   University of British Columbia
                                                                              and
                                                           King’s College London
                                                                              and
                                                         University of Lethbridge
This page intentionally left blank                          CONTRIBUTORS
Dov M. Gabbay
King’s College London, UK.
dov.gabbay@kcl.ac.uk
Michael E. Hobart
Western Washington University, USA.
hobartme@earthlink.net
Dale Jacquette
The Pennsylvania State University, USA.
dlj4@psu.edu
Ralph Jessop
University of Glasgow, Scotland, UK.
r.jessop@crichton.gla.ac.uk
Gordon R. McOuat
University of King’s College, Halifax, Canada.
gmcouat@dal.ca
William J. Mander
Oxford University, UK.
william.mander@hmc.ox.ac.uk
Tim Milnes
University of Edinburgh, Scotland, UK.
tmilnes@staffmail.ed.ac.uk
Amirouche Moktefi
IRIST (Strasbourg) and Archives Poincar ́ (Nancy), France.
                                          e
amirouche.moktefi@gersulp.u-strasbg.fr
Bert Mosselmans
Roosevelt Academy, The Netherlands.
b.mosselmans@roac.nl
Maria Panteki
Thessaloniki, Greece.
nioka@math.auth.gr
Shahid Rahman
University of Lille, France.
shahid.rahman@univ-lille3.fr
Juan Redmond
Univesity of Lille, France.
juanredmond@yahoo.fr
xiv                                Contributors
Joan L. Richards
Brown University, USA.
Joan Richards@brown.edu
Laura J. Snyder
St. John’s University, USA.
snyderl@stjohns.edu
David Sullivan
The Metropolitan State College of Denver, USA.
sullivad@mscd.edu
James van Evra
University of Waterloo, Cananda.
vanevra@watarts.uwaterloo.ca
Ard van Moer
Vrije Universiteit Brussel, Belgium.
Ard.Van.Moer@vub.ac.be
Charissa S. Varma
University of Toronto, Canada.
charissa.varma@utoronto.ca
Fred Wilson
University of Totonto, Cananda.
fwilson@chass.utoronto.ca
John Woods
University of British Columbia, Cananda.
jhwoods@interchange.ubc.ca
                          BENTHAM’S LOGIC
                  Gordon R. McOuat and Charissa S. Varma
At the grand age of 71, celebrated English botanist George Bentham (1800–1884)
found himself in an uncomfortable position during what should have been his rest-
ful retirement years. George Bentham’s successful botanical career, exemplified
by the canonical Handbook of the British Flora (1858) and with J. D. Hooker,
the Genera Plantarum (1862–1883), was grounded in pre-evolutionary taxonomic
styles and practices. These styles and practices valued careful descriptions and
definitions of taxonomic groups based on personal observations, and the care-
ful assessment the relevant taxonomic relationships that brought into sharp relief
species and genera boundaries. The source of George Bentham’s intellectual anx-
iety in early 1870s was the popular new theory of evolution by natural selection,
proposed by Charles Darwin (1809–1882) in his Origin of Species (1859) — a
theory that seemed to be settling into the taxonomic thinking of British natural-
ists, and politely ushering out the pre-evolutionary taxonomic styles and practices
that underpinned George’s work. Darwin’s theory replaced relations of similar-
ity with descent with modification, challenging the idea that species groups are
fixed and unchanging and threatening to undermine centuries of taxonomic tra-
dition. In his 1871 presidential address to the Linnean Society, George Bentham
voiced his concerns with Dawin’s new theory and its relationship to natural his-
tory: “. . . systematic biology has to a certain degree been cast into the background
by the great impulse to the more speculative branches of the science by the pro-
mulgation of the Darwinian theory”.1 There was a revolution afoot in natural
history, and George Bentham soon realised he was on the losing side.
   These pre-evolutionary taxonomic styles and practices formed George Ben-
tham’s “distinctive attitude” regarding the kind of data a botanist could use in
establishing valid taxonomic groups (such as choice of external morphological fea-
tures versus internal features of cell-level anatomy), as well as broader issues on
the nature of groups and the epistemological status of the types of relationships
between taxonomic groups.2 Throughout his botanical career, George Bentham
believed in sharply circumscribed and sufficiently recognisable natural groups.3 In
June of 1827, he drew a tentative connection between species and genera bound-
aries and genealogical relationships, and by 1856 he was prepared to maintain that
   1 George Bentham, “Presidential Address” Proceedings of the Linnean Society of London
1870–1871 (1817): xxxv.
   2 P. F. Stevens, The Development of Biological Classification: Antoine-Laurent de Jussieu,
Nature, and the Natural System. (New York: Columbia University Press, 1994): 105.
   3 P. F. Stevens, The Development of Biological Classification: Antoine-Laurent de Jussieu,
Nature, and the Natural System. (New York: Columbia University Press, 1994): 121.
Handbook of the History of Logic. Volume 4: British Logic in the Nineteenth Century.
Dov M. Gabbay and John Woods (Editors)
 c 2008 Elsevier BV. All rights reserved.
2                          Gordon R. McOuat and Charissa S. Varma
because genealogical relationships could be demonstrated for only a few genera-
tions, any emphasis on ancestral relationships in discussions of species and genera
boundaries were necessarily speculative.4 The theory of evolution by natural se-
lection, it seemed, challenged George Bentham’s view on the distinctiveness of
groups (and consequently the way taxonomists define taxonomic groups), and its
emphasis on descent with modification challenged the traditional taxonomically
relevant relationships.
   But this was not the only revolution in which George Bentham found himself.
Shortly after his first botanical publication — Catalogue des Plantes Indigenes
des Pyrenees et du Bas Languedoc (1825) — George Bentham wrote his Outline
of a New System of Logic (1827). Written during an 1826–1827 stay in London,
the Outline hit the press during a transitional period in a tumultuous chapter
of British logic, between the last years of Scholastic syllogistic logic and the early
years of algebraic logic. On the surface, the Outline appeared to be little more than
a punchy chapter-by-chapter critique of Richard Whately’s popular and polemi-
cal defence of Scholastic logic, Elements of Logic (1826). However, but beneath
the surface lay George Bentham’s preliminary work on definition, division, and
relations, work that reflected the caution and concern George had consistently
exercised in discussing all manner of theoretical or hypothetical matters.5
   The fact that issues involving definition, division, and relations figured in many
of the botanical debates around the time George Bentham was writing the Out-
line certainly played a role in determining which logical issues he would address.
However, it was the two subversive reform movements that employed logic in their
plan — the role of logic educational reform and the role of logic in the legal re-
form — that shaped his radical response. Surprisingly, it was not the connection
to these reform movements that propelled George Bentham’s contributions to the
history of logic into the spotlight. His logical contributions gained recognition
somewhat circuitously and many years after the Outline’s publication, largely as
a consequence of a priority dispute between Scottish philosopher William Hamil-
ton (1788–1856) and British mathematician Augustus De Morgan (1806–1871)
concerning the quantification of the Scholastic predicate.
   This priority dispute began in the late 1840s with the announcement that the
origins and innovations of algebraic logic were found in De Morgan’s lecture “On
the Structure of the Syllogism”, given 9 November 1846 to the Cambridge Philo-
sophical Society. In this lecture, De Morgan proposed a structure for a new form
of logic that included the quantification of the Scholastic predicate. Shortly there-
after, Hamilton and his supporters challenged De Morgan’s priority, claiming that
Hamilton had been teaching this new form of logic, including the quantification
of the predicate, in his Edinburgh classes as early as 1845. As the priority battle
raged on (mostly from Hamilton’s pen), English philosopher and political theorist
Herbert Spencer (1820–1903) and English economist and logician William Stanley
   4 See Bentham Papers, University College London Library ci. 83–87(14–15 June 1827).
   5 P. F. Stevens, “Bentham, George (1800–1884),” in Oxford Dictionary of National Biography,
ed. H. C. G. Matthew and Brian Harrison (Oxford: OUP, 2004).
                                       Bentham’s Logic                                       3
Jevons (1835–1882) complicated matters further by not only announcing in 1873
in the Contemporary Review that this new system of logic had been spelled out
some time earlier by a young botanist by the name of George Bentham, but that
Hamilton had reviewed George Bentham’s Outline in the Edinburgh Review in
1833.6 Our retiring old botanist, it seemed, found himself an unwitting and early
participant in a second revolution, this time in logic.
   George Bentham’s presentation of the quantification of the predicate years be-
fore Hamilton, may have given him his ticket into the history of logic. However,
because histories of logic covering this period tend to structure their narratives on
the development of algebraic logic and the relationship between mathematics and
logic, a quick review of George Bentham’s life makes it is easy to appreciate why
even though he had a ticket, he took a back row seat.7
   Unlike the usual suspects cited in this logical revolution, George Bentham did
not run in mathematical circles and George’s lack of a strong connection to math-
ematical or logical circles makes it hard to establish him as a precursor to the in-
novations that could characterise algebraic logic. His interest in logic was sparked
when he was a teenager, at the industrious hands of his utilitarian uncle, the
formidable Jeremy Bentham (1748-1832). George Bentham’s affair with logic was
decidedly brief, and more often than not, logic played the handmaiden to George’s
botanical thought. Moreover, George Bentham’s first (and last) original logical
publication, Outline of a New System of Logic (1827), did little to establish his
reputation as a logician in logic and mathematical circles. The Outline sold a mere
sixty copies before the press went bankrupt and the remaining copies destroyed.8
For the most part, the Outline was not widely read, and its reviews were at best
lukewarm. Perhaps more importantly, George Bentham did not see his major log-
ical innovations as stemming from mathematical problems or puzzles. Instead, he
followed in the footsteps of the educational, legal, and botanical reformers, seeing
his innovations in logic as helping to solve more general problems in classification.
George Bentham believed his new logic would be more consonant with what peo-
ple — be they educators, lawyers, or naturalists — needed to do in the business
of organising knowledge.
   This English botanist, and for a time promising logician, spent a lifetime holding
fast to ideals of clarity and precision in language and cautioning against speculation
in science. It was these ideals and cautions that placed him on wining and losing
   6 See T. S. Baynes, “Mr. Herbert on Sir Wm. Hamilton and the quantification of the predi-
cate” Contemporary Review 21 (1873): 796-798; W. S. Jevons, “Who discovered the quantifica-
tion of the predicate?” Contemporary Review 21 (1873): 821-824.; and H. Spencer, “The study
of sociology IX — the bias of patriotism” Contemporary Review 21 (1873): 475-502. For the
debate see G. McOuat “The Logical Systematist: George Bentham and His Outline of a New
System of Logic.” Archives of Natural History 30 (2003): 206, M. Filipiuk, ed. George Bentham,
Autobiography 1830-1834. (Toronto: University of Toronto Press, 1997), 484–485.
   7 For example, Kneale and Kneale do not mention George, Styazhkin and Bochenski mention
only his quantification of the predicate.
   8 The press went bankrupt and Bentham was not inclined to rescue the remaining copies.
M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of Toronto
Press, 1997), 271.
4                           Gordon R. McOuat and Charissa S. Varma
sides of two revolutions. The tale of his life, his influences, and his innovations
may perhaps shed some light on the connections between logic and natural history,
and logic in legal and educational reform in Britain during the first half of the
nineteenth century.
   Born 22 September 1800 in Stoke, Plymouth, to Samuel Bentham (1757–1831)
and Mary Sophia Fordyce (1765–1858), George was second son, and third of five
children. George spent the first years of his life in England, and the Benthams
moved frequently between 1805 and 1814, eventually settling in France in 1814,
citing reasons of health, finance, and their children’s education.9 Although none of
the Bentham children had formal schooling, their parents took their education very
seriously. His mother Mary assumed the bulk of this responsibility and completely
supervised their education.10 Both parents encouraged the study of mathematics,
in addition to Latin and Greek.
   George’s most intense period of mathematical study was likely during the first
few years in France. George recalled that his father’s “first care” upon arriving in
France was to find George and Samuel junior a tutor to continue their education.
Samuel secured the services of a Monsieur Chiron “one of the professors at the
college of Saumur, who was at once a good mathematician, and Latin and to
some extent Greek scholar”.11 George recalled happily working through Laplace’s
arithmetic and algebra with his older brother during the winter of 1814, and fancied
himself “pretty well advanced in mathematics, having gone through Euclid and
plane trigonometry, and simple and quadratic equations in Algebra, . . ., spherical
trigonometry, conic sections and fluxions.”12
   After only two years in France, tragedy struck the Bentham household. Samuel
junior died just after his seventeenth birthday, as a result of blood clot triggered
by a fall from a swing in their garden. Though separated by about two years, the
Bentham boys had been inseparable in almost everything else. George recollected
later in his life that “until [Samuel junior’s] illness, I had always slept in the same
room, taken all my lessons with him, gone through the same exercises with the
same books”.13 Not surprisingly, after his brother’s death, George lost the desire
to engage in those activities they once shared, studies in classics and mathematics
were the hardest affected.14
   9 See Catherine Pease-Watkin, “Bentham, Samuel (1757–1831),” in Oxford Dictionary of Na-
tional Biography, ed. H. C. G. Matthew and Brian Harrison (Oxford: OUP, 2004)
  10 See Catherine Pease-Watkin, “The Influence of Mary Bentham on John Stuart Mill,” Journal
of Bentham Studies 8 (2006).
  11 M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of
Toronto Press, 1997), 17.
  12 M. Filipiuk, ed. George Bentham, Autobiography 1830-1834. (Toronto: University of
Toronto Press, 1997), 18.
  13 M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of
Toronto Press, 1997), 29.
  14 Pease-Watkin notes that “The Benthams suspected this outcome for some months, but choice
to hide their suspicions from the children. Samuel wrote to Jeremy stating that ‘till all was over
we concealed all apprehensions from the other children as well as from himself”’ Catherine Pease-
Watkin, “Jeremy and Samuel Bentham — The Private and the Public,” Journal of Bentham
Studies 5 (2002).
                                          Bentham’s Logic                                    5
   Rather than force George to pursue classics and mathematics during this diffi-
cult period, his parents encouraged him to turn attention to other subjects that
had once brought him pleasure, specifically history and geography.15 George im-
mersed himself in history and geography, using of his “natural taste for method
and arrangement” to organise geographical and statistical information into ta-
bles.16 George later credited Jeremy with encouraging and developing his interest
in method and arrangement. And their shared interest in tabulations during this
period did not pass unrecognised. Prussian naturalist, romantic philosopher, and
friend of Samuel Bentham, Alexander von Humboldt (1769–1859) recalled:
       it was his [George Bentham’s] own “natural taste for method and ar-
       rangement, stimulated by uncle’s example and the perusal of some of
       his works” that made him enjoy “tabulating the geographical and sta-
       tistical information . . . as to physical geography, mountain elevations,
       river courses and their basins, etc...17
   Before long, George’s interest in tabulation and classification shifted away from
history and geography, and towards botany.
   In 1817, the same year Jeremy’s work on education reform, the Chrestomathia
(1817), was published, Mary Bentham purchased the third edition of the popular
Flore Fran ̧aise (1815) by French naturalist Jean-Baptiste Lamarck (1744–1829)
              c
and French botanist Augustus Pyrasmus de Candolle (1778–1841). George’s in-
terest was immediately piqued.18 He later wrote:
       I was struck by the analytical tables for the determination of plants,
       which fell in with the methodical and tabulate ideas I had derived from
       the study of some of my uncle’s works and from what I had attempted
       in geography and statistics. . . 19
It was not simply the presence of tables in this botanical work that struck him.
George noted that the tables were arranged according to a methodology he had
  15 George  wrote:
       Samuel realised the impact that his brother’s death would have on George, writing
       of ‘poor George’ and ‘the loss of a brother to whom the attachment was as strong
       as can be’. Indeed many years later, in 1827, George was to write to his elder
       sister Mary Louisa: ‘It is a sad thing to think how those whom I have most loved
       and confided in have been separated from me, my poor brother whom I had never
       quitted a single day till his last fatal illness’
See also pages 29 and 33 of M. Filipiuk, ed. George       Bentham, Autobiography 1830–1834.
(Toronto: University of Toronto Press, 1997), 262.
  16 M. Filipiuk, ed. George Bentham, Autobiography       1830–1834.   (Toronto: University of
Toronto Press, 1997), 30.
  17 M. Filipiuk, ed. George Bentham, Autobiography       1830–1834.   (Toronto: University of
Toronto Press, 1997), xxiv.
  18 M. Filipiuk, ed. George Bentham, Autobiography       1830–1834.   (Toronto: University of
Toronto Press, 1997), 36.
  19 M. Filipiuk, ed. George Bentham, Autobiography       1830-1834.   (Toronto: University of
Toronto Press, 1997), 36.
6                          Gordon R. McOuat and Charissa S. Varma
come to recognise in his uncle’s legal and educational reform writing. What Jeremy
was doing with legal terms, George saw Lamarck and de Candolle doing with
plants. George would later include this observation in his translation of the ap-
pendix to Jeremy’s Chrestomathia.20
   George’s reading of Flore Fran ̧aise led to a small flurry of reading in more
                                       c
philosophical issues in botany, including de Candolle’s Th ́orie  ́l ́mentaire de la
                                                                   e      ee
botanique (1813) — a work that tackled questions of the theoretical status of
classification systems. His reading of Th ́orie  ́l ́mentaire de la botanique led to
                                               e       ee
his reading of the work of de Candolle’s teacher, Pierre Prevost (1751–1839). De
Candolle credited his “logical turn of mind” to Prevost and saw this as leading
to the success of the Th ́orie  ́l ́mentaire.21 In his lectures, Prevost stressed the
                            e     ee
importance of forming the genera in accordance with the real value of relations,
a view Prevost felt could easily apply to the life sciences.22 Attempts by Prevost
and others to put classification on a solid theoretical foundation resonated with
what George was reading in the Chrestomathia. Specifically, what George saw
in Jeremy’s work was an attempt to model educational reform on natural his-
tory. Jeremy’s interest in logical relations, and the relationship Jeremy assumed
between classification and knowledge came out most clearly in his appendix on
nomenclature and classification.
   In 1819, George began what would become an almost four year project — the
French translation the appendix in Jeremy Bentham’s Chrestomathia. For the first
few years, George dedicated little time to this project, perhaps because in 1820,
the Benthams bought the chˆteau de Restincli`res, near Montpellier. The plan was
                                a                   e
to cultivate the land for profit, and Samuel gave George most of the responsibility
of the management of the operation. However, in addition to managing the estate,
George read Scottish philosopher Dugald Stewart’s (1753–1828) scathing attack
on Aristotelian logic and French philosopher Jean le Rond d’Alembert’s (1717–
1783) encyclopaedia, the latter being a work praised by Jeremy, with regard to
the classification systems presented.23 By 1822, George’s work on the translation
picked up speed and he was able to have it ready in time for his visit to London
in 1823. George published his translation of Jeremy’s Appendix, titled “Essai sur
la nomenclature et classification”, en route to London at the Bossanges Press in
Paris. Delighted with his nephew’s efforts, Jeremy encouraged George to expand
on what he felt were substantial and significant additions.
   In spite of his uncle’s encouragements, George was more eager to pursue botany
  20 George would also include Dum ́ril’s Zoologie philosophique (1808) in his list of those that
                                   e
use bifurcating divisions.
  21 J. M. Drouin, “Principles and Uses of Taxonomy in the Works of Augustin-Pyramus de
Candolle.” Studies in History and Philosophy of Biology and Biomedical Sciences 32 (2001):
258.
  22 J. M. Drouin, “Principles and Uses of Taxonomy in the Works of Augustin-Pyramus de
Candolle.” Studies in History and Philosophy of Biology and Biomedical Sciences 32 (2001):
258.
  23 M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of
Toronto Press, 1997), 213.
                                        Bentham’s Logic                                        7
and spent more time meeting up with old friends and botanists than expanding
on his uncle’s logic during this visit.24 Upon returning to France, George resumed
his duties as estate manager and devoted almost all of his time to botany, and by
1825, he deemed himself “thoroughly botanical.”25 His botanical efforts centred on
building up the collection he had begun five years earlier, a collection from which
he wrote the book that would establish his reputation as a serious botanist — Cat-
alogue des plantes indigenes des Pyrenees et du Bas Languedoc.26 Unfortunately,
bad luck befell the Benthams in 1826, George found himself seeking assistance
from his influential uncle, a situation that would direct his attention once more to
issues in logic. 27
   In the first week of August 1826, George travelled to London with his sisters,
with the hope of securing his uncle’s financial support. When it became clear that
he was not going to get financial support from Jeremy and would have to earn a
living, George decided to enrol in Lincoln’s Inn and train as a lawyer — a decision
that infuriated Jeremy.28 Partly to pacify his irate uncle, George agreed to defer
his legal studies and become his uncle’s amanuensis. And so began George’s work
on Jeremy’s unruly unpublished logic papers.29
   Perhaps as an incentive and perhaps to incite interest in these papers, Jeremy
intimated that both James Mill (1773–1836) and his son John Stuart Mill (1806–
1873) had studied these papers and expressed an interest in editing them. George,
however, suspected that the real reason his uncle gave him this particular task was
as an alternative means of compensating him. George reported to his sister that:
        [M]y uncle has imagined that he makes my fortune in giving me his
        logical papers to make a book of; . . . if I succeed in putting it into
        intelligible French, Bossanges [the publisher] may give me something
        for it — but as for a fortune, if I get any from my Uncle it must be in a
        more direct way than through the medium of his manuscripts. Besides,
        he wishes me to do them in English.30
However, it is just as likely that George’s “Essai” proved to Jeremy that his
nephew’s training as a botanist coupled with his appreciation of the theoretical
  24 See M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of
Toronto Press, 1997), 98–153.
  25 M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto : University of
Toronto Press, 1997), 97 and 455.
  26 M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of
Toronto Press, 1997), xxx.
  27 The cultivation of Restincli`res was reasonably successful for a time, but in the end for
                                 e
various reasons, the family returned to England. One factor was the threat of a lawsuit from
neighbouring residents, who objected to Samuel’s irrigation system, which, they claimed, was
diverting the local water supply. See M. Filipiuk, ed. George Bentham, Autobiography 1830–
1834. (Toronto: University of Toronto Press, 1997), 234.
  28 For George’s account of Jeremy’s anger, see M. Filipiuk, ed. George Bentham, Autobiography
1830-1834. (Toronto: University of Toronto Press, 1997), 246–7.
  29 In addition, George worked on his uncle’s papers on Codification and other legal subjects.
  30 M. Filipiuk, ed. George Bentham, Autobiography 1830-1834. (Toronto: University of
Toronto Press, 1997), xxxi.
8                          Gordon R. McOuat and Charissa S. Varma
problems would make George an ideal candidate to champion a Benthamite logic.
And Jeremy Bentham was right.
   While struggling to find some order in Jeremy’s logic papers, helping his own
father, and pursuing a legal career, George began the Outline — a brief and biting
Benthamite critique of Richard Whately’s Elements of Logic (1826). By March
1827, George Bentham had completed his Outline, and his proud uncle assumed
the printing costs.31
   The object of George’s criticism — Whately’s Elements — was the first signif-
icant logic textbook to grace the British University system since Aldrich’s 1691
Logicae Artis Compendium. Up until Whately’s Elements, the two main textbooks
used at Oxford and Cambridge were Sanderson’s 1615 Logicae Artis Compendium
and Aldrich’s 1691 Logicae Artis Compendium. As it turned out, the Elements
was far more than a textbook. Whately was clear about the polemical nature of
his book right from the start. Whately had a religious agenda to push — wanting
to encourage and promote sound reasoning in religious men to counteract religious
scepticism. Whately wrote: “The adversaries of our Faith would, I am convinced,
have been . . . more satisfactorily answered . . . had a thorough acquaintance with
logic been more common than it is”.32 Whately also had an educational reform
agenda to push. Whately saw logic as an agent of university reform and argued
that logic should be compulsory for candidates at Oxford for academic honours.
   In her biography of her father, Whately’s daughter Jane reflected on the role of
the Elements in his educational reform. She writes:
        The task undertaken by [Whately] was one of no ordinary difficulty; it
        was not the originating of a new science, but the resuscitation of an
        old and half-defunct one. The study of logic, formerly pursued with
        great and credible devotion, had, in later years, fallen into disrepute
        among the more intellectual class in the University. It was pursued in
        schools in Oxford merely by committing to heart the technical rules
        of the compendium of Dr. Aldrich. These were by no means without
        their utility as a tough mental exercise, and many an Oxonian might
        remember with gratitude the edge which it gave to his powers of rea-
        soning, particularly if unacquainted with the more valuable discipline
        of mathematics. It was Whately’s great and eminently successful effort
        to raise the study from this inferior condition to something approaching
        a scientific character.33
   The impact of Whately’s Elements was immediate, and its effects were enduring.
In his Historical Sketch of Logic (1851), published less than twenty-five years after
the Elements, Blakey recounted the influence of Whately’s book:
  31 M. Filipiuk, ed. George Bentham, Autobiography 1830-1834. (Toronto: University of
Toronto Press, 1997), xxi.
  32 R. Whately, Elements of Logic, Comprising the Substance of the Article in the Encyclopaedia
Metropolitana; with Additions, &c, (London, B. Fellowes 1831), xxviii.
  33 E. J. Whately, Life and correspondence of Richard Whately, D.D., 2 vols. (1866), 49.
                                         Bentham’s Logic                                        9
        Archbishop Whately’s Elements of Logic, is one of the most important
        and influential logical publications of modern times. It is an able and
        popular exposition of the scholastic logic; and has, in fact, been the
        main instrument in producing the revival of the syllogistic system in
        Great Britain. The work has gone through many editions, and is used,
        more or less, in several seats of learning, as an ordinary text-book for
        logical students.34
In many respects, the Elements owed its existence to the encouragement of
Whately’s friend, teacher, and mentor Edward Copleston, on whose advice Whately
launched his first spirited defence of traditional syllogistic logic in an 1823 article
in the Encyclopaedia Metropolitana. This article was later expanded considerably
and republished in 1826 as Elements of Logic.
   The Elements was regarded as the last successful attempt to breathe life into
a dying discipline. The area of Scholastic logic that benefited the most from
Whately’s pen was the syllogism. Until Whately, the syllogism had not been far-
ing well. Since the seventeenth-century, students educated in the British system
were taught that the syllogism was a better instrument of enquiry and proof than
induction (the syllogism to remain safely on the bookshelves of Oxford and Cam-
bridge), by the close of the eighteenth century, it was becoming clear syllogistic
logic was not flourishing — it was festering.
   Some claim the decline of logic during this period reflects the poor quality of
the logic compendiums in England used in the universities.35 At least one recent
explanation for the poor quality compendiums during this period argued that it
languished because it was not yet a “science”, not yet having a clear theoretical
framework. Van Evra wrote in his “The Development of Logic as Reflected in the
Fate of the Syllogism 1600-1900.” (2000):
        In the early portion of the seventeenth century, logic displayed features
        commonly found in disciplines prior to the emergence of a dominant
        theoretical framework. At the time, the common logic had neither
        a secure theoretical structure externally imposed by pure tradition,
        nor the stabilizing influence of a strong internal theory. In such a
        vacuum, all features of the subject become, potentially at least, equally
        relevant.36
Other histories emphasise the “silencing” of professors and replacing them with
college fellows as being responsible for logic’s decline, claiming that although tutors
became more popular, they lacked the expertise or skill of logic professors and did
  34 R. Blakey, Historical Sketch of Logic: From the Earliest Times to the Present Day (Bailli`re,
                                                                                              e
1851), 454.
  35 See E. J. Ashworth, “Some Notes on Syllogistic in the Sixteenth and Seventeenth Centuries”,
Notre Dame Journal of Formal Logic, XI, no. 1 (1970) 17-33.
  36 Van-Evra James, “The Development of Logic as Reflected in the Fate of the Syllogism 1600-
1900.” History and Philosophy of Logic 21 (2000): 118.
10                           Gordon R. McOuat and Charissa S. Varma
not produce original works during this period, thus causing of the decline of logic.37
   During the seventeenth and eighteenth centuries, logic also came under attack.
Many Renaissance and humanist thinkers of the seventeenth century began to
raise concerns about Scholastic logic’s status as an art and its monopoly on truth.
Promoters of the empirical science of the Scientific Revolution, such as Bacon
and Locke, questioned the application of syllogistic logic to the sciences. One of
their chief concerns was the ability of logical demonstrations to enable discoveries.
Backing the concerns raised by Bacon and Locke, were the “Common Sense”
School of Philosophers in Edinburgh that rose to prominence in the eighteenth
century. These philosophers began to call for educational reform and logic was
one of the subjects on the chopping block. Dugald Stewart in particular, called
for the abolishment of syllogistic logic in the curriculum.
   For Whately, the first step in restoring the logic to its proper place involved re-
sponding to the objections raised from the seventeenth century onwards concerning
logic’s purpose. One of his goals involved debunking the Scholastic assumption
that logic was the art of thinking. Whately’s argument that logic was not simply
an art, but also a science, served two purposes. First, it provided Whately with a
new and powerful response to the seventeenth-century objections to the usefulness
of logic. For Whately, the objections of Bacon, Locke, the Scottish Common Sense
School, and Watts have at least one thing in common — they targeted not the
tools of logic, but their use. Logic was not used to investigate nature (speaking
to Bacon and Locke), it was not the instrument of truth (the Scholastic version),
and it was not the art of rightly employing the rational faculties. Second, the
claim that logic is also a science marks the beginning of a new approach to logic.38
Although Whately was not the first to claim logic as science, his was the first
influential British logic textbook to defend this claim explicitly.39
   Whately’s claim that logic’s goal as a science involved providing “the gener-
alised and abstract representation of all demonstration”, was also the first step
in Whately’s new account of the syllogism. The Scholastics understood the syl-
logism as kind of argument. On Whately’s account, the syllogism was a purely
formal device. Once axiomatized, Whately believed that syllogisms would serve as
a canonical test of the validity of actual arguments, making the syllogism the theo-
retical core of the science of logic.40 From this perspective, Whately’s redefinition
of logic and his re-evaluation of the role of the syllogism have been credited as
  37 Blakey credits this history to Hamilton in his 1833 Edinburgh Review. R. Blakey, Historical
Sketch of Logic: From the Earliest Times to the Present Day (Bailli`re, 1851), 424.
                                                                     e
  38 See J. Van Evra, “Richard Whately and the rise of Modern Logic”, History and Philosophy
of Logic 5, (1984), 1-18; and C. Jongsma, Richard Whately and the Revival of Syllogistic Logic
in Great Britain in the early Nineteenth Century (unpublished dissertation, Toronto 1982).
  39 Blakey notes that Kirwan, in his 1807 “Logic” volume 1 page 1 makes this same claim, see
R. Blakey, Historical Sketch of Logic : From the Earliest Times to the Present Day (Bailli`re,
                                                                                            e
1851), 449. Also see J. Van Evra, “Richard Whately and the rise of Modern Logic”, History and
Philosophy of Logic 5, (1984) 1-18.
  40 See J. Van Evra, “Richard Whately and the rise of Modern Logic”, History and Philosophy
of Logic 5, (1984) 1–18.
                                     Bentham’s Logic                                      11
inspiring the direction of the pioneers of algebraic logic.41 The syllogism, thanks
to Whately, was given a whole new lease on life.
   On the new definition of logic, George agreed wholeheartedly with Whately. In
fact, George began his Outline by claiming:
        In reading the elegant exposition of his views, which Dr. Whately has
        prefixed to his Elements of Logic, I felt that I generally concurred in his
        observations on the utility of Logic, in his refutation of the arguments
        of its detractors — of those who set up Common Sense in opposition to
        Logic, — and in his remarks on the erroneous system proceeded upon
        with regard to this subject in our University Education. The absurdity
        of comprehending, within the province of Logic, every branch of art or
        science to which it may be applicable, will readily be admitted by any
        reader.42
George continued this discussion of the definition of logic as both an art and science
in the beginning of chapter two, where he noted that his uncle had presented a
similar definition of logic in his Chrestomathia a few years before Whately. On
Jeremy’s view:
        [e]very art had a correspondent science: it was a mistake to think
        that the field of thought and action could be divided into a series of
        distinct compartments, some containing an art, some a science, and
        some containing neither the one nor the other. The fact was that,
        ‘Whatsoever spot is occupied by either, is occupied by both: it is
        occupied by them in joint-tenancy.’ The distinction was founded on the
        distinction between practice and knowledge: ‘Practice, in proportion as
        attention and exertion are regarded as necessary to due performance,
        is termed art: knowledge, in proportion as attention and exertion are
        regarded as necessary to attainment, is termed science’. There was no
        ‘determinate line of distinction’ between an art and its correspondent
        science, but where ‘that which is seen to be done’ was regarded as being
        more prominent than ‘that which is seen or supposed to be known’,
        the more likely it was that it would be considered an art, and in the
        opposite case a science.43
The accolades for this new definition of logic in the Outline, did not last long. By
annexing logic to “every branch of human knowledge” George felt Jeremy provided
too broad definition, and by restricting the science aspect of logic to “mere correct
reasoning”, George felt Whately provided too narrow a definition. By recasting
  41 See J. Van Evra, “Richard Whately and the rise of Modern Logic”, History and Philosophy
of Logic 5, (1984) 1-18.
  42 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 1.
  43 J. Bentham, Chrestomathia. ed M. J. Smith and W. H. Burston. (Oxford: Clarendon Press
1983), 59–60.
12                         Gordon R. McOuat and Charissa S. Varma
the definition of logic as “the branch of art-and-science which has for its object
the advantageous application of the human mind to the study of any other branch
of art-and-science” George felt his attempt to make logic entirely general, like a
universal grammar, would be just right.44
   By page five of his introductory remarks, George identified another serious flaw
in Whately’s logic, and his reaction provides the first real taste of a Benthamite
logic. George charged Whately with the failure to distinguish between different
types of entities. Though he raised the concern on page five, George developed
this criticism in Chapter Three, titled “Analytical Outline”. George claimed:
        The remainder of this Analytical Outline is devoted to the definition
        of the processes of abstraction and generalization, which are here very
        aptly distinguished, and to the very useful exposure of the common er-
        ror of ascribing reality to generic terms. This should have been carried
        still farther; he should have exhibited the pernicious effects resulting
        from the realization of those subject matters which D’Alembert first
        called ˆtres fictifs, and what Mr. Bentham has described under the
                e
        name of fictitious entities. He should have pointed out the constant
        but unavoidable fiction which must enter into the composition of any
        discourse, and should have been given some indications by which error,
        in this respect, may be guarded against.45
A significant part of this chapter (fifteen of the twenty-one pages) was dedicated
to presenting a Benthamite classification of entities. George’s account of entities,
as well as his critique of Whately’s discussion of exposition in chapter six, came
straight from Jeremy’s reform work and his unpublished logic papers.
   George was quick to recognise that Whately had fallen into the same trap that
Jeremy claimed snared other significant thinkers. George followed Jeremy in his
claim that explorations into the processes of “abstraction” and “generalization”
were useful in uncovering the error of ascribing reality to generic terms, but such
investigations should have been pushed further to include a distinction between
real and fictitious entities.46
   Jeremy’s distinction between real and fictitious entities marked his first decisive
break from Scholastic logic, one that came early in his career and underpinned some
of his objections in his legal reform work. One of Jeremy’s first targets during his
campaign for legal reform was Oxford’s first Vinerian Professor of Common Law
and Jeremy’s teacher, Sir William Blackstone (1723–1780) and part of Jeremy’s
attack made use of this distinction between real and fictitious entities.47 In his
  44 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 14.
  45 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 29.
  46 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 29.
  47 Blackstone is remembered for, among other things, establishing English law as an academic
discipline. He presented course of private lectures on law at Oxford, which Jeremy attended,
                                         Bentham’s Logic                                        13
Comment on the Commentaries and A Fragment on Government (1776), Jeremy
fixed his radical eye on Blackstone’s Commentaries on the Laws of England (1765-
9), aiming to expose what he saw as a two-fold problem: Blackstone’s application of
the Scholastic method of definition to the terms of jurisprudence, which prompted
Jeremy to present his new classification of entities and his new modes of exposition;
and Blackstone’s conflation of the role of censor and expositor in discussions of
law, which prompted Jeremy to present his new account of methodization.48
   Jeremy identified the first mistake in Blackstone’s discussion of terms of ju-
risprudence, specifically his concept of a “natural right”, was to define the terms
of jurisprudence using the Scholastic method of definition — per genus et differ-
entiam. Definition, on this account, involved first identifying the group to which
the entity in question belongs (or determining the entity’s logical genus), and then
distinguishing the entity under investigation from the others in the group by spec-
ifying the property that makes it what it is and not something else in the group
(or identifying the object’s logical differentia). Blackstone followed the Scholastic
belief that by assigning the logical genus and differentia, a definite meaning is
conveyed, giving “a clear idea of the thing it signifies.”49 The Scholastic method
of definition may be fine for terms like “turnip” and “table”, but for terms such
as “obligation”, Jeremy saw a problem.
   Following Locke, Jeremy claimed that to obtain a “clear idea,” involved ei-
ther having direct sense experience of the entity (a Lockean “simple idea” of a
substance) or by constructing “artificial groupings of sensory ideas” (a Lockean
“complex idea” of a substance). The problem is that even though terms such
as “obligation”, hold the subject position in a proposition, they do not refer to
an entity accessible by direct sense experience, or for which an artificial grouping
could be constructed. Language, for Jeremy, obscured a fundamental distinction
between entities that are real and entities that are fictitious.50
   To help with the exposition of tricky terms, like those of jurisprudence, Jeremy
believed we first needed to recognise that they are fictitious entities and that
fictitious entities cannot be exposited by the method per genus et differentiam.
He states this clearly in the Fragment on Government:
        The common method of defining — the method per genus et differ-
        entiam, as logicians call it, will, in many cases, not at all answer the
        purpose. Among abstract terms we soon come to such as have no su-
        perior genus. A definition, per genus et differentiam, when applied to
and these lectures were later published as Commentaries on the Laws of England (1765–1769).
Blackstone would later to be a justice of the Court of Common Pleas. See Wilfrid Prest, “Black-
stone, Sir William (1723–1780),” in Oxford Dictionary of National Biography, ed. H. C. G.
Matthew and Brian Harrison (Oxford: OUP, 2004).
  48 Blackstone’s confusion of the roles of expositor and censor, see P. Schofield “Jeremy Bentham,
The Principle of Utility, and Legal Positivism,” Current Legal Problems 56 (2003): 1–39.
  49 J. Bentham, ‘A Fragment on Government’, in A Comment on the Commentaries and A
Fragment on Government, ed. J.H. Burns and H.L.A. Hart (London, 1977), 587.
  50 See P. Schofield “Jeremy Bentham, The Principle of Utility, and Legal Positivism,” Current
Legal Problems 56 (2003): 14.
14                           Gordon R. McOuat and Charissa S. Varma
        these, it is manifest, can make no advance: it must either stop short,
        or turn back, as it were, upon itself, in a circulate or a repetend.51
Because the names of real entities refer to entities accessible by sense perception
and can be arranged hierarchically (that is, they can be arranged in descending
orders of generality according to a principle of division of aggregate masses), they
can be defined according to the Scholastic method of definition. In contrast, be-
cause the names of fictitious entities, such as “obligation”, do not refer to entities
accessible by sense perception, and have no superior genus (they cannot be organ-
ised hierarchically according to a principle of division of aggregate masses), they
demand a different method.
   To help appreciate Jeremy’s argument, consider applying the method per genus
et differentiam to the term “obligation”. The first step would involve determining
the superior genus of “obligation”. The only superior genus, according to Jeremy,
for a term like “obligation”, is the universal genus “fictitious entity”. Because we
have identified a superior genus, it appears that we can proceed using the method
per genus et differentiam. However, Jeremy argues that because the species of
the genus “fictitious entities” are so many and so comprehensive, any attempt to
provide a character by which “obligation” can be distinguished from all others,
would lead to an enumeration of properties that may never reach completion.
   Later, in the Chrestomathia, Jeremy presented his new classification of entities,
dividing them into five types: Real entities (perceptible by the senses), Inferential
entities (we believe to have real existence, but imperceptible to our senses, like
God), Fabulous entities (believed by others, but to the existence of which we
can attach no belief, heathen gods), Collective entities (the result of operations
of abstraction and generalisation, forming a class), and Fictitious entities (neither
have, nor is supposed to have any real existence, but which is grammatically spoken
of as real, for example, obligation). George presented this classification of entities
in Chapter Three of the Outline.
   In light of the problem faced when expositing fictitious entities, Jeremy pro-
posed a new method of exposition called “paraphrases”. Jeremy’s first step in
paraphrases rested on a controversial and decidedly unLockean assumption. The
logic of Aristotle and the Scholastics, as well as Locke, was a logic of terms. On the
Scholastic and Lockean account, propositions are the result of combining terms —
terms are fundamental. Jeremy turned this assumption on its head, believing that
propositions came first, and it is by methods of abstraction and analysis, that we
arrive at terms.52 Jeremy’s method of paraphrases brings this controversial as-
  51 J. Bentham, “A Fragment on Government” in A Comment on the Commentaries and A
Fragment on Government, ed. J.H. Burns and H.L.A. Hart (London, 1977), 181n.
  52 Recent scholars have commented on this aspect of Jeremy’s work. See Ogden, in his “Theory
of Fictions”; H.L.A. Hart Essays on Bentham (Oxford: Clarendon Press 1982), 43, and especially
W. O. Quine in his Theories and Things (Cambridge, Belknap Press 1981), “Five Milestones
of Empiricism” in From a Logical Point of View (1961) p. 67-72, and “Russell’s Ontological
Development” The Journal of Philosophy 63, No. 21 (1966), 657-667. Quine, for example, saw
this focus on propositions in Jeremy’s theory of fictitious entities as foreshadowing an innovation
that would gain popularity near the end of the nineteenth century and the beginning of the
                                         Bentham’s Logic                                       15
sumption into the spotlight. For fictitious terms to serve any instructive purpose,
they need to be resolved into Lockean simple ideas, that is, they need raise images
either of perceived substances or emotions. Rather than simply trying to reduce
a fictitious entity to a real entity by a chain of synonyms, paraphrases begins
with a proposition that contained the name of the fictitious entity and translates
that proposition to an equivalent one that uses the names of only real entities.
For Jeremy, if a fictitious entity can be replaced in a proposition by a real entity,
without any loss of meaning, then the fictitious entity is meaningful.
   Take, for example, the application of paraphrases on the term “obligation”. We
begin by combining the name of the fictitious entity “obligation” with other words
to create a proposition, such as “A man is under an obligation to do X”. Then,
we find another proposition equivalent to “A man is under an obligation to do
X”, such as “A man is liable to punishment if he does not do X”. The sentence
“A man is liable to punishment if he does not do X” is equivalent to “A man
is liable to pain if he doesn’t do X”, provided we adopt the Jeremy’s notion of
punishment, namely that punishment is “pain annexed to an act, and accruing on
a certain account, and from a certain source”.53 The idea associated with the word
“pain” is, in Lockean terms, a “simple” idea, and the method of paraphrases has
resolved the sentence containing the term “obligation” into a sentence containing
the simple idea, namely “pain”. So, in the case of obligation, unless an obligation
can be enforced with sanctions and pain, an obligation can be dismissed as “absurd
in logic”.
   Getting back to Blackstone and natural rights, for Jeremy, to see why a “natural
right” is a nonsense term, we must compare the analysis a “natural right” with
a “legal right”. Applying paraphrases, both can be analysed in terms of corre-
sponding duties, but only a legal duty can be analysed further into a simple idea.
Propositions that include legal duty can be translated to propositions that laws
that include the notion of punishment. Legal rights are real rights because they
are produced by existing legal systems. Because there is no corresponding law
with respect to natural duties, claimed Jeremy, natural rights are just imaginary
rights.
   In George’s eyes, Whatley and Sanderson were behaving much like Blackstone,
in placing an undue emphasis in discussions of exposition on one particular mode,
twentieth century with Russell. Quine, however, qualified the historical significance of Jeremy’s
idea:
        Bentham was perhaps the first to see the sentence thus as the primary vehicle of
        meaning. Frege took up the tale. But Russell, in his theory of singular description,
        was the first to put this insight to precise and effective use. Frege and Peano had
        allowed singular description the status of a primitive notation; only with Russell did
        it become an “incomplete symbol defined in use.” What suggested the expedient
        to Russell was not in fact Bentham’s work, it seems, but a use of operators in the
        differential Calculus.”
  53 J. Bentham, ‘A Fragment on Government’, in A Comment on the Commentaries and A
Fragment on Government, ed. J.H. Burns and H.L.A. Hart (London, 1977), 495 in 95n.
16                         Gordon R. McOuat and Charissa S. Varma
namely definition.54 Chapter six of the Outline, titled “Exposition”, George pre-
sented the issues with exposition by framing it in terms of teaching and learn-
ing concepts (an approach Jeremy took in his discussion of exposition in the
Chrestomathia) and outlined Jeremy’s method of paraphrases, in addition to eleven
other Benthamite forms of exposition.
   Like Jeremy, George did not see the problems with the tactical aspect of logic
ending with new classifications of entities and new modes exposition, and he moved
quickly to what he took to be one of the most important subjects in the reform of
the tactical aspect of logic — methodization.
   George began chapter seven, titled “Methodization” with a discussion of the two
operations of methodization — collocation and distribution — but spent the bulk
of this chapter on distribution. According to George, distributive methodization
is performed by three operations:
    1. Dividing an entity into parts: in the case of individuals, this is analysis, and
        in the case of collective entities, this is logical division.
    2. Uniting entities into a whole: In the case of individuals, this is synthesis,
        and in the case of collective entities, this is generalization.
    3. Distribution: “for the performance of this operation, a number of wholes,
        as well as parts, are supposed to be already given; but, as in the case of
        real entities, the exhibiting of wholes and parts might appear to constitute
        the while of this operation, the figure is now changed, the aggregate ideas
        as receptacles into which the several given partial ideas are supposed to be
        placed or distributed.”55
   What George referred to as “distribution” is similar to what Jeremy referred
to as “arrangement” in his critique of Blackstone. For Jeremy, the Benthamite
expositor needed to do more than define entities properly:
        The function of the Expositor may be conceived to divide itself into
        two branches: that of history, and that of simple demonstration. The
        business of history is to represent the Law in the state it has been in,
        in past periods of its existence: the business of simple demonstration
        in the sense in which I will take leave to use the word, is to represent
        the Law in the state it is in for the time being.56
It is important to note that Jeremy did not borrow the term “demonstration” from
logicians and mathematicians, but from naturalists. Jeremy explained his use of
the word “demonstration” in the footnote to this passage:
  54 George organised Whately’s five kinds of definition on a table on pg 95 of G. Bentham,
Outline of a New System of Logic, With a Critical Examination of Dr Whately’s “Elements of
Logic” (London: Hunt & Clarke 1827).
  55 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 101-2.
  56 J. Bentham, ‘A Fragment on Government’, in A Comment on the Commentaries and A
Fragment on Government, ed. J.H. Burns and H.L.A. Hart (London, 1977), p. 414.
                                     Bentham’s Logic                                     17
        The word demonstration may here seem, at first sight, to be out of
        place. It will be easily perceived that the sense here put upon it is not
        the same with that in which it is employed by Logicians and Mathe-
        maticians. In our own language, indeed, it is not very familiar in any
        other sense than theirs: but on the Continent it is currently employed
        in many other sciences. The French, for example, have their demon-
        strateurs de botanique, d’anatomie, de physique experimentale, &c. I
        use it out of necessity; not knowing of any other that will suit the
        purpose.57
What Jeremy referred to as “demonstration”, George called “methodization”.
For Jeremy, there are three kinds of demonstration: arrangement, narration and
conjecture. Jeremy’s notion of “arrangement” seems to correspond roughly with
George’s notion of “distribution”. Of the three kinds of demonstration, Jeremy
saw arrangement as the most difficult:
        Among the most difficult and the most important of the functions of
        the demonstrator is the business of arrangement. In this our Author
        has been thought, and not, I conceive, without justice, to excel; at least
        in comparison of any thing in that way that has hitherto appeared. ‘Tis
        to him we owe such an arrangement of the elements of Jurisprudence,
        as wants little, perhaps, of being the best that a technical nomencla-
        ture will admit of. A technical nomenclature, so long as it is admitted
        to mark out and denominate the principal heads, stands an invincible
        obstacle to every other than a technical arrangement. For to denomi-
        nate in general terms, what is it but to arrange? and to arrange under
        heads, what is it but to denominate upon a large scale? A technical
        arrangement, governed then in this manner, by a technical nomencla-
        ture, can never be otherwise than confused and unsatisfactory. The
        reason will be sufficiently apparent, when we understand what sort of
        an arrangement that must be which can be properly termed a natural
        one.58
Arrangement consists of three tasks: distributing the entities into different groups,
for the purpose of a general survey; determining the order in which those groups
will be brought to view; and finding a name for each of them. Jeremy’s discussion
of arrangement in his legal and his education reform revealed the influence of
methodology articulated by Swedish naturalist Carl Linnaeus (1707–1778).
   Jeremy worried that the language of jurisprudence was corrupt — different ob-
jects grouped under one name, and similar objects under different names. This
corruption resulted in poor systems.59 Concerns about naming and the problem
  57 J. Bentham, ‘A Fragment on Government’, in A Comment on the Commentaries and A
Fragment on Government, ed. J. H. Burns and H. L. A. Hart (London, 1977), 414 footnote 21.
  58 J. Bentham, ‘A Fragment on Government’, in A Comment on the Commentaries and A
Fragment on Government, ed. J.H. Burns and H.L.A. Hart (London, 1977), 414.
  59 S. Jacobs “Bentham, Science and the Construction of Jurisprudence” History of European
Ideas 12 (1990): 585.
18                         Gordon R. McOuat and Charissa S. Varma
of synonymy was a not just a problem for jurisprudence. Many eighteenth-century
naturalists raised this type of concern, and both Jeremy and George saw in the
work of these naturalists, especially the botanists, a solution.60 For Jeremy, nat-
uralists knew that in order for their objects of investigation to be understood and
useful, they must be classified and specifically defined. Natural history provided
the model for Jeremy’s reform.61
   Although not the botanist George was, Jeremy was no stranger to natural his-
tory and natural history classification systems. Jeremy shared a practical interest
in botany with George’s mother Mary, who was a rather accomplished botanist
herself. Jeremy and Mary exchanged letters and botanical specimens for many
years, and Jeremy himself kept a small garden.62 In addition to botanical classi-
fications, Jeremy Bentham refers in his writings to the Linnaean inspired nosolo-
gies of French botanist and Professor of Medicine, Fran ̧ois Boissier de Sauvages
                                                                  c
(1706–1767) and Scottish Professor of Chemistry and Medicine, William Cullen
(1710–1790). Jeremy’s introduction to nosology was likely due to another mem-
ber of the Fordyce family — George’s maternal grandfather, Scottish physician,
George Fordyce (1736–1802). George Fordyce was a student of Cullen’s, and lec-
tured in London.63 Fordyce’s lectures drew from the work of Cullen, and Jeremy
attended some of those lectures.
   Jeremy suggested using Linnaeus’s botanical classification system — the sexual
system of plants — as a guide to arrangement or classification.64 As a teenager,
Linnaeus was fascinated with the sexuality of plants, and after many years of
observing the great diversity of sex organs of plants, he decided to make this
the basis of his botanical classification system in Systema Naturae (1735).65 Lin-
  60 See L. Daston, “Type Specimens and Scientific Memory.” Critical Inquiry 31(2004): 153–
182.
  61 The following three papers suggest this reading of Jeremy’s legal and educational reform
work: W. C. Mitchell “Bentham’s Felicific Calculus” Political Science Quarterly 33 no.2 (1918):
161–183.; S. Jacobs “Bentham, Science and the Construction of Jurisprudence” History of Eu-
ropean Ideas 12 (1990): 583–594.; A. Perreau-Saussine, “Bentham and The Boot-Strappers of
Jurisprudence: The Moral Commitments of a Rationalist Legal Positivist” Cambridge Law Jour-
nal, 63 no.2 (2004): 346–383.
  62 M. Filipiuk, ed. George Bentham, Autobiography 1830–1834. (Toronto: University of
Toronto Press, 1997), xx.
  63 N. G. Coley, “Fordyce, George (1736–1802),” in Oxford Dictionary of National Biography,
ed. H. C. G. Matthew and Brian Harrison (Oxford: Oxford University Press, 2004).
  64 Jeremy mentions Linnaeus in many of his writings. Jacobs notes, for example Jeremy Ben-
tham, The Correspondence of Jeremy Bentham, 1, 1752 − 76. ed.T. L. S. Sprigge (London:
Athlone Press, 1968), p. 105; also A Comment on the Commentaries and A Fragment on Gov-
ernment, ed. J. H. Burns and H. L. A. Hart (London, 1977), 415, 416, 418–419; and in Principles
of Morals, p. 273 note yl. There are also many references to Linnaeus in the Chrestomathia.
  65 For more on Linnaeus, see W. Blunt The Compleat Naturalist ; A Life of Linnaeus (New
York, Viking Press 1971); G. Eriksson, “Linnaeus the Botanist.” in Linnaeus: The Man and his
Work, ed. F. Fr ̈ngsmyr. (Berkley, University of California Press, 1985); S. Lindroth, “The Two
                a
Faces of Linnaeus” in Linnaeus: The Man and his Work, ed. F. Fr ̈ngsmyr (Berkley, University
                                                                   a
of California Press 1985); J. Larson Reason and Experience: The Representation of Natural
Order in the Work of Carl von Linn ́. (Berkeley, University of California Press 1977); E. Mayr,
                                    e
The Growth of Biological Thought: Diversity, Evolution, and Inheritance (Cambridge, Belknap
                                         Bentham’s Logic                                        19
naeus divided all flowering plants into twenty-three classes (the twenty-fourth being
Cryptogamia that included flowerless plants like mosses) according to the number,
relative length, arrangement, etc., of the male organ or stamen. These classes were
then divided into orders based on the female parts, or pistils. That Jeremy would
have been familiar with Linnaeus’s work comes as no surprise. Linnaeus’s system
was widely regarded as the most practical system available from 1737-1810, and
was the most widely used.66
   Like Linnaeus, Jeremy saw the practical benefit of singling out a particular
character to use as a division principle in an artificial system. Jeremy believed that
the consistent use of an established principle of division would result in meaningful
groups organised in a coherent way. Borrowing a page from Linnaeus, Jeremy
organised all offences hierarchically according to his principle of utility.
   Jeremy also saw the merit of arranging entities hierarchically. Linnaeus was
not the only example to which Jeremy appealed on the topic of hierarchical ar-
rangements. Jeremy drew attention to d’Alembert, who famously organising his
ˆtres fictifs hierarchically in his “tree of knowledge”.67 Jeremy maintained that
e
the hierarchical tree structure effectively organised the terms according to three
relations: logical identity, logical diversity, and practical dependence. This ar-
rangement provides the reader, or in the case of the organisation of the terms
of jurisprudence, the Legislator, with “an insight — the more clear, correct, and
extensive the better, - into the matter of every [...] branch of art and science.”68
   A closer inspection of Jeremy Bentham’s classification systems in his legal work,
however, demonstrates a far more consistent application of a division principle
based on a single character than that found in Linnaeus’s systems. But this is
not the only difference between Linnaeus and Jeremy. Jeremy seemed to posit
a stronger relationship between the division principle employed and the essence
of the group when he claimed that a strict adherence to a fundamental division
principle would provide a “natural” arrangement for his classification systems:
        That arrangement of the materials of any science may, I take it, be
        termed a natural one, which takes such properties to characterize them
        by, as men in general are, by the common constitution of man’s na-
        ture, disposed to attend to: such, in other words, as naturally, that is
        readily, engage, and firmly fix the attention of any one to whom they
        are pointed out. The materials, or elements here in question, are such
        actions as are the objects of what we call Laws or Institutions.69
Press of Harvard University Press 1982).
  66 At least part of the reason for its popularity was because “the information necessary for the
construction of more natural systems had not yet been assembled for synthesis” W. Blunt The
Compleat Naturalist; A Life of Linnaeus (New York, Viking Press 1971), 244.
  67 See J. Bentham, Chrestomathia. ed M. J. Smith and W. H. Burston. (Oxford: Clarendon
Press1983), 257.
  68 See J. Bentham, Chrestomathia. ed M. J. Smith and W. H. Burston. (Oxford: Clarendon
Press1983), 218–220.
  69 J. Bentham, ‘A Fragment on Government’, in A Comment on the Commentaries and A
Fragment on Government, ed. J. H. Burns and H. L. A. Hart (London, 1977), p. 415.
20                             Gordon R. McOuat and Charissa S. Varma
It is possible that Jeremy’s understanding of a natural system is more a ves-
tige of Locke than a modification of Linnaeus. Locke agreed with Aristotle and
the Scholastics that essences serve two distinct purposes: a classificatory pur-
pose, that is, the essences of entities should help classifying entities into kinds;
and an explanatory purpose, that is, the essence of entities should help explain
the properties and behaviours of the entities we observe in the world. However,
Locke maintained that each purpose is served by a distinct type of essence, the
explanatory purpose by real essences 70 and the classificatory purpose by nominal
essences.71 For Locke, nominal essences are nothing more than the criteria that
we create and use to mark off the members of a group. Nominal essences are the
boundaries set by us, not by reality. Reality can supply phenomenal resemblances,
but for Locke, resemblances do not constitute natural, real boundaries. At best,
phenomenal resemblances indicate underlying structural resemblances.72 Jeremy’s
principle of utility provided a “natural” classification of offences, and this principle
alone can “render satisfactory and clear any arrangement.” In contrast, Linnaeus,
like many naturalists during this period, called arrangements based on a single
character, “artificial” systems. Artificial systems stood in contrast to “natural”
systems, systems that group organisms according to many characters.
   A further point George stressed in his discussion of methodization was that both
distribution (arrangement) and generalisation are needed for the construction of a
  70 Locke defines real essences in the following way in An Essay Concerning Human Under-
standing:
       Real essences . . . may be taken for the very being of anything, whereby it is what it
       is. And thus the real internal, but generally (in substances) unknown constitution
       of things, whereon their discoverable qualities depend, may be called their essence.
       This is the proper original signification of the word, as is evident from the formation
       of it; essentia, in its primary notation, signifying properly, being. And in this sense
       it is still used, when we speak of the essence of particular things, without giving
       them any name. [3:3:15]
In other words, something is a real essence just in case it provides the sufficient condition(s) for
the explanation of the properties of entities, whether the entities in question are substances or
not, so defined.
  71 Locke defines nominal essences in the following way An Essay Concerning Human Under-
standing:
       Nominal essences . . . [have] been almost wholly applied to the artificial constitution
       of genus and species. It is true, there is ordinarily supposed a real constitution of
       the sorts of things; and it is past doubt there must be some real constitution, on
       which any collection of simple ideas co-existing must depend. But, it being evident
       that things are ranked under names into sorts or species, only as they agree to
       certain abstract ideas, to which we have annexed those names, the essence of each
       genus, or sort, comes to be nothing but that abstract idea which the general, or
       sortal (if I may have leave so to call it from sort, as I do general from genus), name
       stands for. And this we shall find to be that which the word essence imports in its
       most familiar use. [3:3:15]
In other words, something is a nominal essence just in case it sets a boundary to the class in
such a way as to justify our application of a name.
  72 For a more detailed discussion of this claim see Mackie, J., L., Problems From Locke, (Oxford
University Press, London 1976), 134–6.
                                        Bentham’s Logic                                        21
successful system. He provided the following example to help appreciate how these
operations work together, as well as to show his distinction between division and
distribution. George’s example began with the naturalist’s use of generalization:
        A botanist visits a country with whose productions he is as yet unac-
        quainted; he sees a number of plants which resemble one another very
        strongly, and which differ considerably from any other plants which he
        has seen or heard of; he discovers successively several of these sets of
        plants, and by generalization he forms as many new species, charac-
        terized by the properties he has observed in these several individual
        plants.73
Then George moved on to a naturalist’s use of distribution:
        On referring to his books, he compares the several properties there
        given as characteristics of general classes, which those which are pos-
        sessed by his several new species, and thus decides to which of these
        general classes the species in question belong. This many be termed
        distributing those species under their superordinate genera.
        Suppose that by the addition of these new species, some one class (or
        logical genus) may now consist of so many, as to render it difficult
        to compare them or to retain their distinctive properties without the
        help of some intermediate classes; these classes may be formed either
        by dividing the genus into subclasses, or by distributing the species
        into groups. In this case division and distribution appear so closely
        allied, as scarcely to be distinguishable otherwise than by the form of
        expression.74
By emphasising the necessity of both generalization and distribution, this passage
reflects the impact George’s training as a botanist had on his logical thinking.
That he would have emphasised both operations and used such an example is not
surprising, given his experience as a botanist. Historian of biology M ̈ller-Wille   u
argued that Linnaeus discussed a division of labour in natural history, between
the “ ‘collectors (collectores)’ who ‘primarily cared for the number of species’ by
collecting, describing and drawing plant specimens, and ‘taxonomists (methodici )’,
who ‘primarily cared for the classification and denomination of plants’.”75
   For George, a great deal of confusion surrounded the tools used in constructing
a classification system. For Jeremy, the principal tool for arrangement or distri-
bution of real entities was bifurcation (sometimes referred to as “bipartition” or
  73 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 102–3.
  74 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 103.
  75 S. M ̈ ller-Wille, “Gardens of paradise” Endeavour 25 No. 2 (2001), 49. See also S. Lindroth,
          u
“The Two Faces of Linnaeus” in Linnaeus: The Man and his Work, ed. F. Fr ̈ngsmyr (Berkley,
                                                                              a
University of California Press 1985) and J. Larson Reason and Experience: The Representation
of Natural Order in the Work of Carl von Linn ́. (Berkeley, University of California Press 1977).
                                                e
22                         Gordon R. McOuat and Charissa S. Varma
“exhaustive division”). Jeremy traced this tool back to Porphyry’s commentary
on Aristotle’s logic, but also saw this tool wielded by some late eighteenth-century
naturalists. George too found instances of bifurcate divisions in natural history. As
mentioned earlier, George saw the dichotomous key typified in the Flore Fran ̧aise        c
of Lamarck and de Candolle as an instance of distribution using bifurcate divisions.
In the chapter titled “M ́thode Analytique” of the Flore Fran ̧aise, Lamarck began
                           e                                         c
with the following division: Fleurs dont les  ́tamines & pistils peuvent ais ́ment
                                                   e                                  e
se distinguer (flowers whose stamen and pistils are distinguishable) or Fleurs dont
les  ́tamines & pistils sont nuls, ou ne peuvent se distinguer (flowers whose sta-
     e
mens and pistils are absent or indistinguishable). Amateur naturalists (as George
was when he first read the Flore) were quick to praise dichotomous identification
keys, as they had the virtue of allowing anyone to identify plants with relative
ease, as compared to the artificial system of Linnaeus76 (a system that organised
plants based on their sexual characters) or the natural systems of Michel Adan-
son, Bernard de Jussieu, and Antoine Laurent de Jussieu.77 This is because with
bifurcation, large groups are eliminated at each stage by using mutually exclusive
characteristics, making it easy to find the name of a plant, provided the reader
can identify the requisite parts. This method stands in contrast to the methods of
many of the natural systems during the eighteenth century, systems that required
the reader to have a great deal of prior botanical knowledge under their belt.
   Both Jeremy and George understood bifurcation as taking a class and subjecting
it to progressive dichotomous divisions. In other words, every member of class A
has the character X in common, and can be divided into two subclasses, B and C,
depending on if the member of the class A has the character Y or not (so all the
members of B have Y in common and all the members of C do not), then every
member of class B has the character Y , and can be divided into two subclasses,
D and E, depending on if the member of the class B has the character Z or not
(so all the members of D have Z in common and all the members of E do not),
and so on. The primary virtue of this method is that results in a classification
that is both exclusive and exhaustive. Jeremy continued by claiming that if, when
analysing classes and assigning them to divisions to which names are then applied,
bifurcation will complement definition by giving a precise definition of a name in
relation to the properties that distinguish the class it designates. Jeremy discussed
the virtues of bifurcation in Introduction to the Principles of Morals and Legislation
(1781):
        When a number of objects, composing a logical whole, are to be con-
        sidered together, all of these possessing with respect to one another a
        certain congruency or agreement denoted by a certain name, there is
        but one way of giving a perfect knowledge of their nature; and that
  76 Although Lamarck didn’t adopt Linnaeus’s artificial system, his Flore was one of the first
French works to include the Linnaean nomenclature.
  77 J. M. Drouin, “Principles and Uses of Taxonomy in the Works of Augustin-Pyramus de
Candolle.” Studies in History and Philosophy of Biology and Biomedical Sciences 32 (2001):
264-5.
                                         Bentham’s Logic                                       23
        is, by distributing them into a system of parcels, each of them a part,
        either of some parcel, or, at any rate, of the common whole. This can
        only be done in the way of bipartition. . . 78
While Jeremy encouraged the use of bifurcation whenever possible in arrangement
(since classification was designed to assist the logical operation of exposition),
George was very careful to outline the limits of bifurcation in his chapter on
methodization.79
    George’s careful discussion of the restricted use of bifurcation began by draw-
ing a connection between Whately’s and Sanderson’s emphasis on definition and
Whately’s and Sanderson’s general rules for division in his discussions of method-
odization. The problem, as George saw it, was that Whately’s and Sanderson’s
sets of rules apply only to bifurcate divisions, and many things that cannot be
effectively organised by bifurcation. Bifurcation, for example, cannot be applied
to fictitious entities for the same reason the method of definition cannot be ap-
plied fictitious entities — fictitious entities lack a common genus and so cannot be
organized according to “any exhaustive plan of arrangement, but must be picked
up here and there as they happen to occur”.80 George also mentioned two further
restrictions on bifurcation. First, George reminds us that a second purpose of
classification was to aid retention, often times this purpose is best served by mul-
tifurcate divisions rather than bifurcate divisions.81 George assumed Sanderson
was also aware of this latter restriction on the application of bifurcation:
        But upon reflection, he [Sanderson] was probably led to consider, that,
        if the successive bifurcate divisions of logical aggregates were carried
        on to the lowest stage necessary for classification, and if, at the same
        time, names were given to every one of the intermediate aggregates
        so formed, — this operation would, on many occasions, be attended
        with more labour than advantage [our italics], and would require too
        great a strain on the memory. He therefore added the restriction that
        dichotomy was only to be made use of when it could be easily performed
        [our italics]; and that it should not be sought out on all occasions with
        superstitious strictness with which the Ramaeans followed it.82
Both the Benthams note the impracticality of applying this method to botanical
classification. Jeremy wrote:
  78 J. Bentham, An Introduction to the Principles of Morals and Legislation, ed. J. H. Burns
and H. L. A. Hart (London, 1970), 187.
  79 George also talks about division in his “Essai” see G. Bentham, “Essai sur la nomenclature et
la classification des principales branches d’art-et-science” (Paris, Bossange Freres 1823). 70–72.
  80 J. Bentham, An Introduction to the Principles of Morals and Legislation, ed. J. H. Burns
and H. L. A. Hart (London, 1970), 53 note c.
