NOTES
1 That may suggest an ‘anything goes’ attitude toward philosophical questions that I neither hold nor approve of. I would insist that when debate
over a philosophical question - say, the question whether I have hands - ends in deadlock, it does not follow that there is no truth of the matter;
or that we don’t know the truth of the matter; or that we ought to suspend judgement; or that we have no reason for thinking one thing rather than
the other.
2 See David Lewis, On the Plurality of Worlds (Oxford, Blackwell, 1986),86-92.
3 It seems that many find the second question too hard. Seldom does an analytic philosopher of religion defend the eternal torment of the
damned. Among those who discuss the question at all. Richard Swinburne, ‘A Theodicy of Heaven and Hell’ in The Existence and Nature 'of
God, ed. by Alfred Freddoso (Notre Dame, Indiana, Notre Dame Press, 1983) is typical: he offers no ‘theodicy of Hell’ but only a reason why the
damned may not enjoy the delights of Heaven. But Peter Geach rises to the challenge: in Providence and Evil (Cambridge, Cambridge
University Press, 1977), he claims that ‘someone confronted with the damned would find it impossible to wish that things so evil should be
happy’ (139). Grant that they shouldn’t be happy; but why wouldn’t it be best to destroy them? Wouldn’t ‘the work of the Divine Artist… . be
permanently marred if the surd or absurd element of sin were a permanent element of it’? (140) In reply Geach speculates that time forks, Hell in
one fork and Heaven in the other; so that the blessed in Heaven cannot say that Hell was, or is, or will be. But why does this leave the work of
the Artist - the entire work - unmarred?
4 George Schlesinger, ‘The Problem of Evil and the Problem of Suffering’, American Philosophical Quarterly 1 (1964), 244-247; Peter Forrest,
‘The Problem of Evil: Two Neglected Defences’, Sophia 29 (1981), 49-54.
16
5 Alvin Plantinga,. God, Freedom, and Evil (New York, Harper & Row, 1974), 10, 27-29; Plantinga, ‘Self-Profile’ in Alvin Plantinga, ed. by James
Tomberlin and Peter van Inwagen (Dordrecht, Reidel, 1985),35,42.
6 That follows from my meta-ethical position, subjectivism with bells and whistles. See David Lewis, ‘Dispositional Theories of Value’, Aristoteliall
Society Supplementary Volume 63 (1989), 113-137 (reprinted in this volume as Chapter 7). It’s necessary to consider the value judgement taken
in an ‘unrigidified’ form (see 132-133) but there’s nothing wrong with that.
7 See Sylvain Bromberger, ‘An Approach to Explanation’ in Analytic Philosoplhy: Second Series, ed. by R. J. Dutler (Oxford, Blackwell, 1965) on
the distinction between kinds of predicaments.
8 Adams, ‘PIantinga on the Problem of Evil’ in Alvin Plantinga, 240.
9 God, Freedom, alld Evil, 30
10 Steven Doer, ‘The Irrelevance of the Free-Will Defence’, Analysis 38 (1978), 110112, suggests that the question why evil sometimes causes
hann belongs to the department of theodicy that is concerned with the problem of natural evil. If so, the playpen problem falls outside our present
topic. However, I note that in that case, we must dismiss the hypothesis that natural evil is the evil-doing of Satan and his cohorts. (See God,
Freedom, and Evil, 58-59.) For why does the evil-doing of Satan and his cohorts cause harm? God could have put Satan and his cohorts in the
playpen along with Stalin.
11 The closure principle is a generalization of the ‘Rule Beta’ that plays a leading role in Peter van Inwagen’s defence of incompatibilism in All
Essay 011 Free Will (Oxford, Oxford University Press, 1983);‘it first appears on page 94. The closure principle says that the logic of , Unfree’ is a
‘normal’ modal logic; see Brian Chellas, Modal Logic: All Introduction (Cambridge, Cambridge University Press, 1980), 114115. We can see from
Chellas’s Theorem 4.3(4) that the closure principle is equivalent, inter alia, to this combination of four principles: RE: if’A iff B’is valid, so is
‘Unfree A iff Unfree B’, N: ‘Unfree T is valid, where T is an arbitrary tautology, M: ‘Unfree (A & B)’ implies ‘Unfree A and Unfree B’, and C: ‘Unfree
A and Unfree B’ implies ‘Unfree (A & B)’. The compatibilist must therefore challenge one of the four, most likely C: and Michael Slote has done
so in ‘Selective Necessity and the Free-Will Problem’, JOllmal of Philosophy 79 (1982), 5-24. 12 Plantinga, God alld Other Minds (Ithaca, Cornell
University Press, 1967),135. But later he concedes that this was too short a way with compatibilism: ‘Self-Profile’, 45-47, and ‘Reply to Robert M.
