1.3.1 On the Number and Motion of the Heavenly Bodies

Theophrastus raises six main questions about the heavenly bodies. First (5a5–21), how can the variety of heavenly motions be explained? Second (5a21–3), what is the number of heavenly spheres? Third (5a23–8), why do the heavenly bodies not pursue rest rather than motion? Fourth (5a28–b7), why are they not ensouled like animals? Fifth (5b7–10; 5b26–6a5), why do they not enjoy an activity better than circular motion? Sixth (6a5–13), is circular motion essential to them? The first two questions concern the number of heavenly bodies, questions three and five concern their motion, and the fourth and sixth questions concern their nature. I begin with the number and motion of heavenly bodies before turning to their nature in the next section. Against a widespread view, I will argue that none of the questions suggest that Theophrastus gives up the first unmoved mover or the Aristotelian framework in general.

As for the number of heavenly bodies Theophrastus notes that there are several heavenly bodies in circular motion (5a15–16), and that their motions are ‘in some sense contrary’ (5a16). If this is right, there cannot be only a single mover; otherwise, there would be only a single motion (5a17–18). Rather, there should be several movers, one for each motion (5a19). But then it is not clear how there is ‘consonance’ (to sumphōnon) among the heavenly bodies that were all meant to ‘follow the best desire’ (5a20-1).Footnote ²⁴ The second question, as to the cause (aitia) of the number of spheres is closely connected. Since the motion of a heavenly body depends on its number of spheres, to explain the variety of motions, more needs to be said both about the number of (unmoved) movers and about the number of spheres.

These questions again recall Aristotle’s Meta. Λ.8. For there, Aristotle is motivated to introduce 55 unmoved movers in addition to the first mover (1074a10–12) by observing the variety of planetary movements (1073a23–34), and he ties the number of movers to the number of spheres (1074a14–6). Moreover, when Theophrastus says that the required account of the cause of the number of spheres is not ‘the one of the astronomers’ (5a23), one thinks of the fact that Aristotle corrects the number of spheres given by the astronomers Eudoxus and Callippus by introducing retrograde spheres (1073b38–74a5). Thus, not only does the first philosopher, rather than the astronomer, state the cause of the number of spheres (Laks and Most Reference Laks and Most1993, 35), but it is also the first philosopher who comes closer to the right number of spheres.

If anywhere, a departure from Aristotle might be suspected in Theophrastus’ worry about the consonance of the heavenly bodies. As I read Theophrastus, when he says that heavenly bodies follow ‘the best desire’, he means that they follow the desire for the best, namely, the first unmoved mover. There is a tension, then, between this claim and the explanation of the motion of heavenly bodies by appeal to several unmoved movers, and hence, several objects of desire, distinct from the first unmoved mover. Or at any rate, it is far from clear how the many principles (archai, 5a19) can be integrated with the prime mover in such a way that all the heavenly bodies and their motions fit into a unified system.

Undoubtedly, Theophrastus identifies a lacuna in Aristotle’s account as we find it in Λ. For notoriously, Aristotle does not comment on the relation between the additional unmoved movers and the first unmoved mover. But the natural response to the lacuna is not to abandon the first unmoved mover. After all, an easier alternative would be to give up on the fifty-five additional unmoved movers.Footnote ²⁵ More bravely, but in keeping with the framework, one could forge a philosophical link between the prime mover and the other unmoved movers, as notably Michael Frede has suggested.Footnote ²⁶

What is more, Aristotle seems to recognize the need for further work. For he says that it is ‘plausible (eulogon) to assume’ that the number of unmoved movers is the same as the number of spheres and leaves ‘the necessity’ (to anagkaion) of the claim to ‘stronger’ people to establish (1074a15–17). Hence, the problem which Theophrastus highlights is in an area where Aristotle is himself tentative, and to which responses within the framework are available. Plausibly, then, Theophrastus’ first two questions constitute an internal criticism. They are not questions for which we find obvious answers in Aristotle (or Theophrastus), but they do not undermine the Aristotelian-Theophrastean framework.

A similar point can be made concerning the questions about the motion of heavenly bodies. Theophrastus asks first why they do not pursue rest rather than motion (5a23–5), and then why they do not share in an activity better than circular motion (5b7–10; 5b26–6a5). The former question is answered later (7b12–13) in line with Aristotle (see, e.g., Phys. VIII.8, 264a27–8): rest is a privation of motion and a worse state to aspire to. But this underscores the difficulty in explaining why the heavenly bodies cannot perform an activity that is even better than (circular) motion, given that their cause of motion is the best thing (5b7–8; 5b26–7). The only candidate mentioned is the activity of the soul, namely, thought (dianoia) (5b8–10), and Theophrastus himself seems to offer the answer: as far as the prime mover is concerned (6a1–2),Footnote ²⁷ it would impart a better activity, but the heavenly bodies are not ‘capable of receiving’ it (5b28–6a1).

Laks and Most (Reference Laks and Most1993, 41) claim that this explanation in terms of the deficiency of the receiver is Platonist rather than Aristotelian. But Aristotle himself notes the limits imposed on the activities of an entity by its own deficiencies, for instance, when he contrasts the extent to which humans can engage in thinking with the extent to which the prime mover can (Meta. Λ.7, 1072b14–30). In particular, the matter of a material entity imposes limitations on it, for instance, when he says that continuous motion would exhaust perishable things because of their matter (Meta. Θ.8, 1050b26–8). It would not be a sign, then, of Theophrastus’ veering away from Aristotle if he made a similar point about heavenly bodies: because they are material, they cannot engage in the best possible activity, namely, pure thinking, as the prime mover can, according to Aristotle’s Meta. Λ.9, 1074b33–5.

Of course, this move does not fully solve the problem. For one may ask why the heavenly bodies cannot engage in some psychological activity worse than thinking but still better than circular motion. As before, my claim is not that Theophrastus raises worries for which either he or Aristotle can give a straightforward answer. I suggest rather that his questions are of the sort that can be discussed, and potentially answered, within the Aristotelian framework. The fact that there are resources in Aristotle to spell out the beginnings of Theophrastus’ own answer is sufficient to support this suggestion.

The same goes for Theophrastus’ second reason why the heavenly bodies do not engage in the best activity, namely that this view would minimize or abolish all difference (diaphora) (6a3–5). This would threaten the priority of unmoved movers over the heavenly bodies since priority presumably presupposes some difference between the relata. But the elimination of differences would have an even more devastating general consequence, namely that things are no longer be knowable. Since Theophrastus addresses the topic at length in his treatment of differences (8a27), I defer a discussion of this last point to the next section. For now, let us instead look at the questions about the nature of the heavenly bodies. It is primarily here that interpreters have seen evidence of Theophrastus’ abandonment of the prime mover.

1.3.2 Does Theophrastus Abandon the Prime Mover?

Both the fourth and the sixth questions about heavenly bodies concern their nature. The fourth (5a28–b7) asks why the heavenly bodies, if they have a striving (ephesis) for an object, are not ensouled (5b2), and do not have life (zōē) (5b3). The sixth question (6a5–14) is whether ‘circular motion is in the essence (ousia)’ of the first heaven (6a7–8). In my view, although both questions probe the Aristotelian account of first principles in important respects, neither suggests that Theophrastus should have felt compelled to give up the prime mover.

Theophrastus has good reasons to think that the heavenly bodies must be ensouled: striving and desire, just like perception, presuppose a soul, as in the case of animals (5b3-7). This emerges dramatically for heavenly bodies since they strive for the best (5a28-b1), presumably because the desire for the best, namely, the prime mover, is least apt to be a function only of the body. According to Proclus in his commentary on Plato’s Timaeus, Theophrastus himself held that the heaven is ensouled (Diehl 122, 13–14). It has been argued, therefore, that Theophrastus doubts or even gives up the prime mover. For, if the heavenly bodies are ensouled, they must be in motion (5b3) independently of the unmoved mover (Ross and Fobes Reference Ross and Fobes1929, 46; Grumach Reference Grumach1932, 61–2; Reale Reference Reale1964, 49–50; Laks and Most Reference Laks and Most1993, 38). Therefore, the prime mover is not needed as a cause of motion and can be dispensed with (Ross and Fobes Reference Ross and Fobes1929, xxv; Grumach Reference Grumach1932, 63; Reale Reference Reale1964, 155–6; van Raalte Reference van Raalte1988, 194; Reference van Raalte1993, 477; Sharples Reference Sharples1998, 273–4; Sorabji Reference Sorabji1998, 204–6).

Several responses have been offered (see Lefebvre Reference Lefebvre2015, 42–4). On one view, Theophrastus raises aporiai without affirming a solution, and hence without committing himself to rejecting the prime mover (Ellis Reference Ellis1988; cf. Gutas Reference Gutas2010, 38–43). On another view, he attacks not Aristotle but rather a Platonizing version of Aristotle (Berti Reference Berti and Migliori2002, 340–1). But if the implications of the problem raised by Theophrastus are grave enough to threaten the role of the prime mover, it is hard to believe that it is not a serious criticism of Aristotle. Hence, a third strategy due to Lefebvre (Reference Lefebvre2015) is more promising. For he aims to undercut the claim that Theophrastus’ position implies the redundancy of the prime mover.

In particular, Lefebvre (Reference Lefebvre2015, 55–8) shows that Theophrastus does not speak of heavenly bodies as self-moving. But, if the heaven is merely moving, not self-moving, it is not excluded that the heaven is moved by the prime mover. Although I am sympathetic to this proposal, it does not neutralize the threat arising from Theophrastus’ fourth question, namely that the heavenly bodies are in motion in virtue of being ensouled. For if being ensouled is sufficient for being in motion, then there is no need for another cause of motion, regardless of further issues concerning self-motion.

