by Peter Geach

(The Philosophical Review LXIX (1960) pp. 221-225)
(Some revisions added by the author included in The Nature of Human Action ed. Myles Brand, Scott, Foresman, and Co. 1970)

The statement that an act x was voluntary, or intentional, or done with intent, or the like, on the part of an agent A has often been analyzed as a causal statement that x was initiated by some act of A's setting himself to do x, or an act of intending to do x, or the like. Likely there has been a reaction against this type of analysis; it has been held (in my opinion, quite rightly) that the attempt to identify and characterize these supposed acts of bare will always run into insuperable difficulties. To avoid such difficulties, some Oxford philosophers, whom I shall call Ascriptivists, have resorted to denying that to call an act voluntary, intentional, and so forth, is any sort of causal statement, or indeed any statement at all. In this note I shall try to expound and to refute Ascriptivism.

Ascriptivists hold that to say an action x was voluntary on the part of an agent A is not to describe the act x as caused in a certain way, but to ascribe it to A, to hold A responsible for it. Now holding a man responsible is a moral or quasi-moral attitude; and so, Ascriptivists argue, there is no question here of truth or falsehood, any more than there is for moral judgments. If B agrees or disagrees with C's ascription of an act to A, B is himself taking up a quasi-moral attitude toward A. Facts may support or go against such a quasi-moral attitude, but can never force us to adopt it. Further, the Ascriptivists would say, there is no risk of an antinomy, because ascription of an act to an agent can never conflict with a scientific account of how the act came about; for the scientific account is descriptive, and descriptive language is in a quite different logical realm from ascriptive language. Though it has not had the worldwide popularity of the distinction between descriptive and prescriptive language, the Ascriptive theory has had quite a vogue, as is very natural in the present climate of opinion.

Now as regards hundreds of our voluntary or intentional acts, it would in fact be absurdly solemn, not to say melodramatic, to talk of imputation and exoneration and excuse, or for that matter of praise and reward. Ascribing an action to an agent just does not in general mean taking up a quasi-legal or quasi-moral attitude, and only a bad choice of examples could make one think otherwise. (As Wittgenstein said, when put on an unbalanced diet of examples philosophy suffers from deficiency diseases.)

Again, even when imputation and blame are in question, they can yet be distinguished from the judgment that so-and-so was a voluntary act. There are savage communities where even involuntary homicide carries the death penalty. In one such community, the story goes, a man fell off a coconut palm and broke a bystander's neck; the dead man's brother demanded blood for blood. With Solomonic wisdom the chief ordered the culprit to stand under the palm tree and said to the avenger of blood, "Now you climb up and fall off and break his neck!" This suggestion proved unwelcome and the culprit went free. Though the vengeful brother may still have thought the culprit ought to have been punished, his reaction to the suggested method of execution showed that he knew as well as we do the difference between falling-off-a-tree-and-breaking-someone-else's-neck voluntarily or intentionally and just having it happen to you. To be sure, on his moral code the difference did not matter - his brother's death was still imputable to the man who fell on him - but this does not show that he had no notion of voluntariness, or even a different one from ours.

I said that Ascriptivism naturally thrives in the present climate of opinion; it is in fact constructed on a pattern common to a number of modern philosophical theories. Thus there is a theory that to say "what the policeman said is true" is not to describe or characterize what the policeman said but to corroborate it; and a theory that to say "it is bad to get drunk" is not to decribe or characterize drunkenness but to condemn it. It is really quite easy to devise theories on this pattern; here is a new one that has occurred to me. "To call a man happy is not to characterize or describe his condition; macarizing a man" (that is, calling him happy: the words "macarize" and "macarism" are in the O.E.D.) "is a special nondescriptive use of language. If we consider such typical examples of macarism as the Beatitudes, or again such proverbial expression as 'happy is the bride that the sun shines on; happy are the dead that the rain rains on,' we can surely see that these sentences are not used to convey propositions. How disconcerting and inappropriate was the reply. 'Yes, that's true,' that a freind of mine got who cited 'happy are the dead that the rain rains on' at a funeral on a rainy day! The great error of the Utilitarians was to suppose that 'the greatest happiness of the greatest number' was a descriptive characterization of a state of affairs that one could aim at; but in fact the term 'happiness' is not a descriptive term: to speak of people's happiness is to macarize them, not to describe their state. Of course 'happy' has a secondary descriptive force; in a society where the rich were generally macarized, 'happy' would come to connote wealth; and then someone whose own standards of macarism were different from those current in his society might use 'happy,' in scare-quotes so to say, to mean 'what most people count happy, that is rich'...." There you are; I make a free gift of the idea to anybody who likes it.

