Stage la

If every mover is moved, then this proposition is true either perse or per accidens. If per accidens, then it is not necessary, for what is true

per accidens is not necessary.

30 In the parallel passage in Aquinas's commentary on Aristotle's Physics, he provides an explanation that might be attached to this second premiss: 'Nothing that is per accidens is necessary; for what is in anything per accidens is not in it necessarily but can fail to be in it—like musicianship in a builder.' That is, if it's true that the builder is a musician, it's true not per se but per accidens. 'Therefore, if movers are moved per accidens, it follows that they can fail to be moved [without failing to be movers]' (In Phys. VIII: L9.1043). See also In Phys. VIII: L9.1042, quoted just below.

Therefore, that no mover is moved is contingent. But if a mover is not moved, it does not act as a mover (non movet) (as the adversary

31 Cf. the parallel: 'But since you [the adversary] claim that every mover is moved, it follows that if movers are not moved, they do not act as movers' (In Phys. VIII: L9.1043).

Therefore, that nothing is moved is contingent, since if nothing acts as a mover, nothing is moved. This, however, Aristotle considers to be

impossible—namely, that there be a time when there is no moving.

32 Physics VIII 1, 250bll-252a4.

Therefore, the first was not contingent, since what is false and impossible does not follow from what is false and contingent. And so the proposition 'Every mover is moved by something else' was not

true per accidens. (13.97)

I think that the first lemma, which stage la is designed to reject, is an unlikely interpretation of the adversary's thesis anyway, because if a mover is moved only per accidens, then it isn't moved qua mover, in its capacity as a mover. That this is what Aquinas means by being moved per accidens can be seen in a parallel passage in his end p.66


© Copyright Oxford University Press, 2006. All Rights Reserved

Physics commentary, where he presents the distinction not in terms of truth, but in terms of the inherence of properties: 'If everything that is moved is moved by something that is moved—i.e. if every mover is moved—this can occur in two ways. In one way, so that a mover's being moved is found in things per accidens— I mean, so that the mover does not do its moving on account of its being moved (as if we were to say that a musician is a builder not because he is a musician but per accidens)—or [in the second way] so that it is not per accidens that the mover is moved, but perse' (In Phys.

34 The translation of this passage in Blackwell et al. 1963: 516 leaves the words ut movens moveatur untranslated, and mistranslates the words propter id quod movetur.

On this interpretation, I think, any worthy adversary of the sort Aquinas is envisaging would disown the per accidens lemma without a fight.

If I'm right in suggesting that the rejection of the first lemma is a mere formality, then the fact that stage la is flawed has less practical importance than it would have otherwise. As Aquinas reads his adversary's thesis, in maintaining that every mover is moved, the adversary must be maintaining either that every mover is, by its very nature as a mover, something that is itself moved in its moving something, or that every mover is, merely as it happens, something that is itself moved whenever it moves something. But no worthy adversary would accept that reading of the thesis. The first premiss of la (in lines 1-2) is, no doubt, intended to be tautological, unobjectionable to the adversary; but the only interpretation on which it is so is this one: 'Every mover is moved either perse or per accidens.' Aquinas's casting the exhaustive disjunction in terms of the truth perse or per accidens of the adversary's universal thesis leaves the first premiss highly implausible.

Still, if we ignore the dubious moves and the unnecessary technicalities of stage la, we might be able to make out an acceptable line of argument here, one that could be sketched this way: if the adversary's thesis is true per accidens, then a mover's being moved is contingent, in which case it is possible that no mover is moved. But if no mover is moved, then—on the adversary's thesis—nothing acts as a mover, and in that case there is no moving. So, if the adversary's thesis is true per accidens, it is possible that there be no moving at all. But, as Aristotle claims, that there be no moving at all is /'mpossible. Now, what is impossible does not follow from end p.67

what is contingent, and so a mover's being moved cannot be contingent. Therefore, the adversary's thesis is not true per accidens.

It's in the announcement that Aristotle considers it to be 'impossible . . . that there be a time when there is no moving' (lines 7-8) that Aquinas springs his trap for the adversary, who is plainly expected to be in this respect an Aristotelian. This is the hypothesis of beginningless motion whose inclusion in G2 is a necessary condition of its being characterized as 'the most efficacious way'. Attributing it explicitly to Aristotle is important dialectically, but the explicit attribution is important also because Aquinas himself thinks it's not impossible that there once was no motion at all. He believes, of course, that in fact all motion began with creation (although he argues, contrary to most of his contemporaries, that that cannot be known except by recourse to

revelation, which is off limits here

35 For Aquinas's position in the medieval controversy over the possibility of a beginningless universe, misleadingly (but universally) designated de aeternitate mundi, see e.g. SCG II.30-8. See also Wippel 1984b; Wissink 1990; and, more generally, Dales 1990; Dales and Argerami 1991; Kretzmann 1985.

). For purposes of G2, then, Aquinas, like his adversary, fully accepts this Aristotelian hypothesis.

The proposition embedded in the conclusion of la, 'Every mover is moved by something else', isn't expressly identified as the adversary's thesis—which is called 'the first' in line 9—but Aquinas's use of the past tense in lines 9 and 11 strongly suggests that he is taking 'Every mover is moved' as simply an

abbreviation for'Every mover is moved by something else'.

36 In the parallel passage in his Physics commentary he retains the original version of the adversary's thesis: 'Therefore, it follows that at some time nothing is moved. However, that is impossible, because it was shown above [VIII 1, 250bll-252b6] that it is necessary that there always be motion. But this impossibility does not follow from our having supposed that movers are not moved. For if it is per accidens that a mover is moved, it will be possible that movers not be moved; and nothing impossible follows from the positing of what is possible. Therefore, we are left with the conclusion that the other thing from which it follows—I mean, that every mover is moved—is impossible' (In Phys. VIII: L9.1043).

