Basic Logical and Metaphysical Problems with "Intelligent Design" Theory

by Chris Cogan (Copyright 2005)

Feedback, discussion, comments, questions: Chris Cogan, ccogan@ou.edu

Besides the argument from the Bible ("The Bible tells me so"), the main argument for creationism is really just an argument against evolution, which is the argument to the effect that living things are too complicated to have arisen by naturalistic processes. This can be expanded in various ways, and often is, but that is nearly all irrelevant to the point I want to make here, which is that there is a problem with this entire category of argument, a problem that is inherent in the premises, and that makes all such arguments end up being self -refuting.

Note that I said self-refuting, not merely unsound or invalid for some other reason. By self-refuting, I mean that the required premises to support a claim that something is too complicated to have arisen by natural processes unguided by a mind or mind-like "designer" are such as to make the conclusion impossible to sustain. A further implication of the argument I will develop is that there is simply no limit, in principle, to the complexity that can arise by natural processes. That is, as a general principle:

There is, as a matter of logic, absolutely no upper bound on what may occur naturally.

Finally, I will show that the arguments from design cannot be salvaged by appeals to probability. That is, it is never the case that complexity (even "specified" complexity) can be used by itself in such a way as validate the likelihood of an intelligence behind it.

This may seem like a tall order, but it is actually a lot easier than I would have guessed ten or twelve years ago. Most of the work is just a matter of clearing away the overgrowth of foggy concepts and common misconceptions about complexity and design.

Let's start with a simplified, almost schematic, form of the argument from design:

1. There is some level of complexity beyond which any further complexity must be the result of intelligent design.

2. Whatever this level of complexity is, it is way below the level of complexity of most (or at least some) living organisms.

3. Therefore, most (or at least some) living organisms are the result of intelligent design.

My argument is indirect: We start by hypothesizing that the first and second premises are true, and then show that assuming these has consequences that are fatal.

Suppose that a bacterium (which is actually a common example of creationists) is too complex to have arisen naturally and that we can therefore infer that there is (or was) an intelligent designer who designed this bacterium.

Now, let's consider the designer. Although ordinary creationists are quite willing to specify various attributes of their designer, including omniscience and omnipotence, the "Intelligent Design" brand of creationists tend to be a little more discreet in their pronouncements except when they are proselytizing (which, despite their claims of scientific objectivity, occurs frequently even in their supposedly "scientific" books and articles, but it is still not as prominent in their case as it is among the traditional young-Earth creationists). In general, the "Intelligent Design" creationists would like to hide their presuppositions about the designer, because they hope to sneak "Intelligent Design" into the public schools at taxpayer expense.

However, taking a leaf from Dembski, even if we assume no special traits for our imagined designer, it is still true that he must have at least the intelligence of an ordinary human being, otherwise he will not be able to do much in the way of design, intelligent or otherwise. People with low-wattage brains and non-human animals don't do much designing, unless it's extremely simple (as when chimps will strip a twig of leaves in order to stick it down an ant-hole).

Actually, considering that no human on Earth can yet design and actually make a full-fledged living organism from scratch, we probably need to suppose that the designer would be a good deal more intelligent, or at least much more knowledgeable, than any ordinary human is.

Further, I'm sure that both the regular creationists and the "Intelligent Design" creationists will claim that their designer is in fact very intelligent, and very knowledgeable. They would claim, in anything but one of their "scientific" pieces, that the designer is in fact omniscient, though how they could possibly argue this from the level of complexity of living things (or even the whole universe and living things) is beyond me (couldn't a less-than-omniscient being nevertheless have enough knowledge to design living organisms?).

But, given that they must claim (for their religious purposes, if not for "scientific" reasons) that their designer is at least superhuman, and at least more complex than even a human being (including a human being's brain and its function).

Fine, the designer is more-complex than the things he/she/it/they design.

Here's where the problem arises. The initial argument was that living things are too complex to have arisen by unintelligent means alone. But, using that exact same reasoning, we can now argue that the designer him-/her-/it- self is too complex to have arisen by unintelligent means. Therefore, it must have had a designer, too.

Am I done? Can I say, "QED" and go home now?

Not quite, because there is at least one "emergency exit" that creationists can try to use at this point, which is the bizarrely arbitrary claim that the imagined designer that they hypothesize is eternal and never arose by any means at all.