  81 J. Bentham, Chrestomathia ed M. J. Smith and W. H. Burston. (Oxford, Clarendon Press
1983), 252.
  82 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 110–1.
24                         Gordon R. McOuat and Charissa S. Varma
        The quantity of surface necessary to the exhibition of such a diagram,
        presents another circumstance, by which, long enough before the num-
        ber of the extreme branches had reached to any such number as forty
        thousand, as above, not to say the tenth or the hundredth part of it,
        the bar of impracticality would be opposed.83
Further, George was writing the Outline at a time when botanists were racing to
organize and classify the ever increasing numbers of known plants. With thou-
sands of plants species already described and thousands more estimated to exist,
memorization of species names was simply out of the question, so designs for ef-
ficient storage and retrieval systems were the top priority. While dichotomous
keys initially held some promise, especially considering how helpful they were
for the purpose of identification, George was keenly aware of the many serious
flaws resulting from the use of a bifurcating system as a classification system
for tens of thousands of plants. The number of dichotomous divisions would be
enormous, and revising the system in light of new discoveries of species would
be tremendously difficult. It is not surprising for George, then, to find that in
natural history classification systems, natural systems and even certain artificial
systems (like Linnaeus’s sexual system of plants) adopted multifurcate divisions
rather than bifurcating divisions in their classification systems. Further, in some
instances, bifurcation loses its virtue of being exhaustive. For example, natural
history classification systems need to be able to accommodate the new species are
always being discovered.84 For example, in the “Essai” George looks at the effect
that the discovery of intermediate groups has on the kind of key he constructed,
arguing “our ignorance of so many beings, by implication as yet undiscovered,
would be even more damaging to finer subdivisions of the key”.85 Jeremy was also
aware of the problem of the number of species not being fixed had on bifurcation:
        Take, for example, Natural History, and therein Botany. Forty thou-
        sand was, some years ago, stated as the number of supposed different
        species of plants (exclusive of varieties) at that time more or less known
        to the botanic world. But, at that time, the utmost knowledge obtained
        of them by any person was not, to any such degree clear, correct, or
        complete, as to enable him, in this way, to show, of every one of them,
        in any such concise mode, its points of agreement and disagreement
        with reference to every other. And even if, in and for any one year, the
        distinctive properties of the whole multitude of individuals contained
        in the whole multitude of species then known, could have been exhib-
        ited in this systematic form, the sketch given them, if with regard to
        the whole number of species of plants then existing it professed to be,
  83 J. Bentham, Chrestomathia. ed. M. J. Smith and W. H. Burston. (Oxford, Clarendon Press
1983), 252.
  84 J. Bentham, Chrestomathia. ed. M. J. Smith and W. H. Burston. (Oxford, Clarendon Press
1983), 252.
  85 P. F. Stevens, The Development of Biological Classification: Antoine-Laurent de Jussieu,
Nature, and the Natural System. (New York: Columbia University Press, 1994): 103.
                                    Bentham’s Logic                                  25
       and even if it really were, an exhaustive one, would, in and for the next
       year, no longer posses that quality.
These concerns about bifurcation led George to suggested five considerations for
discussions of divisions:
   1. The particular end in view, or object of the division. Knowing this will help
       distinguish between practical or logical division.
   2. The nature of its subject.
   3. The source or principle of the division.
   4. The mode of division, whether it is a contradictory bifurcate division or a
       loose, irregular multifurcate division.
   5. The extent to which the division should be carried out.86
George did suggest that, in some cases, a classification system with multifurcate
divisions may turn out to be as complete and distinct as a bifurcate system, and
we can test its completeness by applying bifurcation to the same subject matter.87
George used the class of animals Vertebrata to illustrate his point. If the goal is
simply to give a general idea of the kind of animals that fall in the class Vertebrata
(in other words, it is not necessary to define exactly the nature of the several
subclasses or species), the class can be multifurcately divided into four classes:
Mammiferae are those who suckle their young; Birds are animals which have
wings and feathers; Fish are those who have fins and live in the water; Lizards
are those little animals with four legs and scaly skin.88 For the naturalist, who
aims for the all-comprehensiveness of her classes and sub-classes, and requires all
classes and subclasses to be distinctly characterized, divisions must be performed
using bifurcation. Using the above example, this can be done in three successive
bifurcate divisions: every member of class Vertebrata are either endowed with lungs
(Mammifera, Birds, Reptiles) or not (Fish). Every member of class of animals not
endowed with lungs is either endowed with mammae (Mammifera) or not (Birds,
Reptiles). Every member of class of animals endowed with mammae is either
endowed with wings (Birds) or not (Reptiles). This division is complete and its
logical species are distinct.89
   It is in George’s chapter eight, titled “Propositions”, where we find his famous
quantification of the predicate. George prepared us for the quantification of the
  86 G. Bentham, Outline of a New System of    Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt &  Clarke 1827), 104-6.
  87 G. Bentham, Outline of a New System of    Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt &  Clarke 1827), 111.
  88 G. Bentham, Outline of a New System of    Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt &  Clarke 1827), 111–2.
  89 G. Bentham, Outline of a New System of    Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt &  Clarke 1827), 113-4.
26                          Gordon R. McOuat and Charissa S. Varma
predicate early on by introducing us to the confusion that can arise from not being
clear about relations. In his Chapter Five, titled “Terms”, George drew attention
to one kind of confusion that can arise with respect to relations, a confusion that
arises when we trade terms between disciplines, specifically between Whately’s
logical notion of genera and species versus a naturalist’s notion:
       As to predicating the “whole essence” of a subject, in one term, that
       is impossible, unless that term be a strict synonym. Dr. Whately does
       not appear to have been aware of this; he implies, that, if we predicate
       the genus, we predicate part of the essence; and if the species, we
       predicate the whole essence; considering species in the sense in which
       naturalists employ the word, in which case it is, in fact, a logical genus
       with reference to individuals.90
In Scholastic logic and in Whately’s logic, wrote George:
       [l]ogical species and genus are only relative, not absolute terms, and the
       use of them always implies subalternation; that is, relation, either to a
       superordinate genus, or to a subordinate species or individual. . . As to
       the particular sense in which naturalists make use of the word species,
       it is very different from the logical sense of the word, the only one in
       which it should be made use of on the present occasion.91
The naturalist’s notion of genera and species, in contrast, is absolute, that is, genus
and species have a fixed place in a fixed hierarchy of nested sets. George was also
quick to notice that Jeremy’s emphasis hierarchical tree structures bring into the
spotlight the confusions surrounding the logical relations of logical identity, logical
diversity, and practical dependence. The goal for George would be to translate the
Aristotelian forms of propositions that contain inclusive relations into equivalent
propositions that contained relations of identity and diversity.92 In this respect,
we can begin to see why George would consider the quantification of the predicate.
   George gave a fuller discussion of his ideas concerning relations in his chapter on
propositions. The idea that propositions should be divided into two terms with the
relations of logical identity, logical diversity, or subalternation is first introduced
in a footnote on page 123, but explored in the body of the text a few pages later:
       [e]very simple proposition may be reduced to the expression between
       two ideas: the two ideas are represented by the two terms (subject and
       predicate); the relation itself by the copula. The species of relation here
       referred to are those of identity and diversity [note: The diversity ex-
       pressed in logical propositions is always absolute, and not contingent.
  90 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 69.
  91 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 69–71.
  92 N. I. Styazhkin, History of Mathematical Logic from Leibniz to Peano. (Cambridge: M.I.T.
Press, 1969), 148.
                                         Bentham’s Logic                                         27
       Resemblance and difference, and other modifications of diversity men-
       tioned in the classification of entities, are not taken into consideration
       in the syllogistic process. Diversity here signifies non-identity.] and if
       logical subalternation.93
   For George, the only possible relations within the terms of a proposition are
identity, diversity, and logical subalternation, where identity is expressed by the
copula “is”, diversity by the copula “is not”. However, for the copula verb to be an
identity between classes, the subject and predicate must be of the same quantifiable
form — hence the quantification of both the subject and the predicate.
   George began the task of reworking the classification of the forms of propositions
by expressing the two terms of a judgment by X (subject) and Y (predicate), their
identity by the mathematical sign “=”, diversity by the sign “ ”, universality by
the words “in toto”, and partiality by the words “ex parte” (for brevity he prefixed
the letters t and p as signs of universality and partiality). What resulted were not
the four familiar Aristotelian forms of judgment, but eight possibilities dependent
on the way in which the quantity and quality of the subject will be combined with
the predicate:
   1. X in toto = Y ex parte or tX = pY
   2. X in toto Y ex parte or tX pY
   3. X in toto = Y ex toto or tX = tY
   4. X in toto Y ex toto or tX tY
   5. X in parte = Y ex parte or pX = pY
   6. X in parte Y ex parte or pX pY
   7. X in parte = Y ex toto or pX = tY
   8. X in parte Y ex toto or pX tY
George then took those eight forms and reduced them to five, rather than the four
traditional Scholastic forms AEIO.94 The difference between George’s classification
of propositions and the Scholastic classification of propositions is more than simple
a difference in number of forms of proposition. George was thinking extensionally,
in terms of nested sets of classes and reducing propositions and syllogisms to an
  93 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 127-8.
  94 G. Bentham, G. Bentham, Outline of a New System of Logic, With a Critical Examination
of Dr Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 133-4. George realised
that 1 is equivalent with 7, and 2 with 8, because it is immaterial in his logic which is mentioned
first, x or y. Similarly, he thinks we can eliminate 2 as 4 expresses the same meaning in a manner
more suitable to the process of deduction. Therefore, thought Bentham, we are left with only
five types of judgment.
28                         Gordon R. McOuat and Charissa S. Varma
identity between members of classes. One consequence of this move was that
the familiar notions of “resemblance” and “essence” no longer belonged to logical
classification.95
   George was aware that earlier logicians had acknowledged the possibility of
quantifying the predicate term, but chose not to. George claimed that such an
omission led to fallacies. George wrote:
       Logicians make no mention of the first form, which they consider as
       useless, and they say that the predicate (or the second term of the
       proposition) is never distributed (note: Elements p, 42) (that is, uni-
       versal ). I should think however, that this assertion can scarcely be
       logical. Many fallacies arise from the considering of terms as synony-
       mous which are not so in reality; and it may be found as advantageous
       to reduce perfect identity to a logical form; as partial identity, or per-
       fect or partial diversity.96
This innovation, the quantification of the predicate, also made George re-think
the axiomatization of syllogisms Whately presented in the Elements. One of the
virtues of Whately’s logic is that he provided an axiomatic presentation of the
rules for syllogisms — a virtue George happily acknowledged and adopted in his
Outline. George wrote:
       But, if every legitimate syllogism must, by definition of the word, be
       self-evident without the help of these rules, might we not suppress
       altogether a system which requires so much labour to understand it,
       so much strain on the memory to keep it in mind, and which, after all,
       rather takes away from, than adds to, this self-evidence? Might not we
       substitute a few plain and simple axioms, the truth of which cannot
       be denied, and which may be found to contain, in general terms, every
       principle upon which a syllogistic conclusion can be founded?97
Traditionally, the students of Scholastic logic had to memorise mnemonic devices
in order to remember the valid forms of syllogisms. Whately replaced the cum-
bersome memory tricks with a small set of axioms, an innovation that not only
rendered distinguishing syllogisms into moods and figures useless, but spoke to
Whately’s attempt to see logic as a science with syllogisms as its theoretical core.
   In his Chapter Nine, titled “Deduction”, George discussed this axiomatization.
Like Whately, George set out to axiomatize syllogistic logic, and even credited
the first two of his own four axioms to Whately. George cited the two relevant
axioms of Whately as: if two terms agree with one and the same third, they
  95 See G. McOuat “The Logical Systematist: George Bentham and His Outline of a New System
of Logic.” Archives of Natural History 30 (2003).
  96 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 135.
  97 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 155.
                                         Bentham’s Logic                                      29
agree with each other; if one term agrees, and another disagrees with one and the
same third, these two disagree with each other.98 George then revised these two
axioms in accordance with his new analysis of propositions, that is, he replaced
the relationship of agreement in Whately’s axioms with the relation of equality.
Consequently, George’s first two first two axioms read as: things which are equal
to the same, are equal to one another; and when two things, one only is equal to
a third, and the other is not equal to that third, these two things are not equal to
each other.
   George’s new analysis of propositions also prompted his addition of two more
axioms.99 For George, Whately’s axioms are sufficient for a logic that does not
quantify the predicate, but if we do quantify the predicate, then we become aware
of that rules are necessary for those syllogisms where, for example, the middle term
is universal in both premises. George claimed that “for it is then only that the
two extremes are precisely equal to the same mean.”100 George two new axioms
are as follows: parts of a part are parts of the whole of that part (that is, of the
whole of which that part is a part); and when the whole of a class is said to be
equal to, or different from, the whole or any part of another class, it is meant that
every individual referred to by the first class, is the same as, or different from, any
individual referred to by such whole or part of such other class.101
   George’s innovations in this chapter regarding the analysis of propositions marked
a shift away from relations of “resemblance” and “essence”, and a move towards
a logic of classes — an extensional logic with relations of identity and diversity —
that would be taken up by the pioneers of algebraic logic.
   To conclude, during the 1870s, our retired botanist found himself in a strange
position in two revolutions. In spite of the fact that he was President of the
Linnean Society of London, George was seen as being mired in the pre-evolutionary
styles and practices of a by-gone era during the revolution in natural history.
In spite of the fact that his only significant logic publication — the Outline of
a New System of Logic — was not widely read, George was seen as going well
beyond his masters Richard Whately and Jeremy Bentham during the revolution
in logic. What remained constant in George were his ideals of clarity and precision
in language, and cautioning against speculation in science. The tale of George
  98 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827).
  99 George wrote:
       In order to obviate all these difficulties of expressions with regard to the concordance
       and the disagreement of collective entities, that I have thought it necessary to
       modify the expression of Dr. Whately’s two axioms, and to add two others, the
       last of which may perhaps be considered rather as an explanation of the relations
       of collective entities to one another, than strictly speaking, as an axiom.
G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr Whately’s
“Elements of Logic” (London: Hunt & Clarke 1827), 158.
 100 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 156.
 101 G. Bentham, Outline of a New System of Logic, With a Critical Examination of Dr
Whately’s “Elements of Logic” (London: Hunt & Clarke 1827), 155-6.
30                           Gordon R. McOuat and Charissa S. Varma
Bentham’s life, his influences, and his innovations show that logic and natural
history were intimately connected during the nineteenth century, but not in ways
normally discussed in the histories of either discipline.
                                 ACKNOWLEDGEMENTS
We are very much indebted to Philip Schofield and Michael Quinn for their in-
sights and assistance while we were looking at Jeremy Bentham’s unpublished
logic papers at University College, London. We would like to thank the librarians
at Robarts, Gerstein, and Thomas Fisher Rare Books Libraries at the University
of Toronto for their invaluable help. Charissa Varma owes particular thanks to
Jennifer Keelan, Martha Harris, and Sharon Brown. This research was supported
by a grant from the Social Sciences and Humanities Research Council of Canada,
we are grateful to the University of King’s College, Halifax, which administers this
grant, and to the taxpayers of Canada.
                                      BIBLIOGRAPHY
 [Ashwroth, 1970] E. J. Ashworth. Some Notes on Syllogistic in the Sixteenth and Seventeenth
    Centuries, Notre Dame 11, no. 1, 17–33, 1970.
 [Ashworth, 1979] E. J. Ashworth. The ‘Libelli Sophistarum’ and the Use of Medieval Logic Texts
    at Oxford and Cambridge in the Early Sixteenth Century. Vivarium, 17:134, 1979.
 [Ayers, 1981] M. R. Ayers. Locke versus Aristotle on Natural Kinds. Journal of Philosophy 78:
    247-72, 1981.
 [Ayers, 1994] M. R. Ayers. Locke: Epistemology and Ontology. London: Routledge, 1994.
 [Ayers, 1999] M. R. Ayers. Locke. London: Routledge, 1999.
 [Baynes, 1873] T. S. Baynes. Mr. Herbert on Sir Wm. Hamilton and the quantification of the
    predicate Contemporary Review 21: 796-798, 1873.
 [Bentham, 1823] G. Bentham. Essai sur la nomenclature et la classification des principales
    branches d’art-et-science Paris: Bossange Freres, 1823.
 [Bentham, 1827] G. Bentham. Outline of a New System of Logic, With a Critical Examination
    of Dr Whately’s Elements of Logic. London: Hunt & Clarke 1827.
 [Bentham, 1817] G. Bentham. Presidential Address Proceedings of the Linnean Society of Lon-
    don 1870-1871, xxxiv-lxxviii, 1817.
 [Bentham, 1968] J. Bentham. The Correspondence of Jeremy Bentham, 1, 1752–76. ed. T. L.
    S. Sprigge, London: Athlone Press, 1968.
 [Bentham, 1970] J. Bentham. An Introduction to the Principles of Morals and Legislation, ed.
    J.H. Burns and H.L.A. Hart, London: Oxford University Press, 1970.
 [Bentham, 1977] J. Bentham. A Fragment on Government, in A Comment on the Commentaries
    and A Fragment on Government, ed. J.H. Burns and H.L.A. Hart London:Oxford University
    Press 1977.
 [Bentham, 1983] J. Bentham. Chrestomathia. ed M. J. Smith and W. H. Burston. Oxford:
    Clarendon Press, 1983.
 [Blakey, 1851] R. Blakey. Historical Sketch of Logic: From the Earliest Times to the Present
    Day. London: Bailli`re, 1851.
                        e
 [Coley, 2004] N. G. Coley. G. Fordyce (1736–1802), in Oxford Dictionary of National Biography,
    ed. H. C. G. Matthew and Brian Harrison, Oxford: Oxford University Press, 2004.
 [Blunt, 1971] W. Blunt. The Compleat Naturalist: A Life of Linnaeus (New York, Viking Press
    1971)
 [Bochenski, 1961] J. Bochenski. A History of Formal Logic. New York: Chelsea Publishing Co.
    1961.
                                      Bentham’s Logic                                         31
[Daston, 2004] L. Daston. Type Specimens and Scientific Memory. Critical Inquiry 31: 153-182,
   2004.
[Drouin, 2001] J.-M. Drouin. Principles and Uses of Taxonomy in the Works of Augustin-
   Pyramus de Candolle. Studies in History and Philosophy of Biology and Biomedical Sciences
   32: 255-275, 2001.
[Eriksson, 1985] G. Eriksson. Linnaeus the Botanist. In Linnaeus: The Man and his Work,
   edited by F. Fr ̈ngsmyr. Berkley: University of California Press, 1985.
                   a
[Ferreiros, 2001] J. Ferreiros. The Road to Modern Logic—An Interpretation The Bulletin of
   Symbolic Logic 7: 441-484, 2001.
[Filipiuk, 1997] M. Filipiuk, ed. George Bentham, Autobiography 1830-1834. Toronto: Univer-
   sity of Toronto Press, 1997.
[Fr ̈ngsmyr, 1985] F. Fr ̈ngsmyr, ed. Linnaeus: The Man and his Work. Berkley: University
   a                      a
   of California Press, 1985.
[Grattan-Guinness, 1984] I. Grattan-Guinness. Notes on the Fate of Logicism from ‘Principia
   Mathematica’ to G ̈del Incompletability. History and Philosophy of Logic 5: 67-78, 1984.
                       o
[Grattan-Guinness, 1991] I. Grattan-Guinness. The Correspondence Between George Boole and
   Stanley Jevons, 1863-1864. History and Philosophy of Logic 12: 15-35, 1991.
[Howell, 1961] W. S. Howell. Logic and rhetoric in England, 1500-1700. New York: Russell &
   Russell, 1961.
[Howell, 1971] W. S. Howell. Eighteenth-century British logic and rhetoric. Princeton: Princeton
   University Press, 1971.
[Jacobs, 1990] S. Jacobs. Bentham, Science and the Construction of Jurisprudence History of
   European Ideas 12: 583-94, 1990.
[Bentham, ] Jeremy Bentham Papers, University College London Library ci, cii.
[Jevons, 1873] W. S. Jevons. Who discovered the quantification of the predicate? Contemporary
   Review 21: 821-824, 1873.
[Jolley, 1999] N. Jolley. Locke: His Philosophical Thought. Oxford: Oxford University Press,
   1999.
[Jongsma, 1982] C. Jongsma. Richard Whately and the Revival of Syllogistic Logic in Great
   Britain in the early Nineteenth Century. Unpublished dissertation, Toronto, 1982.
[Kneale and Kneale, 1962] W. Kneale and M. Kneale. The Development of Logic. Oxford:
   Clarendon Press 1962.
[Larson, 1977] J. Larson. Reason and Experience: The Representation of Natural Order in the
   Work of Carl von Linn ́. Berkeley: University of California Press, 1977.
                            e
[Lindroth, 1985] S. Lindroth. The Two Faces of Linnaeus In Linnaeus: The Man and his Work,
   edited by F. Fr ̈ngsmyr. Berkley: University of California Press, 1985.
                   a
[Locke, 1993] J. Locke An essay concerning human understanding; abridged and edited by John
   W. Yolton. London: Dent, 1993.
[Mackie, 1976] J. L. Mackie. Problems From Locke. London: Oxford University Press, 1976.
[Mayr, 1982] E. Mayr. The Growth of Biological Thought: Diversity, Evolution, and Inheri-
   tance. Cambridge: Belknap Press of Harvard University Press, 1982.
[McOuat, 2003] G. R. McOuat. The Logical Systematist: George Bentham and His Outline of
   a New System of Logic. Archives of Natural History 30: 203-223, 2003.
[Merrill, 2005] D. D. Merrill. Augustus De Morgan’s Boolean Algebra History and Philosophy
   of Logic, 26: 75-91, 2005.
[Mitchell, 1918] W. C. Mitchell. Bentham’s Felicific Calculus Political Science Quarterly 33
   no.2: 161-183, 1918.
[M ̈ller-Wille, 2001] S. M ̈ller-Wille. Gardens of Paradise Endeavour 25 No. 2: 49-54, 2001.
   u                        u
[Ogden, 1982] C. K. Ogden. Theory of Fictions in H.L.A. Hart Essays on Bentham. Oxford:
   Clarendon Press, 1982.
[Stevens, 2003] P. F. Stevens. George Bentham (1800–1884): the life of a botanist’s botanist
   Archives of Natural History 30: 189–202, 2003.
[Pease-Watkin, 2002] C. Pease-Watkin. Jeremy and Samuel Bentham–The Private and the Pub-
   lic, Journal of Bentham Studies 5: 2002.
[Pease-Watkin, 2004] C. Pease-Watkin. Bentham, Samuel (1757–1831), in Oxford Dictionary of
   National Biography, ed. H. C. G. Matthew and Brian Harrison, Oxford: Oxford University
   Press, 2004.
[Pease-Watkin, 2006] C. Pease-Watkin. The Influence of Mary Bentham on John Stuart Mill,
   Journal of Bentham Studies 8: 2006.
32                            Gordon R. McOuat and Charissa S. Varma
[Peckhaus, 1999] V. Peckhaus. 19th Century Logic between Philosophy and Mathematics The
   Bulletin of Symbolic Logic 5: 433-450, 1999.
[Peckhaus, 2003] V. Peckhaus. The Mathematical Origins of 19th Century Algebra of Logic
   Entwurf f ̈r den Beitrag zum History of Modern Logic-Projekt, ed. Leila Haaparanta, Helsinki.
             u
   2003.
[Peckhaus, 2004] V. Peckhaus. Calculus ratiocinator versus characteristica universalis? The two
   traditions in logic, revisited History and Philosophy of Logic, 25, 3–14, 2004.
[Perreau=Saussine, 2004] A. Perreau-Saussine. Bentham and The Boot-Strappers of Jurispru-
   dence: The Moral Commitments of a Rationalist Legal Positivist Cambridge Law Journal, 63
   no.2: 346–383, 2004.
[Prest, 2004] W. Prest. Blackstone, Sir William (1723–1780), in Oxford Dictionary of National
   Biography, ed. H. C. G. Matthew and Brian Harrison, Oxford: Oxford University Press, 2004.
[Quine, 1980] W. V. O. Quine. Five Milestones of Empiricism in From a Logical Point of View.
   Boston: Harvard University Press, 1980.
[Quine, 1981] W. V. O. Quine. Theories and Things. Cambridge: Belknap Press, 1981.
[Quine, 1966] W. V. O. Quine. Russell’s Ontological Development The Journal of Philosophy
   63, No. 21: 657-667, 1966.
[Schofield, 2003] P. Schofield. Jeremy Bentham, The Principle of Utility, and Legal Positivism,
   Current Legal Problems 56: 1–39, 2003.
[Smiley, 1973] T. Smiley. What is a Syllogism? Journal of Philosophical Logic 2 (1973): 136-154,
   1973.
[Spencer, 1873] H. Spencer. The study of sociology IX - the bias of patriotism Contemporary
   Review 21: 475-502, 1873.
[Stearn, 1959] W. T. Stearn. The Background of Linnaeus’s Contributions to the Nomenclature
   and Methods of Systematic Biology. Systematic Zoology 8: 4–22, 1959.
[Stevens, 1994] P. F. Stevens. The Development of Biological Classification: Antoine-Laurent
   de Jussieu, Nature, and the Natural System. New York: Columbia University Press, 1994.
[Stevens, 2004] P. F. Stevens. Bentham, George (1800–1884), in Oxford Dictionary of National
   Biography, ed. H. C. G. Matthew and Brian Harrison, Oxford: Oxford University Press, 2004.
[Styazhkin, 1969] N. I. Styazhkin. History of Mathematical Logic from Leibniz to Peano. Cam-
   bridge: MIT Press, 1969.
[van Heijenoort, 1974] J. van Heijenoort. Subject and Predicate in Western Logic. Philosophy
   East and West 24: 253-268, 1974.
[van Heijenoort, 1992] J. van Heijenoort. Historical Development of Modern Logic. Modern
   Logic 2: 242-255, 1992.
[van Evra, 1977] J. van Evra. A Reassessment of George Boole’s Theory of Logic. Notre Dame
   Journal of Formal Logic 18: 363-77, 1977.
[van Evra, 1984] J. van Evra. Richard Whately and the rise of Modern Logic, History and
   Philosophy of Logic 5: 1-18, 1984.
[van Evra, 2000] J. van Evra. The Development of Logic as Reflected in the Fate of the Syllogism
   1600-1900. History and Philosophy of Logic 21: 115-34, 2000.
[Whately, 1866] E. J. Whately. Life and correspondence of Richard Whately, D.D. London:
   Longmans, Green, 1866.
[Whately, 1831] R. Whately. Elements of Logic, Comprising the Substance of the Article in the
   Encyclopaedia Metropolitana; with Additions, &c,. London: B. Fellowes, 1831.
[Yolton, 1985] J. W. Yolton. Locke, an introduction. New York: B. Blackwell, 1985.
[Yolton, 1968] J. W. Yolton. Locke and the Way of Ideas Oxford: Clarendon Press, 1968.
                          COLERIDGE’S LOGIC
                                        Tim Milnes
                       1   WHY READ COLERIDGE’S LOGIC?
In 1834, Thomas De Quincey wrote of a fellow essayist, philosopher and opium
addict that ‘logic the most severe was as inalienable from his modes of thinking as
grammar from his language.’1 This assessment of Samuel Taylor Coleridge reveals
more, perhaps, than De Quincey intended. It not only indicates the importance of
logic to Coleridge’s thought, but also the unconventional use he made of it. One
of Coleridge’s main intellectual ambitions, indeed, was to unite the disciplines of
grammar and logic. This task, daunting enough in itself, was however, only part
of a much broader undertaking. The reconciliation of grammar, or language, with
logic could only be effected through a universal organon that encompassed subjects
as diverse as poetry, epistemology, natural philosophy, hermeneutics, metaphysics
and theology: in short, as he saw it, all aspects of human life. Even in the work
which he devoted to the subject, the Logic manuscript, which, like so many of
the tantalising, unfinished or fragmentary works that littered his career, remained
unpublished at his death, Coleridge showed little inclination to separate logical
questions from ethical and existential concerns. The reader who approaches Co-
leridge for the first time with the assumption, as Quine puts it, that ‘logic in the
strictest sense is quantification theory, and a logical deduction in the strictest sense
consists in establishing a quantificational implication,’ is likely to find Coleridge’s
treatment of logic exotic, often baffling, sometimes even frustrating.2
   Why then, as James McKusick asks, ‘should we care about Coleridge’s Logic, a
book that was unpublished in his lifetime and had no contemporary influence?’3
Most of its readers today are professional academics attached to English Litera-
ture departments whose interest in Coleridge is primarily literary, historical, or
a mixture of both. Such commentators have on the whole, long since discarded
what Owen Barfield disparagingly refers to as ‘biographical/comparative’ and ‘bi-
ographical/psychological’ approaches to Coleridge’s thought that characterised
   1 Thomas De Quincey, ‘Samuel Taylor Coleridge,’ The Collected Writings of Thomas de
Quincey, rev. ed. David Masson, vol. 2 (Edinburgh: Adam and Charles Black, 1889) 153.
   2 Willard Van Orman Quine, Elementary Logic, rev. ed. (Cambridge, MA: Harvard University
Press, 1980) 116.
   3 James C. McKusick, ‘Coleridge’s “Logic”: A Systematic Theory of Language,’ Papers in the
History of Linguistics, eds. Hans Aarsleff, Louis G. Kelly and Hans-Joseph Niederehe (Amster-
dam: John Benjamins Publishing Company, 1987)479-480.
Handbook of the History of Logic. Volume 4: British Logic in the Ninteenth Century.
Dov M. Gabbay and John Woods (Editors)
 c 2008 Elsevier BV. All rights reserved.
34                                          Tim Milnes
much scholarship in the early twentieth century, preferring instead to concen-
trate on the materiality of either Coleridge’s language or the historical conditions
through which it was determined.4 This adjustment in focus has been linked to
developments in critical theory that make any attempt to assess what Coleridge
thought appear hazardous, to say the least.5 For his part, McKusick responds to
his own challenge by arguing that the Logic is one of Coleridge’s ‘most coherent
and intellectually sophisticated works,’ and ‘arguably the most important theory
of general linguistics produced in England during the early nineteenth century.’6
Whether such claims stand up to scrutiny remains to be seen. Nonetheless, what
can be averred at this stage, albeit tentatively, is that it is the very same qualities
that make Coleridge’s logic appear so exotic and alien to a modern reader that
make it worthy of attention, and then not merely as some historical curio, but as
highlighting a significant part of the genealogy of modern European thought, one
that is too easily forgotten.
   This claim will come as no surprise to Coleridge’s commentators, who have
long since established that Coleridge’s reputation rests upon more than a slim
collection of poems and an influential if gnomic apology for the creative imagina-
tion. His wide- ranging writings, collected in the Bollingen edition — which, after
half a century’s work, has recently been completed — reflect the ebbing of the
eighteenth-century faith in empirical science and reason and the nineteenth cen-
tury’s dawning interest in religion, history and idealism. Coleridge’s world is that
of post-revolutionary optimism swiftly followed by war, paranoia and the search
for a new resolution. For Coleridge, one of the lessons of the age is the inevitable
failure of any attempt to reform society upon purely abstract principles. Instead,
modernity must be allowed to evolve as a higher synthesis of the new energies
of reason and science with the best traditions present in the collective and often
unconscious experience of the people.
   In this new synthesis, logic plays an important but ambivalent role. On one
hand it is the levelling instrument of the revolution, a principle of abstract un-
derstanding, which, overplayed by the Enlightenment philosophes, is all too easily
appropriated by the discourse of calculated utility, by a system of capital that
hi-jacks the language of Hume and Bentham. The logic of Enlightenment, as
Wordsworth expresses it in his poem ‘The Tables Turned,’ represents the ‘med-
dling intellect’ by which ‘We murder to dissect.’ On the other hand, and against
Wordsworth, Coleridge argued that logic could be rehabilitated once it had been
reconnected with an older wisdom, with the higher logic he found in scripture,
ancient philosophy, and above all in the new German idealism.
   4 Owen Barfield, What Coleridge Thought (Connecticut: Wesleyan University Press, 1971) 3.
Nigel Leask’s The Politics of Imagination in Coleridge’s Critical Thought (Basingstoke: Macmil-
lan, 1988) remains one of the best studies of the historical and political context of Coleridge’s
philosophy.
   5 Thus, a decade after Barfield’s work, Jerome Christensen sought in Coleridge’s Blessed
Machine of Language (Ithaca: Cornell University Press, 1981) ‘not [...] to reproduce what
Coleridge thought,’ but ‘to produce the way Coleridge writes’ (15-16).
   6 McKusick, ‘Coleridge’s “Logic”’ 480.
                                      Coleridge’s Logic                                      35
   How this rehabilitation was to be effected was already the subject of debate.
By the end of the eighteenth century there was a widespread feeling that logic
was in need of a serious overhaul. Traditional logic was struggling to adapt to the
epistemological implications of an accelerating culture of scientific discovery. In
particular, before Mill developed an inductive logic, there seemed to be a disjunc-
ture between empirical psychology and the canons of logical deduction. Coleridge
registers this friction, noting with regret that Aristotle’s ‘Organum,’ which had
survived largely intact for centuries, had recently been deposed in France and
‘made to give way to the Logic of Condillac,’ although it would ‘be more accu-
rate perhaps to say that the Study of Logic altogether is exploded in France, for
Condillac’s Book is rather psychological than logical.’7 Written in 1803, these
remarks form part of the preparatory notes Coleridge made outlining a plan for
his own ‘History of Logic.’ It is a full twenty years, however, before he writes to
John Taylor Coleridge, announcing the readiness of the Logic, which at this stage
still bears the working title, ‘ “The Elements of Discourse, with the Criteria of
true and false Reasoning, as the ground-work and preparation for Public Speaking
and Debate — addressed to the Students and Candidates for the Pulpit, the Bar,
or Senate”.’ By this time, Coleridge has apparently made up his mind about the
subordinate status of logic within the hierarchy of intellectual disciplines. The
‘Elements,’ he claims, is ‘a Work of Logic for [...] forensic purposes, denying its
applicability, as a positive Organ, to all subjects [...] in which the absolute Truth
is sought for,’ and of use ‘in all subjects of discussion or inquiry, in which the
Truth relatively to the Sense and Understanding of man in all his social and civil
Concerns and Functions is alone required or of pertinence.’8 Nonetheless, and de-
spite Coleridge’s careful demarcations, it was to prove difficult to confine logic to
the purely ‘forensic purposes’ of intellect, a problem that was to have significant
ramifications for his conception of the relationship between logic, truth, and ‘man
in all his social and civil Concerns.’
                            2   THE POLITICS OF LOGIC
The civic function of logic is of vital importance to Coleridge. As Heather Jackson
has noted, after the French Revolution the concept of logic becomes highly politi-
cised, one of the ‘red-flag terms,’ like method, theory or even philosophy, which
could be relied upon to send sections of British society into paroxysms of anxi-
ety.9 Increasingly disaffected with the ‘logic’ of French rationalism, but equally
contemptuous of the British tradition of ‘common sense,’ Coleridge comes to the
   7 Samuel Taylor Coleridge, Shorter Works and Fragments, eds. H. J. Jackson and J. R. de J.
Jackson, vol. 1 (Princeton University Press, 1995) 128.
   8 Samuel Taylor Coleridge, ‘To John Taylor Coleridge,’ 5 June 1823, letter 1335 of Collected
Letters of Samuel Taylor Coleridge, ed. Earl Leslie Griggs, vol. 5 (Oxford: Clarendon Press,
1956-71) 275.
   9 See Heather J. Jackson, ‘Coleridge’s Lessons in Transition: The “Logic” of the “Wildest
Odes”,’ Lessons of Romanticism: A Critical Companion, eds. Thomas Pfau and Robert F.
Gleckner (Durham: Duke University Press, 1998) 221-222.
36                                         Tim Milnes
conclusion that what both lack is logic, in its strictest sense. Around the time
he writes his ‘Outlines of the History of Logic,’ he makes a notebook entry that
reflects upon the decline of the discipline in both Britain and France:
        Of Logic & its neglect, & the consequent strange Illogicality of many
        even of our principal writers — hence our Crumbly friable Stile/each
        Author a mere Hour-Glass/ — & if we go on in this way, we shall soon
        have undone all that Aristotle did for the human Race, & come back
        to Proverbs & Apologues — /The multitude of Maxims, Aphorisms, &
        Sentences & their popularity among the French, the beginners of this
        Style, is it some proof & omen of this?10
Coleridge’s particular concern with the ‘Crumbly friable Stile’ of the modern writer
in this entry is telling. Deprived of a unifying logic, language becomes atomised,
contingent, ‘each Author a mere Hour-Glass’ of running particulars. In this way,
Coleridge hopes to show that ‘The multitude of Maxims, Aphorisms, & Sentences’
to be found in the French materialist writers who supported the revolution, so
far from being the neplus ultra of logic, was thoroughly illogical. Once again, he
found the neglect of ancient thought, and the declining reputation of Aristotle
in particular, to be symptomatic of a deep intellectual malaise. The virtue of
Aristotle’s logic, in contrast to the psychologising of the French followers of Locke,
was its understanding of the deep and noncontingent connection between thought
and language, a theme he examines at length in his 1818-1819 lectures on the
history of philosophy:
        So Aristotle first of all determined what were the laws common to
        all coherent thinking, and therein he founded not only the science of
        logic, but with it he made general throughout all the civilised world
        the terms of the connection: “we”, “me”, “our”, [“us”], our “ands”,
        and our “thes”, and our “therefores”, and so forth.
In contrast, he continues, like ‘oriental writing’ in which ‘thought is put on thought
with little other connective than “ad[d]” for “and”,’ the ‘new French writings’ are
‘aimed at destroying all the connections of thought’ as well as ‘all the connec-
tions of society and domestic life.’11 At this point, Coleridge’s concern for the
links between language, logic and civic harmony breaks the surface of his writ-
ing. Above all, ancient Greek thought demonstrated the compatibility between
principles of democracy, social cohesion and logic in a way that did not depend
upon the reductive, aggregative kind of philosophy that had, as he saw it, reduced
France and much of Europe to dust. Indeed, as he argues in the ‘Outlines of
the History of Logic,’ ‘Ethics & argumentative Metaphysics were the Offspring of
Democracies whom superior Courage & superior Intellect had rendered victorious
  10 Samuel Taylor Coleridge, The Notebooks of Samuel Taylor Coleridge, ed. Kathleen Coburn
et al., vol. 1 (London: Routledge, 2002) 1759 (text).
  11 Samuel Taylor Coleridge, Lectures 1818-1819 On the History of Philosophy, ed. J.R. de J.
Jackson, vol. 1 (Princeton University Press, 2000) 234.
                                     Coleridge’s Logic                                    37
over Despots.’12 In this way, having ‘scramble[d] the familiar codes,’ as David
Simpson puts it,13 disassociating method, theory and logic from Jacobinism, Co-
leridge felt able to have his radical cake and eat it, arguing that ‘the origin of
Logic’ was in ‘a democracy when the Varieties of Character & moral habit find
sufficient Space & free playroom.’14 The challenge facing modernity then, was one
of recovering that democratic intellect in an age when the increasing dominance of
a mechanised and levelling understanding had rendered democracy itself danger-
ous. To this end, Coleridge believed, philosophy must be enlisted in the process of
reform, binding particular, empirical or abstract truths to spiritual, practical, or-
ganic Truth. It is on this basis that The Friend mounts its politically conservative
defence of ‘theory’:
       The Friend, however, acts and will continue to act under the be-
       lief, that the whole truth is the best antidote to falsehoods which are
       dangerous chiefly because they are half-truths: and that an erroneous
       system is best confuted, not by an abuse of Theory in general, nor
       by an absurd opposition of Theory to Practice, but by a detection of
       the errors in the particular Theory. For the meanest of men has his
       Theory: and to think at all is to theorize.15
The faintly unpatriotic air of the final remark is neutralised by Coleridge’s insis-
tence that the ultimate destiny of all theory is the ideal. The past excesses of
theory, philosophy, method, and logic will be overcome not by rejecting theory as
such, but by pursuing theory to its limit as an ideal of reason: in other words, to
the point where it merges with being. This notion of atonement through higher
synthesis, through the reconciliation of positions apparently opposed or contradic-
tory, is crucial to Coleridge’s changing view of logic, as the radical philosophies of
the young Unitarian poet and lecturer give way to the more harmonious visions of
the German idealist. For the latter figure, increasingly reliant upon opium, upon
his doctor and landlord in Highgate for managing that addiction, and upon friends,
students and the flattering attention of curious visitors for his sense of self-worth,
feelings of guilt combine with a sense of lost opportunities and wasted strength to
colour his view of human intellect. In the struggle to comprehend the failures of
revolution and his own personal inadequacies, ideas of expiation and the search
for a higher unity become the dominant figures of Coleridge’s later thought: in
particular, they shape his belief that the quantificational logic of an overweening,
‘mechanical’ understanding must be redeemed by the higher, qualitative logic of
a divinely sanctioned summum bonum.
   Indeed, one of the few constant elements throughout Coleridge’s career is his
Christian faith, and his refusal to treat questions of faith and reason separately.
  12 Coleridge, Shorter Works 1 127.
  13 David  Simpson, Romanticism, Nationalism, and the Revolt Against Theory (University of
Chicago Press, 1993) 60.
  14 Coleridge, Shorter Works 1 128.
  15 Samuel Taylor Coleridge, The Friend, ed. Barbara E. Rooke, vol. 1 (Princeton University
Press, 1969) 189.
38                                          Tim Milnes
Logic, like any dimension of human intelligence, can only ultimately be compre-
hended in terms of Christianity. Christianity, however, is not a ‘subject’ for phi-
losophy; indeed, it is questionable whether it is susceptible of ‘comprehension’ at
all. As he puts it in Aids to Reflection, ‘Christianity is not a Theory, or a Spec-
ulation; but a Life. Not a Philosophy of Life, but a Life and a living Process.’16
The notion that logic has an integral role to play within a Christian theosophy,
and that only as part of that broader theosophical order can logic itself be demon-
strated to its fullest extent, is one of the larger gulfs that lie between the modern
reader and Coleridge’s work. One useful way of approaching this problem is to
remain alert to how elastic the concept of ‘logic’ remains in this period. Reflecting
on his own early education in Biographia Literaria, Coleridge reveals a climate in
which, as the boundaries between disciplines are being contested and reformed,
definitions are up for grabs. Recalling his old schoolmaster at Christ’s Hospital,
James Bowyer, he writes:
       I learnt from him, that Poetry, even that of the loftiest, and, seemingly,
       that of the wildest odes, had a logic of its own, as severe as that of
       science; and more difficult, because more subtle, more complex, and
       dependent on more, and more fugitive causes.17
Comments such as this hint at Coleridge’s ambition to forge a link between the
logical and the existential, between ratiocination and ‘life.’ This aspiration, how-
ever, wrestled with a more positivist impulse to regulate any new illumination
under the auspices of a master discipline that embodied the codes of traditional
philosophy. Turning to his early schooling in the rules of poetic diction, Coleridge
recounts his dissatisfaction with conventional authorities on prosody:
       But as it was my constant reply to authorities brought against me from
       later poets of great name, that no authority could avail in opposition to
       Truth, Nature, Logic, and the Laws of Universal Grammar; actuated
       too by my former passion for metaphysical investigations; I laboured at
       a solid foundation, on which permanently to ground my opinions, in the
       component faculties of the human mind itself, and their comparative
       dignity and importance.18
These attempts on one hand to poeticise logic and, on the other, to underpin
poetry with logic might appear, at least potentially, to conflict. For Coleridge,
they are two sides of the same coin, as the common principles of both are to
be found ‘in the component faculties of the human mind itself.’ And yet, even
though Coleridge writes Biographia in 1816 with the confidence of someone who,
having studied long and hard the lessons of German idealism, is able to flourish
  16 Samuel Taylor Coleridge, Aids to Reflection, ed. John Beer (Princeton University Press,
1995) 202.
  17 Samuel Taylor Coleridge, Biographia Literaria, eds. James Engell and Walter Jackson Bate,
vol. 1 (Princeton University Press, 1995) 9.
  18 Coleridge, Biographia 1 22
                                      Coleridge’s Logic                                  39
as his trump card Kant’s account of the ‘comparative dignity and importance’ of
the faculties of understanding and reason (of which more later), his concern with
the relationship between logic and the language of poetry indicates the continuing
influence of a different tradition. As Paul Hamilton argues, the imported German
philosophy of self-consciousness to which Coleridge was first introduced around
1800 had to find an accommodation in his thought with a concern with linguistic
propriety — specifically, with the normativity of ‘ordinary’ language — that he had
inherited from late eighteenth-century British philosophies of language and literary
theory.19 It is within this tradition that Coleridge’s original logical speculations
had originally taken shape.
                                    3   ETYMOLOGIC
The question of the relationship between logic and language is one to which Co-
leridge returns throughout his life, with varying results. Here, as in so many
areas, his views change over time. An 1800 letter to William Godwin, however,
reveals one preoccupation that was to remain central to his thinking: the idea
that language is not merely a system of conventions, but a manifestation of divine
providence:
       I wish you to write a book on the power of words, and the processes
       by which human feelings form affinities with them — in short, I wish
       you to philosophize Horn Tooke’s System, and to solve the great Ques-
       tions [...]. “Is Logic the Essence of Thinking?” in other words — Is
       thinking impossible without arbitrary signs? & — how far is the word
       “arbitrary” a misnomer? Are not words &c parts & germinations of
       the Plant? And what is the Law of their Growth? — In something of
       this order I would endeavour to destroy the old antithesis of Words &
       Things, elevating, as it were, words into Things, & living Things too.20
Coleridge’s initial willingness to identify logic with ‘arbitrary signs’ reveals the
increasing influence of John Home Tooke’s work, particularly The Diversions of
Purley, upon his thought at this point. And yet, his desire ‘to philosophize Horn
Tooke’s System’ shows that he is already trying to link Tooke’s etymological theo-
ries to a metaphysical order based on organic principles, according to which words
are ‘parts & germinations of the Plant.’ The idea that logic and language grow
from the same seed underscores Coleridge’s conviction that the linguistic sign is
not arbitrary in its reference, and thus that etymology and logic are fundamen-
tally engaged in the same enterprise: that of understanding their common ‘germi-
nation.’ Following Stephen Prickett, we can roughly divide Coleridge’s thinking
about language into three phases: (1) an early stage, up to around 1800, when
his theories are still dominated by the ahistorical, materialistic associationism of
  19 Paul Hamilton, Coleridge’s Poetics (Basil Blackwell, 1983) 3.
  20 Coleridge, ‘To William Godwin,’ 22 September 1800, letter 352 of Letters 1 625-626.
40                                          Tim Milnes
David Hartley; (2) a relatively brief period in the first decade of the nineteenth
century, during which the influence of Tooke made him increasingly ‘aware of the
illogical complexities of language,’ and a final phase (3) in which, under the sway
of German idealism, he developed a metaphysics of language as constantly evolv-
ing, ‘with words related not so much to things as changes in human consciousness
itself.’21
(1)       Seen in this frame, Coleridge’s comments to Godwin are pivotal, in that
they attest to his gathering sense that Hartley’s claim that words are simply im-
pressions that ‘excite Ideas in us by Association, and [...] by no other means’ was
damagingly reductive, a poor explanation of ‘the power of words, and the pro-
cesses by which human feelings form affinities with them.’22 Similarly, Hartley’s
vision of the growth of language as fundamentally metaphorical, producing mean-
ing as new relations of ‘likeness’ replaced older, ‘literal expressions,’ sat uneasily
with Coleridge’s belief that the principle of words and the ‘Law of their Growth’
must be rooted in a power more dignified than that of contingent association.23
In this light, Hartley’s assumption that the bond between language and the world
was a secure but inscrutable part of God’s design is cold comfort. In response,
Coleridge’s plan ‘to destroy the old antithesis of Words & Things, elevating, as
it were, words into Things, & living Things too,’ should be read as beginning a
move away from a correspondence theory of language, towards a view of words as
in some way constitutive of thought. Words, as he writes in a notebook entry of
1810, ‘are not mere symbols of things & thoughts, but themselves things.’24
    Coleridge’s reaction against Hartley thus helps to form his own contribution to
the ongoing search for a ‘natural’ language of humanity. By rejecting Hartley’s
conventionalist linguistics of association, he aligns himself with eighteenth-century
nonconventionalists like Thomas Reid and Lord Monboddo, who argued that the
origins of language were to be found in the nascent concepts of inarticulate cries
in primitive societies.25 Indeed, Coleridge’s involvement in the debate over natu-
ral language feeds directly into his famous argument with Wordsworth on poetic
diction. Like Wordsworth and Shelley, Coleridge was given to drawing parallels
between the inchoate lyricism of early societies and the embryonic linguistic skills
of children. However, he drew the line at making a fetish of everyday language,
or prizing childish utterance as worthy of poetic emulation. Natural language was
emphatically not ‘ordinary’ language. Accordingly, in the second volume of Bi-
ographia Literaria, he denies that everyday language of rural folk should in be
taken any way as normative:
   21 Stephen Prickett, Words and The Word (Cambridge University Press, 1986) 147.
   22 David  Hartley, Observations on Man, vol. 1 (1791, Poole: Woodstock Books, 1998) 268.
   23 Hartley 1 292.
   24 Coleridge, Notebooks 3 3762.
   25 Monboddo, in turn, aligns himself with Rousseau when he claims ‘that, before men used
language, they conversed together by signs and inarticulate cries [...]. And that the first languages
had a greater greater deal of prosody, or musical tones [...]’ (On the Origin and Progress of
Language, 2nd ed., vol. 1 [Edinburgh, 1774] x).
                                   Coleridge’s Logic                                      41
       The best part of human language, properly so called, is derived from
       reflection on the acts of the mind itself. It is formed by a voluntary
       appropriation of fixed symbols to internal acts, to processes and re-
       sults of imagination, the greater part of which have no place in the
       consciousness of uneducated man [. . . ].26
More revealing still is Coleridge’s own advice to the poet:
       But if it be asked, by what principles the poet is to regulate his own
       style, if he do not adhere closely to the sort and order of words which
       he hears in the market, wake, high-road, or plough-field? I reply;
       by principles, the ignorance or neglect of which would convict him of
       being no poet, but a silly or presumptuous usurper of the name! By
       the principles of grammar, logic, psychology!27
The origins of language are ‘primitive,’ then, in a metaphysical, not a social, sense.
Just as the principles of poetic language are bound up with logic, grammar and
psychology, the origins and fates of words are bound up with the universal Logos.
The appearance of ‘grammar’ in Coleridge’s list of poetic desiderata is significant,
notwithstanding its philosophical subordination, for it betrays the extent to which
Tooke’s influence on Coleridge persisted, despite the immersion in transcendental-
ism that helped to produce Biographia.
(2) For Coleridge, Tooke’s work had two major implications, one of which was
deplorable, the other welcome. The first — the hub of Tooke’s argument — was
the linguistic deflation of ideas. As he argues in Diversions, any truly empirical
‘consideration of Ideas [...] will lead us no farther than to Nouns: i.e. the signs
of these impressions, or names of ideas.’28 Coleridge found this reduction unac-
ceptable, an error driven, as he saw it, by a regrettable hangover from empiricism,
namely the assumption that ‘Words are the signs of things.’29 Against Tooke’s
linguistic atomism, Coleridge proposes a dynamic model of language based upon
the power of the verb, an agency he links to the act of consciousness and the Lo-
gos. As Mary Anne Perkins argues, the Logos, presenting the underlying spiritual
unity of truth, language and being, is ‘the “key” to understanding every area of
his thought after 1805.’30 Vital to this idea is the conception of being as an act,
not a thing. Reality and thought are formed by verb substantive, the ‘I am,’ or
living word of God. The most basic presupposition of thought then, is unity, not
difference. Coleridge makes this point unambiguously in one of his 1819 lectures
on philosophy, when he tackles the question: what is thought?
  26 Coleridge, Biographia 2 54.
  27 Coleridge, Biographia 2 281.
  28 John Home Tooke, Epea Pteroenta or the Diversions of Purley, rev. ed. by Richard Taylor,
vol. 1 (London, 1829) 49.
  29 Tooke, Diversions 1 18.
  30 Mary Anne Perkins, Coleridge’s Philosophy: The Logos as Unifying Principle (Oxford:
Clarendon Press, 1994) 3.
42                                          Tim Milnes
       We are [so] apt to use words derived from external objects, that I verily
       believe that many a man before he asked himself the question would
       imagine his mind made of thoughts, as a wall is of bricks, till he asks
       whether he ever separated a thought from a thought, whether there
       was a meaning in the word “thought,” except as the mind thinking in
       such a situation; whether a notion of any plurality, or anything we can
       call construction, takes place in our experience. We are aware of no
       such thing [...].31
Such speculations, however, inevitably raise the question: is logic subordinate to
grammar? Despite the claims of his apologists, this is a conundrum that Co-
leridge never quite resolves. McKusick, for instance, defends the coherence of
Coleridge’s belief that the apparent dependence of logical canons upon rules of
grammar merely indicates the existence of a higher logic in the form of the Lo-
gos, in which truth and meaning have not yet been alienated. Language is prior
to syllogistic logic, in other words, but the Logos is the spirit of a higher rea-
son in words. Accordingly, language speaks us, but only because the spirit of
God, and thus humanity, already inhabits language.32 Further, McKusick agrees
with Heather Jackson that it is in this way, by drawing the deflationary sting
from Tooke’s theory of language, that Coleridge is able to welcome the second
major implication of Tooke’s work: the constitutive role of language in thought
and, correspondingly, the centrality of etymological inquiry to any philosophical
investigation. As Coleridge himself wrote in literary correspondence published in
1821:
       Etymology [...] is little else than indispensable to an insight into the
       true force, and, as it were, freshness of the words in question, especially
       of those that heve passed from the schools into the marketplace, from
       the medals and tokens [...] of the philosopher’s guild or company into
       the current coin of the land.33
Coleridge accepts Tooke’s etymology as a method of inquiry, but sides with Tooke’s
opponents like Monboddo in arguing that, far from being accidental, the structures
of language are determined by ‘logical categories which are themselves intrinsic to
thought.’34 Indeed, Coleridge’s disagreement with Tooke even finds its way into
his lectures on Shakespeare, as J. Tomalin’s contemporary notes testify:
       Home Tooke had called his book Epea Pteroenta, winged words. In
       Coleridge’s judgement it might have been much more fitly called Verba
       Viventia, or “living words” for words are the living products of the
       living mind & could not be a due medium between the thing and the
  31 Coleridge, Lectures 1818-1819 2 577.
  32 James  C. McKusick, Coleridge’s Philosophy of Language (New Haven: Yale University Press,
1986) 50.
  33 Coleridge, Shorter Works 2 927-928.
  34 McKusick, Coleridge’s Philosophy of Language 42.
                                      Coleridge’s Logic                                       43
       mind unless they partook of both. The word was not to convey merely
       what a certain thing is, but the very passion & all the circumstances
       which were conceived as constituting the perception of the thing by
       the person who used the word.35
Etymology then, as Jackson puts it, becomes a ‘tool won from the enemy’: effec-
tively, ‘etymologic.’36 This in turn enables Coleridge to revive the Enlightenment
quest for a theory of universal language based on the metaphysical primacy of the
Absolute act of self-creation made incarnate in the divine Logos. By the time he
writes Aids to Reflection in the late 1820s, Coleridge has even abandoned earlier
talk of words as ‘living things,’ denying that words are ‘things’ at all: ‘For if words
are not things, they are living powers, by which the things of most importance to
mankind are actuated, combined, and humanized.’37
   It is worth pausing for a moment to consider further the significance of this turn
in Coleridge’s theosophy of language. So long as he had sought to ‘philosophize’
Tooke’s system by rendering it more dynamic, by drawing truth and meaning closer
together, Coleridge’s conception of ‘etymologic’ had had the potential to lead to
conclusions similar to those of Tooke. Increasingly linked to meaning and to the
everyday activity of interpretation, Coleridge’s concept of truth, for a moment,
begins to appear more holistic and linguistic than metaphysical or absolute. In a
notebook entry of December 1804, for instance, he once again attempts to improve
on Tooke by demonstrating how both ‘Word’ and ‘Truth’ are etymologically related
to terms of action:
       Word, Werden — that which is — <Worth, wirthy — Rede, Redlich>
       — Truth is implied in Words among the first men. Tale = tale. Word,
       wahr, wehr — truth, troweth, throweth i.e. hitteth = itteth = it is it.
       The aspirate expresses the exclamatia of action. Through, & Truth —
       Etymol.38
In his haste to demonstrate that truth and language originate in a primordial
‘action,’ however, Coleridge chooses not to explore the further implications of his
remark that ‘Truth is implied in Words among the first men.’ The possibility that
  35 Samuel Taylor Coleridge, Lectures 1808-1819 On Literature, ed. R.A. Foakes, vol. 1 (New
Jersey: Princeton University Press, 1987) 272-273.
  36 See Heather J. Jackson, ‘Coleridge, Etymology and Etymologic,’ Journal of the History of
Ideas 44.1 (1983) 81. Similarly, James McKusick argues in Coleridge’s Philosophy of Language
that ‘Home Tooke was a seminal influence throughout most of Coleridge’s intellectual career’
(39).
  37 Coleridge, Aids 10.
  38 Coleridge, Notebooks 2 Text 2354. In her note to this passage, Kathleen Coburn argues that
Coleridge was misled by Adelung’s dictionary, which ‘suggests, wrongly, that wahr is identical
with the past tense, war, of the verb “to be”.’ As for ‘truth, troweth, throweth’: ‘The first two
words are obviously connected, but neither of them with the third.’ Coleridge’s speculations
on the origins of ‘hitteth,’ meanwhile, ‘seem devoid of any philological basis, nor is there any
etymological connexion between “through” and “truth”.’
44                                          Tim Milnes
truth might simply be the presupposition of communication, the basic precondi-
tion for language to have meaning at all, is passed over. In Coleridge, linguistic
explanations must ultimately rest upon ontological or metaphysical theories. Not
for the last time, the potential of ‘etymologic’ is sacrificed to Coleridge’s quest for
ultimate etymologoi.
(3)     This brings us to Prickett’s ‘third’ stage in Coleridge’s changing views on
language. Parallel to his etymological speculations, Coleridge developed a meta-
physical picture of language evolving through a spiritual process of dialectic. Ety-
mologic was certainly one way of comprehending this evolution. But while etymo-
logic helped one to understand the outward signs of this progression — by, as it
were, putting the process into reverse — it could provide no insight into its basic
principles. Consequently, when Coleridge writes Logic (ostensibly a handbook on
logic but, as McKusick notes, equally a theory of language) and Aids to Reflection
in the 1820s, the role of etymologic is reduced to that of under-labourer, a mere
supplement of theosophy. Thus, when Coleridge declares in the Preface that one of
the main aims of Aids to Reflection is to highlight the advantages of knowing words
in ‘their primary, derivative, and metaphorical senses,’ these varieties of meaning
are seen as determined by a prior metaphysical order of unity, antithesis and syn-
thesis.39 As Michael Kent Havens notes, the ‘three types of meaning correspond
respectively to a primal unity of language, the ongoing process of distinction or
desynonymization, and the reuniting of distinguished meanings in a newly holistic
language.’40
   Desynonymization is a key element in Coleridge’s theory of language.41
Coleridge sees the development of language as an organic process whereby equiva-
lences between terms are broken down, just as branches and leaves emerge from an
originally undifferentiated trunk. As Coleridge explains in one of his philosophical
lectures of 1819, one must be conscious
       that the whole process of human intellect is gradually to desynonymise
       terms, that words, the instruments of communication, are the only
       signs that a finite being can have of its own thoughts, that in proportion
       as what was conceived as one and identical becomes several, there will
       necessarily arise a term striving to represent that distinction.42
This emergence of difference, however — a process that is both conceptual and
historical — is merely the furtherance of a richer unity already inherent in the
totality of the organism, or language. For Coleridge then, the fate of formal logic
is bound up with the progress of language as it strives throughout history to
  39 Coleridge, Aids 1.
  40 Michael Kent Havens, ‘Coleridge on the Evolution of Language,’ Studies in Romanticism 20
(1981): 167.
  41 Paul Hamilton argues that Coleridge repressed his argument concerning desynonymy in Bi-
ographia Literaria because of his concerns about how its potentially radical political implications
might be interpreted. See Coleridge’s Poetics 72.
  42 Coleridge, Lectures 1818-1819 1 212.
                                     Coleridge’s Logic                              45
become more fully itself, growing through stages of spiritual improvement towards
the vanishing point of ultimate reunification with the Logos in higher synthesis.
Consequently, the condition of a language at any given time is an expression of the
stage of a culture’s spiritual and intellectual development. In an 1818 lecture to the
London Philosophical Society, Coleridge contrasts the primitive unity of ancient
Greek with the atomistic, aggregative logic of northern European, or ‘Gothic’
languages. In this way, Greek societies, leaning ‘to the manifold and popular, the
unity in them being purely ideal, namely of all as an identification of the whole’
differed from the ‘northern or Gothic nations,’ in which ‘the individual interest
was sacred’:
       In Greek the sentences are long, and structure architectural, so that
       each part or clause is insignificant when compared with the whole. But
       in the Gothic [...] the structure is short, simple, and complete in each
       part, and the connexion of the parts with the sum total of the discourse
       is maintained by the sequency of the logic, or the community of feelings
       excited between the writer and his readers.43
In this light, it is the privilege of modernity to complete the dialectic by which
these traditions are united. Only in the greatest writers, however, does Coleridge
find clear evidence of the growth of the Logos through language, and only in
Shakespeare does the fallen status of human language as a system of arbitrary
signs achieve partial redemption. As he writes in his lecture notes, Shakespeare’s
language is neither the original ‘Language of Nature’ that ‘was with the Thing,
 it represented, & it was the Thing represented,’ nor the artificial language of the
civilized intellect, but
       a something intermediate, or rather it is the former blended with the
       latter, the arbitrary not merely recalling the cold notion of the Thing
       but expressing the reality of it, & as arbitrary Language is an Heir loom
       of the Human Race, being itself a part of that which it manifests.44
For Coleridge then, as he puts it in Biographia, ‘besides the language of words,
there is a language of spirits,’ and ‘the former is only the vehicle of the latter.’45
The position of logic in this arrangement, however, remains ambiguous: formally,
a mere stepping-stone of abstraction in language’s return to the Logos, it is also
destined, in its higher, metaphysical incarnation, to form the Logos itself.
                                4  THE IDEA OF LOGIC
Coleridge’s views on Shakespearean language in 1818 show how his reading in
German philosophy is leading him to think of language and logic in more frankly
  43 Coleridge, Lectures 1808-1819 2 231.
  44 Coleridge, Lectures 1808-1819 1 429.
  45 Coleridge, Biographia 1 290.
46                                      Tim Milnes
idealistic terms. And it is principally the influence of German thought, particularly
that of Kant, which leads to Coleridge’s hesitation over whether to confine ‘logic’
to the abstract understanding or renew it as a higher discipline of knowledge. Not
for the first time, the dilemma facing Coleridge was between revolution and reform.
Having already resolved to recover reason from the Jacobins, rehabilitating it as
spiritual insight, could he do the same for logic? In both cases, it was German
idealism that suggested the path to redemption. In Biographia, Coleridge recounts
how his early reading of Kant introduced him to a new way of thinking about the
nature of experience:
       I began then to ask myself, what proof I had of the outward existence
       of any thing? Of this sheet of paper for instance, as a thing in itself,
       separate from the phenomenon or image in my perception. I saw, that
       in the nature of things such proof is impossible; and that of all modes of
       being, that are not objects of the senses, the existence is assumed by a
       logical necessity arising from the constitution of the mind itself, by the
       absence of all motive to doubt it, not from any absolute contradiction
       in the supposition of the contrary.46
The use of the term ‘pheænomenon’ here — together with the way in which he dis-
tinguishes the type of thought process he describes from empirical ‘perception’ on
one hand, and, on the other, formal questions relating to ‘absolute contradiction’
— reveals the influence of Kant’s method upon Coleridge’s changing conception
of ‘logical necessity.’ The kind of argument that Coleridge identifies with the new
‘logic’ is nowadays, following Kant, widely classed as a transcendental argument,
whereby the truth of a statement in dispute is shown to be necessarily implied by
the same conditions that constitute the possibility of an indubitable statement.
What varies between different transcendental arguments, however, is the inter-
pretation of ‘necessarily.’ Kant’s arguments have long been criticized for their
tendency to blur the distinction between conceptual necessity and mere psycho-
logical incorrigibility. If anything, however, Coleridge is even more inclined than
Kant to fall into what Frege would later deplore as the conflation of formal and
psychological arguments, equating ‘logical necessity’ with the limits on thought
dictated by ‘the constitution of the mind itself.’ Moreover, as demonstrated by
a schema of the human faculties sketched by Coleridge around the same time, he
saw this new kind of knowledge not (as Kant did) as the foundation of knowledge
in general, but as a ‘logical’ intermediary connecting empirical perception and
spiritual truths accessed through intuition:
       We may usefully distribute all our knowleges into four sorts [? or]
       Orders. First, those derived from the senses by the aid of the Under-
       standing.
       Second, those derived from the Understanding by reflection on its own
       acts and processes.
  46 Coleridge, Biographia 1 200-201.
                                      Coleridge’s Logic                                     47
       Third: those derived from the pure Reason.
       Fourth: those derived from the Conscience.47
The new logic is covered by the ‘second’ order of knowledge, which is the province
of the understanding. Coleridge’s scheme, it should be noted, is hierarchical: the
knowledge of understanding is subordinate to that of reason, and that, in turn, to
Conscience (which he often refers to as practical reason). This architectonic lies
behind Coleridge’s ambition to complete his great work on the intellect, the work
that in a contemporaneous letter to John May he describes as the ‘Logosophia:
or on the Logos, divine and human, in six Treatises.’ Logic, accordingly, takes its
place in volume two:
       The first, or preliminary treatise contains a philosophical History of
       Philosophy [...]. The second Treatise is [...] on the science of connected
       reasoning, containing a system of Logic purified from all pedantry &
       sophistication, & applied practically to the purpose of ordinary life,
       the Senate, Pulpit, Bar, &c. [...] The III. (Logos Architectonicus)
       on the Dynamic or Constructive Philosophy — preparatory to the IV.
       or a detailed Commentary on the Gospel of St John [...]. The Vth.
       [...] on the Panthesists and Mystics [...] The Vlth on the Causes &
       Consequences of modern Unitarianism.48
And yet, the master method of this system remains the ‘logic’ of transcendental
argument — only, in Coleridge’s hands, Kant’s transcendental method is extended
from the theoretical or purely cognitive into the realm of the practical. As he
explains in Biographia, the use of spatial intuitions as postulates in geometry
supplies philosophy with an example of the foundation ‘from which every science
that lays claim to evidence must take its commencement. The mathematician
does not begin with a demonstrable proposition, but with an intuition, a practical
idea.’49 Conceived as inherently moral, the foundational intuition reveals the
original unity of knowing and being. It is in this context that Coleridge urges his
favourite Delphic injunction upon his reader as the ‘first postulate’ of philosophy:
       The postulate of philosophy and at the same time the test of philo-
       sophic capacity, is no other than the heaven-descended know thyself!
