Thus ‘power’ is so used in ordinary language that only the absence of past necessary conditions of an act destroy the power to perform the act. But it is also important to remember that our language is so used that, although the absence of some past necessary conditions destroys this power, the absence of others does not. When I decide at t2 not to swim at t3 and my so deciding is a sufficient condition of my not swimming at t3, then a past necessary condition of my swimming at t3 is absent (viz., my not deciding not to swim at t3), but this decision does not render me powerless to swim at t3. To maintain the contrary position, as Taylor does by accepting PP, is to employ a language which, unlike ordinary language, fails to reflect the difference between my not swimming at t3 (i) because I decided not to at t2, and (ii) because I was drugged or bound at t2.
6. Assessment of Taylor’s Position. In light of all of the considerations adduced in this paper, I would conclude that Taylor has failed to provide us with good reasons for accepting PF, PP, or P, or fatalism, or any particular “fatalistic” conclusions such as “It’s not in my power to swim at t3 (because the water is not turbulent at t4).” On the contrary, I believe I have shown that their acceptance is tantamount to a radical revision of our ordinary language which is, at best, inconvenient and, at less than its best, as in the case of Taylor who does not realize that he is merely revising our language, productive of confusion.
7. The Law of Excluded Middle. I cannot refrain from adding a comment on the views of Taylor and Cahn regarding the law of excluded middle. Taylor had thought that one might avoid his “fatalistic” conclusions by revising the traditional interpretation of the law so that a statement about a future situation is neither true or false until the time of the situation, although, presumably, it possesses one of the two truth-values then and thereafter.12 Cahn, however, has argued that even this radical step will not enable us to avoid the “fatalism” to which Taylor’s acceptance of P leads,13 and Taylor finds his arguments convincing.14 Both Taylor and Cahn talk as though the very idea of this radical step were intelligible, and I wish to indicate that it is not.
From “ ‘There is (either tenselessly or in any tense you like) a large brown cup on the table at 10 A.M. (Tuesday, July 8, 1964)’ is true,” it follows that the cup is large, that it is brown, that it is on the table, and that all of these things characterize the cup at 10 A.M. What does not follow is that the truth that the large brown cup is on the table at 10 A.M. is itself large or brown or on the table or temporally located at 10 A.M. As we in fact use our language, it makes no more sense to attribute a time to truth than it does to ascribe to it a size, a color, or a place. Once again we find that Taylor and Cahn are engaged in a radical revision of ordinary language, one which seems to make no sense and to which they have given no sense.
NOTES
* Richard Taylor, “Fatalism,” The Philosophical Review, 71 (1962): 56-66; John Turk Saunders, “Professor Taylor on Fatalism,” Analysis, 23, 1 (1962): 1-2; Brice Aune, “Fatalism and Professor Taylor,” The Philosophical Review, 71, 4 (1962): 512-519; Richard Taylor, “Fatalism and Ability,” Analysis, 23, 2 (1962): 25-27; Peter Makepeace, “Fatalism and Ability, II,” ibid., 27-29; John Turk Saunders, “Fatalism and Linguistic Reform,” ibid., 30-31; Raziel Abelson, “Taylor’s Fatal Fallacy,” The Philosophical Review, 72, 1 (1963): 93- 96; Richard Sharvy, “A Logical Error in Taylor’s ‘Fatalism’,” Analysis, 23, 4 (1963): 96; John Turk Saunders, “Fatalism and the Logic of ‘Ability’,” Analysis , 24, 1 (1963): 24; Richard Taylor, “A Note on Fatalism,” The Philosophical Review, 72, 4 (1963): 497-499: Richard Sharvy, “Tautology and Fatalism,” this JOURNAL, 61, 10 (1964): 293-295; Steven Cahn, “Fatalistic Arguments,” ibid., 295-305; Richard Taylor, “Comment,” ibid., 305-307.
1 “Comment,” p. 307.
2 Cf. “Fatalistic Arguments,” p. 301. I presume to construe Cahn’s talk of “physical” relations so that it may be broadened to cover empirical relations.
3 “Comment,” p. 306.
