Logic and Language in Early Modern Philosophy Michael Losonsky (In Cambridge Companion to Early Modern Philosophy, ed. D. Rutherford, Cambridge University Press.) I. Introduction In their monumental work The Development of Logic Martha Kneale and William Kneale maintain that during the seventeenth century logic was in “decline as a branch of philosophy” (1962, p.345, also 298). But an era that included Leibniz, who according to the Kneales “deserves to be ranked among the greatest of all logicians” (1962, p.320), as well as Locke, who dismisses formal logic as “learned Ignorance” while writing "the first modern treatise devoted specifically to philosophy of language" (Kretzmann 1967, p.379; also 1968), suggests drama and excitement, not decline. While traditional logic was indeed in decline, logic itself was being transformed into modern mathematical logic. Moreover, the turn away from formal logic was also a dramatic turn to natural language for insight and solutions to the problems of philosophy. These two turns, the mathematical and linguistic turns of early modern philosophy, are defining features of seventeenth-century European philosophy. II. Early Modern Logic In 1626, the Dutch logician Franco Burgersdyck (1590-1635) maintained that there were three kinds of logicians: Aristotelians, Ramists and Semi-Ramists. While Aristotelians continued to develop Aristotle’s logic of categorical syllogisms and immediate inferences, Ramists sought alternative logics that captured reasoning that traditional Aristotelian logic ignored. Semi-Ramists, also called “Philippo-Ramists” after Luther’s collaborator Philipp Melanchton (1497-1560), sought a synthesis of traditional and alternative logics, which included the search for formal methods to capture non-
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Logic and Language in Early Modern Philosophy
Michael Losonsky
(In Cambridge Companion to Early Modern Philosophy, ed. D. Rutherford, Cambridge
University Press.)
I. Introduction
In their monumental work The Development of Logic Martha Kneale and William
Kneale maintain that during the seventeenth century logic was in “decline as a branch of
philosophy” (1962, p.345, also 298). But an era that included Leibniz, who according to
the Kneales “deserves to be ranked among the greatest of all logicians” (1962, p.320), as
well as Locke, who dismisses formal logic as “learned Ignorance” while writing "the first
modern treatise devoted specifically to philosophy of language" (Kretzmann 1967, p.379;
also 1968), suggests drama and excitement, not decline. While traditional logic was
indeed in decline, logic itself was being transformed into modern mathematical logic.
Moreover, the turn away from formal logic was also a dramatic turn to natural language
for insight and solutions to the problems of philosophy. These two turns, the
mathematical and linguistic turns of early modern philosophy, are defining features of
seventeenth-century European philosophy.
II. Early Modern Logic
In 1626, the Dutch logician Franco Burgersdyck (1590-1635) maintained that
there were three kinds of logicians: Aristotelians, Ramists and Semi-Ramists. While
Aristotelians continued to develop Aristotle’s logic of categorical syllogisms and
immediate inferences, Ramists sought alternative logics that captured reasoning that
traditional Aristotelian logic ignored. Semi-Ramists, also called “Philippo-Ramists” after
Luther’s collaborator Philipp Melanchton (1497-1560), sought a synthesis of traditional
and alternative logics, which included the search for formal methods to capture non-
syllogistic reasoning (Ashworth, pp.16-17; Freedman, pp.86-7; Risse, vol. II, pp.516-517,
Nuchelmans, p.104). It is useful to follow Burgersdyck and divide early modern logic into
traditional logic, alternative logic and attempts to synthesize the two.
A. Traditional Logic: Aristotelians
By the thirteenth century, Aristotelian logic consisted of two parts: old and new
logic. Old logic (logica vetus) consisted of Aristotle’s Categories and On Interpretation,
supplemented with Porphyry’s Isagoge as a general introduction to the Categories. The
remaining four texts of Aristotle’s Organon were known as the new logic (logica nova):
Prior Analytics, Posterior Analytics, Topics and On Sophistical Refutations. These four
texts were preserved in the original Greek in Sicily and southern Italy and they were also
brought to Muslim Spain in the eleventh and twelfth centuries in Greek as well as in Arab
translations.
