Thursday, July 31, 2008

Steve Lovell Reviews Erik Wielenberg's God and the Reach of Reason

A redated post.

God and the Reach of Reason: C.S. Lewis, David Hume, and Bertrand Russell by Erik J. Wielenberg

Reviewed by Steve Lovell

I had the honour of being one of the two reviewers of this book, and also the unexpected pleasure of being recognized in the acknowledgements both as an anonymous reviewer and as an influence in my own right. Although I disagree with a number of Wielenberg’s conclusion, I wholeheartedly recommended the publication of the book and would recommend it to anyone with an interest in C.S. Lewis and philosophy.

In essence the book attempts to bring the Lewis, Hume and Russell into dialogue. However, Wielenberg makes no apologies for focussing his attention mainly on Lewis, who he rightly declares to have been unjustly neglected by professional philosophers or at least by professional philosophers working in their professional capacity. The book consists of four main chapters. These examine (1) The Problem of Evil, (2) The Arguments for God’s Existence, (3) The Miraculous, and (4) The Design Argument and the Nature of True Religion.

I begin with my thoughts on the fourth chapter. Having spent the first three chapters focussing on areas of disagreement between them, here Wielenberg looks for areas of agreement between the three protagonists. He finds all three reject the design argument, and that all three favour the separation of church and state. There are is also some interesting discussion of “Lewisian” epistemology and the relationship between his avowed commitment to following the evidence and the views expressed in “On Obstinacy in Belief”. Wielenberg also spends some time in getting to the bottom of Hume’s “Dialogues Concerning Natural Religion” and argues that the views expressed by the character Philo are those closest to Hume’s own.

In all these matters there is much to recommend to the reader. If I have any criticism here at all, it is only that I preferred reading the more straightforwardly philosophical and dialectical material in the previous three chapters. It is to those chapters that I now turn my attention.

The Problem of Evil

In this chapter Wielenberg brings Lewis into dialogue with Hume. The problem of evil is carefully and thoughtfully explained through the use of some well chosen quotes from both Lewis and Hume, and Lewis’s response is then examined. Wielenberg’s exposition of Lewis’s response is extremely sympathetic and he goes to great lengths to show the richness and power of some of Lewis’s ideas here. In the end, however, he thinks that even a supplemented version of Lewis’s response is inadequate. It fails, according to Wielenberg, because it cannot account for “non-victim improving natural child suffering”. (His argument here is inspired by a famous passage from The Brothers Karamazov.) However, the sense in which Lewis’s response fails is merely that it has failed to explain this form of suffering. But as Wielenberg allows, from the fact that Lewis has not provided (and that we cannot provide) an explanation does not follow that no such explanation exists. Ultimately, then Wielenberg thinks that while the problem of evil gives some reason to doubt the existence of God, the argument is far from conclusive.

Arguments For the Existence of God.

In this chapter Wielenberg considers three arguments for the existence of God: The Argument from Reason, The Moral Argument, and the Argument from Desire. All three arguments are close to my heart, and Wielenberg has interesting things to say about each. I begin with the argument from desire.

The argument runs thus:

  1. All normal human beings have an innate, natural desire (Joy) that is for some thing x, where x lies beyond the natural world.
  2. Every desire that is innate and natural to all normal human beings can be satisfied.
  3. Sp: Joy can be satisfied (from 1 and 2).
  4. If Joy can be satisfied, then there is something that lies beyond the natural world.
  5. Therefore, there is something that lies beyond the natural world (from 3 and 4) (p. 110)

Wielenberg grant’s premise (1) and looks for support for (2). This support he finds lacking, and I am inclined to agree. He then looks at an alternative formulation of the argument according to which if there are natural desire that cannot be satisfied, then life is absurd. This, it would appear, is a bullet that Wielenberg is willing to bite. This doesn’t surprise me, although I don’t think it follows from Wielenberg’s claim that atheists and theists tend to disagree about the meaning (or lack of meaning) of life that the Argument from Desire, so formulated has no weight. I’ll admit that, in that formulation it is unlikely to persuade the hardened atheist, but what about the agnostic?

Wielenberg also goes on to consider a Bayesian version of the argument from desire, according to which theism provides the best explanation for the existence of a natural desire which cannot be satisfied by anything in this world. Against this argument Wielenberg offers an interesting, and fairly plausible evolutionary account of how such a desire might arise. He admits, however, that it leaves the argument to “God or Absurdity” untouched.

The Argument from Reason

In this section Wielenberg considers the argument with which readers of this blog are likely to be familiar, that of Miracles chapter 3. He also interacts, although not that much, with Victor’s book, which he rather lamely describes as “quite well done”.