       [...]. And this at once practically and speculatively. For as philosophy
       is neither a science of the reason or understanding only, nor merely
       a science of morals, but the science of being altogether, its primary
       ground can be neither merely speculative or merely practical, but both
       in one.50
  47 Coleridge, Shorter Works 1 412.
  48 Coleridge, To John May,’ 27 September 1815, letter 976 of Letters 4 589-590.
  49 Coleridge, Biographia 1 250. See also Aids 136: ‘we proceed, like the Geometricians, with
stating our postulates; the difference being, that the Postulates of Geometry no man can deny,
those of Moral Science are such as no good man will deny.’
  50 Coleridge, Biographia 1 252. On similar grounds, Coleridge later criticises Aristotle for
48                                           Tim Milnes
As a result, Coleridge’s treatment of logic bears the strain of two competing imper-
atives: first, that philosophy should be unified and complete, a seamless panoply
of thought and faith, and second, that ‘logic’ as such falls within the remit of the
understanding alone. In particular, it is when he tries to define a metaphysical
logic on the basis of the model of transcendental argument inherited from Kant
that Coleridge runs into difficulties. It proved impossible to dovetail the relatively
modest, conceptual aims of transcendental argument with the higher dialectic of
knowing and being planned for the Logosophia, involving as the latter did ques-
tions of will, faith, and the overcoming of logical categories and conceptual clarity
associated with the mere understanding.
    This tension is one reason why Logic, putatively Coleridge’s study of ‘the sci-
ence of connected reasoning’ in the understanding and the propaedeutic to the
‘Dynamic or Constructive Philosophy,’ remains incomplete. Coleridge divides his
subject into the ‘organon or logice organica, heuristica [...] the criterion or logice
dialectica, and the canon or logice simplex et syllogistica.’51 In other words, logic
subdivides into the metaphysical logic of discovery, the transcendental logic of psy-
chology, and the purely formal logic of the syllogism. However, only the sections
on canonic and dialectic logic were completed. On the face of it, this absence
is surprising, given the fact that Logic is full of apologetic digressions where Co-
leridge teasingly anticipates the transition between conceptual understanding and
spiritual insight. Indeed, at times it seems as if he cannot wait to cut through the
dry formalities in order to penetrate the metaphysical order behind reason.52 In
this respect, Logic bears an uncanny resemblance to that earlier and more notori-
ous case of deductive interruption, Biographia, in which Coleridge left incomplete
his metaphysical argument designed to underpin a poetics grounded in the prin-
ciples of imagination. The reasons behind Coleridge’s omission of organic logic,
together with his decision not to include ‘polar logic’ in the rubric of the work,
will be considered later.
                    5 THE CANON, OR LOGIC AS SYLLOGISM
Coleridge’s tendency in the Logic to digress from pure or formal logic into meta-
physics is one symptom of his impatience to broaden the subject and branch out
into what he sees as deeper and more urgent questions. Another is his painstaking
articulation of the place of logic within the hierarchical structure of knowledge.
Noting that the ancient Greeks considered ‘the mind in the threefold relation,’
having ‘confounded Science with Philosophy,’ when in truth, ‘Philosophy is the middle state
between Science or Knowledge and Wisdom or Sophia’ (Samuel Taylor Coleridge,Table Talk, ed.
Carl Woodring, vol. 1 [Princeton University Press, 1990] 173-174).
  51 Samuel Taylor Coleridge, Logic, ed. J.R. de J. Jackson (Princeton University Press, 1981)
52.
  52 See, for example, Logic 87: while ostensibly still treating of ‘common or syllogistic logic,’
Coleridge admits to drifting into territory concerned with ‘the last and concluding stage of our
present subject, viz. the transition of the dialectic into the organic [...].’
                                     Coleridge’s Logic                                   49
that is, relative to the evidence of reason, understanding and the senses, he ranks
the metaphysical disciplines accordingly:
       A — Noetics = the evidence of reason
       B — Logic = the evidence of the understanding
       C — Mathematics = the evidence of sense
Under the heading ‘physics,’ he lists:
       D — Empiric = evidence of the senses
       Scholium. The senses = sense+sensation+impressions.53
In this Kantian division of labour, ‘Logic’ is clearly defined as the province of the
understanding, mediating between Mathematics and Noetics:
       Thus by the mathematic we have the immediate truth in all things nu-
       merable and mensurable; or the permanent relations of space and time.
       In the noetic, we have the immediate truth in all objects or subjects
       that are above space and time; and, by the logic, we determine the
       mediate truths by conception and conclusion, and by the application
       of all the world to the senses, we form facts and maxims of experience
       which is one of the two provinces in and on which the formal sciences
       are to be employed and realised.54
The distinctly un-Kantian definition of ‘Noetics’ as concerned with ‘the immediate
truth in all objects or subjects that are above space and time’ betrays, once again,
Coleridge’s impatience with the Critical Philosophy and his tendency to equate
Kant’s work with logic as the science of understanding. On this picture, Logic is
limited to only one facet of human knowledge because it expresses only one side of
the power of self-consciousness. As he expresses it in marginal comments to Kant,
‘in Logic the mind itself being the Agent throughout does not take itself into
question in any one part. It is a Teller which does not count itself; but considers
all alike as Objective, because all alike is in fact subjective.’55
   This lack of self-consciousness betrays a naivete inherent in formal logic, one
which Coleridge attempts to explain in the introductory chapters of Logic in both
conceptual and historical terms. Here, the first emergence of logic is figured as an
adolescent stage through which the fallen or finite human mind must develop in its
journey towards a reunion with its origin in the plenitude of the Logos, the unity
of being and act. ‘In the infancy and childhood of individuals (and something
analogous may be traced in the history of communities),’ he claims, this logical
phase is preceded by a kind of ‘happy delirium, the healthful fever, of the physical,
moral, and intellectual being, Nature’s kind and providential gift to childhood,’ in
  53 Coleridge, Logic 43-44.
  54 Coleridge, Logic 36.
  55 Samuel Taylor Coleridge, Marginalia, eds. H. J. Jackson, George Whalley, et al, vol. 3
(Princeton University Press, 1980-2001) 275.
50                                      Tim Milnes
which ‘the first knowledges are acquired promiscuously.’56 As it grows, the infant
mind / society develops the ability to trace similarities between things, and thus
form abstractions. The ‘earliest products of the abstracting power’ are words,
which also ‘become the first subject matter of abstraction; and consequently the
commencement of human education.’57 As language becomes increasingly abstract
and complex, it nonetheless retains an organic relation to its origin. As Coleridge
explains, the community of parts of speech ‘is a necessary consequence of their
common derivation or rather production from the verb substantive.’ For ‘the verb
substantive (“am” [...]) expresses the identity or coinherence of being and act. It
is the act of being. All other words therefore may be considered as tending from
this point [...].’58
   This organic solution to the relationship between logic and language can be
seen as Coleridge’s contribution to a debate about the possibility of a ‘universal
grammar.’ Inherited from Port Royal Logic, the notion of a purely rational lan-
guage had been undermined in the late eighteenth century by the nominalism of
Tooke and, more damagingly, by Kant’s insistence that logic was the domain of
understanding rather than reason. As we have seen, however, Coleridge, ever the
philosophical magpie, manages to borrow what suits him from both these thinkers.
He audaciously combines an adapted form of Kantian logic with a metaphysical
rendering of Tooke’s etymology to argue for the ideal unity of grammar and logic.
This line of argument is pursued further in the Magnum Opus:
       in fact, the science of grammar is but logic in its first exemplification,
       or rather in its first product, Loγoς, discursus, discourse, meaning [...]
       either, i.e. thoughts in connexion, or connected language, and the
        primary distinctions of identity and alterity, of essence and form ,
       of act and of being, constituting the groundwork and, as it were, the
       metaphysical contents and preconditions of logic.59
The claim that ‘the science of grammar is but logic in its first exemplification’
is also Coleridge’s main premise in Logic, where he argues that one cannot un-
derstand logic without understanding language, and that one cannot understand
language without understanding the psychological and metaphysical act that un-
derlies being. Tellingly, the relatively short chapter on ‘Pure Logic or the Canon’
that opens the main body of Logic (53-59) is swiftly followed by a more substan-
tial chapter ‘On the Logical Acts’ (60-103), in which the reader is urged not ‘to
think too meanly of the process’ of reasoning by syllogism: ‘If A − C, and B be
contained in A, then B is contained in C. This [...] is a truth of mind: a somewhat
that in the mind actually exists, as any object of the senses exists without us.’60
  56 Coleridge, Logic 7-8.
  57 Coleridge, Logic 15
  58 Coleridge, Logic 16-17.
  59 Samuel Taylor Coleridge, Opus Maximum, eds. Thomas McFarland and Nicholas Halmi,
(Princeton University Press, 2002) 208.
  60 Coleridge, Logic 65.
                                     Coleridge’s Logic                                            51
Armed with his idealistic form of etymology, Coleridge outlines the metaphysi-
cal and grammatical foundations of canonic or syllogistic logic in a logical primer
inscribed for his son, Derwent, in a copy of the 1818 edition of The Friend:
       All Logic, as far as it is pure Science and of course purely formal, is
       reducible to three Acts of the Intellect — i.e. to the first Figure of
       Syllogism. These are: 1. Seclusion. 2. Inclusion. 3. Conclusion —
       and of these the first is the whole, as far as any effort of Judgement
       is concerned. The 2nd and 3rd are self-evident. To think absolutely or
       indefinitely is impossible, for a finite mind at least. To think (Ding,
       denken; res, reor) is to thingify. Thing = The Ing [...] is a somewhat
       set apart — thus Ingle = the Hearth. Conceive the indistinguishable
       all of our Perceptions, Conceptions and Notions as a vast Common
       — In or from this I seclude a determinate portion [...] by assigning
       it’s Terminus = Terminus Major sive communis. 1. All men (= the
       portion) are mortal = Term. Major. Within this common ring-fence I
       include with a fence of it’s own Socrates = Terminus Minor.61
Having set out to demonstrate the basis of syllogistic logic in ‘three Acts of the
Intellect,’ Coleridge suddenly sweeps his reader away on an etymological excur-
sion whereby, first, the origin of thinking is linked to ‘thingifying,’ and secondly,
the process of thingifying is itself traced, via the suffix ‘ing,’ to the way in which
we ‘ring-fence’ the vast, indistinguishable ‘Common’ of our perceptions in logical
reasoning. What seems at first a whimsical flight of speculation is entirely char-
acteristic of Coleridge’s thinking on this subject, and consistent with his other
writings. The same premise that logic, grammar and mind emerge from the pri-
mordial unity of an act forms the basic assumption of Logic. By 1818, Coleridge
is confident that he has answered the questions he posed to Godwin fifteen years
earlier.
   Canonic logic, however, constitutes only one part of Coleridge’s tripartite con-
ception of logic proper. A full account of logic, he maintains, must include an
account of the relationship between logic and truth, and for the logician, ‘all
truth, and consequently all true knowledge, rests on the coincidence of the object
with the subject.’62
                  6 THE CRITERION, OR LOGIC AS DIALECTIC
This brings us to the form of logic Coleridge terms the ‘criterion’ or ‘dialectic.’
The appearance of the latter term, together with Coleridge’s identification of the
central concern of dialectical logic as the problem of the ‘coincidence of subject
and object,’ attests to Kant’s influence on this part of Logic. Indeed, it has long
  61 Coleridge, ‘To Derwent Coleridge,’ November 1818, letter 1152 of Letters 4 885. Coleridge
presents the same etymological account of ‘thing’ in Logic 114-115: ‘The “ing” [...] is the universal
exponent of whatever is inclosed, bounded.’
  62 Coleridge, Logic 37.
52                                            Tim Milnes
been accepted that how one reads the Logic as a whole will be in no small part
determined by how one views Coleridge’s relation to Kant. Controversy over the
exact use Coleridge makes of Kant’s philosophy dates back to contemporaries such
as Henry Crabb Robinson, who wrote the following, not uncritical, assessment:
         To Kant his obligations are infinite, not so much from what Kant has
         taught him in the form of doctrine as from the discipline Kant has
         taught him to go through. Coleridge is indignant at the low estimation
         in which the Post Kantians affect to treat their master. At the same
         time Coleridge himself adds Kant’s writings are not metaphysics, only
         a propaedutic. Were Coleridge in Germany he would not be suffered to
         hold this language; he would be forced to make his election between the
         critical and the absolute philosophy, or he would be equally proscribed
         by both.63
Robinson’s parting shot — the claim that, were he writing in Germany, Coleridge
would not be allowed to have it both ways, borrowing from the Critical Philosophy
where it suited him while simultaneously adopting the more full-blooded idealism
of Kant’s successors, Fichte and Schelling — prefigures the kind of debates over
Coleridge’s debts to Kant that have occupied scholars and commentators for at
least the past century. The fact that Logic is widely seen as Coleridge’s most
explicitly ‘Kantian’ work means that much of this controversy has centred on the
manuscript originally intended as primer for the ‘forensic purposes’ of discourse.
Ranged on one side are those who argue that Coleridge’s handling of logic contains
little new of significance, either (as Rene Wellek argues) because Coleridge sim-
ply misunderstood Kant’s logic or (as J.R. de J. Jackson maintains) because an
exposition of Kant was the limit of his aims in this particular work, or indeed (as
G.N.G. Orsini allows) for both these reasons.64 On the other side of the argument,
however, is a tradition of criticism that has defended Coleridge as an innovator in
logic who improves on Kant, whether (as John Muirhead claims) because of the
way his use of triadic conception of reason logic trumps the dichotomic logic of the
first Critique, or (as James McKusick, Tim Fulford, Gerald McNeice, and Thomas
McFarland have all proposed) because, by playing ‘Tooke,’ as it were, to Kant’s
‘Locke,’ Coleridge successfully introduces a ‘linguistic turn’ into transcendental
idealism.65
   63 Coleridge, Table Talk 2 485.
   64 In Immanuel Kant in England 1793-1838 (Princeton University Press, 1931), Rene Wellek
claims that Coleridge failed to see ‘that nothing of the Kantian epistemology can be preserved
in a new system’ (80). However, in his Editor’s Introduction to the standard edition, J.R. de
J. Jackson counters that Coleridge ‘does not offer any new arguments’ in the Logic because it
is ‘essentially a popularisation of the Critique of Pure Reason’ (lxii). G.N.G. Orsini combines
both points of view in Coleridge and German Idealism (Southern Illinois University Press, 1969),
arguing that while Logic ‘is professedly an exposition of Kant’ (115), Coleridge was hampered by
the fact that he ‘did not fully grasp, or perhaps did not fully accept’ (256) Kant’s argument for
the transcendental unity of apperception in the second edition of the Critique of Pure Reason.
   65 As long ago as 1930, in Coleridge as Philosopher (George Allen & Unwin Ltd., 1930), John H.
Muirhead praised Coleridge’s proto-Hegelian ‘triadic logic’ as ‘an attempt to carry the dialectic
                                       Coleridge’s Logic                                      53
   Whatever one’s position on this perennial topic, the imprint of Kant’s influence
on Logic is undeniable. Having covered ‘canonic’ or syllogistic logic in Part One
in a tidy fifty pages, and with no third section on ‘organic’ logic ever written,
Part Two of Logic, on ‘The Criterion or Dialectic,’ occupies the remaining one
hundred and sixty four pages of the standard edition, thus making up the bulk
of the work as a whole. With chapter titles such as ‘Judicial Logic, Including the
Pure Aesthetic’ and ‘On Synthesis a priori,’ The ‘Dialectic’ makes little attempt
to disguise its reliance upon the Critique of Pure Reason.
   The concern of dialectic, as Coleridge makes clear, is with truth as the ‘coinci-
dence’ of subject and object. However, he notes, this is an ambiguous question.
The question, ‘what is truth?’ could mean three things, quite apart from the
verbal definition of truth, whereby ‘ “If you ask for the meaning of the word [...]
I reply, ‘the coincidence of the word with the thought and the thought with the
thing’”.’ First, the question could mean, what is truth, relatively to God? In this
case, the question has either no meaning or admits of but one reply, viz. “God
himself.” God is the truth, the identity of thing and thought [...]’ But this can-
not be a ‘criterion’ of truth, since a criterion implies an abstraction, which God
is not. ‘Truth, therefore, is its own criterion [...].’66 Secondly, the question may
refer to the ‘coincidence between thought and the word’ most often implied in
the ‘communication’ of truth. While acknowledging that ‘the criteria must vary
with the occasion and the purpose,’ Coleridge links this kind of question with
the search for what ‘metaphysicians have named the thing-in-itself,’ independent
of all human conception, a search that he claims permits of ‘no answer at all.’67
Thirdly, however, ‘[w]e may mean subjective truth, in which truth may be defined
[as] the coincidence between the thought and the thinker, the forms, I mean, of
the intellect.’ And this is the very office of logic as dialectic, which ‘as far as it
concerns the knowledge of the mere form abstractedly from the matter,’ presents
‘the criterion of truth, that is, of formal truth [...].’68
   Significantly, this has the consequence that, strictly speaking, logic is ‘not philo-
sophical; for logic [...] consists in the abstraction from all objects. It is wholly
and purely subjective,’ and ‘has no respect to any reality independent of the
mind.’69 The dialectic of logic is necessary, however, as a propaedeutic and a guard
against attempts to identify the ‘thing-in-itself,’ to rush into the ontology of truth
without having first established one’s grounds in the psychology of truth. For
of Kant’s thought a step farther and turn criticism against the critic’ (89). More recently, Tim
Fulford in Coleridge’s Figurative Language (MacMillan, 1991) (106), Gerald McNiece in The
Knowledge that Endures: Coleridge, German Philosophy and the Logic of Romantic Thought
(MacMillan, 1992) (21), and Thomas McFarland — in his editorial ‘Prolegomena’ to Opus Max-
imum — have agreed with McKusick’s assessment in Coleridge’s Philosophy of Language that,
in McFarland’s words, ‘[m]uch more clearly than Kant, Coleridge sees that epistemological ques-
tions cannot be resolved without recourse to a prior analysis of the linguistic structures that
constitute the means of intellectual inquiry’ (ccxxix).
  66 Coleridge, Logic 111.
  67 Coleridge, Logic 112-114.
  68 Coleridge, Logic 112.
  69 Coleridge, Logic 127.
54                                        Tim Milnes
Coleridge, the worst offenders in this respect are materialist philosophers such as
Hartley and Godwin, who through their neglect of logic unwittingly substituted
philosophical Hypopoeisis for philosophical ‘Hypothesis.’ As he argues in an 1809
notebook entry, ‘Hypothesis’ consists in ‘the placing of one known fact under oth-
ers as their ground or foundation. Not the fact itself but only its position in a
[...] certain relation is imagined.’ Where both the position and the fact are imag-
ined, it is Hypopoeisis not Hypothesis, ‘subfiction not supposition.’70 Increasingly
influenced by Kant, Coleridge came to see logic as the art of supposition, prelim-
inary to philosophy. Thus, the architectonic of Kant’s Transcendental Aesthetic
and deduction of the categories of understanding became exercises in a prepara-
tory discipline of logical hypothesis whereby the mind shed itself of the naive
objectivism that would only produce subfiction or Hypopoeisis. Later, in Aids to
Reflection, he associates this discipline with an awareness of forethought, writing
that it ‘is at once the disgrace and the misery of men, that they live without fore-
thought. Suppose yourself fronting a mirror. Now what the objects behind you
are to their images at the same apparent distance before you, such is Reflection
to Fore-thought.’71
    A crucial element in the hypothesis of dialectic is transcendental method. Al-
ready in Biographia Literaria, Coleridge shows how Kant’s demonstration of the
necessary conditions of experience has affected his insight into the Hypopoeisis of
the materialists. The errors of Hartley and associationist philosophy, he argues,
        may be all reduced to one sophism as their common genus; the mis-
        taking of the conditions of a thing for its causes and essence; and the
        process by which we- arrive at the knowledge of a faculty, for the fac-
        ulty itself. The air I breathe, is the condition of my life, not its cause.
        We could never have learnt that we had eyes but by the process of
        seeing; yet having seen we know that the eyes must have pre-existed
        in order to render the process of sight possible.72
In other words, knowing that eyes are the precondition of sight requires more than
sight itself, it requires a kind of reasoning that goes beyond the mere ‘conditions of
a thing,’ to the ‘cause’ that makes it possible. Later in Biographia, Coleridge gives
the same point a more recognizably Kantian formulation: ‘We learn all things
indeed by occasion of experience; but the very facts so learnt force us inward on
the antecedents, that must be pre-supposed in order to render experience itself
possible.’73 In Logic, following Kant still more closely, he terms this method
of reasoning ‘transcendental,’ noting that ‘the term “transcendental” means the
same as “sciental,” but with an additional significance. All knowledge is excited
or occasioned by experience, but all knowledge is not derived from experience,
such, for instance, is the knowledge of the conditions that render experience itself
   70 Coleridge, Notebooks 3 text 3587.
   71 Coleridge, Aids 12.
   72 Coleridge, Biographia 1 123.
   73 Coleridge, Biographia 1 142.
                                      Coleridge’s Logic                            55
possible [...].’74
   Thus far, Coleridge’s logic of dialectic looks similar to Kant’s transcendental
method. This is why much of the later sections of Logic in particular read as free
translations and adaptations from the Critique of Pure Reason. It is when the
‘Criterion’ is placed within Coleridge’s broader conception of philosophy, however,
that serious differences emerge. Coleridge learns from Kant the power of transcen-
dental argument, or thinking by way of presuppositions, but he has a very different
conception of the status and nature of those presuppositions. ‘Fore-thought’ for
Coleridge represents not just the conceptual or even psychological conditions that
make thought or experience possible, but the half-hidden presence of an Idea that
bears life, encompassing the very dimensions of will, love, and intuition that Kant
had taken such care in the first Critique to isolate from transcendental deduction.
Thus, like Kant, Coleridge is apt to hypostatize the preconditions of thought, as-
suming that the thing presupposed must itself have foundational status. But while
for Kant this foundation is the ‘I think’ or transcendental unity of apperception,
for Coleridge it is the ‘I am,’ the self-inaugurating word, or Logos.
   Unsurprisingly, Coleridge finds it difficult to connect Kant’s transcendental
method to the principles of his theosophy. The closest he comes to success is
in the ‘Essays on the Principles of Method,’ written for the 1818 edition of The
Friend (first published in 1809). ‘Method,’ he notes, ‘implies a progressive transi-
tion’ in reasoning. ‘But as, without continuous transition, there can be no Method,
so without a pre-conception there can be no transition with continuity. The term,
Method, cannot therefore, otherwise than by abuse, be applied to a mere dead
arrangement, containing in itself no principle of progression.’75 Thus, the presup-
position or ‘pre-conception’ underlying Method proper cannot simply be a truth
of formal logic, much less an empirical fact. It must be an Idea with a ‘life’ of its
own; it must, in short, be an organism, capable of growing and seeding further
thought. Coleridge calls this ‘leading Thought’ in Method the ‘Initiative.’76 To-
wards the end of his life, in conversation with Henry Nelson Coleridge, he refers
to this beefed-up transcendental method as ‘structive’ or ‘Synthetic’:
       There are three ways of treating any subject. 1. Analytically. 2.
       Historically. 3. structively or Synthetically. Of these the only one
       complete and unerring is the last. [...] You must begin with the philo-
       sophic Idea of the Thing, the true nature of which you wish to find
       out and manifest. You must carry your rule ready made if you wish
       to measure aright. If you ask me how I can know that this idea — my
       own invention — is the Truth, by which the phenomena of History are
       to be explained, I answer, in the same way exactly that you know that
       your eyes are made to see with — and that is — because you do see
       with them.77
  74 Coleridge, Logic 146.
  75 Coleridge, Friend 1 457.
  76 Coleridge, Friend 1 455.
  77 Coleridge, Table Talk 1 364-366.
56                                        Tim Milnes
The deployment once again of the ‘sight’ analogy indicates Coleridge’s continuing
belief in the link between Kant’s transcendental argument and the seminative and
self-authorizing power of the Ideal, ‘the philosophic Idea of the Thing.’ And yet,
an ‘Idea’ is more than just a logical construct, it is a postulate with an existential
component. Every ‘initiative’ of reasoning involves the input of the volitional,
emotional and spiritual life of an individual. Thus, Coleridge opens the Magnum
Opus with the declaration that ‘[i]n every science something is assumed, the proof
of which is prior to the science itself,’ adding that ‘the one assumption, the one
postulate, in which all the rest may assume a scientific form [...] is the Existence
of the Will [...].’ Linking logic to the Will would be the task of the polar logic of
the Divine Tetractys: in turn, bridging the logic of understanding (the canon and
the criterion) and this higher logic of reason was to be the task of the ‘organon or
logice organica.’78
                  7  THE ORGANON, OR LOGIC AS DISCOVERY
This brings us to the missing logic of Logic. In long footnote to the ‘Aphorisms
on Spiritual Religion’ in Aids to Reflection in which he presents his conception of
polar logic as the schema of a ‘Noetic Pentad’ (to which I shall return later),79
Coleridge directs his more logical-minded reader to the final section of his Logic:
       In the third and last Section of my ‘Elements of Discourse’; in which
       (after having in the two former sections treated of the Common or
       Syllogistic Logic, the science of legitimate Conclusions; and the Critical
       Logic, or the Criteria of Truth and Falsehood in all Premisses) I have
       given at full my Scheme of Constructive Reasoning, or ‘Logic as the
       Organ of Philosophy,’ in the same sense as the Mathematics are the
       Organ of Science [...].
Not for the first time, Coleridge’s redirected reader is destined to be disappointed.
Yet later in the same footnote, Coleridge hints at the reasons behind the absence
of the ’organon’:
       As this third Section does not pretend to the forensic and compara-
       tively popular character and utility of the parts preceding, one of the
       Objects of the present Note is to obtain the opinions of judicious friends
       respecting the expedience of publishing it, in the same form, indeed,
       and as an Annexment to the ‘Elements of Discourse,’ yet so as that
       each may be purchased separately.80
As the appeal to ‘judicious friends’ (his usual coded alert for a project that has
stalled) shows, Coleridge was already thinking of the ‘organon’ as sufficiently dif-
ferent to the ‘canon’ and ‘criterion’ to warrant publication in a separate volume.
  78 Coleridge, Opus Maximum 5, 11, 33.
  79 Coleridge, Aids 180.
  80 Coleridge, Aids 183.
                                     Coleridge’s Logic                                    57
The ostensible reason for this is clear enough: the first and second sections of
Logic are ‘forensic’ and ‘comparatively popular,’ insofar as they are intended to
be formal guides to the principles of reasoning for young professionals. They do
not (explicitly at least) presuppose any theological doctrine. However, logic con-
ceived as ‘Constructive Reasoning,’ as the ‘Organ of Philosophy,’ or ‘instrument
of Discovery and universal Method in Physics, Physiology, and Statistics’ was, as
Coleridge came to see, a method that required content.81 At this point, we realise
the extent of Coleridge’s ambitions in logic: in addition to what he saw as the
older logic of syllogism and exposition, he wanted to create a new logic that was
not only fit for modern science, but managed to anchor the growth of knowledge
in an ancient tradition of Christian metaphysics. It was to be a synthetic logic of
discovery, explaining and regulating the expansion of human consciousness. Such
was the goal of polar logic. The job of organic logic was to connect this higher
scientific/metaphysical logic to the formal/transcendental logic of the canon and
the criterion. Coleridge thought of this task as the attempt to build a bridge be-
tween reason and understanding. Ultimately, it seems, the organon proved to be
a bridge too far.
   As with transcendental method, Coleridge adapted the distinction between rea-
son and understanding from Kant. For Coleridge, Kant’s act of desynonymisation
could not have been more timely. He saw the confusion of the human mind with
the merely formal and mechanical understanding as the great mal de siecle, a
problem that loomed with the same menace as it had when Kant published the
first Critique. Indeed, if anything, the situation was worse in 1830, as Coleridge
complains in On the Constitution of the Church and State, declaring that thanks to
industry and utilitarianism, ‘we live [...] under the dynasty of the understanding:
and this is its golden age.’82 Coleridge’s later works signal a return to Kant after
the ill-fated flirtation with the idealism of Schelling and Fichte in Biographia Lit-
eraria. As Gian Orsini notes, one result of this is that after 1815 Coleridge places
less emphasis on the restorative power of the imagination. Instead, ‘the proper use
of Reason serves to correct the errors of the Understanding.’83 Indeed, Thomas
McFarland argues that the distinction between understanding and reason ‘was the
keystone in the arch of Coleridge’s thought; it was the ultimate, the ne plus ultra,
of all his mentation.’84 While this might be a slight exaggeration, it is certainly
true that Coleridge hoped, ‘by converting the reason of Hume and Diderot into
understanding, and marching to relieve the beleaguered garrison of Christianity
under the generalship of a new conception of reason [...] to wrest “reason” from
the hands of the Antichrist [...].’85 As early as 1812 Coleridge is writing in the
  81 Coleridge, Aids 183.
  82 Samuel  Taylor Coleridge, On the Constitution of the Church and State, ed. John Colmer
(Princeton University Press, 1976) 59.
  83 Orsini 140.
  84 Thomas McFarland, ‘Aspects of Coleridge’s Distinction Between Reason and Understanding,’
Coleridge’s Visionary Languages: Essays in Honour ofJ.B. Beer, eds. Tim Fulford and Morton
D. Paley (Cambridge: D.S. Brewer, 1993) 168.
  85 McFarland, ‘Aspects’ 171.
58                                        Tim Milnes
margins of his copy of Moses Mendelssohn of
       [t]he unspeakable importance of the Distinction between the Reason,
       and the Human Understanding, as the only Ground of the Cogency
       of the Proof a posteriori of the Existence of a God from the order of
       the known Universe — . Remove or deny this distinction, and Hume’s
       argument from the Spider’s proof that Houses &c were spun by Men
       out of their Bodies becomes valid.86
This innovation also had profound implications for Coleridge’s conception of logic.
Unlike Kant, Coleridge conceives of pure reason as constitutive and cognitive as
well as regulative and moral. Consequently, he believes that the canons of logic
must be demonstrably based upon higher principles of reason. This is because
reason as a faculty is not just superior to understanding, but a superintending
presence for all the faculties. As he argues in the Statesman’s Manual, ‘the Reason
without being either the Sense, Understanding or the Imagination contains all
three within itself, even as the mind contains its thoughts, and is present in and
through them all.’87 As Coleridge explains in The Friend, reason differs from
understanding in two key respects. First, it is immutable: ‘the understanding
may be deranged, weakened, or perverted,’ he maintains, ‘but the reason is either
lost or not lost, that is, wholly present or wholly absent.’88 Second, reason is
consubstantial with its objects: ‘Thus, God, the Soul, eternal Truth, &c. are the
objects of Reason; but they are themselves reason [...]. Whatever is conscious
Self-knowledge is Reason; and in this sense it may be safely defined the organ of
the Super-sensuous [...].’ Understanding, on the other hand, ‘supposes something
that is understood ’; it is the organ of the sensuous, depending upon a combination
of logic and the objects of experience.89
   The logic of the understanding then, derives its validity from principles or ideas
of reason, which, being ‘Super-sensuous’ and verbal in nature, supersede logical
canons. There are few areas where the subordination of formal logic to Coleridge’s
notion of rational principle is more glaring than in his treatment of predication
and the law of contradiction. Predication for Coleridge is the product of reflec-
tion, which is itself a lower power of intuition. Since logic depends upon the
pre-reflective unity in judgement, an act of mind, ‘neither the subject nor its
predicate,’ taken separately or together contain the principle of ‘their reality or
objective being.’ This principle is only realised ‘when the mind bears witness to
its own unity in the subject represented to it, and this act with this consciousness
of the same is conveyed or expressed in the connective “is”.’90 At the level of logic
or understanding then, it is through the connective or copula ‘is’ that the mind
affirms not only its own reality and unity, but its power to create that reality. As
  86 Coleridge, Marginalia 3 848.
  87 Samuel  Taylor Coleridge, Lay Sermons, ed. R. J. White (Princeton University Press, 1972)
69-70.
  88 Coleridge, Friend 1 153.
  89 Coleridge, Friend 1 156.
  90 Coleridge, Logic 79.
                                   Coleridge’s Logic                                59
Coleridge puts it, ‘the mind is, and it is a form, and it is formative,’ and predi-
cation is a reflection of this formative act of consciousness.91 Barfield explains this
idea further in What Coleridge Thought:
       The moment of predication is the moment in which the presence of rea-
       son in the understanding is manifested in its effect, but only as effect.
       To meditate faithfully on the principle of contradiction, upon which
       predication and syllogism are based, is to have one’s attention drawn
       [...] to the effective reality of Reason as a universal and constitutive
       principle.92
   However, as we saw with Coleridge’s etymological treatment of the syllogism, the
act of consciousness, indeed the act of being itself, is formed through language. ‘If
we ask ourselves how we know anything,’ he notes in Logic, ‘that rose, for example,
or the nightingale hidden in yonder tree [...] what are these but the goings from
the subject, its words, its verb? The rose blushes, the nightingale sings.’93 Thus,
just as the identity of subject and object is found in self-consciousness, ‘the first
rule of logic is in accordance with the first rule of grammatical syntax — that the
case which follows the verb substantive, and as such stands for the objective, is the
same with the case that precedes, that is, the subjective case.’94 The principles of
reason that govern the logic of understanding are both ideal and linguistic.
   Coleridge’s insistence on the metaphysical and linguistic roots of logic leads him
to distinguish between ‘natural’ and ‘verbal’ applications of logic. The former, in
which there is ‘an actual inclusion’ of the predicate by the subject, is ‘almost
automatic and spontaneous’ to the understanding. It is ‘a science of nature’s
rather than of our conscious self.’ Verbal logic however, is an echo of the divine
Logos, ‘partaking of this communicative intelligence’ through ‘ideas or communi-
cable forms.’95 It is within the latter concept of logic, rooted in the primordial
verbal performance, the ‘I am’ of the communicative Logos, that Coleridge believes
the canons of predication and contradiction must ultimately be framed. Thus, if
the principle of contradiction, which Coleridge gives as ‘impossibile est idem simul
esse [ac non] esse’ (it is impossible for the same thing at the same time to be and
not be), is to be a principle, ‘there ought to be something that should follow’:
       [B]ut what is to follow in the present instance? How can we arrive from
       this negative to a positive? Through another negative? As whatever is
       not ens and non ens [being and not-being] at the same time is possi-
       ble[?] But the rule that two negatives make a positive is grammatical,
       not logical.96
  91 Coleridge, Logic 80.
  92 Barfield 190.
  93 Coleridge, Logic 82.
  94 Coleridge, Logic 80.
  95 Coleridge, Logic 95-96.
  96 Coleridge, Logic 89.
60                                         Tim Milnes
    At the same time, the ‘rival position,’ the principle of identity, or “‘whatever
is, is”,’ is ‘merely a repetition during the act of reflection of the term “is”. I
mean that it expresses no more than my consciousness that I am reflecting, that
is, consciously reflecting the truth of being.’97 Thus, ‘[i]nstead of truth, principle,
or axiom, it is in reality a mere narration of a fact: I reflect on being.’ What
troubles Coleridge, like Hume and Kant before him, about the so-called principles
of contradiction and identity is that they tell us nothing new about the world, and
actually ‘encourage the error of presenting as particular truths the mere exponents
of our universal or essential consciousness.’ Ultimately, these axioms are analytic,
but the copula ‘is’ is itself ‘grounded, not in our reflection, or the analytic unity,
but in the synthetic.’98 Their validity derives from a unity whose origins lie in
the emergence of the universal consciousness into language and being through the
communicative Logos, in the form of the verb substantive, the ‘I am.’ The only
human faculty that can even approach this power is the intuitive reason.
    We can now see that the missing logic in Coleridge’s philosophy, the organon or
logic of discovery, was to be a logic that enabled the mind to make the transition
from understanding to reason, from ‘natural’ to ‘verbal’ logic, and from predication
and contradiction to the more fundamental act of the divine consciousness, which,
echoing in the human mind, subtended the division between analysis and synthesis.
We can also now appreciate why Coleridge runs into trouble. In effecting the
shift from ‘dialectic’ or transcendental argument to a metaphysical logic worthy
of theosophy and the ideas of reason, one might expect Coleridge to show how
argument by way of presupposition (dialectic) cannot account for the totality of
its own presuppositions. Thus, by effectively presupposing its own supersedence,
it would pave the way for a higher ‘polar logic’ of alterity. Indeed, it is for just this
proto-Hegelian move, as he saw it, that Muirhead applauds Coleridge, claiming
that the latter succeeds in turning ‘criticism against the critic.’99
    However, while this may be fine in theory, in Coleridge things are not so straight-
forward. First, Coleridge does not confine the criterion or transcendental logic to
questions of understanding, that is, to questions regarding the way in which con-
cepts are applied to experience, but frequently deploys the method in the realm
of ‘Noetics,’ or intuitive reason. Indeed, Coleridge routinely uses transcendental
argument in support of ideas, not concepts (what he calls ‘conceptions’). Thus, in
the Magnum Opus, the existence of ideas of reason, which are ‘contradistinguished
alike from the forms of the sense, the conceptions of the understanding, and the
principles of the speculative reason by containing its reality as well as the peculiar
form of the truth expressed therein,’ is established by transcendental argument,
since ‘[t]hat without which we cannot reason must be presumed [...] as the ground
  97 Coleridge, Logic 89.
  98 Coleridge, Logic 89.
  99 Muirhead 89. See also Kathleen Wheeler, ‘Coleridge’s Theory of Imagination: a Hegelian So-
lution to Kant?’ The Interpretation of Belief: Coleridge, Schleiermacher and Romanticism, ed.
David Jasper (London, 1986) 22: ‘Coleridge’s conclusion is one with Hegel’s: “The Subjectivity
of Reason is the great error of the Kantean system”.’
                                      Coleridge’s Logic                                     61
of the reasoning.’100
   Second, there is nothing in Coleridge to compare with Hegel’s argument in
the Phenomenology of Spirit that any ‘dialectic’ based on transcendental method
inevitably gives way to ‘dialectic’ based upon negation.101 As I have shown else-
where, this is because the conception of alterity that Coleridge locates at the
heart of his higher logic of reason is not driven by the power of negation, but by
the altogether obscurer agency of the Pythagorean Tetractys, which (among other
things) involves a metaphysics of Will.102 As he indicates in the passage from Mag-
num Opus mentioned above, the ‘source’ of ideas is ‘neither in the reason without
the Will nor in the Will without the reason.’ What this means is that Coleridge
cannot, like Hegel, demonstrate how the Kantian dialectic itself presupposes neg-
ativity. Instead, he relies upon para-philosophical means such as illustrations,
flashes of insight, aids to reflection and above all the willingness of his reader to
be guided, in order to lead his audience towards an illumination that cannot be
attained without an element of volition, or faith. In this respect, as Katherine
Miles Wallace points out, Coleridge saw his relation to the reader more as a guide,
albeit a ‘Chamois-hunter,’ than as a preceptor.103 As he argues in his Appendix
to the 1831 edition of Aids to Reflection, within Noetics, practical and theoretic
reason must work together:
        The Practical Reason alone is Reason in the full and substantive sense.
        It is reason in its own Sphere of perfect freedom; as the source of
        ideas, which Ideas, in their conversion to the responsible Will, become
        Ultimate Ends. On the other hand, Theoretic Reason, as the ground
        of the Universal and Absolute in all Logical Conclusion, is rather the
        Light of Reason in the Understanding [. . . ].104
As Coleridge found in the case of organic logic, one can gesture towards the light of
reason in the understanding as the Absolute of logical conclusion, but one cannot
demonstrate it. In the end, Coleridge’s failure to develop an organon betrays the
extent to which his interest in ‘dialectical’ logic or transcendental argument is
motivated by his unwarranted assumption that this form of reasoning was merely
a preparatory exercise for metaphysical illumination, as if Kant’s method was
merely the stepladder needed to reach the first step on the marble staircase of
‘polar logic.’
 100 Coleridge, Opus Maximum 270-271. Emphasis added.
 101 See  G.W.F. Hegel, Phenomenology of Spirit, trans. A.V. Millar (Oxford University Press,
1977) 29.
 102 Tim Milnes, ‘Through the Looking-Glass: Coleridge and Post-Kantian Philosophy,’ Com-
parative Literature 51.4 (1999): 309-323.
 103 Catherine Miles Wallace, The Design of the Biographia Literaria (London: George Allen
& Unwin, 1983) chapter 1: ‘The Chamois Hunter.’ Coleridge uses the metaphor in the 1818
Friend 1 55: ‘Alas! legitimate reasoning is impossible without severe thinking, and thinking is
neither an easy nor an amusing employment. The reader, who would follow a close reasoner to
the summit and absolute principle of any one important subject, has chosen a Chamois-hunter
for his guide.’
 104 Coleridge, Aids 413.
62                                       Tim Milnes
                  8   POLAR LOGIC AND THE NOETIC PENTAD
This brings us to the logic that transcended the science of understanding cov-
ered by Logic: the logic of reason. Indeed, with polar logic we are immediately
confronted by the question of what logic is, if it is not to be the articulation of
demonstrable principles of reasoning. As we have seen, this is more than just a
problem of nomenclature, in that its answer is bound up with Coleridge’s esti-
mation of the legitimate reach of philosophy within human life. Ever conscious
of what he saw as the hollow rationalism that lay behind the French Revolution,
Coleridge is vigilant about the dangers of an unregulated understanding. Con-
sequently, logic plays a systematically ambiguous role in his thought, potentially
both a menace and a comfort. On one hand, as the instrument of understanding,
it threatens to break out of its allocated canonical and dialectical functions. Un-
fettered, formal logic has dangerous tendency to become reified, collapsing back
into a damaging irrationalism. Thus, in Aids to Reflection, Coleridge complains of
‘Ideas or Theories of pure Speculation, that bear the same name with the Objects
of Religious Faith’ being taken for those objects themselves thanks to the natural
tendency of mind to form
       certain Essences, to which for its own purposes it gives a sort of notional
       Subsistence. Hence they are called Entia rationalia: the conversion of
       which into Entia realia, or real Objects, by aid of the Imagination, has
       in all times been the fruitful Stock of empty Theories, and mischievous
       Superstitions.105
For this reason, he maintains, following Kant, the purely spatio-temporal cate-
gories of understanding should not be extended ‘beyond the sphere of possible
Experience. Wherever the forms of Reasoning appropriate only to the natural
world are applied to spiritual realities, it may be truly said, that the more strictly
logical the Reasoning is in all its parts, the more irrational it is as a whole.’106
   Upon this reasoning, Coleridge bases his argument that the irrationalism of his
age is merely the reflex of an overextended faculty of understanding. One way in
which this pyrrhic triumph of logic is expressed is pantheism, a heresy with which,
as Thomas McFarland has shown, Coleridge struggles throughout his career.107
Dictating the fragments that were to form the uncompleted manuscripts for Lo-
gosophia or Magnum Opus, Coleridge claims to have demonstrated in the Logic
‘that Dichotomy, or the primary Division of the Ground into Contraries,’ though
‘the necessary form of reasoning as long as and wherever the intelligential faculty of
Man [weens] to possess within itself the center of its own System,’ can easily excite
a ‘delusive conceit of Self-sufficiency,’ the ‘inevitable result’ of which, as with ‘all
consequent Reasoning, in which the Speculative intellect refuses to acknowledge a
higher or deeper ground than it can itself supply, is [...] Pantheism.’108
 105 Coleridge, Aids 167.
 106 Coleridge, Aids 254.
 107 Thomas McFarland, Coleridge and the Pantheist Tradition (Oxford University Press, 1969).
 108 Coleridge, Opus Maximum 104-106.
                                    Coleridge’s Logic                              63
   On the other hand, Coleridge’s careful restriction of his critique to the dangers
threatened by ratiocination dominated by ‘Dichotomy,’ reminds us of his attraction
to logic, and his desire to free it from the confines of the understanding. Thus,
while in Logic he urges his reader not to ‘think too meanly’109 of formal logic,
elsewhere he criticises Robert Leighton for assuming that an argument ‘in point of
Logic, legitimately concluded’ is the same as a truth of reason. On the contrary,
‘the truths in question are transcendent, and have their evidence, if any, in the
Ideas themselves, and for the Reason; and do not and cannot derive it from the
conceptions of the understanding [...].’110 Similarly, in his marginalia to Johann
Christian Heinroth’ s Lehrbuch der Anthropologie, he complains of the ‘Followers
of Fichte, Baader, Schelling and Steffens’ that ‘they neglect Logic — or rather do
not understand what Logic is. Thus, what Kant asserted as an assumption for the
purposes of a formal Science, Heinroth asserts as a matter of fact. It is a necessary
fiction of pure Logic [. . . ].’ 111
   Where all these thinkers fail, according to Coleridge, is in their inability to
recognise the point where dichotomic logic (the logic of understanding) must give
way to polar or trichotomic logic (the logic of reason). The same distinction
informs the transition that Coleridge envisages in his 1815 letter to John May
between the preparatory discipline of the Logic or ‘Elements of Discourse’ and
that of the ‘Dynamic or Constructive Philosophy’ or ‘Logos Architectonicus.’ The
exposition and articulation of the latter was to be one of the main tasks of the
Magnum Opus. The limitations of the former, he claims, are evident in Emanuel
Swedenborg’s Prodromus Philosophiae:
       The Reasoning in these pages might be cited as an apt example of
       the inconvenience of the Dichotomic Logic: which acts in a contrary
       direction to the prime end and object of all reasoning, the reduction of
       the Many to One [...]. Two terms in manifest correspondence to each
       <other> are yet opposed as contraries, without any middle term: the
       consequence of which is, that one [...] of the [...] two becomes a mere
       negation of the other [...].112
For Coleridge, duality, opposition, difference, contradiction, all presuppose a more
fundamental unity. As he continues to explain, the negation in dichotomic logic is
‘a mere act of the mind, arising from a defect of perception.’ It is therefore nec-
essary to establish a logic that is able to express the way in which contraries are
always mediated by a ‘middle term,’ yoking dualities within a dynamic unity. The
only way this can be done, he maintains, is with a logic of ‘Trichotomy,’ similar
to that involved in ‘Pythagorean Tetractys.’113 Kant had introduced trichotomic
logic to eighteenth-century though by arguing that all logical oppositions presup-
pose a fundamental unity of consciousness in apperception, but as usual this did
 109 Coleridge, Logic 65.
 110 Coleridge, Marginalia 3 516.
 111 Coleridge, Marginalia 2 1002.
 112 Coleridge, Marginalia 5 459.
 113 Coleridge, Marginalia 5 459.
64                                         Tim Milnes
not go far enough for Coleridge. In his marginalia to Richard Baxter, Coleridge
complains that Kant takes trichotomy ‘only as a Fact of Reflection — [...] in which
he seems to anticipate or suspect some yet deeper Truth latent & hereafter to be
discovered.’ Baxter, on the other hand, is commended not just for prefiguring
Kant’s arguments a century earlier, but for grounding trichotomy ‘on an absolute
Idea presupposed in all intelligential acts.’114 Coleridge makes the same point in
a different way when, in The Statesman’s Manual, he distinguishes three kinds of
necessity. There is, as he puts it, ‘a threefold Necessity’:
       There is a logical, and there is a mathematical, necessity; but the latter
       is always hypothetical, and both subsist formally only, not in any real
       object. Only by the intuition and immediate spiritual consciousness
       of the idea of God, as the One and Absolute, at once the Ground
       and the Cause, who alone containeth in himself the ground of his own
       nature, and therein of all natures, do we arrive at the third, which
       alone is a real objective, necessity. Here the immediate consciousness
       decides: the idea is its own evidence, and is insusceptible of all other.
       It is necessarily groundless and indemonstrable; because it is itself the
       ground of all possible demonstration. The Reason hath faith in itself,
       in its own revelations.115
In this passage we can see how the principal forces in Coleridge’s thought conspire
to ensure the subordination of ‘formal’ (both logical and mathematical) neces-
sity: the prioritisation of intuitive reason and ‘immediate spiritual consciousness’;
the insistence on unity; the idea of the self-grounding and therefore fundamental
agency of consciousness in its act of ‘faith,’ and above all the primacy of a revela-
tory God ‘as the One and Absolute.’ For Coleridge, these principles confirm the
ascendancy of reason over understanding, and thus of intuitive Ideas over logical
rules. The law of contradiction, for example, is seen as secondary to the indemon-
strable principle (itself the engine of the growth of language through desynonymy)
of difference within unity, or distinction without division. Thus, in a fragment
on consciousness from around 1816, he notes that ‘all Ideas, when interpreted
into Conceptions [...] must appear to involve contra-dictions’ to those who never
move beyond the understanding: ‘in other words, as the Ideal Power can only
[...] manifest itself for the Senses in opposite Forces, so the Idea [...] relatively to
Speculation, can only be conceived by incompatible Conceptions [...].’116
    Difference, in other words, will always remain secondary to a fundamental unity,
which means that while difference might appear as contradiction in the understand-
 114 Coleridge, Marginalia 1 347-348.    In a further note, Coleridge links ‘the principle of
Trichotomy’ to the ‘Polar Logic’ of ‘Giordiano Bruno’s Logica Venatrix Veritatis’ and the
‘Pythagorean Tetractys’. For a meditation on the significance of Bruno’s sudden appearance
in this note, see Barfield’s Appendix to What Coleridge Thought.
 115 Coleridge, Lay Sermons 32.
 116 Coleridge, Shorter Works 1 430.
                                      Coleridge’s Logic                                      65
ing, it retains a one-ness in the Reason.117 Thus, dysynonymising terms in Church
and State, Coleridge remarks on ‘the essential difference between opposite and
contrary,’ arguing that unlike contraries ‘[o]pposite powers are always of the same
kind, and tend to union, either by equipoise or by a common product.’118 ‘Op-
position,’ however, is prior to contrariness; just as all contradictions presuppose
a unity, all contraries presuppose an opposition. Indeed, without this principle,
Coleridge’s theory of desynonymy as the means by which thought enriches and
renews itself through differentiation in language would lose its foundation: the
undifferentiated Logos, from which language descends and to which it seeks to
return.
   Nor is this the only consequence of the superiority of the ‘polar’ logic of ‘op-
positions’ over the formal logic of ‘contraries.’ In Church and State, Coleridge’s
opposite / contrary distinction is finally deployed to explain how the principles of
the state, like those of language, can be traced to the two opposite or ‘antagonist’
powers of ‘Permanence and of Progression.’119 Similarly, in his contributions to
J.H. Green’s lectures on aesthetics, Coleridge maintains that ‘the Beautiful ex-
cludes the distinct consciousness [...] of the forms of the Understanding — for
these are determined by a logical necessity’ and do not consider ‘an ultimate end’
but merely the means or ‘the mode of the conspiration of the Manifold to the One.’
However, he continues, ‘the direct contrary is the character of the Beautiful. The
manifold must be melted into the one,’ just as a ‘beautiful Piece of Reasoning’ is
‘not beautiful because it is [...] understood as truth; but because it is felt, as a
truth of Reason, i.e. immediate, and with [...] a facility analogous to Life.’120
   Above all, however, the impetus behind polar logic was theological. As McFar-
land observes, having already established through trichotomic logic that no two
positions in a dyad are equal, and that ‘trinal conceptions arise inevitably from
dyadic conceptions,’ it seems obvious to Coleridge that the next step is to replace
the ‘I-Thou’ dyad implicit in most unreflective conceptions of the relationship be-
tween the self and God, with an account that would vindicate the ‘reasonableness
of the Trinity.’121 This brings us back to the ‘Noetic Pentad,’ which Coleridge
models on the Pythagorean Tetractys. Included in a footnote to the ‘Aphorisms
on Spiritual Religion’ in Aids to Reflection, this schematism incorporates the dy-
namics of polar logic in a relational model of reality that Coleridge hopes will give
a philosophical underpinning to Trinitarianism. He first proposes adopting the
terms ‘objective and subjective reality, &c. as substitutes for real and notional,
and to the exclusion of the false antithesis between real and ideal,’ the relationship
between the two being that of ‘Thesis’ and ‘Antithesis.’122 Once again, however,
the relation between these points is dictated not by negation, as in Hegel, but by
 117 Indeed, Perkins claims that the principle of difference-in-unity ‘is the very foundation of
Coleridge’s system’ (39).
 118 Coleridge, Church and State 24.
 119 Coleridge, Church and State 24.
 120 Coleridge, Shorter Works 2 1313.
 121 McFarland, ‘Prolegomena,’ Opus Maximum cxlii.
 122 Coleridge, Aids 178.
66                                         Tim Milnes
a dynamic produced by Coleridge’s conflicting needs. On one hand, he wishes to
install the identity of mind and world as the foundation of knowing and being; on
the other, he seeks to secure a dynamic of mediation between the two that will
obviate any suggestion of a return to pantheism. As a result, the Pentad is organ-
ised according to Coleridge’s metaphysics of language, centring on the Prothesis
as ‘Verb Substantive,’ the Absolute copula or ‘I am,’ which expresses ’the identity
or co-inherence of Act and Being’:123
                                         1. Prothesis
                                [Verb Substantive: ‘I am’]
            2. Thesis                   4. Mesothesis             3. Antithesis
     [Substantive: ‘thing’]          [Infinitive: ‘to act’]  [Verb: ‘I act, undergo’]
                                         5. Synthesis
                                   [Participle: ‘acting’]
   It is here that the significance of Coleridge’s desynonymisation of ‘opposite’
and ‘contrary’ becomes clear, since it is only through the oppositions of polar
logic, and not the contraries of the logic of equation and contradiction, that the
dynamic relationship between world and self, thesis and antithesis, can come into
being. Thus, the subordination of the laws of formal logic to the principle of
difference-in-unity allows Coleridge, not for the first time, to have it both ways.
Accordingly, while the Prothesis is installed as a noumenous foundational identity,
the ‘Punctum invisibile, et presuppositum [the invisible and presupposed point],’
by which the Pythagoreans ‘guarded against the error of Pantheism,’ a mediating
term is introduced in the form of the ‘Mesothesis,’ expressing the ‘Indifference’ —
but not the identity — of subject and object.124 Inserting the mesothetical point
as ‘the Indifference of the two poles or correlative opposites,’125 as Perkins points
out, enables Coleridge finally to sever his links with Schelling by showing that the
Absolute foundation is the identity of ‘unity and distinction’ and not (as Schelling
claimed in his Identity Philosophy) the identity of ‘identity and distinction.’126
                                  9 BEYOND LOGIC
Ultimately, Coleridge’s philosophy aspires to move beyond argument, beyond even
polarity, into a horizon where knowledge merges with power. As the 1816 fragment
on the ‘four sorts of knowledge’ makes clear, only the ‘fourth’ source of knowledge,
the conscience, enables the reflective intellect to recover its original unity with the
transcendent agency of Will.127 Consequently, we run the risk of misrepresenting
Coleridge’s logic if we ignore the role it plays within a broader logosophic system
 123 Coleridge, Aids 180.
 124 Coleridge, Aids 180-181.
 125 Coleridge, Aids 179.
 126 Perkins 65.
 127 Coleridge, Shorter Works 1 412.
                                      Coleridge’s Logic                              67
that incorporates conscience, imagination, and faith. ‘We (that is, the human
race) live by faith,’ he maintains in The Statesman’s Manual ; faith ‘is scarcely
less than identical with its own being,’ thus ‘it is the Copula — it contains the
possibility of every position, to which there exists any correspondence in reality.
It is itself, therefore, the realising principle, the spiritual substratum of the whole
complex body of truths.’128
   Statements such as this, as we have seen, are typical of Coleridge’s curious
willingness to continue to deploy transcendental arguments (concerning ‘the pos-
sibility of every position’) in areas where argumentation itself might appear to be
ceding ground to volition. Nonetheless, for much of his career, as Coleridge strug-
gled to reconcile the claims of philosophy with those of his religion, the balance of
his thinking shifted increasingly towards the noncognitive. ‘For a very long time
indeed I could not reconcile personality with infinity,’ he recounts in Biographia,
‘and my head was with Spinoza, though my whole heart remained with Paul and
John.’129 It was to be the power of personality that finally triumphed in this
contest, as the teachings of Paul and John overcame the limitless philosophical
planes of Spinoza. Will, rather than reason, becomes the keystone of Coleridge’s
thought as he endeavours to vindicate his belief that ‘[t]he Ground of Man’s nature
is the Will in a form of Reason.’130
   Indeed, for Coleridge, even the alterity that determines Noetics, or the science of
reason, is itself the product of Absolute Will. In an unpublished fragment dating
from around 1818-1819, Coleridge maintains of the Will that ‘being causative
of alterity it is a fortiori causative of itself[,] and conversely the being causative
of itself it must be causative of alterity [...]. Consequently the Will is neither
abstracted from intelligence nor can Intelligence be conceived of as not grounded
and involved in the Will [...].’131 The trichotomic logic upon which all logic depends
thus ultimately rests upon an alterity grounded in the Will. One consequence of
this is that Coleridge conceives of the most fundamental relations governing reality
as personal relationships rather than logical relations.
   This is most powerfully expressed in the familial model deployed to explain the
emergence of consciousness (both human and Absolute) in Magnum Opus. Here,
Coleridge distinguishes three relationships — mother/child, father/son, I/thou —
whose interdependence forms the condition of possibility for the communicative
Logos. Without these relationships, indeed, it is impossible to explain how differ-
ence emerges from identity. ‘The whole problem of existence,’ he argues, ‘is present
as a sum total in the mother: the mother exists as a One and indivisible some-
thing.’ Alterity, and with it language, is only made possible by the intervention of
the father, introducing the difference-in-unity expressed in the Logos: ‘The father
and the heavenly father, the form in the shape and the form affirmed for itself are
blended in one, and yet convey the earliest lesson of distinction and alterity. There
 128 Coleridge, Lay Sermons 18.
 129 Coleridge, Biographia 1 201.
 130 Coleridge, Shorter Works 2 1368.
 131 Coleridge, Shorter Works 1 779.
68                                         Tim Milnes
was another beside the mother, and the child beholds it and repeats [...].’ Finally,
having left the maternal knee, ‘the child now learns its own alterity’.132 Viewed
this way, personality is no less than the key to the relation between unity and
difference. Only when one thinks of the most fundamental relations as personal
relationships, Coleridge argues, can one understand how alterity itself is possible.
Only through the metaphysics of personality is it possible, by acknowledging that
the claims of otherness touch us at the deepest level of our being, to resist the lure
of hyperrationalism or Spinozism.
    This is not to say that our personal relationship with God is identical to our
relationships with other people. Being perfect, God’s personhood is prothetical,
it is ‘personeity, differing from personality only as rejecting all commixture of im-
perfection associated with the latter.’ This installation of God as ‘at once the
absolute person and the ground of all personality,’133 represents Coleridge’s most
mature attempt to bridge religion and philosophy by reconciling ‘personality with
infinity.’ By making the alterity outlined in his ‘higher’ polar logic or noetic de-
pendent upon the willing relationships that sustain the relations between persons,
Coleridge avoids collapsing these relations into an undifferentiated foundation that
could once again be made the exclusive property of philosophy. For the later Co-
leridge, personhood is prior to being, just as faith is prior to knowing.
    Thus, in the unpublished ‘Essay on Faith,’ written in 1820, Coleridge defines
‘Faith’ as ‘Fidelity to our own Being as far as such Being is not and cannot become
an object of the sense,’134 arguing that ‘even the very first step [...] the becoming
conscious of a Conscience, partakes of the nature of an Act [...] by which we take
upon ourselves [...] the obligation of Fealty.’135 As Steven Cole points out, in ‘The
Essay on Faith,’ ‘Coleridge offers his fullest, and most compelling, explanation
of how the idea of personhood is contextually enacted,’ based on how “‘fidelity
to our own being” establishes a relation our being has to the being of others.’136
But Coleridge goes further even than this. He intimates that the obligation of
acknowledgement is the most fundamental precondition of all recognition; whether
perceiving nature or other persons, we are bound to enter into a moral relationship
that presupposes an element of will, and thus faith. Consequently, ‘Conscience is
the root of all Consciousness, and a fortiori the precondition of all Experience’; it
‘is a witnessing respecting the unity of the Will and the Reason effected by the
Self-subordination of the Will, as = Self, to the Reason, as = the Will of God.’137
    In essence, as Anthony John Harding indicates, what Coleridge attempts in the
‘Essay on Faith’ and the Magnum Opus is to reverse Kant’s proof of the exis-
 132 Coleridge, Opus Maximum 131-132. As McFarland notes, this account of the Logos or ‘I
am’ was crucial in freeing Coleridge from Schelling’s dyadic, ‘either/or’ conception of Absolute
Identity (cxxxi).
 133 Coleridge, Opus Maximum 176-177.
 134 Coleridge, Shorter Works 2 834.
 135 Coleridge, Shorter Works 2 836.
 136 Steven E. Cole, The Logic of Personhood: Coleridge and the Social Production of Agency,’
Studies in Romanticism 30 (1991) 103-105.
 137 Coleridge, Shorter Works 2 837-838.
                                      Coleridge’s Logic                                   69
tence of free will from the moral law, basing the moral law on the existence of
free will. In this, he was influenced by recent German criticism of Kant’s sys-
tem as crypto-Spinozism, and in particular Jacobi’s argument in Concerning the
Doctrine of Spinoza in Letters to Herr Moses Mendelssohn (1785) that ‘[f]aith is
the element of all human cognition and activity.’138 It was Jacobi’s work that
encouraged Coleridge in his conviction that ‘[t]he conscience [...] is not a mere
mode of our consciousness, but presupposed therein.’139 And yet, Coleridge was
not prepared to swallow Jacobi’s antidote to philosophy, what the latter termed
in his 1815 Preface to David Hume on Faith as a ‘knowing not-knowing.’140 In-
stead, as Harding indicates, by establishing conscience as the means by which
‘the self becomes aware of its own existence,’ Coleridge does ‘what Kant did not
do for himself, that is, establish a priori the possibility of recognising other hu-
man beings as themselves possessed of conscience and selfhood.’141 Ultimately,
Jacobi’s insistence on the priority of personhood and faith in human knowledge
is transformed by Coleridge into a metaphysics of personality designed to prevent
the higher logic of Noetics from folding into pantheism. Against Kant’s stricture
that there could never be a theology of reason, Coleridge envisages religion and
philosophy in perfect equipoise, the logic of understanding blending with a logic
of reason that is itself part logic, part revelation. Such is the place of logic in
a theosophy according to which reason must always incorporate the illumination
of faith, as Coleridge’s Appendix to On the Constitution of the Church and State
makes clear:
        Finally, what is Reason? You have often asked me; and this is my
        answer;
        Whene’er the mist, that stands ’twixt God and thee
        Defecates to a pure transparency,
        That intercepts no light and adds no strain —
        There Reason is, and there begins her rein!142
                          10    COLERIDGE’S LOGIC TODAY
The story of Coleridge’s influence on modern logic is not one likely to detain the
historian of ideas for very long. This is almost entirely due to the fact that the
Logic manuscripts lay almost unnoticed for much of the nineteenth century, and
were first published, in excerpted form, in Alice Snyder’s 1929 Coleridge on Logic
 138 Friedrich Heinrich Jacobi, The Main Philosophical Writings and the Novel Allwill, trans.
George di Giovanni (McGill-Queen’s University Press, 1994) 234.
 139 Coleridge, Opus Maximum 73.
 140 Jacobi 545.
 141 Anthony John Harding, Coleridge and the Idea of Love: Aspects of Relationship in Co-
leridge’s Thought and Writing (Cambridge University Press, 1974) 189-91.
 142 Coleridge, Church and State 184.
70                                          Tim Milnes
and Learning.143 Even then, a further 62 years elapsed before J.R. de J. Jackson
produced a text for the Bollingen Collected Coleridge that accurately reflected the
content of the Logic manuscripts. Rather like Bentham’s contemporary work on
logic and language, Coleridge’s logic suffered from the chaotic state of the writer’s
corpus at the time of his death. While in the case of Bentham this was largely due
to carelessness about publication, with Coleridge the causes were excessive caution
and procrastination. This, combined with the fact that Coleridge tends to avoid
the technicalities of logic in his published works — typically deferring full expo-
sition with an apology and a promissory note for the ‘Elements of Discourse’ —
means that little of Coleridge’s logical theory was available in the decades follow-
ing his death. His influence in the spheres of ethics and the theory of government
(famously impressing John Stuart Mill), theology (encouraging Newman and the
Oxford movement with his defence of Trinitarianism), philosophy (transmitting his
ideas via Carlyle to Emerson and the American Transcendentalists), and aesthet-
ics (single-handedly inventing the concept of practical criticism that would later
be developed by LA. Richards) is immense and well documented.144 In logic, how-
ever, the dissemination of Coleridge’s thought, at least until recently, has largely
been limited to footnotes and the inferences of his more attentive readers.
    That said, since the end of the nineteenth century a succession of critics has
deplored the neglect of Coleridge’s logic. Among these commentators there is
near unanimity in the view that Coleridge’s single greatest achievements in this
field stem from his exploration of the interconnectedness of logic, language, and
the noncognitive matrices of faith and personhood. Snyder set the tone by ar-
guing that, more than his efforts ‘to schematise the dialectic movement of the
reason,’ Coleridge’s ‘negative criticisms of the lower faculty, the understanding
[...] threw out informal suggestions for which thinkers today are still expressing
their gratitude.’145 Less patronisingly, Muirhead embraced Coleridge’s rejection
of conventional logic, praising ‘a method which proceeds, as he expresses it, by
“enlargement” instead of by “exclusion,” and by inner development instead of by
mere external synthesis,’146 as well as Coleridge’s anticipation of the principle — of
which later idealists such as Bradley, Green and Bosanquet made so much — that
 143 Alice D. Snyder, Coleridge on Logic and Learning (New Haven: Yale University Press, 1929).
 144 Coleridge’s impact on these areas is too vast, and the literature devoted to the subject too
extensive to document here. However, for a comprehensive study of Coleridge’s influence on Mill
and the thought of nineteenth-century Britain, see Ben Knights, The Idea of the Clerisy in the
Nineteenth Century (Cambridge University Press, 1978). For his influence on Cardinal Newman,
see Philip C. Rule, ‘Coleridge and Newman: The Centrality of Conscience,’ The Fountain Light:
Studies in Romanticism and Religion in Honor of John L. Mahoney, ed. Robert J. Barth (NY:
Florida University Press, 2002). Basil Willey’s ‘I.A. Richards and Coleridge,’ in LA. Richards:
Essays in his Honor, eds. Rueben Brower, et al. (Oxford University Press, 1973) assesses the role
Coleridge played in Richards’ conception of the function of criticism, while the poet’s relation to
American transcendentalism is examined by Kenneth Marc Harris in ‘Reason and Understanding
Reconsidered: Coleridge, Carlyle and Emerson,’ Essays in Literature 13.2 (1986). The wider
impact of Coleridge’s thought is mapped in Coleridge’s Afterlives, 1834-1934, eds. Jane Wright
and James Vigus (Palgrave, forthcoming).
