The Five Ways and the Argument from Composition: A Reply to John Lamont

THE FIVE WAYS AND THE ARGUMENT FROM COMPOSITION: A REPLY TO JOHN LAMONT ANTOINE COTE University of Ottawa Ottawa, Ontario, Canada I N AN ATTEMPT to vindicate the celebrated "Five Ways," John Lamont tries to show that Aquinas's arguments for an uncaused cause are successful provided they are understood as resting on an argument from composition.' Lamont further seeks to show that an uncaused cause must be immaterial and unique. In this paper, however, I shall argue that even if we accept the translation of Thomas's various proofs into an argument from composition, such an argument need in no way be thought of as implying the existence of an uncaused cause. Further, I shall show that Lamont's argument for the immateriality of the uncaused cause is problematic and his argument for its uniqueness unconvincing. I Lamont gives the following version of the proof of the existence of an uncaused cause. i. There are effects. ii. To be an effect is to have a cause. iii. Nothing can cause itself. iv. Premise (i) states that there are effects that occur in the worldeffects A, B, C, D, etc. Each of these effects, since it is an effect, has a cause which is different from itself. Now consider the group of all the 1 John R. T. Lamont, "An Argument for an Uncaused Cause," The Thomist 59 (1995): 261-77. 123 124 ANTOINE C6TE effects that there are; call this group X. Since the parts of X are effects, X itself is an effect. Since it is an effect, it has a cause. This cause cannot be itself an effect. For if it were it would be a part of X, and so X would cause itself, which is impossible. Thus the cause of X is not itself an effect. Thus there is an uncaused cause of all effects. (262-63) Let us take a cursory look at the argument before turning to Lamont's justification of it. The first problem arises when Lamont asks us to "consider the group of all the effects that there are." Now such a group is what in more conventional terminology one would call a class or a set. By definition all the "parts," that is, the members of this set, are effects, but it would be absurd to suggest that the set itself is an effect. Lamont does state in a footnote that he has purposely avoided using the term sets, preferring to speak of wholes or groupings. His reason for doing so is that "[s]ets are abstract objects that are not subject to causal influences."2 This implies that groups are subject to such influences . But this is precisely what Lamont has to prove. He must prove that the group of all the effects that there are is indeed a group in the requisite sense and not a purely mental construct. Such a proof seems all the more necessary when one considers that two out of the three examples of invalid cases of the composition argument given by Lamont in his paper involve precisely such an illicit inference from members of a class to the class itself. It would be unfair, however, to reject Lamont's attempt without first examining the details of his argument. I therefore turn to section I of his paper, where the main arguments are to be found. One usually thinks of composition arguments as fallacies. Lamont successfully shows that not all composition arguments are fallacious. For instance, although the first of the following two examples is clearly a case of an invalid use of the composition argument, the second is not: -Every atom in the universe is smaller than a basketball. -Therefore all the atoms in the universe are smaller than a basketball. (264) 2 Ibid., 270 n. 9. A REPLY TO JOHN LAMONT 125 -Each person in the crowd is east of the river. -Therefore the crowd is east of the river. (265) The main problem with composition arguments, according to Lamont, is that no one has succeeded in establishing any universal rules for distinguishing valid from invalid ones. Although it would obviously be preferable if such rules did exist-we would know whether...