32
INTRODUCTION TO JEAN BURIDAN’S LOGIC
the various divisions of supposition.
The theory of supposition should not be assimilated to formal logic,
but to the philosophy of logic; it is the mediæval theory of reference. The
claim that supposition, or any relation R, is in fact a reference-relation
is justified if the R-relation is sufficiently similar to our uses of ‘reference.’
There are two main uses in contemporary philosophy of language. The first,
associated with Davidson, takes the R-relation to figure in a compositional
theory of how sentences acquire their meaning and truth-value, especially
if in this theory the R-values are the referents. As we shall see in Section
6.8, the theory of supposition does provide an account of the truth-value of
sentences, although not their meaning. Hence this first way is inconclusive.
The second way, with Quine, is definitive. On this approach we
take the variables of quantification to be the paradigm case of the sort of
term having an R-value. Now this approach is usually rejected out of hand
because there are no variables of quantification in supposition theory: terms
are bound, not variables. Yet such rejection is facile; there may be no overt
variables of quantification, but we can look at the mediæval treatment of
variable-binding operators in the language. In particular, consider the case
of anaphoric pronouns whose antecedents are quantified terms, such as ‘he’
in “Some man has a daughter and he loves her.” This is a case of ‘relative
supposition’ (discussed in detail in Section 6.4) in which the supposition
of ‘he’ is determined by the supposition of ‘Some man.’ Equally there are
various indexing devices available; in the example, gender disambiguates the
particular antecedent to each pronoun. Such antecedents can be multiplied
and indexed at will in supposition theory. Hence supposition has the key
characteristics of a reference-relation, and may be treated as such.
Supposition theory is a theory of reference. It is a unified theory,
which has as its goal to specify what a term is used to talk about in a given
sentence.
52
The various divisions of supposition illustrate the ways in which
a term may supposit for something. Buridan’s division of supposition can
be put in outline form as follows:
52
The claim that supposition theory is in fact a unified theory, and that it should be
construed as I suggest in Section 6, is a matter of controversy. Some of the arguments
are taken up in the following sections. My interpretation has the virtue of making
supposition theory a reasonable philosophical enterprise, as well as fitting in well
with other branches of mediæval logic, such as fixing coreferentiality by the Doctrine
of Distribution in the theory of the syllogism.
The distinction of referential and
attributive uses of the particular sign given below is, it seems to me, well-confirmed
by the remarks made by Buridan in his discussion of determinate and non-distribute
confused supposition. but the reader should not that the interpretation I offer is a
matter of dispute.
c Peter King, from Jean Buridan’s Logic (Dordrecht: D. Reidel 1985) 3–82.
INTRODUCTION TO JEAN BURIDAN’S LOGIC
33
SUPPOSITION
Improper
Proper
Material
Personal
Discrete
Common
Relative
Absolute
Natural
Accidental
Determinate
Confused
Distributive
Non-Distributive
Of course, we must specify the scope of Buridan’s theory; this is the point of
the first division, the distinction between proper and improper supposition.
A term as improper supposition when it is used rhetorically in a sentence and
not literally. The theory of supposition is developed only for the literal uses
of terms, as indeed are modern theories of reference, and both are equally
far from explaining the meaning of ‘literal’ or understanding non-literal uses
of language.
53
Personal and material supposition illustrate the kind of thing a term
may refer to in a sentence, namely its ultimate significate or itself, whether
as inscription, utterance, or concept. the various divisions of personal sup-
position specify how many of its ultimate significates a term may stand for:
exactly one (discrete); at least one (determinate); several (non-distributive
confused); all present instances (distributive); all past, present, and future
instances (natural). The status of a term, as defined in TS 6.1.1, requires
this interpretation. The theory of supposition, of course, does not say ex-
actly which things a term stands for; like any semantic theory, that is left
to the facts of the matter to decide—terms may fail to refer, the reference
will change with changes in the world, and so on. But it does specify the se-
mantic function a term may have in different sentential context. This point
is important: supposition theory will explain the semantic role of terms in a
53
This is surprising, since Buridan holds that the speculative sciences each study a
single Mental term, to which other terms are related as individual men in an army are
related to their leader (QM 4.3 fol. 14ra and QSP 1.2 fol. 3rb–va): such attributions
to a primary element are semantic but not exactly ‘literal,’ and we would expect a
theoretical account of such uses.
c Peter King, from Jean Buridan’s Logic (Dordrecht: D. Reidel 1985) 3–82.