Thus, we have an extremely complex whatsit, one that had no origin. Are you beginning to see the problem with their whole theory, even if we assume that such a thing is possible (and they cannot show that it is)? It means that complexity has no privileged status, that things that just are, without causation, can be indefinitely complex, and complex in a very special way, at that. It is no ordinary complexity that they imagine their designer to have, but specifically the complexity of a very complex and well-ordered and extremely well-developed mind, with complex information-storage and processing facilities, ones well beyond our own or even beyond what we can even readily imagine.

Obviously, by their own premise, such a thing could not arise by chance, by naturalistic processes, by evolution, or by any non-designer process at all, so their only option is to claim that it always existed, and that it was always very complex.

But, the idea of such a being existing without having come into being is at least as problematic as the idea that living things arose by naturalistic processes, because it assumes that Existence (all that exists, including their designer) is, at its very foundation, extremely complex.

And yet, here they are touting just how unlikely it is for something extremely complex to arise naturally, while at the same time claiming that the very basis of existence is itself just naturally extremely complex, that it is extremely complex, moreover, without ever having become so complex from a simpler state of Existence.

So, couldn't the methodological naturalist say, at this point, "Well then, perhaps you are right. Perhaps life didn't evolve to become so complex, but perhaps it simply naturally occurred as an expression of the pre-existing complexity at the metaphysical base of Existence, right? If your designer didn't have to become complex, perhaps life didn't become complex either, but just happened that way, just as you yourself claim Existence just happened to have this extreme complexity eternally present in it."

The other option is to stick with their initial premise and admit that their designer had to have a designer too, by the same argument that leads to the conclusion of a designer to begin with. But this doesn't work for them because it requires an infinite series of designers, each more complex than the previous one. That will hardly suite their requirements of theism, however, so they have to reject it and settle for the eternal-designer view.

So, eternal designer it is, even though this idea is at least as bizarre as the initial claim that complexity requires design.

But, eternal or not, such a designer still needs an explanation. If we can't explain how it came to be complex, then we need a deeper ontological explanation as to why it is complex, even if it didn't get complex from some simpler state.

Further, this is required by the implications of their own premise that life is too complicated to have arisen by chance. That is, to support this premise, they need some sort of deeper basis. Empiricism certainly will not help them in this case, because, empirically, we see that it is simply not true. Simple things become more complex with no or little help all the time, including many non-living things. If you order a deck of cards by color, by suit, and by face value, you will have a relatively simple order to the deck of cards. Now, if you shuffle it by merely unintelligently mixing the cards up (or by letting animals or wind do it), you will have a much more complexly "structured" deck of cards. Much more information will be required to describe its exact order. I used just a couple of sentences above, but for the shuffled deck I'd probably have to specify each card in sequence in order to provide an equally complete specification of the sequence. Why? Because it's so much more complex.

Obviously, if we have enough cards, we can produce any degree of complexity imaginable by simply shuffling the cards. And yet, this process of shuffling cards is one of the least intelligent processes we can imagine. Indeed, that's one of the reasons we do it; to get rid of the "intelligent" ordering of the cards that may remain from the last game, or from the way the cards were packaged by the manufacturer.

Even Dembski is clever enough to realize that complexity is not, by itself, an indicator of intelligence. Indeed, he uses the example of an election official in New Jersey (Caputo) part of whose job it was to select by random means who would appear first for some office on the ballots: A Republican or a Democrat. Out of 41 times he supposedly did this, 40 were all the same (I don't remember if they were all Democrats or all Republicans). Dembski then uses this simple pattern as evidence of intelligence behind the choosing process, ruling out randomness and unintelligent processes.

Demski treats this very simple pattern as an example of specified complexity, which is a little strange, considering that it was a nearly-mechanical repetition of a single option out of a total of only two initial options (it would have been more "complex" in a way if there had been, say, fifty different possibilities instead of only two).

But, when simplicity (the Caputo example) indicates intelligent design and mere complexity (shuffled cards) indicates unintelligent randomization, we have a problem that brings into question the whole idea that we can argue from complexity to intelligence.

This is one reason why the standard arguments from complexity fail. Dembski (and others), realizing this had to come up with something better. Dembski came up with "specified" complexity, which he defines in various sometimes contradicting ways, but which is a variation on the idea of ordered complexity, which in some ways is intermediate between absolute simplicity (the same item repeated in exactly the same way many times) and absolute complexity (complete randomization).