The Prior Analytics and On Sophistical Refutations were particularly influential
because the former expanded on the meager discussion of syllogisms in the logica vetus
and the latter focussed on fallacies and paradoxes or insolubles, which were mostly new
topics. These fallacies and paradoxes based on syntactic and semantic ambiguities of
ordinary language were discussed under the headings Sophismata or Insolubilia and
motivated syntactic and semantic studies that in the thirteenth-century came to be known
as “modern logic [logica moderna]” in order to distinguish it from “ancient logic [logica
antiqua],” which hewed more closely to the traditional topics of terms, propositions,
immediate inferences, and syllogisms. The most important “modern logic” text for 300
years was Peter of Spain’s Summulae Logicales, written about 1245. It was still being
used in the seventeenth century, by which time it had enjoyed no less than 166 printed
editions (Kneale and Kneale, p.234; Peter of Spain 1972; Ashworth 1974, p.2).
All the books of Aristotle’s Organon were core texts of the European university
curriculum during the fourteenth and fifteenth centuries except the Topics (Ashworth
2
1988, p.143). The Topics discussed practical problems of reasoning, specifically, of how
to find the materials “out of which arguments are constructed” (Topics 105a20) and “how
we are to become well supplied with these” materials (101b13-14). What characterizes
the post-medieval period and informs seventeenth-century philosophy is that the practical
and epistemological perspective of the Topics moves to center stage.
But the medieval logician’s preoccupation with evaluating deductive arguments
with categorical propositions still thrived in the early modern period, particularly in
Roman Catholic countries. As just mentioned, Peter of Spain’s textbook was still being
reprinted in the seventeenth century. Outstanding contemporary contributions to
traditional logic include those of the Portuguese Jesuit Pedro da Fonseca (1528-1599),
whose Institutionum Dialecticarum was published in 53 editions between 1564 and 1625,
the Polish Jesuit Martin Smiglecki (1563-1618), whose monumental 1600-page Logica
(1618) was reprinted three times in Oxford, and the Italian Jesuit Girolamo Saccheri
(1667-1733), whose Logica demonstrativa was published in 1697 (Ashworth 1974,
pp.19-20; 1988, p.163; Nuchelmans, p.103). Thomas Wilson (1524-1581) introduced
scholastic terminology, including the term “proposition,” into English, in his Rules of
Reason (1551), but major traditional logic texts continued to be written in Latin,
particularly John Wallis’s (1616-1703) Institutio Logicae (1687) and Henry Aldrich’s
(1647-1710) Artis Logicae Compendium (1691).
B. Alternative Logic: Informal Logic, Induction and Scientific Method
Traditional logic texts mentioned inductive reasoning, but generally had very little
to say about it. Logicians in the early modern period saw this as a serious defect and
sought alternative logics that would capture non-deductive argumentation. This search
for alternatives during the period has been dismissed as not part of the development of
logic in the sense of the study of deductive validity, but as part of the “new study of
heuristic methodology” (Kneale and Kneale, p.310). If we are guided by early modern
3
conceptions of logic, however, it is a mistake to ignore the new study of methodology in a
history of the period.
1. Rudolph Agricola (1444? - 1485)
Agricola’s De inventione dialectica libri tres, written by 1480 and published in
1515, became a widely used textbook in the sixteenth century. This success marks the
decline of traditional logic (Ashworth 1988, 152-3; Jardine 1988, 181; Mundt 1994, 108-
117). The title already captures important characteristics of the alternative logic
movement. Traditionally, the term “dialectic” was used narrowly for the study of probable
reasoning, not demonstrative reasoning. Moreover, the art of judgment, which is the
evaluation of deductive inferences, was the centerpiece of logic, not the art of invention,
which focusses on the invention or construction of arguments. Agricola’s title upsets
these priorities. Dialectic now is used to cover all forms of argumentation, and invention
becomes the centerpiece of dialectic, which Agricola characterizes as “the art of speaking
with probability [probabiliter] on any question whatsoever, insofar as the nature of the
subject is capable of infusing conviction” (1992, 212, also 228). This also includes
deductive arguments, which he treats as limiting cases of probable reasoning, intentionally
blurring the line between induction and deduction (1992, p.242; also Ong 1983, p.102)
For Agricola invention is “thinking out the middle term or argument” (1992, p.16).
