INTRODUCTION TO JEAN BURIDAN’S LOGIC
67
formed, are respectively affirmative or negative in quality.
91
This is the key
to Buridan’s semantic characterization of the dictum de omni et nullo: his
talk of supposition is entirely in the formal mode. Syllogistic is a branch
of the theory of formal consequence, and so the test for the acceptability of
a syllogism is whether it satisfies the Uniform Substitution Principle, but
the principle applies only if certain relations among the supposition of the
terms obtain. In a particular sentence the relations are made explicit by
the nature of the syncategorematic terms present, and Rule 1 and Rule 2
state how the terms must be semantically related, i. e. in terms of their
supposition.
The clause ‘by reason of the same thing for which that third term
supposits’ is the central semantic contribution of the dictum de omni et
nullo, because it makes explicit the coreferentiality required for the same
term appearing in different sentences. Coreferentiality is the underpinning
of the theory of inference. Now there are many ways of fixing reference;
how can we tell if the clause ‘by reason of the same thing for which that
third term supposits’ is satisfied? That is: when are different occurrences
of a middle term coreferential? There are three cases. First, the third
term in question may be a singular referring expression, that is, a discrete
term, and here there is no trouble, for a discrete term can supposit only
for a single thing, as a matter of semantics. A syllogism with a discrete
term as syllogistic middle is called an ‘expository syllogism’ (Theorem III-4
in TC 3.4.23–25). Second, if the middle is a common term, then we may
simply stipulate that the syllogistic extremes are called the same or not
the same for the same thing(s) for which the middle term supposits; this
stipulation must be visible in the sentences, since the syllogism is a formal
consequence, and consists in adding an identificatory relative-term to the
minor (Theorem III-6 in TC 3.4.29–30).
The third case also involves a common middle term ,but where there
is no such stipulation of suppositional sameness. We do not simply want
coextension here; we want coextension as a matter of the semantics, for
a syllogism is a formal consequence—indeed, a matter of the semantics
obvious from the syntactic form. The answer is given by the theory of
distribution.
Buridan does not specifically address distribution while discussing
the syllogism, but he does not need to: it is covered in his account of com-
mon personal supposition. A term is distributed if in a sentence it is taken
91
We have seen Buridan make the same point in the material mode, for example when
he says that an affirmative sentence “indicates” that its terms supposit for the same
thing(s), as determined by the correspondence truth-conditions listed in Section 6.9.
c Peter King, from Jean Buridan’s Logic (Dordrecht: D. Reidel 1985) 3–82.
68
INTRODUCTION TO JEAN BURIDAN’S LOGIC
to supposit for all it signifies, that is, if it is used to talk about or refer to
everything it signifies. The most obvious case of distribution is where a term
is joined with a distributive sign (a universal quantifier), not in the scope
of a negation. Buridan’s rules for distributive and non-distributive supposi-
tion, discussed above, explicitly state when a term is said to be distributed
and when it is not. The theory of distribution, and hence of syllogistic, is
therefore of widespread applicability, but if we confine ourselves to simple
sentences as on the Square of Opposition, we may give some rules for distri-
bution: universals distribute subjects, negatives distribute predicates, and
no other terms are distributed. Hence in an A-form sentence the subject
alone is distributed, in an E-form sentence the subject and the predicate are
distributed, in an I-form neither subject nor predicate is distributed, and in
an O-form sentence the predicate alone is distributed.
92
The motivation for the theory of distribution is clear; we want to
avoid the case in which one extreme is called the same or not the same
as part of what the middle term supposits for, while the other extreme is
called the same or not the same as part of what the middle term supposits
for, while the other extreme is called the same or not the same as the other
part of what the middle supposits for. In that case there is no connection
between the extremes through the middle, and no inference will hold by the
semantics alone. Distribution is a way of making sure that the foregoing
case does not occur, by talking about everything the middle term supposits
for. Indeed, it obviously follows from the theory of distribution that no
syllogism made up with two negative premisses is acceptable (Theorem III-
2 in TC 3.4.15), as Buridan notes. The principles governing the doctrine of
distribution are given in Theorem III-7 and Theorem III-8 (TC 3.4.34–36).
8.3 Reduction and Proof-Procedure
Buridan follows tradition in taking the first four moods of the first
figure to be evident of ‘perfect’: Barbara, Celarent, Darii, Ferio. Their ac-
ceptability is shown directly by the preceding principles; what I have called
Buridan’s “proof-procedure” for syllogisms is his way of reducing all other
syllogisms to these.
93
In each of the various forms of assertoric and modal
92
Classically, a syllogism is acceptable if and only if (i ) the middle term is distributed
exactly once; (ii ) an extreme term is distributed at most once; (iii ) if the conclusion
is negative exactly one premiss is negative; (iv ) if the conclusion is affirmative neither
premiss is negative.
93
The traditional name for each mood indicates the method of reduction in the following
way: (i ) the initial letter of the name indicated which of the four basic first-figure
syllogisms it is reduced to; (ii ) the first three vowels characterize the quantity and
quality of each premiss (and all other vowels are ignored); (iii ) the letter ‘s’ following a
c Peter King, from Jean Buridan’s Logic (Dordrecht: D. Reidel 1985) 3–82.