Pain, Pleasure and Evolution:
An Analysis of Paul Draper's Critique of Theism

Dennis Jensen

P.O. Box 110843
Aurora, Colorado 80011

From: Perspectives on Science and Christian Faith 51.1 (March 1999): 40-46)

In April of 1997, Paul Draper of Florida International University presented a paper, "God, Evil and Evolution," at the annual regional conference for the Society of Christian Philosophers held in Boulder, Colorado. This was a preliminary study presenting three arguments for a cumulative case against theism to be developed in a forthcoming book.

Arguments that relate evolution to the existence of God will always be of special interest to Perspectives on Science and Christian Faith readers. So before his book comes out, it will be of value to look at some of Draper's basic claims.

In his first two arguments, Draper hoped to show a greater probability for naturalism over theism by arguing that we should expect certain known facts to obtain on the assumption of naturalism more than on theism. His third argument is that naturalism is more intrinsically probable than theism.

Evolution as Evidence Against Theism

In the first part of his paper, Draper looks at the notion of theistic evolution and comments that while evolution does not disprove theism, he believes it does give evidence against it. He claims that with the assumption of the falsehood of special creation, and because special creation has some probability given theism and none given naturalism, "evolution is...more probable on naturalism than on theism." ("Special creation" means that complex living things did not descend from simpler ones but were created independently.)

Even this first meager step in Draper's argument, however, can be questioned. If the assumed falsehood of special creation can be used in this way to count against theism, then it would not be hard to think of a naturalistic myth (if I might call it that) that could count against naturalism and to the same degree. A naturalistic myth would be the alternate of a supernaturalistic myth, like special creation. In this case, our naturalistic myth would need to have a creation account that does not involve intelligent agency.

Imagine that a great cosmic tree once grew in a cosmic ocean and dispersed its seeds upon the earth. Each seed essentially became a species of plant or animal and populated the earth. Intelligent agency is specifically excluded in this myth, just as it is specifically included in special creation. Intelligent agency can be plugged into or removed from such myths and accounts very easily. For example, all we would need to do to make special creation a naturalistic myth would be to say that each species developed independently from inorganic matter but with no creator to do the work. We recognize today that the probability of such a chance event is infinitesimal, but that does not make it any less a naturalistic explanation.

Following Draper's reasoning, since our naturalistic myth has some probability given naturalism and none given theism, and since this myth is known to be false, evolution would be more probable on theism than on naturalism. But since we also assume that the falsehood of special creation makes evolution more probable on naturalism than theism, we would have to conclude that evolution is no more or less likely given naturalism or theism. (That is, it is no more or less likely if no other evidence is being considered.) Obviously, just coming up with naturalistic or supernaturalistic myths will not affect the probability of naturalism or theism.

Draper next critiques Christian evolutionists, like Diogenes Allen, Ernan McMullin, and Howard Van Till, who argue that God's transcendence or God's power and wisdom preclude not only special creation but even the creation of the first life form. God, being so incomprehensibly great in wisdom and power, simply set the universe going so that it would produce life on its own. Draper at one point responds—and this is possibly his best response on this issue—that it is difficult to imagine that a religion that professes the claim of an incarnation must in principle reject any direct divine activity in the creation.

But speaking more generally, to argue from what God should or should not do seems to me to be a method that can too easily be grounded in subjectivity. Duane Gish sometimes complains when scientists try to give any scientific explanations at all for certain phenomena. He appears to think it inappropriate of God to use physical processes in such cases. God should just miraculously and inexplicably "make it so." Augustine thought even six days of creation was gratuitous. What God should do, he thought, was to create everything instantly. Maybe it is our idea of what God should do or how God should do it that is mistaken. Perhaps waiting billions of years does not bother God like it does us. But maybe intervening at certain points in the origin and evolution of life does not bother God either.

Christians should have no difficulty accepting the possibility of theistic evolution in principle. I would accept it if I believed that the scientific evidence warranted it. I think that evangelical exegetes like Henri Blocher have shown us that there is certainly no good biblical grounds to reject at least the more conservative forms of theistic evolution. And possibly all forms can be accepted exegetically.

