The Revolutions of Wisdom "d0e11404"

Focal Meaning, Proportional Analogy, and Homonymy in Aristotle's Science

Heavy and hot are, then, two terms of great theoretical importance where, diagnosing confusions in their use, Aristotle aims to establish and adhere consistently to a single univocal definition but in practice encounters difficulties in following through this programme. Elsewhere, however, as we said, he uses the concept of focal meaning, which preserves the centrality of a primary significance but allows a cluster of others to be related to it. We cannot do justice here, clearly, to the intricacies of this important concept, and of the related but distinct notion of proportional analogy, but one of Aristotle's canonical examples will serve as the briefest of introductions.^[95] "Healthy," we are told,^[96] is said primarily in relation to health itself, but also derivatively of signs of health (as when a blooming complexion is said to be healthy) or of what promotes or preserves health (as when regular exercise and a kind of climate are said to be healthy). "Healthy" is not to be understood and explicated in the same way when said of a climate or of exercise as when said of a patient who has recovered from illness, but the term is not merely ambiguous or homonymous, since all the other uses are to be connected with a primary one in relation to however we define health itself. This allows for what I have been calling semantic stretch, while it still privileges a primary application.

Here, then, is a device of great power and scope which Aristotle in fact uses repeatedly and to particular effect, as we noted, in connection with some high-level metaphysical principles such as essence and being. In his Physics and elsewhere such concepts as place,

, or what it is to be "in" something, and contact,

, are elucidated

[95] Apart from Owen's own discussions, noted above, n. 67, see also J. Barnes 1971, Hamlyn 1977–78, Tarán 1978, Ferejohn 1980, Irwin 1980–81, Fine 1982.

[96] Metaph. 1003a34ff., cf. Top. 106b33ff., Metaph. 1060b37ff. See, for example, Owen 1960/1986, pp. 192ff., 198ff., 1965a/1986, pp. 259ff., and cf., e.g., Mackinnon 1965.

― 199 ―

in such a way.^[97] It is characteristic of his discussion to move outwards from a central, familiar, unpuzzling usage, gradually widening the range of what is to be included under the original rubric.

He proceeds in a similar way when clarifying the concepts of matter and form, and potentiality and actuality, for example, via the notion of proportional analogy, though this is not a kind of focal meaning so much as an alternative to it.^[98] Thus matter—where the term he coined,

, originally meant just wood, of course—is used first of the stuff physical substances are made of, but also of the substratum of change more generally.^[99] That there is something that underlies and survives change is illustrated by such straightforward cases as a man becoming pale or educated, but the idea is then applied not just to the bronze the statue is made from, but also more problematically, in embryology, to the matter—the menses—that Aristotle holds to be supplied by the mother to the embryo.^[100] Again, matter is said to individuate members of the same species, which are the same in form but numerically distinct,^[101] and he feels entitled also to speak of intelligible matter in, for example, mathematics.^[102] Two identical triangles used in a geometric proof are differentiated by their intelligible matter: not the triangles I draw on the blackboard, but the triangles we have specified and are reasoning about. As matter is what is characterised by form, and the genus receives determination from the species (also

), the genus too can be called matter.^[103] But as that last example particularly illus-

[98] As becomes clear at EN 1096b26ff., for example, when he discusses "good."

[99] As, for example, in Ph. 1.6ff., 189a11ff., especially 7, 191a7ff., 9, 192a3ff.

[100] See, for example, GA 727b31ff., 729a10ff., 28ff.; cf. G. E. R. Lloyd 1983a, p. 97.

[101] Metaph. 1034a7f.

[102] See, for example, Metaph. 1036a9f., 1037a4ff., 1045a33ff.; cf. 1059b14ff., 1061a28ff., and cf. Lear 1982, p. 181.

[103] See especially Metaph. 1024b4ff. (but cf.b9ff.), 1045a34f., 1058a21ff., cf. 1016a24ff., 1023b2, 1038a5ff., 1071a36ff., Ph. 200b7f., GC 324b6ff.