Adams’, 371-372, both in Alvin Plantinga. My complaint here applies only to his earlier view.
13 Remember how much the laws of nature must be ‘fine-tuned’ before they even permit life. See John Leslie, Universes (London, Routledge,
1989), 4–6, 27-65.
14 John Bishop, ‘Compatibilism and the Free Will Defence’, Australasian Journal of Philosophy 71 (1993), 104-120. Note that Bishop’s theodicy
offers another solution to the playpen problem - one that is not available within free-will theology narrowly construed.
15 Or some vague approximation to an expected value. I don’t suppose an incompatibilist will thiilk that free choices have well-defined
probabilities; but neither will he want to abandon altogether the idea that some free choices are more likely than others, and so contribute more
weightily to the prospect of a certain gamble on freedom.
16 Maybe God is a satisficer; maybe it is not part of His benevolence, rightly understood, that He must actualize the very best of His options.
See Robert M. Adams, ‘Must God Create the Best?’ Philosophical Review 81 (1972), 317-332, reprinted in Adams, The Virtue of Faith (Oxford,
Oxford University Press, 1987). The more of a satisficer God is, of course, the easier it will be for Him not to regret a gamble that turns out badly.
17 See Robert M. Adams, ‘Middle Knowledge and the Problem of Evil,’ American Philosophical Quarterly 14 (1977), 109-117, reprinted in The
Virtue of Faith; Anthony Kenny, The God of the Philosophers (Oxford, Oxford University Press, 1979), 61-71; and Plantinga, ‘Self-Profile’, 48-50.
18 This is not the fallacy of counterfactual transitivity. Instead, it repeatedly invokes the inference
If it were that A, then it would be that B;
if it were that A & B, then it would be that C;
therefore if it were that A, then it would be that B & C which is uncontroversially valid.
19 Let O(W) be the option that God strongly actualizes at world W, assuming that W is a world where God exists. We say that God can actualize
world W if the following counterfactual is true (here at our actual world): if it were that O(W), then W would be actualized. If, in addition, O(W)
holds at no world except W and hence strictly implies that W is actualized, we say that God can strongly actualize W; if not, we say that God can
weakly actualize W. These definitions differ from Plantinga’s, but they are equivalent; see the statement and proof of ‘Lewis’s Lemma’, in his
‘Self-Profile’, 50-51.
Assume that God is able to leave something unsettled. What God leaves unsettled comes out differently at different possible worlds, but not
because of any difference in what God does. That is: God has an option 0 that holds at two different worlds V and W, so that O(V) = O(W) = O.
Then one or both of these two worlds is a world that He cannot actualize, either weakly or strongly. Else we would have two true counterfactuals
with the same antecedent and conflicting consequents: if it were that 0, V would be actual; if it were that 0, W would be actual; but V and W
17
cannot both be actual. That would mean that 0 was not an entertainable supposition, contrary to the assumption that it is one of God’s options.
Thus we refute ‘Leibniz’s Lapse’, the thesis that for any world (or any world in which God exists), God can actualize that world.
(Susanna Siegel has observed that the ‘lapse’ may be badly named. For Leibniz could invoke his principle of sufficient reason to argue that God
is unable to leave anything unsettled. In that case it would be no lapse for Leibniz to conclude that God is, after all, able to actualize any world.)
Note that this refutation of Leibniz’s lapse does not require us to say anything specific about what it is that God can leave unsettled, and why He
might want to leave it unsettled. But one case to keep in mind is the case that He might leave a creature’s action unsettled, because He values
incompatibilist freedom.