 145 Snyder 13.
 146 Muirhead 86-87.
                                        Coleridge’s Logic                                     71
‘what is necessary and at the same time possible must be real.’147 More recently,
as well as McKusick’s championing of Logic as a landmark of nineteenth-century
linguistics, Nicholas Reid has published a series of articles in which he argues
that by indexing meaning to an imaging process rather than to fictional entities
called images, Coleridge evades the Fregean-Wittgensteinian attack on psychol-
ogism, presenting a view of logic and language, which, ‘admittedly shorn of its
idealist and theistic roots, is ripe for revival.’148
   Apologias such as the above notwithstanding, it is difficult to see how, for the
foreseeable future at least, Coleridge’s position in the history of logic, sandwiched
between Kant and Mill, can avoid appearing to many as an embarrassing example
of romantic overreaching. This is partly because of the way in which the status
of logic as a ‘discipline’ is contested in his writings, but mainly because of how
he challenges philosophy to justify itself at the bar of ‘life.’ Yet it is for pre-
cisely this reason, I would argue, that we should applaud Coleridge and continue
to read him. Driven by an urgent desire to define a new index of rationality for
a post-revolutionary age, the sheer range and audacity of Coleridge’s theosophi-
cal resuscitation of logic in an era dominated by the spectre of irrationalism can
reacquaint a modern reader with the historical conditions of what Habermas calls
the plurality of the voices of reason.149 In particular, his endeavours remind us
how, before Frege, logic briefly assumed an existential form, which, however out-
landish it might appear today, rightly refused the Humean severance of thought
that analytic philosophy was later to inflict on human intellect.
   Admittedly, one result of this is that navigating what McFarland aptly describes
as ‘the reticulation of Coleridge’s thought’ is never without its difficulties.150 In
particular, as his later work strives to adjudicate between competing conceptions
of reason (instrumental, dialogic, intuitive, incarnate, practical), Coleridge’s in-
creasingly elaborate attempts to square the circle between faith, logic and com-
munication acquire a dogmatic character. Suspicious of foundationalism, Coleridge
originally grounds thought and being in a linguistic act. Yet, rather than leave rea-
son to the pragmatics of communication, to the everyday function of ‘etymologic,’
he hypostasizes this act in the form of a metaphysical principle. Instead of dis-
pensing with philosophical foundations altogether, Coleridge ultimately installs a
 147 Muirhead   110.
 148 Nicholas  Reid, ‘Coleridge, Language, and the Imagination,’ Romanticism on the Net,22,
May 2001, 31 March 2006 www.erudit.0rg/revue/ron/2001/v/n22/005977ar.html. See also
Nicholas Reid, ‘Form in Coleridge, and in Perception and Art More Generally,’ Romanticism on
the Net 26, May 2002, 31 March 2006 www.erudit.0rg/revue/ron/2002/v/n26/005699ar.html,
and ‘ “That Eternal Language,” or Why Coleridge was Right about Imaging and Meaning,’
Romanticism on the Net 28, November 2002, 31 March 2006 www.erudit.org/revue/ron/2002/
v/n28/007208ar.html.
 149 See J ̈ rgen Habermas, ‘The Unity of Reason in the Diversity of its Voices,’ Postmetaphysical
          u
Thinking: Philosophical Essays, trans. William Mark Hohengarten (Cambridge: Polity Press,
1992): ‘[T]he unity of reason only remains perceptible in the plurality of its voices — as the
possibility in principle of passing from one language to another — a passage that, no matter how
occasional, is still comprehensible’ (117).
 150 McFarland, Pantheist Tradition xl.
72                                           Tim Milnes
super-foundationalism. Thus, the ‘copula’ that the verb substantive, the Prothesis
or ‘I am’ presents in thought becomes a ground of grounds, a logical, grammatical,
psychological and metaphysical principle underwriting logic and theory of method
alike. But is this principle itself logical? The closing words of Barfield’s study will
serve just as well here, in that they elegantly capture the question that Coleridge
bequeaths to modern logic. What finally matters when considering Coleridge’s
logic, Barfield concludes, is ‘whether there is indeed a sense in which it is proper
to characterise as a nuclear — or polar — logic the exactness of the understanding,
not blurred or cancelled, but pierced to its empty heart at each moment, by the
energy of imagination as the bearer of related qualities.’151
                                       BIBLIOGRAPHY
 [Barfield, 1971] O. Barfield. What Coleridge Thought. Wesleyan University Press, 1971.
 [Burnett, 1773–92] J. Burnett [Lord Monboddo]. On the Origin and Progress of Language. 6
    vols. Edinburgh, 1773-92.
 [Christensen, 1981] J. Christensen. Coleridge’s Blessed Machine of Language. Cornell Univer-
    sity Press, 1981.
 [Cole, 1991] S. E. Cole. The Logic of Personhood: Coleridge and the Social Production of
    Agency. Studies in Romanticism 30 (1991): 85-111.
 [Coleridge, 1995] S. T. Coleridge. Aids to Reflection. Ed. John Beer. Princeton University Press,
    1995.
 [Coleridge, 1983] S. T. Coleridge. Biographia Literaria. Eds. James Engell and Walter Jackson
    Bate. 2 vols. Princeton University Press, 1983.
 [Coleridge, 1956–71] S. T. Coleridge. Collected Letters of Samuel Taylor Coleridge. Ed. Earl
    Leslie Griggs. 5 vols. Oxford: Clarendon Press, 1956-71.
 [Coleridge, 1969] S. T. Coleridge. The Friend. Ed. Barbara E. Rooke. 2 vols. Princeton Univer-
    sity Press, 1969.
 [Coleridge, 1972] S. T. Coleridge. Lay Sermons. Ed. R.J. White. Princeton University Press,
    1972.
 [Coleridge, 1987] S. T. Coleridge. Lectures 1808-1819 On Literature. Ed. R.A. Foakes. 2 vols.
    Princeton University Press, 1987.
 [Coleridge, 2000] S. T. Coleridge. Lectures 1818-1819 On the History of Philosophy. Ed. J.R.
    de J. Jackson. 2 vols. Princeton University Press, 2000.
 [Coleridge, 1981] S. T. Coleridge. Logic. Ed. J.R. de J. Jackson. Princeton University Press,
    1981.
 [Coleridge, 1980–2001] S. T. Coleridge. Marginalia. Eds. H. J. Jackson, George Whalley, et al.
    6 vols. Princeton University Press, 1980-2001.
 [Coleridge, 2001] S. T. Coleridge. The Notebooks of Samuel Taylor Coleridge. Ed. Kathleen
    Coburn, et al. 5 vols. Routledge, 2002.
 [Coleridge, 1976] S. T. Coleridge. On the Constitution of the Church and State. Ed. John
    Colmer. Princeton University Press, 1976.
 [Coleridge, 2002] S. T. Coleridge. Opus Maximum. Eds. Thomas McFarland and Nicholas
    Halmi. Princeton University Press, 2002.
 [Coleridge, 1995] S. T. Coleridge. Shorter Works and Fragments. Eds. H. J. Jackson and J. R.
    de J. Jackson. 2 vols. Princeton University Press, 1995.
 [Coleridge, 1990] S. T. Coleridge. Table Talk. Ed. Carl Woodring, 2 vols. (Princeton University
    Press, 1990)
 [De Quincey, 1889] T. De Quincey. The Collected Writings of Thomas de Quincey. Rev. ed.
    David Masson. 14 vols. Edinburgh, 1889.
 [Fulford, 1991] T. Fulford. Coleridge’s Figurative Language. MacMillan, 1991.
  151 Barfield 193.
                                      Coleridge’s Logic                                      73
[Habermas, 1992] J. Habermas. Postmetaphysical Thinking: Philosophical Essays. Trans.
   William Mark Hohengarten. Polity Press, 1992.
[Hamilton, 1983] P. Hamilton. Coleridge’s Poetics. Basil Blackwell, 1983.
[Harding, 1974] A. J. Harding. Coleridge and the Idea of Love: Aspects of Relationship in
   Coleridge’s Thought and Writing. Cambridge University Press, 1974.
[Harris, 1986] K. M. Harris. Reason and Understanding Reconsidered: Coleridge, Carlyle and
   Emerson. Essays in Literature 13.2 (1986): 263-282
[Hartley, 1791] D. Hartley. Observations on Man. 2 vols. 1791, Woodstock Books, 1998.
[Havens, 1981] M. K. Havens. Coleridge on the Evolution of Language. Studies in Romanticism
   20 (1981): 163-183.
[Hagel, 1977] G. W. F. Hegel. Phenomenology of Spirit. Trans. A.V. Millar. Oxford University
   Press, 1977.
[Jackson, 1998] H. J. Jackson. Coleridge’s Lessons in Transition: The “Logic” of the “Wildest
   Odes”. Lessons of Romanticism: A Critical Companion. Eds. Thomas Pfau and Robert F.
   Gleckner. Duke University Press, 1998. 213-224.
[Jackson, 1983] H. J. Jackson. Coleridge, Etymology and Etymologic. Journal of the History of
   Ideas 44.1 (1983): 75-88.
[Jacobi, 1994] F. H. Jacobi. The Main Philosophical Writings and the Novel Allwill. Trans.
   George di Giovanni. McGill-Queen’s University Press, 1994.
[Knights, 1978] B. Knights. The Idea of the Clerisy in the Nineteenth Century. Cambridge
   University Press, 1978.
[Leask, 1988] N. Leask. The Politics of Imagination in Coleridge’s Critical Thought. Macmillan,
   1988.
[McFarland, 1993] T. McFarland. Aspects of Coleridge’s Distinction Between Reason and Un-
   derstanding. Coleridge’s Visionary Languages: Essays in Honour of J.B. Beer. Eds. Tim
   Fulford and Morton D. Paley D.S. Brewer, 1993. 165-180.
[McFarland, 1969] T. McFarland. Coleridge and the Pantheist Tradition. Oxford University
   Press, 1969.
[McKusick, 1987] J. C. McKusick. Coleridge’s “Logic”: A Systematic Theory of Language.
   Papers in the History of Linguistics. Eds. Hans Aarsleff, Louis G. Kelly and Hans-Joseph
   Niederehe. John Benjamins Publishing Company, 1987. 478-489.
[McKusick, 1986] J. C. McKusick. Coleridge’s Philosophy of Language. Yale University Press,
   1986.
[McNiece, 1992] G. McNiece. The Knowledge that Endures: Coleridge, German Philosophy and
   the Logic of Romantic Thought. MacMillan, 1992.
[Milnes, 1999] T. Milnes. Through the Looking-Glass: Coleridge and Post-Kantian Philosophy.
   Comparative Literature 51.4 (1999): 309-323.
[Muirhead, 1930] J. H. Muirhead. Coleridge as Philosopher. George Allen & Unwin Ltd., 1930.
[Orsini, 1969] G. N. G. Orsini. Coleridge and German Idealism. Southern Illinois University
   Press, 1969.
[Perkins, 1994] M. A. Perkins. Coleridge’s Philosophy: The Logos as Unifying Principle. Claren-
   don Press, 1994.
[Prickett, 1986] S. Prickett. Words and The Word. Cambridge University Press, 1986.
[Quine, 1980] W. V. O. Quine. Elementary Logic. Rev. ed. Harvard University Press, 1980.
[Reid, 2001] N. Reid. Coleridge, Language, and the Imagination. Romanticism on the Net 22.
   May 2001. 31 March 2006. www.erudit.0rg/revue/ron/2001/v/n22/005977ar.html
[Reid, 2002] N. Reid. Form in Coleridge, and in Perception and Art More Generally. Roman-
   ticism on the Net 26. May 2002. 31 March 2006. www.erudit.0rg/revue/ron/2002/v/n26/
   005699ar.html
[Reid, 2002a] N. Reid. That Eternal Language, or Why Coleridge was Right about Imaging
   and Meaning. Romanticism on the Net 28. November 2002. 31 March 2006. www.erudit.org/
   revue/ron/2002/v/n28/007208ar.html
[Rule, 2002] P. C. Rule. Coleridge and Newman: The Centrality of Conscience. The Fountain
   Light: Studies in Romanticism and Religion in Honor of John L. Mahoney. Ed. Robert J.
   Earth. Florida University Press, 2002. 231-255.
[Simpson, 1993] D. Simpson. Romanticism, Nationalism, and the Revolt Against Theory. Uni-
   versity of Chicago Press, 1993.
[Snyder, 1929] A. D. Snyder. Coleridge on Logic and Learning. Yale University Press, 1929.
74                                          Tim Milnes
[Tooke, 1829] J. H. Tooke. Epea Pteroenta or the Diversions of Purley. Rev. ed. by Richard
   Taylor. 2 vols. London, 1829.
[Wallace, 1983] C. M. Wallace. The Design of the Biographia Literaria. George Allen & Unwin,
   1983.
[Wellek, 1931] R. Wellek. Immanuel Kant in England 1793-1838. Princeton University Press,
   1931.
[Wheeler, 1986] K. Wheeler. Coleridge’s Theory of Imagination: a Hegelian Solution to Kant?
   The Interpretation of Belief: Coleridge, Schleiermacher and Romanticism. Ed. David Jasper.
   London, 1986. [16-40]
[Willey, 1973] B. Willey. LA. Richards and Coleridge. LA. Richards: Essays in his Honor. Eds.
   Rueben Brower, et al. Oxford University Press, 1973. 227-236.
[Wright and Vigus, forthcoming] J. Wright and J. Vigus, eds. Coleridge’s Afterlives, 1834-1934.
   Palgrave, forthcoming.
          RICHARD WHATELY AND LOGICAL
                                       THEORY
                                      James Van Evra
                                    1  INTRODUCTION
Richard Whately (1787–1863) is better known for his influential defense of logic
than for adding anything to the formal core of the subject. That defense, set out
in his Elements of Logic [EL] of 1826,1 appeared at the beginning of a period
of significant change in the subject. The popular logic2 that Whately inherited
had been the subject of unremitting criticism at the hands of major thinkers for
more than two centuries. As a result, the received logic, which consisted largely
of texts that were formally defective and contained little evidence of theoretical
insight, was in disarray. Soon after the publication of Elements, by contrast,
the first wave of great 19th century works in logic appeared, works by George
Boole and Augustus De Morgan in Britain, Bernard Bolzano on the continent,
and Charles Sanders Peirce in the United States. Many factors contributed to this
reversal of fortune, including, perhaps most importantly, the growing recognition
of interesting ties between logic and mathematics. Although Whately’s defense
of the subject preceded the mid century shift, it was nonetheless more than a
defense of the existing tradition. Rather, Whately’s contribution to the process
was a reconception of the broad theoretical context within which logic occurs. In
effect, his theory cast logic in a form more congenial to the formal developments
soon to come. In addition to its impact on the core conception of the subject, his
reconception had the further effect of clarifying the boundary conditions on logic,
something sorely needed in a period in which logic had often been uncritically run
together with epistemology or rhetoric. Settling logic’s boundary conditions, in
turn, served to better define the scope and limits of the neighboring disciplines (in
particular rhetoric, the topic of another of Whately’s works, Elements of Rhetoric
[1828]).
    1 The date of the first monograph edition; the work first appeared in 1823 in several parts in
a popular cyclopedia. See Jongsma 1983 for a full account of its provenance.
    2 While they contain no innovations, popular texts are nonetheless a good indicator of the
state of logic at that time. E. J. Ashworth puts the point well: “. . . the textbook-writers and
schoolteachers of a period may be as important as the leading intellectuals, for it is by these
minor figures that all innovations are accepted, altered, and made into the new commonplace.
To concentrate solely on the great thinkers is to obscure the reality of the university and school,
of the mainstream orthodoxy which lies behind these thinkers and which feeds them.” [Ashworth,
1985: LIV]
Handbook of the History of Logic. Volume 4: British Logic in the Nineteenth Century.
Dov M. Gabbay and John Woods (Editors)
 c 2008 Elsevier BV. All rights reserved.
76                                         James Van Evra
                                        2 BIOGRAPHY
2.1 Life and Works
Richard Whately was born in London on February 1, 1787, to an established family
whose members held positions of responsibility in government service, banking, law
and the clergy. Throughout his life, he is described as being a brilliant, independent
thinker who bore little routine allegiance to established tradition. The youngest
(by six years) of nine children, his independence was at first a matter of necessity;
often isolated and left to his own interests, he learned to read and write at an
early age, studied nature on solitary walks, and displayed prodigious ability at
arithmetic calculation. Later in life his independence took the form of advocacy
of less than uniformly popular causes, such as ending legal disadvantages borne
by English Catholics, and, while serving as president of a royal commission on the
Irish poor (1835/6), arguing for making improvements to agriculture rather than
forcing the impoverished into workhouses.
    After early schooling near Bristol, Whately entered Oriel College in 1805. His
life in Oxford was a series of successes, beginning with a double second BA (in
Classics and Mathematics) in 1808. The degree was one of the first double degrees
awarded there, and one of its components, mathematics, was just beginning to
gain acceptance in Oxford as an independent field of study. He was elected fellow
of Oriel in 1811, awarded an MA in 1812, and became a private tutor. In 1822,
he left Oxford to become a parish priest, but returned in 1825 as principal of St.
Alban Hall. He became professor of Political Economy in 1829, and in 1831 left
Oxford permanently to become Archbishop of Dublin, a post he occupied until his
death on October 8, 1863.3
    Throughout his working life, Whately was known more for his talent as a teacher
(“no don,” it was said at the time, “was ever less donnish”)4 than for his pub-
lished scholarship. However, while not extensive in quantity, and while his two
primary works were texts, his writings were nonetheless influential; they are cited,
for instance, 237 times in the Oxford English Dictionary. Among them, three are
particularly noteworthy: the two Elements already mentioned, and his early His-
toric Doubts Relative to Napoleon Buonaparte [1819], an ironic critique of David
Hume’s argument on miracles based on the claim that the same argument could
be used to cast doubt on the existence of historical figures such as Napoleon.
    Logic played a relatively minor role in the full scope Whately’s working life; he
was primarily a theologian and philosopher. However, the fact that he was not a
logician is unremarkable, for at that time, no one was a logician.5 What interest he
had in logic had two origins, one theological, the other pedagogical. Regarding the
former, while at Oxford, Whately came under the influence of Edward Copleston,
    3A  full account of Whately’s life can be found in E. Jane Whately 1875.
    4 Cf. Brent 2004/5. Mary Prior [1967] likens him to an early 19th century P. T. Geach.
    5 None of the text writers in the tradition were primarily logicians. Much later, Peirce (cor-
rectly) claimed that he was the first since the middle ages to completely devote his life to logic.
Cf. [Fisch, 1985, xviii].
                               Richard Whately and Logical Theory                                77
the founder of a group (the Noetics) devoted to defending Christian doctrine on
the ground of its presumed reasonableness, and to producing clergy capable of
providing a reasoned defense of the teachings of the Church of England. Whately
considered logic to be an important part of the university curriculum insofar as it
strengthened rationality to such ends:
         The cause of Truth universally, and not least, of religious Truth, is ben-
         efited by everything that tends to promote sound reasoning and facili-
         tate the detection of fallacy. The adversaries of our Faith would, I am
         convinced, have been on many occasions more satisfactorily answered
         . . . had a thorough acquaintance with Logic been a more common
         qualification than it is. [1826, xxvii]6
    The other source of his interest in logic stems from changes then taking place in
Oxford. At the time, logic was being forced to prominence as a result of curricular
reform, one effect of which was that more students were required to study it. At the
same time, the university came under attack from external critics for promoting
the teaching of what they took to be a worthless subject. With logic so much a
matter of contention, there was obvious need for a clear account of the nature of
the subject and a rationale for studying it. Whately provided both.7
2.2       Reception
Elements was one of the most influential logic texts of the 19th century. When
referring to the background against which his own work was set, for instance,
George Boole immediately recognized Elements as the standard text in logic.8 In
reaction to Elements, De Morgan called Whately the “restorer of logical study in
England”,9 and Peirce’s interest in the subject was initially spurred by a reading of
the work at age 11.10 Beyond such direct influence, Elements was broadly popular,
appearing in nine editions during Whately’s lifetime, and in many reprint editions
    6 This was a common sentiment at the time. Others (e.g. Wesley quoted below) held the same
view. Unlike them, however, Whately sharply distinguished the science itself from its presumed
benefits.
    7 A more complete account of the reform process and reactions to it can be found in my [1984,
sec. 2.3].
    8 In Laws of Thought, Boole says “that portion of this work which relates to logic presupposes
in its reader a knowledge of the most important terms of the science, as usually treated, and of
its general object. On these points, there is no better guide than Archbishop Whately’s Elements
of logic [Boole, 1854, pref.].
    9 [De Morgan, 1860, 247]
   10 Peirce’s first encounter with logic, as Max Fisch [1982, xviii] describes it, occurred when
Charles found a copy of Elements in his brother’s room, and asking what logic was, got a simple
answer, stretched himself on the carpet with the book open before him, and over a period of
several days absorbed its contents. Since that time, he often said in later life, it had never been
possible for him to think of anything, including even chemistry, except as an exercise in logic.
And so far as he knew, he was the only man since the middle ages who had completely devoted
his life to logic.
78                                       James Van Evra
throughout the nineteenth and early twentieth centuries.11 It was so popular
in the United States that, as Whately (correctly) observed, “. . . it is in use, I
believe, in every one of their Colleges” [9th ed., xviii] The reason for its popularity
is clear: Elements contains a well reasoned and engaging defense of logic generally,
an accurate presentation of what was then its formal core, and a coherent theory
of the subject.
                     3 WHATELY’S CONCEPTION OF LOGIC
3.1 Logic as Whately Found it
Whately wrote Elements in part as a critique of a long standing tradition in
common logic on the one hand, and in reaction to the pointed criticisms of logic
which arose in reaction to that tradition on the other. His point throughout was
that logic had been misrepresented by the authors of the texts and misunderstood
by the critics.
   The tradition in question began as a revival of Aristotelianism in the late six-
teenth century. By that time, the sophistication of Medieval logic had been lost
to critics (e.g. the “Humanists”) who deplored the imposition of what they con-
sidered barbarous technicalities on the language, and yearned instead for a return
to the elegant Latin of Cicero and Seneca. By the mid-sixteenth century, popular
works such as the Dialectic of Peter Ramus [1543] appeared which played to liter-
ary fashion by freely drawing illustrations from favored Latin authors. Rather than
conceiving of logic as having to do with semantic problems, forms of argument, or
paradoxes as it had been during the medieval period, logic was now described as
“an art which teaches one to dispute well”. As a result, such
       Lack of interest in, and open hostility to, the older logical methods
       were widely accompanied by an increased concern with rhetoric, so
       that logic, for long the ‘art of arts’, was now required to be patterned
       on logically untutored thought and speech in a way as free as possible
       from subtle technicalities and rigour. [Thomas, 1964, 300]
   Later works in the tradition identified the syllogism as the central formal feature
in logic, but often gave faulty accounts of such basic topics as rules for identifying
middle terms, and for distinguishing major and minor premisses. Clearly, the
atmosphere of continuing hostility to the older logic had taken its toll:
  11 So popular was Elements that it was featured in an 1867 poem by W. S, Gilbert: (“Sir
Macklin”. Cf [Gilbert, 1968, 96]):
       Then I shall demonstrate to you,
       According to the rules of Whately
       That what is true of all, is true
       Of each, considered separately.
                              Richard Whately and Logical Theory                              79
       Having discarded the whole corpus of late medieval novelties and dis-
       coveries, it is clear that [the authors] left themselves little room for
       manoeuvre. In strictly formal logic, about all that they set themselves
       to do was to give an account of the syllogistic. Their normal ill success
       is a measure of the general loss of logical nerve at this time. [Ibid., 302]
   Prominent works in the tradition include those by Robert Sanderson [1615],
Richard Crakanthorpe [1622], The mathematician (and teacher of John Locke)
John Wallis [1687], and Henry Aldrich [1691]. Of these, versions of the first and
especially the last were still in prominent use in Britain in the early 19th century.
Indeed, Whately acknowledged Aldrich’s (“concise, but generally accurate”) Com-
pendium as the source of the formal core of Elements. When attention turned to
theory, however, any allegiance Whately felt to his predecessor vanished.
   To give some indication of what texts of the period were like, what follows is a
brief characterization of those by Sanderson and Aldrich. Both texts begin with
essentially the same description of logic. Here, in translation, is Aldrich’s version:
       Logic, which by synecdoche (or the figure which takes the part for the
       whole) is denominated Dialectics, has been called the art of reasoning;
       or, an instrumental art, directing the mind into the knowledge of all
       intellectual things. For this reason it ought to be the first of all dis-
       ciplines, as being necessary to the acquirement of the rest. [Jackson,
       1836, 1n]12
   As we will see, such an optimistic estimate of logic’s role is an example of what
for critics was a sure sign of its abject failure.
   Sanderson’s Logicae Artis Compendium [1615] was popular throughout its his-
tory and still in use in the late 18th and early 19th centuries.13 Little of it is
devoted to what would now be considered logic. An account of the figures and
rules of the syllogism, for instance, occupies just three (of 357) pages. Further,
the syllogism is interpreted, following Aristotle, as a kind of discourse, but not
one based on a distinct conception of logical form (there is, for instance, no formal
representation of the syllogism in the text). Also, Sanderson recognizes only three
syllogistic figures on the Philoponian arrangement of syllogistic terms. The re-
mainder of the work is devoted to topics which, while related to logic (e.g. speech,
or the mind), are no longer considered proper parts of the subject. Sanderson’s
conception of logic is broad enough, in fact, to include everything to which logic
might apply, including all of (material) existence (and non-existence!) The syllo-
gism appears only derivatively, i.e. as a sub-class of second intentions (used to
group things taken materially. (The other sub class being demonstration.))
  12 Sanderson, more than a century earlier, put it as follows: “Logica, quae & Synecdichice Di-
alectica, est ars instrumentalis, dirigens mentam nostrum in cognitionem omnium intelligibilum”.
(1) Obviously, little had changed in the intervening period.
  13 Cf. [Ashworth, 1974]. Jeremy Bentham studied Sanderson as an undergraduate, and used it
in designing a new curriculum for logic in the schools. See also [Bentham, 1827, 13, 22].
80                                         James Van Evra
    Befitting its status as a compendium, is Aldrich’s Artis Logicae Compendium
[1691] is spare (in most versions, approx. 50 pages in length), and contains little
more than the basic components of an Aristotelian text, i.e. a classification of
terms and propositions, an account of syllogistic structure, a few pages on method,
and an appendix on fallacies. While slight, Aldrich’s text contains evidence that
by the latter stages of the 17th century, at least some formal progress had been
made:
        [Aldrich] defines the terms in the Philoponian way and works out the
        consequences accurately and completely, arriving at twenty-four moods
        in four figures, tables them in full, and effects a deductive rejection of
        the 232 invalid moods. To come on these few pages after the logical
        rags and tatters of the previous two centuries is to be presented with
        a creation of haute couture. [Thomas, 1964, 310–311]14
    Though not part of the same tradition, other 17th century works displayed much
the same character. Arnauld and Nicole’s Port Royal logic (La Logique, ou l’art
de Penser (1662)) was, like Aldrich compendium, formally accurate in its (brief)
presentation of the syllogism, but was otherwise in the thrall of Cartesian epis-
temology, and promoted Descartes’ critical attitude toward logic as traditionally
conceived.15
    The same extravagant claims made for logic in the 17th century can be found in
the 18th century as well. In his “Address to the Clergy”, for instance, John Wesley
(a translator of Aldrich in addition to being a noted theologian) characterized logic
in this way:
        Some knowledge of the sciences also, is, to say the least . . . expedi-
        ent. Nay, may we not say, that the knowledge of one, (whether art
        or science), although now quite unfashionable, is even necessary next,
        and in order to, the knowledge of the Scripture itself? I mean logic.
        For what is this, if rightly understood, but the art of good sense? of
        apprehending things clearly, judging truly, and reasoning conclusively?
        What is it, viewed in another light, but the art of learning and teach-
        ing; whether by convincing or persuading? What is there, then, in the
        whole compass of science, to be desired in comparison of it? [Wesley’s,
        Works, 1756. Jackson ed. 1872, Vol. 10. 483]
3.2 The Critics
As Whately received it, logic was caught in the rational equivalent of a perfect
storm, i.e. a set of circumstances that converged to make progress in the sub-
ject highly unlikely. During the same period in which logic was being described
  14 Whately incorporated this feature of the Compendium, with some modifications, in Elements.
It is one of the few features of the work that he used without extensive alteration.
  15 See [Descartes, quote below, 82].
                            Richard Whately and Logical Theory                               81
as playing a role in gaining knowledge, (while at the same time being based on
modes of thought fixed by tradition), major thinkers in the great outwash from the
Renaissance were busy promoting their own new methods for gaining knowledge
(including the fundamentals of the scientific method as we now know it) and mak-
ing major additions to the theory of knowledge itself. To them, not only did the
logic of terms, propositions and syllogisms play no role in uncovering new truths,
it was an example of what they were intent on leaving behind, i.e. a device based
on discredited tradition that did nothing but merely reshuffle knowledge already
in hand. Here is Whately’s description of the situation:
        By representing logic as furnishing the sole instrument for the discovery
        of truth in all subjects, and as teaching the use of intellectual faculties
        in general [members of the preceding tradition] raised expectations
        which could not be realized, and which naturally led to a reaction.
        [1826, vii]
   The reaction is well known. Some critics focused on the general Aristotelian
context. Thus Francis Bacon:
        For as water ascends no higher than the level of the first spring, so
        knowledge derived from Aristotle will at most arise no higher again
        than the knowledge of Aristotle. And therefore, though a scholar must
        have faith in his master, yet a man well instructed must judge for
        himself. [Bacon, 1605, 20]
   And in the 18th century, Thomas Reid:
        [Aristotle’s]’s works carry too evident marks of pride, vanity and envy
        which have often sullied the character of the learned. He determines
        boldly things above all human knowledge.
        He delivers his decisions oracularly, and without any fear of mistake.
        Rather than confess his ignorance he hides under hard words and am-
        biguous expressions, of which his interpreters can make what they
        please. [Reid, 1843, 553]16
   Others focused on what they took to be the nature of logic. Thus Locke’s
famous remark that
        If the syllogism be taken as the only proper instrument of reason and
        means of knowledge, it will follow that before Aristotle, there was not
        one man that did or could know anything by reason, and that since the
        invention of the syllogism there is not one in ten thousand that doth.
        But God has not been so sparing to men to make them barely two-
        legged creatures and left it to Aristotle to make them rational. [Locke,
        1706, 264]
  16 In his review of Elements, John Stuart Mill [1828, 138] indicated that even before Whately,
such extreme views had already lost favor.
82                                       James Van Evra
   The main criticism directed at logic, however, and one repeated frequently, was
that it could not do what had been claimed for it. Again in Bacon’s words,
       The syllogism consists of propositions, propositions of words are the
       signs of notions. If, therefore, the notions (which form the basis of
       the whole) be confused and carelessly abstracted from things, there is
       no solidarity in the superstructure. Our only hope, then, is induction.
       [Bacon, 1620, 316]
   And in a similar vein, Descartes says that
       I noticed that as far as logic was concerned, its syllogism and most
       of its other methods serve rather to explain to another what one al-
       ready knows, or even, as in the case of Lully, to speak freely and with-
       out judgement of what one does not know, than to learn new things.
       [Descartes, 1637, 14] 17
   Comments such as these are only a small sampling of the widespread contempt
for logic. In a period of intense activity in epistemology, the logicians’ claims
seemed both antiquated and false. The point constantly made was that there
are better ways to gain knowledge than those offered in logic, methods such as
Baconian induction, or the new methods of Descartes or Locke.
   Such criticism gave rise, in turn, to a variety of works, identified as logics, that
were designed substantially to modify, or completely replace, syllogistic texts.
The common rationale was that if the syllogism could not lead to knowledge, then
syllogistic texts should be replaced by accounts of what could. Thus, for instance,
as Rolf George points out,
       Denis Diderot’s article on logic in the Encyclopedie, the most widely
       consulted work of the 18th century, has nothing whatever to say about
       logic. It is claimed here, simply, that reasoning is a natural ability,
       and that to conduct logical inquiries is like “setting oneself the task of
       dissecting the human leg in order to learn how to walk” (Encyclopedie,
       Logique; [George, 2002, 36]).
   Other works, including Arnauld’s [1662] work already mentioned, Issac Watts’
Logick, or the right use of reason [1725], and Elements of logick [1748], by William
Duncan, relegate the syllogism to a minor role, or discard it altogether, and replace
it with the fruits of 17th century Epistemology.
   While such alternatives to the syllogistic manuals were popular, they did not en-
tirely displace the 17th century texts. Support for the standard texts, for instance,
remained particularly strong in Oxford. That support, in turn, led inevitably to
renewed criticism, and the debate over logic that resulted was an important factor
in the genesis of Elements.
  17 ‘Lully’ is Ramon Lull (1232? –1315), best known as the author of an Ars Magna [1274].
                              Richard Whately and Logical Theory                            83
3.3      Whately’s alternative
Whately made two initial assumptions that set his view of logic apart from those
of his forbears. The first is that logic is a science on a par with other prominent
sciences (e.g. chemistry, physics or algebra). While logic had been considered to
be equally an art as well as a science, that is, Whately considered it to be primarily
a science, and an art only derivatively and in application (see [1826, 1, 127–28]).
In calling logic a science, Whately meant that it is a discipline based on clear
theoretical principles.18
    For Whately, it was the longstanding lack of a strong theoretical foundation,
by contrast, that helped to explain why logic had progressed so little in preceding
centuries, for, as he says, no science progresses unless it is founded on such prin-
ciples [1826, 2]. And, while centuries of stagnation suggested to its critics that
logic had become moribund, Whately remained optimistic about its prospects.
One had to recall, he suggested, that sciences such as physics and chemistry had
been dormant for long periods before receiving the theoretical support needed for
scientific progress [1826, 11].
    The same lack of ‘right principles’ also explained, according to Whately, why
logic had been the subject of so much criticism, for without a proper foundation,
it had been subjected to the claims and demands of common opinion, but not
science:
        The vanity . . . by which all men are prompted unduly to magnify their
        own pursuits, has led unphilosophical minds . . . to extend the bound-
        aries of their respective sciences, not by the patient development and
        just application of the principles of those sciences, but by wandering
        into irrelevant subjects. . . . none is more striking than the misappli-
        cation of logic, by those who treated it as the ‘art of rightly employing
        the rational faculties’, or who have intruded it into the province of
        natural philosophy, and regarded the syllogism as an engine for the
        investigation of nature. [1826, 6–7].
    The second assumption Whately made is that logic is immediately concerned
with language. Earlier logicians had held that language serves logic only as a
representative of thought, which in turn is the primary subject.19 Whately, on the
other hand, took language itself to be the primary subject:
        Logic is entirely conversant about language: a truth which most writers
        on the subject, if indeed they were fully aware of it themselves, have
  18 Sir William Hamilton criticized Whately for suggesting that the definition of logic as a
science introduced something new. He pointed out that it had often been so described in the
past [Hamilton, 1833, 131 ff.]. Hamilton was right; there had been considerable debate in the
16th and 17th centuries concerning whether logic was an art or a science. As then understood,
however, ‘science’ was understood in the broad Aristotelian conception of that term (i.e. one
that encompassed any classificatory scheme). Whately’s conception of science, by contrast, is of
the modern theoretical variety, and not the Aristotelian.
  19 See [Aldrich,, Jackson, 1836, 3]
84                                         James Van Evra
       certainly not taken care to impress on their readers. Aldrich’s definition
       of logic, for instance, does not give any hint of this. [1826, 56n]
    And,
       . . . since logic is wholly concerned with the use of language, it follows
       that a Syllogism (which is an argument stated in regular logical form)
       must be “an argument so expressed, that the conclusiveness of it is
       manifest from the mere force of the expression”, i.e. without consider-
       ing the meaning of the terms . . . [1826, 88]20
    While placing language in such a central role is reminiscent of earlier Hobbesian
nominalism, the view is also a precursor of the view that it is possible to deal with
a variety of subject areas as they are represented in language.
    Within the scope of these two broad assumptions, the heart of Whately’s view is
that logic is a ‘generalized and abstract statement of all demonstration whatever’
[1826, 34]. In terms of the sort Boole would later use, it is ‘a method of analyzing
that mental process which must invariably take place in all correct reasoning’
[1826, 11]. What makes it abstract is that, unlike those who treat the syllogism as
one kind of argument among many,21 Whately separates logic from its intended
field of application. Thus thinking of the syllogism as a particular sort of argument
is, Whately says (in obvious reply to Locke), a “mistake no less gross than if any
one should regard Grammar as a peculiar Language, and should contend against its
utility, on the ground that many speak correctly who never studied the principles
of grammar” [1826, 11, 13/14]. In his conception, by contrast, using logic as an
abstract analytical device ‘is like using chemical analysis to examine the elements
of which any compound body is composed’ [1826, 11/12].
    Perhaps the clearest example of Whately’s conception of the abstract nature of
logic can be seen in the analogy he draws between logic and mathematics:
       All numbers must be numbers of some things. But to introduce into the
       science any notice of the things respecting which calculations are made,
       would be evidently irrelevant, and would destroy its scientific charac-
       ter: we proceed therefore with arbitrary signs representing numbers in
       the abstract. So does logic pronounce on the validity of a regularly con-
       structed argument, equally well, though arbitrary symbols have been
       substituted for the terms: and, consequently, without any regard to
       the things signified by those terms. [1826, 13–14]
    And,
  20 The word ‘force’ in the passage is likely a misprint of the word ‘form.’ Cf. [1826, 37] for a
similar passage using the latter term.
  21 Whately cites as an instance the rhetorician George Campbell, who juxtaposed syllogistic
and moral reasoning.
                          Richard Whately and Logical Theory                      85
      In order to trace more distinctly the different steps in the abstracting
      process by which any particular argument may be brought into the
      most general form, we may first take the syllogism stated accurately
      . . . and then somewhat generalize the expression by substituting (as
      in algebra) arbitrary unmeaning symbols for the significant terms that
      we originally used. [1826, 35]
   Within this context, Whately considered the syllogism to be a purely formal
device for the expression of any argument, as opposed to being an argument of a
particular kind. In Whately’s words, an argument, when cast into the proper form
      is called a syllogism; which is . . . not a particular kind of argument,
      but only a peculiar form of expression, in which every argument may
      be stated. [1826, 24]
   Again, logic,
      which is, as it were, the grammar of reasoning, does not bring forward
      the regular syllogism as a distinct mode of argumentation, designed to
      be substituted for any other mode; but as the form to which all correct
      reasoning may be ultimately reduced; and which, consequently, serves
      the purpose . . . of a test to try the validity of any argument. [1826,
      11–12; See also 124n]
   Unlike earlier theories, that is, Whately held the syllogism to be an abstract
canonical form, into which (perhaps elliptically stated) arguments may be formu-
lated for validity testing. So regarded, the syllogism has more in common with
algebraic expressions containing variables, than with arguments as they appear in
ordinary language. Whately later strengthened this conception of the syllogism as
a normal form by insisting on a completely tenseless interpretation of the copula
(see [1826, 57]).
   Defining the syllogism as abstract in relation to arguments permitted Whately
to adopt a somewhat more liberal approach to the relation between the syllogism
and various sorts of common arguments than that found in the 17th century texts.
He allowed, for instance, (where the earlier logicians had not) the possibility that
other sorts of arguments might have their own rules:
      . . . rules have been devised for ascertaining the validity of [hypothet-
      ical syllogisms] at once, without bringing them into categorical form.
      [1826, 108]
   And, although he held that the categorical syllogism remains the ultimate
ground for the validity of arguments, no longer were various kinds of arguments
treated as de facto incomplete syllogisms.
   Whately’s formal development of the syllogism (in a section occupying less than
one fifth of the whole work) bears an external resemblance to earlier logics. On
86                                   James Van Evra
closer inspection, however, the effects of his theoretical stance are immediately
apparent. Thus he initially restricts the scope of the discussion to reasoning and
argumentation, and then only as they occur in language. In addition, while he
begins in the traditional manner with a classification of terms, his classification
differs fundamentally from any that could be regarded as Aristotelian. He counts,
for example, the traditional distinctions (e.g. between univocal and equivocal
terms, or terms of first and second intention) not as classifications of terms, but of
ways in which they are used. Whately’s classification is confined to just one major
division (singular and common terms), all others being relegated to derivative
status.
   One consequence of the abstract character of logic, according to Whately, was
that it has value even when it has no direct application [1826, ix, 20]. Judging
logic by its direct effects (as those who held that it was an art had done), he says,
reflects a confusion about the nature of theories. Thus finding fault with logic for
not making people think better ‘is as if one should object to the science of optics
for not giving sight to the blind’ [1826, 12]. Obviously, Whately’s association of
logic with pure science was based on the recognition of similarities between the
structural properties of theories in areas such as optics with those in logic.
   Whately’s identification of argument forms containing variables as the proper
focus for logic now seems to be nothing more than a mundane recognition of the
obvious. In the early 19th century, it was nothing of the kind. While there had
been earlier logicians whose work included abstract components, Whately’s theory
explicitly recognized this as a defining feature of logic.
3.4 Other aspects of Whately’s approach to logic
Beyond Whately’s general conception of logic as a formal science with similarities
to algebra, etc., there are two more specific features found in Elements that are
worthy of mention. The first concerns his identification of logical individuals as
things which are incapable of logical division, which replaces the traditional notion
of individuals as essentially simple things [1826, 68]. This interpretation is later
prominent in the works of both Boole and De Morgan). Second, he includes (for
the first time since the Port-Royal logic), the principle of conditionalization:
      A conditional proposition . . . may be considered as an assertion of
      the validity of a certain argument; since to assert that an argument
      is valid, is to assert that the conclusion necessarily results from the
      premisses, whether those premisses be true or not. [1826, 110]
3.5 Whately on method
Separate from the core of Elements, but appended to it, Whately includes a lengthy
section (50 pp.) entitles ‘Dissertation on the province of reasoning’, which concerns
such traditional topics as induction, the discovery of truth, and inference and proof.
                          Richard Whately and Logical Theory                      87
Unlike earlier treatments of the same topics, in which they are run together with
logic proper, Whately explicitly separates the two:
       Logic being concerned with the theory of reasoning, it is evidently
       necessary, in order to take a correct view of this science, that all mis-
       apprehensions should be removed relative to the occasions on which
       the Reasoning-process is employed — the purposes it has in view —
       and the limits within which it is confined. [1826, 205]
   In this way he separates logic from the substantive theory of reasoning (a point
which Mill later failed to appreciate), the former being concerned with the latter
as theory to model, i.e. in this case to typical ways in which reasoning is used
and justified. Logic itself remained for Whately the formal theory of the theory of
reasoning.
   Nowhere is this separation more evident than in Whately’s treatment of induc-
tion. Attempting to establish a relationship of precedence or dependence between
logic and induction (a continual arguing point in the 17th and 18th century), is,
says Whately, a mistake:
       Logic takes no cognizance of induction . . . as a distinct form of argu-
       ment. . . . The essence of an inductive argument (and so of the other
       kinds which are distinguished for it) consists not in the form of the Ar-
       gument, but in the relation which the subject-matter of the Premises
       bears to that of the Conclusion. [1826, 124]
   And, as Whately notes on several occasions, the mistake of treating the syllo-
gism as a kind of argument rather than as a common logical form has a common
origin with the mistake of treating induction as a form of argument.
             4 CRITICAL REACTION TO WHATELY’S THEORY
Reactions to Elements began to appear shortly after its first monograph publica-
tion. While there was general agreement that Whately had mounted an admirable
defense of logic, and that he was responsible for stimulating new interest in the
subject, specific reaction to his revision of the context within which logic is done
fell into two groups. Those still committed to the Aristotelian style of logic with
all its accoutrements interpreted Whately as being one of them due to his reten-
tion of the syllogism, but found his conception of logic as a pure formal science
too spare. Those who had moved beyond Aristotelian logic, by contrast, tacitly
adopted his formal style, but (again owing to his retention of the centrality of the
syllogism), regarded him as still a member of the older tradition.
   The first substantial review of Elements was written by Jeremy Bentham’s
nephew, George Bentham (1800–1884). His book, entitled Outline of a new system
of logic, with a critical examination of Dr. Whately’s “Elements of logic”, [1827],
is an account of his own ideas on logic (which include the first use of equations
88                                   James Van Evra
and symbolic quantifiers in the formalization of syllogistic premisses), combined
with a running commentary on Elements. To Bentham, Elements was simply the
‘last and most improved edition of the Aristotelian system’ [Bentham,, 1827, vii],
rather than something fundamentally new. While Bentham lauds Whately for
having done much to ‘divest the science of that useless jargon, of those unmeaning
puerilities, with which it had been loaded by the schoolmen’, [Bentham, 1827,
v] he nonetheless finds Whately’s conception of logic too confining. Reflecting
older views, Bentham advocated a return to the earlier view that logic should be
concerned with the acquisition of knowledge as well as its formal representation.
Logic, he says, is the ‘branch of art-and-science which has for its object the ad-
vantageous application of the human mind to the study of any other branch of
art-and-science’ [1827, 14]. As such, logic should be concerned with both deduc-
tion and induction. Being confined only to deductive reasoning, Bentham held,
Whately’s theory is too restricted to be of use.
   In the following year, the young John Stuart Mill’s review of Elements appeared
in the Westminster review. Like Bentham, Mill complimented Whately for having
written a ‘clear exposition of the principles of syllogistic logic, and vindicating it
against the contemptuous sarcasms of some modern metaphysicians.22 Then, prior
to expressing reservations similar to those of Bentham, Mill makes a comment that
would be repeated in the 20th century: ‘[Whately] has written rather excellently
concerning logic, than expounded in the best manner the science itself” [1828,
138]. Mill’s criticism of Whately, repeated later in his System of logic [1843], again
concerns the abstract nature of Whately’s conception of logic. In opposition to
Whately, for instance, Mill says that logic can remove ambiguity, for the ‘analysis,
to which it subjects any process of reasoning, affords the readiest and most certain
means by which a latent ambiguity in any one of the term employed, or the tacit
assumption of any false or doubtful propositions, can be detected’ [1828, 144].
   Mill’s main complaint, however, was that while Whately was on solid ground
when he dealt with terms and propositions as they are used in syllogisms, ‘Aris-
totelians did not stop here, nor confine within these narrow bounds the dominion
of their science’ [1828, 154]. Rather, he says, they included also instruction for
the right employment of words as an instrument for the investigation of truth. As
an alternative, Mill offers the prospect of the development of a logic of induction.
Once again, Whately was being judged by the standards of the older tradition.
   Even Sir William Hamilton, no friend of Oxford logic generally, grudgingly
admitted Whately’s salutary influence on the study of logic by saying that just
when logic seemed dead,
       . . . a new life was suddenly communicated to the expiring study, and
       hope at least allowed for its ultimate convalescence under a reformed
       system. [Mill, 1828, 199]
   After ascribing the revitalization to the publication of Elements, however, Hamil-
ton nonetheless dismisses Whately (and others) by saying that they relied exclu-
  22 [Mill, 1828, 137]
                            Richard Whately and Logical Theory                               89
sively on Aldrich, who in turn knew little of the Aristotelian heritage of the sub-
ject. The new works, he says, show ignorance of the Greek commentaries on the
Organon, of the scholastic, Ramistic, Cartesian or Wolfian approaches, or the Kan-
tian dialectic.23 In addition, Hamilton accused Whately of contradicting himself
by saying that logic was about the regularities which underlie reasoning on the
one hand, and that logic is about language on the other.24 Once again, Whately’s
theory was judged to be too narrow.
   One further 19th century commentator deserves mention. Robert Blakey, the
author of the first modern work entirely devoted to the history of logic, character-
ized Whately’s legacy in the following manner:
       A great change has been effected in Oxford of late years, and almost
       solely through the labours of Dr. Whately. Since the publication of
       his Elements, many excellent works have made their appearance from
       that venerable seat of learning. . . . [Blakey, 1851, 427]
   By the 20th century, the changes Whately advocated had become so routinely
assumed that he was no longer associated with them. When he is mentioned at
all, 25 it is as a syllogistic logician who defended the subject but added nothing
to it.26 Arthur Prior’s assessment, for instance, is that ‘in the early 19th century
the common logic was rescued from oblivion by Richard Whately but was not
enlarged by him.’27 Mary Prior is equally explicit: ‘Whately’s achievement is not
so much in logic as in moral metalogic; he explained what logicians should have
been doing.’ [M. Prior, 1967, 287]
   The conclusion that Whately is to be counted among the “old” logicians also
gains support from the view that he had little impact on subsequent events in
logic. Mary Prior, for instance, says that
       Between 1826, the year Whately’s Elements of Logic was published,
       and 1860, George Boole, De Morgan, and John Stuart Mill were writ-
       ing. It is therefore natural to expect to find adumbrations of their work
       in Whately, but in his systematic and formal treatment of logic there
       are remarkably few. [M. Prior, 1967, 287]
   I suggest, on the contrary, that there are adumbrations of the work of later
logicians in Whately, but they are in the context in which he locates the subject
rather than in his ‘formal treatment of logic’.
  23 [Hamilton,  1833, 127]
  24 See [Hamilton, 1833, 135].
  25 He is mentioned, for instance, neither in the Kneale’s Development of logic [1961], nor in
Bochenski’s History of formal logic [1961]. Also interesting is the fact that while Louis Liard
[1878], and C. I. Lewis [1918], mention Bentham’s commentary on Elements, neither mentions
Elements itself.
  26 A conclusion drawn also in two theses which deal with Whately. See [Jongsma, 1983], and
[Brody, 1967].
  27 [A. Prior, 1967, 541]
90                                        James Van Evra
           5   WHATELY’S ROLE IN THE DEVELOPMENT OF LOGIC
There are those (most noticeably Thomas Kuhn in the mid 20th century) who
think that change in a science is exclusively the result of transitions between its
major theories. Richard Whately, among others, stands as a counterexample to
such a view. He is one of those individuals who have an affect on a science without
making a direct addition to the theories at its core. In this sense, his contribution
is more like Francis Bacon’s in the 17th century, or Whewell’s or Mill’s in the 19th ,
than that of Newton or Frege. Influence of Whately’s sort on a science is often
all but invisible; By our standards, his conception of the broader theory of logic is
so much like our own that his role in its inception is easily forgotten. Indeed, the
extent of his contribution is obvious only in comparison with earlier conceptions
of the subject. When that comparison is made, however, Whately stands out as
having had prescient insight into what logic would eventually become.
                                      BIBLIOGRAPHY
 [Aldrich, 1836] H. Aldrich. 1691 Artis Logicae Compendium. See [Mansel, 1852] or [Jackson,
    1836].
 [Arnauld and Nicole, 1662] A. Arnauld and P. Nicole. La Logique ou l’Art de Penser, 1662. In
    J. Buroker (Tr. and Ed.), Logic or the art of thinking, Cambridge, 1996.
 [Ashworth, 1615] J. Ashworth. Introduction to Sanderson, 1615 [1985].
 [Bacon, 1605] F. Bacon. The Advancement of Learning, 1605. In Advancement of learning and
    novum organum, New York, 1899.
 [Bacon, 1620] F. Bacon. Novum Organum, 1620. In [1899], above.
 [Bentham, 1827] G. Bentham. Outline of a New System of Logic, 1827. Bristol: Thoemmes,
    1990.
 [Blakey, 1851] R. Blakey. Historical Sketch of Logic. Edinburgh, 1851.
 [Bochenski, 1961] I. Bochenski. A History of Formal Logic. Notre Dame, 1961.
 [Boole, 1847] G. Boole. The Mathematical Analysis of Logic . . . , 1847. Cambridge, 1952, 49–
    214.
 [Boole, 1854] G. Boole. An Investigation of the Laws of Thought, 1854. Cambridge and London:
    1954, New York.
 [Boole, 1952] G. Boole. Studies in Logic and Probability, R. Rhees (ed.), London, 1952.
 [Brent, 2004/5] R. Brent. Whately, Richard. Oxford dictionary of national biography. Oxford,
    2004/5.
 [Brody, 1967] B. Brody. The Rise of the Algebra of Logic. Diss., Princeton, 1967.
 [Copleston, 1810] E. Copleston. A Reply to the Calumnies of the Edinburgh Review against
    Oxford ldots. Oxford, 1810.
 [Crakanthorpe, 1622] R. Crakanthorpe. Logicae Libri Quinque, London, 1622.
 [De Morgan, 1847] A. De Morgan. Formal logic. London, 1847.
 [De Morgan, 1860] A. De Morgan. Logic. The English Cyclopedia. London, 1860. In P. Heath
    (ed.), On the Syllogism and Other Logical Writings, London, 1966.
 [Duncan, 1748] W. Duncan. The Elements of Logick. London, 1748.
 [Fisch, 1982] M. Fisch (gen. ed.). The Writings of Charles Sanders Peirce. Vol. 1, Bloomington,
    1982 –1985.
 [Garland, 1980] M. Garland. Cambridge before Darwin. Cambridge, 1980.
 [George, 2002] R. George. The rise of modern logic (with J. Van Evra). In D. Jacquette (ed.),
    A Companion to Philosophical Logic. Oxford, 2002.
 [Hamilton, 1833] W. Hamilton. Logic, in reference to the recent English treatises on that sci-
    ence. Edinburgh Review, 1833. In Discussions on Philosophy and Literature, pages 117—173,
    London, 1866.
 [Hiscock, 1960] W. Hiscock. Henry Aldrich, Oxford, 1960.
                             Richard Whately and Logical Theory                               91
[Howell, 1971] W. Howell. Eighteenth Century British Logic and Rhetoric. Princeton, 1971.
[Jackson, 1836] T. Jackson (ed.). A Compendium of Logic, Translated and Abridged from
   Aldrich. London, 1836.
[Jongsma, 1983] C. Jongsma. Richard Whately and the Revival of Syllogistic Logic in Great
   Britain in the Early Nineteenth Century. Dissertation, Toronto, 1983.
[Kenney, 1959] W. H. Kenney. John Locke and the Oxford Training in Logic and Metaphysics.
   Dissertation, St. Louis, 1959.
[Kirwan, 1807] R. Kirwan. Logick, or an Essay on the Elements, Principles and Different Modes
   of Reasoning. 2 vols., London, 1807.
[Kneale and Kneale, 1962] W. Kneale and M. Kneale. The Development of Logic. Oxford, 1962.
[Lewis, 1918] C. I. Lewis. A Survey of Symbolic Logic. Berkeley, 1918.
[Liard, 1878] L. Liard. Les Logicians Anglais Contemporains. Paris, 1878.
[Locke, 1706] J. Locke. An Essay Concerning Human Understanding, 1706. 5th ed.; reprinted
   1961, London.
[Macfarlane, 1916] A. Macfarlane. Lectures on British Mathematicians. New York, 1916.
[Machen, 1926] A. Machen. The Anatomy of Tobacco. New York, 1926.
[Mallet, 1927] C. Mallet. A History of the University of Oxford. Vol. 3, Modern Oxford, London,
   1927.
[Mansel, 1852] H. Mansel. Artis Logicae Rudimenta, from the text of Aldrich. 2nd ed., Oxford,
   1852.
[Mill, 1828] J. S. Mill. Review of Whately, “Elements of Logic”. Westminster Review : 137–172,
   1828.
[Playfair, 1808] J. Playfair. Review of La Place, “Traite de Mechanique celeste”. Edinburgh
   Review, 11: 149–284, 1808.
[Philodicaeus, 1829] Philodicaeus. A Brief Appeal to the University of Oxford on the Classifi-
   cation of Merit: With some Hints to the Rev. H. H. Woodgate on Logic. Oxford, 1829.
[Prior, 1967] A. Prior. Modern logic: The Boolean period. Encyclopedia of Philosophy, Macmil-
   lan and Free Press, vol. 4, 541. New York, 1967.
[Prior, 1949] A. Prior. Logic and the Basis of Ethics. Oxford, 1949.
[Prior, 1967] M. Prior. Whately, Richard. Encyclopedia of Philosophy. Macmillan and Free
   Press, vol.8, pages 287–288. New York, 1967.
[Reid, 1843] T. Reid. The Works of Thomas Reid. London, 1843.
[Sanderson, 1615] R. Sanderson. Logicae Artis Compendium, 1615 . E. J. Ashworth (ed.),
   Bologna, Editrice CLUEB, 1985.
[Stewart, 1814] D. Stewart. Elements of the Philosophy of Human Nature. Vol.2, London, 1814.
   [Vol. 1, 1792; Vol. 3, 1827.]
[Thomas, 1964] I. Thomas. Medieval aftermath: Oxford logic and logicians of the seventeenth
   century. In Oxford Studies Presented to Daniel Callus (Oxford Historical Society, new series),
   pages 297–311, Oxford, 1964.
[Van Evra, 1984] J. Van Evra. Richard Whately and the rise of modern logic. History and
   Philosophy of Logic, 5: 1–18, 1984.
[Wallis, 1687] J. Wallis. Institutio Logicae, ad Usus Accomodata. Oxford, 1687.
[Ward, 1965] W. Ward. Victorian Oxford. London, 1965.
[Watts, 1725] I. Watts. Logick, or the Right Use of Reason. London, 1725.
[Wells, 1897] J. Wells. Oxford and its Colleges. London, 1897.
[Whately, 1875] E. J. Whately. Life and Correspondence of Richard Whately D.D.. 3rd ed.,
   London, 1875.
[Whately, 1826] R. Whately. Elements of Logic. London, 1826.
This page intentionally left blank   THE LOGIC OF SIR WILLIAM HAMILTON:
  TUNNELLING THROUGH SAND TO PLACE
                        THE KEYSTONE IN THE
                            ARISTOTELIC ARCH
                                          Ralph Jessop
                       1 BACKGROUND AND INTRODUCTION
        Every learner in science, is now familiar with more truths than Aristo-
        tle or Plato ever dreamt of knowing; yet, compared with the Stagirite
        or the Athenian, how few, even of our masters of modern science, rank
        higher than intellectual barbarians! (Hamilton, ‘Philosophy of Percep-
        tion’, p. 40).1
The Enlightenment initiated the modern world as the product of a hitherto un-
surpassed devotion to reason and scepticism. The scientific successes, ideals of
progress, material advancement, hopes of social amelioration and freedom of the
Enlightenment, were accompanied by catastrophic failures, conspicuous atrocities
perpetrated in the name of reason and authority, and increasing fears of a dreadful
new age of barbarism. Several Western countries incurred massive rifts, upheavals,
wars, and profound societal changes that impinged upon or were feared to be the
results of Enlightened thought. Following the shock-waves of the American and
French revolutions, as some of the first effects of the industrial revolution were
beginning to be felt, divisions at the heart of the Enlightenment between reason
and scepticism resurfaced in varying guises in England, France, Germany, and
Scotland. During the 18th century Scotland had undergone major economic and
political changes that both weakened the country’s autonomy and yet liberalised
its intelligentsia in ways that helped foster that great flourishing of intellectual
talent we now call the Scottish Enlightenment. This intellectual movement laid
the groundwork for so many succeeding cultural and material changes across the
world. A number of its leading lights were members of the University of Glasgow.
   1 Sir Willam Hamilton, Works of Sir William Hamilton, with an introduction by Savina
Tropea, 7 vols (Bristol: Thoemmes Press, 2001), vol. 1, Discussions. All references to Hamilton’s
articles are to this reprint edition and in this form. Page numbers correspond to, Discussions on
Philosophy and Literature, Education and University Reform, 2nd edn (London: Brown, Green
and Longmans; Edinburgh: MacLachlan and Stewart, 1853).
Handbook of the History of Logic. Volume 4: British Logic in the Nineteenth Century.
Dov M. Gabbay and John Woods (Editors)
 c 2008 Elsevier BV. All rights reserved.
94                                       Ralph Jessop
It was here, in one of the homes in the Professors’ Court of this University that
William Stirling Hamilton was born on 8 March 1788.
   The cultural antecedents within Hamilton’s family background are interesting.
In the late 17th century, two ancestors were leading Covenanters; in the 18th ,
several became somewhat distinguished academics. One of Hamilton’s namesakes
became the Professor of Divinity and later Principal of the University of Edin-
burgh — intriguingly he ‘acquired a high reputation [. . . ] for theological erudi-
tion’ (Veitch, p. 5).2 But it was in medicine that Hamilton’s direct male ancestors
excelled. His grandfather, Thomas Hamilton, a professor of medicine at the Uni-
versity of Glasgow, was fairly close to some of the more eminent medics at the
University, such as William Cullen, Joseph Black, and William and John Hunter.
However, Thomas was also frequently in the company of Adam Smith and James
Watt, since not only were they connected through their respective roles within the
University, but they were also members of the literary Anderston Club, presided
over by the classical scholar and, to some extent still renowned, Professor of Math-
ematics, Robert Simson. Hamilton’s father followed in his own father’s academic
footsteps but, having been Professor of Anatomy from his early 20s, he died young,
aged just 31.
   William Hamilton’s academic lineage, the mainly Glasgow-based Enlightenment
figures of his father’s and grandfather’s acquaintance, and the general educational
ethos contributed to by a good number of the University’s alumni and profes-
sors during the century of Hamilton’s birth, probably played important roles in
helping to mould the academic he would later become. Certainly it does seem
as though Hamilton looked back into his past and may have found there sources
of inspiration with regard to his somewhat pugilistic critical approach to philoso-
phy, his legendarily extensive erudition, distinctive and in many ways exemplary
pedagogical style, understanding of the nature of philosophy, and (in the works
of Thomas Reid and Dugald Stewart) subject matter of extensive later study.
Particularly with regard to Reid (who was the Professor of Moral Philosophy at
Glasgow and is generally recognised as the founding father of the Common-Sense
school), Hamilton’s own development of Reidian philosophy is a significant factor
that I shall briefly return to later. However, although Hamilton’s intellectual in-
heritance from mainly Glasgow-based Enlightenment scholars must have helped
shape his intellect, the educative role of his mother, Elizabeth Hamilton, should
not be forgotten.
   Hamilton was just 2 years old when his father died. His mother played a crucial
role in his educational development. Elizabeth probably imbued in him a great
keenness to excel, while balancing against this her various attempts to ensure that
he did not develop too fast by, for example, returning him to school education
in England following a period at Glasgow University and affording him ample
leisure time during vacations to enjoy various physical pursuits and the compan-
ionship of other boys and his younger brother Thomas (who later gained some
fame as a writer and the author of the novelistic account of pre-industrial Glas-
   2 John Veitch, Memoir of Sir William Hamilton (Edinburgh and London: Blackwood, 1869).
                              The Logic of Sir William Hamilton                                 95
gow, Cyril Thornton (1827)).3 Hamilton’s early childhood and overall educational
experiences under the supervision of a strong mother were therefore, as it were,
the complete obverse of the deeply unpleasant regime so notoriously inflicted by
James Mill and Jeremy Bentham on John Stuart Mill, the man who would later
become Hamilton’s greatest antagonist (notably some nine years after his death).
Mill’s one-time famous attack on Hamilton in his longest philosophical work, the
Examination of the Philosophy of Sir William Hamilton, still stands far in excess of
virtually any other attempt to disparage his standing as a philosopher.4 Since Mill,
most commentators who in one way or another berate Hamilton, either merely add
minor footnotes to Mill’s Examination or uncritically accept his authority. Since
Mill’s Examination — peppered with numerous misreadings of Hamilton — is in
so many ways misleading, it deserves a thorough critical reassessment which I have
not judged to be appropriate or even possible within the scope of this chapter.5
   Hamilton’s principal biographer, John Veitch, claims that ‘no son could cherish
greater regard or a more loyal affection for a mother than he did’ (Veitch, p. 12).
However, Hamilton’s early letters to his mother often suggest a surprisingly direct
and at times high-handed manner towards her (see Veitch, pp. 25-6). Fiery, im-
perious, peremptory, Hamilton’s style of writing in these letters possibly indicates
a certain fierceness of temperament tolerated or even enjoyed by his mother. Sev-
eral of his somewhat more mature letters suggest increasing tenderness towards
her (see letter to his mother, dated 27 November 1807, Veitch, p. 31). By the time
of her death in 1827, which profoundly affected him, Hamilton had lived with her
for almost his whole life and there can be little doubt that he was deeply attached
to her (Veitch, pp. 134-5). Veitch claims that Hamilton wrote to his mother ‘with
the familiarity of an equal in point of years, without reserve, and often strongly’
but he credits her with having been conscious of ‘those qualities of mind which
became afterwards so remarkable’ and he accords to Elizabeth the praise of re-
solving ‘to give him every advantage of education which lay in her power’ (Veitch,
p. 27).
   Hamilton’s early school education was mainly at the Glasgow Grammar School,
followed by a brief spell in the Latin and Greek classes at the University of Glasgow
in 1800. He was at this time just twelve years old. Though it was not uncommon
for boys to attend Scottish universities at such a young age, Hamilton was certainly
   3 Thomas Hamilton, The Youth and Manhood of Cyril Thornton, ed. by Maurice Lindsay
(Aberdeen: The Association for Scottish Literary Studies, 1990).
   4 John Stuart Mill, An Examination of Sir William Hamilton’s Philosophy and of the Prin-
cipal Philosophical Questions Discussed in his Writings (London: Longman, Green, Longman,
Roberts & Green, 1865).
   5 For one attempt to reconcile the logics of Mill and Hamilton prior to Mill’s Examination,
see, [Alexander Campbell Fraser], ‘Province of logic and Recent British Logicians’, North British
Review, 33 (Nov. 1860), 401-427. There were also several attempts to defend Hamilton following
Mill’s attack — for example, see, [Alexander Campbell Fraser], ‘Mill’s Examination of Sir William
Hamilton’s Philosophy’, North British Review, 43 (Sept 1865), 1-58. For a brave attempt to
defend Hamilton against Mill, though not specifically addressing his logic, see Dallas Victor Lie
Ouren, ‘HaMILLton: Mill on Hamilton: A Re-examination of Sir William Hamilton’s Philosophy’
(unpublished doctoral thesis, University of Minesota, 1973).
96                                   Ralph Jessop
on the younger side of the norm. But Elizabeth Hamilton decided against William
continuing his studies at Glasgow in the following academic session and he and
Thomas were therefore moved in 1801 to schools in England. After returning to
Scotland both boys entered the University of Glasgow in 1803, where William
seems to have performed well in Latin, though ‘In the classes of Logic and Moral
Philosophy Hamilton was greatly distinguished, having in each carried off the
highest honour of the year, which was then [. . . ] awarded by the votes of the
members of the class’ (Veitch, p. 21).
   Hamilton’s most notable teacher from this time was Professor George Jardine
(1742-1827) in the Logic class. Jardine’s teaching made a lasting impression on
him (Veitch, p. 21). As his studies continued, he studied medicine from 1804,
studying botany and anatomy from 1805. His medical studies continued at Glas-
gow throughout 1806 and, during the winter of 1806-7, at Edinburgh. However, his
book purchases from around this time included a fairly broad array of philosoph-
ical, medical, and historical works (see Veitch, p. 24). Though greatly impressive
by today’s academic standards for undergraduates, the breadth of his reading was
very much in line with the generalist nature of Scottish educational practice and
was not particularly atypical of other students who would later become eminent
scholars and writers. By the time of Hamilton’s death he had amassed some ten
thousand volumes, around eight thousand of which were purchased by the Univer-
sity of Glasgow where they are currently held in a special collection. Within this
collection there are about one hundred and forty editions of Aristotle’s works and
a good number of the texts he reviewed, some of which display neat manuscript
marginalia at times evincing a peculiar degree of care in, for example, comparing
earlier and later editions of works by Archbishop Richard Whately — one of the
Oxford logicians whom Hamilton repeatedly criticised.