However, with the help of Aristotle, we can see that Theophrastus’ claim that the heaven is ensouled need not imply that there is no need for the prime mover. As a preliminary, we may note that, in De caelo, Aristotle, too, says that that the heaven is ensouled (see, e.g., DC II.2, 285a29–30). Hence, even if the prime mover is incompatible with ensouled heavenly bodies, at most Theophrastus would have chosen one of the options developed (also) by Aristotle, not given up on Aristotle’s larger research programme. But more importantly, there is good reason to think that the prime mover is compatible with ensouled heavenly bodies. For in his account of desire (orexis) in DA III.10, Aristotle argues that the object of desire (to orekton) is an unmoved mover, whereas the capacity of desire (to orektikon) is a moved mover. Both are required to explain animal motion (433b13–18), but the object of desire is the first mover because it is unmoved (433bb11–12). Similarly, one might think that the unmoved mover and the capacity of desire in the heavenly bodies jointly explain their motion but that the unmoved mover is first in the explanatory order because it is unmoved.

This application to the cosmological case is not trivial, and Theophrastus’ fourth question hints at a fair criticism of Aristotle. For in Λ.7, he does not state the implications of the claim that the unmoved mover moves as an object of desire and thought (1072a26) for the nature of heavenly bodies. Moreover, he does not explain exactly how his views on desire could make the implication that the heavenly bodies are ensouled compatible with the view that the prime mover is their cause of motion.Footnote ²⁸ At the same time, the fact that we can find, in Aristotle, the ingredients for a possible response to this problem should make us doubt that Theophrastus means to present an insuperable difficulty that forces him to abandon the unmoved mover.

In response, the opponent might grant that the claims about ensouled heavenly bodies in Theophrastus’ Metaphysics do not commit him to abandoning the prime mover but insist that Proclus’ report shows that he did abandon it (Sharples Reference Sharples1998, 274). For Proclus says not only that, for Theophrastus, the heaven is ensouled but also that ‘it is reasonable (eikotōs) for Theophrastus to say that the soul is a principle of movement, without positing anything before it, and to believe that one should not search for a principle of what is a principle’.Footnote ²⁹ This suggests that there is no principle of motion prior to the soul, nor therefore to the soul of the heaven, which seems to rule out that the prime mover is needed to move the heavenly bodies.

However, as Lefebvre (Reference Lefebvre2015, 49–55) emphasizes, we should read these claims in the context of Proclus’ treatment. This context is a discussion of the ‘generation of soul’ (psuchogonia) (Diehl 120, 18–19): against Plato, Theophrastus says that the soul is not generated, which is compatible with the soul being moved in place by the prime mover (Lefebvre Reference Lefebvre2015, 54). One could respond that the claim, as Proclus formulates it, is not concerned only with generation but broadly with ‘movement’ (kinēsis) (Diehl 122, 12) including locomotion. But even if this is right, the statement by Proclus need not imply that Theophrastus is arguing that Aristotle’s prime mover should be given up. For the cited claim stems from Theophrastus’ De caelo (Diehl 122, 17), and Aristotle does not seem to rely on the unmoved mover in his De caelo either.Footnote ³⁰ It may be due to the explanatory context that Theophrastus does not mention the unmoved mover in his De caelo.

Furthermore, the broader principle that, according to Proclus, Theophrastus uses against Plato is that, in order to avoid an infinite regress, we should not seek the causes of first things (Diehl 120, 10–13). But this principle remains intact, even if, in addition to the soul of heavenly bodies, one posits an unmoved mover. For the unmoved mover, as Theophrastus himself notes in his Metaphysics (5a4–5), stops the regress. Overall, therefore, the report from Proclus is far from decisive. If we want to understand Theophrastus’ position on the prime mover, we must do so from within the treatise that is dedicated to first philosophy, namely, his Metaphysics, not from an allusion to his De caelo. And this treatise, together with material from Aristotle, does not suggest sufficient grounds for Theophrastus to show that Aristotle’s account should not have included the prime mover.

Similar considerations hold for Theophrastus’ sixth question, whether circular motion is essential to the first heaven (6a7–8). Initially, Theophrastus states two options. The first option is that circular motion is essential such that, if the heaven stopped moving, it would be destroyed (6a7–9). The second option is that circular motion belongs accidentally to the first heaven (6a9–10) since it belongs to it ‘by some striving and desire’. But he notes that, even on the second option, if the desire is ‘connate’ (sumphuton) to the first heaven, its circular motion will not be accidental. Theophrastus cautiously approves of the view that the desire is essential to the first heaven (‘nothing prevents that there are some such entities’, 6a11–12), before restating the initial problem whether removing the motion (not the desire) would destroy the heaven (6a12–14). Later, he seems to affirm that the heaven would be destroyed because its circular motion is ‘according to its substance’ (kata tēn ousian) (10a14).

Some scholars have claimed that, if Theophrastus affirms that circular motion is essential to the heavenly bodies, he must think that the heavenly bodies are self-moved, without any role left to play for the prime mover (see, e.g., Ross and Fobes Reference Ross and Fobes1929, xxv; van Raalte Reference van Raalte1988). Lefebvre (Reference Lefebvre2015, 55–8) has already shown that Theophrastus does not say that the heavenly bodies are essentially self-moved, but merely that they are essentially in motion. Hence, there is no easy inference from the essence of heavenly bodies to the redundancy of the prime mover. Indeed, in my view, the argument for the redundancy of the prime mover is flawed.

For the essence of the heavenly bodies may involve the prime mover. Hence, they may be essentially moved and there is a need for a prime mover. Theophrastus himself gives us the materials for this view. If the desire is essential to the heavenly bodies, this implies that they are not moved accidentally. But the desire in question cannot just be any desire but must be a desire for the best, namely, the prime mover. Hence, if the desire for the prime mover is in the essence of the heavenly bodies, their essence includes the prime mover. Thus, in the text of Theophrastus, there is a readymade view according to which the heavenly bodies are essentially moved and there also must still be a prime mover.

Nonetheless, the sixth question is a challenge. For Aristotle does not say explicitly whether the motion of the heavenly bodies is essential to them. He comes closest in Meta. Θ.8, where he argues that they are not moved ‘according to potentiality’ (kata dunamin) (1050b21) and contrasts them with perishable substances whose substance is potentiality (dunamis) rather than actuality (energeia) (1050b27–8). But these suggestive remarks do not amount to a full account. Moreover, he does not explain the consequences of this view for the relation between heavenly bodies and the prime mover. Theophrastus’ call for clarification, then, is well motivated. But this does not imply that he is putting forward a fatal criticism.

In the last two sections, I have argued that Theophrastus’ questions about the heavenly bodies constitute internal criticisms of Aristotle: criticisms for which there is no clear-cut answer in Aristotle but that do not undermine the general framework either. I will now extend this reading to the questions about the entities at the centre of the universe.

1.3.3 The Entities at the Centre of the Universe

The issue of the reach of the prime mover becomes even more pressing when one thinks about its relationship with entities further removed from it than the heavenly bodies. For, if the reach of the prime mover does not extend all the way to the centre of the universe, the connection between the prior and posterior entities in the world in its entirety (to pan) seems broken. Theophrastus brings up this topic in discussing the entities at the centre of the universe (ta peri to meson). These are the four elements (Ross and Fobes Reference Ross and Fobes1929, 47; Laks and Most Reference Laks and Most1993, 39) or perhaps all sublunary entities subject to generation (van Raalte Reference van Raalte1993, 222; Henrich Reference Henrich2000, 273; cf. Aristotle’s GC II.9, 335a25). At any rate, we will see that Theophrastus’ questions can be applied to both the elements and perishable substances.

Theophrastus’ main question is why, unlike heavenly bodies, entities at the centre of the universe are not ‘capable of striving’ (ephetika) for the prime mover, although they are subject to motion (5b10–13). He offers two possible answers: either the sublunary entities are ‘incapable’ (adunata) (5b13) of striving for the prime mover or the prime mover does not ‘reach’ (diiknoumenou) them (5b13–14). Since the second option implies a weakness of the prime mover, he rejects it as ‘absurd’ (atopon) (5b14–15). Rather, sublunary entities cannot strive for the prime mover because they are ‘something non-receptive (adekton) and disconnected (asundeton)’ (5b18).Footnote ³¹ This answer coheres with his explanation why the heavenly bodies cannot share in an activity better than circular motion, namely that they are not ‘capable of receiving it’ (5b28–6a1). As discussed, Theophrastus seems to agree with Aristotle that the matter of an entity imposes limitations on what it can do. In the case of sublunary entities, they are not even capable of striving for the unmoved mover.

It is sometimes suggested that Theophrastus’ question misses the mark because, according to Aristotle, the elements imitate the circular motion of the heavens by their cycles of transformations (GC II.10, 337a3–4; Meta. Θ.8, 1050b28–30) (Ross and Fobes Reference Ross and Fobes1929, 47; Tricot Reference Tricot1948, 10, n. 2; Laks and Most Reference Laks and Most1993, 40). However, in these passages the elements are said to imitate the motion of the heavenly bodies, not (the activity of) the prime mover. Indeed, the cycle of transformation of elements as well as the cycle of generation of organisms are explained by the motion of the first heaven and the sun, not directly by the prime mover (Meta. Λ.6, 1072a9–18; cf. GC II.10). Hence, Theophrastus’ question why they are not moved directly by striving for the prime mover is pertinent.Footnote ³²

Building on these worries, Theophrastus raises a follow-up question, namely, whether the entities at the centre of the universe are parts of it, and if so, how (5b19–21). If they are not capable of striving for the prime mover, they are ‘like pushed away from the most noble things (tōn entimotatōn) not only in place but also in accordance with activity (energeia), if indeed circular motion is such’ (5b21–3). Since their activity is characterized as circular motion, the ‘most noble things’ seem to be the heavenly bodies. The claim, then, is that the sublunary entities are removed from heavenly bodies not only in space but also because they engage in a less noble activity. He adds that the locomotion and changes of sublunary entities into each other are caused accidentally by the motion of the heavenly bodies (5b24–5).