Wielenberg’s interpretation of Lewis here is interesting. Passages in Lewis that Victor and I have tended to depreciate (that stuff on evolution) plays a more central role in Wielenberg’s interpretation. According to Wielenberg’s Lewis, naturalism cannot account for reason because it cannot account for intentionality, and it cannot account for intentionality because there is no way that evolution could turn the stimulus/response relationship into the thought/object-of-thought relationship. One of the problems with this line of interpretation is that it makes Lewis’s discussion of the relationship between Cause-Effect explanations and Ground-Consequent explanations of thought simply irrelevant. Wielenberg is aware of this, but still thinks his interpretation is clearly correct and finds no other line of argument in the relevant passages from Lewis (see footnote 111 to page 96).

Wielenberg’s strategy from here is to admit that we don’t have a naturalistic explanation of intentionality, but to point out that not having one isn’t the same as there not being one. Wielenberg notes the similarity with the theist’s response to the problem of evil, described above. Although he doesn’t say as much, this presumably means that Wielenberg accepts that evil is (non-conclusive) evidence against God if, and only if, intentionality is (non-conclusive) evidence against naturalism.

The Moral Argument

The dialectic in the discussion here is rather involved, and I will not attempt to capture it in detail here. Wielenberg’s discussion takes some interesting moves, including juxtaposing the moral argument, the Euthyphro dilemma and Lewis’s arguments against Dualism. Two arguments against Dualism are considered. One of these is based on the idea that the bad power would have to pursue badness for its own sake but that we have no experience of this, and indeed it seems impossible. Wielenberg dismisses this argument quickly, finding an apparent counter-example in Augustine’s account of his early life. Personally, I’m completely unconvinced by Wielenberg here. An a priori version of this argument runs: If action is to be rational it must be aimed at some end and if that end is sufficient to explain the action, then the agent must regard that end as good or worthwhile. But an end cannot be regarded as good or worthwhile merely on the grounds that it is morally wicked. Therefore, the action is not performed solely because it is morally wicked. This argument goes back at least to Aristotle. Wielenberg considers much this argument, but thinks that either the premises aren’t a priori or that the Dualist will be happy to accept that the bad power acts irrationally rather than rationally. I’m not so sure. While the concept of agency may allow for actions without ends, I don’t think we can make sense of agents who perform actions simply and solely because they are bad … and bad according to a standard they accept, bad by that agents own lights, because they are actions that the agent sees as bad.

However, although the discussion of Dualism is interesting, the big issues are elsewhere. Wielenberg’s main line of response to the Moral Argument is that naturalism has the resources to accommodate moral realism. He claims that moral truths are necessary truths, no more in need of explanation than the truths of logic, and that our evolutionary origins have served to bring us into contact with these truths. The view Wielenberg is advancing here is rather underdeveloped, and while I have yet to read the book he has written on this issue, I don’t hold out much hope for this general line of thought.

Firstly, it seems obvious that moral truths are not analytic truths, so if they are necessary truths they are synthetic. This is itself is enough to cause some naturalists to go pale. However, Wielenberg also seems committed to (at least some) moral truths being known a priori. This seems to commit Wielenberg to the Synthetic A Priori. Now, naturalists haven’t generally looked too kindly on the Synthetic A Priori, and one can see why. Synthetic truths are substantive truths, and it is hard to see how we can get at such truths without investigating how things are in the world. The obvious way to get around this is to claim that we are appropriately configured to recognize or hardwired to believe such truths. Wielenberg seems to think evolution could have done this:

“Each of the following cognitive seems likely to be selected for by evolution. The first is the capacity to recognize oneself as a bearer of certain fundamental rights – for example, the right not be killed for no reason and the right not to be exploited by others. Beings that recognize that they have such rights are more likely to resist treatment that would render them less likely to pass on their genes to the next generation.” (p. 90)

Well, yes I suppose so. But what work is being done here by the word “recognize” that wouldn’t be also done by “believe”? In short, Wielenberg has offered an evolutionary explanation for the origins of moral beliefs, but he has not offered an explanation of either moral knowledge. On this account the beliefs are simply those we’d have whether or not there are any moral facts with which they may correspond.

So, at least in this volume, Wielenberg has not provided a good alternative explanation for either moral truth or moral knowledge. So, as far as this discussion goes, Lewis’s moral argument comes out pretty much unscathed.


In this chapter, Wielenberg brings Hume and Lewis into dialogue on the topic of miracles. Wielenberg’s interpretation of Hume is different from my own. He sees both the argument and the intended conclusion rather differently from me. But no matter. According to Wielenberg, Hume seeks to establish that

“It is never reasonable to believe that a miracle has occurred on the basis of religious testimony.” (p 130)

For Wielenberg, “Religious testimony” is a technical term

“Testimony that is intended to support a particular system of religion.” (p. 127)

So, fleshing this out, the intended conclusion is that

It is never reasonable to believe that a miracle has occurred on the basis of testimony that is intended to support a particular system of religion.