4 Loc. cit.
5 Compare G. J. Warnock, “Every Event Has a Cause,” in A. G. N. Flew, ed., Logic and Language, Second Series (Oxford, Basil Blackwell, 1955), pp. 95-111.
6 It is considerations of the kind here adduced that led me early in the “fatalism” controversy to comment that “the undesirable fatalistic implications” of Taylor’s arguments “are not detailed by Taylor.” Compare “Professor Taylor on Fatalism,” p. 1.
7 For example, cf. the following: “Comment,” p. 306; “Fatalistic Arguments,” pp. 299 and 303.
8 “A Note on Fatalism,” p. 498.
9 “Fatalism and Ability,” p. 26.
10 “Fatalistic Arguments,” p. 300.
11 “Fatalism and Linguistic Reform,” p. 31.
12 “Fatalism,” pp. 63-65.
13 “Fatalistic Arguments,” pp. 303-305.
14 “Comment,” p. 307.
13
FALLACIES IN TAYLOR’S “FATALISM”
CHARLES D. BROWN
RICHARD TAYLOR has presented two versions of his cunning argument on fatalism—one in Philosophical Review 1 and an expanded version in his Metaphysics;2 but they are essentially the same argument, and they are equally fallacious. Because of the wide attention this argument has already received, I will not present yet another summary of it; rather I refer the reader to Taylor’s presentations and to Steven Cahn’s defense and elaboration of Taylor’s argument.3 In spite of Cahn’s vigorous defense of the “fatalistic argument,” two smirking defects remain: (1) an equivocation on ‘necessary condition’ and (2) a confusion of the reciprocity of necessity and sufficiency with a necessary and sufficient condition. Both defects result in Taylor’s begging the question of fatalism at the outset of his argument.
In his second and third presuppositions, Taylor presents what we may for the moment take to be a standard interpretation of logically sufficient conditions and logically necessary conditions. “The ingestion of cyanide ensures death” can be represented by ‘C ⊃ D,’ in which case, “C” is the sufficient condition (as logical antecedent) of “D.” “Oxygen is essential for life” can be represented as ‘L ⊃ O,’ in which case, “O” is the necessary condition (as logical consequent) of “L.” There has been some confusion over Taylor’s additional statement that these antecedents and consequents may be “logically unconnected,” 4 or “logically unrelated.”5 Raziel Abelson, for example, observes the seeming contradiction of such logically unrelated logical conditions, and he attributes the conflict to Taylor’s confusion of logical relations with causal relations.6 But Taylor denies having introduced causation into his argument. The only plausible alternative interpretation of the statement that antecedents and consequents are not logically connected (or related) seems to be that the assertions “L ⊃ O” and “C ⊃ D” are not tautologies—that the consequents are not analytic out of the antecedents. As Taylor has set up his argument, however, I doubt that even so mild a claim as this one can be successfully maintained.
THE JOURNAL OF PHILOSOPHY, VOL. 62, NO. 13, 1965.
The fourth presupposition, Taylor asserts, is just a summation of the second and third presuppositions. It shows the reciprocity of necessary conditions and sufficient conditions in logical implications. That is, if the ingestion of cyanide is sufficient to produce death, then death is a necessary consequent of the ingestion of cyanide; and if the presence of life is sufficient evidence for asserting the presence of oxygen, then the presence of oxygen is a necessary consequent of the presence of life. But it is strange, indeed, to say that the presence of oxygen is a necessary consequent of the presence of life. We would more properly say that its presence is a necessary condition for life. Then, should we not also say that death is a necessary condition for the ingestion of cyanide? This, clearly, is absurd. This is the first passage in which Taylor obviously kicks dust into our eyes by leading us from the conditional mode of thinking about logical antecedents and consequents into thinking about necessary preconditions; but he pulle
d on his dust-kicking boots in the second and third presuppositions. The damage is done by the seemingly innocuous “prepositional shift,” from the logically standard ‘necessary [sufficient] condition of ’ to ‘necessary [sufficient] condition for.’ The word ‘for’ is used also in the second and third presuppositions, but the implications of its use are not fully evident until the reciprocity of necessity and sufficiency is introduced. Logically, the consequent of an implication is the necessary condition of (i.e., proceeding from) the antecedent. To replace the ‘of’ with ‘for’ is not only to change the direction of the “proceeding,” but also to invoke nonlogical necessary conditions. If Taylor insists on his grammatical freedom to use ‘for’ in presuppositions two and three (though I see no legitimate reason for his doing so), he must still restrict his argument to a univocal interpretation. But he does not so restrict himself; and that he freely employs a shift of prepositional direction is obvious in the following analysis.