Note also that the refutation does not presuppose Molinism. If the difference between worlds V and W concerns the action of a free creature,
Molinism says that one of the conflicting counterfactuals is true and anti-Molinism says that neither is true; but what matters for the refutation is
just that they can’t both be true. Call a world utopian if it contains significantly free creatures, none of whom ever freely do evil. Once we know
that there are some worlds that God cannot actualize, we are in a position to speculate that every utopian world is one of these unactualizable
worlds. That is a weak version of the hypothesis of depravity. Plantinga’s own version of the hypothesis, in The Nature of Necessity, 186-189, is
stronger by a quantifier shift. His hypothesis is that every possible creature P suffers from trans-world depravity: that is to say (almost), there is
no world God can actualize where P exists and is significantly free and never freely does evil. (I omit another unimportant strengthening, and I
omit Plantinga’ s use of essences as surrogates for possibilia.)
20 Compare C. B. Martin’s idea of a finkish disposition: as it might be, the solubility of something that would instantly cease to be soluble if ever it
were put into solvent. Martin discussed finkish dispositions years ago in Sydney, and in ‘Powers and Conditionals’, presented at the University of
North Carolina in 1968. 1 agree with Martin that finkish dispositions are possible, and that they refute a simple conditional analysis of
dispositions. What I deem absurd is not finkishness per se, but finkishness applied to God’s powers.
21 See John Bigelow, The Reality if Numbers (Oxford, Oxford University Press, 1988), 133; and ‘Real Possibilities’, Philosophical Studies 53
(1988), 38, where supervenience on what things exist turns into supervenience on what things exist and holV tlley are arranged (i.e. arranged in
patterns of instantiation). Bigelow’s principle is a weakened form of C. B. Martin’s principle that truths require truthmakers; see D. M. Armstrong,
‘CO B. Martin, Counterfactuals, Causality, and Conditionals’ in Cause, Mind, and Reality: Essays Honoring C. B. Martin, ed. by John Heil
(Dordrecht, Kluwer, 1989).
22 A fortiori, for what it’s worth, they violate the analysis I advanced in Counterfactuals (Oxford, Blackwell, 1973). For on my analysis, the truth of
counterfactuals is supervenient on being.
23 A charge of deception gives us a third solution to the playpen problem. Again I reply that outright deception is not required to create a
playpen.
24 Cited in Plantinga, God, Freedom, and Evil, 27.
25 I myself would say that suppositions contrary to essence are entertainable because essence is a flexible matter; it’s no contradiction that a
being is, loosely speaking, God but is not, strictly speaking, God because of one lapse from omniscience; just as it is no contradiction that a
glass is, loosely speaking, empty but is not, strictly speaking, empty because of one remaining drop of beer. Not essentialism per se, but only an
especially rigid version of essentialism stands in the way of supposing counterfactually that Descartes lacks his essence, or that God lacks His.
Nor need we explain this in terms of my theory of counterparts; the same flexibility is available on rival approaches to modal metaphysics, except
for one approach that lacks adherents. See my On the Plurality if Worlds, Ch. 4.
26 an the analysis I offer in Counterfactuals - which, however must remain bracketed so long as we suspend disbelief about Molinism - the hamand-
eggs principle amounts to assuming, roughly, that any closest B-world to any closest A-world to ours must be an A&B-world. The analogy of
similarity distance to spatial distance quickly reveals counterexamples. Then why is the ham-and-eggs principle plausible offhand? Maybe we
mistake the double counterfactual
If it were that A, then: if it were that B . . .
for a single counterfactual with a conjunctive antecedent
If A and B, then. . . .
27 Campsall’s Notabilia, 7, in The Works of Richard of Campsall ed. by Edward A. Synan (Toronto, Pontifical Institute of Mediaeval Studies,
1982), Vol. II, 40.
28 ‘On Ockham’s Way Out’, Faith mid Philosophy 3 (1986), 235-269.
29 According to Calvin Normore, personal communication.
30 ‘Is the Existence of God a “Hard” Fact?’, Philosophical Review 76 (1967), 492-503.
18
31 Notabilia, 8, in The Works of Richard of Campsall, 41.
32 For contrary views, see Plantinga, ‘On Ockham’s Way Out’; and Terence Horgan, ‘Counterfactuals and Newcomb’s Problem’, Journal of
Philosophy 78 (1981),331-356.