My greatest concern is simply that theistic evolutionists not reject other scientific possibilities merely because of their philosophical and less than certain theological assumptions. To follow methodological naturalism without being open to non-naturalistic explanations seems to me to be unscientific. We should certainly seek naturalistic explanations. But we also need to be open to non-naturalistic explanations, particularly as we increasingly fail to find naturalistic ones. For example, we must never cease looking for nonintelligent causes for the origin of life. God might have created as the above theistic evolutionists say, but as we continue to fail to find such explanations, we should consider the possibility of direct activity by an intelligent agent. It is simply unscientific to presuppose that there cannot be such an agency discoverable by science. Methodological naturalism closes us off from legitimate hypotheses that we need to consider.

Draper's next claim is that if methodological naturalism is as successful as some believe (he thinks it is not), then its ...success would obviously be...much more likely on naturalism than on theism." But this is not at all obvious. Wes Morriston, philosophy professor at the University of Colorado, responded to Draper's claim by pointing out that we simply have no reason to believe that God would not usually use such naturalistic processes rather than intervene more directly in nature. Thus we still have found no greater likelihood for naturalism over theism.

From his assumed proof that evolution is more probable on naturalism than on theism, Draper goes on to attempt to show that it is significantly more probable on naturalism. He does this by trying to show that we do have strong reasons to believe God would carry out special creation. Using the old design argument he says:

 Since we know antecedently that present day organisms are produced only by reproduction and yet such organisms have not always existed, it is, independent of the evidence for evolution, quite likely that the resemblance between the causes of complex machines and complex living things consists in the fact that both are specially created, machines by human beings and organisms by God.

But this does not follow at all. God might create living things like one creates a machine, using, however, chance processes to originate life and an evolutionary mechanism to change from one species to another. Or God might specially create only the first life forms and then allow evolutionary processes to carry out the further development of life. Or God might specially create the first life forms and then be more directly involved in the changes from one major life form to another (progressive creation). All of these possibilities are compatible with Draper's definition of evolution and none fit his definition of special creation (given earlier). (He defines "evolution" as "`descent with modification' of all complex organisms" with one mechanism for such change being "natural selection operating on random genetic mutation.")

The Distribution of Pleasure and Pain as Evidence Against Theism

Draper presents a second major argument against theism by looking at the distribution of pleasure and pain in the world under evolutionary naturalism as compared to evolutionary theism. He says that under evolutionary theism "it would not be at all surprising" that the driving mechanism of evolution would be goal directed and that pleasure and pain would be given special treatment to be respectively expanded and limited for moral purposes.

But it also would not be at all surprising if pleasure and pain were not given special treatment. God might use an apparently very nongoal-directed mechanism like natural selection as long as it can be foreseen to achieve the desired final goal, namely, the production of an organism with sufficient mental capacity to become human. And, of course, simply producing the lower organisms might also be a goal in itself. On the prehuman level, God's goals might have nothing to do with pleasure and pain. Pleasure and pain as they are found in the animal world are just a part of the mechanism to allow survival of the individual and species.

But then, why is so much pain allowed in the animal world? Could not God have achieved this goal without so much pain? Let's consider three possible answers.

First, consider that the existence of sentience, awareness, is such an enormous problem for naturalistic evolution that it is difficult to imagine that any animal could possess sentience unless it were uniquely given by God or a similar supernatural source. Thus, whether animals actually feel pain can seriously be questioned. With advances in our understanding of microbiology, the purely mechanistic character of life has become increasingly prominent. From what we see of them, animals are, more than anything else, essentially complicated machines. Such machines can be programmed to survive by avoiding harm and this will appear to us as pain responses. But it is unimaginable that a complicated machine made of nonconscious matter can in itself feel anything at all. We readily recognize that the simplest life forms probably lack sentience and thus the awareness of pain. If this is true of higher animals, then there is no animal pain.

Secondly, from a theistic world view, it might seem to be more likely that the higher animals would be sentient than not. But it also seems likely that animals might possess only enough awareness of pain to fulfill its survival value. An animal is programed to avoid such pains even though it would not be fully aware of pain as we are.

Thirdly, if animals do in any way feel pain like humans do, then God would surely provide compensation for such. And compensation does not mean simply giving pleasure in an afterlife to make up for past suffering. It would have to be something much deeper and worthy of the notion of compensation or redemption. But animals are not free to refrain from doing evil, they cannot develop moral character, and they are not deserving of punishment or reward. Thus there would be no reason for any special arrangement of pleasure or pain in the nonhuman animal world even on the assumption that they are sentient and as aware of pain as we are. The purpose of prehuman life is fulfilled and the problem of animal pain is answered: Animal pain occurs for survival purposes, animals must survive so that humans will come into being, and the pain that must occur is negated by its compensation or redemption.