Guided by the traditional theory of syllogisms, Agricola sees the first step in argument
construction as the search for true propositions that involve one of the terms of the
conclusion and a new term. Agricola’s project is to design a method for finding such
propositions for a given conclusion. Once a list of propositions is generated, the arguer
then tries various pairs of propositions until a syllogism for the conclusion is found.
The tool for locating propositions is a list of topics or loci that apply to all things.
According to Agricola, each topic is a “common mark [nota] of a thing, through which it
is possible to discover what can be shown, which is also what is probable, with regard to
4
any particular thing” (1992, 20). Each topic is “a refuge and treasure chest” wherein “all
tools for fixing belief [faciendae fidei] are stored” (ibid.). Agricola’s list consists of 24
topics ordered hierarchically in order to make it easier for the arguer to work
systematically through them.
To illustrate his method, Agricola considers the case of someone arguing for the
proposition “A philosopher needs a wife” (1992, 414). Agricola runs through his 24
topics collecting various properties of philosophers, including that philosophers are pale
and thin, but in the end settles on the property of aiming to live virtuously. He then does
the same for “wife,” focusing on various virtues of a wife, including that they desire and
can bear children. This suggests various propositions that together with the proposition
that philosophers seek to live virtuously can be put together in a syllogism with the
conclusion that philosophers need wives (1992, 414-422).
Agricola recognizes that this is not an automatic procedure for finding premises.
“The value of these exercises,” he writes, “is primarily this: the description [descriptio]”
that can be built by going through the topics and that will be structured by the topics for
easy overview (1992, 422). The arguer needs to build a rich description involving the
terms of the conclusion, and then the arguer will have a fund of information, “lined up for
battle” (1992, 424). In this manner Agricola shifts the focus of logic to methods for
generating descriptions of objects.
Agricola includes a brief discussion of induction by simple enumeration, but he
distinguishes between a complete induction or enumeration of all instances, in which case
the conclusion must be true if all the premises are true, and an incomplete induction,
where the conclusion need not be true even if the premises are true (1992, 322-328).
There are good and bad incomplete inductions and he suggests that the difference between
a good incomplete and a complete induction is a matter of degree (1992, 318)
With the exception of including in his examples of syllogisms ones that include
singular propositions, which were ignored by traditional Aristotelian logic, Agricola makes
5
no contributions to the theory of deduction. But as we have seen, Agricola’s significance
lies elsewhere. By attending to informal reasoning and the problems of argument
construction, Agricola can legitimately be seen as a precursor of the field of informal logic.
Moreover, by turning to non-deductive reasoning and methods for discovering truths
about an object, including empirical truths, Agricola tills the soil for the cultivation of
inductive logic and the logic of scientific reasoning.
2. Ramus (1515-1572)
Agricola’s logic was taught at the University of Paris by Johann Sturm (1507-
1589), and it is likely that one of Sturm’s students was Pierre de la Ramée, better known
as Ramus. Legend has it that the title of his master’s thesis was “Whatever Aristotle
Stated is False [Commentitia],” indicating Ramus’ reputation as a radical critic of Aristotle
which lasted well into the seventeenth century.1 His Dialectique (1555), the first
published logic book in French, and the Latin version Dialecticae Libri Duo published a
year later, were very popular texts (nearly 250 editions were published) that took broad
swipes at Aristotelian logic. However, it needs to be pointed out that Ramus follows
tradition in important respects, more so than Agricola. For instance, his discussion of
argumentation sharply distinguishes deductive and inductive arguments and when
discussing argumentation, he focusses almost exclusively on deduction.
But there are good reasons for Ramus’ reputation. While he begins his logic in
good Aristotelian fashion with a discussion of categories and classification, his list of nine
Gaukroger (2001, 5) argues that Bacon aims to “transform the epistemological activity of the
philosopher from something essentially individual to something essentially communal.” This is an
important corrective to our understanding of Bacon and helps explain his interest in language, but it is
also true that for Bacon social interaction is a source of errors, and that human beings need forms of
reasoning that adhere to nature, not convention.