   If by this time Hamilton was beginning to distinguish himself as a student of
marked ability at the University of Glasgow, probably the most conspicuous edu-
cational advantage Elizabeth bestowed upon her son, was her determination that
he should complete his university education at Oxford. In 1807 he secured a Snell
Exhibition and entered Balliol College where he continued his studies until taking
his Bachelor of Arts in 1810 — he of course obtained a First. His Oxford days
seem to have been highly stimulating — certainly he made many acquaintances
during this period and he read voraciously. The vast extent of Hamilton’s learn-
ing became somewhat legendary from around the time of his final examination at
Oxford. According to one account, ‘He allowed himself to be examined in more
than four times the number of philosophical and didactic books ever wont to be
taken up even for the highest honours [. . . ]. Since that time [. . . ] there has been
no examination in this University which can be compared with his in respect to
philosophy’ (Veitch, p. 60). However, his first career was not in philosophy but
instead in law.
   He became a member of the Bar in 1813 and having returned to Edinburgh he
lived with his mother and his cousin, whom he later married in 1829. His wife
was a devoted companion and without her hard work as an amanuensis, perhaps
                             The Logic of Sir William Hamilton                          97
little of Hamilton’s lectures would have survived. Though as yet we know too
little about her, Lady Hamilton must have been or become through her marriage
to William, one of those many women during the 19th century whose knowledge
of literature and philosophy far excelled the attainments accredited to them by
posterity. Now that Hamilton was an advocate attempting to make a living at
Edinburgh, with rather too few cases to attend to, he began to investigate his
family history, though not as a light hobby but with a real purpose. Family
tradition had it that William was an indirect descendant of Sir Robert Hamilton
of Preston, a staunchly fierce Covenanter who died in 1701, after which time the
baronetcy was not assumed by the heir and from thenceforth had lapsed into a
mere family memory. After three years of research Hamilton finally presented
a case to the Edinburgh Sheriff which proved that he was the heir-male to his
Covenanting ancestor. Henceforth, William Stirling Hamilton became Sir William
Hamilton, Baronet of Preston and Fingalton (Veitch, p. 69). This may seem a
curious moment in Hamilton’s personal history but no doubt he was motivated by
several practical considerations, not least of which must have had to do with social
and career advancement. From what I can gather from the occasionally sketchy
accounts of his life by Veitch and Monck, Hamilton had sufficient employment as
a lawyer but was only moderately successful: law was ‘but a secondary pursuit for
him’ and instead he haunted the Advocates library with the bibliophilic zeal of an
antiquarian (Veitch, p. 75).6
   If he was less suited to the law than he might have been, his politics were also
an obstruction to great material success since he was a Whig, the ruling party of
the day Tory. As Veitch assesses Hamilton’s politics, he was ‘a man of progress’
and liberal principles, though little if at all involved in party politics of any kind
(Veitch, p. 78). Of course he knew and socialised with many of the leading Scots
of the day, Sir Walter Scott, Thomas de Quincey, Francis Jeffrey, J.G. Lockhart,
Macvey Napier, and many others, but he also had a fair number or European
friends from Russia, France, and Germany. He visited Leipzig in 1817 and again
travelled to Germany in 1820, visiting libraries in Berlin and Dresden. He was
largely instrumental in the Advocates Library’s purchase of an extensive collection
of valuable German works. His interest in German literature and philosophy, which
would later grow to unrivalled proportions among his contemporaries, dates from
around this time.
   Also in 1820 Hamilton applied for the Moral Philosophy professorship at Ed-
inburgh University. Although he had strong support for this chair, not least of
all from the elderly Dugald Stewart, John Wilson (better known as the famous
‘Christopher North’ of Blackwood’s Magazine) secured the post due to the Town
Council’s patronage, though he was by no means a suitable candidate. In many
ways this was quite scandalous and seems to have been entirely due to Hamil-
ton’s Whig politics (Veitch, pp. 96-103). Hamilton had to settle for a poorly paid
Professorship of Civil History to which he was appointed in the following year
   6 W.H.S. Monck, Sir William Hamilton (London: Sampson Low, Marston, Searle & Rivington,
1881).
98                                          Ralph Jessop
and he had to wait for a further 15 years before he secured a post in Philosophy.
Finally, in 1836, though with only a small majority over the other candidate, he
was elected to the Chair of Logic and Metaphysics at the University of Edinburgh.
Veitch gives a fairly thorough account of Hamilton’s appointment and the strong
testimonials that supported him, but the narrowness of his majority and his fail-
ure to secure the earlier appointment of the Moral Philosophy chair indicate that
University appointments in Scotland were handled in an altogether shameful man-
ner — Hamilton was without doubt one of the most philosophically erudite and
talented men in Scotland at that time but party politics, personal preferences,
and unfounded doubts about his religious beliefs were allowed to prevail. In some
ways little had changed since the more understandable yet equally non-academic
rejection of David Hume by the University of Glasgow in the previous century,
nor since the huge debacle that erupted in 1805 when John Leslie was accused of
being an infidel and as a result nearly failed to secure the Mathematics Chair at
Edinburgh in 1805 because he had endorsed Hume’s theory of causality (see Veitch
pp. 183-210).7
    Though there is much to relate about Hamilton’s life from this time on, like
many scholars and dedicated teachers he led an industrious and comparatively
uneventful life, though not unmarred by damaging vicissitudes, such as the deaths
of his son in 1836, his brother Thomas in 1842, and a daughter in the winter of
1844-5. From the time of his appointment to the Logic and Metaphysics Chair
in 1836 until around the mid 40s, Hamilton was clearly working far too hard and
under a great deal of personal strain. Then, in July 1844, aged 56, he suffered
a physically debilitating stroke that partially paralysed him for the rest of his
life. Two years later he became embroiled in a controversy with Augustus de
Morgan concerning their respective quantification systems. Academically, this is
undoubtedly the most troublesome and embarrassing moment in Hamilton’s career
and several have agreed that he behaved rather foolishly. According to William
Kneale, Hamilton was ‘a pedantic Scottish baronet’ who was ‘properly ridiculed by
De Morgan’.8 General opinion about the affair has been that de Morgan came out
on top. 9 Possibly the intensely desperate times of the mid 40s in Britain generally,
a prevailing sense of crisis in Scotland following the massive upheaval due to the
disruption of the Scottish Kirk in 1843, the various bereavements Hamilton had
suffered, and his loss of physical vigour may be factors that ought to be considered.
Hamilton may have acted in an imperious but also a somewhat desperate manner
towards de Morgan but one cannot help but wonder whether his judgment merely
faltered amidst a context of great personal and social difficulties.
    The details of the de Morgan controversy are tediously complex, the complexity
exacerbated by Hamilton’s forensic analyses of the events and his sustained suspi-
    7 On Leslie see, Ralph Jessop, Carlyle and Scottish Thought (Basingstoke: Macmillan, 1997),
pp. 36-9.
    8 William Kneale, ‘Boole and the Revival of Logic’, Mind, 57 (Apr. 1948), 149-75 (p. 152n).
    9 For example, see, William Kneale and Martha Kneale, The Development of Logic (Oxford:
Clarendon, 1962), pp. 353-4.
                               The Logic of Sir William Hamilton                            99
cions over some years that de Morgan had plagiarised or at least was in some way
trying to steal all credit from him. This issue lasted for many years well beyond
Hamilton’s death and although Hamilton backed down considerably from his first
accusations of plagiarism in 1847, neither men really gave up. But, as de Morgan’s
fame rather prospered as a result of the conflict, Hamilton’s prestige diminished.
Peter Heath has analysed and recounted this controversy in admirable detail and
therefore I shall defer to his general understanding of the whole affair, if not to
his tendency to agree that de Morgan’s assessment of Hamilton’s quantification
is right.10 Some, though perhaps not all, of the relevant letters and other doc-
uments were collected together by Hamilton.11 In very general terms, it would
seem that both de Morgan and Hamilton did not fully understand each other’s
respective positions and arguably ‘the two systems are not only distinct from, but
opposed to each other.’12 As I shall argue later in agreement with Robert Fogelin,
de Morgan was mistaken concerning a fundamental point and Hamilton’s system
can thus be shown to be consistent and much more robust than many who sub-
scribed to de Morgan’s standpoint have assumed it to be. One of the most fruitful
and important outcomes of the controversy was the effect it had on George Boole
whose interest in and subsequent mathematization of logic was in no small part
inspired by the rather public disagreement between de Morgan and Hamilton.13
Heath insightfully remarks of the de Morgan controversy, that de Morgan’s notes
‘contain the following, which might well serve (and was perhaps so intended) as
an epigraph for the whole encounter: “Two French squadrons at B — cannonaded
each other — why? Because each took the other for Russians. ‘Then why did
they fight?’ Said a little girl’.14
   Although Hamilton was not affected mentally by the stroke he suffered in 1844
and managed to continue in his Chair at Edinburgh for a good many years after,
it is fair to say that he was greatly impeded by this disablement. Indeed, it is
arguable that his standing as a philosopher may have suffered more from this
than from Mill’s Examination, since one of the greatest problems in studying
Hamilton’s works has always been simply this: he produced no magnum opus
either in metaphysics or logic. It is reasonable, if yet somewhat whimsical, to say
that, had he not been struck down in 1844 he would have at some stage during
his remaining years brought his philosophical endeavours together in at least one
definitive and fully mature volume. But this did not happen and after his paralysis
he produced relatively little until his death at Edinburgh on 6 May 1856, aged 68.
   Although Hamilton did not produce a fully definitive work on metaphysics or on
logic, he nevertheless did write rather extensively, producing material which, in the
  10 Peter Heath, ‘Introduction’ in, Augustus de Morgan, On the Syllogism and Other Logical
Writings (London: Routledge & Kegan Paul, 1966), vii-xxxi (pp. xii-xx).
  11 Miscellaneous Writings and Correspondence, Works, vol. 7.
  12 [H.L. Mansel], ‘Recent Extensions of Formal Logic’, North British Review, 15 (May 1851),
90-121 (p. 95n).
  13 See, Luis M. Laita, ‘Influences on Boole’s Logic: The Controversy between William Hamilton
and Augustus De Morgan’, Annals of Science, 36 (1979), 45-65 (p. 61; p. 65).
  14 Heath, p. xvi.
100                                       Ralph Jessop
recently reprinted edition of his works by Thoemmes Press, fills seven volumes. In
addition to this, he also produced an extraordinarily, many would say excessively
footnoted edition of the Works of Thomas Reid in which he first published his
fragment on Logic, the ‘New Analytic of Logical Forms’ in 1846.15 His footnotes
in Reid’s Works are at times quirky and needlessly pedantic. However, by sharp
contrast with this, towards the end of his life he produced a much less cumbersome
and indeed rather elegant edition of the Works of Dugald Stewart.16 However, long
before these scholarly works Hamilton contributed a series of substantial articles for
the Edinburgh Review all of which, despite his failing health, he managed to publish
in a single volume entitled Discussions on Philosophy and Literature, Education
and University Reform in 1852. The first of these essays, the ‘Philosophy of the
Unconditioned’ (1829) immediately struck readers as largely unintelligible or overly
philosophically sophisticated, mystical even in its complexity17 — but it was this
article that properly launched Hamilton’s career and first made him famous as
the first truly eminent Scottish philosopher since Dugald Stewart who, after many
years of ill health, had died in the previous year.
   In ‘Philosophy of the Unconditioned’ Hamilton reviews the work of Victor
Cousin and his attempt to establish an eclectic philosophy of the Infinito-Absolute.
The main thing to note here is that Hamilton constructs his Law of the Condi-
tioned, a law prescribing the domain of positive knowledge or the realm of what
may be said to be knowable — in some ways it might be regarded as a forerun-
ner of Ayer’s logical positivism, though Hamiltonian positivism is a far cry from
rejecting either metaphysics or theology. A full account of this article is not ap-
propriate here — suffice to say that Hamilton argues that all knowledge lies in a
mean between two extremes of unknowables or inconditionates, which is to say all
things that may be described as absolute or as infinite comprise the boundaries
of knowledge and are thus strictly incomprehensible to us. The absolute and the
infinite are posited as contradictories, neither of which can be positively construed
to the mind but one of which, on the basis of the laws of excluded middle and
non-contradiction, must obtain — though which of the two obtains is incognisable.
Hence, Hamilton inaugurates what I call his doctrine of nescience, or learned ig-
norance as the ‘consummation’ of knowledge (‘Philosophy of the Unconditioned’,
p. 38). As all thought, and hence all knowledge, is conditional, of the plural, phe-
nomenal, limited, Hamilton declares: ‘To think is to condition; and conditional
limitation is the fundamental law of the possibility of thought’ (‘Philosophy of the
Unconditioned’, p. 14).
   Educated in philosophical discourses of the much more sedate and even-tempered
stateliness that typified so much earlier 18th century and contemporary philosoph-
  15 Thomas Reid, The Works of Thomas Reid, preface, notes, and supplementary dissertations
by Sir William Hamilton (Edinburgh: MacLachlan, Stewart; London: Longman, Brown, Green
and Longmans, 1846).
  16 The Collected Works of Dugald Stewart, ed. by Sir William Hamilton, 11 vols (Edinburgh:
Thomas Constable, 1854-60).
  17 ‘Francis Jeffrey to Macvey Napier, 23 November, 1829’, in Macvey Napier [Jnr.], ed., Selec-
tions from the Correspondence of the Late Macvey Napier (London: Harrison, 1877), p. 68.
                               The Logic of Sir William Hamilton                               101
ical prose from Francis Hutchison, to Hume, Smith, Reid and Dugald Stewart,
few at the time may have been able to understand what Hamilton wrote. His
highly concentrated style, densely philosophical critical argumentation, frequent
references to various German philosophers including Kant, Fichte, and Hegel —
though all this must have dazzled and excluded many, few can have failed to catch
the sense of excitement that pervaded ‘Philosophy of the Unconditioned’. But
this must have been more the case with his second major article ‘Philosophy of
Perception’ (1830) which was written in an even more vigorous and racy style than
his first. He now turned his attention to scepticism, principally to the scepticism
of Hume, though one could be forgiven for missing this since more prominently
‘Philosophy of Perception’ launches an extraordinarily vitriolic attack on the man
who had gained such a high reputation as a philosopher and who for many years
continued to be admired by established English philosophers and writers such as
John Stuart Mill and Leslie Stephen and whose job, following his death, Hamilton
failed to get in 1820 — Thomas Brown. Hamilton savaged Brown for leading phi-
losophy back into the morass of Humean scepticism and for fundamentally failing
to grasp the true import of Thomas Reid’s critical philosophy of Common Sense,
which, according to Hamilton, had given a successful, if nonetheless relatively
unsophisticated answer to Hume.
    So much needs to be said about this article but I shall confine myself to just
a few points: in ‘Philosophy of Perception’ Hamilton develops his own version of
Common Sense philosophy in the form of a doctrine of perception which holds that
in the act of perception the self and the not-self were instantaneously revealed in
one indivisible moment of cognition — he calls this theory of perception natu-
ral dualism or natural realism and maintains that, contradistinguished from all
representationist theories of perception that in one way or another tend towards
scepticism, natural dualism is a theory of immediate or presentative perception.
Leaving aside all consideration of just how natural dualism was proffered by Hamil-
ton as the best and most successful counter-argument to what some may think
of as a straw man scepticism of purely theoretic indeterminacy — the absolute
scepticism of Hume — I want to draw attention to Hamilton’s non-Kantian notion
of the relativity of knowledge by means of just one quotation:18
       Relatives are known only together: the science of opposites is one.
       Subject and object, mind and matter, are known only in correlation
       and contrast [. . . ]. Every conception of self, necessarily involves a
       conception of not-self: every perception of what is different from me,
       implies a recognition of the percipient subject in contradistinction from
  18 Hamilton has often been mistakenly thought of as borrowing heavily from Kant. However,
in several places he is critical of Kant and his relativity of knowledge needs to be distinguished
from Kant’s. On this see, Manfred Kuehn, ‘Hamilton’s Reading of Kant: A Chapter in the Early
Scottish Reception of Kant’s Thought’, in George MacDonald Ross and Tony McWalter, eds,
Kant and His Influence (Bristol: Thoemmes Antiquarian Books, 1990), 315–347 (pp. 333–45).
It should also be noted that Hamilton’s notion of the relativity of knowledge is more subtle and
more complex than Mill represents it as being. On this see, John Veitch, Hamilton (Edinburgh
And London: Blackwood, 1882), pp. 201–222.
102                                        Ralph Jessop
       the object perceived. [. . . ]. In Perception, as in the other faculties, the
       same indivisible consciousness is conversant about both terms of the
       relation of knowledge. (‘Philosophy of Perception’, pp. 50–51).
Although the details of Hamilton’s theory of the relativity of knowledge deserve
separate examination, in the above we can see indications of a notion that is of
critical importance to understanding Hamilton’s logic, namely, the notion that
subject and object exist only in correlation with one another, such that opposites,
or the distinguishable terms of subject and predicate in a given proposition, may
be thought of as being held together in a relationship of equation or non-equation
— in this lies the germ of Hamilton’s emphasis on the relativity of concepts and
his quantification of the predicate.
   The influence that Hamilton’s two articles on the unconditioned and perception
had on Victorian thought is also a subject that deserves separate study. Noah
Porter, professor of moral philosophy and metaphysics (1846), and later president
of Yale, wrote effusively and at some length in Veitch’s biography of the consid-
erable extent of Hamilton’s influence on American students (Veitch, pp. 421-8).
However, although Porter thought that Hamilton had been a positive religious
force in American thought, there is another side to this story. Emphasising the
vastness of our ignorance in ‘Philosophy of the Unconditioned’, while regarding
this as a prompt for our wonderment and faith, the most significant direct and
lasting effect of the ‘Philosophy of the Unconditioned’ is perhaps best assessed
in terms of the influence it had upon Henry L. Mansel, a prominent follower of
Hamilton who wrote several articles defending Hamilton’s logic, developed his own
version of it in his Prolegomena Logica where he specifically acknowledges his debt
to Hamilton, and, more popularly, in his Bampton Lectures, gave rise to a doctrine
of Christian Agnosticism.19 But as the agnostic movement developed during the
19th century, Hamilton’s ‘Philosophy of the Unconditioned’ in comparison with
its transmutation into Mansel’s Christian Agnosticism can also be seen as having
a profound effect on anti-Christian agnostics such as Thomas Huxley (Darwin’s
bulldog). As Sheridan Gilley and Ann Loades nicely put it: ‘Huxley saw in Mansel
the suicidally honest theologian, sitting on an inn sign and sawing it off.’20 Hamil-
ton’s importance to the growth of agnosticism, although not widely known, has
certainly been established not only by more recent scholarship but also in some of
the earlier responses to Hamilton.21 Hence, though firstly, inspiring a new religious
piety and apparent salvation from scepticism, but secondly, becoming infused into
  19 For example, see Henry Longueville Mansel, Prolegomena Logica: An Enquiry into the Psy-
chological Character of Logical Processes (Oxford, 1851) pp. xi–xii; ‘The Philosophy of the
Conditioned: Sir William Hamilton and John Stuart Mill’, Contemporary Review, 1 (1866), 31-
49; 185-219; [Bampton Lectures], The Limits of Religious Thought (London : John Murray,
1858).
  20 Sheridan Gilley and Ann Loades, ‘Thomas Henry Huxley: The War between Science and
Religion’, The Journal of Religion, 61 (1981), 285-308 (p. 297).
  21 Bernard Lightman, The Origins of Agnosticism: Victorian Unbelief and the Limits of
Knowledge (Baltimore and London: John Hopkins University Press, 1987), p. 16; Robert Flint,
Agnosticism (Edinburgh and London: Wm Blackwood, 1903).
                             The Logic of Sir William Hamilton                           103
succeeding waves of religious doubt and the growth of agnostic principles more
powerfully damaging to orthodox belief than even Hume’s or Voltaire’s more full-
frontal atheistic attacks on religious belief, the full significance that Hamilton’s
‘Philosophy of the Unconditioned’ would come to have on Victorian society, phi-
losophy, literature, and culture, was nothing short of immense.22
   But Hamilton’s influence can also be seen through certain of his friends or ac-
quaintances and, of course, his students. One of his former students, who was
profoundly influenced by him, became the Professor of Moral Philosophy at the
University of St. Andrews, James Frederick Ferrier, a philosopher whose personal
and philosophical connections with Hamilton run deep and whose work is now
just as undeservedly but even less well known than that of Hamilton. In litera-
ture, Hamilton was a major influence on E. S. Dallas whose two volume work of
literary theory, The Gay Science, positively teems with Hamiltonian philosophy
— but Dallas has also now shrunk into the shadows and is barely known.23 George
Davie claims that Hamilton inspired a number of brilliant scholars including Fer-
rier and the physicist so greatly admired by Einstein, James Clerk Maxwell.24
The Maxwell connection is particularly interesting since a recent work of literary
criticism has opened up a whole new field of study — the Victorian relativity
movement. Christopher Herbert argues in Victorian Relativity that the cultural
and philosophical antecedents of Einstein’s special theory of relativity are to be
found in a succession of Victorian philosophers and thinkers, the writings of whom
have been largely neglected for over a century. Interestingly, just as scholars who
have written on agnosticism have traced its origins in the 19th century back to
Hamilton, Herbert also finds in Hamilton’s theory of the relativity of knowledge
a significant starting point for his study, seeing Hamilton at the beginning of a
major Victorian re-invention of the fundamentally Protagorean relativist rejection
of absolutism.25 First suggested in ‘Philosophy of the Unconditioned’, developed
in ‘Philosophy of Perception’, and, as we shall see, a key component of Hamil-
ton’s logic, Hamiltonian relativism intriguingly re-positions Hamilton as one of
the 19th -century’s key avante garde thinkers.
   It is too easy to see Hamilton in caricature and as of marginal importance,
since this is where the later Victorians placed him and where scholars of the 20th
century left him. However, for all that he was hampered by a corrupt system
of university patronage and indeed may not have been the most self-promoting
of figures, within the context of British philosophy and literature of the 1830s,
  22 See, Ralph Jessop, ‘Carlyle’s Agnosticism: An Altar to the Unknown and Unknowable God’,
Literature and Belief, 25 (1&2) (2005), 381-433 (pp. 395-404).
  23 E.S. Dallas, The Gay Science (London: Chapman and Hall, 1866).
  24 See, George Elder Davie, The Democratic Intellect: Scotland and Her Universities in the
Nineteenth Century, Edinburgh University Publications, History Philosophy & Economics: 12
(Edinburgh: Edinburgh University Press, 1961; repr. 1982), p. 261; John Hendry, James Clerk
Maxwell and the Theory of the Electromagnetic Field (Bristol and Boston: Adam Hilger, 1986),
p. 112.
  25 Christopher Herbert, Victorian Relativity:     Radical thought and Scientific Discovery
(Chicago and London: The University of Chicago Press, 2001), pp. 35-7.
104                                         Ralph Jessop
Hamilton was a leading and highly dynamic writer who played an important part
in re-invigorating Reidian philosophy while also turning British attention arguably
in a different but by no means unrelated direction towards German philosophy.
It must be noted that in early 19th -century Britain very few people could read
German and Hamilton was one of almost a handful of intellectuals during the 1820s
and 30s who had any direct knowledge of the works of Kant and other German
philosophers such as Hegel. Hamilton was one of a tiny number of Germanizers
who promoted the study of German philosophy and, given the rise of neo-Kantian
and neo-Hegelian philosophy in Britain amongst philosophers who owed some debt
to Hamilton, in this alone his influence was probably much more profound than
has generally been appreciated.
    Another major Germanizer was Hamilton’s one-time friend, the literary man of
letters, Thomas Carlyle, who knew Hamilton from at least the late 1820s. Carlyle,
though a literary giant of the Victorian period and well beyond, of proportions it
would be difficult to exaggerate, is yet another figure who is now increasingly little
known and poorly understood. Not much is positively known about their friend-
ship but at the time of its publication in 1829 Carlyle read Hamilton’s ‘Philosophy
of the Unconditioned’ with admiration and in later life he wrote a highly rever-
ent Reminiscence of Hamilton which is published in Veitch’s biography, followed
by a fairly intimate letter from Carlyle written in 1834 in which he says, ‘Think
kindly of me; there are few in Scotland I wish it more from’ (Veitch, pp. 121–7).
There are undoubtedly many interesting parallels between Carlyle and Hamilton,
some of which were detected in Carlyle’s own lifetime by his close friend David
Masson.26 I have discussed some of the interconnections between Carlyle and
Hamilton elsewhere.27
    Hamilton produced a good number of other articles for the Edinburgh Review,
including the following selection: ‘On the Revolutions of Medicine in Reference
to Cullen’ (1832); ‘On the Study of Mathematics as an Exercise of the Mind’
(1836) — controversially, though rather in keeping with his Scottish predecessors
attitudes about mathematics, Hamilton did not think that mathematics was a good
exercise of the mind as part of a liberal education and instead advocated philosophy
while in several ways indicting the emphasis on mathematics at Cambridge;28 ‘On
the Patronage and Superintendence of Universities’ (1834); ‘On the State of the
English Universities with More Especial Reference to Oxford’ (1831). Several of his
articles were pointedly critical of Oxford and Cambridge and it must be said that
Hamilton undoubtedly set himself up for retributive attacks from some scholars
in response to his various denunciations of the established ancient universities of
  26 David Masson, Recent British Philosophy: A Review with Criticisms including some Com-
ments on Mr Mill’s Answer to Sir William Hamilton, 3rd edn (London: Macmillan, 1877),p.
69.
  27 For example see, Jessop, Carlyle and Scottish Thought, pp. 28-31. Also see Alex Benchi-
mol, ‘William Hamilton’ in, The Carlyle Encyclopedia, edited by Mark Cumming (Cranbury,
NJ:Farleigh Dickinson University Press, 2004), 207-9.
  28 For example compare, Richard Olson, ‘Scottish Philosophy and Mathematics 1750-1830’,
Journal of the History of Ideas, 32 (1971), 29-44 (pp. 41-4).
                           The Logic of Sir William Hamilton                      105
England. More directly germane to our subject matter and also involving his
first of many salvos against English logicians, is his ‘Logic: The Recent English
Treatises on that Science’ (1833).
   In this article he gave yet another virtuoso performance that at once established
his reputation as one of the foremost logicians of his day. But, as with some of
his other slashing remarks on contemporary scholars and, as he saw it, the present
philistinism of learning in Britain, in ‘Logic’ he was unsparing in his treatment
of several Oxonian logicians. Supposedly reviewing eight recent publications on
logic, he focuses almost exclusively on Whately’s Elements of Logic, making only
brief reference to George Bentham’s Outline of a New System of Logic (1827). As
some commentators have noted, the cursory treatment of Bentham is odd since a
system for quantifying the predicate is given in the work of this nephew of the more
famous Jeremy Bentham.29 Hamilton’s paternity of this doctrine has often been
called in doubt but since most of the specific characteristics of his quantification
are largely, if not exclusively, peculiar to him, and since his own theory is so
thoroughly grounded in a painstaking explication of the grounds for quantifying
the predicate, I shall not engage with the tortuous historical complexities and shall
instead attempt in the following sections to outline how Hamilton arrives at his
quantification system and what that system itself is. Furthermore, it is likely that
Hamilton was so dismissive of Bentham’s Outline that he simply did not read all
of the text and cast it aside in order that he might focus on the main logician, his
principal target, Richard Whately.
   Though Hamilton intimates a certain degree of respect for the natural abilities
of the authors under review, he denounces their lack of genuine originality — the
source of Hamilton’s ire is clear and is a much repeated complaint elsewhere in his
work, namely, the inadequacy of the authors’ learning:
       None of them possess — not to say a superfluous erudition on their
       subject — even the necessary complement of information. Not one
       seems to have studied the logical treatises of Aristotle; all are ignorant
       of the Greek Commentators on the Organon, of the Scholastic, Ramist,
       Cartesian, Wolfian, and Kantian dialectic. (‘Logic’, p. 129).
And so he continues to cut and hew his way through the inadequate learning of
contemporary Oxonians. But, as Whately’s Elements stood pre-eminent, it is this
work that Hamilton takes to task and he proceeds to make point after forensic point
against Whately, all the time demonstrating his own vastly superior knowledge of
swathes of the literature of logic and outlining several features of his own system
of logic.
   As part of his sustained attempt to claim priority in discovering his system of
quantification, in a much later footnote in Discussions Hamilton asserts that on
the basis of a rather vague authority — ‘the tenor of the text’– the ‘Logic’ article
shows that he ‘had become aware of the error in the doctrine of Aristotle and the
logicians, which maintains that the predicate in affirmative propositions could only
  29 Heath, p. xvi.
106                                           Ralph Jessop
be formally quantified as particular’ (‘Logic’, p. 162n). More strongly, and with
much better justification, he writes in an appendix in Discussions, ‘Touching the
principle of an explicitly Quantified Predicate, I had by 1833 become convinced
of the necessity to extend and correct the logical doctrine upon this point. In
the article on Logic [. . . ] the theory of Induction there maintained proceeds on
a thoroughgoing quantification of the predicate, in affirmative propositions’ (Dis-
cussions, p. 650). Certainly the 1833 ‘Logic’ article does suggest this but it does
not display the full quantification system (see ‘Logic’, p. 163). As I shall indicate
later, although the quantification system was not fully articulated until the publi-
cation of his ‘New Analytic’ in 1846, a good deal of Hamilton’s treatment of logic
in his lectures does proceed painstakingly towards the quantification. That he was
teaching this system during the late 1830s, several years ahead of de Morgan’s
quite different quantification, seems to have been generally agreed and is patently
evident in the wonderfully clear account given by one of his students, Thomas
Spencer Baynes, whose winning essay written for a competition set by Hamilton
in 1846 was later published as An Essay on the New Analytic of Logical Forms in
1850.30
   In ‘Logic’ Hamilton very much lays out his stall. He begins with a brief definition
of logic that he later elaborates:
       Nothing, we think, affords a more decisive proof of the oblique and
       partial spirit in which philosophy has been cultivated in Britain, for
       the last century and a half, than the combined perversion and neglect,
       which Logic — the science of the formal laws of thought — has expe-
       rienced during that period. (‘Logic’, p. 119).
Just a few years after writing this, Hamilton commenced his Lectures on Logic as
the Professor of Logic and Metaphysics at Edinburgh. These were posthumously
published in four volumes, two on Metaphysics in 1859 and two on Logic in 1860.
In the Lectures on Logic, Hamilton spends a lot more time carefully explaining his
above definition of logic as ‘the science of the formal laws of thought’. Though
his definition of logic in the lectures is interesting and deserves discussion, I shall
merely summarise some main points here.
   Hamilton basically holds that Logic proper or Formal Logic is Abstract logic,
dealing only with necessary inference and devoid of all adventitious or extra-logical
matter or contingent considerations. However, the nature of Logic as a pure sci-
ence had been greatly misunderstood in Britain: ‘Bacon wholly misconceived its
character in certain respects; but his errors are insignificant, when compared with
the total misapprehension of its nature by Locke’ (LL.I.29).31 The British had
  30 Thomas Spencer Baynes, An Essay on the New Analytic of Logical Forms, Being that which
Gained the Prize Proposed by Sir William Hamilton, in the Year 1846, for the Best Exposition
of the New Doctrine Propounded in his Lectures; with an Historical Appendix (Edinburgh:
Sutherland and Knox; London: Simpkin, Marshall, and Co., 1850).
  31 All references in this form are to the Lectures on Logic, ed. by H.L. Mansel and John Veitch
(Edinburgh and London: Blackwood, 1860), vols I or II; vols 5 or 6 in the Thoemmes reprint
edition of Hamilton’s Works.
                           The Logic of Sir William Hamilton                      107
mistakenly departed from a certain general agreement on the formal nature of
Logic among ancient and more recent German logicians. They had ignorantly
or perversely deviated from centuries of collective wisdom and while German lo-
gicians from the time of Leibniz had probably done more than most to further
the science, Hamilton seems to have regarded the more recent and more serious
perversions and confusions in the work of recent English authors, in particular
Whately, as a sort of further entrenchment of the very kind of misconstrual most
likely to hinder the science (LL.I.40). It was also reprehensible that such recent
British scholars should be ignorant of the German logicians: ‘Great Britain is,
I believe, the only country of Europe in which books are written by respectable
authors upon sciences, of the progress of which, for above a century, they have
never taken the trouble to inform themselves’ (LL.I.33).
   Distinguishing between Special or Concrete Logic (logic in its particular ap-
plications or instantiations in the several arts and sciences) and Pure or General
or Abstract Logic, for Hamilton, Pure Logic has to be contradistinguished from
any particular subject or discipline, the object-matter of which must necessarily
be contingent due to the nature of the topics it addresses or to which Logic is
being in some sense applied or put into practice (LL.I.56). However, while Spe-
cial Logic is dismissed this is not to say that practical matters to do with Logic
must be altogether excluded from consideration (see LL.I.61-2). He coins the term
‘Modified Logic’ to describe what he argues had been improperly called Applied
Logic by Kant and some other German philosophers, defining Modified Logic as
‘a science, which considers thought not merely as determined by its necessary
and universal laws, but as contingently affected by the empirical conditions under
which thought is actually exerted’ (LL.I.60). Although Hamilton’s treatment of
Modified Logic in the second volume of his Lectures on Logic as the correlative
second main branch of Abstract Logic, is certainly interesting, I have chosen not
to discuss it but instead focus on what for Hamilton was clearly of much greater
immediate importance, namely, Formal or Pure Logic. He insists that Pure Logic
really comprises the whole of Abstract Logic — Modified Logic is ‘a mere mix-
ture of Logic and Psychology’; ‘There is in truth only one Logic, that is, Pure
or Abstract Logic’, ‘Modified Logic being only a scientific accident, ambiguously
belonging either to Logic or to Psychology’ (LL.I.63).
   With such points in mind, he provisionally defines Logic as ‘the Science of the
Laws of Thought as Thought’ (LL.I.4). Extruding the contingent, inasmuch as
this is possible, Hamilton claims that Logic is only concerned with those phenom-
ena of formal (or subjective) thought that are necessary or ‘such as cannot but
appear’ as opposed to the contingent phenomena of thought, or ‘such as may or
may not appear’ (LL.I.24). Hence, through his final introduction of the notion
of necessary laws being the sole province of Logic, Hamilton asserts that ‘Logic,
therefore, is at last fully and finally defined as the science of the necessary forms
of thought’ (LL.I.24). Though this is Hamilton’s final definition of Logic, he takes
care to explain the sense or ‘quality’ of ‘necessary’ by extensively quoting Wilhelm
108                                        Ralph Jessop
Esser’s System der Logik.32 Hence, to summarize Esser’s arguments as translated
by Hamilton: the necessity of a form of thought is contradistinguished from con-
tingency by being subjective, which is to say that a necessary form of thought
‘must be determined or necessitated by the nature of the thinking subject itself’
as opposed to being determined objectively; since we are incapable of conceiving
the possibility of its non-existence, such incapacity warrants the notion that the
form of thought in question is original to, or a constitutive feature of, the thinking
subject or human mind, and thus the subjective necessity of a form of thought
‘must be original and not acquired’; as necessary, subjective, and original in the
sense explained, the form of thought that comprises the object-matter of Logic
cannot necessitate on some occasions and not on others, and thus it must also be
universal ; and finally, further enriching the sense or quality of ‘necessity’, accord-
ing to Esser, ‘if a form of thought be necessary and universal, it must be a law;
for a law is that which applies to all cases without exception, and from which a
deviation is ever, and everywhere, impossible, or, at least, unallowed’ (LL.I.24-5).
With the sense of ‘necessity’ within the phrase ‘necessary forms of thought’ hereby
explained in terms of something subjectively determined by the human mind as
subject and not objectively determined or extraneous to the mind, original and
not acquired, universal, and a law, Hamilton gives what he regards as his ‘most
explicit enunciation of the object-matter of Logic’ as: ‘Logic is the science of the
Laws of Thought as Thought, or the science of the Formal Laws of Thought, or
the science of the Laws of the Form of thought; for all these are merely various
expressions of the same thing.’ (LL.I.25-6).
   Many more recent formal logicians would generally agree that formal Logic,
as contradistinguished from informal Logic and the subject matter or topics of
Rhetoric, is entirely or principally concerned with deductive arguments and thus
with valid inference and the necessary laws that pertain to or determine validity.
But we need to take notice of the state of logic in 1830s Britain as Hamilton
understood it: his contemporaries appeared to him to be woefully misguided,
unscholarly in their reading of the Aristotlean tradition, and blissfully ignorant of
the more recent German tradition in logic since the time of Leibniz. Hamilton’s
peers were, as far as he could tell, largely unaware of how the recent German
writers including and after Kant had far excelled the Oxford logicians in their
knowledge and understanding of the subject. Furthermore, it seems clear that
part of Hamilton’s crusade was against the prevailing tendency of philosophy in
Britain towards the increasingly prevalent mediocrity and barbarity of his times
that almost inevitably was following in the wake of and was implicitly collusive
with an era of rapidly advancing materialism. Pervaded from its outset by reason
and scepticism, during the early decades of the 19th century the Enlightenment
was evolving into new forms of a more socially pervasive utilitarian rationality and
sceptical subversions of faith. Acutely conscious of such trends in philosophical
  32 Hamilton translates for his students many passages from Esser. His inclusion of these nu-
merous quotations implies the commencement of a substantial realignment in British philosophy
with German logicians.
                            The Logic of Sir William Hamilton                      109
discourse Hamilton took considerable pains to establish his standpoint as one
much more closely in line with the German approach that regarded Logic as a
pure science.
                            2 DOCTRINE OF CONCEPTS
       Logic [. . . ] is exclusively conversant about thought,–about thought
       considered strictly as the operation of Comparison or the faculty of
       Relations; and thought, in this restricted signification, is the cognition
       of any mental object by another in which it is considered as included,–
       in other words, thought is the knowledge of things under conceptions.
       (LL.I.40).
Shortly after the above quotation Hamilton distinguishes between the act of con-
ceiving (conception) and the thing conceived (concept) and briefly suggests a sim-
ilar distinction with regard to perception, somewhat tentatively coining the term
‘percept’ some 40 years before its first recorded use in the Oxford English Dictio-
nary. He takes particular care to point out that Logic is concerned with ‘thought
considered as a product; that is, as a concept, a judgment, a reasoning’ (LL.I.74).
The operation of Comparison, or as he later calls it, the Faculty of Comparison,
produces Concepts, Judgments, and Reasonings. However, Hamilton argues that
Concepts and Reasonings are modifications of Judgments, ‘for the act of judging,
that is, the act of affirming or denying one thing of another in thought, is that
in which the Understanding or Faculty of Comparison is essentially expressed’
(LL.I.117). This means that for Hamilton, ‘A concept is a judgment’ and as such
it collects together or is ‘the result of a foregone judgment, or series of judgments,
fixed and recorded in a word,–a sign’ which may be supplemented or extended by
additional attributes, themselves judgments (LL.I.117). Thus, as a concept in a
sense fixes a judgment or series of judgments, collecting together various attributes
within a single term, so also can it be analysed into these components or amplified
by the annexation of further attributes.
   An important point to note here that will later be of fundamental significance to
Hamilton’s quantification of the predicate, is that for Hamilton an oft-ignored and
oft-violated postulate or principle of Logic is that the import of the terms used in
a judgment or reasoning should be fully understood or made explicit, which is to
say that ‘Logic postulates to be allowed to state explicitly in language all that is
implicitly contained in thought’, that Logic demands licence to make explicit the
full import of any particular concept or term, much as it attempts to do in making
overt all of the steps and relations involved in a given process of reasoning or ar-
gument. Though this fundamental postulate is simple in its statement, Hamilton
clearly regarded its significance as central to his project to evolve or develop Logic
as a Pure science: ‘This postulate [. . . ], though a fundamental condition of Logic,
has not been consistently acted on by logicians in their development of the science;
and from this omission have arisen much confusion and deficiency and error in our
110                                            Ralph Jessop
present system of Logic’ (LL.I.114). Interestingly, in one place he makes the postu-
late more precise by replacing the term ‘implicit’ with ‘efficient’ (NA.LL.II.270).33
The analysis of concepts into their often implicitly-held component attributes as
also their amplification by means of adding new attributes, effectually constitutes
an adherence to this postulate. Indeed, arguably, Hamilton’s emphasis on the im-
portance of being rigorously explicit elevates this postulate to the status of what
one might call a meta-theoretical principle of explicitness.
   With reference to the Latin ‘concipere’ he explains that traditionally concep-
tion indicated ‘the process of embracing or comprehending the many into the one’,
‘the act of comprehending or grasping up into unity the various qualities by which
an object is characterised’ (LL.I.120). Within our consciousness this process is
typified by two cognitions, one immediate and ‘only of the individual or singular’,
the other ‘a knowledge of the common, general, or universal’ (LL.I.122). Thus ‘a
Concept is the cognition or idea of the general character or characters, point or
points, in which a plurality of objects coincide’ (LL.I.122, and see, 122-3). Mir-
roring his notion of the formation of language, which he outlines in his Lectures
on Metaphysics,34 as a synthesising process that moves from chaos to the con-
struction of general and universal terms which enable a complex and iterative
relationship with the individuals or particulars into which such concepts may be
analysed — a process of composition and decomposition that he occasionally de-
scribes as organic — Hamilton regards the formation of concepts as a complex
process involving human agency or the exertion of an ‘act of Comparison’ upon
an otherwise chaotic or confused array of presentations. Reminiscent of his highly
significant relativist maxim, ‘To think is to condition’, Hamilton is thus acutely
aware of the role of human agency or volition in the process or mental activity
involved in the formation of concepts as unities consisting of various points drawn
together by an act of comparison and the implicit intentionality within this action
‘of discovering their similarities and differences’ (LL.I.123). This awareness of the
part played by human agency or volition, and the artificiality and partiality of
concepts, becomes clearer when he invokes ‘the act called Attention’, by means
of which certain objects and qualities forming any given concept become strongly
highlighted (LL.I.123). As Hamilton explains the mental operations of attention
and abstraction, these two processes involved in the formation of concepts ‘are,
as it were, the positive and negative poles of the same act’, by means of which as
some objects and qualities become highlighted, others ‘are thrown into obscurity’
(LL.I.124, 123). He is claiming that the thinking subject’s point of view or per-
spective, or conditioning role is crucial to the formation of concepts, and as we
shall see later, point of view also plays an important role with regard to Hamilton’s
treatment of propositions.
   What Hamilton is describing here, and in what follows this observation about
the act of abstraction and attention, is the relativity of concepts, and thereby he
  33 All references in this form are as above to the Lectures on Logic with ‘NA’ prefixed to indicate
the ‘New Analytic of Logical Forms’ as given in vol.II.249-317.
  34 Lectures on Metaphysics, II.327-332.
                          The Logic of Sir William Hamilton                       111
begins to point up certain implicit features of concepts that emphasise their fluidity
or their adaptability, as well as their inherent inadequacy or insufficiently explicit
nature. For Hamilton the act of comparison combined with that of abstraction
and attention reduces in consciousness the multiple (or really differing objects of
consciousness) into the unity that is a concept (see LL.I.124). By throwing out of
view the non-resembling marks or characters or points that in reality individuate
one individual from another, and by attending to the resembling points alone, we
treat these resembling points as though they were identical and thereby synthesise
them into a unity, a concept through which we think of the individuals or to which
we relate these individuals. Implicitly, in opposition to mechanistically grounded
(or mechanically modelled) theories of the mind, for Hamilton the formation of
concepts in the mental act of conception is a single or indivisible, quasi-organic
process of thought that may be analysed into components for the purpose of speak-
ing about and better comprehending concepts (see LL.I.133). No doubt integral to
or consistent with Hamilton’s natural dualism, he regards our knowledge of what
is presented and represented in consciousness — the phenomena of consciousness
— as ‘a direct, immediate, irrespective, determinate, individual, and adequate
cognition’ (LL.I.131). Such cognitions of the phenomena of consciousness, though
virtually self-sufficient as cognitions, do not, for Hamilton, constitute thought, or
so it would seem, since repeatedly he construes thought in terms of a relational
or relative process in which the thinking subject is more distinctively active or
operates as an agent conditioning the objects of thought. Hence, by contrast with
the mere phenomena of consciousness as self-sufficient cognitions, by means of
which the human mind may be thought of as, if not merely sentient, then active
but unthinkingly or non-rationally so, ‘A concept, on the contrary, is an indirect,
mediate, relative, indeterminate, and partial cognition of any one of a number of
objects, but not an actual representation either of them all, or of the whole at-
tributes of any one object’ (LL.I.131). This means that concepts are non-absolute
or ‘not capable of representation as absolute attributes’ (LL.I.137). Earlier, in
attempting to explain the inadequacy of concepts and their relativity, Hamilton
pointed up the partiality of concepts, their dependence upon selective ‘represen-
tation of a part only of the various attributes or characters of which an individual
object is the sum’ — for example, as we think Socrates through a particularly
small range of attributes, say, man, biped, animal, our representation of him will
be proportionately less adequate than when we think him through a greater range
of attributes. Hence, a concept, construed as a collection of attributes to which
the concept refers, are always one-sided or partial and to a greater or lesser extent
inadequate by contrast with the plenitude of attributes that constitute any given
individual.
   However, this partiality or incompleteness of concepts, arising due to the com-
bined act of abstraction and attention that conceiving necessarily involves, while
it suggests the relativism of the thinking subject’s conditioning influence on the
objects of thought, does not seem to be the relativity that Hamilton particularly
wishes to emphasise. Instead, the relativity of concepts to which he draws our
112                                      Ralph Jessop
attention inheres in the relational nature that a concept does not overtly make
explicit but rather, in a sense, disguises by means of the fiction and even illusion
of unitariness we tend to impose upon the collection of more or less resembling
but nonetheless differing individual attributes that comprise any given concept.
According to Hamilton, ‘A concept or notion, as the result of a comparison, neces-
sarily expresses a relation. It is, therefore, not cognisable in itself, that is, it affords
no absolute or irrespective object of knowledge, but can only be realised in con-
sciousness by applying it, as a term of relation, to one or more of the objects, which
agree in the point or points of resemblance which it expresses’ (LL.I.128). Although
Hamilton’s account of the relativity of concepts is far from being unproblematic, he
asserts that the passage just quoted resolves ‘the whole mystery of Generalisation
and General Terms’ and thus that his notion (that a concept expresses a relation,
and is not therefore an ‘absolute or irrespective object of knowledge’) resolves the
disputes between Conceptualists and Nominalists. That concepts constitute ab-
solutes or objects of knowledge is an ‘illusion’ that fundamentally arises due to
our conversion of similarity into identity, and thereby, according to Hamilton, ‘the
real plurality of resembling qualities in nature is factitiously reduced to a unity in
thought; and this unity obtains a name in which its relativity, not being expressed,
is still further removed from observation’ (LL.I.128).
   This may sound as though Hamilton is saying that concepts are fundamen-
tally fallacious; that they are fallaciously unities, actually pluralities. However,
it needs to be noted that while he may be saying that the view that concepts
are absolute objects of knowledge, or that they express unities, arises due to our
conversion of similarity into identity, the illusion of concepts as objective abso-
lutes thus generated is not due to some inherent fallaciousness within our mental
faculties. Rather, this illusion is due to an insufficiently rigorous attention to the
internal components, structure, and relational processes involved in any given in-
stance of conceiving as deeply integrant to the characteristic features and nature of
the product of such thought, namely, a concept. Furthermore, as he later argues,
concepts, formed by comparison and hence relative or expressive of a relation, ‘can
only be thought of in relation to some one of the individual objects they classify’,
but as such ‘they fall back into mere special determinations of the individual ob-
ject in which they are represented. Thus it is, that the generality or universality of
concepts is potential, not actual’ (LL.I.134). Though Hamilton does not go on to
explain that this implies that concepts have to be regarded always as in some sense
provisional and thereby capable of modification, supplementation, or some other
adjustment, and available for examination from differing points of view, this rel-
ativist attack on the implicit absolutism of concepts construed as non-relativistic
objects of knowledge, is further pursued by Hamilton’s excursus into the role of
language in a way that further brings to the fore the sense in which concepts are
provisional artefacts, the establishment and permanence of which is dependent on
the human subject — that concepts are necessarily subjective.
   He apologizes for his digression into the extra-logical domain of metaphysics, a
metaphysics which he admittedly here only sketches (LL.I.131). However, it would
                          The Logic of Sir William Hamilton                        113
seem that, for Hamilton, in order to analyse the products of thought (concepts) as
principal components within demonstrative arguments, we need to understand the
relativity of concepts by making this explicit. To do so may involve some digression
into metaphysical discourse, and even the enunciation of a metaphysical standpoint
consistent with or flowing out of Hamilton’s natural dualist position, a position
more or less contentious for some readers but of substantive interest in relation to
the Scottish Common Sense tradition in philosophy out of which Hamilton devel-
oped the term ‘natural dualism’. Clearly concerned that his constriction of Logic
to the necessary laws of thought is being violated by metaphysical considerations,
he also apologizes for digressing onto a similarly extra-logical consideration of lan-
guage. However, just as metaphysical discourse is at the interface between Logic
and the nature of the object-matter of Logic, an important part of Hamilton’s
description of the generation of concepts as illusorily absolute unities or objects of
knowledge, though actually relative, is his eloquent and fairly extensive treatment
of language. He attempts to explain how a name or term or linguistic sign some-
how fixes a concept within our consciousness as though it expresses an absolute
object of knowledge and not a process or relational bundle of attributes brought
together by some contributing act or relationship of agency by means of which the
thinking subject participates with or conditions the phenomena of consciousness
to make its otherwise chaotic or confused and unfixed plenitude meaningful.
   His digression onto the relationship between language and thought further deep-
ens the relativism of his whole approach as he brings to the fore the reciprocal
nature of language and thought: ‘Considered in general, thought and language
are reciprocally dependent; each bears all the imperfections and perfections of the
other; but without language there could be no knowledge realised of the essential
properties of things, and of the connection of their accidental states’ (LL.I.137).
He prioritises thought over speech or language but does so in such a way as to
suggest that this priority really pertains to the origination of the phenomenon of
language rather than being a necessary condition of all speech (see LL.I.138). Be
that as it may, Hamilton’s more important concern has to do with describing the
reciprocal relationship between thought and language, the process of conception
and the claim that, were it not for our ability to fix ‘and ratify in a verbal sign’ all
of the constituents of a concept, it would otherwise ‘fall back into the confusion
and infinitude from which it has been called out’ (LL.I.137). He illustrates this
for his students with some nice metaphors, such as the following:
      You have all heard of the process of tunnelling, of tunnelling through
      a sand bank. In this operation it is impossible to succeed, unless every
      foot, nay almost every inch in our progress, be secured by an arch of
      masonry, before we attempt the excavation of another. Now, language
      is to the mind precisely what the arch is to the tunnel. The power of
      thinking and the power of excavation are not dependent on the word
      in the one case, on the mason-work in the other; but without these
      subsidiaries, neither process could be carried on beyond its rudimentary
      commencement (LL.I.139).
114                                    Ralph Jessop
Hamilton elaborates upon the metaphor of tunnelling through a sandbank, but
while his simple claim may be altogether unexceptionable (that without language
thought would at best remain in the most elementary and fragile state of almost
total impermanence), it is interesting to note that what his tunnelling metaphor
most strongly indicates is the relationship that the thinking subject is caught
within, a relationship between the ever-shifting sand of a multitudinous plenum
and the crafting of signs that, unlike the relative nature of concepts with regard to
their constituent parts, may seem to render in some sense or to some degree per-
manent for consciousness the collected attributes of any given concept, but which
still leaves such concepts relative and factitiously unitary. But as the thinking
subject’s struggle to secure or make permanent through language the confusion
and infinitude of the phenomena of consciousness is here highlighted, Hamilton’s
awareness of Logic’s contest with a disintegrative atomism or absolute relativism,
at once suggests his consciousness of the volitional nature of reason and the abyss
of a more thoroughgoing relativism that threatens to undermine the warrantabil-
ity of his project entirely. But though this exaggerates the danger of Hamilton’s
relativism with regard to his own project, it draws to our attention the acute sense
that Hamilton has of logic’s relationship through language to the surrounding and
teeming chaos of the universe within which our intellects struggle to achieve order
— logic, like language, is for Hamilton a process of inching forward, of a tunnelling
through sand only made possible by constructing arches to hold back and make
orderly the chaos that perpetually threatens to engulf us. Any elementary con-
cepts we might have the capacity to form without the assistance of language, which
Hamilton admits may be a possibility, would be ‘but sparks which would twinkle
only to expire’, implying that without language, and we must add, without logic,
whatever thought might be possible is barely worth considering (LL.I.139). It is
therefore of the greatest moment for Hamilton that the logic we construct should
be robust and built upon proper foundations established through the most exact-
ing scrutiny of the work of others — this is not just a task for those who profess
to be logicians; it is also a task involving sound architectural skills, considerable
scholarship, the craftsmanship of the master builder, and a critical engagement
with and demolition of the crumbling and imperfect buildings of the past.
   But it all must be carried out with an acute consciousness of the materials one
has to work with and the instability of the substance that only logic can hold in
place. Hamilton has, more or less wittingly, but nonetheless in a most profound
way, highlighted the imperfect, inchoate, factitious, anthropocentric, volitional,
indeterminate, and inherently relative nature of concepts. However, in doing so
he has overburdened the tenability of the fixative term or word or sign to such
an extent that the permanence or ratification he seems to claim we are capable
of establishing or ascribing to concepts — their ‘acquired permanence’ as he later
describes it (LL.I.225) — begins to look questionable. That this should be the case
— that Hamilton’s conceptualism is in fact a dismantling of Logic’s object-matter
into a purely formal kind of object that describes a process only artificially or
analogically rendered as a material object consisting of identifiable components —
                           The Logic of Sir William Hamilton                        115
should perhaps not greatly surprise us, given that this idealism or immaterialism
is so clearly congruent with his doctrine of nescience, and several other aspects of
his philosophical position related to this doctrine.
   But this is not to say that Hamilton actually undercuts the whole point of
examining concepts and attempting to evolve a more complete science of the nec-
essary laws of thought. Rather, it is to his credit that, as he persists with his
enquiry, though the relativism and in general the perilously fragile condition of
concepts (as factitiously wholes, and indeed thereby factitiously permanent fixtures
in consciousness), is brought to his students’ attention, he effectually engages in a
quasi-Kantian critique of reason that at once points up the queerness, or idealism,
of Logic’s object-matter, while yet exploring what can be made determinate and
placed under the regulation of irrefragable laws with regard to thought, an entity
that he brings before us with great authenticity not as something already fixed or
identical or closely analogous to material entities and physical nature (or to con-
ceptions of physical nature that construe the material mechanistically), but rather
as an object-matter evincing an indeterminacy with which it behoves exacting
logical scrutiny to engage.
   Having pointed up the indeterminate and relative nature of concepts he goes on
to discuss the three main relations of concepts, namely, the relation they hold to
their objects, to their subject, and to each other. The first relation, to their objects,
is of course encapsulated by the term ‘quantity’ since all concepts are said to consist
of a greater or lesser number of attributes (the objects of a concept). However,
as is now well known but, according to Hamilton, had been largely overlooked
by many contemporaneous and earlier logicians, the quantity of a concept can be
distinguished into two different kinds, denominated by the terms ‘extension’ and
‘intension’ (or Hamilton’s more frequently used term ‘comprehension’). Although
the terms ‘extension’ and ‘intension’ are well known to present-day logicians, that
Hamilton regarded his contemporaries as being largely ignorant of these terms and
that their importance to his own attempts to improve traditional logic is so great,
provides at least two good reasons for elucidating his treatment of these terms
here.
   He claims that the distinction between extension and intension ‘forms the very
cardinal point on which the whole theory of Logic turns’ (LL.I.119). He buttresses
this claim by repeatedly returning to the significance of his distinction between and
treatment of extension and intension in several later lectures, for example, when he
argues that propositions can be distinguished as intensive or extensive depending
on whether the subject or the predicate is respectively the containing whole (see,
LL.I.231-3). However, as we shall see, the way in which he handles this distinction
further deepens his underlying notion concerning the relativity of concepts to show,
by demonstrating how extension and intension are correlatives of one another, that
a relativistic analysis of concepts and arguments is possible and indeed further
evolves the science of Pure Logic. However, importantly, the relativistic analysis
that Hamilton’s coordination of extension and intension enables, is one that is
nonetheless anchored in the fundamental rule of containment, namely, the axiom
116                                    Ralph Jessop
which ‘constitutes the one principle of all Deductive reasoning’, ‘that the part of a
part is a part of the whole’ (LL.I.119, 145, 144). It is perhaps worth pointing out
that this axiom is also given in two Latin phrases by Hamilton, one of which is:
‘Prœdicatum prœdicati est prœdicatum subjecti’. The Editor of his lectures points
out in a footnote that this is ‘A translation of Aristotle’s first antipredicamental
rule as given in the Categories (see LL.I.144).
   Hamilton’s thorough and fairly extended explication of extension and intension
in his Lecture VIII may be summarised as follows: a concept is a thought that
embraces or, in the sense explained earlier, brings into unity an indefinite plurality
of characters and it is also applicable to an indefinite plurality of objects about
which it may be said, or through which these objects may be thought. As such,
a concept is a quantity of two different and opposed kinds, denoted by the terms
‘intension’ and ‘extension’ (LL.I.140-52).
   ‘Extension’ refers to the external quantity of a concept, being determined by
the number of objects — concepts or realities — to which the concept may be
applied or which it classifies and hence under which these entities are said to be
contained (a concept’s extensive quantity is comprised of the number of objects
that can be thought mediately through the concept). The extensive quantity of a
concept is also referred to as its sphere or breadth and ‘the parts which the total
concept contains, are said to be contained under it, because, holding the relation
to it of the particular to the general, they are subordinated or ranged under it.
For example, the concepts man, horse, dog, &c., are contained under the more
general concept animal’ (LL.I.145). When these parts of a concept’s extension are
exposed, this is called Division.
   ‘Intension’ refers to the internal quantity of a concept, being determined by the
number of objects — concepts or realities — that constitute the concept and hence
in which these entities are said to be contained (a concept’s intensive quantity
is the conceived sum of the attributes that constitute it, formed into a whole
or unity in thought). The intensive quantity of a concept is also referred to as
its comprehension or depth and ‘the parts [. . . ] which go to constitute the total
concept, are said to be contained in it. For example, the concept man is composed
of two constituent parts or attributes, that is, of two partial concepts,–rational and
animal ; for the characters rational and animal are only an analytical expression
of the synthetic unity of the concept man’ (LL.I.143-4). When these parts or
characters of a concept’s intension are exposed, this is called Definition.
   According to Hamilton, logicians ‘have exclusively developed’ the extensive
quantity of concepts. However, he asserts that the extensive and intensive quanti-
ties comprise ‘the two great branches of reasoning’ and that intension ‘is at least
of equal importance’ in comparison with extension (LL.I.144-5). This claim is sig-
nificant in that, placing intension and extension on an equal footing, as two main
branches of reasoning, immediately brings the analysis of concepts under, as it
were, dual aspects or two main perspectives from which concepts may be viewed,
analysed, and through which Hamilton can further elaborate his thesis concerning
the relativity of concepts by bringing any given concept’s extensive and intensive
                            The Logic of Sir William Hamilton                      117
quantities into relation with one another — which is precisely what he does at
this stage in his Lectures as also in Discussions and his ‘New Analytic of Logical
Forms’.
   Hamilton seems to suggest at this stage what he will later make much more ex-
plicit, that the axiom or fundamental and sole principle of all Deductive reasoning,
that a part of a part is a part of the whole, originates with regard to the intensive
quantity of concepts — he certainly introduces this axiom of containment by il-
lustrating how the intensive quantity of a given concept, such as is signified in the
term ‘man’, may be analysed into ever diminishing parts contained in, or implicit
within, the concept ‘man’, ‘till we reach attributes which, as simple, stand as a
primary or ultimate element, into which the series can be resolved’ (LL.I.144). If
he is suggesting that the axiom of containment has been, as it were, translated
for application to the extensive quantity by the notion expressed as ‘whatever is
contained under the partial or more particular concept is contained under the total
or more general concept’, it is unclear whether he is at all troubled by the idea
that the necessity of relation implicit in the axiom of containment, particularly as
expressed with regard to a concept’s intension, is grounded on an analogy with
physical containment that he elsewhere rejects as unwarranted — that is, with re-
gard to the assumption that mind and body are analogically related (LL.I.145).35
However, he has effectually described concepts as necessarily factitious or artificial
constructions, given a kind of sufficient or provisionally adequate permanence by
means of language. He has also described concepts as relative continua in thought,
the analysis of which is itself, purposive or tendentiously conducted in order that
we can, as I indicated earlier, both speak about and thereby better comprehend
concepts (see LL.I.133). Thus, Hamilton can perhaps claim some licence in deploy-
ing terms such as ‘contain’ without issuing caveats to guard against some of the
assumptions antithetical to his general philosophical standpoint of natural dualism
that, as a term analogically related to physical containment, ‘contain’ itself may be
said to contain or imply. However, Hamilton is relying upon a traditional technical
language of containment to elucidate the intensive and extensive quantities which
he will later claim is better replaced by the more accurate ‘substantive verb, (is,
is not)’ to express the equation or affirmation or negation of identity between a
given concept and the objects constitutive of it or to which it may be related (or
through which other concepts or particulars may be thought) (LL.I.154). As we
shall see later, he does regard the axiom or sole principle of all Deductive reasoning
(that a part of a part is a part of the whole), as rather crucially originating in
a thought unshakeably natural to us, namely, our knowledge of the quantity of
intension and the ways in which intensive containment provides as it were a natu-
ral grounding for all purely logical inference — Hamilton, strenuously striving to
craft Logic into a pure science, abstracted from all extra-logical matters, cannot
resist bringing the laws of thought into an intimate relation with the natural, the
human as irrevocably an interrelated whole consisting of the self and the not-self.
   But this aside, he applies the axiom of containment to both the intensive and
  35 Discussions, pp. 61-2.
118                                     Ralph Jessop
the extensive quantities of concepts. Both may be said to contain attributes, but
they crucially differ in that, while the intensive quantity is said to be contained in
a concept, the extensive is said to be contained under it. Thus, while both of the
quantities of extension and intension contain in differing senses, and while both
are determined by wholes that may be identified and thereby quantified, ‘These
two quantities are not convertible. On the contrary, they are in the inverse ratio of
each other; the greater the depth of comprehension [intension] of a notion the less
its breadth or extension, and vice versa’ (LL.I.142). This inverse ratio relationship
between extension and intension is well illustrated by Hamilton as follows:
       When I take out of a concept, that is, abstract from one or more of
       its attributes, I diminish its comprehension [intension]. Thus, when
       from the concept man, equivalent to rational animal, I abstract from
       the attribute or determination rational, I lessen its internal quantity.
       But by this diminution of its comprehension I give it a wider exten-
       sion, for what remains is the concept animal, and the concept animal
       embraces under it a far greater number of objects than the concept
       man (LL.I.147).
Hamilton explains the way in which this inverse ratio operates in fuller detail,
pointing up as he does so that as we continue the analytic process outlined in
the above quotation, the diminishment of a concept’s intensive quantity, with the
correspondent amplification or expansion of its extension, must finally result in
‘that concept which all comprehension and all extension must equally contain, but
in which comprehension is at its minimum, extension at its maximum,–I mean
the concept of Being or Existence’ (LL.I.149). And by contrast with this, when a
concept’s intension is at its maximum, its extension being then at its minimum,
the concept we must end up with is that of an individual, ‘the concept being
a complement of the whole attributes of an individual object, which, by these
attributes, it thinks and discriminates from every other’ (LL.I.148).
   His notion that, when extension is at a maximum and thereby intension at a
minimum we must end with existence, raises the interesting topic of the existential
import of concepts and propositions. However, leaving this aside, it should now be
clear that, having described concepts as generally relative, Hamilton is laying bare
the sort of things that this relativity enables with regard to how we may view and
analyse concepts. In an ingenious table in his Discussions, which the editors of the
Lectures appended to the end of Lecture VIII, Hamilton further exposes how their
relational or relative nature enables us to view any given part of a concept under
different relational aspects, for example: as species of a genus from one end, the
genus of a species from its opposite end; how the expansion of a concept’s extension
implies a coordinate reduction in its intension; the sense in which we may say that
each part ‘in opposite respects, contains and is contained’ (LL.I.153); and how ‘the
real identity and rational differences of Breadth and Depth’ become exposed, such
that it becomes more apparent that extension and intension ‘though denominated
quantities, are, in reality, one and the same quantity, viewed in counter relations
                            The Logic of Sir William Hamilton                       119
and from opposite ends. Nothing is the one, which is not, pro tanto, the other’
(LL.I.153).
   When explicating these points to do with a concept’s extension and intension in
his Lectures Hamilton used ‘a circular machine’ which he had devised. Similar to a
now-familiar child’s toy consisting of a circular base with a wooden rod at its centre
onto which are placed differently coloured circular discs of various circumferences,
the discs of Hamilton’s machine represented concepts that could be placed on the
central rod representing the individual to which these concepts might refer. Each
disc was placed on the rod in the correct order to illustrate several things to do
with the relation of extension to intension, the completed device standing as an
inverted cone. This ‘machine’ is represented and discussed by Veitch (see Veitch,
pp. 250-2). As though trying to reinforce Hamilton’s claim that his doctrine of
the quantification of the predicate had been taught in his Lectures prior to de
Morgan’s quantification, the Editors’ insertion of the table and the explanatory
pages from his Discussions at the end of Lecture VIII reveal how his distinction
between extension (Breadth) and intension (Depth), and the way in which he
brings these two quantities into relation with one another, form the basis of his
quantification of the predicate. That Hamilton’s editors thought it necessary to do
this, strongly suggests their awareness of how easy it might be not to grasp from
his lectures, the extent to which Hamilton’s treatment of Concepts and the two
principal quantities of extension and intension lead to and are indeed an integral
part of Hamilton’s quantification of the predicate. However, the editors’ insertion
of this table is perhaps best understood as an attempt to bring to the fore the
very sort of thing that it may be assumed Hamilton explained at this stage in his
Lecture VIII by use of his ‘circular machine’. I have reproduced Hamilton’s table
below.
   The table displays the lines of Breadth and Depth, the conceptual range as
Ideal, the objective or particular range of individuals or singulars as Real, and gives
lines of direction to indicate affirmation and negation (which I shall not attempt
to explain). The table distinguishes between individuals or singulars (z, z , z ) and
classes (A, E, I, O, U ). The highest genus or widest attribute is given as A, A, A,
etc., the subaltern genera and species as E, I, O, U , and the lowest species or
narrowest attribute as Y . Each class is represented as ‘a series of resemblances
thought as one’, symbolised by the same letter to denote that they are thought of
as one, though really distinct or differing from one another in some respect(s) (this
being intimated by the vertical lines separating each letter). This is in line with
Hamilton’s notion that a concept is relative in the sense that it brings into a unity
in thought what are really discrete though resembling characters. The narrowest
attribute (Y ) is shown as a simple term constituted by the individuals z, z , z .
