¹ In the interpretation of this argument against chance, I follow J. Cooper in maintaining that the results that Aristotle considers as occurring either teleologically or by chance are those that benefit organisms: ‘Aristotle on Natural Teleology’, in Schofield, M. and Nussbaum, M. (eds.), Language and Logos: Studies in Ancient Greek Philosophy (Cambridge, 1982), pp. 197–222, esp. pp. 207–9CrossRefGoogle Scholar .

² An influential view is that Aristotle's metaphysics is based on his biology and that organisms provide the fundamental model for his metaphysical and scientific conceptions. Marjorie Grene has provided perhaps the most popular example of the ‘biological’ interpretation of Aristotle: A Portrait of Aristotle (Chicago, 1963)Google Scholar . Montgomery Furth provides a recent defence of the ‘biological’ interpretation, and he maintains that Aristotle's metaphysics of substance is a ‘deep theoretical foundation’ for his biological sciences: Substance, Form, and Psyche: An Aristotelian Metaphysics (Cambridge, 1988), e.g. p. 5Google Scholar . Others, such as Wilfrid Sellars, have argued for the primacy of artefacts in the development of Aristotle's philosophical conceptions: ‘Aristotle’s Metaphysics: An Interpretation’, Philosophical Perspectives (Springfield, 1967), pp. 73–124Google Scholar . William Jacobs argues contra Grene for the importance of artefacts in the development of Aristotle's, overall philosophy: ‘Art and Biology in Aristotle’, Paideia: Special Aristotle Issue (1978), 16–29Google Scholar . Jacobs, however, refuses to align himself completely with Sellars, because he believes that some of Aristotle's philosophical concepts depend at least in part on observations of ‘everyday human abilities’ that may be exercised outside the arts: p. 27. Jacobs contends that organisms have ontological priority for Aristotle, but that artefacts are epistemologically more basic. Artefacts provide the point of departure, when it comes to making sense of reality, since Aristotle takes them to be ‘more knowable to us’: p. 26.

³ The translations throughout are mine.

⁴ An item in a series is ‘for the sake of’ the end (or subsequent stage) of the series only if its reason or explanation for being in the series is that it contribute to bringing about or maintaining the end. This is a consequence of Aristotle's idea that the end ‘hypothetically necessitates’ the items or conditions that bring it about and support its existence. Their hypothetical necessitation by the end explains their presence in the series, because when an end hypothetically necessitates its conditions it necessitates them as means to the end (e.g.Phys. 2.9): cf. Cooper, J., ‘Hypothetical Necessity’, in Gotthelf, A. (ed.), Aristotle on Nature and Living Things (Pittsburgh, 1985), pp. 151–67Google Scholar . In the case of human action it would seem obvious to most of us that the explanations for actions are the projected ends of those actions. It is not obvious that in nature the processes that bring about natural results are explained by those results, even if processes of the same sort always yield results of the same sort and always prove to be beneficial to those results. This is because most of us today are teleologists about action but not about nature. Therefore, an analogy between the processes of action and of nature, which only acknowledged that in both sorts of processes there are series that yield results and that the series benefit or support the results, would not be impressive in establishing the goal-directedness of nature. David Charles perhaps thinks that Aristotle's argument from analogy is successful because he thinks that Aristotle has only a modest goal in mind in drawing the analogy between action and nature, ‘Teleological Causation in the Physics’, Aristotle's Physics, ed. Judson, L. (Oxford, 1991), pp. 101–28, esp. pp. 114–15Google Scholar . Charles finds two major forms of teleology in the Physics: agency and functionality. In agency the goal is an object of an agent's desire and choice, the value of which the agent is aware of, and also the means for achieving it. Nothing of this sort holds in the case of organisms, which engage in no psychological states in the development of, e.g., their bodily parts. Parts of organisms exist for the sake of the whole because they serve a ‘function’ of some sort in the life of the organism, and for Charles having a function within an organism seems to amount to contributing ‘to the survival or flourishing of that organism’ (p. 107), or being ‘beneficial to the organism's survival or flourishing’ (n. 5). Functional teleology seems to allow that if an item in a system is beneficial to the system as a whole, it is for the sake of that system. ‘Being for the sake of’ could mean, then, as little as ‘being beneficial for’. When Aristotle draws his analogy between action and nature, Charles thinks that what is common in the analogy to both action and nature can accommodate both sorts of teleology. Thus the analogy need accommodate nothing more than functional teleology. This neutral common character consists in merely sharing two features of teleology: (a) in sequences where there is a goal, the items of the sequences are for the sake of the goal; (b) the order of the items is for the sake of the goal (pp. 114–15). For Charles there is no deciding in the Physics between the two modes of teleology, agency and functionality, over the issue of which of them is the more basic mode of teleology for Aristotle, because the key to deciding between them lies in the meaning of ‘for the sake of’, which Aristotle leaves unanalysed in the Physics (p. 116). Yet, pace Charles, functional teleology is not a kind of teleology Aristotle would recognize, since for Aristotle ‘being for the sake of’, whatever it may ultimately amount to, could not simply be something as weak as ‘being beneficial for’. It is true that ‘being for the sake of’ entails ‘being beneficial for’. But, as is indicated by the notion of hypothetical necessitation, what it is for something to be for the sake of an end is that it exists only because it is the means to that end and only thereby is it also something beneficial for that end.

