Definition of Terms
In many of the cases so far considered, Aristotle would justify the implicit or explicit comparisons he himself uses by referring to the general rule, of which both items compared can be seen as instances, a rule which can, or should in principle, be supported independently. But the broader questions that Aristotle's theory of meaning and his demand for precision and the literal raise concern also his reaction to and criticism of many of the complex and problematic theoretical terms that his predecessors and contemporaries used in their natural philosophical speculation, whether or not Aristotle saw these as, or as involving, metaphor. In some instances he proceeds in the way we might expect from his criticisms of the obscurity of metaphor and the like and from his general statements requiring the strict use of terms: that is, he goes all out to purge the terms of ambiguity and vagueness and to establish a single clear-cut definition, even though the strain
that this imposes on some parts of his scientific enterprise are, at times, as we shall see, considerable. In other cases, however, he allows that a term may be "said in many ways,"
The pair heavy/light had been used in ordinary Greek primarily of what is difficult or easy to carry, though in both cases with a fair range of other meanings or applications as well, including difficult, and easy, more generally.[68] But signs of the strain under which the naive conception was coming are already visible in pre-Socratic philosophy, where various correlations are proposed with other pairs of opposites (such as dense/rare) or with the elements as well as with movements,[69] and
[67] See especially Owen 1957/1986, 1960/1986, 1965a/1986.
[69] See, for example, Parmenides fr. 8.56ff., Empedocles fr. 21, together with the admittedly often tendentious reports and criticisms in Aristotle (e.g., De generatione et corruptione [GC ] 314b20ff., 315a3ff., 10f.), and in Theophrastus (De sensu, Sens. , e.g., 59ff.). Whether or not weight is a primary property of the atoms for Leucippus and Democritus is much disputed (see below, Chap. 5 n. 41), but it appears from passages in both Aristotle (GC 326a9f.) and Theophrastus (Sens. 61) that heavy and light were sometimes referred to the size of atoms. Compounds, however, could also be distinguished by the proportion of atoms to void or the amount of void they contain. Theophrastus also reports (Sens. 62, 68) that light and heavy were correlated or associated with rare and dense (as also still in Aristotle Ph. 217b11ff.).
where Aristotle complains with some justice that the capacities in question were generally left undefined.[70]
Plato in the Timaeus first follows up the popular association with below and above and emphatically rejects the idea that this second pair relates to two distinct regions in the universe.[71] The universe is spherical, so it makes no sense to talk of one part of the sphere being above or below another. Imagining—boldly—a thought experiment in which someone stands in the heavens at the interface of fire and air and forces a larger, and a smaller, quantity of fire towards the air (i.e., towards the centre), he says that it is obvious that the smaller quantity will be moved more easily.[72] It then will be "lighter" and tend "upwards," the larger will be "heavier" and tend "downwards"—though "downwards" in this case is to the periphery, "upwards" away from it, whereas the ordinary Greek assumption was that, on earth at least, more fire is "lighter."[73] What is light in one region, Plato is prepared to say,[74] is the opposite of what is light in the other. Several aspects of the interpretation of this text remain highly disputed,[75] but it is beyond doubt that Plato has radically redefined heavy and light: they do not just depend on the quantities of the material concerned but, like up and down, are relativised to where in the universe you are or to which element is in question.
Aristotle, in turn, is no less emphatic that certain conventional views are mistaken. Modifying Plato's idea of the importance of the element in which the real or imagined weighing takes place, he distinguishes between the two simple bodies that are heavy (or light) absolutely (that is, earth and fire) and the two that are so only relatively
[70] De caelo (Cael. ) 308a3f.; cf. Theophrastus Sens. 59f., who makes partial exceptions of Democritus and Plato.
[71] Ti. 62c ff.
[72] Ti. 63b–c.
[73] This is stated to be obviously true and to hold universally, by Aristotle at Cael. 308b3ff., 13ff., 18ff., in the course of his quite exceptionally polemical criticisms of the theories in the Timaeus .
[74] Ti. 63d–e.
[75] See, for example, Solmsen 1960, pp. 275ff.; Hahm 1976, pp. 59ff., 70; O'Brien 1984.
