Skip to main content

Could God Have Made the Big Bang? (On Theistic Counterfactuals)

  • Duncan Macintosh (a1)

That the universe began in a big bang is often believed by theists to confirm divine creation ex nihilo. But Quentin Smith claims that it means God must not exist. For if he does, there is an earliest state E of the universe. God made E. E is ensured either to contain animate creatures or to lead to an animate state. For God would know that an animate universe is better than an inanimate one, and that even a minimally morally good being would be obliged to create one if he could. And God, being at least minimally mor-ally good, and all-powerful, would be able and inclined to ensure the existence of one (p. 53). But science says that E is inanimate since the big bang singularity (E) involves the life-hostile conditions of infinite temperature, curvature and density; also that it is inherently unpredictable and lawless so that there is no guarantee it will emit particles that will evolve into an animate state. Thus £ is not ensured to lead to an animate state (p. 53), and thus God could not have made E. So, God does not exist (p. 54). Smith: “There are countless logically possible initial states of the universe that lead by a natural and law-like evolution to animate states and if God had created the universe he would have selected one of these” (p. 58).

Hide All


1 Smith Quentin, “Atheism, Theism and Big Bang Cosmology,” Australasian Journal of Philosophy, 69 (1991): 4866. All page references are to this work unless otherwise noted.

2 And if he was going to make several universes, beginning each with big bangs, he would so make them that at least one eventually contains life.

3 A referee thinks “no chance” too strong. Surely there is some chance; for she might succeedin cutting the whole lawn. But this is due to ambiguity in the sense of ‘chance’. There is chance qua logically possible outcome, and qua finite likelihood of outcome. It can be a possible outcome of her wanderings that the lawn gets cut, without there being a finite chance of that. There are infinitely many ways she could fail; so there is no finite chance of her ever randomly succeeding. E.g., she could start going in a circle after the first second, or half-second, or… In any case, that is the notion I seek to illustrate with this example. (Contrast it with the next, where we assume some finite chance of success without your further interfering.) Even if there is no finite chance of her succeeding by random motions, your guidance is not neededTor her to succeed, only for her to be guaranteed to do so, either certainly, or with some finite chance.

4 Leslie John reports, in his Universes (London and New York: Routledge, 1989), p. 181, that A. R. Peacocke “has been arguing… that God might… have ereated up to infinitely many universes, confident that at least some… would become life-containing just by chance,” which sounds rather like the last option.

5 Chancy or worse—see note 3 above.

6 A referee wonders how God is here to be understood to have created the living universe. God seems merely its guarantor, taking no action. But Smith only requires that God ensure life, not directly create it. Besides, God doer take an action: he starts a process that raises the probability of life beyond what it would have been had he done nothing; and his readiness to step in ensures life, even if the big bang only makes it slightly probable. But if life results without God's having to intervene in his big bang, is he responsible for life, especially if it was unforeseeable that his making the big bang would yield it? Did he ensure it? Yes, for he made the process which had a chance of making life, which otherwise would have had no chance of existing, so that life's existence counterfactually depends on him. And his being ready to step in if things go badly means he gives a guarantee; he is the ensurer.

7 A referee asks, “If God would have intervened had things begun to go poorly, then how could it even have been ‘potentially’ self-defeating to have created a world in which things might have gone… poorly were God a definite nonintervener?” If it was always part of God's intention in making the universe with a big bang that, if it did not autonomously yield life, he would make it do so, then no act of his tended to defeat his aim to make life. For his initial act of creation was the action, as defined by its intention, of making-life-by-a-big-bangor-directly-(if required). So both making the big bang and leaving it alone, and making it and fixing it, would have counted as executing his initial intention. Thus, if he had had to intervene, that would have been consistent with his initial intention, not a repair on its consequence, and so not a repair of a self-defeating action. I agree, given the assumption about God's intention. His making a big bang, which then happened not to yield life, would only be self-defeating if he intended to make life with a big bang, intending not to have to interfere later.

8 A referee thinks there is a respectable—if not compulsory—sense in which God works against the natural tendencies of a big bang if its chances of emitting life are, say, one in a billion, and he sticks life in its sequel. But since there is no law saying that if there is a big bang, there is no life, God does not change a law, nor violate one. Since, then, there is nothing for God to be interfering with, his making life in the sequel of a big bang is no interference with it. Is he not altering the odds of life being in its sequel? Not necessarily. Maybe the reason the odds are one in a billion that there will be life after a big bang is that God puts life after one of every billion big bangs. Besides, it is consistent with the odds of one's emitting life being one in a billion, that in a given run of big bangs, there be two in a billion; other runs could contain fewer than one to balance out the numbers.

9 A referee suspects paradox. “Could God determine… what is to result from an indeterministic process, without destroying its essential randomness…?” But what makes the process involving the big bang and later life indeterministic is that there is no law taking the initial states of the big bang into life-involving states. That is consistent with God's making—causing, determining—both the initial state and the later, living one. Suppose I put down some cards in random order. There is determinism in the relation between me and each card, for I caused each to be put down; but not between one card and another, for no card caused another to be put down.