Though simple or singular in the sense that it has no extension or has a minimal
extension, it is dichotomised by using a thick line ‘|’ to denote ‘not’ — that is to
say, if ‘Y ’ is thinkable, as a strict logical necessity one must also be able to think
‘not Y ’, but the narrowest attribute must otherwise be a singular attribute and
not a class (as in, say, U ). This is not to say that some given attribute represented
120                                      Ralph Jessop
                         Table 1. Schemes of Two Quantitites
    (LL.I.152; for further detail see, pp. 152-6, and Discussions, pp. 699-701).
by ‘Y ’ could not be translated into ‘A’ in another table since any ‘Y ’ is only the
least or narrowest attribute in relation to U and the other letters above this, and
the least must be in thought, if not in fact, a unity or attribute directly relatable
to the individual constituents as shown beneath it. In other words, while the
table posits a narrowest attribute Y , as something not consisting of any other
resembling Y , closer examination of Y ’s intension could involve translating it into
another table of the same form in which it might be translated into A. As the
thick line ‘|’ denotes ‘not’ (e.g. to contradistinguish between say, ‘animal’ and ‘not
animal’, the thinner lines ‘merely discriminate one animal (A) , from another (A)’
(LL.I.156). Hamilton’s elucidation of the table more clearly begins to indicate just
how his understanding of the relativity of concepts, and what a concept consists in
with regard to the two quantities of extension and intension, combine to illustrate
the necessity of quantifying the predicate: ‘A is only A, not A, A, A, &c.; some
Animal is not some Animal; one class of Animals is not all, every, or any other;
this Animal is not that; Socrates is not Plato; z is not z . On the other hand, E
is EA; and Y is Y U OIEA; every lower and higher letter in the series coalescing
uninterruptedly into a series of reciprocal subjects and predicates, as shown by
the absence of all discriminating lines. Thus Socrates (z ), is Athenian (Y ), Greek
(U ), European (O), Man (I), Mammal (E), Animal (A)’ (LL.I.155).
   This needs to be further elucidated as follows: reading from left to right, the
                           The Logic of Sir William Hamilton                        121
first A contains under it the concepts EIOU Y , and the full range of all individual
and hence actual Athenians, represented by ‘z, z , z .’ However, the second A
contains under it EIOU , and not Y (‘|Y ’), and hence neither z nor z nor z . The
first A is a term identical qua term to all of the other As in the row of As, and
it resembles (according to the thinking subject) the other As to such an extent as
to be unified in thought and fixed as a unity by means of the single classificatory
or general term ‘A’. However, just as the first A and the second A are in all
respects identical, except that the first A contains under it Y (and hence, in this
case, z ) whereas the second A does not, similarly each of the other As may be
differentiated from one another and from the first A by means of reference to their
respective depths or intensive quantities. This implies that each of the terms A
is, as Hamilton puts it, ‘only A, not A, A, A, &c. [. . . ] one class of Animals is
not all’ (LL.I.155). This suggests the basis of quantification of the subject term in
a simple proposition such as ‘Some Animals are Mammals’. But, as we shall see
shortly, it also suggests the basis of quantifying the predicate term.
   Suppose that: A represents the term ‘Animal’; E represents the subordinate
or species ‘Mammal’ to A as the genus Animal; I represents the subordinate or
species ‘Man’ to E as the genus Mammal. For ease of explanation it is necessary
to follow Hamilton and ignore a reasoning that starts by quantifying the class
with the universal term ‘All’, for it can easily be seen that were we to commence,
in Breadth, with ‘All A’ (‘all Animal’) then, contained under this class are ‘All
E and some not E’ (‘all Mammal and some not Mammal’), which unnecessarily
complicates explication of how Hamilton’s table demonstrates the need to quantify
the predicate. Now, if we only attend to the range encompassed by the first five
As, then this implies we must only be looking at some A and not all (in this case
the first five). What is contained under these five As is the range of all Es (i.e. ‘not
E’ (‘|E’) has been excluded by abstracting or ignoring the sixth A). Translating
this selection of the first five As and the range of Es listed under them, we have
‘some Animals are all Mammals.’ This process can of course be continued by
repeating the process of removing one attribute or class of E, so that the next
step in the reasoning process that examines what is contained under E, involves
a further abstraction or removal of, in this case, the 4th E, resulting in: ‘Some E
(Mammals) are all I (Man)’, and so on.
   If this begins to illustrate how reducing the extension of a concept deepens or
expands its Depth, and vice versa — the inverse ratio principle Hamilton claims to
exist between extension and intension — the table also illustrates that, in effect,
the same reasoning can be applied whether we are arguing in Depth (i.e. with
regard to the intensive quantity) or in Breadth (i.e. with regard to the extensive
quantity). This, according to Hamilton, is because, ‘Though different in the order
of thought, (ratione), the two quantities are identical in the nature of things,
(re). Each supposes the other; and Breadth is not more to be distinguished from
Depth, than the relations of the sides, from the relations of the angles, of a triangle’
(LL.I.154). The table illustrates how the same reasoning can be applied to both
the extension and intension of a concept, so long as both the subject term and
122                                      Ralph Jessop
the predicate term in all of the resulting propositions are quantified consistently
in accordance with the correct understanding of what the table indicates at each
step in the reasoning with regard to containment (in or under). Hamilton’s own
illustration of this is as follows:
        In effect it is precisely the same reasoning, whether we argue in Depth
        — “z is, (i.e. as subject, contains in it the inherent attribute), some Y ;
        all Y is some U ; all U is some O; all O is some I; all I is some E; all E
        is some A — therefore, z is some A:” or whether we argue in Breadth
        — “Some A is, (i.e. as class, contains under it the subject part), all E;
        some E is all I; some I is all O; some O is all U ; some U is all Y ; some
        Y is z — therefore, some A is z .” The two reasonings, internally
        identical, are externally the converse of each other; the premise and
        term, which in Breadth is major, in Depth is minor. (LL.I.154)
But, as this displays how the inverse ratio principle operates, and how so relating
extension and intension reveals that it is only externally that the two quantities
determine a difference in reasoning whereas, with each term consistently quantified
using ‘some’ or ‘all’ as appropriate, subjects and predicates can swap position to
yield formally identical conclusions (‘z is some A’ and ‘some A is z ’), Hamilton
is quick to point out in this more developed treatment of the topic explicating
his ‘Schemes of the Two Quantities’ table that: ‘In syllogisms also, where the
contrast of the two quantities [extension and intension] is abolished, there, with
the difference of figure, the differences of major and minor premise and term fall
likewise.’ (LL.I.154). With these few words Hamilton is touching on just how,
through examining concepts extensively and intensively to show the inverse ratio
relationship between these two quantities, the resultant quantification of all terms
in a reasoning that hinges on the identification of the subject and predicate terms
of a proposition, enables a significant simplification of traditional Formal Logic.
However, he is also touching on a claim, the significance of which he brings to
the fore much later in his Lectures, namely, that ‘In fact, the two quantities and
the two quantifications have by logicians been neglected together ’ (LL.I.155). As
attending to the quantities of extension and intension reveals their ‘real identity
and rational differences’, such that it becomes more apparent that they are ‘in
reality, one and the same quantity’, Hamilton will later show, through careful
stages of his teaching of logic in the Lectures, that this identification of the two
quantities is highly significant with regard to both propositions and syllogisms
(LL.I.153; compare Discussions, pp. 701–2). However, before the full relevance of
Hamilton’s inverse ratio principle with regard to extension and intension can be
illuminated with regard to propositions and syllogisms, it is necessary to outline
some of the other points he makes in his Lectures concerning concepts.
    If by the end of Lecture VIII it is beginning to emerge, particularly with the ed-
itors’ helpful insertion of Hamilton’s ‘Schemes of the Two Quantities’ table given
above, that Hamilton is carefully working towards his quantification of the predi-
cate, in some of the Lectures that follow, he still has much to say that will further
                           The Logic of Sir William Hamilton                        123
reinforce his notion that concepts are relative, not simply in the sense that they
are relational (or wholes constituted by and referring to a plurality of resembling
entities, or relative with regard to their objects), but also in the sense that our
construction and grasp of the depth (intension) and breadth (extension) of any
given concept is dependent upon a subjective interrelation of clearness/obscurity
and distinctness/indistinctness. This introduces the notion of the quality of a
concept (see LL.I.157).
   Thus, having elucidated the inverse ratio relation between the extension and
intension of a concept, he goes on to consider the subjective relation of concepts (i.e.
the relation to the subject that thinks a concept) in relation to their clearness and
distinctness as similarly relative terms determining the quality of a given concept.
He explains the terms ‘clearness’ and ‘distinctness’ at some length, acknowledging
as he does so a considerable debt to Leibniz (see LL.I.159–65).
   Hamilton points out that while intensive distinctness is at a maximum when we
reach simple notions that are thereby indefinable, and while extensive distinctness
is at a maximum when, to quote his translation of Esser, ‘we touch on notions
which, as individual, admit of no ulterior division’, such distinctness is only ideal
and that in fact this ideal distinctness is something we are always approaching
but never in reality attaining (LL.I.170). As this ideal distinctness is regarded
by Esser as an incentive to re-analyse the intension and extension of concepts,
it is clear that Hamilton is suggesting to his students that this relativity and
incompleteness or non-absolute condition of concepts is an important aspect of his
overall approach to the study of Logic, a crucial incorporation of an awareness of
the ultimate indefinability and non-absolute dimension of Logic’s object matter.
And in this Hamilton is arguably being entirely consistent with other aspects of
his metaphysics and overall philosophical position.
   In Lecture X he has more to say about the imperfection of concepts, this time
returning to the problem of language. Concepts have, as it were, a propensity to be
obscure and indistinct and these vices are due, partly to their very nature as wholes
that bind together ‘a multiplicity in unity’, and partly from their dependence upon
language as that which fixes concepts in consciousness (LL.I.172). He explains the
problem of language by means of an illustrative analogy with methods of exchange
in countries lacking an established currency. Thus, language operates much like
the handing over of unquantified bags of precious metals which may or may not be
closely scrutinised to see if they yield the value they purportedly signify — on most
occasions the language user takes on trust that a particular term binds together
what it seems to claim for itself, namely, that it does in fact represent a multitude
of entities collectively amounting to a certain sum or value; but at other times,
this will not be the case. This analogy of course teems with significance but there
seem to be two main points that Hamilton is attempting to emphasise: firstly,
‘that notions or concepts are peculiarly liable to great vagueness and ambiguity,
and that their symbols are liable to be passed about without the proper kind, or
the adequate amount, of thought’ (LL.I.173-4); and secondly, that an important
distinction, originated by Leibniz, can be made with regard to our knowledge that
124                                   Ralph Jessop
divides cognition into the blind or symbolical and intuitive (see LL.I.180-86). In
short, Leibniz’s notion of symbolical knowledge refers to concepts as terms taken
to signify entities obscurely and imperfectly presented to the mind but which
may potentially be exposed or made explicit, though we cannot think all of the
ingredients that comprise the symbol or term used to refer to them. By contrast
with this, intuitive knowledge is of these ingredients themselves inasmuch as this
is possible. Hamilton fails to explain the significance of this distinction except to
claim that thereby Leibniz and his followers in Germany superseded or overcame
‘the whole controversy of Nominalism and Conceptualism,–which, in consequence
of the non-establishment of this distinction, and the relative imperfection of our
philosophical language, has idly agitated the Psychology of this country [Britain]
and of France’ (LL.I.179). However, it should be fairly obvious that by means of
this distinction between symbolical and intuitive knowledge, Hamilton is providing
a warrant for treating concepts in a purely formal manner in order to examine more
closely the relations between concepts considered blindly or symbolically, while
at the same time drawing attention to the interface between what the symbols
represent with regard to the imperfect but real condition of our intuitive knowledge
of the particulars to which concepts relate, and through which we both think these
particulars and constitute our concepts. Once again, the relativism, imperfection,
inchoateness, and mutability of concepts and the phenomenal and plural nature of
our knowledge as relative, relational, provisional, and so on, is being emphasised
by Hamilton, while at the same time he attempts to establish a domain or object
matter of Logic at once stable, constricted, and discriminated from the material
or actual, though both domains of the intuitive and symbolical are held in relation
to one another as mutually informative and only theoretically discrete.
   Though Hamilton is clearly aware that his various points concerning the rel-
ativity of concepts is foregrounding matter that might easily be thought of as
extra-logical, having distinguished between symbolical and intuitive knowledge, in
Lecture XI he appropriately turns to what he regards as ‘The Relation proper’
of concepts, namely, their relation to each other — something that can be repre-
sented symbolically and diagrammatically and which thereby establishes a set of
relationships familiar to logicians. Again he discusses this in relation to the two
principal quantities of extension and intension. Taking extension first, he outlines
five principal relations: Exclusion, Coextension, Subordination, Co-ordination,
and Intersection. All of these he illustrates with simple circle diagrams, a prac-
tice used by some earlier philosophers, such as, according to Hamilton, the late
16th century Christian Wiese, and of course developed later by John Venn. It is
needless to reproduce Hamilton’s diagrams (see LL.I.189; 256), but it is important
to note that of all of these relations between concepts considered with regard to
their extension, ‘those of Subordination and Co-ordination are of principal impor-
tance, as on them reposes the whole system of classification’ (LL.I.189-90). He
elucidates Subordination and Co-ordination with crystal clarity. However, as yet
further evidence of his interest in and even fascination with relativism, he contin-
ues to draw into consideration: that there is no absolute exclusion in the relation
                            The Logic of Sir William Hamilton                     125
between concepts known as Exclusion (LL.I.188); the ways in which our perspec-
tive on concepts results in re-describing a genus as a species and a species as a
genus; how speculatively, if not practically, we may always divide a concept ad
infinitum (LL.I.192-3); how different abstractions result in different relations of
genus and species in both subordination and co-ordination; and how, admitting
that there can be both a highest genus and a lowest species, neither of which
are convertible into a species nor a genus respectively, within Subordination there
are gradations of genus and species known as subalterns or intermediates such
that, the genus lower than the highest genus, can become a species, whereas the
species higher than the lowest species, can become a genus. This, as Hamilton
points out all comes from Porphyry’s Introduction to Aristotle’s Categories, but it
is nonetheless notable that Hamilton should emphasise the non-absolute aspects
of the relations between concepts and the dependence of these various reversals
of species and genus on how we regard them (LL.I.196). Point of view is indeed
something he is at pains to emphasise as important to several matters, not least
of which is the relation of whole to part and the distinction, which he regards
as erroneous, between Logical and Metaphysical wholes, these being, contra to
previous logicians, ‘equally logical’ (see LL.I.201-2).
   Hamilton is evidently working towards some important claims to do with the
whole-part relationship that will give priority to the intensive quantity of concepts
which in part relies upon his notion of Involution (see LL.I.202-3). Skipping over
the various kinds of whole that Hamilton elucidates, the notion that a genus con-
tains its species either potentially or actually (see LL.I.205-6), along with various
other points of interest, it is in Hamilton’s treatment of Comprehension or the
quantity of intension in Lecture XII that he first makes fully explicit one of his
major disagreements with traditional logic, a disagreement nevertheless that he
has been hinting at in one way or another from a fairly early stage. He claims that
the relations of Involution and Co-ordination have been:
      altogether neglected by logicians: and, in consequence of this, they
      have necessarily overlooked one of the two great divisions of all reason-
      ing [. . . ]. In each quantity there is a deductive, and in each quantity
      there is an inductive, inference; and if the reasoning under either of
      these two quantities were to be omitted, it ought, perhaps, to have
      been the one which the logicians have exclusively cultivated [i.e. the
      deductive reasoning in extension]. For the quantity of extension is a
      creation of the mind itself, and only created through, as abstracted
      from, the quantity of comprehension [intension]; whereas the quantity
      of comprehension is at once given in the very nature of things. The
      former quantity is thus secondary and factitious, the latter primary
      and natural. (LL.I.217-8).
Now, it must be noted that by the term ‘inductive’ Hamilton departs from what we
might call the standard treatment of or distinction between induction and deduc-
tion. Thus, he does not mean that kind of inference in which the conclusion may
126                                    Ralph Jessop
be said to exceed or amplify the content of the premises. He is not talking here
about probable reasoning or judgements that go beyond what is delivered in the
premises — his restriction of induction to a form of necessary reasoning is therefore
substantially different from the empirical reasoning or induction of the physical
sciences. Instead, as he makes clear in a later lecture, he is treating induction in a
formal sense as differentiated from an informal or material sense, to mean a process
of reasoning from all of the parts to the whole that these parts entirely constitute
(see LL.I.319-26). This aside, that he is now distinguishing as he does in the above
quotation between extension and intension may seem to be at odds with the claim
he made earlier about concepts being factitiously unities, but really pluralities —
now it would seem they are only partially factitious, projected, mind-dependent
and mind-generated, but partially ‘primary and natural’. Also, this prioritisation
of intension seems to sit uncomfortably with his attempt to hold, by means of
his inverse ratio relation between the two quantities, that extension and intension
are non-convertible but equivalent with regard to the reasoning that we apply to
both quantities. Leaving aside this weaker objection, that he has now introduced
a distinction between extension and intension that prioritises intension as the real
and primary, creates problems for Hamilton that he does not satisfactorily resolve.
For example, is not the intensive quantity also a binding together of resembling
but nonetheless discrete entities held in a relationship of containing and contained,
and thus is not intension every bit as factitious as extension? Furthermore, how
can an abstraction from the intensive quantity be satisfactorily described as fac-
titious, given that it is just that, an abstraction from what is not factitious but
rather ‘in the very nature of things’ ? Perhaps, by ‘factitious’ Hamilton is merely
asserting that, in relation to the thinking subject, extension is a product of the
mental action of abstracting from the real, intensive quantity, and in this sense
the product that is a concept’s extensive quantity may be more or less arbitrarily
or selectively constructed, that extension is both factitiously a unity, really a plu-
rality of resembling attributes, and that its factiousness in relation to the reality
of a given concept’s intension is a matter of degree. That is to say, both with
regard to the degree of resemblance between a concept’s attributes in extension,
and with regard to how close its extension is to its definition, its distinctness, or
the thinking subject’s grasp of its intensive quantity, the extensive quantity may
be more or less factitious.
   Be that as it may, Hamilton does seem to be making an important distinction
and claim here that may be more sympathetically understood as an attempt to
ground in reality the formation of our concepts as emanating from or evolving out
of a deeper or more intimately connected relationship of resemblance between the
parts of a whole than is tenable of the resembling attributes in extension. He does
this by pointing up the difference between the senses of ‘contain’: in extension
a concept contains under it, its various attributes in subordination; whereas in
intension, a concept contains in it, its various attributes as constitutive of and as
the definitional properties that make the concept what it is, just as the multitude
of parts go together to constitute any given individual. To distinguish between
                          The Logic of Sir William Hamilton                        127
these two different senses of ‘contain’, Hamilton introduces the notion of Involution
mentioned above, which he explains as follows:
      In the quantity of comprehension [intension], one notion is involved in
      another, when it forms a part of the sum total of characters, which
      together constitute the comprehension of that other; and two notions
      are in this quantity co-ordinated, when, whilst neither comprehends the
      other, both are immediately comprehended in the same lower concept.
      (LL.I.220).
He gives two illustrations of the Involution relationship between a concept’s at-
tributes, the second of these pointing up the inter-relatedness of concepts as in-
volving and involved, such that we may be said to think a certain concept only in
and through that of another concept:
      In this quantity [of intension] the involving notion or whole is the more
      complex notion; the involved notion or part is the more simple. Thus
      pigeon as comprehending bird, bird as comprehending feathered, feath-
      ered as comprehending warm-blooded, warm-blooded as comprehend-
      ing heart with four cavities, heart with four cavities as comprehending
      breathing with lungs, are severally to each other as notions involving
      and involved. (LL.I.223).
Suggesting as this does, a relative overlapping and even partial integration of one
concept with another, such that they mutually imply and yet collectively consti-
tute the individual or unity they define, he immediately follows this somewhat
sketchy account of Involution, by differentiating between this relativist aspect
(containing and contained) and the non-relative (since not necessarily containing
and contained) co-ordination:
      Again, notions, in the whole of comprehension, are co-ordinated, when
      they stand together as constituting parts of the notion in which they
      are both immediately comprehended. Thus the characters oviparous
      [egg producing] and warm-blooded, heart with four cavities, and breath-
      ing with lungs, as all immediately contributing to make up the com-
      prehension of the notion bird, are, in this respect, severally considered
      as its co-ordinate parts. These characters are not relative and co-
      relative,–not containing and contained. For we have oviparous animals
      which are not warm-blooded, and warm-blooded animals which are
      not oviparous. Again, it is true, I believe, that all warm-blooded ani-
      mals have hearts with four cavities [. . . ], and that all animals with such
      hearts breathe by lungs and not by gills, But then, in this case, we have
      no right to suppose that the first of these characters comprehends the
      second, and that the second comprehends the third. For we should be
      equally entitled to assert, that all animals breathing by lungs possessed
      hearts of four cavities, and that all animals with such hearts are warm-
      blooded. They are thus thought as mutually the conditions of each
128                                     Ralph Jessop
      other; and whilst we may not know their reciprocal dependence, they
      are, however, conceived by us, as on an equal footing of co-ordination.
      (LL.I.223)
Hamilton hints that the significance of Involution and Co-ordination will come to
the fore when he later moves on to tackle the syllogism and, seemingly regarding
this as virtually a digression, his explanation of Involution in Lecture XII is un-
fortunately rather brief and might have been more fully elucidated, particularly
given that it may be much more crucial than perhaps Hamilton realised, to how we
might best understand at least one major aspect to do with the co-identity rela-
tionship within universal propositions which I shall discuss in Section 4. However,
he says just about enough to suggest what one might call a grounding relativism in
which his few words on this, though offered to explain the wholeness of a concept’s
intensive quantity, verge on further pointing up the artificial nature of the funda-
mental laws of logic of identity, excluded middle, and non-contradiction. If this is
less evident in the above quotation, it is more so in how he relates the involution
within a concept’s intensive quantity to his non-linear notion of the formation or
evolution of general and particular terms:
      Our notions are originally evolved out of the more complex into the
      more simple, and [. . . ] the progress of science is nothing more than a
      progressive unfolding into distinct consciousness of the various elements
      comprehended in the characters, originally known to us in their vague
      or confused totality [the condition in which they may be said to be
      maximally complex].
      It is a famous question among philosophers,–Whether our knowledge
      commences with the general or with the individual,–whether children
      first employ common, or first employ proper, names. In this con-
      troversy, the reasoners have severally proved the opposite opinion to
      be untenable; but the question is at once solved, by showing that a
      third opinion is the true,–viz. that our knowledge commences with the
      confused and complex, which, as regarded in one point of view or in
      another, may easily be mistaken either for the individual, or for the
      general. [. . . ]. It is sufficient to say in general, that all objects are
      presented to us in complexity; that we are at first more struck with
      the points of resemblance than with the points of contrast; that the
      earliest notions, and consequently, the earliest terms, are those that
      correspond to this synthesis, while the notions and the terms arising
      from an analysis of this synthesis into its parts, are of a subsequent
      formation. But though it be foreign to the province of Logic to de-
      velop the history of this procedure; yet, as this procedure is natural to
      the human mind, Logic must contain the form by which it is regulated.
      It must not only enable us to reason from the simple and general to
      the complex and individual; it must likewise, enable us to reverse the
      process, and to reason from the complex and individual to the simple
                            The Logic of Sir William Hamilton                      129
       and the general. And this it does by that relation of notions as contain-
       ing and contained, given in the quantity of comprehension [intension].
       (LL.I.221-2).
In short, the notion of Involution takes us once again into the relativism of Hamil-
ton’s metaphysical epistemology and theory of language which he outlines more
extensively in his Lectures on Metaphysics.36 And, if he is, perhaps rather shak-
ily, attempting to give priority to the reality of intension over and against the
factitiousness of extension, he at least does so in a way that coheres interestingly
with his theory of language with regard to the iterative or relative or non-linear
origination of general and particular terms as a natural process of thought and
indeed reasoning. It must be noted, however, that Hamilton is here attempting
to foreground not the relativity of concepts but rather, that concepts considered
intensively enable the natural tendencies in reasoning and thought generally ‘to
reason from the simple and general to the complex and individual [and also] re-
verse the process, and [. . . ] reason from the complex and individual to the simple
and the general’ — this, he clearly sees as a particular virtue of the intensive
quantity, namely, that, as contradistinguished from the extensive quantity, the
reversals in the direction of reasoning that we do in fact or naturally make, can
only legitimately be carried out with regard to the intensive quantity.
   Although the somewhat fuzzy logic of this notion of Involution within, or charac-
terising, the relation of the internal components of a given concept, or its intension
(as contradistinguished from Subordination in extension), is not handled as fully
as one might wish, it does seem that Hamilton is attempting to ground in an in-
determinate reality of chaos or confusion, concepts as originating in, wrested out
of, and partaking in a significant degree of relativism. However, I would claim
that, though fluxive, quasi-organic, and arguably suggesting an indivisible and
interminably iterative process, the involution relation that Hamilton seems to be
suggesting inheres within a concept’s intension, warrants, or is itself peculiarly
suggestive of, the distinction of constituent parts comprising wholes and the de-
fensibility and inherent reasonableness of regarding such parts as being contained
in the whole such that their relations one to the other may be understood as nec-
essary and as such fundamental to Pure Logic as the science of the necessary laws
of thought. The naturalness of this reasoning as grounded in the notion that the
parts of parts involve and are involved, and thereby permit reversing the process of
reasoning, is brought out or hinted at by Hamilton in several ways. For example,
just what he seems to think of as Involution may be being suggested in what he
says about partes integrantes with regard to Mathematical, or the Quantitative, or
Integrant Whole (see LL.I.204). Perhaps the naturalness and primacy Hamilton
is trying to claim of intension, as contrasted to the factitiousness of extension,
is being indicated in the natural involutions of physical entities as we encounter
them in, say, the human body, but also as may be postulated of all material phe-
nomena as the relative, plural, and confused constituents of consciousness which,
  36 See, Lectures on Metaphysics, II.319-27.
130                                     Ralph Jessop
conditioned by thought, we naturally or inevitably collect together (or grasp) as
resembling particulars. Organised into wholes, as part of the natural processes of
thought and our conditioning propensity to cognise, these wholes avail, and indeed
require, a corresponding analysis into their particulars and a relation of these to
the concept they constitute, our act of grasping or even perceiving them as col-
lected together into the unity of the concept, being the synthetic reconstruction
of the particulars constituting it, albeit to some degree provisionally. For Hamil-
ton, it is as though a power of the mind — the faculty of comparison – is itself
reinforced by and originates in the involution relationship as something that at
once suggests the greater chaos or confused complexity into which the involutions
of the parts seem to be capable of collapsing, while also suggesting the sharper
discrimination into non-involved particulars that are thereby not constitutive of
but which rather divide a concept into its extensive quantity. If so, Involution at
once suggests and is itself the suggestion of what a concept’s intension may be
evolved into by the native processes of conception and its negative corollary, the
reasoning of an analysis only prevented from disintegrating into infinite divisibility
by an aptly pragmatic limitation on the extent of this science’s explicitness, which
Hamilton explicitly imposes (see LL.I.192-3; 210).
   But in order to go at least some way towards understanding why Hamilton is
now claiming that the reasoning of intension is natural, or has a basis in reality
— and indeed in a reality that he regards as otherwise confused or chaotic and
theoretically if not practically open to infinite divisibility in its unconditioned state
as the raw and unorganised mere phenomena of consciousness — and hence why
he regards intension as prior or superior to the quantity of extension, we need to
note at least something of the extent to which he regards the history of logic since
Aristotle as flawed. We also need to note that Hamilton quite pointedly regards his
prioritisation of intension as a significant contribution towards placing the keystone
in the arch of the Aristotelian logic. He claims that logicians following Aristotle
rather surprisingly neglected the process of reasoning to do with intension, even
though they explicitly stated and relied upon the axiom: ‘The character of the
character is the character of the thing; or, The predicate of the predicate is the
predicate of the subject’ (LL.I.218). However, according to Hamilton, Aristotle
understood the application of this axiom:
      In fact I think it even possible to show in detail, that his whole analysis
      of the syllogism has reference to both quantities, and that the great
      abstruseness of his Prior Analytics, the treatise in which he develops
      the general forms of reasoning, arises from this,–that he has endeav-
      oured to rise to formulæ sufficiently general to express at once what
      was common to both kinds;–an attempt so far beyond the intelligence
      of subsequent logicians, that they have wholly misunderstood and per-
      verted his doctrine. They understood this doctrine, only as applied to
      the reasoning in extensive quantity; and in relation to this kind of rea-
      soning, they have certainly made palpable and easy what in Aristotle
      is abstract and difficult. But then they did not observe that Aristotle’s
                           The Logic of Sir William Hamilton                       131
       doctrine applies to two species, of which they only consider one. [. . . ].
       This mistake,–this partial conception of the science,–is common to all
       logicians, ancient and modern: for in so far as I am aware, no one has
       observed, that of the quantities of comprehension and extension, each
       affords a reasoning proper to itself; and no one has noticed that the
       doctrine of Aristotle has reference indifferently to both. (LL.I.218-9).
                         3 DOCTRINE OF JUDGMENTS
       In the proposition Men are animals, we should be allowed to determine
       whether the term men means all or some men,–whether the term ani-
       mals means all or some animals; in short, to quantify both the subject
       and predicate of the proposition. This postulate [‘To state explicitly
       what is thought implicitly’] applies both to Propositions and to Syllo-
       gisms. (NA.LL.II.252).
In elucidating some aspects of Hamilton’s ‘Schemes of the Two Quantities’ table
and certain features of his doctrine of Concepts in the previous section, I have
so far been indicating something of how his development and understanding of
extension and intension underpinned his quantification of the predicate. Before
examining the quantification itself, we need to look at a selection of some of the
other construction work.
   By ‘judgment’ Hamilton means ‘proposition’ or the forming of a proposition,
such as ‘water rusts iron’, by means of the process of judging (LL.I.227). What
is relative within a judgment or proposition has at least the appearance of being
much more stable than the relativity of concepts discussed at various stages above,
though the very fact that Hamilton calls attention to propositions as judgments,
keeps in focus how it is that a proposition may be regarded as at least related
to some mental agency, a judgment being the product of the act of judging (akin
to a concept being the product of conception). Introducing his students to the
standard terms of ‘subject’, ‘predicate’, and ‘copula’, he points out that the subject
is the determined or qualified notion, whereas the predicate is the determining or
qualifying notion, the relation of determination between them being signified by
the copula. He therefore defines a proposition as:
       the product of that act in which we pronounce, that, of two notions
       thought as subject and as predicate, the one does or does not consti-
       tute a part of the other, either in the quantity of extension, or in the
       quantity of Comprehension. (LL.I.229).
Hamilton is clearly regarding subject and predicate as being held in a relation
of quantity in accordance with the axiom that a part of a part is a part of the
whole, but he is also, consistent with his treatment of extension and intension,
incorporating both of these quantities, and this is how he does so:
132                                    Ralph Jessop
       The first great distinction of Judgments is taken from the relation of
       Subject and Predicate, as reciprocally whole and part. If the Subject
       or determined notion be viewed as the containing whole, we have an
       intensive or Comprehensive proposition; if the Predicate or determin-
       ing notion be viewed as the containing whole, we have an extensive
       proposition. (LL.I.231-2).
Hence, whether a proposition is extensive or intensive depends on how we view
the Subject and the Predicate. Of course Hamilton acknowledges that in single
propositions it is rarely clear which way the proposition is to be viewed, since
it is generally unclear whether it is the Subject or the Predicate that is being
regarded as the containing whole. However, according to Hamilton, ‘It is only
when propositions are connected together into syllogism, that it becomes evident
whether the subject or the predicate be the whole in or under which the other is
contained’ — he thus regards the distinction of propositions into extensive and
intensive as highly general, though of great importance, since ‘it is only in subor-
dination to this distinction that the other distinctions [he is about to introduce]
are valid’ (LL.I.233). The other distinctions that relate to the extensive/intensive
distinction with regard to the subject and predicate relation may be summed up
as: the distinction between Categorical (or simple propositions, in which, follow-
ing the practice of Aristotle’s followers, the predicate is either simply affirmed or
denied of the subject, as in ‘A is B’, or ‘A is not B’) and Conditional propositions,
Conditional propositions being further distinguished into Hypothetical (‘if A then
B’), Disjunctive (‘A or B’, or ‘D is either B, or C, or A’), and Dilemmatic or
Hypothetico-disjunctive (‘If X is A, it is either B or C’) (see LL.I.233-242).
   Although it is crucial to grasp that Hamilton is showing his students how in
principle the subject and predicate terms may be viewed as reciprocally related to
one another in any given proposition as expressions solely concerned with a whole-
part relationship, either in extension or intension, it is needless to elucidate this
any further here. More importantly, is Hamilton’s major point of difference with
Aristotle and ‘The doctrine of Logicians’ concerning the division of propositions
into four classes or species, since it is in this that Hamilton may be said to be
making his most explicit statement within his Lectures so far, concerning how
his system and much if not all of his previous discourse on logic has been working
towards and in turn will rely upon a thoroughgoing quantification of the predicate,
a quantification which notably traditional logic had failed to achieve due, not only
to exclusively focusing on extension, but also to the establishment of a class of
proposition that admitted vagueness or ambiguity with regard to the subject term’s
quantification and thereby disabled quantification by failing to make explicit the
quantity pertaining to the subject term which, according to Hamilton, is ‘involved
in every actual thought’ though at times not in its linguistic expression (LL.I.244).
But, as we shall see, for Hamilton, as this notion that quantity is either explicit or
implicit enables the removal of a class of propositions in which the subject term is
not quantified in expression, it also underpins the whole notion that the predicate
term’s quantity may also be made explicit.
                            The Logic of Sir William Hamilton                        133
     Following Aristotle, logicians traditionally divided propositions with regard to
their extensive quantity by categorising them as: Universal or General ; Particular ;
Individual or Singular ; and, Indefinite. According to Hamilton these terms were
applied with the following meanings: in Universal propositions ‘the subject is
taken in its whole extension’; in Particular propositions ‘the subject is taken in
a part, indefinitely, of its extension’; in Individual propositions ‘the subject is
at a minimum of extension’; and in Indefinite propositions, ‘the subject is not
[. . . ] overtly declared to be either universal, particular, or individual’ (LL.I.243).
With regard to quantification generally, Hamilton claims that commonly only the
Subject is regulated, whereas the predicate ‘Aristotle and the logicians do not
allow to be affected by quantity; at least they hold it to be always Particular
in an Affirmative, and Universal in a Negative’ (LL.I.244). However, he claims
that this doctrine is untenable, incomplete, that it resulted in confusion, and that
this confusion and incompleteness is partly due to logicians paying insufficient
heed to the fundamental principle of explicitness that I explained earlier. At this
stage in his lectures, Hamilton might have simply gone straight to quantifying
the predicate to show that, by making the quantification of the predicate term
explicit, it is no longer requisite that we consider whether the proposition is being
considered extensively or intensively. However, he is much more careful to take
his students through the evolution of his logic as it involves a critique and a
significant modification of traditional logic’s assumption that the predicate term
in an affirmative proposition is always Particular whereas in a negative proposition
it is always Universal.
     By contrast with the Aristotelian doctrine of the logicians, which divides propo-
sitions into the four classes or species of Universal, Particular, Individual, and
Indefinite, Hamilton proposes that they should be differentiated quite differently,
and he does this largely by redefining ‘indefinite’ and thereby eradicating as a
distinct class traditional logic’s Indefinite propositions, while retaining the notion
of indefiniteness as describing, within any given whole, an indeterminate range
of quantities sufficiently competent to be classed as the quantifier of particularity
we normally express by the term ‘some’ — in other words Hamilton’s ‘indefinite’
constitutes that species of judgments/propositions the logicians formerly called
‘Particular’.
     His redefinition of ‘indefinite’ seems to amount to this: for the logicians ‘indefi-
nite’ meant little more than that the extensive quantity of the subject (universal,
particular, or individual) was unexpressed and thus indefinite in the sense of being
unclear or inexplicit; by contrast with this, for Hamilton, ‘indefinite’ refers more
directly to the quantity expressed by the terms ‘some’, ‘many’, and various other
expressions so long as they designate ‘some indefinite number less than the whole’.
Hence, his definition of ‘indefinite’ refers to any quantity within a whole, ranging
from (possibly) a singular to a number less than the whole to which the indefinite
term refers (LL.I.246). It should perhaps be noted here that Hamilton’s definition
of ‘indefinite’ may not at first sight seem to differentiate itself sufficiently from the
quantity of Individual judgments or propositions, nor from Universal propositions.
134                                          Ralph Jessop
As Fogelin claims, Hamilton ‘held complicated views on the quantifier some’, in
which, while he sometimes referred to ‘some’ as meaning ‘some but not all’, he
also used the definition ‘some perhaps all’, and as Fogelin argues, there seem to
be good grounds for saying that Hamilton purposely incorporates both readings
of ‘some’ (Fogelin, 153).37
   These complicated views on the meaning of Particulars, as that class of proposi-
tions which are indefinite, only come to the fore in Hamilton’s ‘New Analytic’ (see
NA.LL.II.279-80). However, that by ‘some’ Hamilton meant ‘some but not all’,
and that he distinguishes between Individual and Particular Propositions does
seem to be fairly clear if we confine attention solely to Lecture XIII. Not only
does he give little or no indication in Lecture XIII that ‘some’ might mean ‘some
perhaps all’, but he also makes it very clear that a proposition may be expressed
using the indefinite article, and thereby be classed as a Particular proposition, as in
one of his examples, ‘An Englishman generalised the law of gravitation’; whereas,
when expressed using a proper name, it will be an Individual proposition, as in
‘Newton generalised the law of gravitation’ (LL.I.247). This, however, merely il-
lustrates what is required to transform an ordinary language statement from being
a Particular proposition to an Individual one and Hamilton’s explicitness principle
ought to allow that all such indefinite propositions, not involving a plurality, must
be expressible as individual or, where the ‘some’ really means ‘all’, as universal
propositions.
   Though Hamilton may not have squarely tackled certain problems to do with
his definition of the quantifier ‘some’, it is at least possible to detect in his virtual
eradication of the traditional logic’s class of Indefinite propositions, that he is
making the treatment of the quantity of propositions more internally consistent
by accommodating the indefinite within Particular propositions by removing the
obstruction to them being accommodated within Particular propositions, namely,
the mistaken condition of their failure to express whether they were universal,
particular, or individual — a failure that, Hamilton might have pointed out, was
entirely due to the logicians not grasping that, in an extensive proposition, where
the quantity of the subject is unexpressed it must either be a part (and hence
particular) of the predicate as whole; or, in an intensive proposition, its quantity
must be capable of being expressed as the containing whole determined by the
predicate as part of that whole. To re-state the quotation given earlier: ‘If the
Subject or determined notion be viewed as the containing whole, we have an
intensive or Comprehensive proposition; if the Predicate or determining notion be
viewed as the containing whole, we have an extensive proposition’ (LL.I.231-2).
By itself this will not imply what quantity ought to be appended to the subject
of a given proposition where that quantity remains unknown. However, this is
  37 Robert J. Fogelin, ‘Hamilton’s Quantification of the Predicate’ in, Philosophical Interpre-
tations (Oxford: Oxford University Press, 1992), first published in Philosophical Quarterly, 26
(1976), 149-65. All references to Fogelin are to this edition. Fogelin also supplies in this edition a
second article entitled ‘Hamilton’s Theory of Quantifying the Predicate — A Correction’ (166-8),
also first published in Philosophical Quarterly, 26.
                          The Logic of Sir William Hamilton                       135
irrelevant as the traditional class of Indefinite propositions regards the proposition
in extension alone, and as I have claimed in support of Hamilton, viewing the
traditional Indefinite proposition as extensive, determines that the subject term
must be particular in relation to the predicate term as the determining whole
containing the subject term. What Hamilton has achieved here is the removal of an
anomalous class of propositions that is only possible by means of his understanding
of the whole-part relationship in propositions viewed extensively and intensively.
However, although much more might be said about all this, there is another crucial
factor to removing the traditional class of problematic and confusing indefinite or
indeterminate propositions, namely, his introduction of the terms ‘predesignate’
and ‘preindesignate’.
   It is easy to see how the application of these terms to the traditional class of
Indefinite propositions warrants a discrimination between propositions that may
thereby be deemed to be either logically adequate or inadequate. When a propo-
sition (presumably considered as an external expression that may thus be more or
less precise, as contradistinguished from a judgment, construed here as the mental
process) articulates its quantity by prefixing terms such as ‘all’ and ‘some’, this
is a predesignate proposition; whereas when a proposition does not articulate its
quantity, the proposition is called by Hamilton, preindesignate. Hence, the subject
and predicate terms of a proposition may also be either predesignate or preindes-
ignate terms. But, for Hamilton, though in a proposition’s external expression
one or both of its terms may be preindesignate (the quantity often being ‘elided in
its expression’), the unexpressed quantity is always involved in thought (definite
or indefinite ‘quantity being involved in every actual thought’, though not always
marked by a quantifier) (NA.LL.II.250; LL.I.244). Hence, adhering to his princi-
ple of explicitness, such preindesignate terms may be translated into predesignate
ones — which is to say, that unexpressed quantities may always be expressed at
least in principle and indeed should be expressed in adherence to the fundamental
principle of explicitness. One might therefore say that, with regard to the tradi-
tional class of Indefinite propositions, if in fact it is impossible to determine the
predesignate term(s) (as quantified in thought) of a given statement purporting
to be a proposition but which is somehow quite indeterminate in its meaning by
means of its use of a preindesignate term(s) (hence unquantified in expression),
then such a statement must be deemed to be non-propositional and inadequate
for consideration within a reasoning or argument.
   The differences between propositions, with regard to quantity, according to
Hamilton, arise, on the one hand, ‘from the necessary condition of the Internal
Thought’ (when we consider them specifically as Judgments), and on the other
hand, ‘merely from the accidental circumstances of [a proposition’s] External Ex-
pression’ (when we consider them as propositions) (LL.I.243). Thus, he charac-
terises three classes of proposition as properly adequate for logical consideration:
Universal judgments or propositions, ‘in which the whole number of objects within
a sphere or class are judged of,–as All men are mortal, or Every man is mortal’ ;
Individual judgments or propositions, in which ‘the whole of a certain sphere is
136                                    Ralph Jessop
judged of, but in which sphere there is found only a single object, or collection
of single objects,–as Catiline is ambitious,–The twelve apostles were inspired’, the
individual(s) in question here constituting what Hamilton describes, with possible
oblique reference to his notion of involution, ‘determinate wholeness or totality
in the form of oneness [or] indivisible unity’; and, Particular judgments or propo-
sitions, ‘in which, among the objects within a certain sphere or class, we judge
concerning some indefinite number less than the whole,–as Some men are virtu-
ous–Many boys are courageous–Most women are compassionate. The indefinite
plurality, within the totality, being here denoted by the words some, many, most’
(LL.I.245-6). As he explains by means of reference to the example cited above
concerning the conversion of a Particular proposition (‘An Englishman . . . ’) into
an Individual proposition (‘Newton . . . ’), although the logicians are right to treat
Universals and Individuals as convertible, their correspondence one with the other
is not merely due to ‘the oneness of their subject’, but rather: ‘The whole distinc-
tion consists in this,–that, in Universals and in Individual Judgments, the num-
ber of the objects judged of is thought by us as definite; whereas, in Particular
Judgments, the number of such objects is thought by us as indefinite’ (LL.I.246).
Hence, the major distinction between propositions in terms of quantity is between
the definite (universal or individual) and the indefinite (particular). This distinc-
tion between definite and indefinite quantity, Hamilton declares most forcefully in
his ‘New Analytic’: ‘definite and indefinite are the only quantities of which we
ought to hear in Logic; for it is only as indefinite that particular, it is only as
definite that individual and general, quantities have any (and the same) logical
avail’ (NA.LL.II.250).
   Thus, bearing in mind that, since Universals and Individuals with regard to
quantity, constitute one class of definite propositions, whereas Particulars consti-
tute the only alternative quantity and thus class of indefinite propositions, all that
needs to be added to this twofold definite (Universal and Individual) and indefinite
(Particular) distinction between propositional forms to yield the traditional (yet
Hamiltonized) fourfold distinction, is the standard distinction between affirmation
and negation, known as the quality of the proposition. Hamilton briefly outlines
the notion of quality, which he regards as an unfortunately ambiguous and yet gen-
erally accepted term to denote affirmation and negation (LL.I.250). Important to
his quantification of the predicate, he sensibly argues against some contemporary
and earlier logicians who held that affirmation and negation properly belong to
the copula and not to the subject nor to the predicate terms. Drawing attention to
the non-literal or non-grammatical sense in which the copula should be regarded
as expressing the form of the relation between subject and predicate, he argues
against certain previous and some modern logicians, that negation does not belong
to the predicate term but rather to the copula, allowing him to treat subject and
predicate as being held together in a reciprocal relation of whole to part, such that
in a negative judgment a part is taken out of a whole, whereas in an affirmative
judgment a part is put into a whole (see LL.I.251-4). All that thus belongs to the
subject and to the predicate, for Hamilton, is their respective definite (Universal or
                            The Logic of Sir William Hamilton                      137
Individual) or indefinite (Particular) quantities; but as this enables a distinction
between propositional forms according to their quantity (definite or indefinite),
traditionally given as Universal or Particular, when these are both distinguished
according to their quality (affirmative or negative), this results in the traditional
fourfold A, E, I, O distinction of propositional forms: A (universal affirmative); E
(universal negative); I (particular affirmative); O (particular negative).
   He represents the forms A, E, I, O by using simple circle diagrams (see LL.I.255).
I shall not provide these since they merely illustrate the quantity and quality as-
pects of the four propositional forms and since I shall shortly provide one of Hamil-
ton’s later tables from the ‘New Analytic’ which, by deleting one of the diagrams
as redundant and adding one that expresses co-extension, incorporates these circle
diagrams into a new set of four figures to which he relates both the traditional
forms A, E, I, O, and his additional forms. Although it is highly likely that Hamil-
ton pointed out to his students the quantifications of the predicate and the subject
terms implied by the diagrams, for example, that the Universal Affirmative (A)
in traditional logic implied an indefinite or Particular predicate, the text does
not at this stage make any explicit reference either to the assumptions concerning
quantification nor does it advance the thoroughgoing quantification that results in
Hamilton doubling the traditional four propositional forms A, E, I, O. It must be
noted, however, that much later in the Lectures he does assert that ‘The nineteen
useful [syllogistic] moods admitted by logicians, may [. . . ] by the quantification of
the predicate, be still further simplified, by superseding the significance of Figure’
(LL.I.402). Although his quantification of the predicate does not seem to have
been made fully explicit, he was clearly at the very least intimating aspects of it
to his students some years before his controversy with de Morgan.
   Lecture XIV is the last lecture on Hamilton’s doctrine of Judgments. Before he
terminates the lecture, he makes at least one significant claim worth mentioning,
namely, his rejection of Modal propositions as a separate class and his argument
that, for example, the modal proposition ‘Alexander conquered Darius honourably’
ought to be treated as merely a complex proposition in which the mode is regarded
as part of the predicate. As Hamilton points out, the predicate can be more or
less complex and there is no need for the Aristotelian logic’s modal propositions
as ‘modified by the four attributions of Necessity, Impossibility, Contingence, and
Possibility. [. . . ] in regard to these, the case is precisely the same; the mode is
merely a part of the predicate, and if so, nothing can be more unwarranted than on
this accidental, on this extra-logical, circumstance to establish a great division of
logical propositions’ (LL.I.257). Once again Whately comes under fire concerning
this, as also when Hamilton moves on to discuss and outline some basic points
concerning the subject of the conversion of judgments or propositions, such as
when the subject and predicate are transposed in a categorical proposition (see
LL.I.258–9; 262–3). I shall come back to the subject of conversion briefly later,
but for the time being I shall leave Hamilton’s lectures and the commencement of
his introduction to reasoning and the syllogism, and instead leap forward to his
later work in the fragmentary but nonetheless insightful and more mature work, his
138                                     Ralph Jessop
‘New Analytic of Logical Forms’. For, within Lecture XIV Hamilton has arrived at
a significant stage in laying the groundwork for his quantification of the predicate.
Although by this stage he has yet much more to construct, it would seem that he
has reached a critical point that has by now established a warrant for providing
the thoroughgoing quantification that I shall illustrate in the following section.
                   4 QUANTIFICATION OF THE PREDICATE
       THIS NEW Analytic is intended to complete and simplify the old —
       to place the keystone in the Aristotelic Arch. (NA.LL.II.249)
It is abundantly clear in the ‘New Analytic’ that Hamilton’s principle of explic-
itness, or rather the thoroughness of his adherence to this principle, is a major
driving force behind his quantification of the predicate. He cites the postulate
or principle as something insisted upon by Logic, but insufficiently adhered to by
previous logicians, claiming on the basis of this principle, ‘that, logically, we ought
to take into account the quantity, always understood in thought, but usually, and
for manifest reasons, elided in its expression, not only of the subject, but also of
the predicate, of a judgment’ (NA.LL.II.250; and see p. 252). Making explicit
the quantities of both the subject and the predicate in a judgment or proposition
facilitates regarding the subject and predicate terms as comprising ‘an equation’.
It is important to understand how it can be said that the whole-part relationship
in extension or intension may be thought of as constituting either an equation or
non-equation. Though in some places Hamilton does not seem to be as clear as
he might have been on this point, at other places his explanation is substantially
more helpful than the often rather submerged hints in the Lectures. For example,
in the ‘New Analytic’ Hamilton’s treatment of the subject of the Conversion of
Categorical propositions clarifies what he thinks erroneous in traditional logic’s
Conversion of propositions with regard to quantity.
   He regards the doctrine of Conversion as ‘beset with errors’ but that these errors
are generated from two principal ones — Hamilton is worth quoting at length here:
       The First cardinal error is,–That the quantities are not converted with
       the quantified terms. For the real terms compared in the Convertend
       [the original proposition], and which, of course, ought to reappear with-
       out change, except of place, in the Converse [the proposition converted],
       are not the naked, but the quantified terms. This is evident from the
       following considerations:
       1o , The Terms of a Proposition are only terms as they are terms of
       relation; and the relation here is the relation of comparison.
       2o , As the Propositional Terms are terms of comparison, so they are
       only compared as Quantities,–quantities relative to each other. An
       Affirmative Proposition is simply the declaration of an equation, a
       Negative Proposition is simply the declaration of a non-equation, of its
                                                                          139
                    The Logic of Sir William Hamilton
terms. To change, therefore, the quantity of either, or of both Subject
and Predicate, is to change their correlation,–the point of comparison;
and to exchange their quantities, if different, would be to invert the
terminal interdependence, that is, to make the less the greater, and
the greater the less.
3o , The Quantity of the Proposition in Conversion remains always the
same; that is, the absolute quantity of the Converse must be exactly
equal to that of the Convertend. It was only from overlooking the
quantity of the predicate [. . . ] that two propositions, exactly equal
in quantity, in fact the same proposition, perhaps, transposed, were
called the one universal, the other particular, by exclusive reference to
the quantity of the subject.
4o , Yet was it of no consequence, in a logical point of view, which of
the notions collated were Subject or Predicate; and their comparison,
with the consequent declaration of their mutual inclusion or exclu-
sion, that is, of affirmation or negation, of no more real difference than
the assertions,–London is four hundred miles distant from Edinburgh,–
Edinburgh is four hundred miles distant from London. In fact, though
logicians have been in use to place the subject first, the predicate last,
in their examples of propositions, this is by no means the case in ordi-
nary language, where, indeed, it is frequently even difficult to ascertain
which is the determining and which the determined notion. [. . . .]
The Second cardinal error of the logicians is, the not considering that
the Predicate has always a quantity in thought, as much as the Sub-
ject; although this quantity be frequently not explicitly enounced, as
unnecessary in the common employment of language; for the deter-
mining notion or predicate being always thought as at least adequate
to, or co-extensive with, the subject or determined notion, it is sel-
dom necessary to express this, and language tends ever to elide what
may safely be omitted. But this necessity recurs, the moment that, by
conversion, the predicate becomes the subject of the proposition; and
to omit its formal statement is to degrade Logic from the science of
the necessities of thought, to an idle subsidiary of the ambiguities of
speech. [. . . .]
1o , That the predicate is as extensive as the subject is easily shown.
Take the proposition,–All animal is man, or, All animals are men. This
we are conscious is absurd, though we make the notion man or men as
wide as possible; for it does not mend the matter to say,–All animal is
all man, or, All animals are all men. We feel it to be equally absurd as
if we said,–All man is all animal, or, All men are all animals. Here we
are aware that the subject and predicate cannot be made coextensive.
If we would get rid of the absurdity, we must bring the two notions
into coextension, by restricting the wider. If we say,–Man is animal,
140                                     Ralph Jessop
      (Homo est animal ), we think, though we do not overtly enounce it, All
      man is animal. And what do we mean here by animal? We do not
      think,–All, but Some, animal. And then we can make this indifferently
      either subject or predicate. We can think,–we can say, Some animal is
      man, that is, Some or All Man; and, e converso,–Man (some or all) is
      animal, viz. some animal.
      It thus appears that there is a necessity in all cases for thinking the
      predicate, at least, as extensive as the subject. Whether it be ab-
      solutely, that is, out of relation, more extensive, is generally of no
      consequence; and hence the common reticence of common language,
      which never expresses more than can be understood — which always,
      in fact, for the sake of brevity, strains at ellipsis. (NA.LL.II.257-9)
This lengthy quotation brings several things to our attention: that Hamilton re-
gards the quantifying terms as integrant components of the ‘naked’ terms or con-
cepts or judgments used in any given proposition — and hence whenever the
‘naked’ terms are transposed in some conversion, their clothing (the quantifier)
goes with them; the terms of a proposition only exist qua terms as related to one
another by a comparison of their respective quantities, and that their quantities
will be identical or equated in an affirmative proposition, non-identical or not
equated, but rather related by some confliction to do with their quantities, in a
negative proposition — violation of this relation of comparison or equation/non-
equation is to change without warrant the relation being asserted in their very
predication; that it is mistaken to deem a proposition to be Universal or Par-
ticular solely by attending to the status of the quantification of its subject; and,
Hamilton’s reliance on his principle or postulate of explicitness as the fundamental
principle of Logic is invoked as warranting exposure of what must (on pain of oth-
erwise thinking an absurdity) be thought with regard to the predicate’s quantity
and thereby its relation of equation or non-equation with the quantified subject.
   This last point concerning how in thought if not in linguistic expression (due to
our propensity in linguistic expression to elide quantities into the ‘naked’ terms)
strongly suggests a mental and in this sense private language of subject-predicate
equation/non-equation, a mental relation of the respective quantities of both sub-
ject and predicate, that is often but not always behind the scenes and tantamount
to the fundamental necessary laws of logic themselves. However, Hamilton seems
to be at pains to describe these laws as in some sense natural, informing actual
discourse, and in turn operating as the standard against which rhetorical utter-
ance may be tested and to which rhetoric may be reduced without attempting to
modify logic’s laws to suit grammar: ‘We should not do as the logicians have been
wont,–introduce and deal with [‘the rhetorical enouncements of common speech’]
in their grammatical integrity; for this would be to swell out and deform our sci-
ence with mere grammatical accidents; and to such fortuitous accrescences the
formidable volume, especially of the older Logics, is mainly owing. In fact, a large
proportion of the scholastic system is merely grammatical’ (NA.LL.II.262). The
                          The Logic of Sir William Hamilton                       141
tendency in Hamilton’s logic to correct and simplify traditional logic, thus marks
a major departure from regarding logic as an attempt to capture the vagaries and
complexities of grammatical rules, real-life argumentation, and rhetoric in order
to keep logic focused on the laws of necessary inference. However, although this
approach warrants the unnatural sounding expressions that result from quantify-
ing the predicate, as in ‘All men are some mortal’, Hamilton is not such a purist as
to eschew utterly the notion of Pure Logic’s relevance to and capacity to translate
‘the rhetorical enouncements of common speech’, nor is his quantification of the
predicate justified solely by means of reference to a merely dogmatic assertion that
quantity is ‘always understood in thought’ though often elided in its linguistic ex-
pression, since his arguments in support of quantifying the predicate incorporate
appeals to common instances in which the quantification is made explicit: ‘in fact,
ordinary language quantifies the Predicate so often as this determination becomes
of the smallest import’ (NA.LL.II.259).
   Be all that as it may, it becomes clear in the ‘New Analytic’ that one of Hamil-
ton’s major achievements with regard to the complex and, as he often asserts,
confusing doctrines, rules, and practices of the logicians has to do with how his
system effectually sweeps away various different types of conversion. With the
establishment of his quantification of the predicate the only defensible type of
conversion is the simple conversion he advocates. His simple conversion relies
wholly upon the predicate’s quantification being made as explicit as the quan-
tification of the original subject term. Hence, simple conversion merely involves
whatever transposition of terms is possible so long as the respective quantifiers of
the original subject and predicate terms remain attached to these terms in order
to retain in conversion any given proposition’s meaning as an equation or non-
equation of the quantities of the two terms in the original proposition. Over and
again Hamilton emphasises the erroneousness of earlier species of conversion, both
with regard to affirmative and negative propositions (see NA.LL.II.256-76). The
logicians had missed ‘the one straight road’ of conversion, simple conversion, and
Hamilton makes the ambitious but clearly defensible claim that if, by means of
his quantification of the predicate, he is right in having reduced all species of con-
version to the simple conversion he advocates, then ‘the whole doctrine of logical
Conversion is superseded as operose and imperfect, as useless and erroneous. The
systems, new and old, must stand or fall with their doctrines of the Conversion of
propositions’ (NA.LL.II.276).
   Though some elements of Hamilton’s construction of his system of quantifying
the predicate have been overlooked above, I think I have given an ample outline of
his notions concerning the quantification and how this radically supersedes much
of the traditional logic while yet incorporating and building upon at least some
part of it. However, we now need to look at Hamilton’s quantification procedure
in some more detail and the best place to commence this is by examining the table
supplied at Appendix V section (d) in the second volume of his Lectures on Logic,
which, with only a few minor adjustments, I have replicated below as Table 2.
   The only significant modification I have made to Hamilton’s table, is to adopt
142                                                   Ralph Jessop
       Table 2. Application of Doctrine of Quantified Predicate to Propositions
                                  New Propositional Forms — Notation
         Fig.l.                   Fig. 2.                                                       Fig. 4.
                                                             Fig. 3.
 Affirmative.
                [A fA]        C                : F    All Triangle is all Trilateral [fig. 1].
    (1)
                [A f l ]      C                  A   All Triangle is some Figure (A) [fig. 2].
    (ii)
                              A                :C     Some Figure is all Triangle [fig. 2].
                [Ii A ]
    (3)
                [I f l ]      C               •, B    Some Triangle is some Equilateral (I) [fig. 4].
    (iv)
 Negative,
    (v)         [A n A ]      C                : D    Any Triangle is not any Square (E) [fig. 3].
                              c
                [A n l ]                      •, B   Any Triangle is not some Equilateral [fig. 4].
    (6)
                [I il A ]     B                : C    Some Equilateral is not any Triangle (O) [fig. 4].
    (vii)
    (8)         [I n I]       C.               ,B     Some Triangle is not some Equilateral [tig. 4].
  Key:
  1,3,6,8              = Hamilton's forms
  ii, iv, v, vii       = Aristotelic or traditional forms. Hence, using the scholastic letters A, E, I, O: A
                          (Universal affirmative) = ii; E (Universal negative) = v; I (Particular affirmative) =
                          iv; O (Particular negative) = vii.
  A = universal; I = particular; f = affirmation; n = negation;, = some; : = all.
                       = the affirmative copula (is)
                       = the negative copula (is not)
  The two arrow lines above indicate which of the terms is subject and which predicate thus: where the
  proposition is being read in Extension, the thick end of the arrow line denotes the subject, the thin end
  the predicate; where the proposition is being read in Intension, subject and predicate are reversed and
  hence the thick end of the arrow line denotes the predicate, the thin end the subject.
  To illustrate how to read Hamilton's symbolic representation of a proposition in Table 2, note that the
  following should be read thus:
  C: B»^~,A
  Extensively: All C is some A — i.e. All C is contained tinder some A
  Intensively: Some A is all C - i.e. Some A contains in it all C.
                          The Logic of Sir William Hamilton                       143
his later method of symbolising the Universal and Particular terms using the letters
A for Universal, and I for Particular (enclosed in square brackets). This alteration
should also be of some help to readers who may wish to make comparisons between
Table 2 and Hamilton’s more detailed ‘Table of the Mutual Relations of the Eight
Propositional Forms on Either System of Particularity’ to which I shall refer later
(see NA.LL.II.284).
   Table 2 is interesting in several ways. Firstly, the various configurations of A
(universal/definite) and I (particular/indefinite) under Affirmative and Negative
regularly display all of the possible permutations of a thorough quantification of
both subject and predicate as neatly displayed in the symbols given in square
brackets such as [A f A]. Secondly, the arrow lines indicate which of the terms
is the subject and which the predicate, depending on whether the proposition is
to be read as extensive or intensive — hence, the terms may be easily transposed
according to Hamilton’s method of simple conversion without any alteration in
the proposition’s meaning using the same symbolic notation for both an extensive
and an intensive reading (though it must be said that by this stage in Hamil-
ton’s logic, having made so much of extension and intension to establish the dual
perspectives from which propositions may be viewed, how thus the subject and
predicate terms may be transposed, and how the number of syllogistic forms can
thereby be amplified, the extension-intension distinction seems to fall out of ac-
count as superseded by the full quantification itself). Thirdly, using Hamilton’s
own terminology and categorisation, the four circle diagrams need to be thought
of as expressing not four but three possible principal relations of: Toto-total Coin-
clusion (fig. 1); Toto-total Coexclusion (fig. 3); and, brought together under one
class of counter-related relations, Incomplete Coinclusion and Coexclusion (fig. 2
and fig. 4). Fourthly, Hamilton’s example of proposition (1) uses the symbol ‘Γ’,
but since it only appears in this proposition, it is not immediately clear which
of the other propositions may be said to be its contradictory or negation, though
with a simple change of terms proposition (v), describing the relation of Toto-total
Coexclusion (fig. 3), seems to be the most obvious contradictory of proposition (1)
as describing Toto-total Coinclusion (fig. 1). These last two points require further
explanation and as we shall see this will involve some discussion of a major source
of difficulty and controversy concerning Hamilton’s quantification.
   Firstly, my third point above: in Hamilton’s system the relationship between
subject and predicate in each proposition needs to be thought of as a relationship
of mutuality. This becomes much more clear when we take note of Hamilton’s
‘Observations on the Mutual Relation of Syllogistic Terms in Quantity and Quality’
at Appendix V (e) in the Lectures on Logic (see NA.LL.II.285). With reference to
the circle diagrams in Table 2 above, these relations can be given as follows:
         1. Toto-total Coinclusion (fig. 1) — the relation of ‘coidentity, ab-
            solute convertibility or reciprocation’.
         2. Toto-total Coexclusion (fig. 3) — the relation of ‘non-identity,
            absolute inconvertibility or non-reciprocation.
144                                     Ralph Jessop
         3. Incomplete Coinclusion and its counter-relation Incomplete Coex-
            clusion (fig. 2 and fig. 4) — the relations of ‘partial identity and
            non-identity, relative convertibility and non-convertibility, recip-
            rocation and non-reciprocation’. Under this counter-related pair
            of Incomplete Coinclusion and Coexclusion, Hamilton details all
            of the propositional forms he regards as intermediaries between
            the extreme opposites of proposition (1) Toto-total Coinclusion,
            and proposition (v) Toto-total Coexclusion as:
            (ii)   Toto-partial coinclusion (fig. 2)
            (3)    Parti-total coinclusion (fig. 2)
            (iv)   Parti-partial coinclusion (fig. 4)
            (6)    Toto-partial coexclusion (fig. 4)
            (vii)  Parti-total coexclusion (fig. 4)
            (8)    Parti-partial coexclusion (fig. 4)
                                                            (see NA.LL.II.285)
As Hamilton argues in his Lectures, in keeping with traditional logic a universal (A
or E) may be treated as an individual or as a universal proposition — hence, the
Toto-total Coinclusion and Coexclusion relations represented in Table 2’s propo-
sition (1) [fig. 1], and proposition (v) [fig. 3] respectively, may be translated
in two ways as referring either to: two single/individual entities (A and B) that
are (1) identical or (v) non-identical; or, two groups (collections) of things, all of
which in each collection (A and B) are sufficiently resembling to be asserted as
(1) identical or (v) non-identical — hence Fogelin seems to be right to claim that
‘Hamilton develops his theory of universal propositions on an existential (rather
than a Boolean) interpretation)’, though arguably, as seems to be the case with
his specific illustrations of propositions (1) and (v), he does also accommodate a
priori truths constituted by wholes that may be thought of as having no existential
import as physical realities, the whole and thus individual that is ‘All Triangle’
existing in thought, though its ontological status, as a purely ideal whole, is by no
means unrelated to the whole that might be any more or less complete collection
of real triangles as instantiations of the unity of thought that is ‘All Triangle’ (see
Fogelin, p. 152). Incidentally, the complex and interesting topic of Hamilton’s un-
derstanding of the existential import of propositions, which I have merely touched
on here, is not perhaps as one at might first think, since Hamilton is conscious that
‘the Logician has a right to suppose any material impossibility, any material fal-
sity; he takes no account of what is objectively impossible or false, and has a right
to assume what premises he please, provided they do not involve a contradiction
in terms’ (LL.I.322; also see, p. 338; p. 360).
   Though in his Lectures Hamilton seems to keep to just one sense of the quanti-
fier ‘some’ it becomes clear in the ‘New Analytic’ that particulars, though in each
and every case an indefinite quantity, may be indefinite in two different senses,
                           The Logic of Sir William Hamilton                        145
either: as Indefinite Definitude, in the sense expressed by ‘some, at least’ (which
is to say, ‘at least one, possibly more, but not all’); or, as Definite Indefinitude, in
the sense of ‘some, at most’ (which is to say, ‘some, perhaps all, but not less than
one’) — hence, the Incomplete Coinclusion and Coexclusion relations represented
in Table 2’s fig. 2 and fig. 4, may also be translated in two different ways to ac-
commodate this difference in the possible senses of the Particular quantifier ‘some’.