This fuller version of the adversary's thesis can be inferred from the abbreviated version along with the Aristotelian principle 'Everything that is moved is moved by something else (Omne quod movetur ab alio movetur)'— OQM, I'll call it. The OQM principle is tacit and underived in G2, but it occurs explicitly as the first premiss of Gl, where it is supported by three sub-arguments; and so Aquinas is end p.68


© Copyright Oxford University Press, 2006. All Rights Reserved

within his rights to be assuming it here.

37 The OQM principle has been much discussed. For a good critical review of the literature and a very well-informed account of the OQM principle in medieval philosophy generally and Aquinas particularly, see Weisheipl 1965; also Lobkowicz 1968 and Weisheipl 1968 (a reply to Lobkowicz 1968).

I consider the tacit OQM to be acceptable within G2, especially because MacDonald's published analysis and appraisal of G1 and the First Way shows the acceptability of the OQM principle in this context (MacDonald 1991a: 128-32).

Consequently, although there are good reasons to worry about some features of stage la, I think we can extract an acceptable line of argument from it, at least for the sake of argument G2. I wish I could say the same of the much more important stage lb, which focuses on the second lemma, the

stronger, likelier interpretation of the adversary's thesis.

38 As I read SCG 1.13, stages la and lb of G2 are separated by what may be considered an ad hominem supplement to la: 'Again, if any two [characteristics] are conjoined per actidens in something, and one of them is found without the other, it is probable that the other can be found without that one. For example, if being white and being musical are found in Socrates, and in Plato one finds being musical without being white, it is probable that in some other man one can find being white without being musical. Therefore, if being a mover and being moved are conjoined in something per accidens, while being moved is found in something [else] without its being a mover, it is probable that being a mover is found [in some third thing] without its being moved. And one cannot raise as a counter-instance against this a case of two things of which one depends on the other and not vice versa (as is clear in the case of substance and accident), for [ex hypothesi] these [characteristics under discussion] are conjoined not per se [as accident is conjoined with substance] but per accidens' (13.98; cf. In Phys. VIII: L9.1044). Since this argument concludes to the probability of an unmoved mover, which is not something Aquinas wants to deny, I would bring out its ad hominem character by continuing it along the following lines: many things are moved although they are not movers, and so if'Every mover is moved' is true per accidens, then, probably, there is a mover that is not moved. But the adversary flatly denies that there is an unmoved mover, and so the adversary must also deny that 'Every mover is moved' is true per accidens.

Stage lb

But if the aforementioned proposition ['Every mover is moved by something else'] is true perse, then, similarly, something impossible or absurd follows. For the mover must be moved either with the same species of motion as that with which it acts as a mover, or with another. If with the same, then it will have to be the case that what alters is altered, and, further, that what heals is healed, that what teaches is taught—and with the same knowledge. But this is impossible;

end p.69


© Copyright Oxford University Press, 2006. All Rights Reserved for it is necessary that the one who is teaching have the know ledge, but it is [also] necessary that the one who is learning not have it. And so the same thing will be had and not had by the same thing—which is impossible. But if [the mover] is moved in accordance with another species of motion—e.g. so that what alters is moved locally, and what moves something locally is increased, and so on as regards the others—then, since there are finitely many genera and species of motion, it will follow that this cannot go on ad infinitum. And so there will be a first mover that is not moved by something else. Unless, perhaps, someone might say, 'Suppose there is recapitulation of this sort: once all the genera and species of motion have been used up, there must be a return to the first again, so that if what moves something locally is altered, and what alters it is increased, then what increases that is, again, moved locally'. But from this will follow the same as before—namely, that that which acts as a mover in accordance with some species of motion is moved in accordance with the same species, albeit not directly, but indirectly (non immediate sed mediate). Therefore, we are left with having to posit some first [mover] that is not moved by anything extrinsic [to it].

39 Cf. Physics VIII 5, 256b27-257al4; In Phys. VIII: L9.1046-9. The textual material quoted in the immediately preceding note intervenes between stages la and lb.

The genera of motion at issue here are just the three I mentioned earlier: locational change, quantitative change, and qualitative change (or alteration). After picking out alteration as the genus, the case developed in lines 5-7 picks out two subaltern species of alteration—healing and teaching—and then a most specific species of teaching: teaching some particular item of knowledge. For the argument to have any chance of deriving the absurdity it aims at, it must focus on most specific species of motion, as Aquinas himself insists in his commentary on the relevant passage in Aristotle's Physics: 'It is plainly impossible that a mover be moved with the same species of motion [as that with which it acts as a mover]. For it is not enough to stop at some subaltern species [of motion]; instead, one will have to go on through the process of division all the way to the individuals—i.e. to the most specific species. For example, if someone is teaching, he is not just being taught, he is teaching and being taught the same thing. I mean that if he is teaching geometry, he is being taught that same [bit of geometry] (hoc idem)' (In Phys. VIII: L9.1046).

There is no logical absurdity in your being altered in one way while altering something else in another way—in, for instance, your end p.70

Kretzmann, Norman , (deceased) formerly Susan Linn Sage Professor Emeritus of Philosophy, Cornell University, New York

Study Aid

Study Aid

This Book Is One Of The Most Valuable Resources In The World When It Comes To Getting A Scholarship And Financial Support For Your Studies.

Get My Free Ebook

Post a comment