Of these three possibilities, only the second seems to me to be adequate. Here, too, we must include the notion of compensation mentioned in the third scenario. On the one hand, it is difficult to imagine that the higher animals feel nothing at all. On the other hand, even with full compensation for any undeserved pain, it is hard to imagine God allowing such animals to live for millions of years with excruciating pain just so that we might come into being.

More should now be said about the distribution of pleasure and pain on the human level. Draper insists that for the evolutionary theist, pain and pleasure must have some (presumably perceivable) moral goal and not just a survival value. Yet the pain and pleasure we perceive usually only incidentally has moral value; usually it is only what you would expect assuming evolutionary naturalism.

One response might have something to do with the Fall, the first entrance of sin in the human race. Unfallen man and woman might have had pain of a kind, though, like their prehuman ancestors, not (normally) of the type that could produce the anguish it can create in an individual today. And then pain would only have been enough to prevent harm to a person. Or perhaps there was no pain at all before the Fall. Humans were protected as they obeyed God's commands ("Do not step over that cliff," "Do not touch that fire," "Do not eat that fruit," etc.) and as God would give special protection.

Having said all of this, I feel very sure that if there had been no Fall, there would still have been some instances of pain fully as horrible as we know it today. These would have been tied to a spiritual purpose, such as in Job's case, but not necessarily under any perceivable pattern.

With the Fall, we find pain and pleasure more clearly present and tied to survival purposes. Perhaps humans had to return to something more of their prehuman nature. No longer would they have the direct communication and presence of God to protect them. They would need the motivation of natural pleasure and pain to keep them alive and reproducing. Yet with their fully human mental capacities, and possibly as part of God's punishment or curse, their pain would not be mitigated as it is with the other higher animals.

Although other theodicies apply as well, certainly the Jobian theodicy applies to such apparently random pain. Everyone must suffer pain in order to answer Job's question: "Will I remain true to the God who deserves my commitment though I'm drawn emotionally (not rationally) to reject God for allowing this?" It is also important that we notice that even if the first sin had no effect on human nature, a random distribution of pain and pleasure will be inevitable to the degree that people are free to inflict or bestow them on others.

Concerning theodicies generally, Draper objects that they might explain some facts but they mystify others. But is not it possible that several theodicies might answer what one alone cannot? Jesus pointed out that some pains are punishments. When questioned about some who died in recent catastrophes, he warned some of his listeners that unless they repent they will perish as well. Other pains, he said, were meant simply to bring glory to God by their results. It was not anyone's sin that caused one man to be born blind. Being blind all of one's life is indeed a small price to pay when such honor and esteem is brought to God with its removal, especially in lieu of the fact that God promises compensation or redemption for all pain that we undeservedly endure in this life.

During the question and answer period, Draper admitted that a multiple theodicy approach might have potential, his only qualification being that these theodicies would have to exhibit coherence. It is not difficult to think of such "gap-filling" theodicies, though I am sure Draper would question their adequacy when we look at individual examples.

It is interesting that Draper thinks that pleasure should greatly exceed what is "biologically necessary" on an evolutionary theistic view. Christian theism would make it an important point to claim that in our normal secular existence we should not find much satisfying pleasure. We will only have psychological anguish and discontent as we are separated from God. Our greatest pleasure is to come from relationship with God. If pleasure in our normal secular existence were to "greatly exceed what is biological necessary," we would be too anesthetized to our need for God.

I have succumb to the temptation to answer Draper by presenting theodicies, though I could only touch on a few. I think that to do so shows more clearly that there are good reasons to believe that God has morally sufficient reason for allowing the evils and the arrangement of evil and good Draper is concerned about. But I would also claim that if God has morally sufficient reasons for allowing evil, we should not expect to know what they are. "It logically impossible to bring about certain goods without at least the existence of the evils we find in our world," Draper comments. But if this is true, then, contra Draper, no finite amount of evil in this life can provide any "evidence favoring naturalism over theism." Those achieved goods might sufficiently outweigh the evil in the world and they might be too significant to do without. We simply have no way of knowing this is untrue or likely untrue.

It is like a child undergoing a painful surgery. The child thinks that because the parents allow this, they cannot be good and they must not really love her after all. The child simply does not have the mental capacities to comprehend a parent's knowledge. What the child takes as good, probabilistic evidence should not be considered such at all.