The difference between the logical “condition of ” and the nonlogical “condition for” becomes even more obtrusive in the fifth presupposition, wherein Taylor states that we cannot perform any act in the absence of any condition (i.e., precondition) necessary for the performance of that act. He says, “This is no law of logic, and in fact cannot be expressed even in the contemporary modal logics, but it is manifestly true.”7 He persists, however, in using the words ‘condition necessary for’; but since the fifth presupposition by his own assertion has nothing to do with logic or with logical relations (since it cannot even be logically expressed), he must be referring to some nonlogical necessary condition. Nonlogical (practical?) necessity is normally construed in terms of causation; but Taylor states that “Our problem has been formulated without any reference whatever to causation.”8 I suspect that, in view of the prepositional shift, Taylor should make only the more modest claim that he avoided using the words ‘clause’ and ‘causation.’ (I fail to see how ‘necessary condition for’ can be legitimately interpreted other than causally.)
But we can even do without the causal interpretation (i.e., we can avoid using ‘cause’ and ‘causation’) and still defeat the argument within the scope of Taylor’s first six presuppositions. The fifth presupposition gives us practical grounds for distinguishing between antecedent necessary conditions and consequential necessary conditions. When Taylor moves from logical necessity to practical necessity, ‘antecedent’ and ‘consequent’ refer to temporal priority and posteriority, respectively, rather than to sufficiency and necessity, respectively.9 This shift of reference is entailed by the prepositional shift. In logically formulated statements, priority and posteriority are accounted for by the component statements. Thus, ‘C ⊃ D’ could be rendered, “If one takes cyanide then he will soon die (i.e., death will be a practical necessary consequent).” And ‘L ⊃ O’ can be rendered, “If there is life then there must have been oxygen (i.e., oxygen was a practical necessary antecedent).” The dual “necessity” as regards the consequent of each of these implications should be obvious.
Taylor invokes logical necessary condition to defend a supposed indifference as between practical antecedent necessary conditions and practical consequential necessary conditions. His denial of any connection between logical necessary conditions and practical necessary conditions in presupposition five, however, precludes this defense.10 In both of the situations hypostatized by Taylor, the occurrence of the naval battle is, as the implicative consequent, the logical necessary condition. But the position of ‘P’ in ‘S ⊃ P’ and ‘O ⊃ P’ is quite irrelevant to Taylor’s conclusion that “P” necessitates “O.’ Where, then, does the argument derive its apparent force? The occurrence of the naval battle in the first situation is, in addition to being the logical necessary condition, the practical antecedent necessary condition, whereas in the second situation it is, additionally, the practical consequential necessary condition. The “force” of the argument depends merely upon Taylor’s (or his readers’) transferring the practical necessity, per se, along with the logical necessity from the first situation to the second. That this is an illicit transferal is evident from the fact that they are two different kinds of practical necessity—a distinction quite in accordance with Taylor’s sixth presupposition, if we can also accept his fifth. But the overriding difficulty is Taylor’s apparent attempt to assimilate practical necessity to logical necessity via the prepositional shift, and the resultant referential shift, smuggled in by his use of ‘for.’ The same confusion is propagated by Cahn in his “firming up” of Taylor’s argument and also in his argument that the rejection of the law of the excluded middle leads to the impossibility of any action since, for both, he adopts Taylor’s presuppositions.11
Nor is this all. In the very first paragraph of his statement of the first situation,12 Taylor makes the reading of the headline (S) a necessary and sufficient condition of the occurrence of the battle (P) and, conversely, the occurrence of the battle a necessary and sufficient condition of the reading of the headline. That is, there is no possibility of encountering the logical necessary condition without being able to deduce the logical sufficient condition. But this relationship is nowhere stated in his presuppositions. The relevant data are necessary condition, sufficient condition, and the reciprocity of necessity and sufficiency—not the conjunction of necessity and sufficiency.13 Thus, Taylor further presupposes determinism, itself, in constructing his first situation; and, in identifying the forms of the arguments of the two situations, he has merely transferred this seventh presupposition to his second situation. The controversy reduces to this: if one presupposes determinism, one must consistently (and determinedly) try to defend it; but one need only be consistent to do so.14