⁵ Sedley, David has recently argued for this view in ‘Is Aristotle's Teleology Anthropocentric?’, Phronesis 36 (1991), 179–96CrossRefGoogle Scholar . Sedley's thesis is that nature as a whole is basically centred on human beings and that in various ways natural processes are ultimately for the sake of humans. Even the weather is for the sake of humans, since among its goals is the promotion of the crops that are themselves for the sake of human consumption. Part of Sedley's defence of his thesis seems to depend upon an argument of something of the following sort: since art imitates nature, art and nature have common ends, and since art is for the sake of humans, so too must nature be for the sake of humans. Sedley's argument is not merely concerned with the relatively uncontroversial view that art imitates the ends of natural human activities, as would be the case in the art of weaponry, which imitates the end of the natural human activity of fist-fighting. In such cases of imitation, art and nature could easily be seen to aim in common at the good of humans. He also argues for the much more controversial view that art imitates the ends of non-human processes, such as the growth of certain plants, which then on his view must too be ultimately for the sake of humans. Agriculture in the cultivation of certain crops completes what nature is unable to complete, and in so doing agriculture takes up the end of nature: the nurture of certain plants, which ultimately is for the sake of human consumption. In his defence of this view of agriculture, Sedley appeals to fr. 11 of the Protrepticus, in which Aristotle defends his belief that art completes nature by maintaining that some seeds require the assistance of agriculture to come to maturity (p. 188). My speculations about the examples of art's imitation and completion of nature owe a great deal to Sedley's considerations.

⁶ Aristotle might regard nature as more basic than art in more than one way. Natural kinds are metaphysically more basic than artificial kinds, because natural kinds are real and artificial kinds are not, since artefacts are not substances (Met. 1043b21–2). Also, Aristotle considers it probable that the sciences and arts have been lost and regained often over eternity (Met. 1074b10–12). Accordingly, nature is temporally prior to art because natural kinds are eternal and remain constantly and uninterruptedly in existence (e.g.G.A. 731b24–732al), whereas the arts pass in and out of existence.

⁷ Because knowledge of the alphabet is exact and complete, the grammarian uses the letters of the alphabet in his writing without any deliberation (N.E. 1112a34). Aristotle seems, then, to regard the lack of deliberation in the arts to apply to cases in which the artistic activity has a fixed routine that is repeated in the same way every time it is engaged in.

⁸ David Charles does not think that Aristotle gives any ‘explanatory priority’ to action or to nature in the analogy between action and nature (p. 115). But this seems to be false, since Aristotle argues from the goal-directed character of action to that of nature on the basis of the analogy between them (199a8–12). He does not simply draw a symmetrical analogy between action and nature, but he draws the analogy for the clear purpose of arguing for the goal-directedness of nature. As far as I know, he never argues for the goal-directedness of action, nor does he ever argue from the goal-directedness of nature to that of action on the basis of the analogy between action and nature or on any other basis. He provides no such arguments because he thinks that the truth of the goal-directedness of action needs no argument.