Intrinsic probability

Draper closes his paper with some thoughts about how plausible or intrinsically probable (probable independently of grounds for belief or disbelief) theism is to naturalism. He begins by saying that "it is obvious to me" that with the lack of any evidence for or against supernatural beings, we should consider that it is not likely that such highly specified beings exist. Interestingly, while reading this portion of his paper during his lecture he interjected a comment by Alvin Plantinga. "What's obvious is not obvious," Plantinga said. In no other portion of this critique will the truth of Plantinga's observation be more certain.

Draper says that the more specific a claim is, the lower its intrinsic probability and the higher the intrinsic probability of its denial. Belief that an all knowing, powerful, and good person who created but is not part of the natural world ("theism") is more specified than the mere belief that something which is not physical nor emergent from the physical can affect the physical universe ("supernaturalism"). Theism, he says, entails but is not entailed by supernaturalism. If theism is true, then supernaturalism must be true, but not vice versa. There might be spirits or other nonphysical beings that can affect our world without God existing. Since naturalism is not less likely than supernaturalism, and theism is more specified than supernaturalism (see Fig. 1), it follows that naturalism is more likely than theism, Draper claims.

If it is true that the more highly specified an existential claim the less likely its intrinsic probability, then the theist might find this quite valuable in confronting some contemporary nontheistic claims. A multiple universe argument to explain the apparent "fine tuning" of our universe to account for life—and even more so intelligent life—would be one of the first casualties. If our universe did not just happen to possess the precise parameters of many of the laws and other characteristics it happens to have, then life would never have begun. Because it is so extremely unlikely that our universe should possess these precise parameters, it is extremely unlikely that life could ever exist. But if we conjecture that there are a large number of actual existent universes, the probability increases as that number increases that we will eventually run into one that does by chance have the needed characteristics. Because one happened to eventually work, the hypothesis says, here we are.

The many universes claim is a highly specified claim. If Draper's principle is correct and if this claim is pure, unverifiable conjecture, then as more universes are conjectured, the claim becomes more and more unlikely. But what about Draper's claim itself? Does intrinsic probability correspond inversely to specificity? As we will see shortly, I believe Draper has not adequately unpacked the notion of specificity and is mistaken in some important points.

In the question and answer period following Draper's presentation, Dr. Michael Tooley (University of Colorado, Philosophy Department) and Draper appeared to agree that unverifiable or unverified conjunctive claims decrease in probability and disjunctive ones increase in probability as claims are added to the conjunction or disjunction. It is less likely that there is a ghost in this corner and that corner than that there is a ghost in just this corner. (Any other term for a factual but unverifiable or unverified entity—insofar as it is unverified—could be used in the place of the term "ghost.") But it is more likely that there is a ghost in this corner or that corner than in just this corner. (Notice that the statement, "there is a ghost in this corner" is equivalent to "there is a ghost in this corner and there either is or is not a ghost in that corner.")

The claim that there is no ghost in this corner is at least as unwarranted as the claim that there is one. Both claims assume grounds for belief when no evidence is present to provide it. Likewise the claim that there is no ghost in this corner and in that corner is a stronger claim than that there is no ghost in this corner. It is less likely than that there is no ghost in just this corner. "There is no ghost in this corner or there is no ghost in that corner" is more likely than "there is no ghost in this corner."

"There is a ghost in this corner" is more likely than "there is a ghost in this corner who has power .18." (No power to all possible power would be 0 to 1.) "There is no ghost in this corner" is less likely than "there is no ghost in this corner who has power .18." To say that there is no ghost in this corner with power .18 is a weaker claim requiring less warrant.

Let's symbolize the statement, "there is a ghost in the corner" with G and the power of this ghost we can indicate by affixing the numbers from .00 to 1.0. Thus G.03 would mean there is a ghost in the corner with power .03. To say that there is no ghost in the corner, ~G, is the same as saying ~(G.00 v G.01 v G.02...v G.98 v G.99 v G1.0) which is the same as saying (by DeMorgan's Theorum) ~G.00 . G.01 ~ G.02 ....~G.98.~G.99.~G1.00. So to say that there is no ghost in the corner is to make a number of conjunctive claims; namely, that there is no ghost in the corner who has one particular power and there is no ghost in the corner who has another designated power, and there is no ghost in the corner who has another designated power, etc. And the more claims that are made without warrent, the less likely is the claim to be true than a simpler claim.