― 200 ―

trates—where whether the genus is matter, or is just like matter, is disputed^[104] —the point at which Aristotle is using the term in an "as if" way (that is, in a way he has to recognise as such) may be quite unclear and controversial.

Thus in a variety of contexts, dealing especially with the fundamental notions that underpin the whole of his philosophy and science, Aristotle offers a kind or kinds of analysis that while certainly not in direct contradiction to anything in his logic, nevertheless represent a certain relaxation of the requirements of univocity and universal, perse , predication laid down in his accounts of definition and demonstration.^[105] The balance between these two points is delicate—the more so as we do not have an extended formal discussion of focal meaning and so have to rely on the scattered comments that occur in texts that deploy the notion. But clearly, first, there is no question of Aristotelian metaphora being involved, in the sense of the transference of a term from one field to another. Nor, secondly, is focal meaning a matter of a comparison to be justified by reference to an (in principle) independently verifiable general rule exemplified in the particular cases compared. Thirdly, while the extent to which focal meaning is proposed by Aristotle as a tertium quid between what he calls synonymy (i.e., univocity) and homonymy is disputed,^[106] at the very least it is marked out

[104] See, for example, Balme 1962a; A. C. Lloyd 1962, 1970; Wieland 1960–61/1975, pp. 136f.; Rorty 1973, 1974; Grene 1974; M. J. White 1975; Lear 1982, p. 181; I. Mueller 1987.

― 201 ―

from other cases of homonymy in that a systematic relationship can be exhibited between primary and peripheral significances. Nevertheless, fourthly, despite his evident dislike of some modes of reasoning based on likenesses, and despite the demand for the strictest univocity in all terms used in demonstrative reasoning, focal meaning and proportional analogy tacitly mark a departure from that ideal in many key concepts. This is not the reintroduction of imagery, but it is a loosening of the straightjacket of univocity, an implicit recognition (maybe) that the requirements specified for definition and predication in the Posterior Analytics are an ideal .^[107]

― 202 ―

The tension here mirrors and indeed exemplifies a further deep-seated tension within Aristotle's divergent statements on the relationship between philosophy and dialectic. Often that relationship is expressed in terms of a series of contrasts: the philosopher works—or can work—on his own, the dialectician in conjunction with a partner;^[108] the philosopher deals with truth, the dialectician with opinion—for dialectical syllogisms reason from generally accepted views, demonstrative ones from premises that are true and primary.^[109] Yet on other, admittedly rarer occasions, Aristotle recognises a fundamental role for dialectic, in the sense of the critical scrutiny of received opinions, as a means of securing the primary principles of each science.^[110] But the snag is that the primary principles used in demonstrations, including definitions, are required to be better known than and prior to the conclusions.^[111] To get round the difficulty Aristotle would no doubt invoke his distinction between what is "better known to us" and what is "better known simpliciter,"^[112] but quite how the move from the first to the second is to be made, or how we are to recognise we have accomplished it when we have accomplished it, can be problematic^[113] —

[108] See, for example, Top. 155b7ff.

[109] See, for example, Top. 100a18ff., 105b30f., Metaph. 1004b17ff.; cf. Rh. 1355a33ff.

[111] As at APo. 1.2.71b20ff., referred to above, n. 105.

[112] As at APo. 1.2.71b33ff.; cf. also Top. 142a6ff., Ph. 184a16ff., Metaph. 1029b3ff., EN 1095b2ff.

[113] See, for example, Wieland 1962/1970, pp. 69ff.; Mignucci 1975, pp. 30f.; S. Mansion 1979.

― 203 ―

particularly when we are dealing with cases that may involve focal meaning or proportional analogy. In any event the chief point that remains is that what is presented in the Posterior Analytics has to be seen as an ideal to which no more than approximations are to be expected in some key areas of inquiry.