¹ One very obvious objection to the kind of theodicy which is here being contemplated is that there is a good deal of suffering which does not give rise to such desirable responses. I have tried to cope with this objection in an earlier paper, ‘An Examination of the “Soul-Making” Theodicy’, American Philosophical Quarterly, 7, 1970, pp. 124–130.Google Scholar

² For simplicity's sake, I shall henceforth simply say ‘theodicists’ or ‘the theodicist’ when I wish to refer to theodicists of the contemplated kind.

³ McCloskey, H. J., ‘God and Evil’, The Philosophical Quarterly, 10, 1960, pp. 108–109CrossRefGoogle Scholar. This article is reprinted in Pike, Nelson, ed., God and Evil (Engle-wood Cliffs, N.J., 1964).Google Scholar

⁴ Penelhum, Terence, Religion and Rationality (New York, 1971), p. 240.Google Scholar

⁵ For a different reply from the one I am about to give, see ‘An Examination of the “Soul-Making” Theodicy’, op. cit., pp. 121–123. The earlier reply seems to me less satisfactory than the one given above.

⁶ John Rawls has stated a similar principle in a different context: ‘…a person is required to evidence his belief… when he proposes to take action on the basis of it which substantially interferes with the interests of other persons’. This occurs in ‘Outline of a Decision Procedure for Ethics’, Thomson, Judith J. and Dworkin, Gerald, eds., Ethics (New York, 1968), p. 68.Google Scholar

⁷ ‘Rational’ here should not be understood to entail ‘prudent’ or ‘not contrary to one's non-epistemic interests’. Rather ‘rational’ in this context should be taken to mean the same as ‘not epistemically unwarranted’. This leaves open the possibility that a man could both truly believe that it would be imprudent for hun not to believe in God and yet hold God's existence as an item of faith.

⁸ On the traditional view of basic propositions, some are basic only relative to one individual (e.g. ‘It now looks to Dore as though he sees a tree outside his window’) while others are basic for every rational man (e.g. ‘It is prima fade wrong to torture people’). The believer in God's omniscience may wish to claim that every true proposition is basic relative to God. If this claim were correct, then the first of the two propositions just cited would be basic relative to God as well as relative to Dore.

⁹ I claim only that this is a necessary, not a sufficient, condition. Gettier-type counter-examples necessitate a more elaborate statement of conditions which are both necessary and sufficient for knowledge. My present purpose does not require a full analysis of knowledge.

¹⁰ This is only necessary, not sufficient, for adequate evidence. For suppose that it would be irrational for S not to believe p at t because p is basic relative to S at t. Then a fortiori it is true of just any proposition, ø, that it would be irrational for S to believe ø at t and not to believe p at t. But this fact does not entitle us to say that S has evidence for p at t.

¹¹ Suppose that it would be irrational for S not to believe p at t. Then it is plausible that no matter how many non-equivalent disjunctive propositions containing p as a disjunct we might specify, it would be irrational for S not to believe each one of them at t. Now at this point someone may wish to argue as follows: ‘Suppose that it would be irrational for S not to believe these disjunctive propositions at t only because it would be irrational for S not to believe p at t. (Suppose that it would not be irrational for S not to believe any of the disjuncts other than p which are contained in these propositions.) Then it would beirrational for S to believe any of these propositions at t and not to believe p at t. Hence there would be an indefinitely large number of non-equivalent propositions which would be such that (a) it would be irrational for S not to believe each of them at t and (b) it would be irrational for S to believe any of them and not to believe p.’ But is (b) really true of these propositions? Would S necessarily be irrational were he to believe them without believing p, i.e. would he be irrational were he to believe them on faith? It is not crucial that these questions have a negative answer. For we can elaborate on my thesis about adequate evidence in such a way as to show that none of the envisaged disjunctive propositions entitles us to say that S has adequate evidence for p and, hence, that none is part of the kind of series being contemplated above: S has adequate evidence for p at t only if (a) it would be irrational for S not to believe (non-equivalent) q at t; (b) it would be irrational for S to believe q at t and not to believe p at t; and (c) it is not the case that (b) is true in virtue of the fact that it would be irrational for S not to believe q at t only because it would be irrational for S not to believe p at t.

¹² It is sometimes said that even if p is non-basic, S does not need evidence for p, in my sense of ‘evidence’, in order to be entitled to claim to know that it is true, since S is entitled to claim to know that p is true if p has withstood skilful attempts to falsify it. But what about the proposition (call it ‘P’) that p has withstood those attempts? Surely, barring any other ground for being entitled to claim to know (non-basic) p, S would not be entitled to claim to know that p is true unless he were entitled to claim to know that P is true. But here we need to answer the question what entitles S to claim to know that P is true. If the answer is that S is entitled to claim to know that P itself has withstood skilful attempts to falsify it, we are started once again upon an unacceptable regress.