Supernature is a disjunction of a myriad of possible entities. Thus it is much more probable that there is this entity or that one or that other one, rather than just this one alone. It is more likely that at least one of these possible entities exists than that only one exists with the existence of any others being unstated. (A v B v C) is more probable than A .(B v ~B).(C v ~C). For example, if ghosts are defined as nonmaterial beings, then the existence of ghosts, or of God, or of gods, or of Buddhas (appropriately defined) would be more likely than the existence of only one of these (for example, God) with the existence of all the others being unstated. So, supernaturalism is more intrinsically likely than theism.

Naturalism is the denial of all these entities. If it were to deny only one, like atheism denies theism, then we should think that naturalism is as likely as its alternative. But because naturalism makes such a strong claim, because it says there is no God, and there are no Buddhas, and there are no ghosts, etc., it must be far less likely than supernaturalism. Without evidence to support it, it is less likely as the number of conjunctive claims increases. To be wrong in just one of these conjunctive claims is to negate the whole of naturalism.

If supernaturalism claimed a conjunction of specified entities, then it would be as antecedently unlikely as naturalism. Furthermore, because naturalism denies so many specified existentials, it is far less likely than atheism which denies only one specified existential claim. Likewise naturalism is far less likely than theism to the degree that theism makes only one specified claim.

Now Christian theism and the theism of many other religions do deny the existence of some of the supernatural entities enumerated earlier. Early Christians were often called atheists because they denied the existence of the Greek/Roman gods. Most forms of theism affirm the existence of usually a number of other supernatural entities like angels, demons, spirits, etc. To the degree any form of theism is conjunctive in its affirmation of the existence of certain supernatural entities and in its denial of the existence of others—to that degree it is closer to naturalism in plausibility. Most forms of theism would be a little more likely than naturalism because they would leave open the question of whether certain entities exist, neither affirming nor denying them; they would not make a definite denial of their existence like naturalism does.

Because theism is definitely less intrinsically likely than agnosticism (because agnosticism makes essentially no claims at all), does not that count for something? Does not it at least count against theism? Not if theism was never meant to be ascribed any probability before consideration of evidence. Not if the probability for theism is rather the conclusion of an evaluation of evidence of some kind. Now to say that we have no probability for a belief or no reason to believe it is not to say that there is reason to disbelieve it. There is a difference between saying that there is no probability available to us to know that it is true and saying that there is zero probability that it is true. (In this latter sense we usually speak of probability as being 0 to 1; that is, having no possibility of being true to being absolutely certain. A probability of .5, or 1 in 2, or 1/2, would mean that it is as likely as not.) Before we have any evidence, any reason to believe or disbelieve, agnosticism is the appropriate position.

Theism is more intrinsically improbable than the above might suggest, but only to the degree that it is made up of a number of necessary and thus conjunctive claims. And of course, as we have noted earlier, highly specified conjunctive claims like the multiple universe view are likewise highly intrinsically improbable until some evidence is provided to support them.

If the existence of something is highly intrinsically improbable, then it still should not take an excessive amount of evidence to make it probable. Before there was any evidence for the existence of quarks, say, the fact that their existence was highly intrinsically improbable had no effect upon scientists' later evaluation of the evidence for their existence. We should remember that every entity is highly specified. It is only unspecified to the degree we lack knowledge about it. Evidence makes it more specified and more probable.

I should mention that, if looked at within a causal context, I think a good case can be made for the claim that theism is significantly more intrinsically probable than naturalism (though I question whether this should any longer be called intrinsic probability). Is it more likely that this universe should exist on its own, in the various forms it has existed in through time, or that a comparatively very simple Being is the source of this universe? I think that this intuition is one of the very basic reasons many (most?) people who believe in God do so.

At the close of the conference professor Draper joked that his position, agnosticism, should be the winner among those presented since the other two major speakers, William Rowe and Stephen Wykstra, an atheist and a theist, both presented such good cases. It might be said that the entirety of this paper has been essentially a case for agnosticism since our main concern has been only to refute atheistic arguments and to put us at a kind of level playing field. Evidence for theism has only incidentally been included. In the last section especially, I pointed out that it is agnosticism above all other positions that has the highest intrinsic probability. Dr. Draper might very likely disagree with my arguments but I can only hope that he would be attracted to these conclusions, if he does indeed finds agnosticism attractive.

Professor Draper calls himself a hopeful agnostic. I would hope that some of my comments might support his hopefulness to reject what I think are some mistaken arguments for atheism and with that to better reevaluate some of the very strong arguments for theism and Christianity.