Evidence Considered: Intelligent, Optimal, and Divine Design

This is a second excerpt from Evidence Considered: A Response to Evidence for God (available here)

Evidence for God is a book edited by William Dembski and Michael Licona that presents fifty arguments for faith from the Bible, history, philosophy, and science. In this excerpt, I respond to the twenty-first chapter by Richard Spencer entitled: “Intelligent, Optimal, and Divine Design.” This is one of the chapters in the science section of the book. The endnotes have been removed.

Richard Spencer attended the Nature of Nature Conference at Baylor University in 2000 in which he heard speakers talk about intelligent design as though it were synonymous with optimal design. He believed this to be wrong and made a comment to that effect. He then crafted this comment into an essay, which the editors included in Evidence for God.

Spencer saw a challenge in the incompatibility of divine design and the apparent lack of optimization in nature. He, therefore, created a narrative to explain it. To me, it is clear that his hypothesis exists to explain why God’s design is indistinguishable from what we would see if there were no God. In other words, Spencer is plugging a hole that only exists if you assume God’s existence. Without evidence for God, his discussion has no foundation or even any meaning.

Spencer starts with the following: “If something has been intelligently designed, people often expect to see structures that are perfectly crafted to perform their individual tasks in the most elegant and efficient way possible (e.g., with no extra components). This expectation is incorrect not only for human design but also for divine design.” Spencer attributes imperfect divine design to “the use of secondary agents (including physical laws), the reuse of common design elements, the adaptive nature of biological organisms, and the fact that we don’t fully know the purposes of the Creator.”

It is unclear to me how anyone can know so much about divine design. What possible vantage point could we have that would allow us to assess divine design in this way? Nonetheless, I will go through Spencer’s four reasons for imperfect divine design: Secondary agents; reuse of common design elements; adaptive systems; and unknown purposes. The divine reuse hypothesis is particularly interesting because there is a scientific prediction involved that is testable as we shall see.

Secondary agents

Spencer observes: “In human design, we frequently have to do things in ways that are suboptimal simply because the complexity and magnitude of the overall task preclude spending time and attention on each detail that would be required to execute an optimal design … This limitation does not, of course, affect divine design.”

“Nonetheless, a similar limitation does affect divine design. It arises whenever design employs secondary agents.” Spencer then gives an example of designing a microprocessor with a computer-aided design (CAD) tool, which he then notes might be a better design in many senses, undermining his point a little.

In any event, he continues:

In the same way, but for different reasons, God usually makes use of secondary agents to accomplish his work. Such secondary agents include physical laws since these laws do, at least sometimes, define or help to define structures in nature. For example, there are physical laws and properties of matter that determine the physical structure of certain objects, and once the laws and properties are in place, God does not need to individually create each atom, cell, or higher-level object. Having created physical laws, God is constrained by them unless he specifically chooses to suspend them. As a logical possibility, God is, of course, free to suspend the physical laws he has instituted. Yet I don’t know a single unequivocal example in which he has done so. This is not to deny miracles. I am simply saying that I don’t know of any examples of miraculous structures in nature, and that includes biological structures.

He does not specify what “secondary agents” God uses apart from physical laws. I certainly cannot think of any. So what is Spencer claiming? His vagueness here means there are multiple possible scenarios, but none that I can see help his cause. Is he saying that the world looks as though it were created through natural causes because God chose to do it that way? In this case, he is left positing an explanation for why God made it look like He does not exist. If all we need is one positive demonstration of God’s existence, why is Spencer focusing on this area which seems devoid of God’s influence, even if it was because God used “secondary agents”? And observe the statement “God usually makes use of secondary agents.” Are we to suppose that Spencer has witnessed God in multiple universes with a variety of creation styles, and from his observations, this is “usually” how God does it? Apparently not, because Spencer then says he has never seen any unequivocal case of God intervening in nature at all. To the apologist, this may indicate that God chose to use a secondary agent. To the scientist, God is an unnecessary assumption in this scenario. This does not disprove God’s existence, but it certainly does not prove it either. The ball remains firmly in the theist’s court.

Reuse of common design elements

Next, Spencer says:

Given that God uses secondary agents to bring about physical structures, we can expect to see certain patterns and processes repeated in many places and used in different ways, even though the design may not be optimal for each individual application. In addition, any designer, divine or otherwise, is certainly free (and likely) to reuse structures and to implement similar functions in similar ways, although this will not always be the case. The appearance of similar structures in many different systems, particularly when those structures are not optimal for each situation, is frequently cited as evidence for macroevolution … But it is also exactly what you would expect to see for a system constructed using secondary agents under divine control.

Again, evolution is the natural law that drives this, and that is how it looks. But Spencer “expects” that this is just what a divine being would do. Once again the apologist is left explaining why God acts in a way that is indistinguishable from a Godless nature. Also, God’s use of secondary causes does not exonerate him from poor design since he must have made all the secondary causes too. If God is omnipotent and omniscient then for Him to think through all the consequences of creating and using physical laws would presumably constitute no effort at all.

There is another crucial point that I need to make on the “divine reuse” hypothesis. Spencer claims that divine reuse and evolution would both account for the observed structural similarity across different systems. That may be true qualitatively, but it is not so quantitatively. Here Spencer is making a claim that is amenable to empirical testing, and so we should look at the similarities more carefully and see whether we can distinguish between these two hypotheses.

One can look at a particular gene that occurs across different species (for example, the gene for hemoglobin A) and deduce a family tree based on how many differences there are. Identical genes are more closely related than genes separated by one mutation, which in turn are more closely related than genes separated by two mutations, etc. So the gene gives you a family tree. The significance of the tree will depend on the hypothesis you are testing, but the existence of the tree is an empirical fact.