The significance of these points should become clear later, but in the meantime, I
want to claim that while Hamilton made a great deal of the importance of Logic
as a Pure science that must not become intermixed with any grammatical, linguis-
tic, rhetorical, or other material concern (inasmuch as this is practically possible),
as Fogelin rightly points out, ‘Hamilton acknowledges both interpretations of the
quantifier some, but only insists that each interpretation must be examined in
order to capture all the everyday inference patterns a logician should study’ (see
Fogelin, p. 153). It does seem as though Hamilton is attempting to make his
system sufficiently accommodating, such that it can encompass different readings
of ‘some’ (or can embrace different degrees of indefiniteness). His system can also
be applied, not simply to both individual and universal quantities (as the only
definite quantities), but also, on the one hand: to a priori truths or universals,
the wholeness of which is as an unanalysable individual/singular, their ontological
status being ideal and thus potentially existing only in thought, though actually
never out of relation to, real entities; and, on the other hand and more conspicu-
ously, his system can be applied to those universals that we might better describe
as general terms, the a posteriori nature of which implies that at best they are
only approximate or provisional universals which may admit of some exceptions
without nevertheless losing their applicability in a syllogistic reasoning as universal
terms.
   Now, to come to the second main observation I want to make about Table 2:
Hamilton’s proposition (1) is diagrammatically represented by figure 1 in Table 2
and figures 2, 3, and 4 all look as though they illustrate relations that must stand
counter to proposition (1), but since the example Hamilton gives uses a term ‘Γ’
that he does not replicate elsewhere in Table 2, it is not immediately clear which
of the other propositions may be said to be its contradictory or negation. Figure
1 absolutely equates C with Γ and thus describes an absolute identity or Toto-
total coinclusion relation between C and Γ. This is also illustrated by Hamilton’s
example of a possible proposition that might express this relation between the
judgments or concepts C and Γ: ‘All Triangle is all Trilateral’. However, surely
both C and Γ collect together respectively the entire class of C (all possible shapes
and sizes of triangles) and Γ (all possible Trilateral figures) and, linked by the
copula, the proposition ‘All C is all Γ’ brings these two quantities into a relation
of comparison? If so, does this mean that de Morgan’s interpretation of proposi-
tion (1) is right, namely, that it expresses a complex proposition constructed by
compounding ‘every C is Γ’ and ‘every Γ is C’ (de Morgan, p. 257)?
   In Fogelin’s first attempt to express the meaning of Hamilton’s proposition
(1), he accepts de Morgan’s interpretation and (though shown using A and B
146                                     Ralph Jessop
as terms) Fogelin thus translates proposition (1) as ‘All C is Γ and all Γ is C’
(Fogelin, p. 151). On this reading of proposition (1), according to de Morgan, it
is contradicted by one of either proposition (6) or (vii), which we may here give,
not as in the illustrative propositions given in Hamilton’s Table 2 but, to conform
to the letter symbols used in proposition (1), as follows: (6), ‘some Γs are not
Cs’; (vii), ‘some Cs are not Γs’ — these are, for de Morgan, the contradictories of
proposition (1), ‘All C is all Γ’ (de Morgan, p. 257). To be sure, as is clearly seen
simply by comparing figure 1 (illustrating proposition (1)) and figure 4 (illustrating
propositions (6) and (vii)), both (6) and (vii) must stand in a negative relation to
proposition (1). In a helpful but somewhat complex and possibly inaccurate table,
Hamilton asserts that (6) and (vii) stand as what he calls unilateral contraries of
(1) — this seems to suggest that he would accept that (6) and (vii) do severally
contradict (1). However, as contraries neither (6) nor (vii) properly constitute a
contradiction of (1) (see, NA.LL.II.284). Now, it might at first sight seem odd
that neither Fogelin nor (in his original critique of Hamilton) de Morgan, mention
proposition (v) as the contradictory of (1), since, translating this into symbols
consistent with (1) and in line with de Morgan’s reading, proposition (v) should
be read as stating ‘No C is Γ’ and indeed, to be consistent with de Morgan’s
translation of proposition (1), he ought to have translated this as ‘No C is Γ and
No Γ is C’. However, de Morgan does mention proposition (v) in a later footnote
as being offered as the contradictory of (1) by ‘an eminent defender’ of Hamilton’s
system (de Morgan, p. 258n1). But de Morgan brushes this aside as not being
‘in the system’. Quite what de Morgan means by this is rather unclear, especially
since as Fogelin points out, ‘de Morgan is just wrong in suggesting that the system
of propositions do not pair up into proper contradictories’, and he goes on to list,
(1) and (8); (ii) and (vii); (3) and (6); (iv) and (v) as contradictory pairs. But still,
it may seem puzzling why (1) Toto-total coinclusion and (v) Toto-total coexclusion
should not be thought of as contradictories. Proposition (v)’s relation to (1), is
displayed in one place by Hamilton in such a way as to suggest that (v) and (1)
mutually contradict one another (see NA.LL.II.286). However, in another place
Hamilton gives their relation as one of bilateral contraries — this is to say that
both the Toto-total coinclusion of the terms in (1) as shown in figure 1 of Table 2,
and the Toto-total coexclusion of the terms in (v) as shown in figure 3 of Table 2,
stand not as contradictories of one another but both potentially false; which is of
course to say, that if (1) is true though (v) must be false and vice versa, since (1)
and (v) may both be false they are not strictly contradictories (see NA.LL.II.284).
   Be all that as it may, as Fogelin rightly points out, the trouble lies with proposi-
tion (8) which de Morgan rejects as having no contradictory within the system. To
establish against de Morgan’s rejection of proposition (8) that Hamilton’s system
is comprehensive and not inconsistent, we need to be able to answer the question:
what is the true contradictory of (8) in the system? This is an important ques-
tion since if there is no contradictory of (8) within the system, de Morgan is right
to assert that it has not been generated from any necessary laws of thought but
rather, as de Morgan so derisorily claims, on the basis of ‘an arbitrary extension
                          The Logic of Sir William Hamilton                         147
of the application of language’ (de Morgan, p. 258n1). However, as intimated
above, Fogelin ably answers de Morgan’s rejection of (8) by arguing that its true
contradictory is (1). Incidentally, although de Morgan can be fulsome in his praise
of Hamilton — but then so can Mill — de Morgan’s treatment of Hamilton’s
quantification scheme is at times rather scurrilously worded. It would seem that
both de Morgan and Mill were, rather excessively, much given to resort to more
or less veiled abusive ad hominem attacks on Hamilton, and I cannot help but
comment here that while Hamilton’s frequent denunciations of others, including
de Morgan, must have to some extent provoked such responses, his achievements
and reputation as a logician have most certainly suffered unduly from the cheap
rhetorical tricks of opponents whose conduct ought to have been exemplarily fair,
not solely as a mark of respect for Hamilton’s considerable endeavours but also as
a generally more virtuous way of conducting their discourse — paradoxically, it
would seem that rather too often winning the argument is much more important
to those who should be most concerned with striving to resolve it satisfactorily.
   Fogelin’s defence of Hamilton is an incisive attempt to redress the balance and
although I shall not rehearse the full extent of his critical examination of Hamil-
ton’s system, to date it stands as one of the strongest defences of Hamilton’s
quantification of the predicate. Fogelin’s defence is carried out in part by translat-
ing the controversial proposition (8), which de Morgan also claims with complete
disdain was erroneously, and, so he implies, foolishly offered by one of Hamilton’s
defenders as the true contradictory of (1) (de Morgan, p. 259n3). As Fogelin
rightly says about proposition (8), ‘it is this proposition that has been the con-
stant source of confusion’ (Fogelin, 151). However, as Fogelin rightly attempts
to show in his first article on Hamilton’s quantification, proposition (8) is indeed
the contradictory of proposition (1). But, Fogelin’s second article on Hamilton
makes an important correction to his first attempt to establish that proposition
(8) genuinely contradicts proposition (1), and it is therefore to this second article
that I shall now refer.
   According to Fogelin’s reading of proposition (1) this should be interpreted as:
‘Anything that is an A is identical with anything that is a B’ — which ‘means
that there is but one thing that is an A, one thing that is a B, and these things
are identical’ (Fogelin, p. 167). This is to say, as Fogelin argues, that de Morgan’s
interpretation of proposition (1) — ‘All A is B and all B is A’ — is wrong.
However, de Morgan was right in his interpretation of proposition (8) and hence
(8) can be stated as ‘Some A is not some B’, which may then be translated as the
contradictory of (1) as: ‘Something that is an A is not identical with something
that is a B’. Fogelin goes on to demonstrate just how this interpretation renders a
certain syllogism (which on his previous interpretation of proposition (8) is invalid),
can be shown to be valid using his second (and de Morgan’s original) reading,
thereby proving that Hamilton’s system is ‘saved from inconsistency’ (Fogelin, p.
168). Although the relation of Hamilton’s quantification system to syllogisms is
of course important, instead of looking at this I want to enrich Fogelin’s argument
somewhat by considering his interpretation of proposition (1) as meaning ‘that
148                                     Ralph Jessop
there is but one thing that is an A, one thing that is a B, and these things are
identical’.
   To return to Hamilton’s own example, ‘All Triangle is all Trilateral’: as express-
ing the relation of Toto-total coinclusion, this expresses the nearest thing (along
with Toto-total coexclusion) Hamilton would call an absolute, which is to say that
the subject and predicate terms are co-inclusively related as maximally similar
such that they may be said to be absolutely convertible one with the other — they
are thus absolutely reciprocal. Whatever, if anything, might be said to differentiate
‘all Triangle’ from ‘all Trilateral’, such that they can be clearly if not distinctly
separated into two entities (and this may merely be that the terms themselves are
different signs both of which signify the same single entity), as two terms brought
into a mutual relation of coidentity, this coinclusion is nonetheless one in which all
material difference has been abstracted from thought, such that the co-identities
are indeed absolute and the proposition that articulates this is thus effectually an
implicit denial of their non-identity. Hence, such propositions asserting Toto-total
coinclusion that involve merely a terminal differential assert a co-identity between
two terms, and as such these propositions can be most directly contradicted by
the assertion of the most minimal difference in their quantity, since with regard to
their quantity alone their co-identification implies a unity or singularity — hence
the assertion of parti-partial coexclusion in proposition (8) is this most direct
contradiction of the unity expressed by proposition (1) in terms of two unities
being co-identical, since the very ground of (1) being the expression of absolute
identity is at once negated by proposition (8)’s assertion that the co-identities are
co-excluding.
   But, we need to keep in mind that, for Hamilton, there is no absolute exclusion
in that relation between concepts known as Exclusion. Since he rejects absolute
exclusion, but also since Hamilton so permeates his system with correlations of
one sort or another, it seems fair to regard his system as ultimately one in which,
as there can be no absolute exclusion, there can also be no absolute inclusion or
perfect identity/unity. Rather, the universals of, say, the Toto-total Coinclusion
proposition (1), as relating at least two terms together, are assertions of either:
an approximate (adequate) but non-maximal co-identification; or, an absolute co-
identification that is total, but only to the extent that the subject and predicate
being equated in the proposition rely upon some merely nominal/terminal dif-
ferential. For Hamilton it would seem that some differential is the minimum
requirement for any concept, judgment, or reasoning to be possible, whether this
differential is actual (thus rendering the universal approximate), or is so crucially
dependent upon the merely terminal as to render all other distinction between
them impossible. Even when proposition (1) may seem to be an affirmation or
assertion of perfect identity, for (1) to exist as a proposition or material expression
of the unity of thought in which A is identified with B, it must consist of at least
two entities. As such, ‘All A is all B’ is most directly contradicted by the assertion
of the sole proposition that most adequately breaches or contradicts the relation
of coinclusion affirmed by ‘All A is all B’. For (1) and (8) to be contradictories,
                           The Logic of Sir William Hamilton                        149
the ‘all’ in (1) quantifies both subject and predicate as individuals absolutely co-
identified (though the ‘absolute’ here, must involve some differential), while the
‘some’ in (8) must quantify the subject and predicate of (8) to be at least the
individuals referred to in (1) also, which (8) must then be asserting are coexclu-
sive to an extent beyond the necessary differential making (1) possible, such that
these individuals in (8), contradicting (1), may be said to be co-exclusive. The
contradictory of proposition (1), as before, is hence the assertion of Parti-partial
Coexclusion in proposition (8), ‘Some A is not some B’, or as Fogelin interprets
this, ‘Something that is an A is not identical with something that is a B .
   Now, the unity of A and B’s coinclusion or co-identification in thought, being
expressed as a coinclusion relationship between the two terms ‘A’ and ‘B’, is the
expression of a bringing into unity concepts which, as Hamilton took much trou-
ble to explain in his Lectures, are themselves relative, since a concept consists of
disparate entities thought as one by means of the degree of resemblance between
their several attributes — this is why in the bulk of actual cases of proposition
(1) expressing Toto-total coinclusion we need to think of their co-identity as ap-
proximate. However, the best examples of proposition (1) may be thought of as
propositions in which the subject and predicate terms are concepts or singular-
ities/individuals in which the constituents that respectively define them, which
is to say their intensive quantities, are not mere bundles of pluralities possibly
thought erroneously or merely approximately as constituting two unities. Instead,
in a best case example of proposition (1) the subject and predicate terms will be
constituted by attributes that so overlap or co-inform one another in meaning that
they comprise what one might just as well call a true whole, a whole the parts of
which are as notions involving and involved. Thus, the intensive quantities of both
the subject term and the predicate term in a best case example of proposition (1)
will be involuted wholes, which are in turn related to one another by Involution
to form what we might call a true whole. Such true wholes are best exemplified
by a priori truths in which each of the terms is involved and involving in each
other. Now, while this may not be fully satisfactory, some such extrapolation from
at least Hamilton’s notion of involution and some other elements of his work on
Logic, seem to go a long way to justify Fogelin’s reading of the real meaning of
proposition (1), namely, that it is the assertion that there is but one thing that
is an A, and one thing that is a B, and these things are identical. The unities
co-identified in a best case example of proposition (1), as involuted wholes, them-
selves both involved and involving one another, bespeak the nearest true whole or
unity that Hamilton can admit into his system. Hence, the only possible and most
efficient and immediate contradiction of the coidentity of A and B, yet again, must
be the assertion of Parti-partial coexclusion between ‘A’ and ‘B’ as expressed in
proposition (8), for this is not to find a mere single exception within A or B that is
not a co-identical attribute of both, but rather this is to declare the non-identity or
co-exclusion of the two things that are A and B, and thus (8) is the contradiction
of the unity that proposition (1) asserts.
150                                    Ralph Jessop
    However, this story, complicated enough, is not yet fully resolved. For Hamilton
lists propositions (1) and (8) as Compossible, which is to say, that they are mutually
but not existentially contradictory of one another (i.e. though contradictory they
can coexist). Intriguingly this Leibnizian notion of compossibility is expressed by
means of empty spaces in the relevant columns of Hamilton’s ‘Table of the Mutual
Relations of the Eight Propositional Forms’ (see NA.LL.II.284). But if (1) and (8)
are compossible does this not run contra to Fogelin’s strong defence of Hamilton’s
system? As I understand it, on Fogelin’s reading, the contradictoriness of (1) and
(8) is existential. So, how can Hamilton be right to class (1) and (8) as compossible
if they must be regarded as existentially contradictories?
    What we are looking for is an interpretation of (1) and (8) that shows that
they mutually contradict one another and yet may be said to be coexistent. One
way in which (1) and (8) may be said to be compossible has to do with the sense
in which (1) expresses an absolute unity, and the sense in which (8) expresses
indefiniteness. For Hamilton, (8) expresses either one of two different species of
indefinitude. Thus, all that (8) may be asserting against (1) is the very differential
implicit within (1)’s terminal difference between its subject and predicate. For,
while (8) may operate in some instances and be used within a syllogism as if it is the
existential contradictory of (1), as soon as it in fact loses its Particular/indefinite
status and becomes either an individual or a universal in both quantities — as soon
as we know that its terms have to be quantified as universal/individual (hence def-
inite) — it must immediately be transformed into proposition (v), ‘Any A is not
any B’/‘No A is any B’. Of course, were we able to introduce, as a known, partial
definitude, (8) would similarly metamorphose into either (6) or (vii). Thus, while
(1) and (8) do mutually contradict one another as conflicting assertions, the indef-
initeness of (8) and the strictly non-absolute nature of even a best case example of
(1) does not warrant knowledge of existential contradiction (mutual annihilation),
though such coexistence may be for us unimaginable or inconceivable.
    But this does not resolve the problem since it seems to amount to saying that
we can regard (8) as functionally the existential contradiction of (1) though possi-
bly not its true contradictory but rather the condition of (1)’s existence and vice
versa, and hence that (1) and (8) are propositions the compossibility of which is
dependent on our epistemological limitation. Now, granted that (1) is the asser-
tion of unity, (8) the assertion of divisibility, it might be the case that Hamilton’s
classification of (1) and (8) as compossible simply serves to remind us of the very
epistemological limitation — our nescience — with regard to anything absolute
and anything infinite similar to what he enounced in ‘Philosophy of the Uncon-
ditioned’ by means of his Law of the Conditioned. But, the problem with this
is that if Hamilton’s classification of (1) and (8) as compossible relies solely on
the possibility that their mutual contradiction is already inherent within both of
these propositional forms, then the attribution of compossibility is pointless since
it merely reiterates what we might regard as an implicitly admissible distinct-
ness between the subject and predicate terms in (1), and were it the case that (8)
only functioned to perform this reiteration, as Hamilton himself suggests, (8) itself
                          The Logic of Sir William Hamilton                        151
would be useless (see Discussions, p. 695).
   But these points notwithstanding, the compossibility of (1) and (8) may reside
quite simply in Hamilton’s principle of explicitness, as follows: in logic we must be
allowed to express what we must be allowed to think and therefore we can think and
express the contradiction even of that which seems to disallow contradiction (for
example, ‘All Triangle is all Trilateral’) — and in this sense (8) is the expression
of what contradicts (1), even though its contradiction (‘Some Triangle is not some
Trilateral’) may be or seem to us to be an expression lacking all existential import
and merely derived as the incomprehensible negative of an assertion of a necessary
truth and absolute unity. If this is so, then the compossibility of (1) and (8) is
simply the coexistence of what formally contradicts but does not thereby entail
existential contradiction. This does not rule out that in fact a given particular
instance of (1) and (8) may existentially contradict one another, but since the
indefinite status of (8) cannot provide positive knowledge, excepting that it asserts
co-exclusion where (1) asserts co-identity, (1) and (8) have to be regarded as
formally compossible.
   For all its seeming weirdness, baulked at and ridiculed by de Morgan, and for
all the difficulty of attempting to resolve this problem of compossibility, I think
Hamilton is right to defend proposition (8) as importantly the declaration that
a whole of any kind is divisible, or that the assertion of an indivisible unit may
be negated by the assertion of divisible plurality (see Discussions pp. 695-6). As
such (8) does contradict or is the negation of proposition (1) as the declaration
or ‘judgment of indivisibility, of unity, of simplicity’ (see Discussions, p. 695).
However, retaining (1) and (8) as formally compossible draws attention to a certain
freedom of thought within Hamilton’s system that underscores the importance of
being able to think and thus express our ability at least to entertain the negation
of any assertion of an absolute or indivisible whole. This liberal approach is
one that endorses the enormous importance of the dialectical nature of logic and
argumentation as it recalls from obscurity the seemingly innocent, impotent, and
merely theoretic, but sophisticated, potent, and practically useful parti-partial
negative of otherwise unquestionable absolutes. If I am right so to interpret at least
part of the significance of Hamilton’s defence of proposition (8), and thus of the
complete range of his thoroughgoing quantification system, what we seem to have
between (1) and (8) is something akin to the contradictory relationship between
absolute and infinite that, as it first brought Hamilton fame with his controversial
and yet highly potent and profoundly influential Law of the Conditioned, placed
the conditioned nature of legitimate thought within a frame or field of reference
at once vertiginously complex and yet competent to enable a rational negotiation
with and within an awareness of this complexity.
152                                     Ralph Jessop
                   5 THE SYLLOGISM: SOME IMPLICATIONS
       The science now shines out in the true character of beauty,–as One at
       once and Various. Logic thus accomplishes its final destination; for as
       ‘Thrice-greatest Hermes,’ speaking in the mind of Plato, has expressed
       it — ‘The end of Philosophy is the intuition of Unity.’ (NA.LL.II.252)
Hamilton’s treatment of the syllogism deserves a whole chapter in its own right —
I shall only be able to deal with it rather briefly in this section. In the Lectures
Hamilton takes a considerable amount of time to explain many detailed aspects of
the syllogism to his students. He distinguishes and displays four different classes
of syllogism — the Categorical, Disjunctive, Hypothetical, and the Hypothetico-
disjunctive (see LL.I.291-2). The categorical syllogism is also displayed in both
extensive and intensive forms, something he will later capture in his symbolic
notion as indicted in Table 2 earlier and as fully detailed in the final table of the
second volume of Lectures (LL.I.295-300).38 Importantly, in drawing attention
to the different reasonings between extensive and intensive syllogisms, where the
copula signifies respectively ‘contains under’ and ‘contains in’, Hamilton makes
the point that from what can be observed of the inverse ratio relation between
extension and intension with regard to syllogisms, ‘it is not to the mere external
arrangement of the terms, but to the nature of their relation, that we must look in
determining the character of the syllogism’ (LL.I.300; and compare, p. 348). This
is important to how he will proceed to regard the syllogism within the lectures
but it also bears within it the necessity of quantifying the predicate, even though
at this stage he does not seem to have produced the full system of quantification
as given in Table 2 earlier. With extensive and intensive syllogisms differentiated,
Hamilton constructs three rules, in place of Whately’s six, three for extensive, and
three for intensive syllogisms, the intensive merely inverting the extensive rules
(see LL.I.305-6; 315).
   In all this we can see Hamilton working, as it were, from the ground up — as we
shall see shortly, his ultimate position will be even more simple or general as, on the
basis of his thoroughgoing quantification system, he develops a single general rule
or Canon governing all valid syllogisms in both affirmative and negative moods.
With regard to what is happening in the Lectures it is therefore important to
remember both their instructional function and that some aspects in the Lectures
are later superseded, such as, for example, his later rejection of the Rule of Reason
and Consequent in favour of just three main Rules of Identity, non-Contradiction,
and Excluded middle (see LL.I.290n).
   In his treatment of the syllogism in the Lectures, Hamilton introduces his stu-
dents to the usual suspects: the four figures, moods, the ingenious mnemonics of,
for example, Barbara, Celarent, Darii, the formal fallacies, and so on, a great deal
  38 Compare William Thomson, An Outline of the Necesary Laws of Thought: A Treatise on
Pure and Applied Logic, 3rd edn. (London: Longman, Brown, Green and Longmans, 1854),
pp. 245–9. Thomson fully endorses and offers some explanation of this table which he also
reproduces.
                           The Logic of Sir William Hamilton                     153
of which he lays out with painstaking detail (see LL.I.394-468). He also discusses
the various forms of conversion, but since I have already touched on this subject in
the previous section and how Hamilton’s quantification effectually displaces other
kinds of conversion with his simple conversion, it is needless to say anything more
about his treatment of it in the Lectures (see LL.I.262-5; NA.LL.II.264-76). Using
simple circle diagrams to illustrate the relations between the extremes or subject
and predicate terms of the conclusion and the middle term, his explanations of
syllogisms must have given his students an excellent grounding in the differing
figures and moods. However, as so often occurs in Hamilton’s expositions of the
traditional logic, he marks some significant differences between his treatment and
that of both his predecessors and contemporaries. One major example of this is
his rejection of the fourth figure which is, of course, simply shown, in extension,
as follows:
                                         P is M
                                         M is S
                                          S is P
He argues that, though the fourth figure can be shown to be valid, ‘the logicians,
in consequence of their exclusive recognition of the reasoning in extension, were
not in possession of the means of showing that this figure is a monster undeserving
of toleration, far less of countenance and favour’ (LL.I.424). I shall not rehearse
Hamilton’s arguments against the fourth figure, except to note that he shows
that in this figure there is an unwarranted switch from reasoning in extension to
intension or vice versa and thus it performs ‘a feat about as reasonable and useful
in Logic, as the jumping from one horse to another would be reasonable and useful
in the race-course. Both are achievements possible; but, because possible, neither
is, therefore, a legitimate exercise of skill’ (LL.I.427). But, Hamilton’s principal
reason for rejecting the fourth figure is that it involves a mental process ‘which is
not overtly expressed’ — in other words, when we adhere rigidly to the principle of
explicitness, the fourth figure’s reliance upon an intermediary conclusion becomes
evident (see LL.I.427-8).
    There are many other interesting features in the lectures worth mentioning,
such as his treatment of the difference between Induction and Deduction, which I
briefly touched on in Section 2 above. He discards what is now much more typ-
ically classed as Inductive argument, instead regarding Deduction and Induction
both as formal forms of demonstrative reasoning — logicians had ‘corrupted and
confounded’ logical deduction ‘governed by the necessary laws of thought’ with
contingent matter and probability (LL.I.325; and see 319–26). This formal ap-
proach, according to Hamilton, is more in keeping with Aristotle’s understanding
of induction (see LL.I.325–6). It also has important implications for how Hamilton
regards both the Sorites and Enthymeme, both of which he discusses in terms of
their formal characters as syllogisms. He does of course explain that the Sorites
became associated with that sophism or informal fallacy commonly referred to or
illustrated by means of the examples of piling up grains of sand until what was
154                                     Ralph Jessop
once maintained to be a small quantity becomes large (a Progressive Sorites), and
also the famous bald man example (Regressive Sorites) (see LL.I.376–8; 464–6).
According to Hamilton, the Sorites only became associated with such sophistic
or fallacious reasoning some time in the 15th century and the failure of logicians
to incorporate the Sorites as a legitimate chain-syllogism was all down to their
exclusive concentration on extension and not keeping in mind that, for Aristotle
‘all our general knowledge is only an induction from an observation of particulars’
(LL.I.377; 380; and see, 366–85).
   His treatment of the Enthymeme is similarly interesting and informative, ar-
guing that it is only the external form of the enthymeme that may be said to be
imperfect or incomplete. As Hamilton rightly shows, an enthymeme is not merely
an argument in which one of the premises is missing or suppressed; it may also
be that the conclusion has been suppressed/omitted. But, whether the major or
minor premise or the conclusion is not made explicit, this does not, for Hamilton,
warrant calling an enthymeme a special or defective syllogism — it ‘constitutes no
special form of reasoning’, nor did Aristotle maintain that it did (LL.I.387). The
enthymeme illustrates an important principle that pervades so much of Hamilton’s
approach to logic, namely, that the mere verbal accident of elision (or in the case of
the enthymeme we might say more or less deliberate omission, often serving pur-
poses of persuasion which Hamilton somewhat oddly does not address directly)
is something that the logician, in staunchly adhering to the principle of explicit-
ness, can make explicit as something that is in thought, though not in expression
(for Hamilton’s treatment of the enthymeme, see LL.I.386-94). He provides some
nice examples of enthymemes but I shall only quote one of these since it involves
a nice quip against Hegel that suggests something of Hamilton’s capacity for at
least occasional touches of humour: ‘There is recorded [. . . ] a dying deliverance of
the philosopher Hegel, the wit of which depends upon [its] ambiguous reasoning.
‘Of all my disciples,’ he said, ‘one only understands my philosophy; and he does
not.’ But we may take this for an admission by the philosopher himself, that the
doctrine of the Absolute transcends human comprehension’ (LL.I.398). There is
in this also more than a hint of not only Hamilton’s opposition to Hegel but also
to absolutism more generally.
   To his credit Hamilton reminds his students of certain points he made much
earlier in the lectures to do with the status of propositions with regard to their
discrete components, pointing out that a syllogism is an integrated mental act and
that it ought therefore to be thought of not in a merely mechanical manner:
       It is [. . . ] altogether erroneous to maintain, as is commonly done,
       that a reasoning or syllogism is a mere decompound whole, made up
       of judgments; as a judgment is a compound whole, made up of con-
       cepts. This is a mere mechanical mode of cleaving the mental phe-
       nomena into parts; and holds the same relation to a genuine analysis
       of mind which the act of the butcher does to that of the anatomist.
       It is true, indeed, that a syllogism can be separated into three parts
       or propositions; and that these propositions have a certain meaning,
                          The Logic of Sir William Hamilton                       155
      when considered apart, and out of relation to each other. But when
      thus considered, they lose the whole significance which they had when
      united in a reasoning; for their whole significance consisted in their
      reciprocal relation,–in the light which they mutually reflected on each
      other. We can certainly hew down an animal body into parts, and con-
      sider its members apart; but these, though not absolutely void of all
      meaning, when viewed singly and out of relation to their whole, have
      lost the principal and peculiar significance which they possessed as the
      coefficients of a one organic and indivisible whole. It is the same with
      a syllogism. The parts which, in their organic union, possessed life and
      importance, when separated from each other, remain only enunciations
      of vague generalities, or of futile identities. Though, when expressed
      in language, it be necessary to analyse a reasoning into parts, and to
      state these parts one after another, it is not to be supposed that in
      thought one notion, one proposition, is known before or after another;
      for, in consciousness, the three notions and their reciprocal relations
      constitute only one identical and simultaneous cognition. (LL.I.275-6).
The notion of interrelation and simultaneity in the above is important to how
Hamilton will go on to view the syllogism’s structure as, for example, something
in which it is only mere convention that always places the conclusion last, and
that the relative positions of major and minor premise themselves can easily be
switched around without any loss of meaning. However, I have quoted the above
passage at length since it eloquently apprises us of an important general dimension
of Hamilton’s whole approach to Logic — for all that he is striving for a rigorously
pure science free from the extra-logical and although he will later in the ‘New
Analytic’ speak of a ‘Symbolic Notation [that will display] the propositional and
syllogistic forms, even with a mechanical simplicity’, Hamilton constantly opposes
the mechanical or rigidly structural in favour of considering the syllogism less like
a dismembered material body or constructed building and more like a process —
multiplex, organic, interrelated, and even fluid — and his logic needs to be seen
as an attempt to capture, not only the dual perspectives afforded by extension
and intension, but also a greater sense of the richness and complexity of formal
reasoning (NA.LL.II.251).
   In the previous section I referred to Hamilton’s ‘Observations on the Mutual
Relation of Syllogistic Terms in Quantity and Quality’ at Appendix V (e) in the
Lectures on Logic (see NA.LL.II.285). In this ‘Observations’ section, he provides
the relations between any given proposition’s subject and predicate terms and,
by inverting the order of the negative propositional forms given in the previous
section in Table 2, he displays the best-worst quantification relationships between
each of the four affirmative propositions and their corresponding negatives (see
NA.LL.II.286). I have reconfigured Hamilton’s presentation of these relations as
follows:
156                                      Ralph Jessop
The significance of this schema becomes clear when we note how Hamilton’s system
of quantification brings into consideration certain aspects of syllogistic reasoning
formerly ignored by Aristotelian logic. According to Hamilton:
      Former logicians knew only of two worse relations,–a particular, worse
      than a universal, affirmative, and a negative worse than an affirma-
      tive. As to a better and worse in negatives, they knew nothing; for
      as two negative premises were inadmissible, they had no occasion to
      determine which of two negatives was the worse or better. But in
      quantifying the predicate, in connecting positive and negative moods,
      and in generalising a one supreme canon of syllogism, we are compelled
      to look further, to consider the inverse procedures of affirmation and
      negation, and to show [. . . ] how the latter, by reversing the former,
      and turning the best quantity of affirmation into the worst of negation,
      annuls all restriction, and thus apparently varies the quantity of the
      conclusion. (NA.LL.II.285-6).
I shall not attempt to explain this in detail. Suffice to say that the above schema
of best-worst quantification relationships is used by Hamilton to construct what
he calls his ‘General Canon’ to determine the relationship between the subject (S)
and predicate (P) of the conclusion in a syllogism, on the basis of a best-worst
comparison between the relationships between the subject term and middle term
(M), and the middle term and predicate term of the syllogism. Hence, depending
on the relationship between the subject and predicate terms of a syllogism, the
relationship between S and P in the conclusion will be, for example, totally coex-
clusive, or partially coexclusive, or toto-partially coexclusive, and so on. Hamilton’s
‘one supreme canon of syllogism’ is given as follows:
      General Canon. What worst relation of subject and predicate, subsists
      between either of two terms and a common third term, with which
      one, at least, is positively related; that relation subsists between the
      two terms themselves. (NA.LL.II.285)
He translates this ‘General Canon’ into twelve clear rules governing 36 syllogistic
moods in the first three figures (the fourth figure having been rejected as pointed
                          The Logic of Sir William Hamilton                          157
out above), plus 24 negative moods — these forms are detailed in the final table
given at the end of the Lectures on Logic (see LL.II.475). However, it might be
better to think of these not as rules but rather as instantiations of his General
Canon since, once one has grasped how Hamilton’s system operates, each of them
can be translated from this Canon with relative ease (see NA.LL.II.285-9). I shall
not discuss these instantiations in any detail but will instead use just one of them
to illustrate how Hamilton’s General Canon operates, using an example which
Fogelin uses to demonstrate the correctness of his reading of propositions (1) and
(8) as discussed in the previous section.
   Fogelin gives the following syllogism as an example that he claims Hamilton’s
system allows as valid, yet which can be shown, using a Venn diagram, to be
invalid if translated using his earlier translation of proposition (8), though valid
using de Morgan’s interpretation of this parti-partial coexclusion proposition:
       Some P is not any M
       Some S is some M
  ∴    Some S is not some P
Having already accepted Fogelin’s treatment of the above syllogism which shows
how it can be read as valid (and hence what proposition (8) expressing Parti-
partial coexclusion means, or can be read as meaning), all I want to do here is
merely illustrate how Hamilton’s General Canon can be invoked to produce from
the two premises the conclusion as given. I think that all this will demonstrate is
that Fogelin is right that the above syllogism would be declared valid in Hamilton’s
system as he asserts. However, I am not entirely sure that Hamilton’s table and list
of interpretations of his General Canon are completely free from anomalies, and it
must be said that to check this thoroughly, given the complexities of co-ordinating
various mixtures of extensive and intensive readings with Hamilton’s own symbolic
representation of the negative moods, is beyond the scope of this chapter. Hence,
I shall simply give Fogelin’s example as unequivocally exemplifying adherence to
Hamilton’s General Canon, even though by treating this so simplistically I may be
overlooking certain niceties and complexities concerning how it ought to be used:
  Fogelin’s Example:        Relation:                     Hamilton’s rule from
                                                          General Canon:
  some P is not any M       Parti-total coexclusion       VII.a: A term
  some S is some M          Parti-partial conclusion      parti-totally coexclusive,
  some S is not some P      Parti-partial coexclusion     and a term partially
                                                          coinclusive, of a third
                                                          [M], are partially
                                                          coexclusive of each other.
                                                          (NA.LL.II.288)
   From what I have been so far able to ascertain, though not articulate in full,
certain possible anomalies (or doubts that I still have) notwithstanding, Hamilton’s
General Canon does seem to work.
158                                    Ralph Jessop
   If, in agreement with Fogelin, I am right about this, Hamilton’s system marks a
most significant advancement in the treatment of the syllogism that almost entirely
replaces the previous systems which Hamilton so often decried as being imperfect,
flawed, confused, and misleading. No wonder that Fogelin, though acknowledging
that Hamilton’s achievement may not seem important ‘from a modern point of
view’, regards his quantification system as ‘a radical departure from traditional
theory’ (Fogelin, pp. 163–4). If Hamilton’s ultimate regulation or General Canon
of syllogistic forms is fully adequate, comprehensive, sufficiently general, and flex-
ible, his treatment of the syllogism arguably marks the considerable improvement
on and replacement of traditional logic’s treatment of the syllogism that he himself
believed it did. In place of numerous logical rules and botched attempts to bring
the Aristotelic system out of the chaos in which Hamilton found it, by judiciously
interpreting but not slavishly falling under the spell of Aristotle’s enormous au-
thority, Hamilton may well have developed, albeit within certain limitations, a
system of formal logic that, as it culminated in a simple mechanism for making
all quantities in both affirmative and negative propositions explicit, only required
the capstone of his single or supreme canon of the syllogism, to warrant his grand
claim that he had produced a system whose beauty resided in the very naturalness
of being ‘as One at once and Various.’
                                  6  CONCLUSION
       A mere knowledge of the rules of Rhetoric can no more enable us to
       compose well, than a mere knowledge of the rules of Logic can enable
       us to think well. (LL.I.48-9)
Augustus de Morgan misunderstood the complexities that Hamilton’s system could
either accommodate or was pointing towards. As Fogelin claims, though de Mor-
gan clearly did try to understand Hamilton’s system ‘he failed to do so and com-
mentators since have hardly done better’ (Fogelin, p. 162). According to Fogelin
this has much to do with the strangeness of Hamilton’s language, such as we find in
his non-standard exchanges of ‘all’ for ‘any’. True, these things do create problems,
as also do Hamilton’s now strange symbolism, dense tables, and the fragmentary
nature of much of his work on Logic, excepting the Lectures which are abundantly
clear, always informative, and even occasionally somewhat entertaining. How-
ever, I suspect that the linguistic difficulties have more to do with Hamilton’s
enormous ability to compress the language of his texts into an often overly taut,
quasi-litigious style that is at first off-putting and certainly at times not for the
faint-hearted. However, there are some more substantial reasons why de Morgan
and others have floundered and in the end did Hamilton a great disservice.
   In order to gain more than a foothold in Hamilton’s logic it would seem that
one also has to have a foothold in and possibly be to some extent persuaded by
his metaphysical standpoint of natural dualism, for in this doctrine’s opposition
to absolutism — not least of all in its opposition to absolute scepticism — inheres
                          The Logic of Sir William Hamilton                       159
Hamilton’s relativity and the germ of what becomes a much more pervasive cor-
relativism throughout so much of his writing. Hamilton’s relativity places subject
and predicate into simple relations of equation and non-equation, distinguishes
indefinite quantities into two kinds or degrees of indefiniteness, and is founded
on recognising the significance of perspective with regard to a given concept’s
dual quantification dimensions of extension and intension. All this seems to be
an attempt to frame subject and predicate as the unity of thought from which
they originate. However, the frame is really much larger, and herein lies a great
part of the problem and yet potential of Hamilton’s system. A part of Hamil-
ton’s relativism also brings into play not just perspective (and thereby different
possible readings of propositions), but both the relations between a concept’s at-
tributes and the conditioning nature of the human subject’s agency. The relations
between logic (the laws that constitute the conditions of thought) and language
(the terms without which we would otherwise be incapable of participating in any
productive thought, argument, discourse, analysis, or articulation of logic’s laws,
propositional forms, and so on), and the surrounding chaos or ever-shifting sands
against, and yet in relation to which, logic and language comprise our attempts
to make whole, divide, and recombine this otherwise unintelligible plenum of in-
determinate entities under which we are continually at risk of being submerged
— as these relations form the frame or field within which Hamilton’s logic is con-
ducted, he was constructing a logical system both highly suggestive of a deeper
relativism he eschewed, and yet which he virtually postulated was the condition
within which logic had to operate as the very function of logic and language had
to do with stabilising the multiple within unity. This, perhaps more than anything
else, makes his writing complex. However, much more simply, Hamilton’s enor-
mous erudition — also a factor in the diversity he attempts to bring into order
— remains an intimidating force that few, excepting some of his more devoted
students and followers, have found a congenial companion.
   Hamilton regarded philosophy — particularly logic and metaphysics — as the
greatest gymnastic of the mind and with such a conception of his subject matter
providing a large part of philosophy’s raison d’ˆtre, he was attempting to set his
                                                    e
students on a course of study that would challenge their intellectual abilities to
their utmost through his attempts to make the study of Logic measure up to and
(it must have been his hope) compete with the higher educational, intellectual,
and scholarly standards that had been demanded on the Continent. The Ger-
man logicians in particular, made even the Professors of Britain’s one-time most
prestigious university, Oxford, appear to Hamilton’s eyes as little better than in-
tellectual philistines or dilettantes. In his definition of Logic as a pure science of
the necessary laws of thought, single-handedly Hamilton was laying the ground-
work for improving upon and in many ways replacing the traditional Aristotelian
logic that he regarded as having been sunk into a mass of confusion by centuries
of almost slavish or insufficiently critical adherence to the Stagirite, coupled with
numerous misinterpretations of the true nature of their subject.
   Within the boldness of Hamilton’s emphatic assertions and robust denunciations
160                                        Ralph Jessop
of the errors and confusions that he argues arose due to ‘the passive sequacity of
the logicians [in following] obediently in the footsteps of their great master’, Aris-
totle, there is a genuine sense of his excitement concerning this claim that his
system supersedes all previous systems new and old (NA.LL.II.262). By this stage
in his work on logic it must have seemed to Hamilton as though, after many long
years of arduous industry during which he had been diligently examining, sum-
marising, and critically assessing the logics of others written mainly in Greek,
Latin, English, and German, at last the numerous errors and confusions riddling
centuries of Aristotelian logic could be removed. Through the earlier works of
ancient and scholastic logicians, these errors and confusions had continued to sift
themselves, but they also permeated the work of his Oxonian contemporaries.
But now, after years of clearing away the detritus of former ages and chastising
the dilettantism of more recent logicians, he could with one final push radically
brush aside those aspects of the traditional logic that had significantly impeded
its development or evolution into a system at once more complex and yet more
orderly (see NA.LL.II.252). At last he had constructed a robust system that he
could proudly advocate as the ‘keystone in the Aristotelic Arch’, and while this
keystone had doubled the number of propositional forms, these forms constituted
a structural whole based on the logic of Aristotle and thus at once the tradi-
tional logic was brought one significant step closer to the beauty and perfection
which Aristotle’s work seemed to promise but had not realised (see NA.LL.II.249).
However, Fogelin rightly counters Hamilton’s claim that his ‘New Analytic’ com-
pletes traditional Aristotelian logic — for, instead of merely placing the keystone
in the Aristotelic arch, ‘By introducing an entirely different system of classifying
propositions in virtue of their potential roles in syllogisms, Hamilton made a rad-
ical departure from traditional theory’ (Fogelin, p. 163). The keystone that is
Hamilton’s quantification of the predicate one might thus say, is far from being a
pretentious claim to glory undeserved — rather, Hamilton’s ‘keystone’ is an overly
modest description for a much more radical, yet substantial and more stable con-
struction that involved the destruction and removal of logical ruins from bygone
ages, though it is plain to see that at least some of the foundations were incorpo-
rated to support Hamilton’s arch. But with a final and fitting twist of irony, all
too quickly the course of logic would develop during the 19th century in ways that
left so much of Hamilton’s endeavour far behind as a curiously flawed relic.
                                     BIBLIOGRAPHY
 [Baynes, 1850] Thomas Spencer Baynes. An Essay on the New Analytic of Logical Forms, Being
   that which Gained the Prize Proposed by Sir William Hamilton, in the Year 1846, for the
   Best Exposition of the New Doctrine Propounded in his Lectures; with an Historical Appendix
   (Edinburgh: Sutherland and Knox; London: Simpkin, Marshall, and Co., 1850)
 [Benchimol, 2004] Alex Benchimol. William Hamilton. In The Carlyle Encyclopedia, edited by
   Mark Cumming (Cranbury, NJ:Farleigh Dickinson University Press, 2004), 207-9
 [Dallas, 1866] E. S. Dallas. The Gay Science (London: Chapman and Hall, 1866)
 [Davie, 1961] George Elder Davie. The Democratic Intellect: Scotland and Her Universities in
                             The Logic of Sir William Hamilton                                 161
   the Nineteenth Century, Edinburgh University Publications, History Philosophy & Economics:
   12 (Edinburgh: Edinburgh University Press, 1961; repr. 1982)
[de Morgan, 1966] Augustus de Morgan. On the Syllogism and Other Logical Writings, edited
   with an introduction by Peter Heath (London: Routledge & Kegan Paul, 1966)
[Flint, 1903] Robert Flint. Agnosticism (Edinburgh and London: Wm Blackwood, 1903)
[Fogelin, 1992] Robert J. Fogelin. Philosophical Interpretations (Oxford: Oxford University
   Press, 1992)
[Fraser, 1865] A. C. Fraser. Mill’s examination of Sir William Hamilton’s Philosophy. North
   British Review, 43, 1–58, September 1865.
[Fraser, 1860] A. C. Fraser. Province of logic and recent British logicians. North British Review,
   33, 401–427, November 1860.
[Gilley and Loades, 1981] S. Gilley and A. Loades. Thomas Henry Huxley: The War between
   Science and Religion, The Journal of Religion, 61 (1981), 285-308
[Hamilton, 1853] Sir William Hamilton. Discussions on Philosophy and Literature, Educa-
   tion and University Reform, 2nd edn (London: Brown, Green and Longmans; Edinburgh:
   MacLachlan and Stewart, 1853)
[Hamilton, 1860] Sir William Hamilton. Lectures on Logic, ed. H. L. Mansel and J. Veitch.
   (Edinburgh and London: Blackwood, 1860).
[Hamilton, 2001] Sir William Hamilton. Works of Sir William Hamilton, with an introduction
   by Savina Tropea, 7 vols (Bristol: Thoemmes Press, 2001)
[Hamilton, 1860] Sir William Hamilton. Lectures on Logic, ed. by H.L. Mansel and John Veitch
   (Edinburgh and London: Blackwood, 1860)
[Hamilton, 1990] Thomas Hamilton. The Youth and Manhood of Cyril Thornton, ed. by Maurice
   Lindsay (Aberdeen: The Association for Scottish Literary Studies, 1990)
[Hendry, 1986] John Hendry. James Clerk Maxwell and the Theory of the Electromagnetic Field
   (Bristol and Boston: Adam Hilger, 1986)
[Herbert, 2001] Christopher Herbert. Victorian Relativity: Radical thought and Scientific Dis-
   covery (Chicago and London: The University of Chicago Press, 2001)
[Jessop, 2005] Ralph Jessop. Carlyle’s Agnosticism: An Altar to the Unknown and Unknowable
   God, Literature and Belief, 25 (1&2), 381-433, 2005.
[Jessop, 1997] Ralph Jessop. Carlyle and Scottish Thought (Basingstoke: Macmillan, 1997)
[Kneale and Kneale, 1962] William Kneale and Martha Kneale, The Development of Logic (Ox-
   ford: Clarendon, 1962)
[Kneale, 1948] William Kneale. Boole and the Revival of Logic, Mind, 57 (Apr. 1948), 149-75,
   1948.
[Kuehn, 1990] Manfred Kuehn. Hamilton’s Reading of Kant: A Chapter in the Early Scottish
   Reception of Kant’s Thought, in George MacDonald Ross and Tony McWalter, eds, Kant and
   His Influence (Bristol: Thoemmes Antiquarian Books, 1990), 315-347
[Laita, 1979] Luis M. Laita. Influences on Boole’s Logic: The Controversy between William
   Hamilton and Augustus De Morgan, Annals of Science, 36 (1979), 45-65
[Lightman, 1987] Bernard Lightman. The Origins of Agnosticism: Victorian Unbelief and the
   Limits of Knowledge (Baltimore and London: John Hopkins University Press, 1987)
[Mansel, 1851] H. L. Mansel. Recent Extensions of Formal Logic, North British Review, 15 (May
   1851), 90-121
[Mansel, 1866] H. L. Mansel. The Philosophy of the Conditioned: Sir William Hamilton and
   John Stuart Mill, Contemporary Review, 1 (1866), 31-49; 185-219
[Mansel, 1851] H. L. Mansel. Prolegomena Logica: An Enquiry into the Psychological Character
   of Logical Processes (Oxford: William Graham, 1851)
[Mansel, 1858] H. L. Mansel. The Limits of Religious Thought (London : John Murray, 1858)
[Masson, 1877] David Masson. Recent British Philosophy: A Review with Criticisms including
   some Comments on Mr Mill’s Answer to Sir William Hamilton, 3rd edn (London: Macmillan,
   1877)
[Memoir, 1869] Memoire of Sir William Hamilton London: Sampson Low, Marston, Sarl &
   Rivington, 1881)
[Mill, 1865] John Stuart Mill. An Examination of Sir William Hamilton’s Philosophy and of
   the Principal Philosophical Questions Discussed in his Writings (London: Longman, Green,
   Longman, Roberts & Green, 1865)
[Monck, 1881] W. H. S. Monck. Sir William Hamilton (London: Sampson Low, Marston, Searle
   & Rivington, 1881)
162                                        Ralph Jessop
[Napier, 1877] Macvey Napier, Jnr, ed. Selections from the Correspondence of the Late Macvey
   Napier (London: Harrison, 1877)
[Olson, 1971] Richard Olson. Scottish Philosophy and Mathematics 1750–1830, Journal of the
   History of Ideas, 32 (1971), 29-44
[Ouren, 1973] Dallos Victor Lie Ouren. HaMILLton: Mill on Hamilton: A Re-examination of Sir
   William Hamilton’s Philosophy (unpublished doctoral thesis, University of Minesota, 1973)
[Reid, 1846] Thomas Reid. The Works of Thomas Reid, preface, notes, and supplementary
   dissertations by Sir William Hamilton (Edinburgh: MacLachlan, Stewart; London: Longman,
   Brown, Green and Longmans, 1846)
[Stewart, 1854-60] Dugald Stewart. The Collected Works of Dugald Stewart, ed. by Sir William
   Hamilton, 11 vols (Edinburgh: Thomas Constable, 1854-60)
[Thomson, 1854] William Thomson. An Outline of hte Necessary Laws of Thought: A Treatise
   on Pure and Applied Logic, 3rd edition (London: Longman, Brown, Green and Longmans,
   1854)
[Veitch, 1882] John Veitch. Hamilton (Edinburgh And London: Blackwood, 1882)
   “THE WHOLE BOX OF TOOLS”: WILLIAM
WHEWELL AND THE LOGIC OF INDUCTION
                                     Laura J. Snyder
    William Whewell (1794–1866), well-known polymath, has been criticized for his
view of inductive logic ever since publishing it. His friend, the logician Augustus
DeMorgan, may or may not have intended a cruel reference to Whewell’s lowly
social origins as the son of a master carpenter when he complained, in his 1847
textbook on logic, about certain writers incorrectly using the term “induction” as
including “the whole box of [logical] tools.” Other writers on logic, from John Stu-
art Mill on, have similarly claimed that Whewell did not present a valid inductive
logic. Mill accused Whewell of allowing “no other logical operation [in scientific
discovery] but guessing until a guess is found that tallies with the facts”. In the
twentieth century, followers of Karl Popper made that criticism a compliment,
claiming that Whewell had “anticipated” Popper’s anti-inductive methodology of
conjectures and refutations. Whewell’s response to later criticisms of DeMorgan
is, I believe, as apt as it was then: “I do not wonder at your denying [my induc-
tion] a place in Logic; and you will think me heretical and profane, if I say, so
much the worse for logic”.1 As I will show in this chapter, Whewell’s method of
discoverers’ induction is, as he always claimed, an inductive logic, one that was
strongly influenced by Whewell’s reading of Francis Bacon. Though Whewell’s
inductive logic does include the “whole box of tools”, this is a benefit, rather than
a liability, of his view.
                   1   WHEWELL’S EDUCATION AND CAREER
Whewell was born in 1794 in Lancaster, England, the eldest child of a master
carpenter. His prodigious abilities were recognized early. As a child, the master
of the local grammar school saw his potential and offered to teach him for free
at his facility. Soon after, Whewell was examined by a tutor from Trinity Col-
lege, Cambridge, who predicted that the young boy would one day place among
the top six wranglers (the holders of first-class degrees). Whewell was then sent
to the Heversham Grammar School in Westmorland, some twelve miles to the
north, where he would be able to qualify for a closed exhibition to Trinity. In the
nineteenth century and earlier, these “closed exhibitions” or scholarships were set
aside for the children of working class parents. Whewell studied at Heversham
   1 William Whewell to Augustus DeMorgan, January 18, 1859, in [Todhunter, 1876, 2: 417].
Handbook of the History of Logic. Volume 4: British Logic in the Nineteenth Century.
Dov M. Gabbay and John Woods (Editors)
 c 2008 Elsevier BV. All rights reserved.
164                                      Laura J. Snyder
for two years, and received private coaching in mathematics from John Gough,
the blind mathematician to whom reference is made in Wordworth’s “Excursion”.
Although Whewell did win the exhibition it did not provide full resources for a
boy of his family’s means to attend Cambridge; money had to be raised in a public
subscription to supplement the scholarship money. He thus came up to Trinity in
1812 as a “sub-sizar” (scholarship student).2
   When he arrived at Trinity, Whewell became involved with the Analytical So-
ciety, formed the year before by Charles Babbage, John Herschel and others. The
Analytical Society had been founded for the purpose of reforming the teaching
and practice of mathematics at Cambridge. The members of the society wished to
introduce continental methods of mathematics to a university which had excluded
progress since the time of Newton (specifically, they hoped to introduce Lagrange’s
algebraic and formalistic version of the calculus, which included replacing Newton’s
fluxion dot notation with Leibniz’s “d” notation) (see [Fisch, 1994] and [Becher,
1980].) In addition to his involvement with the Analytical Society, Whewell’s years
as a Cambridge undergraduate were enriched by his friendships with Richard Jones
(the future political economist) and Herschel, each of whom had great effect on his
future thinking about science. He also formed friendships with Hugh James Rose,
Julius Charles Hare, and Connop Thirlwall, men greatly interested in and influ-
enced by the German Romantic movement. Whewell finished his studies in 1816
as Second Wrangler and Second Smith’s Prizeman. After a short post-graduate
period as a private tutor, Whewell won a college fellowship in 1817.
   After winning his fellowship, Whewell was assistant tutor from 1818 until 1823,
and tutor from 1823 until 1838. He was elected to the Royal Society in 1820, and
ordained a priest — as required for Trinity Fellows to maintain their position after
an initial seven year period — in 1825. He took up John Henslow’s Chair in Miner-
alogy in 1828 (when Henslow vacated it for the Botany Professorship), and resigned
it in 1832. During the early 1830s Whewell was instrumental in the formation of
the British Association for the Advancement of Science (BAAS), the Statistical
Section of the BAAS, and the Statistical Society of London. In 1833, at a meeting
of the BAAS, he invented the word “scientist”, an achievement emblematic of his
broader influence upon science and culture. In 1838 Whewell became Professor
of Moral Philosophy. Shortly after his marriage to Cordelia Marshall on 12 Oc-
tober 1841, he was named Master of Trinity College, having been recommended
to Queen Victoria by Prime Minister Robert Peel. He was Vice-Chancellor of the
University in 1842 and again in 1855. In 1848 he played a large role in establishing
the Natural and Moral Sciences Triposes at the University. Whewell engaged in
scientific research, winning a medal from the Royal Society for his work on the
science of the tides. He corresponded with the most eminent scientists of his day.
He published his own translations of German novellas and poetry, and, with his
popular translations of Plato’s dialogues, was partly responsible for a “Platonic
revival” in Britain in the 1850s (see [Turner, 1981, 371–2]). Towards the end of
   2 Whewell’s status as a sub-sizar meant that he had to serve the other boys at dinner. See
[Garland, 1980, 11], and [Rothblatt, 1981].
           “The Whole Box of Tools”: William Whewell and the Logic of Induction               165
his life Whewell was asked to tutor the young Prince of Wales (the future Edward
VII) in political economy.
   As this brief summary of his life indicates, Whewell’s rise from what we would
probably call the middle (or lower-middle) class to the intellectual and social elite
was spectacular. If this rise tended to make him a bit arrogant, and even imperious,
in his later life, it also I think imbued him with a sense of mission. Whewell very
much saw himself as having a “calling” to reform philosophy; he refers to this as
his “vocation”.3 As we will see in the next sections, Whewell’s project, throughout
his career, was to reform inductive philosophy, and to do so by revising Francis
Bacon’s inductive logic.
                              2   REFORMING INDUCTION
From his days as an undergraduate at Trinity College, to his final years spent
as Master of that institution, Whewell considered his project to be the reform
of the inductive philosophy, a reform that was intended to apply to all areas
of knowledge. Whewell intended that his reformed inductive philosophy would
provide the groundwork for the reshaping of more than natural science: morality,
politics, and economics would also be transformed. In this chapter, however, I will
focus only on Whewell’s renovation of induction (details about how he intended
this reform to be applied to the other areas can be found in my [Snyder, 2006]).
   Whewell and his friend Richard Jones together planned to reform induction,
and saw their task as consisting in two parts. The first was defining a “true idea
of induction”. In an early notebook entry, Whewell lamented “that the true idea
of induction has not been generally fixed and agreed upon must I think be very
obvious”.4 The second task was “to get the people into a right way of thinking
about induction”;5 that is, to publicize the nature and value of induction in all
areas of thought. In one of his many letters to Jones, Whewell refers to induction
as the “true faith”, wondering how it can “best be propagated”.6
   Whewell and Jones framed their reforming mission as a battle against those
they referred to as the “downwards mad”, that is, those who preferred a deductive
approach to the logic of the natural and moral sciences.7 One of these “deductive
savages” was Richard Whately, fellow of Oriel College, Oxford, who became arch-
bishop of Dublin in 1831.8 Early that year Jones wrote to Whewell after seeing
   3 In his 1831 review of Herschel’s Preliminary Discourse, Whewell referred to “the history and
philosophy of physical science” as his “vocation” (see [Whewell, 1831, 374]). Later, he wrote to
Herschel that “the reform of our philosophy is the work to which I have the strongest vocation,
and which I cannot give up if I were to try” (letter to Herschel, 9 April 1836, in [Todhunter,
1876, 2: 234]).
   4 See notebook dated 28 June 1830, WP R.18.17 f. 12, pp. v–ix.
   5 William Whewell to Richard Jones, 25 February 1831, WP Add.Ms.c.51 f. 99.
   6 William Whewell to Richard Jones, 25 February 1831, WP Add.Ms.c.51 f. 99.
   7 See William Whewell to Richard Jones, 20 January 1833, WP Add.Ms.c.51 f. 149; and 22
July 1831, Add.Ms.c.51 f. 110.
   8 William Whewell to Richard Jones, 19 February 1832, WP Add.Ms.c.51 f.129.
166                                        Laura J. Snyder
the third edition of Whately’s Elements of Logic, objecting to Whately’s “strange
notion” that induction was a type of deductive reasoning.9 Indeed, Whately had
claimed that all forms of reasoning could be assimilated to the syllogism [Whately,
1827, 207]. Induction, for example, was said to be reasoning in which the major
premise, which is generally suppressed, can be expressed as “what belongs to the
individual or individuals we have examined, belongs to the whole class under which
they come” [Whately, 1827, 209]. Thus, to take Whately’s example, if we find,
from an examination of the history of several tyrannies, that each of them lasted
a short time, we conclude that “all tyrannies are likely to be of short duration”.
In coming to this conclusion we make use of a suppressed major premise, namely,
“what belongs to the tyrannies in question is likely to belong to all” [Whately,
1827, 209]. Whately admitted that some would complain that his notion of in-
duction was too narrow, in that it did not account for how the minor premises
are obtained; i.e., for how it was ascertained that each of the examined tyrannies
were “short-lived”. But he distinguished between “logical discoveries”, which oc-
cur when syllogistic reasoning alone is used to deduce a conclusion from known
premises, and “Physical Discoveries”, which involve more than syllogistic reason-
ing because various methods are used for ascertaining the premises — including
observation, experiment, the selection and combination of facts, abstraction of
principles, and others [Whately, 1827, 234–6]. On Whately’s view logic was con-
cerned only with reasoning from premises, not with ascertaining the premises.
Thus inductive logic was also only concerned with reasoning from premises, or
what Whately called logical, as opposed to physical, discovery.10
    Whewell and Jones saw Whately’s characterization of induction as more than
just a point of logic. Rather, it presented a potential obstacle for the reform
of philosophy they sought. They realized that many people associated induction
with scientific method, even if they were unclear about the precise meaning of
the term. If people accepted Whately’s definition of induction, they might be led
to the erroneous conclusion that science is essentially deductive, concerned only
with deducing conclusions from assumed “first principles”. Whewell and Jones
linked this deductive view of science with that of the scholastic Aristotelians;
indeed, Whately had explicitly framed his work as a defense of Aristotelian logic
against the “confused” views of induction that resulted from its connection with
Bacon.11 But Jones and Whewell were eager to endorse a renovated version of
   9 Richard Jones to William Whewell, 24 February 1831, WP Add.Ms.c.52 f.20.
  10 As Pietro Corsi notes, Whately thus distinguished between two senses of induction: one
having to do with the collection of facts, the other being a form of inference. The first was not
of concern to Whately. Rather, his aim was to demonstrate that the second was “under the
dominion of logic” [Corsi, 1988, 151]. This aim arose out of the context of attacks on the Oxford
curriculum from those, such as John Playfair, who argued that progress in the mathematical
sciences had been impeded by the reliance on Aristotelian logic. Whately and his teacher,
Edward Copleston, defended the curriculum and the prominence of Aristotelian logic within it
(see also [McKerrow, 1987]).
  11 See [Whately, 1827, ix and 9] and [McKerrow, 1987, 172]. In a letter to Jones, Whewell
claimed that Whately and his followers were even worse than Aristotle, “far more immersed in
verbal trifling and useless subtilty” (7 April 1843, WP Add.Ms.c.51 f. 227). Pietro Corsi has
          “The Whole Box of Tools”: William Whewell and the Logic of Induction                167
Baconian inductivism. From their undergraduate days, they referred to Bacon
as the “Master”, and framed their task in terms of revising and popularizing his
inductive logic. Although his view was not without flaws, it was the closest thing
to a true inductive logic. Certainly, it was a vast improvement over the logic of
the scholastic Aristotelians.
   Because of their emphasis on the syllogism, Whewell complained, the scholas-
tics “talked of experiment” but “showed little disposition to discover the truths
of nature by observation of facts” [Whewell, 1860, 48]. In a notebook from 1830,
Whewell wrote of the Aristotelian method that such a method “could lead to no
such truths, and in the development of physical science especially was entirely
barren. . . . The business of speculative men became, not discovery, but argumen-
tation”.12 Later, in his History of the Inductive Sciences, Whewell would refer to
the Middle Ages as a “stationary period” in science. Whewell and Jones believed
that the correct, Baconian view of induction needed to be brought before the pub-
lic in order to prevail against sterile deductive approaches to scientific knowledge.
Whewell expressed his “confidence” to Jones that “by and by the whole world will
think [the deductive definition of science] as nonsensical as we do”. But before
this could happen, he and Jones would need to spread the “true faith”.13
               3   WHEWELL’S ANTITHETICAL EPISTEMOLOGY
The early influence of Bacon strongly inclined Whewell toward an inductive, em-
pirical view of epistemology. At the same time, however, Whewell had a deep
appreciation for a priori, deductive forms of reasoning, due to his interest in math-
ematical sciences. His early work in mechanics exposed him to a physical science
which seemed to incorporate both empirical and a priori elements.14 But it was
his experience studying mineralogy abroad that struck Whewell with the need to
combine empirical and a priori elements in an epistemology and scientific method-
ology. In 1825, when Henslow vacated the chair of Mineralogy at Cambridge to
take up the Botany Professorship, Whewell announced himself a candidate for
the position. Although he had published mathematical papers on crystallogra-
phy, Whewell did not have much empirical knowledge of mineralogy; thus he went
abroad to study with experts such as Friedrich Mohs. (This was not unusual at
the time. When Adam Sedgwick was elected to the Woodwardian professorship of
geology in 1818, he knew little of the discipline and needed to learn it quickly) (see
noted that Whately and the “Oxford Noetics” used Stewart’s writings in defense of Oxford’s
educational system, with its emphasis on Aristotelian logic, a position opposed to Stewart’s own
[Corsi, 1987].
  12 William Whewell, Notebook, WP R.18.17 f. 12, p. 94. See also a notebook dating from
1831-2 (WP R.18.17 f. 15, p. 24). Thirty years later, in the final edition of the Philosophy of the
Inductive Sciences, Whewell criticized Descartes on similar grounds. See [Whewell, 1860, 163].
  13 William Whewell to Richard Jones, 25 February 1831, WP Add.Ms.c.51 f. 99.
  14 Whewell’s first book was a textbook on dynamics, published in 1819. He went on to produce
four more textbooks on mechanics, a book on analytical statics, and a text on the differential
calculus.
168                                           Laura J. Snyder
[Garland, 1980, 96]). Whewell was strongly impressed with the “German school”
in mineralogy, especially with the elegant mathematical treatment given by Mohs
of a science which Whewell had previously considered to be purely empirical. He
wrote to his friend Hugh James Rose that “I am afraid . . . that I may not bring
back my faith as untainted as you have done: for I find my mineralogical super-
naturalismus giving way in some respects. It may be possible to bring about a
union between the two creeds [i.e., the a priori and the empirical], which I hope
will not be such a thing in science as you hold it to be in faith”.15
   Whewell eventually reconciled the empirical and a priori elements of science
in an epistemology that is “antithetical”, in that it expressed what Whewell
called the “Fundamental Antithesis”, or dual nature, of knowledge.16 Accord-
ing to Whewell’s mature epistemology, all knowledge involves both an ideal, or
subjective, element, as well as an empirical, or objective, element. Although his
experience with mineralogy had sparked Whewell’s desire to “bring about a union
between the two creeds” of empiricism and a priorism, Whewell’s notebook writ-
ings on induction prior to 1831 present induction as a purely empirical process,
consisting in enumerative induction of observed instances.17 He had not, up to this
point, found a way to synthesize the a priori elements of scientific knowledge with
the empirical epistemology he wished to follow Bacon in endorsing. By February
of 1831, while working on his review of John Herschel’s Preliminary Discourse
on the Study of Natural Philosophy, Whewell for the first time used a metaphor
indicating his initial attempt to combine these empirical and a priori or purely
rational elements of science. Whewell explained that
       Induction agrees with mere Observation in accumulating facts, and
       with Pure Reason in stating general propositions; but she does more
       than Observation, inasmuch as she not only collects facts, but catches
       some connexion or relation among them; and less than pure Reason
       . . . because she only declares that there are connecting properties,
       without asserting that they must exist of necessity and in all cases.
       If we consider the facts of external nature to lie before us like a heap
       of pearls of various forms and sizes, mere Observation takes up an
       indiscriminate handful of them; Induction seizes some thread on which
       a portion of the heap are strung, and binds such threads together.
       [Whewell, 1831, 379]
   He asked Jones to look over his draft.18 Jones’s assessment proved to be valuable
for Whewell:
  15 William Whewell to H.J. Rose, 15 August 1825, in [Todhunter, 1876, 2: 60]. See also [Fisch,
1991, 67].
  16 Whewell’s epistemology is referred to as “antithetical” in [Fisch, 1985] and [Fisch, 1991].
  17 See especially the earliest draft of Philosophy, dated 28 June 1830, where his view of induction
sounds much like the view expressed later by Mill in his System of Logic, a view that Whewell
strongly criticized (WP R.18.17 f. 12, 98, 104).