One might wonder what the criticism is. Neither the removal of sublunary entities from heavenly bodies nor the fact that they are moved accidentally by them implies that the sublunary entities are not parts of the universe. Indeed, the fact that they are moved by the heavenly bodies presumably implies that they are parts of the same universe to which the heavenly bodies belong. What is more, Theophrastus’ remarks indicate how they are parts, namely, at a level below the heavenly bodies in the sense that they are not directly moved by the prime mover and engage in a less noble activity.

To get a better grip on Theophrastus’ criticism, it is helpful to consider a section at the end of the first critical wave (6a14–7a22). In this passage, Theophrastus imposes a demand of continuity on explanations from first principles: ‘one might consider it good … to explain (apodidonai) the things that follow in sequence (ta ephexēs) straightaway (euthu) and not to proceed up to a point and then stop’ (6a16–18). The positive model is the Pythagorean Eurytus who assigned numbers also to human, horse, and other things, and thus extended his explanation of phenomena to the sensible world (6a19–22).Footnote ³³ By contrast, Theophrastus attacks certain Platonists who ‘stop (katapauontai) after they have proceeded to a certain point’ (6a23–4), that is, extend their explanation from the first principles only to a certain point in the chain of explananda.

For example, Theophrastus criticizes those who posit the one and the indefinite dyad but then generate only numbers and geometrical entities and fail to explain phenomena of the sensible world or explain them only in a cursory fashion by deriving, for instance, place, void, and infinity from the dyad and soul and other phenomena from numbers and the one (6a23–b3).Footnote ³⁴ The main target mentioned by name is Speusippus and his followers (6b5–6), whereas Xenocrates (6b6–9), another Platonist called Hestiaeus (6b9–11),Footnote ³⁵ and Plato (6b11–15) are treated more lightly since they did at least something to extend the explanation from the first principles to the sensible world.Footnote ³⁶

An extreme version of the view under attack, which Theophrastus ascribes to ‘some people’ without naming them, discusses only the principles, not the things they are meant to explain, on the grounds that the truth is found only among principles (6b15–17). According to Theophrastus, this conception of the science of first things contrasts markedly with that of other ‘sciences’ (methodois), where the things that follow from the principles are ‘stronger’ and ‘as if more complete’ (6b17–20). His thought may be that, in other sciences, it is by explaining the explananda that the science comes, so to speak, alive, not by dwelling on the explanantia (cf. Laks and Most Reference Laks and Most1993, 47–8). Theophrastus grants that there is some ‘plausibility’ (eulogōs, 6b21) to the criticized view, given that the science at stake concerns first things and thus the principles themselves (6b20–2). But this concession need not overturn the demand of continuity: even if there is something to be said for focusing one’s attention on the (first) principles, still one has to show how they operate as principles and to do so fully, all the way down to explananda of the sensible world.Footnote ³⁷

Theophrastus’ focus on the Platonists might induce one to think that the passage does not have anything to do with Aristotle and even that the treatise as a whole is not as dedicated to a critique of Aristotle as I have made out. But a different reading is available. Theophrastus sets out certain desiderata against the background of Academic theories, but he introduces them to engage critically with the Aristotelian account (cf. Krämer Reference Krämer1973). In the case of the demand of continuity, this interpretation is particularly attractive since it is closely related to the issue of the reach of the prime mover with which Theophrastus has been pre-occupied: in Aristotelian terms, the demand of continuity asks one to show how the explanatory reach of the prime mover extends all the way to the sensible world.

Still, one should raise the question why he brings up the demand of continuity at this juncture of the treatise. I suggest that Theophrastus motivates the demand of continuity with the help of the Academic background to press his worry about Aristotle’s account of the entities at the centre of the universe. For it is sublunary entities at the centre of the universe which Eurytus, unlike some Platonists, accounted for with his principles. Theophrastus, then, challenges Aristotle to meet the demand of continuity in a similar manner and to show in line with the basic tenets of his framework that the first intelligible entity is a principle of all perceptible entities, not only some of them. This is a pressing demand especially because Aristotle himself insists against Speusippus that the world is not episodic.

Certainly, Aristotle has something to say in response. For arguably, in Meta. Λ.6, he appeals to the prime mover to explain not only the motion of the heavenly bodies but also indirectly the cycles of generation of perishable substances and of the transformation of elements (1072a9–18).Footnote ³⁸ Hence, a weak version of the demand of continuity is met: in an indirect manner, the reach of the first principle is extended all the way to the sensible world, including sublunary substances. But if one is tempted by a stronger version of the demand where the first principle must directly explain the explananda of the sensible world, one will not be satisfied. At least, then, more would have to be said why the strong demand of continuity need not be heeded. Theophrastus’ challenge also shows that, in principle, a stronger, direct link between sublunary entities and the prime mover could be forged. Neither Aristotle nor Theophrastus undertakes this, but the option is not explicitly rejected. Theophrastus, then, puts his finger on another point where the framework may need further development.

Where does all of this leave us? I have argued that Theophrastus’ first critical wave raises issues for Aristotle that require further reflection. Moreover, although I have indicated some of Aristotle’s resources for responding to Theophrastus’ questions, we have seen that he does not explicitly and fully answer them. But I have also argued that the difficulties are not grave enough to suggest that Theophrastus abandoned basic Aristotelian tenets, such as the prime mover. Rather, he raises internal criticisms that call for treatment within the Aristotelian-Theophrastean framework. I will argue that Theophrastus’ strategy is the same in the second and third critical waves that no longer rest only on Aristotle’s theology but also on his account of substance in the middle books of the Metaphysics and his natural-scientific methodology.

2.2.1 Against Materialism and Formalism

In introducing the shapeless and shaped principles, Theophrastus both characterizes them and associates them with groups of thinkers. He says that the shapeless principles are ‘like having a capacity’ (hoion dunamikas) (6b25). This is his first allusion to the treatment of matter as ‘in potentiality’ (en dunamei) to which he will return later (8a10–11). Moreover, he offers an example of thinkers who posit shapeless principles: ‘those who posit fire and earth [as principles]’ (8a25–6). Although their exact identity is unclear,Footnote ⁵³ these must be the natural philosophers who, according to Aristotle, define things with reference only to their matter, in contrast with (Platonist) dialecticians.Footnote ⁵⁴ Similarly, Theophrastus says that they posit only ‘the material (hulikas) principles’ (7a7).

One key feature of these material principles, which will come to the fore in the argument against materialism, is that they are ‘indeterminate’ (aoristous) (7a19). By contrast, the main feature ascribed to shaped principles is determinacy. For Theophrastus claims, with reference to Plato’s Timaeus,Footnote ⁵⁵ that shaped principles must be most of all ‘determined’ (hōristhai) (6b27; 6b28). Although he does not say anything explicit about determinacy, his claim that ‘order and being determined is most peculiar to the most noble things’ (tois gar timiōtatois oikeiotaton hē taxis kai to hōristhai) (6b27–8) sheds light on it. The (shaped) principles are determinate in that they have order (taxis). In the first instance, then, they themselves seem ordered, without any further cause of their order (if they are first principles). But in addition, the determinate principles also impart order to other things.

The latter conclusion is borne out by Theophrastus’ examples of sciences, such as geometry, music, and mathematics (7a1–4), and crafts (7a4–6). In the case of sciences, ‘what comes after the principles, too, follows’ (7a3–4). I take this to be a claim about the entities in the domain of the science: everything in the domain of a science such as mathematics derives from its first principles. Similarly, in the case of crafts, ‘the tools and everything else (ta alla) are in accordance with the principles (kata tas archas)’ (7a5–6). For instance, in the case of medicine, the actions to be performed and the drugs to be used derive from the principle of medicine, that is, health (Ross and Fobes Reference Ross and Fobes1929, 60–1). But these are presumably not just claims about implication. For the principles also dictate the order of theorems in a science or the order of the craftsman’s actions. Thus, the (shaped) principles are themselves ordered and determinate but they also order and make determinate all other things in their domain.

The association of shaped principles with the Timaeus suggests that, when Theophrastus goes on to speak of thinkers who posit ‘all principles as shaped (emmorphous)’ (7a6–7), he has in mind Platonists. Hence, by contrasting formalists with materialists (7a6–7), he follows Aristotle (see, e.g., DA I.1, 403a24–b9) in that he sets up a contest between Platonists who define things with reference to form alone and natural philosophers who define things with reference only to matter. This sheds light on the third group, namely, those who posit ‘both the shaped principles and the principles belonging to matter because completion is in both’ (7a7–9). According to Laks and Most (Reference Laks and Most1993, 50–1), this is a general ‘dualist’ view open to Platonists. But the association of Plato with shaped principles suggests that the third option excludes Platonists and captures only the Aristotelian view (cf. Ross and Fobes Reference Ross and Fobes1929, 61).

One might object that Theophrastus supports the view that ‘completion (to teleon) is in both’ (7a8–9) with a reference to ‘opposites’ (antikeimenōn), which has a Pythagorean or Platonist ring (Laks and Most Reference Laks and Most1993, 50): ‘for the entire substance is as if composed from opposites’ (7a9–10). But Theophrastus says only that substance is ‘as if’ (hoion) composed from opposites. This does not imply that they are strictly opposites in the sense of contraries (see Jaulin Reference Jaulin2015, 145). It remains open to Aristotelians to give their own account of the opposite shaped and shapeless principles, where they are not contraries but rather construed as form and matter, as we will soon see Theophrastus do (Section 2.3.2).