10 Were the odds infinitely bad, God might be in trouble. But Smith himself (in “A Natural Explanation of the Existence and Laws of our Universe,” Australasian Journal of Philosophy, 68 [1990]: 2243) interprets physics as saying there is a finite chance of a big bang's yielding a universe with laws like ours, laws congenial to life; big bangs do not fully underdetermine their ejecta's properties, but make certain properties probable in some degree, give them a small, finite chance of emerging. Note that if this is not true, either in the nature of big bangs, or by God's decree, then Peacocke's method (see note 4, above) is faulty. If the odds are infinitely small that any given universe will contain life, the same is true of any number of universes; so it would be unreasonable to be confidant that some will contain life “just by chance.” What if God plays the big bang game infinitely many times, with an infinitely small chance of life emerging each time; is he likely to win? Alas, this passes my math; I cannot say. But for God to be sure of creating life by making many universes, this must not be like tossing coins until one comes up heads. For heads may never come up, though that is ever less likely. (It is 50 percent likely for each toss, but far less likely for 10 tosses, less likely still for 100, and so on.) Instead it must be like being allowed to pick lottery numbers at random (picking each time only from those one has not yet picked) from finitely many numbers until one picks the winner. So God can leave it open which of the many universes he will make will contain life, but must decree that at least one of them will. (So to make the analogy with semi-unreliable friends accurate, your friends must have agreed that one of them will do what you ask, but they must not have told you which one will do it when you ask.)

11 A referee thinks that while I rightly claim Smith has not shown that God did not create the big bang, all Smith meant was that God could not create the universe via a big bang simpliciter, a point with which my arguments seem to agree. But Smith was trying to argue that the big bang did not ensure life, so God could not have ensured life by making it; and that God could not have ensured life by making the big bang then doing other things. For since big bangs make life un-likely, in making one, he would be defeating his aim of making life, making a mistake he would have to fix. But God would not be so irrational. So he must not exist, given a big bang. I have tried to show that God's life-making actions can be part of the big bang; or that it does not have properties making it self-defeating of God to make it while doing other things too to guarantee life.

12 Bennett Jonathan, “Counterfactuals and Possible Worlds,” Canadian Journal of Philosophy, 4 (1974): 381402; and Davies Wayne, “Indicative and Subjunctive Conditionals,” Philosophical Review, 88 (1979): 544–64.

13 Perhaps not all big bangs yield life; but that is not Smith's objection, though we consider its implications, below.

14 Smith does not object to things holding logically pre-creation. He only doubts the pre-creation existence of conditions making determinate, contingent CFs about creation.

15 Stalnaker Robert, “A Theory of Conditionals,” in Studies in Logical Theory, edited by Rescher Nicholas (Oxford: Blackwell, 1968), pp. 92112;Thomason Richmond and Stalnaker Robert, “A Semantic Analysis of Conditional Logic,” Theoria, 36 (1970): 2342; and Jackson Frank, “On Assertion and Indicative Conditionals,” Philosophical Review, 88 (1979): 565ff.

16 Lewis David, Counter/actuals (Cambridge: Harvard University Press, 1973).

17 A referee objects that this implies that there are, available to be discerned, true CFs at the resembling world, which there could not yet be for Smith, since it does not become the relevant resembler until creation of the actual world. But a CF is not made to hold of one world by that world's most resembling another world where the CF holds. (For what then would be the condition of its holding there?) Rather, a CF is made to hold of a given world by the CF's consequent holding at a world where the CF's antecedent holds and which is otherwise the world most like the given world. And since worlds are just sempiternal possible successions of possible events, so that one apprehends worlds by imagining such successions, the truth-conditions of CFs are sempiternally available for contemplation.

18 Theism and Big Bang Cosmology,” Australasian Journal of Philosophy, 69 (1991): 492503, a reply to Smith. See especially pp. 493-96. Actually, I find Craig difficult to follow on this; he may not have posed the question to himself in quite this way, but may have just argued that CFs concerning future contingents could, in general, be determinate pre-creation. But he seems to find that a reply to something like our worry. Still, maybe I should only say that, if he has an answer to it, it is the argument I now sketch.

19 Proposition (l) appears as “(C)” in his article.

20 Does this, plus logic, make the required CF determinate? For any propositions P and Q, if Q is true, e.g., ‘there is life’, then if P is true, e.g., ‘there is a big bang’, then ‘if P then Q’ is true; thus, ‘if there is a big bang, then there is life'. But this only makes a material conditional (MC) determinate and true, not a CF. For that, we need 'if P had been true, then Q would have been true', which does not follow from the MC. MCs with true consequents will be true whatever else holds at any world, but contingent CFs are true only if certain particular things hold at resembling worlds.

21 The argument of this section in effect develops a point made rather quickly by John Leslie in his Universes, pp. 181-82.

22 Is there not at least a deterministic relation between the universe and God? Yes, but see note 9 above. Also, God's causing events may not involve nomic causation.

23 A referee suggests that this point ("that creating a universe in which the various stages are not nomically/counterfactually following from one another is compatible with giving it a determinate… selected… life-containing form”) is one “that J. J. C. Smart has long been arguing for. Smart concedes that in his B-theory-of-time world the future is determinate; he denies… that this commits him to determinism. See… his… Our Place in the Universe.”

24 A version of this paper was presented at Dalhousie University. For helpful discussion, my thanks to David Braybrooke, Bob Bright, Steven Bums, Susan Dimock, Wayne Fenske, Randall Keen, Bob Martin (especially for “the second way”), Victoria McGeer and Sheldon Wein. I am especially grateful to two anonymous referees.

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

Dialogue: Canadian Philosophical Review / Revue canadienne de philosophie
  • ISSN: 0012-2173
  • EISSN: 1759-0949
  • URL: /core/journals/dialogue-canadian-philosophical-review-revue-canadienne-de-philosophie
Please enter your name
Please enter a valid email address
Who would you like to send this to? *


Full text views

Total number of HTML views: 0
Total number of PDF views: 2 *
Loading metrics...

Abstract views

Total abstract views: 106 *
Loading metrics...

* Views captured on Cambridge Core between September 2016 - 20th November 2017. This data will be updated every 24 hours.