54
INTRODUCTION TO JEAN BURIDAN’S LOGIC
(3) The truth-value of the E-form (e. g. “No S is P ”) and O-form (e. g.
“Some S is not P ”) sentences is determined by (1) and (2).
Now (1) and (2) are rather similar to giving truth-conditions in terms of
set inclusion among the extensions of the terms. But note that they are
stated as necessary conditions for truth, not as sufficient conditions. That is
because of problems with Liar-sentences (TC 1.5.5–1.5.7; Soph. 2 Theorem
12 and Soph. 8 Sophisms 7 and 11 especially).
Recall that Buridan takes sentences to be assertions, and hence the
sentential form requires certain contextual prerequisites be met for the sen-
tence to count as a sentence. An affirmative assertion (sentence) indicates
that its terms supposit for the same, according to the requirements of the
given sentence; a negative sentences indicates the opposite (TC 1.5.1–2).
But this is not all, for Buridan also holds what Hughes has felicitously
named the Principle of Truth-Entailment ([1982]) 110): a sentence ‘p’ and
a sentence of the form ‘A exists,’ where ‘A’ names p, together entail a sen-
tence of the form ‘A is true.’ (Naturally ‘p’ as it occurs in the latter sentences
is in material supposition and only equiform to the original sentence.) This
principle is quite intuitive if we remember that sentences are assertions: an
utterance which actually counts as an assertion contextually presupposes
that the assertion is true. Buridan sometimes expresses this loosely, by say-
ing that the truth of a sentence is nothing other than that very sentence
itself (QM 2.1 fol. 8vb). He is thinking of Mental sentences, of course, but
the point is that ‘truth’ is eliminable: we do not need a truth-predicate.
What are the causes of the truth of falsity of a sentence? Buridan
discusses this question in TC 1.2 and QM 6.8. We have seen (in Section
5.5) how Buridan rejects various suggestions about what a sentence signifies,
especially the complexe significabile. But what then is the cause of the truth
or falsity of a sentence? Some fact or event? But then it seems as though we
have to admit negative facts, or perhaps future facts, and the like. Modern
logicians sidestep some of these difficulties by taking negation as a sentential
operator, so that negated sentences are molecular; their truth-value is then
derived from the truth-values of atomic sentences. Buridan however, admits
as basic two forms of the copula (‘is’ and ‘is not’), and so cannot treat the
truth-value of negative sentences as derivative. Rather, he takes a more
radical line: there is no cause of the truth of a negative, just as there is
no cause of the falsity of an affirmative. The causes of the truth of an
affirmative are equally causes of the falsity of the corresponding negative:
the way things are. And that is the end of the story. Further support for
this view can be found in the fact that Buridan states correspondence truth-
conditions only for A-form and I-form sentences, since there is no cause of
c Peter King, from Jean Buridan’s Logic (Dordrecht: D. Reidel 1985) 3–82.
INTRODUCTION TO JEAN BURIDAN’S LOGIC
55
the truth of the E-form and O-form sentences.
Buridan also has several remarks about the “number” of causes of
truth, which are as one would expect: the sentence “All men are sexists”
has more causes of its truth that “Some men are sexists.” These causes of
the truth of the sentences are related by set-inclusion, it should be noted.
Such principles are crucial for proving equipollence and conversions, as we
shall see in Section 7.3.
7. Consequences
7.1 Conditionals, Inferences, and Consequences
Buridan’s theory of consequences covers material treated by modern
logic under the separate headings of a theory of conditionals and the rules
of inference. These are not distinguished by the logicians of the fourteenth
century;
78
this need not be an error—we need to recall the philosophical
motivations for drawing the distinction initially.
Consequences in modern logic are distinguished as conditionals and
rules of inference; they are not merely equivalent: to show that it is possible
to pass from one to the other a Deduction Theorem is needed, which is not a
trivial matter. It is not obtainable in incomplete systems. Each is specified
syntactically, but conditionality is represented by a sign in the language (as
primitive or a defined abbreviation) appearing in formulae, whose behavior
is given by axioms, while rules of inference are neither stated in the language
nor in the syntactic recursive grammar, but metalinguistically govern the
production of a formula from other formulae. Inference rules are tied to
deducibility and provability, and hence to validity by the notion of syntactic
consequence; conditionals are loosely tied to truth and interpretation.
Buridan, as noted, does not distinguish an object-language from a
metalanguage, so it would be difficult for him to arrive at precisely our
distinction between conditions and rules of inference.
79
Yet aside from the
incompleteness of logical systems, a worry it would be anachronistic to
78
Things were not always so: the twelfth-century philosopher and logician Peter Abelard
clearly distinguishes arguments and conditionals, and even argues for a Principle of
Conditionalization and Deconditionalization—a mediæval “deduction theorem.” But
Abelard’s work seems to have been completely lost to the fourteenth century, for
unknown reasons. The unpublished studies of Christopher Martin on Abelard and
Boethius on conditionals are invaluable.
79
The problem is exacerbated if Spoken language is taken as primary, for the distinction
is usually not drawn verbally; if drawn at all, it typically relies on non-verbal cues.
Buridan mentions such cues in TS 2.6.79 when distinguishing composite and divided
senses.
c Peter King, from Jean Buridan’s Logic (Dordrecht: D. Reidel 1985) 3–82.