Hence, I submit that the third view that accepts both formal and material principles is, in effect, Aristotelian hylomorphism, even though Theophrastus has not yet fully spelled out this position. If this is right, the ensuing arguments against materialism and formalism (7a10–b8) lend indirect support to Aristotelian hylomorphism. For two of the three options are ruled out, with hylomorphism the only one left standing.

Against materialism (7a10–19),Footnote ⁵⁶ Theophrastus notes that its advocates are in an uneasy spot because they accept that the universe and all its parts are ‘in order and reason’ (en taxei kai logōi) (7a12) down to the smallest thing, be it inanimate or ensouled (7a16–17), and yet they claim that the principles are themselves ‘indeterminate’ (aoristous) (7a19). More precisely, the ‘natures’ (phuseis) of all things are ‘determinate’ (hōrismenai), but their principles are indeterminate (7a17–19). Presumably, Theophrastus’ thought is that if the world is ordered, its order must derive from some principle. But it is determinate principles that cause order. Hence, it cannot be the case that all the principles of the world are material.

By contrast, Theophrastus complains about formalism that it posits too much order. He notes at the end of his argument (7a19–b8) that the phenomena ‘demand a delimitation (aphorismon) how far what is ordered extends (mechri posou to tetagmenon), and why more is impossible, or else the transition is for the worse’ (7b6–8).Footnote ⁵⁷ And at the outset, he says: ‘on the other hand, it is difficult to confer reasons (logous) on all things, gathering [them] together towards the for-the-sake-of-which (to heneka tou) in all things, in animals, plants, and the very bubble’ (7a19–22).Footnote ⁵⁸ Whereas materialists falter because they end up with disordered principles, formalists exaggerate the extent to which the world is ordered.

Notably, Theophrastus states formalism in terms of final causation (‘the for-the-sake-of-which’). Plausibly, the reason for this choice is his insistence on the notion of ‘order’ (taxis). As we saw, the key characteristic of formalist principles is that they are determinate, which in turn implies that they cause order. Moreover, in his arguments against both materialists and formalists, Theophrastus assumes that the world exhibits some order. Hence, insofar as the order of the whole world is concerned, the most obvious formalist proposal is that its sole principle is an all-pervasive final cause.

Who is the target of Theophrastus’ anti-formalist argument? Given the earlier mention of the Timaeus, Plato is a natural answer. That said, the ‘greatest example’ (megiston paradeigma) of a final-causal account that Theophrastus cites is an account of the sun causing the succession of the seasons and thus the generations of animals and plants (7b2–5). This account looks like Aristotle’s in GC II.10 and Meta. Λ.6, 1072a9–18. Hence, one might conclude that one of the primary targets of the argument is Aristotle.

However, although the ‘greatest example’ is an example of a final-causal account, it is not an example of formalism. Theophrastus introduces this account with an ‘except if’ (plēn ei) on the back of his criticisms of the radical view that assigns a final cause even to the bubble (7a22–3). The more nuanced final-causal account has it that all the various shapes (morphas) in the region of earth and air ‘come about (sumbainei) by the order and change (metabolēi) of other things’ (6a23–7b1). The Aristotelian-sounding account of the seasons is the greatest example of this nuanced final-causal view, not of the formalist proposal that attributes a final cause even to the bubble.Footnote ⁵⁹

Moreover, recall that the formalist view is that all principles are formal. Thus, on the final-causal construal of formalism, the Aristotelian position would have to be that the only principle is the final cause. This is hardly Aristotle’s view since even in Λ, he argues that matter is also a principle of natural things (see, e.g., Λ.2, 1069b32–4). Hence, I prefer another reading. Theophrastus targets formalist views quite generally, including Plato’s but also a misguided formalist construal of Aristotle that explains everything in terms of final causes and moreover only in terms of final causes.

The implication of Theophrastus’ arguments is that neither the world as a whole nor the entities in the world, such as animals and plants, can be explained exclusively in terms of material causes or exclusively in terms of formal-final causes – and hence that they must be explained in terms of both. In the third section of the second wave, Theophrastus spells out this option in explicitly hylomorphic terms. His investigation of first principles thus turns to the material and formal causes or principles of perceptible things.

2.2.2 Matter and Form

The argument against materialism and formalism is followed by the passage on rest (see Section 2.1) which is in turn followed by a passage on matter and form (8a8–20). These transitions are not smooth, since the discussion of matter and form revisits the hylomorphic option stated before the arguments against materialism and formalism (7a7–9). Thus, the passage on rest seems to intrude into a continuous line of thought. But from another point of view, the sequence makes good sense. For the first part of our section (6b23–7b8) is preparatory to Theophrastus’ positive account: he sketches three options concerning the principles and argues against two of them. By contrast, the passage on rest and the discussion of matter and form state positive conclusions: the first external moving principle of the world is activity (energeia), and the immanent principles of (natural) things are matter and form.

The passage on matter and form, then, opens with a call to give an ‘account’ (logos) ‘how there is a partition (merismos) of things that are (tōn ontōn) into matter and shape’ (8a9–10). This could be taken as an enquiry into the partition of the domain of all entities into material and formal ones, or as an enquiry into the internal partition of entities into their matter and form (Laks and Most Reference Laks and Most1993, 57). Since Theophrastus will rely on craft cases where some matter is shaped or enformed (8a12–14), the second reading is more plausible: he is concerned with a hylomorphic analysis of some entity. He offers two options for the partition. The first option (8a10–11) is that form is ‘something that is’ (on), whereas matter is not but ‘is something that is in potentiality (en dunamei on) and that is led to actuality (eis energeian)’. The second option (8a11–12) is that matter is ‘something that is’ (on) but is indeterminate (aoriston).

These options are not genuinely distinct (cf. Reale Reference Reale1964, 62–5; Laks and Most Reference Laks and Most1993, 57–8; Sharples Reference Sharples2015, 77). In formulating the first option, he qualifies the claim that matter is not something that is and says that it is something in potentiality, which is compatible with the claim on the second option that matter is indeterminate.Footnote ⁶⁰ Moreover, the craft argument given seemingly for the second option relies on the notion of potentiality from the first option. For he says that the ‘generation or substance’ of (the objects of) craft is due to the ‘being shaped’ of matter ‘in accordance with the accounts (logous)’ (8a12–14) and specifies that matter is ‘what is not a this (tode) nor of some quality nor of some quantity but what, as indefinite in the forms, has some potentiality’ (8a17–19). It is the combined account of matter in terms of indeterminacy and potentiality that shows that matter exists. For although, as matter is shaped, there is a ‘transition (metabasis) to the better’ (8a14–15), it must exist since without matter as something indeterminate and potential, there would be no generation (8a15–19).

This account of matter and form is a brief but orthodox statement of Aristotle’s view as we find it notably in Meta. ZH.Footnote ⁶¹ As Burnyeat (Reference Burnyeat2001, 68 n. 141) notes, the characterization of matter as indeterminate and in potentiality is reminiscent of the claim that matter is potentially a ‘this something’ (tode ti) in Meta. H.1, 1042a27–8 (cf. Z.3, 1029a20–1). In general, the treatment of matter as potential and form as actual recalls Meta. H.Footnote ⁶² Similarly, the argument for the existence of matter on the basis of the possibility of generation is close to Aristotle’s argument for the status of matter as substance in H.1 (1042a32–b8). And the craft argument resembles the account of the generation of the composite in Z.7–9 (see, e.g., Z.8, 1033b5–19). Thus, in line with Aristotle, Theophrastus seems to defend the third option from earlier: there are both material and formal principles since generated things have matter and form.Footnote ⁶³

2.2.3 Bad and Good

Shortly, we will see that Theophrastus supplements his hylomorphism with essentialism. But in-between, there is an apparently disconnected section on ‘opposites’ (enantia) (8a21–b9). Theophrastus’ asks, ‘why then the nature and the whole substance of the entire world (tou pantos) consists of opposites, and why the worse almost balances out the better, or rather there is much more of it’ (8a22–5).Footnote ⁶⁴ Scholars have taken Theophrastus to be concerned with a Pythagorean (Tricot Reference Tricot1948, 26, n. 1; Ross Reference Ross and Fobes1929, 65) or Platonist view (Laks and Most Reference Laks and Most1993, 59), which some think he ends up embracing (Laks and Most Reference Laks and Most1993, xxv). But the passage is better read as an appendix to his defence of hylomorphism and does not signal a departure from Aristotle.

Theophrastus asks both why in general the world consists of opposites and why, specifically, there is both good and bad in the world. But he then goes on to treat this double question in line with a ‘nearby’ (eggus) question (7a27), namely, ‘why not all things are good, and not all things are similar (homoia)’ (8b1). Theophrastus does not hold that all things are similar since he insists that being is varied, as we will see in Section 2.3. In this sense, then, he accepts that the world consists of opposites, and presumably also that there is good and bad. But this does not show that he abandons Aristotle and veers towards Platonism, as Laks and Most (Reference Laks and Most1993, xxv) suggest. For, of course, Aristotle too does not think that all things are similar, nor that there is only good in the world (see, e.g., Meta. Θ.9, discussed next).

If anything, one might see a tension with Aristotle’s arguments against opposite principles in Λ.10, where he argues, among other things, that it is false to ‘make all things out of opposites’ (1075a28–9), and that there is no opposite to the first thing (1075b21–2). But this tension is merely apparent. For first, Aristotle attacks the view that all things are only from opposites on the grounds that there must also be a third thing underlying them, namely, matter (1075a31–4). This is a position Theophrastus accepts (see Section 2.2.2). Second, the claim that there are opposites in the world is compatible with thinking that the first entity, namely, the prime mover, does not have any opposite. Theophrastus, as well as Aristotle, can accept both.