One can then do the same for many different genes (fibrinopeptide B, cytochrome C, etc.). So you will end up with many family trees, one for each gene.

In the “divine reuse” hypothesis, there would be no expectation that different, independent genes have the same family tree. For instance, if God reused an element of a kangaroos hemoglobin A gene in a sheep, He would not be required to perform a similar reuse for the fibrinopeptide B gene. He might or he might not. In the evolution hypothesis, the expectation would be that all the different genes have the same family tree since the tree would arise from common descent. Note that there are many possible trees. For example, if we had ten different mammals and no other restrictions, we would expect there to be about a million different possible trees, so each gene we examined would be like rolling a million-sided die. The divine reuse hypothesis says that we would get different results for each different gene: just a random throw of the die. Evolution says that each time we roll this million-sided die, it will come up with the same family tree. In other words, evolution is making a far stronger and more specific prediction than the divine reuse hypothesis.

This experiment has been done multiple times, so the results are in and, of course, all the different genes reveal the same family tree which is what evolution predicted. The theory of evolution was developed before we knew about DNA, so this confirmation is truly remarkable. On the other hand, we should reject the divine reuse hypothesis based on these results. I am sure that a creative Intelligent Design advocate will eventually devise an appendage to the hypothesis that tries to account for it, but we should favor the simpler explanation. And we should note that if divine reuse is the cause, then the divinity in question has gone to extraordinary lengths to make it look like evolution. [I discussed this more here]

Adaptive systems

Spencer continues that adaptive systems “may appear to be less than optimal” because they are “inherently wasteful. In order to be able to adapt to different conditions, the system will virtually always have components that are not being used in a given situation.” This means that “engineers like [Spencer] who design adaptive systems expect to see many components that appear to be wasted or left over from some previous use. Although the appearance of such structures is commonly used to argue that evolution is not under intelligent control, in fact, it is a necessary consequence of adaptive systems. Moreover, since adaptive systems are not infinitely malleable, … this feature of adaptive systems provides evidence for microevolution but not for macroevolution.”

This line of reasoning is also inadequate when one looks at specific designs. For example, many people choke to death each year because our food intake is connected to our air intake and we can easily envisage a person designed in such a way that this weakness does not exist. So either God designed humanity this way, or this feature changed through adaptation. But such a significant change would constitute macroevolution. In other words, the built-in choke hazard is a design flaw that is unrelated to our adaptability, so either God designed it poorly, or macroevolution occurred. Either way, the adaptive systems argument falls flat as an excuse for suboptimal design.

Unknown purposes

Finally, in tacit acknowledgment that the above is inadequate, Spencer falls back on the old idea that the design is done for reasons beyond our comprehension. He says that “we are rarely in a position to fully understand all of the design objectives and constraints.” The challenge with all this is related to the loose definition of God. Spencer ascribes attributes to God to pull together an argument but eventually has to admit that he does not know what God would do and why. This is theistic agnosticism—the belief that God exists, but cannot be known—and is often the end point of any theistic attempt to describe God. Terms like omnipotent, omniscient, unchanging and ineffable are just ways of expressing ignorance. Such terms can have no real meaning.

Spencer goes on to quote Francis Collins approvingly:

It seems to me we should not make the mistake of assuming that God’s perfect will for us is biological perfection, any more than we should assume that God’s perfect will for us is the absence of suffering. It is those occasions when things aren’t perfect that we often learn the most, and when our closeness to God, which is a higher goal even than our own happiness, is most likely to come about. And so perhaps God in a merciful way speaks to us through our imperfections, and we shouldn’t neglect the significance of that. The underlying assumption that we should all be genetically perfect doesn’t necessarily make sense to me.

Spencer agrees “wholeheartedly” and mentions in passing that we must “also remember that the world we are observing is not the original creation. It is a corrupted version of the creation … [T]here is still an unknown factor to deal with since we are not able to observe the original creation at this time.”

Collins, who is a serious scientist (he headed the Human Genome Project) and an earnest Christian, rejects Intelligent Design and Creationism. If you have accepted faith in Christianity, then you need some way to understand the poor design. Collins does this with theistic evolution which excuses God by allowing a secondary cause (evolution, that is macroevolution by Spencer’s definition) and by making it not be a priority for God, which removes the issue, at least when discussed in the abstract. In practice, I cannot imagine what “closeness to God” is brought about by watching your baby choke to death on a grape, for example, for either baby, parent or anyone else. However, if you have not accepted theism, then no contradiction exists at all. So the entire discussion sits squarely inside the realm of theism, rather than acting as evidence for theism. It is brought up by both Spencer and Collins to shore up defenses against an apparent contradiction. But for the atheist, this is a non-issue: we are looking for (and not finding) a reason to accept faith in the first place.

Spencer then suddenly brings in the creationist fall. This seems strangely esoteric. Did Adam and Eve have a built-in choking hazard? If so, then the poor design existed before the fall, and God must take the blame. If not, then what mechanism changed Adam and Eve’s breathing orifice? Did God do that, or did the devil tinker with our design, or was there some very rapid evolution in the course of a single generation? All of these scenarios are beyond the remit of observations and reason, and so can only be believed irrationally. It is difficult to take this kind of argumentation seriously. I cannot understand this comment as it does not seem to fit in with the rest of his argument.

Ultimately, Spencer has proffered no evidence for belief in God. He has merely attempted to defend the implications of a belief in Intelligent Design. He has failed to make this defense, and the logic of reductio ad absurdum would dictate that we now reject the belief. Collins rejects Intelligent Design and squares his belief in God with an acceptance of science in a different way. It is the original belief in God that is concerning us here, and Spencer’s essay fails to make the case.

[Dr Spencer responds to this essay here]