Another theory has it that S is entitled to claim to know that p is true if most people in his society are disposed to assent to it. (It should be added that S does not have strong evidence against p.) But this does not provide an alternative to the theory of basic propositions. For surely, barring any other ground for believing (non-basic) p, S would not be entitled to claim to know that p is true unless he were entitled to claim to know that most people in his society are disposed to assent to it (etc.). And a full answer to the question how S can be entitled to claim to know this latter proposition would presumably have to make reference to basic propositions.

¹³ Of course it might be claimed that ‘the value of the ends which the theodicist says are served by suffering outweighs the suffering’ does not, since it is a value utterance, express a proposition (make a true or false statement) and a fortiori does not express a proposition which might be basic relative to God. But this non-cognitivist position flies in the face of perfectly ordinary ways of speaking. ‘It is true that it is better to give than to receive’ is no more linguistically odd than ‘It is true that the earth is not flat’.

¹⁴ Still, it is worth noting that these considerations, even if true, do not give us a reason for holding that the theodicist does not believe his theodicy. Surely it is far from clear that the theodicist does not really believe some propositions on faith—that he only claims to do so in moments of confused philosophical reflection.

¹⁵ Cases in which (1) S has adequate evidence for p and not-p is basic relative to S or vice versa and (2) both p and not-p are basic relative to S, would be cases in which it would be irrational for S not to believe both p and not-p. But since it would be irrational for S to believe p and not-p, it cannot be true that it would be irrational for S not to do so: S is irrational in not believing some given propositions only if he would not be irrational were he to believe them. Hence the envisaged cases cannot arise.

¹⁶ Since the cases outlined in the preceding footnote cannot arise.

¹⁷ Here I would have what William James calls a momentous option.

¹⁸ I do not wish to claim that p is basic relative to S only if (1) and (2) are true. Presumably the defender of basic propositions will want to say that when (1) and (2) are true, they are basic relative to S. Hence if it were true that, no matter what proposition p may be, p is basic relative to S only if (1) and (2) are true, then the defender of basic propositions would be committed to holding that when (1) is basic relative to S, it seems to S upon reflection that (1) is basic; and so on. And ‘It seems to S upon reflection that it is basic that it seems to S upon reflection that p is basic’ (or, worse still, ‘It seems to S upon reflection that it is basic that it seems to S upon reflection that it is basic that it seems to S upon reflection that p is basic’, etc.) only doubtfully refer to a state of mind which S could get himself into. Anyway, it is preposterous that an infinite number of such propositions is basic relative to S.

A corresponding regress with respect to (2) is impossible. For the proposition (call it ‘ø’) that S does not know of anything such that it would be irrational for him not to take it as adequate evidence for not-p is really the same as the proposition (call it ‘ø′’) that S does not know of anything such that it would be irrational for him not to take it as adequate evidence for not-ø. Necessarily, if ø′ is true then ø is true and necessarily if ø′ is false then so is ø. It might be thought that (a) S could know that it has very frequently been true in the past that when he knew of no adequate evidence for a proposition, not-q, then it subsequently developed that he had such evidence all along and hence that (b) S could know of adequate evidence for not-ø even though he knew of no adequate evidence for not-p. But in fact S's knowing about his past oversights is ipso facto adequate evidence which S has for not-p. Or, at any rate, the evidence is adequate for not-p if it is adequate for not-ø. Since p is basic relative to S only if (2) is true (see footnote 15), it is fortunate for the defender of basic propositions that the contemplated regress cannot occur.

¹⁹ I have not claimed that if it seems to S that it would be irrational for anybody not to believe p whether or not he has evidence for it, then, if condition (2) is satisfied, p is basic relative to anybody. The principle of non-contradiction does not guarantee that if not-Q is basic relative to some, then anybody who believes its contradictory believes what is false. For it is not a necessary condition of a proposition's being basic relative to S that it is true—only that it would be irrational for S not to believe it.

²⁰ The denial of the proposition that the earthly death of a person does not involve his destruction does not appear to be basic, though it may be that there are grounds, e.g. in the form of problems about the identity of a resurrected person with his past self, for rejecting this proposition. Still, these grounds are surely not sufficiently strong and obvious to render it incredible that the theodicist believes on faith that people survive their earthly death.

²¹ This is an enlarged version of a paper which was read at the May, 1971 meeting of the Western Division of the American Philosophical Association.