By contrast, there are two claims which Theophrastus characterizes as ‘paradoxical’ at the end of the passage (8b4–8), and constitute ‘some transcendent wisdom’ (huperbatos tis sophia) (8b8)–9). The first, which ‘appears more paradoxical’ than what has been said so far (8b4–5) is that ‘what is (to on) cannot be without opposites’ (8b5). The second claim, advocated by those ‘who make use of the paradox even more’ (8b6), is that there is ‘what is not and has not been and will not be’ (8b6–8).Footnote ⁶⁵ Contra Laks and Most (Reference Laks and Most1993, 61), the fact that Theophrastus designates both views as ‘paradoxical’ suggests that not only the second but also the first is part of the transcendent wisdom to be rejected. And there is indeed a version of the first claim that Theophrastus should reject, following Aristotle, namely that anything must have an opposite; for the prime mover does not.

But there is still the question why this passage is located where it is. Here, an analogy with Aristotle’s Meta. Θ.9, 1051a4–21, is helpful. In this passage, which follows the conclusion that actuality is prior to potentiality in Θ.8, Aristotle argues that, in the case of a ‘good’ (spoudaia) capacity (1051a4–5), the activity (energeia) is better than the capacity (1051a15), whereas in the case of bad capacities, the activity is worse (1051a15–17). Similarly, I suggest that Theophrastus’ discussion of opposites is an appendix to the treatment of matter and form in terms of potentiality and actuality. For although Theophrastus does not explicitly say so, hylomorphism can explain why there is good and bad: if there is matter and form (and hence, potentiality and actuality), there is also bad and good.

2.2.4 Essentialism

On the back of his defence of hylomorphism, Theophrastus states what sounds like a familiar Aristotelian slogan: ‘that being (to on) is in many ways is clear’ (8b10). But unlike Aristotle, Theophrastus will not draw intensional distinctions between different senses of being,Footnote ⁶⁶ but rather extensional ones between different kinds of entity (cf. Laks and Most Reference Laks and Most1993, 61–2; Henrich Reference Henrich2000, 125; van Raalte Reference van Raalte2015, 89). Moreover, Theophrastus swiftly moves to epistemological concerns. Hence, scholars usually treat the claim about being together with the discussion of knowledge that begins with the parallel claim that ‘knowing (episthastai) is in several ways’ (9a10). But I will argue that the passage following the claim about being is not primarily concerned with knowledge. Instead, Theophrastus completes his account of the principles of the natural world by supplementing hylomorphism with essentialism.

Theophrastus’ first step (8b10–20) is to defend the claim that being is in many ways. As I understand the text, the key support claim is that ‘knowing (to episthastai), therefore, is not without some difference (diaphoras)’ (8b16–17). The argument is simple. Without differences between entities, there is no knowledge. But there is knowledge (implicit). Hence, there are differences between kinds of entities. Therefore, there are several kinds of entities.

The ‘therefore’ (ara) in the supporting claim makes clear that it is itself the conclusion of an argument (8b10–16). This argument involves perception. First, Theophrastus says that ‘perception both beholds (theōrei) the differences and seeks the causes’ (8b10–12), before proposing more cautiously that perception ‘suggests them to thought (dianoiai)’, in some cases, ‘seeking them simply’, and in others, producing a ‘puzzle’ (aporia) to be resolved (8b12–16).Footnote ⁶⁷ Those differences, as Theophrastus clarifies (8b17–20), are differences between universals, namely, between the sub-kinds of a kind, whether higher genera or lower species.Footnote ⁶⁸

Presumably this technical understanding of differences as differentiae that distinguish species of a genus from each other motivates Theophrastus to back away from the strong claim that we perceive the differences. For it seems implausible that we simply perceive differentiae of kinds. All he needs for his present argument is the weaker claim that perception gives us the input on the basis of which we can grasp the differentiae of kinds by thought. In effect, then, the supporting claim rests on a premise about knowledge acquisition: we come to know things by grasping their differences on the basis of perception.

It is not clear how strong the premise about knowledge acquisition is. In particular, it is not clear whether it is a claim about all knowledge. Given the role of perception, one might think that the claim is restricted to knowledge of perceptible things. But even the restricted claim seems more assailable than the conclusion it supports. Plausibly, if there is knowledge of things, (kinds of) things must be differentiable, regardless of the role perception plays in the acquisition of knowledge.

But what matters most for my reading is that Theophrastus uses a claim about knowledge to establish something about the world, namely that there are differences. In my view, the same holds for the second major claim about knowledge in the passage, namely that ‘all knowledge is of the peculiar things (tōn idiōn)’ (8b20–1).Footnote ⁶⁹ The main reason given is that ‘the substance and essence is peculiar for each thing (kath’ hekaston idion)’ (8b21–2).Footnote ⁷⁰ A tacit premise, then, is that knowledge is of essences which, together with the claim that essences are peculiar to their bearers, implies that knowledge is of the peculiar things. Again, assuming that there is knowledge of things, it follows that there are essences. Hence, Theophrastus can spell out the claim about being in more detail: the differences between kinds of things arise from, or perhaps simply are, their (peculiar) essences.Footnote ⁷¹

One might wonder where the cluster of claims about essence on which Theophrastus relies come from. Once again, a natural answer is Meta. Z. In Z.13, 1038b9–10, Aristotle says that the essence of an entity is peculiar to it. Moreover, he insists that definition is of essence and form (see, e.g., Z.5, 1031a11–14; Z.11, 1037a25–9), as well as on the link between definition and knowledge: knowledge is of that of which definition is (see, e.g., Z.15, 1039b27–40a7). Given the brevity of Theophrastus’ remarks, and the lack of defence, he is plausibly relying on the account presented by Aristotle in Z.Footnote ⁷² But this is not to say that Theophrastus uses the claims in the same way as Aristotle. For in Z.13, Aristotle famously appeals to the claim that essence is peculiar to its bearer to argue that universals are not substances and principles of things (1038b8–15). By contrast, if I am right, Theophrastus uses the peculiarity of essence in support of his extensional claim about being, namely that there are different kinds of entities.

One might object to my subordination of the claims about knowledge to the claim about being on the grounds that it sits ill with the rest of the section (8b24–9a9). For Theophrastus develops the thesis that ‘it is characteristic of knowledge that it comprehends (sunidein) the same thing in several things’ (8b24–5). More precisely, a ‘complete’ (teleos) branch of knowledge grasps what is universal and common to multiple kinds of thing, as well as what is peculiar to particular kinds of thing (8b25–7). For example, mathematics grasps what is common to numbers and lines (e.g., that they are quantities), but also what is peculiar to each of them (8b26–7).Footnote ⁷³ Moreover, he notes that the goal of some branches of knowledge is universal, whereas the goal of others (practical and productive ones) is ‘particular’ (en merei) (8b27–9a4).Footnote ⁷⁴ Finally, he allows that what is ‘the same’ in several things allows for degrees of strength: it can be the same in substance, number, species, genus, or most weakly, by analogy (9a4–9).Footnote ⁷⁵ These claims seem designed to spell out the thesis that knowledge comprehends what is the same in several things, without any link to the claim about being.

But in my view, Theophrastus’ interest in knowledge in these lines remains entangled with the claim about being. For, in arguing that a complete branch of knowledge grasps what is common and what is peculiar to different kinds, he elaborates on the earlier claim that there is no knowledge without differences between universals at the level of genera and species (8b16–20), as well as the claim that knowledge is of peculiar things. But these earlier claims were adduced to support the claim about being, as I argued earlier. Hence, the entire passage (8b10–9a9) discusses knowledge with a view to the claim about being.

But how does the claim about being fit into the larger argumentative context? Since it is a claim about differences between kinds of entities, and since these differences are spelled out in terms of essences, I suggest that its main contribution is to add an essentialist layer to hylomorphism. For, although Theophrastus does not explicitly identify essence and form in our passage, given how many Aristotelian assumptions he adopts, presumably he accepts the identification.Footnote ⁷⁶ If this is right, he implicitly adds the claim that the form of an entity is its essence to his hylomorphic analysis of natural things. Thus, he completes the account of the principles of natural things: they include the prime mover as the first external moving cause, matter, and finally form or essence.

Compared with the first critical wave, the second wave is critical in an attenuated sense. For Theophrastus emerges as an advocate of the Aristotelian account of the principles of natural things who defends it against alternatives and misinterpretations. But this is also a role that an internal critic of a theory may play: to safeguard it from nearby views and misconstruals. As I will argue next, the final critical wave (9a10–12a2), which contains Theophrastus’ famous discussion of teleology, is similar in spirit to the second wave in this regard and does not, as some scholars have thought, amount to a rejection of final causation.

3.1.1 From Knowing to Causal Enquiry

Regarding the question how one should pursue each of the things (9a10–11), Theophrastus says that the ‘starting point and most important thing’ (archē kai megiston) is the ‘kindred manner’ (oikeios tropos) (9a11–12). He argues for this claim by appeal to objects of different sciences: the first and intelligible things and the things that are movable and belong to nature, and among the latter, the things that ‘are in power’ (en archēi) and the ones that follow them, all the way down to animals, plants, and inanimate things (9a11–15).Footnote ⁷⁸ For short, for each genus, there is ‘something peculiar’ (ti … idion) (9a15–16).Footnote ⁷⁹ Although the examples are of different objects of sciences, what Theophrastus seems to drive at is that the sciences are in turn differentiated by their differing objects. The ‘manners’ or tropoi, then, are simply the sciences or branches of knowledge (cf. Laks and Most Reference Laks and Most1993, 67–8).Footnote ⁸⁰

Thus, for each domain of objects, there is a ‘kindred’ or ‘peculiar’ (idios, 9a20) science, that is, there is exactly one science that is suitable for investigating each domain. It is clear, then, why the tropos is the starting point for explaining how each of the things is to be pursued. As in the case of being, knowing is in several ways in the extensional sense: there are several sciences or branches of knowledge. This is why Theophrastus can couple the questions how many tropoi there are and in how many ways knowing (to eidenai) is (9a23–4): they are the same question.Footnote ⁸¹ Therefore, he claims that, to know how each thing is to be pursued, one first has to establish what, and how many, the relevant sciences are. This preliminary step allows one to choose the science appropriate for pursuing each of the objects.