⁹ David Charles finds no common teleological form shared by the teleologies of agency and functionality, or at least none of any interest (pp. 106–8 n. 25). Yet Aristotle must think that action and nature share enough features in common so that be may argue from the teleology of action to the teleology of nature on the basis of their analogous features. There would be no point to the argument from analogy unless Aristotle thinks that action and nature share enough in common so that he may argue for the teleological character of nature on the basis of the analogy. Aristotle could acknowledge a common teleological form for both action and nature without thinking that they were identical in every teleological feature, and he could even model the teleology of nature upon that of action without transferring every teleological feature of action to nature. He denies that there is deliberation in nature (199b26–8), and for that matter in art, but he certainly would not deny that action often involves deliberation in the pursuit of its goal. Deliberation is simply not relevant to the teleology that is common to nature and action. Rather, the features of teleological sequences, which are common to action and nature, and which Aristotle may borrow from action, are perhaps those I have suggested of (a) finitude, (b) termination in a goal, (c) which is good.

¹⁰ There is no reason to think that Aristotle intends by ⋯ νο⋯ν ἓχων anything other than ‘one who has reason’ or ‘sense’, where ‘reason’ or νο⋯ς is broadly construed as ‘reasonableness’, and this is the standard way the passage is taken by its English translators: e.g., ‘reasonable man’ (Ross and Barnes); ‘the man who has intelligence’ (Tredennick); ‘he who has an intellect’ (Apostle); ‘the man who has rational intelligence’ (Wheelwright). The phrase ⋯ νο⋯ν ἓχων is an ordinary way of expressing a person of reasonable nature as opposed to one of unreasonable nature. The νο⋯ν of the νο⋯ν ἓχων could not be the ‘mind’ that is the God of the universe, since in the context of this passage from A minor the actions under consideration are series that do not constitute an infinite series because they are limited by an end. God is not involved in any such serial action, since God's state of pure activity is its own end, which is infinite and unchanging (cf. Met. 1071b19–20, 1072b13–30, 1048b18ff.). Nor need Aristotle even intend by the νο⋯ν ἓχων the technical, psychological sense of νο⋯ν of De Anima 3.4–5, which itself might be thought to be modelled on the divine νο⋯ν The activity of the human νο⋯ν in the psychology is the cognition of intelligible forms, which is an activity that has itself for its end, but the actions under consideration in A minor for the performance of the νο⋯ν ἓχων are those that constitute series that are limited by goals.

¹¹ J. Barnes retains W. D. Ross's translation of this passage in his revision of the Oxford translation. There is an alternative tradition concerning 994b 15, in which ⋯ν τοιο⋯τοις stands in place of ⋯ν τοῖς οὖσιν. The latter is the text of Laurentianus 87.12 (A^b) and of Alexander (Al); whereas ⋯ν τοιο⋯τοις is the text of Parisinus 1853 (E) and of Vindobonensis phil. gr. C (J), and it can be inferred that ⋯ν τοιο⋯τοις is also the Greek of the text used by William of Moerbeke in his Latin translation. The two readings are each grammatical, although ⋯ν τοῖς οὖσιν is more plausible or attractive, because it allows for a more forceful and clearer statement on Aristotle's part. E and J are close in their text overall, and Ross holds that they thus belong to the same family and that A^b and Al are each based on different originals, of which the ‘traces of uncial corruption’ in A^b argue for its being older than the original of EJ. Thus, in cases of divergent but plausible readings, Ross follows the principle that the agreement of any two, EJ, A^b, and Al, is decisive, and since A^b and Al both read ⋯ν τοι οὖσιν in disagreement with EJ, Ross decides in favour of that reading for his edition: Aristotle's Metaphysics, vol. 1 (Oxford, 1924), pp. clv–clxviGoogle Scholar . In their editions of the Metaphysics, W. Jaeger and H. Tredennick agree with Ross in their reading of 994b 15, and Tredennick follows Ross in his translation of ⋯ν τοι οὖσιν.

¹² Perhaps another important indication of the primacy of human action in the development of Aristotle's teleology is the fact that he uses in his initial and definitive explanation of final causality in Physics 2.3 an example drawn from human action for his illustration: when someone's walking is for the sake of health (194b33). Perhaps, since this example has health as the end of the action, the action might be taken as the exercise of the medical art and thus as an example from the arts. Yet when we walk for exercise, and not for the sake of getting somewhere, we hardly think that we are engaged in the practice of medicine, nor need we be following the prescription of our physician.

¹³ I should like to thank an anonymous referee for CQ for many helpful comments on a draft version. I also wish to express my gratitude to Wayne State University for the award of the Career Development Chair for 1992–1993, which provided me with the leisure to compose this article.