There is a radical flaw in this whole pattern of philosophizing. What is being attempted in each case is to account for the use of a term "P" concerning a thing as being a performance of some other nature than describing the thing. But what is regularly ignored is the distinction between calling a thing "P" and predicating "P" of a thing. A term "P" may be predicated of a thing in an if or then clause, or in a clause of a disjunctive proposition, without the thing's being thereby called "P." To say, "If the policeman's statement is true, the motorist touched 60 mph" is not to call the policeman's statement true; to say, "If gambling is bad, inviting people to gamble is bad" is to call either gambling or invitations to gamble "bad." Now the theories of nondescriptive performances regularly take into account only the use of a term "P" to call something "P"; the corroboration theory of truth, for example, considers only the use of "true" to call a statement true, and the condemnation theory of the term "bad" considers only the way it is used to call something bad; predications of "true" and "bad" in if or then clauses, or in clauses of disjunction, are just ignored. One could not write off such uses of the terms, as calling for a different explanation from their use to call things true or bad; for that would mean that arguments of the pattern "if x is true (if w is bad), then p; but x is true (w is bad); ergo p" contained a fallacy of equivocation, whereas they are in fact clearly valid.

This whole subject is obscured by a centuries-old confusion over predication embodied in such phrases as "a predicate is asserted of a subject." Frege demonstrated the need to make an absolute distinction between predication and assertion; here as elsewhere people have not learned from his work as much as they should. In order that the use of a sentence in which "P" is predicated of a thing may count as an act of calling the thing "P," the sentence must be used assertively; and this is something quite distinct from the predication, for, as we have remarked, "P" may still be predicated of the thing even in a sentence used non-assertively as a clause within another sentence. Hence, calling a thing "P" has to be explained in terms of predicating "P" of the thing, not the other way around. For example, condemning a thing by calling it "bad" has to be explained through the more general notion of predicating "bad" of a thing, and such predicating may be done without any condemnation; for example, even if I utter with full conviction the sentence, "If gambling is bad, inviting people to gamble is bad," I do not thereby condemn either gambling or invitations to gamble, though I do predicate "bad" of these kinds of act. It is therefore hopeless to try to explain the use of the term "bad" in terms of nondescriptive acts of condemnation; and, I maintain, by parity of reasoning it is hopeless to try to explain the use of the terms "done on purpose," "intentional," or the like, in terms of nondescriptive acts of ascription or imputation.

With this I shall dismiss Ascriptivism; I adopt instead the natural view that to ascribe an act to an agent is a causal description of the act. Such statements are indeed paradigm cases of causal statements: cf. the connection in Greek between αιτια ("cause") and αιτιος ("responsible"). Let us recollect the definition of will given by Hume: "the internal impression we feel and are conscious of when we knowingly give rise to any new motion of our body or new perception of our mind." Having offered this definition of will, Hume concentrates on the supposed "internal impression" and deals with the causal relation between this and the "new motion" or "new perception" on the same lines as other causal relations between successive events. Like a conjurer, Hume diverts our attention; he makes us forget the words "knowingly give rise to," which are indispensible if his definition is to have the least plausibility. If Hume had begun by saying, "There is a peculiar, characteristic, internal impression which we are sometimes aware arises in us before a new perception or new bodily motion; we call this volition or will," then his account would have a fishy look from the outset. To say we knowingly give rise to a motion of mind and body is already to introduce the whole notion of the voluntary; an "internal impression" need not be brought into the account, and is anyhow, I believe, a myth. But without the "internal impression" Hume's account of causality cannot be fitted to voluntary causality; without it we no longer have two sorts of event occurring in succession, but only, on each occasion, one event to which "we knowingly give rise" - words that express a non-Humean sort of causality.

For an adequate account of voluntary causality, however, we should need an adequate account of causality in general; and I am far from thinking that I can supply one. To develop one properly would require a synoptic view of the methods and results of the strict scientific disciplines - a labor of Hercules that far exceeds my powers; and it would take a better man than I am to see far through the dust that Hume has raised. All I have tried to do here is to make it seem worthwhile to investigate non-Humean ideas of causality in analyzing the voluntary, instead of desparately denying, as Ascriptivists do, that voluntariness is a causal concept.