5
5
The seventeenth century was rife with projects for developing unambiguous universal
characters or languages whose syntactic structure closely tracked semantic content. A paradigm
example is John Wilkins’ Essay Towards a Real Character, and a Philosophical Language (1668), who
builds his language out of 40 basic categories, each assigned to a unique pair of consonants and vowels
for speech and a unique inscription for writing. Others who contributed or promoted universal language
projects include Cave Beck (1623-1706?), George Dalgarno (1626?-1687), Jan Komensky (1592-1670),
Francis Lodwick (1619-1694), and Marin Mersenne (1588-1648). See Rossi 1960, Knowlson 1975,
Slaughter 1982, Poole 2003, and Maat 2004.
6 Leibniz 1960, II: p.40; III: pp.400 and 573; IV: pp.438-9; NE III.v.3; and 1988, p.18. Also Lewis
1918, Carnap 1956, p.10, Mates 1972 and 1986, pp.69-73, and Adams 1994, pp.46-50.
7 The theorem appears in both Frege’s Begriffsschrift and Russell and Whitehead’s Principia
Mathematica (1997, p.23)
8 See Ishiguro 1990, pp.17-43 who argues that for Leibniz this was a principle defining the identity of
concepts, not the synonymy of expressions.
9 See Couturat 1901, Keynes 1921, Carnap 1950, Hacking 1975, Daston 1988, and Cussens 2002.
10 This appears to be a departure from Hobbes’ earlier position in De Corpore, according to which there
clearly is such a thing as "reasoning in silent thought without words" (DC I.1.3). On this tension in
Hobbes, see Losonsky 1993a.
11 Arnauld and Nicole go on to claim that because in communication we must associate ideas and
words, “this habit is so strong that even when we think to ourselves, things are presented to the mind
only in the words in which we usually clothe them in speaking to others” (Arnauld and Nicole 1996, 23-
4). This, however, is treated only as a habit, and the conclusion they draw is that they need to examine
how ideas and words are joined, and not how words are presented to the mind and play a role in
reasoning.
12 Another Cartesian work that is devoted primarily to language is Géraud de Cordemoy’s Discours
physique de la Parole [A Philosophical Discourse Concerning Speech] (1668). Cordemoy aims to
examine speech more closely in order to better understand which aspects of language use can be
attributed to the body alone and which require a rational soul.
13 On the relation between Locke and Hobbes, see Rogers 1988.
14 Compare Jolley 1999 (162, also see 144), who writes: "Locke’s major theses concerning the
metaphysics and epistemology of classification can be understood independently of his teachings about
language."
15 On Leibniz and Bhme, see Aarsleff, pp.42-83 and Losonsky 1992 and 1993b.16 In Academiarum Examen, or the Examination of the Academies (1653) the ‘Behemist’ John Webster
(1610-1682) suggests that the language of nature is the universal characteristic that Wilkins and others
were trying to construct. Wilkins and the mathematician Seth Ward (1617-1689) respond in Vindiciae
Academiarum (1654), arguing against the possibility of a language of nature and defending artificial
universal language projects. See Debus 1970, Aarsleff 1982, p.262, and Losonsky 2001, pp.105-110.
17 The mirror metaphor also opens Leibniz’s posthumously published work Unvorgreifliche Gedanken,
betreffend die Ausbung und Verbesserung der teutschen Sprache [Unassuming Thoughts Concerning
the Practice and Improvement of the German Language]: “It is known that language is the mirror of the
understanding” and when understanding flourishes, so does the use of language (1966a, p.449).
18 Leibniz 1960, IV: p.424; 1982, p.275. On the constitutive role of symbols in reasoning in Leibniz’s
philosophy, see Dascal 1987 and Losonsky 2001, pp.160-3 and 171-3.
19 Leibniz 1960, VII: 264. For Leibniz, "[t]hat is said to express a thing in which there are relations
which correspond to the relations of the thing expressed" (1962, VII: 263-4). Also see 1960, VI: 327
and 1962, VI.6: 131.
20 On the distinction between surface and underlying form in Leibniz, see Brekle 1971 and Dascal 1987,