Did Paul Draper Destroy the Contingency Argument?

Reply to Paul Draper

In his recent book, Atheism and the Problem of Evil, Paul Draper examines the overall case for and against traditional theism. Although the problem of evil is the book’s central focus, Draper also considers several arguments for God’s existence, including the argument from contingency. His discussion raises an especially interesting challenge to that argument.

Draper raises several concerns about the contingency argument, but the fundamental challenge is that we need not expect every contingent fact to have an explanation. He offers a principled reason: diverse realities call for explanation, whereas a uniform reality may not call for explanation in the same way. The difference between diversity and uniformity may therefore be relevant to whether something needs an explanation. Thus, Draper thinks there is reason to doubt that every contingent fact has a deeper explanation.

I appreciate the clarity of this challenge. Draper identifies a specific point at which defenders of the contingency argument may need to do more work. Below, I share a few reflections.

1. A modest contingency argument

First, suppose Draper is right that our explanatory principle should initially be restricted to non-uniform contingent facts. We could then formulate a more modest contingency argument:

1. Non-uniform contingent realities call for an explanation.

2. The totality of non-uniform contingent reality cannot ultimately explain itself.

3. Therefore, there is a uniform foundation of non-uniform contingent reality.

That conclusion would still be significant.

Draper told me a year or so ago that he was becoming sympathetic to a fundamental-mind hypothesis in response to the hard problem of consciousness. If a modest contingency argument points toward a uniform foundation, and other considerations point toward a fundamental mind, the combined result would be a fundamentally uniform mind. Interesting arguments can therefore still be developed even within the constraints of Draper’s proposal. (Paul is well aware of this and has suggested some such arguments.)

2. Is uniformity enough?

My more fundamental reply is that I am skeptical that uniformity gets to the root of every explanatory demand.

To illustrate, suppose fundamental reality were contingently and uniformly blue. It would still be reasonable to ask: why blue rather than red? The reality’s uniformity would remove any question about why only one part is blue. But it would not explain why the whole reality has the character it does.

The same point applies to existence. Even if a reality contains no internal variation, we may still ask why it exists at all. If it could have failed to exist, its perfect uniformity does not seem to answer the question of why it is actual.

I have come to think that there are multiple, non-competing reasons to expect a deeper explanation. Draper’s objection helps reveal the importance of distinguishing different sources of explanatory demand. Diversity is one source. When a reality has multiple parts, properties, or arrangements, we naturally ask why those differences obtain. But it does not follow that diversity is the only source of a call for explanation.

Many philosophers have argued that contingency itself invites such a call for a deeper explanation. When a reality could have been otherwise or could have failed to exist, we can ask why it is actual rather than merely possible.

Experience supports this thought. We can imagine many uniform contingent realities that do not exist. For example, imagine a uniformly blue blob that fills all of reality. Such a blob does not exist. But why not? The best answer, it seems to me, is that nothing caused it to exist, and a contingent blob cannot simply exist on its own without a deeper causal explanation.

Thus, I do not see that uniformity removes every possible call for explanation. In particular, it seems to me that contingency itself generates an explanatory demand of its own.

3. Complications for inductive arguments

Draper also raises more modest concerns based on evidence from different sample classes. For example, he observes that candidate necessarily existing entities, such as numbers or laws of logic, are non-concrete, while all known concrete things are contingent. These cases may appear to support the generalization that all concrete things are contingent.

Alexander Pruss and I have discussed objections to arguments of this kind (in our book, Necessary Existence). One difficulty is determining the appropriate reference class. We can divide cases in many different ways, and different classifications may support different inductive conclusions.

For example, one might observe that all known contingent totals are caused. If we apply that pattern to the totality of contingent things, its cause would have to be non-contingent. Otherwise, the proposed cause would itself belong to the totality it was supposed to explain. The result would be that some necessary reality is concrete, which might balance Draper’s argument for the opposite conclusion.

There is also an epistemic complication. A necessary foundation is precisely the kind of reality whose necessity we would not expect to observe directly from within our limited experience. We would ordinarily infer its necessity from its effects or from the explanatory role it plays. We should therefore be cautious about demanding that a necessary foundation resemble the ordinary contingent objects we encounter.

These complications while not decisive, incline me toward more principled defenses of explanatory principles rather than a simple enumeration of observed cases.

4. A productive challenge

Draper’s critique is valuable because it encourages defenders of contingency arguments to identify more precisely what generates a call for explanation. It is not enough to say that a fact “seems unexplained.” We need to distinguish the explanatory significance of complexity, diversity, contingency, and existence.

My present view is that Draper’s appeal to uniformity does remove one explanatory demand. But it does not yet show that a uniform contingent foundation would be explanatorily complete, with no remaining call for a further explanation.

Still, this is exactly the kind of objection that advances the discussion. Draper offers a clear and insightful challenge, and I think the central question deserves further exploration: Does uniformity genuinely remove the call for explanation, or does there remain a deeper call grounded in contingency itself?