But the issue is complicated further by Theophrastus’ addition of an intensional point. He says that one must delimit (aphorisai) ‘what knowing is’ (ti to epistasthai) (9a25–6) and adds that knowing belongs to ‘the things said in several ways’ (tois pleonachōs legomenois), which is why one cannot grasp ‘something universal and common’ to all instances of knowing (9a27–b1). Further, Theophrastus calls the question what knowing is the ‘starting point’ (archē) ‘for these very things’ (pros auta tauta) (9a25), that is, presumably the preceding double question how many tropoi there are and in how many ways knowing is (9a23–4). If the previous question is extensional, Theophrastus’ point seems to be that, to settle how many branches of knowledge there are, one must answer the intensional question what knowledge is, which is, however, difficult since knowing is said in several ways.

Theophrastus does not say any more on this issue.Footnote ⁸² Instead, he moves on to the question that interests us most, namely, ‘to what extent and of which things one must seek the causes, similarly among both perceptible and intelligible things’ (9b2–4). What is the connection between this question and the discussion of knowing? Adopting Ross’ ᾗ (‘wherefore’) at 9b1,Footnote ⁸³ Theophrastus says that it is ‘puzzling (aporon) or not easy to say’ (9b1–2) to what extent and of which things one must seek the causes because knowing is said in several ways (or because knowing is difficult to delimit, which is due to its being said in several ways).Footnote ⁸⁴

Why does the fact that knowing is said in several ways make it hard to say of which things one should seek the causes, and how far one should push the causal search? As we will see shortly, Theophrastus connects the issue of the limits of causal inquiry with a worry about an infinite regress (9b4–5). Moreover, he distinguishes between things that we can ‘behold’ (theōrein) ‘through a cause’ (di’aitiou) and things that we cannot behold through a cause (9b8–11). If we take both the claim that ‘knowing’ (eidenai) is said in several ways and the claim about ‘beholding’ (theōrein) to concern broadly our grasping of things, a link with the limits of causal inquiry emerges. If one has not settled what grasping something involves, it is hard to say where to stop the causal search. For one cannot distinguish things that must be grasped through a cause from others. Moreover, if one makes the wrong assumption about knowing or grasping, namely that it is always causal, one will end up with an infinite regress.

There is, then, a relatively straightforward connection between the intensional worries about knowledge and the limits of causal inquiry. The link with the lengthy extensional discussion of knowledge is less obvious but emerges clearly in Theophrastus’ treatment of the example of astronomy (9b24–10a21). I will return to this example next, but roughly, depending on whether one identifies the science of first things with astronomy and treats it as an entirely ‘natural’ (phusikos) tropos or branch of knowledge or not (10a8–9), one will either stop the causal search once one has reached the first natural cause or pursue the causal chain further to intelligible principles. Hence, how one carves up the objects of science, and thus the sciences themselves, bears on where one sets the limit for causal inquiry.

On the reading I suggest, then, Theophrastus’ discussion of knowledge, starting from the claim that knowing is in several ways (9a10), issues in the central question as to the limits of causal inquiry (9b2–3), to which he will respond in the remainder of the treatise.

3.1.2 Causal Inquiry

I already alluded to one of Theophrastus’ motivations for imposing limits on causal inquiry, namely, to avoid a regress. On the back of his statement of the central question, he says: ‘for the way into infinity is in both cases [i.e., the case of perceptible as well as intelligible things] not favourable (allotria) and destructive of thinking (phronein)’ (9b4–5). Given this threat of a regress, it seems that some limit must be imposed on the extent of causal inquiry: the chain of causes must end somewhere. This insinuates also that there are some entities of which one should not seek the causes, namely, those in which the causal chains terminate.

Theophrastus spells out this account with the distinction between perceptible and intelligible things. Recall that, according to a basic tenet of the Aristotelian-Theophrastean framework, intelligible entities are principles of perceptible ones (see Section 1.2). In line with this tenet, he says that, although both perceptible and intelligible things are ‘starting points’ (archai) in some sense, perceptible things are ‘our’ starting point, whereas intelligible things are the starting point ‘simply’ (haplōs) (9b6–8). He also explicates the latter claim by saying that the intelligible things are the ‘end point’ (telos). This is a familiar Aristotelian thought: in the order of our inquiry, perceptible things are the starting point, whereas in the order of explanation, the intelligible things are the starting point or principle.Footnote ⁸⁵

This thought issues in a lesson concerning the limits of causal inquiry: ‘up to some point’ (mechri tinos) we can ‘behold’ (theōrein) things ‘through a cause’ (di’aitiou), beginning from perceptions (9b8–9). But when we reach the ‘highest’ (akra) and ‘first’ (prōta) things, and hence the termini of the causal search, we can no longer behold them through a cause (9b10–11). Thus, we get a first answer to the central question: one should seek the causes of things only up to the first principles, and one should seek the causes only of things that depend on the first principles, not of the first principles themselves.

Theophrastus also states different options for explaining why we cannot behold the highest and first things through a cause: either they do not have a cause, or we are too weak to grasp it (9b11–13).Footnote ⁸⁶ He himself seems to prefer the first option. For he adds, along the same lines as Aristotle, that ‘presumably that is truer (alēthesteron) that the beholding (theōria) of such things takes place by the intellect (tōi nōi) itself touching (thigonti) and as if clasping (hapsamenōi) [them], which is also why there is no deception about them’ (9b13–16).Footnote ⁸⁷ Thus, whereas there is a ‘beholding’ (theōria) or grasp both of the first principles and of what is explained by them, we grasp the latter through a cause but the former immediately by the intellect touching them. The attribution of this capacity of touch to the intellect suggests that it is not our weakness that prevents us from finding the cause of the first things; rather, they do not have a cause. Thus, the remarks about our cognitive access to the principles supply a further reason why we should impose a limit on causal inquiry. If we attempted to grasp the first principles through a cause, there would not only be an infinite regress, but we would also not adequately grasp the principles themselves.

Moreover, since the perceptible things are the starting point of our inquiry, whereas the intelligible things are the end point (9b7–8), the first principles we cannot grasp through a cause seem to be the intelligible things, whereas the things that we can grasp through a cause are the perceptible things. This sheds light on what the intelligible things are. Given the Aristotelian framework as stated at the outset (Section 1.2), one might think that Theophrastus has in mind only unmoved movers. But this seems too restrictive. For in parallel passages in APo II.19 and Meta. Θ.10, Aristotle is concerned at least also with essences or forms.Footnote ⁸⁸ Further, since the start of the treatise, Theophrastus has supplemented the initial framework with hylomorphism and essentialism. At this point, then, he has plausibly in mind not only unmoved movers but also essences or forms as intelligible first principles.

Overall, then, Theophrastus outlines a general account of the limits of causal inquiry. This account is motivated by concerns both about an infinite regress and about our cognitive access to first (intelligible) principles. The limit imposed on causal inquiry is that one should seek the causes only up to the first principles (and hence that one should seek the causes only of dependent entities, not of primary ones). But there is also a more concrete worry, namely, in which thing exactly one should set the limit. It is this further issue Theophrastus addresses next, with the example of astronomy.

3.1.3 Astronomy and the First Things

Having discussed how we can behold or grasp first principles, Theophrastus adds a further difficulty concerning the limits of causal inquiry, namely, ‘in which thing (en tini) one should make the limit’ (9b19–20). According to Theophrastus, this difficulty is in itself important but also ‘necessary for the particular investigations (tas kath’ hekasta pragmateias), and most of all the greatest ones’ (9b17–19). His examples are ‘the [investigations] of nature’ and ‘the [investigations] that are even more primary’ (9b20–1). This issue recalls the earlier concern with the ‘delimitation’ (aphorismos) of the study of first things from others, especially from natural philosophy (4a2–9). But the ‘limit’ (horos) sought here is not between sciences, but the limit of causal inquiry within a science (cf. van Raalte Reference van Raalte1993, 462).

This point emerges clearly from Theophrastus’ ensuing worries about an infinite regress: ‘for those who seek an account of all things destroy any account and with it also knowing’ (9b21–2). Or better, ‘they seek [an account] of things of which there is none and is not any by nature (pephuken)’ (9b22–4). Both claims recall the central question from earlier (9b2–5). The thinkers he criticizes face an infinite regress because they do not know to what extent to seek an account, and they do not get the domain of things right for which one should seek an account. Theophrastus does not express these claims in explicitly causal terms, but they can surely be understood this way: to seek an account (logos) is to seek a cause.

Thus, Theophrastus is still concerned with the limits of causal inquiry. But whereas before, he was concerned with the abstract principle that one should not seek causes beyond the first principles, he now asks concretely ‘in which thing’ one should draw the limit, that is, which thing one should make the terminus of the causal inquiry.Footnote ⁸⁹ Moreover, although the issue of the limits of causal inquiry is distinct from the question about the delimitation of sciences from each other, the clue of the ensuing passage is that Theophrastus connects these issues with each other. In this way, he also links the question about the limits of causal inquiry with the earlier extensional discussion of knowledge and its tropoi.