(water and air: that is, relative to other elements).[76] He is confident that air is light in comparison with earth and with water, but he raises as a puzzle the question of whether air has weight in air, deciding the issue positively by invoking a purported trial which, he claimed, showed that in air an inflated bladder weighs more than an uninflated one.[77] Evidently here the possibility of carrying out a measurement was appreciated, though its difficulty and delicacy are reflected in the fact that when Aristotle's conclusion was challenged by later commentators, first by Ptolemy and then by Simplicius, they obtained quite different results from the same test.[78]
Aristotle further diverges from Plato in insisting that heavy (and down) are always to be defined in relation to movement to —and light (and up) in relation to movement away from —the centre of the universe, deemed to coincide with the centre of the earth.[79] Plato too had held that the universe and the earth are spherical, but Aristotle now demonstrates the latter thesis with a battery of arguments.[80] Some of these, it is true, are not independent of the issue concerning the nature of heavy and light, for they attempt to show the earth's sphericity as a consequence of the doctrine that the natural movements of the simple bodies are to, or from, the centre of the universe, where Aristotle assumes that heavy bodies do not move downwards in parallel lines.[81]
[76] See, for example, Cael. 4.1.308a7ff., 4.4.311a15ff., and cf. Seeck 1964, pp. 108ff.; Hahm 1976, p. 62.
[77] Cael. 311b9ff.
[78] Simplicius reports Ptolemy's result (that the bladder weighs less) and then proceeds to describe his own attempt to verify the facts at In Aristotelis De caelo commentaria (In Cael. ) 710.24ff. The weight of a bladder inflated with air is also discussed in the Problemata , 25.13.939a33ff., and in Anon. Lond. 31.34ff., 32.22ff.
[79] E.g. Cael. 308a14ff.
[80] Cael. 297a8–298a20; cf. Ti. 62e, 63a. The importance of this point will be the greater if Furley 1976, pp. 97f., is right to argue that the rival atomist account of the natural motion of atoms (in Epicurus, certainly, and in Furley's view also in Democritus), according to which they move perpendicularly "downwards" in space, depends crucially on the doctrine that the earth is flat.
[81] Cael. 296b6ff., 18ff., 297b17ff.
But some offer good independent grounds for his thesis, notably arguments that appeal to astronomical data, first, to changes in the visibility of the stars at different latitudes, and especially in the circumpolar stars that never set,[82] and, second, to the shape of the earth's shadow in eclipses of the moon.[83]
The example of heavy and light vividly illustrates the meaning shifts that occur as theory develops, shifts that are similar in kind to those that have been explored from later science, where one example often cited is that between the notions of mass in Newton and in Einstein.[84] While in assessing just how radical those ancient meaning shifts were it is fair to recognise that the theoretical framework within which heavy and light were entrenched in ancient debate was a good deal less sophisticated than many more modern examples, we should not, on the other side, underestimate just how much of Aristotle's account of both the sublunary and the superlunary region was at stake—a point not lost on some of his ancient critics such as Philoponus.[85] Meanwhile Aristotle's own view of the matter was that he was providing heavy and light with clear, univocal definitions, and ones that incorporated the adjustments to popular notions necessary to take into account the doctrine of the spherical earth.
My second example was the pair hot/cold. Once again Aristotle
[82] Cael. 297b30ff.; cf. further below, Chap. 5 at n. 57.
[83] Cael. 297b24ff.
[84] See, for example, Kuhn 1964/1977, p. 259 n. 30. Aspects of the problems of meaning invariance have been discussed, taking as illustrations the differences between Aristotle and his predecessors on the question of up and down, falling and rising, by Feyerabend 1962, p. 85 (1981a, pp. 85ff.), and by Hesse 1974, pp. 33ff.
[85] In his De aeternitate mundi contra Aristotelem , for which our chief source is the extensive quotations in Simplicius, Philoponus explores, among other things, the difficulties that Aristotle's theory encounters in squaring the doctrine of the four simple bodies with that of the two directions of natural sublunary movement, and he mounts a sustained attack on the Aristotelian doctrine of the fifth element, aether, lacking the primary qualities hot, cold, wet, and dry. See especially Wildberg forthcoming and cf. also the De Aeternitate mundi contra Proclum 13.6 and 13–17, 492.5ff., 512.17–531.21; cf. M. Wolff 1978, p. 156.
complains about the ambiguities of common usage—and about the disagreements among earlier theorists.[86] Sometimes touch is invoked as the criterion, sometimes various effects (melting, burning, and the like) that the substance claimed to be hot, or cold, has on other things, and the conflicts between these criteria are discussed. Thus, boiling water imparts heat better than flame, but flame can burn; again, boiling water, he says, is hotter to the touch than olive oil, but cools and solidifies more quickly.[87] The consequences of unclarity on this, and on the nature of the dry and the wet, are particularly drastic since, as he puts it, "it seems evident that [these four primary opposites] are practically the causes of death and of life, as also of sleep and waking, of maturity and old age, and of disease and health."[88] More even than that, they provide the basis of Aristotle's own essentially qualitative element theory.
