1 Chisholm, “Human Freedom and the Self,” in Reason and Responsibility (2nd ed.), ed. Feinberg, J., pp. 359·366, a revision of his The Lindley Lecture of the University of Kansas (Lawrence: University of Kansas, 1964)Google Scholar. and Taylor, , Action and Purpose (Englewood Cliffs: Prentice-Hall, 1966)Google Scholar. In his first footnote, Chisholm cites Aristotle, Reid, Campbell, and Taylor as predecessors. My scarequotes around 'theory of agency’ reflect my feeling that these views are simply an incorrect theory of agency.

2 If Sam lifts his right arm with his left, he has caused the right arm to rise; but has he raised it? If Sam asks someone to lift his (paralyzed) arm, and this person does so, has Sam caused it to rise? Has Sam raised his arm? I assume an affirmative answer to all three questions. Anyone who disagrees can modify his understanding of ( 1) accordingly, without detriment to anything that follows.

3 Shaffer, Jerome A., Philosophy of Mind (Englewood Cliffs: Prentice-Hall, 1968), pp.87fGoogle Scholar.

4 My use of ‘event’ covers as needed what others might prefer to call sets or sequences of events, states, or processes.

5 Taylor (op. cit., p. 128) but not Chisholm (cf. footnote 8, below) allows that if an agent directly caused an event e, there may have been an event c which caused the agent to cause e, and in just this sense, e may have been caused by some event. But even for Taylor, e cannot have been directly caused by any event.

6 Chisholm, op. cit., p. 363.

7 Goldman, Alvin I., A Theory of Human Action (Englewood Cliffs: Prentice-Hall, 1970), p. 84Google Scholar.

8 The second clause of (R) should be understood as

Rb) Nothing caused Sam's causing his arm to rise.

That is, it asserts that nothing stands in the “caused” relation to Sam's act of

causing his arm to rise (i.e., his act of raising his arm; cf. ( 1) and footnote 2). But if nothing caused this act, then some acts─at least this one ─ are uncaused. Hence, (Rb) requires (MLT).

(Rb) may appear to conflict with the first clause of ( R), which is

(Ra) Sam caused his arm to rise.

They would conflict if Sam's act on this occasion, referred to in (Rb), were identical with the motion of Sam's arm on this occasion, referred to in ( Ra). Suppose we have

e =Sam's act of causing his arm to rise

f =the upward motion of Sam's arm.

(Ra) says that Sam caused f; (Rb) says that nothing caused e. If e and f were identical, if they were the same event, then (R) as a whole would be saying of this one event both that Sam caused it and that nothing caused it. But there is no good reason to think that e and f are identical. Chisholm denies it (op. cit., p. 363); Davidson, may be committed to its denial (see footnote 15, below); and I myself have argued against it (“Individuation of Actions,” The journal of Philosophy, LXVII [August 6, 1970], 520·530)Google Scholar. According to the view I defended there, f would be a proper part of e, caused by other proper parts of e (see footnote 15 below). If these other proper parts of e (or the earliest of them) were themselves wholly uncaused, we would have a case of e─Sam's act ─being uncaused, though f─the motion of Sam's arm ─ was caused by Sam. That is, the two clauses of ( R) are compatible. (Chisholm [op. cit.] has a different view. Though he would agree that e and f are non-identical, he thinks that for Sam to be responsible Sam must have caused both e and f. I find this unnecessary and counter-intuitive: “Sam caused e” translates as “Sam caused himself to raise his arm,” which does not seem appropriate for the normal case.)

9 Davidson, , “Actions, Reasons, and Causes,” The Journal of Philosophy, LX (November 7, 1963), 685–700CrossRefGoogle Scholar; Goldman,op. cit.

10 This “standard objection” was articulated as early as 1956 by A. I, Melden, in “Action,” The Philosophical Review, LXV (1956), 523–541Google Scholar; reprinted in Norman S, Care and Landesman, Charles (eds.), Readings in the Theory of Action (Bloomington: Indiana University Press, 1968), pp. 27–47Google Scholar. More recently, it was offered by Cornman, James, in “Comments” on Davidson's “Agency,” in Agent, Action, and Reason, ed. Binkley, Robert, Bronaugh, Richard, and Marras, Ausonio (Oxford: Basil Blackwell, 1971 ), p. 26Google Scholar. I quote in full because Cornman's positive suggestion is the closest I have seen to my own views:

I tend to agree with Davidson that the difference [between an arm raising and an arm rising] is a difference in cause. Where I am more skeptical is in agreeing with Davidson's classification of such differentiating causes as desires. I tend more to let the classification be determined by scientists, leaving open the question of whether they classify the relevant event as a brain process, or a desire, or perhaps some postulated theoretical entity that fits neither of those two classifications.