NOTES
1 Richard Taylor, “Fatalism,” Philosophical Review, 71, 1 (1962): 56-66.
2 Richard Taylor, Metaphysics (Englewood Cliffs, N.J.: Prentice-Hall, 1963), ch. 5.
3 Steven Cahn, “Fatalistic Arguments,” this JOURNAL, 61, 10 (May 7, 1964): 295-305.
4 Metaphysics, p. 58.
5 “Fatalism,” pp. 57-58.
6 Raziel Abelson, “Taylor’s Fatal Fallacy,” Philosophical Review, 72, 1 (1963): 94-95.
7 Metaphysics, p. 58.
8 Ibid., p. 64. Also, “Fatalism,” p. 63.
9 I do not here contravene Taylor’s sixth presupposition—if he does not. If he suspects any such contravention then, since I am merely elucidating presupposition five, it will be up to him to reconcile the implications of his fifth and sixth presuppositions.
10 Metaphysics, pp. 63-64. “Fatalism,” pp. 61-63.
11 Cahn, op. cit., pp. 304-305.
12 Metaphysics, p. 59. “Fatalism,” p. 59.
13 Bruce Aune, in “Fatalism and Professor Taylor,” Philosophical Review, 71 (1962): 515, contends that Taylor introduces necessity and sufficiency in presupposition four; but Taylor is immune to this charge. That Taylor does beg the question of determinism in setting up the first situation, however, is obvious.
14 Cahn, op. cit., makes several statements in Part I of his article which are clearly inconsistent. In the first paragraph of his section on The Crucial Assumptions (p. 296) he reiterates Taylor’s notion of nonlogical necessity, e.g., oxygen is necessary for life. In the first paragraph of Fatalism and Causation (301) he states that neither part of a nonlogical proposition, e.g., oxygen is necessary for life, “is logically necessary or sufficient for the other, nor is either the cause of the other.” Cahn previously states (296) that his argument to the fatalistic conclusion from nonlogical necessity “is made possible by the fact that, for any two events or states of affairs A and B, if A is a necessary condition of B, then B is a sufficient condition of A, and vice versa, and it matters not at all which of them occurs first in time.” Now, surely, Cahn cannot deny that this latter is a statement of the reciprocity of logical necessity and sufficiency. (He even here avoids the surreptitious “for,” which only makes his use of “for” elsewhere more obviously a propositional
shift.) (1) Since the law of the excluded middle is presupposed by both Taylor and Cahn, the reciprocity of logical necessity and sufficiency either is or else it is not the decisive relationship by which fatalism is “established”; it cannot both be and not be that relationship, as these passages cited in Cahn’s article indicate. (2) If it is the essential relationship then, in view of the prepositional shift, we have equivocation on the one hand and such absurdities as death being a necessary condition for the ingestion of cyanide on the other hand. (3) If it is not the essential relationship then we have no reciprocity and, hence, no fatalistic argument.
PART II
THE ESSAY
14
RENEWING THE FATALIST CONVERSATION
MAUREEN ECKERT
AFTER THE publication in 1967 of Steven M. Cahn’s Fate, Logic, and Time,1 the philosophical debate on fatalism that had raged in the leading professional journals for five years faded from view. Nearly two decades later, however, developments in logic enabled a talented, ambitious undergraduate named David Foster Wallace to revisit the issue and offer a new analysis of its central argument.