Theophrastus proceeds with the example of astronomy (9b24–10a21).Footnote ⁹⁰ He begins with a demand to ‘those who take the heaven to be eternal’ (9b24–5) and also accept other results of astronomy concerning locomotion, sizes, shapes, and distances (of heavenly bodies) (9b24–7). I will call them ‘astronomical thinkers’, and they presumably include Aristotle as well as Theophrastus, given his acceptance of the Aristotelian framework (Section 1.2).Footnote ⁹¹ He demands from the astronomical thinkers that they ‘state both the first movers and that-for-the-sake-of-which and what the nature of each thing is and their position relative to each other and the substance of the whole (hē tou sumpantos ousia)’ (10a2–4). Further, as an astronomical thinker ‘descends’ (hupobainonti) (into the sublunary world), she should do the same concerning the species or parts, all the way down to animals and plants (10a3–5).Footnote ⁹²

One might think that Theophrastus is engaged in hyperbole to ridicule an opponent. But at the end of the treatise, he restates in earnest that one should seek ‘in which things the things that are (ta onta) are and how they relate to each other’ (12a1–2). There is no reason to think that Theophrastus considers these ambitious questions ridiculous. Rather, he suggests that it does not suffice to state the astronomical results precisely because stating these results is not sufficient to answer the more ambitious questions. For, as he continues, if astronomy ‘assists’ (sunergei) in this task but does not concern the first things, ‘the most eminent (ta kuriōtata) and prior things would be other than nature’ (10a5–6). This is because ‘the manner (tropos) is not, as some believe, natural (phusikos) or not entirely’ (10a8–9). Following Ross (Reference Ross and Fobes1929, 70), I take the ‘manner’ or tropos under discussion to be the science of first things. Thus, in a way reminiscent of Aristotle’s Meta. E.1, 1026a27–32, Theophrastus says that the first science is not natural philosophy if astronomy does not study first things.

I will shortly ask whether Theophrastus endorses the antecedent of this conditional. But even setting aside this question, we can see how the discussion of astronomy matters for the limits of causal inquiry. Depending on how one carves up the branches of knowledge, and in particular, depending on whether one thinks that besides natural science, there is another science, one will reach different conclusions as to the limit of causal inquiry. If one thinks that there is no science beyond natural science (because there is no domain of objects beyond the domain of objects of natural science), one will stop the causal inquiry once one has reached the first natural cause, such as the first heaven. Otherwise, one will pursue the causal chain further to the intelligible unmoved mover that is its terminus. Hence, the entity in which (en tini) one ought to place the limit of causal inquiry depends on the range of objects one thinks there are, and thus on the range of sciences and their delimitation from each other.

What, then, is Theophrastus’ own view of the range of sciences? On might think that he rejects the claim that the tropos of first things is not (entirely) natural. For he goes on to claim (kaitoi, 10a9) that being in motion (to kineisthai) is ‘kindred’ (oikeion) or peculiar to nature ‘in general’ (haplōs) and most of all to the heaven (10a9–10).Footnote ⁹³ He then raises the issue whether the heaven is not essentially (kata tēn ousian) in circular motion (10a14), such that one should either not seek an explanation for it ‘or in some delimited (aphorismenon) manner’ (10a16–18). On this view, then, Theophrastus suggests that the limit of causal inquiry is the motion of the (first) heaven and that one should not seek any cause beyond it. Thus, the tropos or science of first things is entirely natural; it is simply astronomy.

But as Theophrastus says (10a19–21), the issue is bound up with the earlier discussion whether circular motion is part of the essence of the first heaven (6a5–14). But I already argued against the common interpretation of this passage that has Theophrastus attack the need for an unmoved mover to explain the motion of heavenly bodies (Section 1.3.2). Moreover, Theophrastus explicitly allows that one seek an explanation for the motion of the heaven ‘in some delimited (aphorismenon) manner’ (10a19). This may be read as an allusion to the view that one should not postulate a further moved moving cause but rather an unmoved one.

Hence, I suggest that Theophrastus endorses the antecedent of the aforementioned conditional: astronomy assists in answering the questions raised for the astronomical thinkers, but it cannot itself answer them because it does not concern the first things. His point, then, is not that one should stop the causal inquiry at the first natural cause but rather that one should not. For this, too, matters for determining the limits of causal inquiry: not to be too cautious and to set the limit before one has reached the first principle.

For short, then, Theophrastus argues that one should limit one’s causal inquiry to things that depend on first principles, but he also seems to think that the first principles are intelligible things, including unmoved movers, and that, therefore, one should not abort the causal search before one has reached these principles. Let us now see how these conclusions apply to the special case of final-causal inquiry.

3.2.1 The Counterexamples

Theophrastus backs up with various examples the claim that some things do not seem to be for the sake of something but rather are coincidentally or by necessity. The examples are drawn both from the superlunary and the sublunary world (10a27–8). He begins with the sublunary cases (10a28–11a1) and returns to the superlunary case only at the end of his treatment of teleology, after he has made his positive proposal about the teleological principles (11b12–23).

The sublunary examples divide into four groups. The first group (10a28–b7) consists of large-scale natural processes, such as the tides, and ‘generally changes (metabolai) and perishings (phthorai) and generations (geneseis) towards something else at various times’ (10b2–4). The second group (10b7–16) consists of odd features of animals, among which ‘some are like vain (mataia)’ (10b7–8), such as male breasts, whereas others are ‘by force or contrary to nature (para phusin)’ (10b14–15). The examples of the third group (10b16–20) are ‘the most important and most apparent (malista dokoun)’ (10b16–17) and concern ‘the growths (trophai) and generations of animals’ (10b17–18) which are ‘not for the sake of anything but are coincidences (sumptōmata) and due to other necessities’ (10b18–19). The fourth group (10b21–11a1) consists of features of plants and inanimate things, about which Theophrastus says that, plausibly, ‘these latch onto certain shapes and differences from each other spontaneously (tōi automatōi) and by the rotation of the whole’ (10bb27–11a1).

Since others have examined these cases and their relation to Aristotle’s biological corpus in detail (see especially Lennox Reference Lennox, Fortenbaugh, Huby and Long1985; Repici Reference Repici1990; Gourinat Reference Gourinat2015), I restrict myself to a few general remarks. First, as Gourinat (Reference Gourinat2015, 158–64) notes, it is not anti-Aristotelian to think that certain phenomena are not for the sake of something but due to coincidence or necessity. After all, in Phys. II.4–6, Aristotle makes room for, and explains, phenomena that come to be by chance and spontaneity. Moreover, in PA I.1, 642a2–14, Aristotle says that there are two kinds of cause (aitiai), ‘the for-the-sake-of-which’ and ‘the from necessity’. Necessity again subdivides into what Cooper (Reference Cooper, Gotthelf and Lennox1987, 259) has called ‘Democritean’ necessity and necessity ‘from a hypothesis’ (ex hupotheseōs) (642a4–14). Whereas the latter seems conditional on a final cause (e.g., if there is to be an axe, it must be made of hard material), Democritean necessity is independent of the final cause and due to the nature of the matter involved.Footnote ⁹⁶ Finally, Aristotle acknowledges that some phenomena may not have a final cause (see, e.g., Meta. H.4, 1044b12; PA IV, 677a17–19, cited in Gourinat Reference Gourinat2015, 161) and hence are not apt for teleological explanation (cf. Repici Reference Repici1990, 207).Footnote ⁹⁷

Second, we can account for the examples of the first, third, and fourth groups from within Theophrastus’ own treatise. Earlier, he voiced a worry about the Aristotelian view that the locomotion and changes of sublunary entities seem caused accidentally by the motion of the heavenly bodies (5b24–5). He now elaborates on this point by noting that large-scale natural processes, the growth and generation of animals, and features of plants and inanimate things, are caused accidentally ‘due to other necessities’ (10b19), that is, presumably, the motion of the heavenly bodies.Footnote ⁹⁸ This time, he does not raise a concern about the accidental causation of these phenomena but rather takes it for granted. Hence, if the earlier worry about accidental causation was a worry about Aristotle, the present endorsement of accidental causation is surely not directed against Aristotle.

This leaves us finally with the examples in group two. Some of them are phenomena that seem ‘vain’: the breasts of males, the ‘emission’ (proesis) of females, the growth of a beard in certain animals as well as the growth of hair in certain places, the size of horns (for instance, in the case of deer that are obstructed by them) (10b7–14). Others are ‘by force or contrary to nature’ such as the copulation of the heron and the life of the day-fly (10b14–16). One may debate whether ultimately, for Aristotle, these phenomena are for the sake of something.Footnote ⁹⁹ But what is more important is that, in principle, Aristotle too seems prepared to admit that certain phenomena do not have a final cause, such as the colour of the eye’s iris (GA V.1, 778a29–b6) (cf. Most Reference Most1988a, 231–2). Moreover, Theophrastus’ examples are precisely the sorts of case where Aristotle too hesitates to assign a final cause. For example, he says that the discharge of females does not contribute (sumballetai) anything to the embryo (GA II.4, 739a20–6) (Repici Reference Repici1990, 209; Gourinat Reference Gourinat2015, 168),Footnote ¹⁰⁰ and that for some animals the ‘prominence’ (exochē) of the horns is ‘useless’ (achrēstos) (PA III.2, 663a8–9) (cf. Gourinat Reference Gourinat2015, 168–9).Footnote ¹⁰¹ The fact that Theophrastus picks up on cases Aristotle finds problematic indicates that they partake in the same research programme, even if they were to disagree on certain cases.Footnote ¹⁰²

Something similar holds for the superlunary case (11b12–23). Theophrastus says that ‘even among the first things many [phenomena] are evidently observed [to occur] also as it happens (hōs etuche)’ (11b12–14). As an example, he mentions ‘what has been said about the changes (metabolas) around earth’ (11b14–15) and a little later adds phenomena ‘in the air and other [places]’ (11b16–17). The reference of ‘what has been said’ (ta lechthenta) (11b14) is to the part of the anti-formalist argument (7a22–b8) where Theophrastus stated a nuanced final-causal account as an alternative to the formalist view that assigns a final cause even to the bubble.Footnote ¹⁰³ According to this nuanced account, the various shapes in the region of earth and air come about due to the order and change of other things, as on the view that the seasons and generations of living things are due to the sun (7b2–5). The phenomena in earth and air occur ‘as it happens’ in that they do not themselves have a final cause (11b15) but are accidental occurrences due to the motion of the heavenly bodies (cf. 5b24–6). But as we have seen (Section 2.2.1), this nuanced final-causal view is plausibly Aristotle’s. Theophrastus’ claim that even among the first things, phenomena occur ‘as it happens’ is in line with, not contrary to Aristotle.Footnote ¹⁰⁴

This picture leaves us with two related questions: against whom does Theophrastus put forward his counterexamples, and in what sense is he critical of Aristotle at all? I will return to these questions after we have looked at his positive account of final-causal inquiry.