  18 William Whewell to Richard Jones, 11 February 1831, WP Add.Ms.c.51 f. 98.
           “The Whole Box of Tools”: William Whewell and the Logic of Induction          169
        I go along with you in your use of the word induction only I fear to
        a certain extent — I do not myself like to oppose it to or contrast it
        with either observation or pure reason — Induction according to me
        and Aristotle (admire my modesty) is the whole process by which the
        intellect gets a general principle from observing particulars or individ-
        uals and in that process both observation and pure reason have a part
        — when observation has collected the facts abstraction. . . . seizes on
        the law or principle and then the inductive process is compleat in all
        its parts.19
    In this passage Jones suggested to Whewell a way to synthesize the empirical
and the ideal in his epistemology: namely, by seeing induction as an act that
includes both observation and pure reason. Whewell seems to have been inspired
by this characterization of induction. He soon began to describe induction as
a process involving observation and reason. The rational element of induction
was provided by certain “conceptions of the mind”. In a notebook dated 1831-2,
Whewell for the first time characterized induction as involving such conceptions:
“induction supposes a power of clearly representing phenomena by the mind as
subordinate to the conceptions of space, number, etc.”20 In another notebook of
this period Whewell described the importance of distinctly conceiving conceptions
in order to make proper inductions,21 and added what appears to be his first
list of conceptions regulating different sciences.22 By 1833 Whewell had come to
believe that, as he explained in a notebook entry, “knowledge implies passive sense
[i.e., observation] and active thought [i.e., reason]”.23 In an address to the British
Association for the Advancement of Science in that same year, Whewell claimed
that “a combination of theory with facts . . . is requisite” in order to discover new
truths” [Whewell, 1834a, xx]. And in a letter to Jones in 1834 Whewell described
his “Philosophy of Induction”, claiming that “you will see that a main feature is
the assertion of ideas and facts as equally and conjointly necessary to science”.24
    By 1837, Whewell was ready to formulate more systematically this position,
and to express it publicly. He did so in his “Remarks on the Logic of Induction”,
which was appended to his textbook Mechanical Euclid. In this essay Whewell
explained that induction requires both an idea provided by the mind and facts
provided by the world [Whewell, 1838, 181]. A “general idea”, which is not given
by the phenomena but “by the mind”, is “superinduced upon the observed facts”
[Whewell, 1838, 178]. In his History of the Inductive Sciences, published that same
year, Whewell similarly noted that both facts and ideas are requisite for the “for-
mation of science”. “Real speculative knowledge”, Whewell claimed, “demands
the combination of the two ingredients” [Whewell, 1857a 1: 5–6]. By the time
   19 Richard Jones to William Whewell, 7 March 1831, WP Add.Ms.c.52 f. 26, emphasis added.
   20 William Whewell, Notebook, WP R.17.18 f. 13, p. 54.
   21 William Whewell, Notebook, WP R.17.18 f. 15, p. 2.
   22 William Whewell, Notebook, WP R.17.18.f. 15, p. 47, in an entry dated 2 July 1831.
   23 William Whewell, Notebook, WP R.18.17 f. 8, facing p. 20.
   24 William Whewell to Richard Jones, 5 August 1834, WP Add.Ms.c.51 f. 174.
170                                    Laura J. Snyder
he published the first edition of the Philosophy of the Inductive Sciences in 1840,
he had worked out more details of the position, and developed an argument sit-
uating his epistemological view as a “middle way” between stark empiricism and
full-blown rationalism.
   Because of the dual or antithetical nature of knowledge, Whewell claimed, gain-
ing knowledge requires attention to both Ideas and Sensations. Both are required
for knowledge: “without our ideas, our sensations could have no connexion; with-
out external impressions, our ideas would have no reality; and thus both ingredients
of our knowledge must exist” [Whewell, 1858a, 1: 58]. An exclusive focus on one or
the other side of the antithesis is to be avoided. Whewell criticized both Kant and
the German Idealists, for their exclusive focus on the ideal or subjective element,
and Locke and his followers of the “Sensationalist School”, for their exclusive focus
on the empirical, objective element.
   Whewell generally referred to these ideas, which comprise the ideal or subjec-
tive element in his antithetical epistemology, as “Fundamental Ideas”. Whewell
explained that “I call them Ideas, as being something not derived from sensation,
but governing sensation, and consequently giving form to our experience; — Fun-
damental, as being the foundation of knowledge, or at least of Science” [Whewell,
1860, 336]. They are supplied by our minds in the course of our experience of the
external world; they are not simply received from our observation of the world
[Whewell, 1858a, 1: 91]. This is why Whewell claimed that the mind is an active
participant in our attempts to gain knowledge of the world, not merely a passive
recipient [Whewell, 1860, 218].
   Although these Ideas are supplied by our minds, they are such that they enable
us to have real knowledge of the empirical world. They do so by connecting the
facts of our experience; this occurs because the Ideas provide the general relations
that really exist in the world between objects and events. These relations include
Space, Time, Causation and Resemblance, among numerous others. By enabling
us to connect the facts under these relations, the Ideas provide a structure or form
for the multitude of sensations we experience [Whewell, 1847, 1: 25]. Thus, for
example, the Idea of Space allows us to apprehend objects as existing in space,
in spatial relations to each other, and at a particular distance from us. Indeed,
we need these Ideas in order to be able to make sense of our sensations: “our
sensations, of themselves, without some act of the mind, such as involves what
we have termed an Idea, have not form. We cannot see one object without the
Idea of Space; we cannot see two without the idea of resemblance or difference;
and space and difference are not sensations” [Whewell, 1858a, 1: 40]. Every
science, Whewell believed, has one or more Fundamental Ideas particular to it,
which provide the structure for all the facts with which that science is concerned
([Whewell, 1858b, 137] and [Whewell, 1858a, 1: 3]). The Idea of Causation is
especially associated with the science of Mechanics, while the Idea of Space is the
Fundamental Idea of Geometry. Whewell explained further that each Fundamental
Idea has certain “conceptions” included within it; these conceptions are “special
           “The Whole Box of Tools”: William Whewell and the Logic of Induction             171
modifications” of the Idea applied to particular types of circumstances.25 For
example, the conception of force is a modification of the Idea of Cause, applied to
the particular case of motion (see [Whewell, 1858a, 1: 184–5, 236]).
   Thus far, this discussion of the Fundamental Ideas may suggest that they are
similar to Kant’s forms of intuition and categories, and there are indeed some
similarities. Because of this, numerous commentators argue that Whewell’s epis-
temology was derived from his reading of Kant, or perhaps a view of German
philosophy refracted through the writings of Coleridge and the Coleridgian circle
at Cambridge, which included Whewell’s close friends Rose and Julius C. Hare.26
However, although there are some similarities between his view of the Fundamen-
tal Antithesis and certain notions of Kant’s and Coleridge’s, there are important
differences, which are frequently overlooked by commentators on Whewell.
   On the one hand, Whewell did read and appreciate Kant, in a time when Ger-
man philosophy was not terribly popular in England (see [Wellek, 1931]). He knew
German well enough to publish translations of German prose and poetry,27 and, if
a story related by his student Isaac Todhunter is true, well enough to pass in Ger-
many as a native-speaker.28 A notebook from 1825 contains reading notes of the
Critique of Pure Reason.29 Like Kant, Whewell was interested in how it is possible
for us to have knowledge which has a universality and necessity that experience
alone can not give it. And his answer, like Kant’s, involves certain conceptions
and ideas which are in some sense a priori, because they are not derived from
experience. As he wrote to Herschel, “My argument is all in a single sentence.
You must adopt such a view of the nature of scientific truth as makes universal
and necessary propositions possible; for it appears that there are such, not only
in arithmetic and geometry, but in mechanics, physics and other things. I know
no solution of this difficulty except by assuming a priori grounds”.30 Moreover,
Whewell rather modestly admitted that his discussions of the Fundamental Ideas
of Space and Time are mere “paraphrases” of Kant’s discussion of the forms of
intuition in the first Critique [Whewell, 1860, 335]. A contemporary translator of
Kant claimed, in fact, that in his Philosophy Whewell “more elegantly expressed”
certain doctrines of Kant.31 These most basic Ideas of Space and Time (in some
  25 [Whewell, 1858b, 187]. Whewell was not always consistent in maintaining the distinction
between Ideas and conceptions. For instance, in this work he referred both to the “Idea of
Number” (p. 54) as well as the “conception of number” (pp. 50 and 56).
  26 On readings of Whewell as a “British Kantian”, see [Butts, 1965a], [Buchdahl, 1991], [Ruse,
1979], and [Marcucci, 1963]. Claims that Whewell’s view is closely linked to that of Coleridge
can be found in [Preyer, 1985], [Cantor, 1991], [Cannon, 1976], and [Sloan, 2003].
  27 In 1847, Whewell edited English Hexameter Translations from Schiller, Goethe, Homer,
Callinus, and Meleager, which contained Whewell’s translation of Goethe’s Herman and
Dorothea.
  28 Isaac Todhunter quotes a letter from the scientist Alexander von Humboldt, who bemoaned
the fact that he had “lost the pleasure” of seeing Whewell in Potsdam because he had told his
servant to admit an English gentlemen, but Whewell was turned away because “you have spoken
German like an inhabitant of the country!” (see [Todhunter, 1876, 1: 411].)
  29 See William Whewell, Notebook, WP R.18.9 f. 13, p. 19.
  30 William Whewell to John Herschel, 22 April 1841, in [Todhunter, 1876, 2: 298].
  31 In Whewell’s library there is a translation of the Kritik der reinen Vernunft by Francis
172                                         Laura J. Snyder
works, Whewell includes the Idea of Cause among these) do function in Whewell’s
epistemology as similar to Kant’s forms of intuition in being conditions of our hav-
ing any knowledge of the world; indeed, Whewell referred to them as “conditions
of experience” [Whewell, 1858a, 1: 268] and “necessary conditions of knowledge”
[Whewell, 1841, 530]. For instance, on his view, experiencing objects as having
form, position and magnitude requires the Idea of Space (see [Whewell, 1844b,
487]). Because of these similarities, Whewell was accused by his contemporaries
(not all of whom had themselves studied Kant) of trying to import Kant into
British philosophy; in his review of the Philosophy DeMorgan expressed surprise
that “the doctrines of Kant and Transcendental Philosophy are now promulgated
in the university which educated Locke” [DeMorgan, 1840, 707]. The logician H.
L. Mansel — who was the closest thing to a British Kantian in those days —
thought Whewell had not gone far enough, criticizing him for his “stumble on the
threshold of Critical Philosophy” [Mansel, 1860, 258].
   However, this was a threshold that Whewell did not intend to cross. There are
important differences between Kant’s transcendental philosophy and Whewell’s
antithetical epistemology. Whewell did not follow Kant in distinguishing between
the a priori components of knowledge provided by intuition (Sinnlichkeit), the Un-
derstanding (Verstand ) and the faculty of Reason (Vernunft). Thus Whewell drew
no distinction between “precepts”, or forms of intuition, such as Space and Time,
the categories, or forms of thought, in which Kant included the concepts of Cause
and Substance, and the transcendental ideas of reason. Further, Whewell included
as “Fundamental Ideas” many which function not as conditions of experience but
as conditions for having knowledge within their respective sciences: although it is
certainly possible to have experience of the world without having a distinct Idea
of, say, Chemical Affinity, we could not have any knowledge of certain chemical
processes without it ([Whewell, 1858a, 2: 2-3] and [Whewell, 1860, 349]). Unlike
Kant, Whewell did not attempt to give an exhaustive list of these Fundamental
Ideas; rather, he believed that there are others which will emerge in the course
of the development of science. Moreover, and perhaps most importantly for his
philosophy of science, Whewell rejected Kant’s claim that we can only have knowl-
edge of our “categorized experience”. The Fundamental Ideas, on Whewell’s view,
accurately represent objective features of the world, independent of the processes
of the mind, and we can use these Ideas in order to have knowledge of these ob-
jective features.32 Indeed, Whewell criticized Kant for viewing external reality as
Hayward (1848) inscribed “The Rev. Dr. Whewell/With the Translator’s respects”. In the text,
the translator added a note to Kant’s Preface to the 2nd edition: Kant wrote “If the intuition
must regulate itself according to the property of the objects, I do not see how one can know
anything with regard to it, a priori, but if the object regulates itself, (as objects of the senses,)
according to the property of our faculty of intuition, I can very well represent to myself this
possibility” (pp. xxv-xxvi). The added note reads “See Whewell’s Philosophy of the Sciences,
vol. ii, p. 479, where this is more freely translated and more elegantly expressed” (p. xxvi).
  32 In 1865 Charles Peirce expressed a similar view of Whewell’s divergence from Kant’s view:
“Dr. Whewell has usually been considered a Kantian. Up to a certain point this is true. He
accepts Kant’s division of the matter and form of our knowledge and also his theory of space
and time but he seems to have cast away from the doctrine of the limits of our knowledge which
           “The Whole Box of Tools”: William Whewell and the Logic of Induction              173
a “dim and unknown region” [Whewell, 1860, 312]. Whewell’s justification for the
presence of these concepts in our minds takes a very different form than Kant’s
transcendental argument. For Kant, the categories are justified because they make
experience possible. For Whewell, though the categories do make experience (of
certain kinds) possible, the Ideas are justified by their origin in the mind of a
divine creator. And finally, the type of necessity that Whewell claimed is derived
from the ideas is very different from Kant’s notion of the synthetic a priori (we
will return to this final point in section VIII).
    Thus we should take seriously Whewell’s frequent denials that his epistemology
is identical or even particularly similar to Kant’s. In his response to DeMorgan’s
review, a privately-published pamphlet, Whewell noted that his critic had gone
too far in associating him with the Kantian philosophy: “It might have occurred
to him, . . . that by the very circumstance of classing many other ideas with
those of space and time, I entirely removed myself from the Kantian point of
view” [Whewell, 1840b, 4]. In his reply to Mansel, later published as part of
the Philosophy of Discovery, Whewell clearly noted that he never intended to
follow Kant’s view, and pointed out ways in which his philosophy differs from
Kant’s [Whewell, 1860, 335–46]. What he particularly admired of Kant’s work
was his having shown the untenable nature of the account of knowledge given by
the Lockean school [Whewell, 1860, 308]. Whewell’s dislike of Locke began early,
even before his 1825 reading of the first Critique: in 1814 he wrote to his former
headmaster, the Reverend G. Morland, that while reading Locke he “grew out of
humour with [him]”.33 In his chapter on “The Influence of German Philosophy
in Britain” in the final edition of the Philosophy, Whewell set up a tripartite
division showing himself to be in between the idealism of the Germans and the
strict empiricism of Mill, standing in here for the Lockean school. Speaking of
himself in the third person, Whewell wrote: “Kant considers that Space and Time
are conditions of perception, and hence sources of necessary and universal truth.
Dr. Whewell agrees with Kant [only] in placing in the mind certain sources of
necessary truth; he calls these Fundamental Ideas, and reckons, besides Space and
Time, others, as Cause, Likeness, Substance, and several more. Mr. Mill, the
most recent and able expounder of the opposite doctrine, derives all truths from
Observation, and denies that there is such a separate source of truth as Ideas”.34
Whewell was, as he insisted, seeking a “middle way” between the excesses of Locke
and Kant.
    It is noteworthy that, while working out his theory, there are no references
to Kant or uses of technical Kantian terminology such as “forms of intuition”,
“categories”, “analytic and synthetic truths” in his letters to Jones, his notebooks,
is the essence of the critical philosophy” [Peirce, 1982, 205].
   33 William Whewell to George Morland, n.d., April 1814, in [Todhunter, 1876, 2: 2].
   34 [Whewell, 1860, 336]. Buchdahl claims in his [1991] that Whewell shares Kant’s “idealist”
or “subjectivist” approach to the Ideas, and therefore also to the knowledge we gain using them.
Butts [1994, 278] maintains that Whewell “owes his theory of science to Kant”. Fisch has a more
nuanced view, claiming that there are “Kantian undertones” in Whewell’s epistemology, but less
than generally supposed (see his [1991, 105]).
174                                        Laura J. Snyder
or his early drafts of the Philosophy.35 This suggests that Kant was not explicitly
on Whewell’s mind as he strove to develop an epistemology that could incorporate
both empirical and ideal elements of knowledge. While there are, as we have seen,
some aspects of Whewell’s view that are similar to Kant’s epistemology, it is not
necessarily the case that Whewell derived these directly from his reading of Kant in
1825. Whewell certainly appreciated Kant’s critique of the Sensationalist school,
and took note of his ideal solution. But his reading of Dugald Stewart, Thomas
Reid and other Scottish philosophers similarly impressed Whewell, and may well
have first exposed him to the type of ideal solution given by Kant.36 Whewell
expressed an appreciation of the work of the Scottish philosophers as early as
1814, when he sent a letter full of their praise to Morland.37 Years later, when
he published the Philosophy, Whewell applauded the “intelligent metaphysicians
in Scotland”, including “Reid, Beattie, Dugald Stewart, and Thomas Brown”, for
arguing against the “sensationalism” of Hume and Locke (see [Whewell, 1860, 214]
and [Whewell, 1847, 2: 309]. He noted that their shared view, “according to which
the Reason or Understanding is the source of certain simple ideas, such as Identity,
Causation, Equality, which ideas are necessarily involved in the intuitive judgments
which we form, when we recognize fundamental truths of science, approaches very
near in effect to [my] doctrine . . . of fundamental ideas . . . ” [Whewell, 1860,
215] (see also [Whewell, 1847, 2: 310–11]). Stewart had argued, for example, that
every act of observation involves an “interpretation of nature”. In his Elements
of the Philosophy of the Human Mind, he explained that “without theory or, in
other words, without general principles inferred from a sagacious comparison of
a variety of phenomena, experience is a blind and useless guide” (cited in [Corsi,
1987, 97]). It seems likely that Stewart was at least as much an influence upon
the development of Whewell’s epistemology as Kant.
   Similarly, Coleridge’s writings did not inspire Whewell’s antithetical epistemol-
ogy. Commentators have made much of the network of friendships at Trinity
including classical and German scholars influenced by Coleridge and German phi-
losophy such as Hare, Rose, Thirwall, F.D. Maurice and John Sterling, and men
of science such as George Peacock, G.B. Airy, Sedgwick, Herschel, and Whewell
(see [Preyer, 1985] and [Cantor, 1991, especially p. 77]. However, friendship, in
these cases, did not imply philosophical agreement. Whewell was especially close
to Hare and Rose, but he did not join them in their enthusiasm for the Romantic
movement in general or its specific views of education, history, poetry or philoso-
phy. Whewell particularly objected to the way the Romantics set up an opposition
between reason and feeling or “enthusiasm”, elevating the latter over the former.
He argued this point with Rose. “Why will you not see that in speculative matters,
though Reason may go wrong if not guided by our better affections, you cannot
  35 This point is also noted by [Fisch, 1991, 105].
  36 Pietro Corsi has suggested, for example, that Whewell’s earliest views of space and time
as conditions of experience may have been sparked initially by his reading of Dugald Stewart’s
Philosophical Essays, prior to his reading of Kant (see [Corsi, 1988, 155]).
  37 William Whewell to George Morland, 15 June 1814, in [Todhunter, 1876, 2: 6].
            “The Whole Box of Tools”: William Whewell and the Logic of Induction              175
do without her? All your efforts not to reason at all will only end in your rea-
soning very ill . . . Finding that Reason alone cannot invent a satisfactory system
of morals or politics, are you not quarrelling with her altogether, and adopting
opinions because they are irrational?”38 In a notebook entry from 1816-17, con-
taining reading notes on Coleridge’s On the Constitution of the Church and State
According to the Idea of Each, Lay Sermons, Whewell accused Coleridge of trying
to “make reason commit suicide” by ceding ground to enthusiasm.39
    At first glance, however, there are similarities between Whewell’s epistemology
and the view expressed by Coleridge in his Preliminary Treatise on Method (1817),
which served as the general introduction of the Encyclopedia Metropolitana. Unlike
some Cambridge admirers of Coleridge, such as Whewell’s friend Rose, Coleridge
himself did not disparage natural science, and intended his treatise on method to
illustrate the application of his philosophy to the scientific study of nature.40 Co-
leridge noted that discoveries of truth are not made by accident, but by the distinct
presentation of an Idea. He claimed that the science of Electricity had progressed
more rapidly than that of Magnetism because the former contained a clear Idea
of Polarity, while the latter had no clear regulative Idea [Coleridge, 1849, 17–21].
Coleridge described the “perfect” scientific method as involving the placing of par-
ticulars under a general conception, which becomes their “connective and bond of
unity” [Coleridge, 1849, 54]. So far, this is not opposed to Whewell’s view of the
role of fundamental ideas in our knowledge. Yet, as Whewell perceived, Coleridge
was attempting to “separate the poles of the Fundamental Antithesis”, by assert-
ing an absolute division between the ideal and the empirical parts of knowledge
[Whewell, 1860, 424–5]. Coleridge drew a distinction between “Metaphysical” and
“Physical” Ideas, explaining that “Metaphysical Ideas, or those which relate to
the essence of things as possible, are of the highest class . . . Physical Ideas are
those which we mean to express, when we speak of the nature of a thing actu-
ally existing and cognizable by our faculties” [Coleridge, 1849, 20]. Metaphysical
Ideas, then, have nothing to do with empirical experience, while Physical Ideas ex-
clusively concern empirical experience. In his Aids to Reflection and The Friend,
Coleridge asserted further that there are two distinct faculties, the Understand-
ing and the Reason, claiming that these different faculties are responsible for our
having knowledge of different types [Coleridge, 1965-80, 9: 252n and 4: 158]. The
Understanding, according to Coleridge, is the “conception of the Sensuous, or the
faculty by which we generalize and arrange the phenomena of perception”; it is the
faculty that deals with our perceptions of material objects, which we gain through
the senses [Coleridge, 1965-80, 4: 156]. The faculty of Reason is “the organ of the
   38 William Whewell to H.J Rose, 29 December 1823, in [Stair Douglas, 1882, 95].
   39 See a notebook entry from 1816-17, containing reading notes on Coleridge’s On the Con-
stitution of the Church and State According to the Idea of Each, Lay Sermons.         Quoted in
[Todhunter, 1876, 1: 349].
   40 See [Coburn, 1975]. Although, to my knowledge, Whewell never referred explicitly to Co-
leridge’s Preliminary Treatise in his letters to Rose or in his notebooks, he surely must have
read it, given his interest in scientific method and given that Rose was one of the editors of the
Encyclopedia Metropolitana.
176                                         Laura J. Snyder
Super-sensuous”, in that it does not depend on the senses. Rather, Reason is the
source of necessary and universal principles of mathematics and science, as well
as the laws of thought. Understanding is the faculty that leads us to Theories,
which, according to Coleridge, describe relations that are only contingent, being
“the result of observation”. Reason is the faculty that leads us to Laws, which he
claims describe necessary relations between things.41 Thus Coleridge maintained
that the Ptolemean system and the Newtonian system were developed by different
faculties: the Ptolemean system by the Understanding, the Newtonian system by
Reason [Coleridge, 1965-80, 9: 252–3].
   Whewell pointedly criticized Coleridge’s view of the discoveries of Ptolemy and
Newton, calling it “altogether false and baseless”. He noted that “the Ptolemaic
and the Newtonian system do not proceed from different faculties of the mind,
but from the same power, exercised more and more completely” [Whewell, 1862,
122]. As we have seen, Whewell believed that all knowledge requires the use of
both sensations and conceptions; he explained that “there is in science no faculty
which judges according to sense without doing more; and no creative or suggestive
faculty which must not submit to have its creations and suggestions tested by
the phenomena” [Whewell, 1862, 123]. Further, Whewell objected to Coleridge’s
denigration of the understanding by associating it with the instinct of animals
(in Aids to Reflection Coleridge compares this faculty to the instinct of bees and
ants). In this way Coleridge made even more explicit his disdain for the faculty of
perception and the empirical facts with which it deals in the attainment of human
knowledge.42 Like Plato, whom Whewell also criticized, Coleridge was guilty of
“disparaging or neglecting facts” [Whewell, 1860, 11]. After writing one of his
papers on Plato, Whewell sent it to DeMorgan, writing “I hope you will think
that in the paper I send you I have demolished the Coleridgean account of Reason
and Understanding”.43
                   4    WHEWELL’S DISCOVERERS’ INDUCTION
Once he had developed his antithetical epistemology, Whewell was able to con-
struct an inductive methodology that accounted for the discovery of laws while in-
corporating both empirical and a priori elements. Whewell’s first explicit, lengthy
discussion of this inductive method — which he called (in an 1859 letter to De-
Morgan) “Discoverers’ Induction” — appeared in the 1840 first edition of the
Philosophy of the Inductive Sciences, Founded Upon Their History.44 His view
of inductive method remained essentially unaltered through his publication of the
  41 [Coleridge, 1849, 21]. As we will see below, on Whewell’s view this distinction between laws
and theories is invalid, because there is no fundamental division between necessary truth and
empirical truth
  42 See [Coleridge, 1965-80, 9: 243–5] and [Whewell, 1862, 123–6]. See also Whewell’s notes on
Coleridge’s Aids to Reflection (WP R.G.13 f. 39, p. 21, quoted in [Levere, 1996, 1686–89]).
  43 William Whewell to Augustus DeMorgan, 2 February 1858, WP O.15.47 f. 23.
  44 William Whewell to Augustus DeMorgan, January 18, 1859, in [Todhunter, 1876, 2: 417].
          “The Whole Box of Tools”: William Whewell and the Logic of Induction               177
second edition of the Philosophy in 1847, and the third and final edition, which
appeared as three separate works between 1858 and 1860 (The History of Scien-
tific Ideas, in two volumes [1858], the Novum Organon Renovatum [1858] and the
Philosophy of Discovery [1860]). In describing Discoverers’ Induction, Whewell
began by noting that the standard view of induction holds that it is “the process
by which we collect a General Proposition from a number of Particular Cases”
[Whewell, 1847, 2: 48]. However, Whewell rejected this overly narrow notion of
induction, casting induction in the light of the Fundamental Antithesis by arguing
that, in scientific discovery, it is not the case that “the general proposition results
from a mere juxta-position [sic] of the cases” (that is, from simple enumeration
of instances) [Whewell, 1847, 2: 48]. Rather, Whewell explained that “there is a
New Element added to the combination [of instances] by the very act of thought
by which they were combined” [Whewell, 1847, 2: 48]. As Jones had suggested
years before, induction was described by Whewell in the various editions of the
Philosophy as an “act of the intellect” which includes both observation and rea-
soning.45 In the Philosophy, Whewell coined the term “colligation” to describe
this “act of thought”.
   Colligation, Whewell explained, is the mental operation of bringing together a
number of empirical facts by “superinducing” upon them a fundamental conception
that unites the facts and renders them capable of being expressed by a general law.
The conception provides the “true bond of Unity” tying together the phenomena,
by providing a property shared by the known members of a class (note that,
in the case of causal laws, this can be the property of sharing the same cause)
[Whewell, 1847, 2: 46]. We have already seen that, in his 1831 review of Herschel’s
Preliminary Discourse, Whewell used a metaphor suggesting that the facts of
nature are like pearls, and the conception is the string upon which the pearls can
be threaded. Without the string we have nothing but an “indiscriminate” heap
of pearls, while with the string we have something ordered and beautiful. In all
editions of the Philosophy, Whewell continued to believe that laws are formed by
connecting facts with a uniting conception.
   In order to colligate facts with a conception, we must first have suitable facts.
Whewell noted that it is necessary to have “already obtained a supply of definite
and certain Facts, free from obscurity and doubt” [Whewell, 1847, 2: 26]. It is
useful at this point to consider what Whewell precisely meant by a “fact”. In some
places Whewell oversimplified matters, as when he claimed that the colligation
of facts involves the establishing of a connection “among the phenomena which
are presented to our senses”.46 This suggests that “facts” are simply observed
phenomena. However, the situation is complicated for Whewell by his antithetical
epistemology. As we have seen, his view of the nature of knowledge entails that
all observation is mediated by our Ideas and conceptions; thus there can be no
  45 See [Whewell, 1858b, 36], and compare to letter from Richard Jones, 7 March 1831, WP
Add.Ms.c.52 f. 26.
  46 [Whewell, 1847, 2: 36]. He was being sloppy here, as the passage follows the pages in which
he described the need to decompose facts.
178                                        Laura J. Snyder
conception-free observation [Whewell, 1858a, 1: 32]. One consequence of this,
Whewell explained, is that we cannot easily separate in our perceptions of things
the element that our mind contributes and the element that comes from outside
our mind [Whewell, 1847, 2: 27]. It is, indeed, impossible to do so [Whewell, 1847,
2: 30]. Nevertheless, we can attempt to make explicit the Ideas and conceptions
that are involved in our perception of specific phenomena. Once we do so, we can
use those facts that have reference to the more “exact Ideas” of Space, Time, and
Cause, or the conceptions associated with these Ideas, such as Position, Weight,
Number, as the “foundation of Science” [Whewell, 1847, 2: 30–2]. This is useful
because these Ideas and conceptions are particularly definite and precise [Whewell,
1847, 2: 38]. The operation by which we separate complex facts into more simple
facts exhibiting the relations of Space, Time and Cause, is called the Decomposition
of Facts [Whewell, 1847, 2: 33]. Whewell described it as a method of “render[ing]
observation certain and exact” [Whewell, 1847, 2: 35]. He outlined a number of
“Methods of Observation” which can be used in this process.47
   Once we have decomposed facts, Whewell explained, we can examine them with
regard to other Ideas and conceptions. We can bind together the facts by applying
to them a “clear and appropriate” conception; although such a conception may not
be as exact and precise as the conceptions associated with Space, Time and Cause,
it must still be clear enough to be capable of giving “distinct and definite results”
[Whewell, 1847, II: 39]. It is necessary to “explicate conceptions”, that is, clarify
them and render them as precise as possible. Conceptions used in colligation
must be not only clear, but also “appropriate” to the facts involved. As Whewell
pointed out, attention is generally not given to this aspect of discovery. He noted
that “the defect which prevents discoveries may be the want of suitable ideas,
and not the want of observed facts” [Whewell, 1840b, 7]. An “appropriate” or
“suitable” conception is one which expresses a property or cause shared by the facts
which it is used to unify. As Whewell put it, conceptions must be “modifications
of that Fundamental Idea, by which the phenomena can really be interpreted”
[Whewell, 1858b, 30]. Often scientists apply an inappropriate conception to a set
of facts, as when astronomers prior to Kepler applied the conception of epicycles
to colligate planetary motions. This was not the appropriate conception, because
it does not describe the way that planets do, in fact, move. Scientific discoveries
are made not merely when accurate observations are obtained, as was the case
after Tycho Brahe’s observations of the orbit of Mars, but when in addition to
accurate observations the appropriate conception is used, as when Kepler applied
the conception of an ellipse rather than that of the epicycle. Whewell observed
that finding this correct conception is often the most difficult part of discovery, the
part that gives it its “scientific value”. As Whewell noted, Tycho Brahe already
  47 [Whewell, 1847, 2: 337–58]. Whewell’s interest in methods of observation was not merely
theoretical, but also practical. In the mid-1830s, Whewell organized a vast international project
of tidal observations for the Royal Society, which awarded him a gold medal in recognition of
this work in 1837. As part of this enterprise Whewell wrote a Memorandum and Directions for
Tide Observations. For more on Whewell’s work on tidal science, see [Snyder, 2002; Cartwright,
1999].
         “The Whole Box of Tools”: William Whewell and the Logic of Induction                  179
had access to the “facts”, but only Kepler was able to determine that these facts
could be correctly colligated by the conception of an ellipse. Another example
used by Whewell is Aristotle’s inability to account for the mechanical forces in the
lever by attempting to use the conception of a circle to colligate the known facts
regarding the proportion of the weights that balance on a lever. Archimedes later
showed that this was an inappropriate conception, and instead used the idea of
pressure to colligate these facts [Whewell, 1847, 1: 71–2]. The problem with other
notions of induction — most notably Whately’s — is that they omit precisely this
crucial and difficult step of finding the appropriate conception (or, in Whately’s
terms, the minor premise).
   How does a scientist discover the appropriate conception with which to colligate
a group of facts? Whewell believed that this was an essentially rational process,
but one that was not susceptible of being put in algorithmic form. What is most
important to this process is that the discoverer’s mind contains a clarified or
“explicated” form of the appropriate conception. The explication of conceptions is
necessary because, although Whewell claimed that the Ideas and their conceptions
are provided by our minds, they cannot be used in their innate form. Indeed, in
an 1841 paper read before the Cambridge Philosophical Society, “Demonstration
that all Matter is Heavy”, Whewell denied that his Fundamental Ideas were innate
ideas in the typical sense of the term. Unlike innate ideas, the Fundamental
Ideas are not “self-evident at our first contemplation of them” [Whewell, 1841,
530]. Later, Whewell introduced the term “germs” to describe the original form
of the conceptions in our minds. In the third edition of the Philosophy, Whewell
explained that “the Ideas, the germ of them at least, were in the human mind
before [experience]; but by the progress of scientific thought they are unfolded
into clearness and distinctness” [Whewell, 1860, 373] (see also [Whewell, 1858b,
30–49] and [Snyder 2006, 55–60]).
   Whewell is a realist about our knowledge of the world, as he reassured Herschel
(who had questioned his commitment to realism in his review of the Philosophy
and the History [Herschel, 1841]).48 Our Idea of Space, Whewell explained, is
what enables us to conceive things as existing in space with spatial characteristics;
but the reason we conceive things as so existing in space is that “they do so exist”
[Whewell, 1844b, 488]. Objects in the world really do exist in spatial relations
to each other, and with spatial characteristics of size and shape. By the 1850s,
Whewell came to argue that our Fundamental Ideas correspond to the world be-
cause both Ideas and world have a common origin in a divine creator. Our minds
contain the germs of these ideas because God “implanted” these germs within
our minds. Eventually, then, Whewell placed his epistemology on a theological
foundation.
   This theological foundation for Whewell’s epistemology was inspired by the work
  48 When a person has knowledge about something, Whewell elsewhere claimed, “we conceive
that he knows it because it is true, not that it is true because he knows it” [1844a, 479]. I thus
disagree with Buchdahl’s reading of Whewell in his [1991].
180                                         Laura J. Snyder
of his childhood friend, the comparative anatomist Richard Owen.49 In the early
to mid-40s, Owen developed his theory of the archetypal vertebrate.50 Owen first
published his theory in 1846 in his work on the Anatomy of Fishes and developed it
further in On the Archetype and Homologies of the Vertebrate Skeleton (1848) and
On the Nature of Limbs (1849). Owen argued that individual vertebrate animals
could be seen as modified instantiations of patterns or archetype forms that existed
in the Divine Mind. This “unity of plan” view allowed Owen to explain homologies,
that is, similar structures that have quite different purposes, such as the wing of a
bird and the forelimb of a quadruped. According to Owen, the similarity of these
structures that were used in such different ways was a consequence of their being
variations on the vertebrate archetype. Similarly, structures without apparent
purpose, such as male nipples, did not contradict the claim that all of creation
was designed; rather they were the result of the application of general archetypes
(for more on Owen, see [Rupke, 1994]).
   Whewell applied Owen’s archetype theory to his epistemology in five chapters
originally written for the Plurality, which Whewell deleted at the last moment
before publication.51 (Whewell took this drastic step on the advice of his friend
Sir James Stephen, who had been Regius professor of history at Cambridge since
1849, and who was concerned that these chapters were too “metaphysical” for
the general reader.) Whewell added a discussion of this point in the last volume
of the final edition of the Philosophy of the Inductive Sciences, the Philosophy
of Discovery, published in 1860; presumably, the intended audience for this book
could be expected to follow metaphysical expositions (see [Whewell, 1860, chapters
30 and 31]). Whewell argued that there were many archetypical Ideas in the
Divine Mind, in accordance with which the universe was created. As he put it,
God “exemplified” in his creation certain Ideas existing in His mind, by creating
the universe in accordance with these Ideas.52 For example, God exemplified an
Idea of Space in his universe by creating all physical objects as having spatial
characteristics, and as existing in spatial relations to each other. The Idea of
Space, then, became on Whewell’s view an archetypical Divine Idea similar to the
Idea of the vertebrate skeleton: it was no less embodied in the physical world. In
creating us in his own image, God implanted us with the germs of the Divine Ideas.
Whewell claimed that “our Ideas are given to us by the same power which made
the world”. We can know the world because God has created us with the Ideas
needed to know it. More precisely, our minds are created with the “germs” from
which the Ideas can develop [Whewell, 1860, 373]. God, then, has implanted within
  49 In his [2003], Sloan argues that Whewell’s philosophy was an influence upon Owen’s devel-
opment of his archetype theory, specifically Owen’s concept of the archetype as both empirical
and transcendental (see pp. 58–9).
  50 For more details regarding the development of Owen’s theory and its appearances in his
lectures in the early 40s, before its publication, see [Sloan, 2003, 55–7].
  51 See William Whewell’s letter to Richard Jones, 30 December 1853, WP Add.Ms.c.51 f. 278.
  52 [Whewell, 1860, 371–379], and Of the Plurality of Worlds, printer’s proofs including five
chapters not included in the published version, WP Adv.c.16 f.27, p. 277. These chapters are
included in the edition of the Plurality of Worlds edited by Michael Ruse [Whewell, 2001].
          “The Whole Box of Tools”: William Whewell and the Logic of Induction              181
us the germs that need to be unfolded into Ideas representing those archetypical
Divine Ideas upon which the universe was created. Gaining a clear view of these
Ideas enables us to have knowledge of both the natural world and its Creator.53
   Because the Fundamental Ideas exist innately within us only in the form of
germs which must be unfolded, the explication of Ideas and conceptions is a nec-
essary part of science. Explication of conceptions is a necessarily social process,
proceeding by “discussion” and “debate” among scientists; explication is not a pro-
cess that can take place solely within the mind of an individual genius. Whewell
noted that disputes concerning different kinds and measures of Force were impor-
tant in the progress of mechanics, and the conception of the Atomic Constitution of
bodies was currently being debated by chemists [Whewell, 1847, 2: 6–7]. Whewell
explained that by arguing in favor of a particular meaning of a conception, sci-
entists are forced to clarify and make more explicit what they really mean. This
is beneficial, whether or not the original expression prevails. If it does not, then
it is replaced by a more accurate or clearer expression of the conception. If it
does, the original expression has been improved by the scientists’ efforts. Thus,
like Bacon, Whewell argued that “the tendency of all such controversy is to diffuse
truth and to dispel error. Truth is consistent, and can bear the tug of war; Error
is incoherent, and falls to pieces in the struggle” [Whewell, 1847, 2: 7]. Whewell
claimed that the explication of conceptions is a “necessary part of the inductive
movement” [Whewell, 1858b, vii]. Indeed, a large part of the history of science
is the “history of scientific ideas”; i.e., the history of their explication and sub-
sequent use as colligating concepts [Whewell, 1858a, 1: 16]. This is why, even
though Whewell worried about his philosophy being considered too metaphysical
by Jones and Herschel, ultimately he believed that metaphysics is a necessary part
of science at all stages; he explained that “the explication or . . . the clarification
of men’s ideas . . . [is] the metaphysical aspect of each of the physical sciences”
[Whewell, 1858b, vii]. Disagreeing with Auguste Comte, who had claimed that
at its highest level of advancement the intellect becomes “scientific” by purging
itself of metaphysics, Whewell argued that successful discoverers differ from “bar-
ren speculators” not by rejecting metaphysics, but by having “good metaphysics”
rather than bad [Whewell, 1858b, vii].
   Although the explication of conceptions is a process that occurs by discussion
and debate among groups of scientists, scientific discoveries are, Whewell noted,
generally made by individuals. Why is it that particular individuals are able to dis-
cover the appropriate conception to apply to a set of facts? Whewell pointed to a
facility for “invention”, the quality of “genius”; yet he strongly denied that there is
anything “accidental” about scientific discoveries. Whewell explicitly opposed the
view popularized by David Brewster in his Life of Newton (1831) that the discov-
ery of the law of universal gravitation was a “happy accident”. Whewell stressed
that discovery is always preceded by much intellectual and scientific preparation.
  53 Just as Owen’s archetype theory was supplanted by Darwin’s theory of evolution, so too
could Whewell’s theological foundation for his epistemology be replaced by an evolutionary epis-
temology. For an early attempt, see [Snyder, 1996].
182                                       Laura J. Snyder
Thus, Whewell noted that Kepler was able to discover the elliptical orbit of Mars
because his mind contained the clear and precise Idea of Space, and the conception
of an ellipse derived from this Idea. Whewell explained that “to supply this con-
ception, required a special preparation, and a special activity in the mind of the
discoverer . . . To discover such a connection, the mind must be conversant with
certain relations of space, and with certain kinds of figures” [Whewell, 1849, 28–
9]. The fact that Kepler’s mind was more “conversant” with the long-explicated
conceptions derived from the Idea of Space, such as that of the ellipse, explains
why he was able to recognize that the ellipse was the appropriate conception with
which to colligate the observed points of the Martian orbit, while Tycho Brahe
(who had made the most precise observations) was not.54
   Another useful quality for the discoverer is a certain facility in generating a
number of possible options for the appropriate conception. Often, before the ap-
propriate conception is applied to the facts, the discoverer must call up in his
mind a number of possibilities. Whewell sometimes used the term “guesses” to
describe this stage of the discovery. Because of this, twentieth-century commenta-
tors, for the most part, have incorrectly viewed Whewell’s methodology as similar
to the “method of hypothesis” (or, as it is now known, the “hypothetico-deductive”
method). (For examples of this interpretation, see [Achinstein, 1992], [Buchdahl,
1991], [Butts, 1987], [Hanson, 1958], [Hempel, 1966], [Laudan, 1971], [Laudan,
1980], [Niiniluoto, 1977], [Ruse, 1975], [Wettersten, 1994], and [Yeo, 2004].) On
that view, there is no rational inference to a hypothesis; its formation is generally
described as a “guess”. However, Whewell claimed that the application or the
selection of the appropriate conception, in Kepler’s case and in all cases of dis-
covery, is not a matter of guesswork. Whewell described this process as being one
in which “trains of hypotheses are called up and pass rapidly in review; and the
judgment makes its choice from the varied group” [Whewell, 1847, 2: 42, emphasis
added]. Thus, even though at a certain point in his investigation (i.e., once he had
inferred that the Martian orbit was some type of oval) Kepler called up in his
mind “nineteen hypotheses” of possible ovals, his choice of the appropriate ellipse
conception was based on “calculations”, and hence on a rational process (and,
certainly, his conclusion that the orbit was oval involved much rational inference).
Nor is the selection of an appropriate conception a matter of mere observation.
Whewell claimed that to choose the appropriate conception requires more than
this: “there is a special process in the mind, in addition to the mere observation
of facts, which is necessary” [Whewell, 1849, 40]. (See also [Whewell, 1831, 379]
and [Whewell, 1860, 256–7].) This “special process of the mind” is inference. In
order to colligate facts with the appropriate conception, “we infer more than we
see”, Whewell explained [Whewell, 1858a, 1: 46]. Whewell claimed that, in the
case of choosing the conception of force to colligate the observed motions of a
needle towards a magnet, an inference, or an “interpretative act” of the mind, is
required (see [Whewell, 1858a 1: 31 and 45]). Inference is required before we see
  54 For a more detailed discussion of how Kepler’s discovery conformed to Whewell’s methodol-
ogy, see [Snyder, 1997].
          “The Whole Box of Tools”: William Whewell and the Logic of Induction         183
that “force” is the appropriate conception with which to colligate magnetic phe-
nomena. Since the selection of the appropriate conception with which to colligate
the facts involves inference, Whewell noted that discoveries are made “not by any
capricious conjecture of arbitrary selection” [Whewell, 1858a, 1: 29], that is not
by guesswork.
   Indeed, selecting the appropriate conception typically requires not just one in-
ference, but a series of inferences. This is why Whewell claimed that discoverers’
induction is a process, involving a “train of researches” [1857a 1: 326]. (See also
[Lugg, 1989].) Whewell explicitly opposed limiting inductive discovery to enumer-
ative inference, writing that “induction by mere enumeration can hardly be called
induction” [1850, 451]. Indeed, enumerative inference could not, in most cases, ac-
count for the discovery of the appropriate conception with which to colligate the
data. Rather, Whewell allowed that any form of valid inference could be used. He
especially stressed the power of analogical reasoning. He extolled the importance
of analogical inference in two reviews from the early thirties, his review of Her-
schel’s Preliminary Discourse and his discussion of the second volume of Lyell’s
Principles of Geology (see [Whewell, 1831, 385] and [Whewell, 1832, 110]. See
also [Whewell, 1837, 2: 391]). Decades later, in his Plurality of Worlds, Whewell
more systematically discussed the importance of defining the varying degrees of
precision and relevance to different kinds of analogies [Whewell, 1855]. Not ev-
eryone was pleased with this liberal notion of inductive reasoning. As we have
seen, DeMorgan complained about certain writers using the term “induction” as
including “the use of the whole box of tools” [DeMorgan, 1847, 216].
   Although inductive discovery is not a matter of accident, according to Whewell,
there is no “logic of discovery” in the sense of an algorithmic procedure. According
to Whewell, “no maxims can be given which inevitably lead to discovery. No
precepts will elevate a man of ordinary endowments to the level of a man of genius”
[Whewell, 1858b, 94]. As we have seen, Whewell’s antithetical epistemology entails
the need for the existence of clear conceptions in the mind of the discoverer; and,
while there are methods to aid in the clarification and selection of the appropriate
conception, there is no algorithm or mechanical method for this process. This
is why invention requires “genius”. However, the fact that Whewell denied the
possibility of a mechanical method of discovery does not entail his rejection of a
rational discovery-method. His discoverers’ induction is a rational method for the
discovery of laws, even though it is not purely mechanical. His method does not
contain rules that are universally applicable but does offer rational “instruments”
for aiding in discovery. Thus, in a letter to DeMorgan, Whewell noted that it
is possible to have an “Art of Discovery”; he quipped “if I had £20,000 a year
which might be devoted to the making of discoveries, I am sure that some might
be made”.55
  55 William Whewell to Augustus DeMorgan, 14 February 1859, in [Todhunter, 1876, 2: 416].
184                                   Laura J. Snyder
                       5   CONFIRMATION OF THEORIES
Once Discoverers’ Induction results in a new hypothesis, Whewell claimed that
there were three ways to confirm it. These confirmation criteria are first, that
“our hypotheses ought to fortel phenomena which have not yet been observed;
at least phenomena of the same kind ” as those which the hypothesis originally
colligated [Whewell, 1858b, 86]; second, that they should “explain and determine
cases of a kind different from those which were contemplated in the formation”
of the hypotheses [1858b, 88]; and third that hypotheses should “become more
coherent” over time [1858b, 91]. These verification criteria are known respectively
as prediction, consilience, and coherence.
   On Whewell’s view, prediction includes a temporal element: a “successful pre-
diction” occurs when some fact unknown at the time that the theory was discovered
is predicted by the theory, and is afterwards found to be true. Such predictions,
when they turn out to be successful, are evidence in favor of the hypothesis. (How-
ever, contrary to Mill’s assertion in his [1963-91, 7: 501], Whewell did not argue
that successful predictions were conclusive evidence for a theory; indeed, as exam-
ples of hypotheses that led to successful predictions, Whewell included several that
were already known to be false, such as the epicycle theory.) Moreover, Whewell
claimed even more strongly that successful predictions were better evidence than
mere explanations of known facts. He maintained that “to predict unknown facts
found afterwards to be true is . . . a confirmation of a theory which in impressive-
ness and value goes beyond any explanation of known facts” [Whewell, 1857a 2:
464].
   Whewell argued for his two claims about predictive success by invoking an
analogy between constructing a true theory and breaking a code. “If I copy a long
series of letters of which the last half-dozen are concealed, and if I guess these
aright, as is found to be the case when they are afterwards uncovered, this must
be because I have made out the import of the inscription” [Whewell, 1860, 274].
Successful prediction of formerly unknown facts, more so than the colligation or
explanation of known facts, Whewell thus argued, is proof that we have broken
the code of Nature, that we have “detected Nature’s secret” [Whewell, 1858b,
87]. Additionally, he compared our ability to colligate facts into hypotheses by his
methodology to knowing the alphabet of the language of Nature. As in learning a
new language, however, this is not enough; we must also know how to construct
intelligible words and sentences using the alphabet, and thus learn how to use
“the legislative phrases of nature” [Whewell, 1858b, 87]. Whewell noted that it is
easier to learn the alphabet than to use these legislative phrases; thus successful
predictions of unknown facts serve as better evidence than colligations of known
facts. Our prediction of some new fact is analogous to the attempt of a language
student to form a proper sentence in her new language. When our prediction is
confirmed, as when the attempted sentence elicits an appropriate response from a
native speaker, it is a sign that we have spoken correctly: Nature is “respond[ing]
plainly and precisely to that which we utter, [and] we cannot but suppose that we
          “The Whole Box of Tools”: William Whewell and the Logic of Induction             185
have in great measure made ourselves masters of the meaning and structure of her
language” [Whewell, 1858b, 87].
   Whewell’s view of the power of prediction seemed to be borne out in 1846, when
a successful prediction appeared to provide stunning evidence for the truth of New-
ton’s theory. In that year, the planet Neptune was discovered after its existence,
position and even its mass had been predicted mathematically. Perturbations in
the orbit of Uranus expected on Newtonian theory had led some to conclude that
there must be an unobserved planetary body external to Uranus’ orbit exerting an
additional gravitational force on the planet. Using Newton’s theory, the French
mathematician U.J.J. Le Verrier calculated mathematically the mass and orbit of
this postulated planet.56 Acting upon Le Verrier’s calculations, astronomers at
the Berlin Observatory found the planet less than one degree from its expected
location. From Newton’s theory, then, it was possible to predict successfully the
existence, position and mass of a previously unexpected planet. This success was
considered further and quite strong evidence for Newton’s theory of Universal
Gravitation. Whewell praised the discovery as a triumph of astronomy, which he
termed the “Queen of the Sciences”. He argued that predictive success is strong
confirmation of a theory, because the agreement of the prediction with what is
found to be true is “nothing strange, if the theory be true, but quite unaccount-
able, if it be not” [Whewell, 1860, 273–4]. If Newtonian theory were not true,
Whewell was suggesting, the fact that from the theory we could correctly predict
the existence, location and mass of Neptune would be bewildering, and indeed
miraculous, equivalent to the feat of a non-speaker of Russian forming an intelli-
gible and meaningful question in that language which elicited a proper response
from a native speaker.
   Whewell did not offer an argument for his claim about the superiority of “new
evidence” over “old evidence”, that is, of successful predictions over explanations
of known facts. He merely stated that “If we can predict new facts which we
have not seen, as well as explain those which we have seen, it must be because
our explanation is not a mere formula of observed facts, but a truth of a deeper
kind” [Whewell, 1849, 60]. The intuition behind this claim, which has been made
by numerous modern commentators, is that if a theorist knows what fact must
be explained by his theory, it is simple for him to “cook” the theory in order to
somehow explain the known fact; whereas, on the other hand, if a theory pre-
dicts a novel (unknown) fact it is not due to the ingenuity of the theorist, but
to the truth of the theory. However, this intuition has been contested in recent
years (see [Brush, 1989; Snyder, 1998]). Interestingly, the discovery of Neptune,
Whewell’s paradigmatic case of a novel prediction, may be considered not only as
a prediction of a new fact, but equally well as an explanation of a known fact,
namely the perturbations of the planet Uranus. The problem for Newton’s theory
was described in this way by Airy in his report to the BAAS in 1832: “I need
  56 At roughly the same time an Englishman, James Couch Adams, also made these calculations,
which very nearly caused an international priority war between the British and the French after
the planet was discovered (see [Smith, 1989]).
186                                      Laura J. Snyder
not mention that there are other subjects (the theory of Uranus, for instance)
in which the existence of difficulties is known, but in which we have no clue to
their explanation” [Airy, 1833, 189]. Thus whether the postulation of an unseen
planet is considered an explanation of a known fact or the prediction of a new
fact depends upon the way it is described, not on any objective features of the
evidence itself or of its relation to the theory. So it would be odd, to say the least,
if the way of describing the evidence made a difference to its evidential value. It
should be noted that Whewell only supported the thesis about the superiority of
new evidence in very strong cases of predictive success, such as the discovery of
Neptune; he explained of such success that “It is a confirmation which has only
occurred a few times in the history of science, and in the case only of the most
refined and complete theories, such as those of Astronomy and Optics” [Whewell,
1857a 2: 464].
    Whewell’s next, and most interesting, confirmation test is consilience. Whewell
explained that “the evidence in favour of our induction is of a much higher and
more forcible character when it enables us to explain and determine cases of a kind
different from those which were contemplated in the formation of our hypothesis.
The instances in which this have occurred, indeed, impress us with a conviction
that the truth of our hypothesis is certain” [Whewell, 1858b, 87–8]. Whewell
called this type of evidence a “jumping together” or “consilience” of inductions.
To understand what Whewell means by this, it may be helpful to schematize the
“jumping together” that occurred in the case of Newton’s law of universal grav-
itation, Whewell’s exemplary case of consilience. In book III of the Principia,
Newton listed a number of “propositions”. These propositions are empirical laws
that were inferred from certain “phenomena” (which are described in the preced-
ing section of book III). The first such proposition, inferred from phenomena of
“satellite motion”, is that “the forces by which the circumjovial planets are con-
tinually drawn off from rectilinear motions, and retained in their proper orbits,
tend to Jupiter’s centre; and are inversely as the squares of the distances of the
places of those planets from that centre”. The result of another, separate induc-
tion from the phenomena of “planetary motion” is that “the forces by which the
primary planets are continually drawn off from rectilinear motions, and retained
in their proper orbits, tend to the sun; and are inversely as the squares of the
distances of the places of those planets from the sun’s centre”. Newton saw that
these laws, as well as other results of a number of different inductions, coincided
in postulating the existence of an inverse-square attractive force as the cause of
various phenomena.57 According to Whewell, Newton saw that these inductions
“leap to the same point”; i.e., to the same law [Whewell, 1858b, 88]. Newton was
then able to bring together inductively (or “colligate”) these laws, and facts of
  57 See [Newton, 1968]. As Ernan McMullin notes, the notion of the “attractive force” was
problematic for Newton; he was never satisfied in his search for an “agent cause” of gravitational
behavior, a “cause of gravity”, as Newton put it. But Newton did succeed in causally unifying
disparate phenomena by the concept of this attractive force (see [2001, 296]). It is this aspect of
Newton’s theory that Whewell pointed to as the exemplar of consilience.
           “The Whole Box of Tools”: William Whewell and the Logic of Induction                    187
other kinds (for example, the event kind “falling bodies”), into a new, more gen-
eral law, namely the universal gravitation law: “All bodies attract each other with
a force of gravity which is inverse as the squares of the distances”. By seeing that
an inverse-square attractive force provided a cause for different kinds of events —
for satellite motion, planetary motion, and falling bodies — Newton was able to
perform a more general induction, to his universal law.
    As I have argued elsewhere [Snyder, 2005; 2006], understanding Whewell’s no-
tion of natural kinds is useful for comprehending his notion of consilience. Whewell
believed that there are natural kinds of objects, such as minerals and species, but
also natural kinds of events, such as diffraction and planetary motion. Consilience
occurs when a theory brings together members of different kinds, showing that
they belong to a more general classification. In the case of event kinds, indi-
vidual types of events are members of the same kind when they share the same
cause.58 Newton discovered that what makes “the orbit of Mars” a member of
the class “planetary motion” is that it is caused to have the properties it does by
an inverse-square attractive force of gravity between Mars and the other bodies in
the universe. He also found that other event kinds share this kind essence. What
Newton did, in effect, was to subsume these individual event kinds into a more
general kind comprised of sub-kinds sharing a kind essence, namely being caused
by an inverse-square attractive force. Consilience of event kinds results in causal
unification. More specifically, it results in unification of natural kind categories
based on a shared cause. Phenomena that constitute different event kinds, such as
“planetary motion”, “satellite motion”, and “falling bodies”, were found by New-
ton to be members of a unified, more general kind, “phenomena caused to occur
by an inverse-square attractive force of gravity” (or, “gravitational phenomena”).
In such cases, according to Whewell, we learn that we have found a “vera causa”,
or a “true cause”, i.e. a cause that really exists in nature, and whose effects are
members of the same natural kind.59 Moreover, by finding a cause shared by phe-
nomena in different sub-kinds, we are able to colligate all the facts about these
kinds into a more general causal law. Whewell claimed that “when the theory, by
the concurrences of two indications . . . has included a new range of phenomena,
we have, in fact, a new induction of a more general kind, to which the inductions
formerly obtained are subordinate, as particular cases to a general population”
[Whewell, 1858b, 96]. He noted that consilience is the means by which we ef-
fect the successive generalization that constitutes the advancement of science (see
[Whewell, 1847 2: 74]).
    Note that consilience is importantly different from the type of reasoning known
as “inference to the best explanation”, even when the “best explanation” is de-
fined as a causal hypothesis. The causal law expressing the essence of different
event kinds is not one postulated merely because it explains or accounts for these
   58 For more details of Whewell’s notion of natural classification, see [Snyder, 2006, chapter
three].
   59 See [Whewell, 1860, 191]. Ruse [1976, 232] agrees that Whewell’s notion of the vera causa
lies at the center of his criterion of consilience. See also [Butts, 1977, 81] and [Ruse, 1975, 161–2].
188                                      Laura J. Snyder
different classes of facts. Rather, in the case of each class, the law has emerged
from a process of inductive reasoning following Whewell’s methodology. Thus the
causal law is not imposed from above as a means of tying together different event
kinds. Instead, it wells up from beneath in each separate case. What is important
in the case of consilience, then, is not merely that a single causal mechanism or
law can explain or account for different event kinds (as the inference to the best
explanation allows), but rather that separate lines of induction lead from each
event kind to the same causal mechanism or law, that there is a convergence of
distinct lines of argument. At each step, there is inductive warrant for the causal
law besides its explanatory utility.
   Whewell discussed a further, related test of a theory’s truth: “coherence”
[Whewell, 1858b, 91]. Coherence occurs when we are able to extend our hypothesis
to colligate a different event kind without ad hoc modification of the hypothesis;
that is, without suppositions which are added merely for the purpose of saving the
phenomena, for which there is no independent evidence. When Newton extended
his theory regarding an inverse-square attractive force, which colligated facts of
planetary motion and lunar motion, to the natural class “tidal activity”, he did not
need to add any new suppositions to the theory in order to colligate correctly the
facts of this event kind (facts about particular tides).60 The situation was rather
different for phlogiston theory (which explained combustion, before the discovery
of oxygen, by the presence of an “essence” in the burning body called “phlogis-
ton”). According to Whewell, phlogiston theory colligated facts about “chemical
combination”. But when the theory was extended to colligate facts about the
weight of bodies, it was unable to do so without an ad hoc and implausible modifi-
cation (namely, the assumption that phlogiston has “negative weight”) [Whewell,
1858b, 92–3]. The emission or particle theory of light was claimed by its adherents
to be capable of colligating the facts of different kinds (indeed, the same facts that
the wave theorists claimed they could colligate). However, according to Whewell
and other proponents of the wave theory, the particle theory could do so only by
the admission of suppositions that were ad hoc or implausible. For example, in
order to account for the facts of the velocity of light (in particular, the uniformity
of velocity), the particle theorists needed to suppose that light particles are emit-
ted with different velocities based on the mass of the luminous body, but that, as
Humphrey Lloyd acerbically put it, “among these velocities is but one which is
adapted to our organs of vision” [Lloyd, 1835, 300–1. See also 296]. (Of course,
the wave theorists were themselves accused of postulating an implausible or ad
hoc “luminiferous ether”; it was argued by Whewell and others that there was
independent evidence for the ether’s existence, and hence that it was not an ad
hoc supposition.) In non-coherent cases such as these, the modifications have no
independent evidence; they are added to the theory solely because they are needed
  60 Facts about the tides were not known systematically, until Whewell engaged in his tidology
research program. Part of the importance of his project, Whewell believed, was to strengthen the
claim that Newton’s law could colligate these facts without modification (see [Whewell, 1834b,
15–17).
          “The Whole Box of Tools”: William Whewell and the Logic of Induction               189
to colligate a new or problematic set of facts. As this discussion suggests, coher-
ence is closely related to consilience. The distinction between them is as follows.
In cases of consilience, two independent colligations of different event kinds yield
the same conclusion. In cases of coherence, the result of one or more colligations
of one or more event kinds is extended, without ad hoc modification, to another
event kind. This latter process generally occurs over time as theoretical systems
are developed.61
   Consilience, then, tells us that we have found a law that colligates members
of a natural kind — either members of a natural object kind or a natural event
kind.62 In the case of object kinds, consilience is, Whewell suggested, evidence
for the truth of a non-causal empirical law such as “all metals are crystalline
when solid”, and thus is evidence that we have grouped certain individuals into
a kind (“metals”) that is natural and hence supportive of the law. In the case
of event kinds, consilience is evidence for having found a causal law, such as the
inverse-square law discovered by Newton. This is at the same time evidence for
the naturalness of the kind “gravitational phenomena”.
   We are now in a position to see why a common argument against Whewell’s
criterion of consilience is flawed. A number of commentators have proposed that
consilience is fundamentally a relativistic or contextual criterion. According to this
view, whether some theory is consilient depends upon our knowledge-state with
respect to the classes involved; specifically, whether we know that certain classes of
facts are members of the same general type of thing (see [Laudan, 1971]; [Forster,
1988]; [Harper, 1989] and [Morrison, 1990]). One critic notes that for Descartes,
planetary and terrestrial motions were phenomena of the same type, because both
were the result of vortices; thus, relative to the Cartesian system, Newton’s theory
was not consilient [Laudan, 1971, 374] (see also [Butts, 1977, 74–5; Fisch, 1985;
Morrison, 1990]). However, as we have seen, the criterion for consilience is that a
theory connects two (or more) different natural kinds into a more general natural
kind in virtue of inductively inferring the same cause for each natural sub-kind.
The issue, then, is not the relativistic one of whether or not we previously knew
or suspected that these sub-kinds all belong to a more general kind, but rather
with whether they do all fit into the same more general kind in virtue of sharing
the same cause, and whether we have reached this conclusion by the proper sort
of inductive reasoning (thus conferring on the conclusion inductive warrant), and
not merely by postulating a common cause hypothetically. Whewell famously
believed that Descartes did not reach his conclusion by proper reasoning. On the
other hand, Newton’s theory was consilient because it satisfied both conditions.
   This erroneous criticism of Whewell’s criterion of consilience may arise from a
confusion of Whewell’s view with that of Herschel. Herschel’s brief discussion of
  61 [Whewell, 1858b, 90–1]. The claim that consilience results in coherence can be found in
[Forster, 1988] and [Harper, 1989]. However, it is more accurate to say that consilience and
coherence are similar means for obtaining the same result, namely simplification in the sense of
(causal) unification.
  62 Herschel recognized this in his review of Whewell’s History and Philosophy [Herschel, 1841,
227]. See also [Ruse, 1976, 240].
190                                        Laura J. Snyder
consilience in the Preliminary Discourse focuses on the subjective, psychological
aspects of consilience, and does not explicitly include the requirement of causal
unification under more general natural kinds (see [Herschel, 1830, 170]). Herschel
explained that “the surest and best characteristic of a well-founded and exten-
sive induction . . . is when verifications of it spring up, as it were, spontaneously,
into notice, from quarters where they might be least expected, or even among
instances of that very kind which were at first considered hostile to them” [1830,
170]. Herschel’s view of consilience is prone to the relativistic criticism. Like Her-
schel, Whewell did at times stress the psychological aspect of consilience, noting
the impact of “the unexpected coincidence of results” [Whewell, 1847, 2: 67]. Nev-
ertheless, he emphasized the logical aspect of consilience, which does not depend
upon the psychological. Whewell’s consilience not only concerns the psychological
element of surprise, but also the logical element of causal unification of different
event or process kinds into more general kinds, in virtue of sharing a common
cause.
    Whewell justified this criterion by claiming that consilience is “a criterion of
reality, which has never yet been produced in favour of falsehood”. Elsewhere
he noted that “there are no instances, in which a doctrine recommended in this
manner has afterwards been discovered to be false” (see [Whewell, 1858b, 90] and
[Whewell, 1860, 192]. He can thus be seen as making an inductive argument in
favor of the confirmation value of consilience: namely, the argument that, since
consilient theories in the past have all turned out to be true, we can infer that
(probably) all consilient theories are true, and thus that our current consilient
theories are (probably) true.63 Whewell has been much ridiculed for this induc-
tive argument by Mill and his twentieth-century followers (see, for example, [Van
Fraassen, 1985, 267]). As Whewell’s detractors today like to point out, the one
positive instance strongly adduced by Whewell for his inductive argument has
been shown to be a counter-instance: Newton’s theory, which seemed irrefutable
in Whewell’s time, has since been shown to be false. Newton’s theory employs
conceptions that, it turns out, are not true of the world, such as absolute space
and absolute time. The cause postulated by the theory, the inverse-square attrac-
tive force of gravity propagated through space and time, is also non-existent (it
is not a vera causa). Thus, although Newton’s theory was highly consilient, more
so than any other theory of Whewell’s time — it brought together different event
kinds, by having inductive warrant for the inference that they shared a common
cause — it turned out to be false. Whewell’s inductive argument does seem to fail
in establishing any confirmatory value for consilience.
    However, even though Newton’s theory turned out to be false, there is something
about the theory which has not been disconfirmed. Newton’s theory still seems to
have been correct in bringing together event kinds that belong to a more general
   63 Whewell also made an additional argument for consilience’s value: he claimed that consilience
is like the coinciding testimony of two witnesses. In my [2005], I show how this argument can be
understood as a form of “common-cause argument”. But I here focus on the argument Whewell
most relied upon.
         “The Whole Box of Tools”: William Whewell and the Logic of Induction     191
kind, showing something true about the natural kind structure of the physical
world, although it was wrong about what the shared kind essence of the natural
kind was. Newton showed that the phenomena of free fall, pendulum motion, lunar
acceleration, satellite motion and planetary motion share a cause, and in this way
constitute a general event kind. This insight has not been proved false, although
the particular shared cause he postulated has been rejected. Einstein’s general
theory of relativity proposed instead that the cause of these phenomena (and
others) is the curved structure of space-time. It may well be that this proposed
cause will be replaced by another; but so far it seems unlikely that a later theory
will separate these different phenomena and attribute to each of them different
causal structures or mechanisms. These event kinds do seem to be sub-kinds of
a more general kind that is connected in nature. Thus the one truly consilient
theory endorsed by Whewell is still considered to have gotten things right, at least
in terms of the natural kind structure of the physical world.64
   Of course, the fact that Newton’s theory was wrong about the particular cause
he postulated does show that Whewell went too far in claiming that consilience is
conclusive evidence for a theory, in the sense of proving that a theory will never be
shown to be false. And it is certainly possible that even the natural kind structure
Newton’s theory proposed might turn out to be wrong. It would have been more
accurate — as well as more consistent with his general view of the progress of
science — for Whewell to have claimed that hypotheses satisfying the condition of
consilience are our best theories, but are still subject to correction. He suggested
this kind of revision when he asserted, referring to Newton’s Rules of Reasoning in
Book Three of the Principia, that “the really valuable part of the Fourth Rule is
that which implies that a constant verification, and, if necessary, rectification, of
truths discovered by induction, should go on in the scientific world. Even when the
law is, or appears to be, most certainly exact and universal, it should be constantly
exhibited to us afresh in the form of experience and observation” [Whewell, 1860,
196, emphasis added].