Why the intentions of the person testifying are important, I do not know. Nor, to be honest, do I understand how this brings in any important difference between religious testimony and testimony simpliciter. But such gripes are relatively unimportant. According to Wielenberg, at the heart of Hume’s argument is a principle he calls the “Probability Principle”:

“We should rate the occurrence of event A as more probable than the occurrence of event B if and only if: The evidence provided by our experience supports the occurrence of A to a greater extent than it supports the occurrence of B.” (p. 128)

The wording of this principle is, I think, rather poor. “Probability” and “Evidential Support” are difficult terms, but given the passages of Hume from which Wielenberg extracts the principle, it seems to be roughly equivalent to the following:

Event A is more probable (more likely to come about) than event B if and only if: we have experienced more events like A than we have experience events like B.

I have many problems with this principle, but I’ll put those aside to focus on the issues raised by Wielenberg. Wielenberg rightly connects this principle with another principle known as “The Uniformity of Nature”, a principle which Lewis discusses in his book on Miracles. Lewis thought that Hume’s argument relied heavily on this principle and that it was not justified in doing so. According to Lewis, if God exists, then we have no guarantees that the future will resemble the past or that unobserved regions of space will resemble those closer to home. Using a version of the argument from reason, Lewis goes on to argue that we can only be justified in trusting our belief in the uniformity of nature if God exists, but then if we believe that God exists, we have no guarantee against miracles. If Lewis is right it would appear that the uniformity of nature and so Hume’s argument are in trouble.

Having already rejected the argument from reason, Wielenberg seems to think that Lewis can’t avoid Hume’s conclusion this way. But here Wielenberg would seem to have forgotten the dialectical situation. It is Hume that needs the Probability Principle and the Uniformity of Nature. To make trouble for Hume’s argument we need only note that the existence of God would make our belief in unqualified versions of these principles unjustified. If this were true, then Hume’s argument would, in effect, be assuming that God doesn’t exist, an assumption which he surely isn’t entitled to make in the context. Wielenberg seems to see that Hume isn’t entitled to take the Probability Principle and the Uniformity of Nature as simple obvious, but he keeps the claim that if there is no good argument for the existence of God then the Probability Principle will stand. It is here that I think Wielenberg has lost his grip on the dialectic.

Wieleberg goes on to discuss Lewis’s case for the “fitness” of the incarnation, and also Lewis’s famous “Trilemma”.


This is only a short review, and in reality there is much more to be said than I’ve said here. While I have mainly been critical of Wielenberg, there are actually many areas where we agree. He gives Lewis a number of small but significant victories in the dialogues with Hume and Russell, and his exposition of Lewis’s work is clear and concise. Wielenberg’s analysis of Lewis’s arguments is sympathetic and his knowledge of Lewis’s writings is clearly extensive, as is illustrated by his excellent use of relevant quotations. Of the few book length attempts to philosophically analyse and assess the main philosophical themes in Lewis’s writings, this is by far the best. Anyone interested in Lewis and Philosophy should certainly add this to their reading list.


exapologist said...

Wow -- that was a very helpful review!

I had a question regarding your point about naturalism and the synthetic a priori: Suppose I come to a philosophical thesis via the common philosophical method of definition-and-counterexample. So, for example, suppose I'm trying to determine the nature of knowledge. I keep refining my analysis in light of hypothetical scenarios until I've added enough clauses to rule out all the counterexamples. The result is a putative analysis of knowledge that captures my intuitions about when we have knowledge and when we don't, and it admits of no counterexamples that I can think of. I then accept the analysis because of these virtues.

Now would *that* count as synthetic a priori knowledge? If so, then why is that problematic from the standpoint of naturalism? But if it's not, then it appears that we don't need it -- we can get along fine with the method of definition- and-counterexample to gain justified beliefs and hypotheses about the world besides ordinary observation and scientific investigation, even on the assumption that naturalism is true.

At any rate, that was my initial reaction to your point.

Nice post!


Anonymous said...


As a stock method of "analysis", I think that counts as analytic, as it's a form of persuasive definition.

It does have following in common with synthetic truth though: it needn't be "obvious".

Does philosophy need the concept of synthetic a priori if there are non-obvious analytic truths?

Well, I doubt I'm going to convince anyone with this, but "yes". To me the essence of "synthetic" is "substantive". If it's possible to know substantive things a priori, then these won't, I think, be available by mere conceptual analysis.

However, that is not to say that analysis (the "definition - and - counterexample method") couldn't be used as part of a method for coming to know such truths. I just think that in coming to your conclusion you'll have to use something synthetic/substantive (even if obvious) at some point.

Personally I don't see this as a problem for the synthetic a priori. I also think there are naturalistic epistemologies (with a heavy evolutionary bais) that can accomodate the synthetic a priori, I only point out that it's not a terribly well received category of knowledge in most naturalist circles.


Anonymous said...

Wow, I think I'm going to have to read that book! I thought it would just be another Beversluis-style 'smackdown'.