3.2.2 The Limits of Final-causal Inquiry

Theophrastus introduces his account of final-causal inquiry as follows: ‘but if not, one must grasp some limits (horous) of the “for-the-sake-of-which” and “for the best” and not posit them simply in all cases’ (11a1–3). The positive claim is clear enough: one must recognize the limits of final causation, and thus of the range of things to which the vanity and axiological principles apply. The ‘but if not’ (ei de mē) presumably picks up the preceding ‘mechanistic’ view that one may be tempted to embrace in light of the fourth group of examples, that is, the features of plants and inanimate things (10b26–11a1). Thus, Theophrastus entertains two responses to the counterexamples, namely, to dispense with final causation altogether or to recognize its limits, where the first response is rejected and the second developed.Footnote ¹⁰⁵

There are limits both in the case of nature as a whole (11a5–7) and in the case of particular animal kinds (11a7–13). Theophrastus defends the limits in terms of the axiological principle. It is doubtful that nature ‘desires the best in all things and gives a share of eternity (tou aei) and what is ordered in all cases where this is possible’ (11a5–7), and similarly that in animals everything is, where possible, arranged for the better (a doubt exemplified by the windpipe) (11a7–13).Footnote ¹⁰⁶ I take it that the ‘doubt’ (distasmon, 11a4) derives from the examples stated previously, and that it extends to the vanity principle as well: if it is doubtful that everything, where possible, is for the best, it is also doubtful that everything has a final cause.

In addition, Theophrastus provides a back-up argument (11a13–18): even if the desire of nature is ‘like this’ (11a14), that is, even if it desires the best in all things, many things do not ‘submit to’ (hupakouon) and are not ‘receptive’ (dechomenon) of ‘the good’ (to eu) (11a15) (cf. 5b28–6a1). Indeed, there are many more such things than things receptive of the good (11a15–16). In support, Theophrastus assumes that it is better to be animate than inanimate (11a17–18) and points out that inanimate things are ‘boundless’ (apeiron) in number whereas there are only few animate things, and that even animate things tend to live only for a short while (11a16–18).

These arguments, if successful, show that there are limits to final causation. But if this is right, there ought to be limits to final-causal inquiry as well. For otherwise one will inquire into the final causes of phenomena that do not have any. Or perhaps one may inquire into the final causes of such phenomena but only up to the point of recognizing that there are not any. Thus, Theophrastus’ account of teleology develops both the earlier anti-formalist argument and his general account of the limits of causal inquiry. Against the formalist, he furnishes further arguments against attributing a final cause to everything and against the view that the only type of explanation is final-causal. In further defence of the limits of causal inquiry, he shows that there are limits also in the special instance of final-causal inquiry.

I have argued that none of these points are directed against Aristotle. Instead, since the treatment of teleology continues the anti-formalist argument, plausibly Theophrastus targets the same opponents as before, namely, the formalists who hold that there are only formal-final causes and who attribute a final cause to everything. I noted that Theophrastus seems to think that this view emerges in the Timaeus (6b27), although he treats Plato also as a proponent of the view that the principles are opposites, as we will see shortly. But again, there is another plausible target given the context of Theophrastus’ treatise, namely, a misleading formalist construal of Aristotle. Especially if Theophrastus is broadly in agreement with Aristotle, as I have argued, it is crucial for him to distance the view he accepts from a nearby account into which one might be tempted to develop it.

However, an objection to this interpretation arises from a section about the Platonists (11a18–b12) that intervenes between the back-up argument for the limits of final causation (11a13–18) and the superlunary counterexample (11b12–23). For, according to Laks and Most (Reference Laks and Most1993, 86), Theophrastus here leaves behind Aristotle and instead accepts a Pythagorean-Platonist account of opposites (enantia). Given his agreement with Aristotle on the limits of (final-)causal inquiry noted so far, this would be a surprising move; and indeed, on inspection, there is little reason to think that Theophrastus joins the Platonist camp.

In the wake of his claim that there are many more things that are not receptive of the good than things that are receptive of it (11a14–18), Theophrastus turns to Platonist views on this topic. First (11a18–26), he seems to reject an extreme view that he attributes to Speusippus according to which the good is ‘rare’ (spanion, 11a18–19; a23) and ‘in few things’ (en oligois, 11a19).Footnote ¹⁰⁷ But his attitude towards the second view (11a27–b12), attributed to ‘Plato and the Pythagoreans’ (11a27) is less straightforward. On this view, there is an ‘opposition’ (antithesis) between the one and the indefinite dyad, where ‘the boundless (apeiron) and the unordered (atakton) and all, so to speak, shapelessness (amorphia) by itself’ is in the dyad (11b2–5). The ‘nature of the whole’ cannot exist without this dyad, and the dyad ‘balances out or even exceeds the other’, that is, the nature of the whole (11b5–7). Overall, then, the two principles (archas) are contraries (enantias) (11b7). As a consequence (11b7–12), not even the god ‘can lead everything to the better’ (11b8–8) (since the dyad is a principle), and hence there must be a limit to the good, and to final causation, in the world.

What is Theophrastus’ stance on this theory? According to Laks and Most (Reference Laks and Most1993, xxv), he embraces it as a non-Aristotelian alternative to the view that assigns a final cause to everything. But if this is right, his acceptance of the Platonist view and his rejection of the Aristotelian view are poorly motivated. For the Aristotelian view is already an alternative to the formalist view that assigns a final cause to everything. Moreover, in his discussion of Speusippus, Theophrastus seems to endorse the Aristotelian view that explains the bad in terms of indefiniteness and matter (11a20–1), just as he did earlier (8a21–b8; see Section 2.2.3). It remains mysterious, then, why he should suddenly feel compelled to accept the much more far-reaching claim that there is an indefinite dyad. Finally, there is no textual marker that signals Theophrastus’ acceptance of the Pythagorean-Platonist view.

Instead, I suggest that Theophrastus rejects the Pythagorean-Platonist view no less than the Speusippan one. There is evidence of this rejection at the beginning of the presentation of the view. For Theophrastus says that Plato and the Pythagoreans say that ‘the distance (apostasin) is long’ (11a26–b1), that is, presumably the distance between the indefinite dyad and the one which they put in ‘opposition’ (antithesin) (11b2).Footnote ¹⁰⁸ In the present context, this implies a long distance between the bad and the good. But if Theophrastus follows the Aristotelian view recalled earlier (11a20–1), which assigns the bad to matter and the good to form, introducing a long distance between good and bad is a problem. For in Aristotelian terms, this means introducing a long distance between form and matter. But on the Aristotelian view, matter and form are unified because one is substance in potentiality and the other in actuality (see Section 2.2.2), which seems ruled out by ‘a long distance’ between them.

If this is right, Theophrastus rejects the Pythagorean-Platonist view and remains committed to the Aristotelian account of final causation and its limits. This does not change in the final lines of the treatise (11b24–12a2). These lines are sometimes treated as the conclusion of the whole treatise (see Laks and Most Reference Laks and Most1993, 22), but they are better read more narrowly as the conclusion of the section on teleology.Footnote ¹⁰⁹ For Theophrastus says first about the issues raised by the superlunary example that they need to be examined further (11b24) and then reiterates that we need to try and grasp ‘some limit (horon)’ both in nature and in ‘the substance of whole’, and of both the vanity principle and the axiological principle (11b25–7). This remark encapsulates the core of his (Aristotelian) treatment of teleology: we must grasp the limits of final causation in the world and, on that basis, limit our final-causal inquiry into the world.

What is more, Theophrastus undergirds the commitment to limits of final-causal inquiry with a further consideration: grasping the limits of final causation is the starting-point of a study of the whole world because it can tell us ‘in which things (en tisin) the things that are (ta onta) are and how they relate to each other’ (12a1–2). This last thought is reminiscent of Aristotle’s Meta. Λ.10, 1075a11–25, one more time. The place of things and their relation to each other depends on the order of the world. But this order in turn depends on the first final cause, that is, the prime mover, which moves the heavenly bodies directly but the sublunary entities only indirectly. Hence, to understand the place of things and their relation, we need to grasp the final-causal structure of the world and its limits.

Since I have mainly confronted opponents who take Theophrastus to abandon Aristotelian teleology, I have emphasized his agreement with Aristotle on the limits of causal inquiry in general and final-causal inquiry in particular. This may make one wonder to what extent he is still a critic of Aristotle, even an internal one. My answer is the same as in Section 2. As in the anti-formalist argument, the main achievement of Theophrastus’ treatment of teleology is to safeguard the Aristotelian account from a misconstrual, arguably more explicitly than Aristotle does himself. In sensing the danger of the misconstrual to which the Aristotelian account might lend itself, he remains critical of Aristotle, although in the friendly spirit of an internal critic who defends the true version of a theory.