11 This case differs from the one cited in footnote 2, in which Sam asks someone to lift the arm; for in that case, Sam brings about the motion of his arm by doing something. Simply coming to have a reason to raise the arm is not a case of doing anything, and so in this case, I would deny that Sam has raised his arm, and also deny that Sam has caused his arm to rise.

12 Using the labels introduced in footnote 8, what Davidson and Goldman assert is that Sam's having a reason to raise his arm caused e, and did not merely cause f. But if e─and all acts─have causes, then (M L T), which says that some do not, is false. If we revert to (B) rather than (B’), then it is only implied that f was caused; the possibility remains open that e was not, and so (ML T) might be true. But now the problem is to understand the relation between e and f, iff was caused by Sam's having a reason to raise his arm, but e was not. The likeliest attempt is to say that Sam's having a reason, a mental event, is itself part of e. To say that e, Sam's act of raising his arm, was uncaused would then be to say that his having a reason to raise the arm was uncaused, though f, the motion of the arm, was caused. But this move makes the notion of “having a reason” even more obscure, and I do not think that Davidson or Goldman or any other proponent of a mental cause theory would find it palatable. Nor does it seem it would really help the libertarian. The latter would like to say that Sam could have refrained from raising his arm even though he had a reason to raise it. Claiming that nothing caused him to have a reason does not appear relevant to saying this.

13 Melden, , op. cit., and Free Action (London: Routledge and Kegan Paul, 1961)Google Scholar. The label ‘contextual’ is Shaffer's, op. cit., p. 94. Brand, Myles, in the introduction to his anthology, The Nature of Human Action (Glenview: Scott, Foresman and Company, 1970), p. 17,Google Scholar reads Melden as requiring the obeying of appropriate rules for an action to have been performed, and so dubs the theory a “Rule-Following” account. On my interpretation, which I think is entirely in accord with the spirit of, say, Chapter 13 of Free Action, Melden's theory is more flexible.

14 For argument, see Hilary Putnam, “Dreaming and ‘Depth Grammar',” in Analytical Philosophy, First Series, ed. R. J. Butler, pp. 211-235.

15 Again using the labels introduced in footnote 8, e ─ Sam's act ─ consists in the volition plus consequent neural and muscular events plus the consequent upward motion of Sam's arm; the last-mentioned summand is the event f. The volition was the initial stage of e, and f was its terminal stage. This is the view I defended in the article cited in footnote 8. An alternative view would treat f not as a part of e, but as a causal consequence of e, and would identify e─Sam's act─simply and entirely with the volition. I indicated the possibility of this alternative view in the article, p. 528; and Davidson may actually have adopted this view in “Agency” (cited in footnote 10); “Doing something that causes my finger to move does not cause me to move my finger; it is moving my finger” (p. 11 ). Although on this alternative, e causes f─Sam's act causes the motion of his arm─and we say that Sam causes f (by his act, e), we do not say that Sam causes e. Sam causes his arm to rise; he does not (normally) cause himself to cause his arm to rise, even on this alternative.

16 Op. cit., pp. 90f. Taylor's position is strange given his admission of the possibility that agents can be caused to act.

17 I wish to thank Professors Alvin Goldman and Dale Gottlieb for extremely helpful comments on earlier versions of this paper. Since this paper was written, similar views have been presented by Arthur Danto, C., Analytical Philosophy of Action (Cambridge University Press, 1973). pp. 63–76Google Scholar (but see his apparent rejection of certain features of this view on pp. 76f., and his earlier comments on volitions as theoretical posits, pp. 54-56, 61); O'Shaughnessy, Brian, “Trying (as the Mental 'Pineal Gland’),” journal of Philosophy, LXX (1973). 365·386Google Scholar; and McCann, Hugh, “Volition and Basic Action,” Philosophical Review, LXXXIII (1974), 451·473Google Scholar. And I have since learned of similar views developed by Sellars, Wilfrid (e.g., in “Thought and Action,” in Freedom and Determinism, ed. Lehrer, Keith (New York: Random House, 1966))Google Scholar and David M. Armstrong, “Acting and Trying,” read at the Rockefeller University in the winter of 1971·72.