In the De generatione et corruptione he presents not only a very full discussion of issues connected with element theory and of rival views to his own, but also a set of definitions of the four primary opposites, to which he believes other qualitative differences (hard and soft, rough and smooth, viscous and brittle, and so on) can be reduced.[89] "Hot," he says, is "that which combines things of the same kind" (
[86] See, for example, PA 648a21ff., 36ff., and cf. also 649b9ff., GC 330a12ff., on dry and wet.
[87] PA 648b12ff., 17ff., 26ff., 30ff. Aristotle further exploits the distinction between what is hot per se and what hot per accidens , e.g., PA 649a5ff., and between what is hot potentially and what hot actually, e.g., PA 649b3ff.
[88] PA 648b4ff.
[89] GC 2.2.329b7ff.
[90] GC 329b26–32.
totle does not proceed per genus et differentiam , but he evidently aims to give clear and distinct characterisations of the four primary opposites. The somewhat abstract nature of his account is, however, striking. Moreover, as soon as we look at the range of types of case where he uses the four opposites, we encounter instances where his initial characterisations seem inappropriate and hard to apply.
This is particularly true when he is discussing the role of vital heat, one of the chief foundations of his whole biology. It is important, from his point of view, that it is heat in question, since this gives him his link with his general physical theory of the elements. But it is a quality that sometimes seems remote both from anything that might be suggested by the definition "that which combines things of the same kind" and from what might be thought to have some justification either in terms of popular usage or, indeed, of appeals to subjective impressions. Thus in one of his several discussions of the main groups of animals[91] he arranges them in a hierarchy according to their methods of reproduction, which are themselves correlated with the four primary opposites. The most perfect animals, the Vivipara, are "hotter and wetter and less earthy by nature"; next come the ovoviviparous animals, the cartilaginous fishes (sharks and rays), which are cold and wet; the third and fourth groups are Ovipara that lay perfect, and those that lay imperfect, eggs, and these are hot and dry, and cold and dry, respectively; and the fifth and final group, the larvae-producing animals such as the insects, are "coldest of all."
We can see why he claims that the Vivipara, which include humans, are the most perfect creatures, and also why they are warmer than, for instance, fish. Yet it is a puzzle why he should claim that the oviparous
fish are cold and dry —though we may notice that he has already used the combination cold and wet for the (superior) ovoviviparous fish.[92] The whole represents a schema that appears to owe more to Aristotle's preconceptions of the hierarchy of the animal kingdom, and especially to his views on the distance of the different groups from humans at the top of that hierarchy, than it does either to empirical considerations, or even to considerations derived from the general definitions of hot, cold, wet, and dry set out in the De generatione et corruptione .
Nor is it only in connection with animal taxonomy and vital heat that questions of this kind arise. At the very heart of the physical theory, some of the correlations proposed between simple bodies and pairs of primary opposites pose problems. Earth, Aristotle suggests, is cold and dry, air hot and wet, water cold and wet, fire hot and dry.[93] The "wetness" of both air and water corresponds, of course, both to the range of the term
[92] Compare, however, the account given of the dietetic qualities of different sea-animals at Vict. 2.48 (L) 6.548.9ff., where fish in general, and also shellfish, are said to be "dry," though the cartilaginous fish "moisten" ([L] 6.550.7f.). But in the discussion in the medical writer many other qualities are also taken into account, and differences are suggested between the flesh of different types of fish.
[93] GC 330b3ff.
[94] Cf. GA 761b18ff. and the discussion of the shape of flame in Theophrastus Ign. 52ff., 35.6ff. While Theophrastus represents flame as generally having a pyramidal shape, he recognises also not just that fire is a kind of movement (see above, Chap. 3 at n. 170), but also that flame is constantly moving and flowing (Ign. 54, 37.3ff.).