But, ultimately, even if we accept Dembski's notion of specified complexity (according to whatever definition he eventually settles on as the "real" one -- if any), the initial problem doesn't go away. If living things have too much of this attribute to occur naturally, then so does the designer, since, presumably, the designer would be something like the ultimate in complexity and order, a kind of metaphysical version of the Mandelbrot set, in which each point a mathematical plane is given a value according to a simple formula iterated many times. The simple algorithm for determining the values of the points on a plane produces what is, by any *ordinary* measure of complexity, a very complex (infinitely complex, in fact) structure that is also extremely ordered (each point on the plain is rigidly determined by the simple formula in a rigidly deterministic and simple way). But, despite the complexity, the ordering of the values for each point according to a simple algorithmic method (which, in its usual forms is "corrupted" a bit by the requirement that an arbitrary terminating value be supplied so the calculation of the value for each point can be guaranteed to stop after a finite amount of calculation), there is an underlying simplicity to it that none of the usual statistical methods for calculating complexity would detect. And, it is no fluke; it is a member of a large class of such instances of extremely specified complexity (complexity far beyond that of the human mind).

Does this work? Does specified complexity, in any form, save the arguments from complexity (by turning them into arguments from "specified" complexity)?

No. They suffer the same difficulties as the original argument except for the one that "specification" is designed to obviate, which was that complexity alone could just be randomization.

True, Dembski "proves" that living things have a special kind of ordered complexity.

But, that is hardly news to anyone, and certainly not to scientists. Indeed, explaining this complexity is part of why the theory of evolution is such a successful theory. Complexity can give living things more options, more different ways to respond to their environments and therefore, assuming that the costs of the complexity are not too high, it can give them an advantage over their otherwise identical but simpler species-mates. But randomness in an organism is generally harmful (though this is not true of all randomness in living things), so the complexity has to have an order to it that produces some sort of functional benefit to the organism.

So, does specified complexity require intelligent design? No. It requires some sort of systematic process for producing it, but it does not require intelligence design. Systematic selection from a continually replenished "pool" of candidates is enough, if the ones that are "selected" are able to contribute more to the pool of candidates than those that the systematic selection process culls out of the "pool."

However, let us suppose that it does require intelligent design. Does it help? No, because the same considerations would still apply to the designer it-/him-/her-/them- self(ves). If specified complexity requires design, then the designer has to exemplify specified complexity and so must also be designed.

-- Unless we can somehow "explain" specified complexity in the designer without resorting to another designer. But, if that is possible, then what's to stop us from using the sane method to explain the specified complexity of living organisms?

Here's where ID folks try to make a relevant distinction between the eternal existence of a designer with specified complexity and the origination of living organisms with specified complexity.

They try, but fail. This marvelously ad hoc differentiation is without justification. If, as they claim (but usually in different terms), Existence is such as to have specified complexity at its very ontological base, then on what grounds can they possibly deny that living organisms might arise in such an Existence by processes that themselves are merely expressions of this metaphysical specified complexity rather than as products of intelligent design?

The answer is: They can't. Any attempt at giving the eternal specified complexity of a designer a privileged status over the metaphysically inherent specified complexity of Existence as such, in order to provide a "creator" for living organisms is bound to fail, because it is simply arbitrary. There is no rational basis for such a special exemption from the premises they themselves have laid out.

Indeed, Dembski seems to have been aware of this problem. At one point he argues that, even if his argument does not show that living things are specifically designed but only that the universe itself exhibits specified complexity and must therefore be designed, then it will have served his purposes. But he didn't notice (or didn't admit to noticing, at least) that this, too would be applicable even to his designer, and that it would therefore mean that his designer either had to be designed or was itself the "product" of unintelligent "causes" (not in the sense of efficient causes, if the designer is eternal, but very bizarre material causes, at least).

Thus, Dembski must either opt for yet another designer to explain the specified complexity of the designer of living things, or he must opt for the equally bizarre claim that such fantastic specified complexity could always have existed, without being caused at all in any ordinary sense.

But then, suppose we accept that such eternal specified complexity is possible (and there is a sense in which it is -- though one not helpful to Dembski or other creationist arguments). Then, can't we argue that the specified complexity of living things is simply an expression of this inherent specified complexity of the metaphysical base of Existence?

If specified complexity can be inherent in an eternally-existing being, why can't the specified complexity of living things be inherent in the nature of the universe itself, so that living things are merely one way in which this specified complexity is expressed or exhibited?

I cannot see any good answer for the ID ("Intelligent Design") advocates at this point. They are faced with either requiring (by their own premises) that their designer also have a designer or, if they claim eternality of the designer's specified complexity, they admit that not all specified complexity must have a designer, in which case they can hardly argue consistently that living things must have had a designer, particularly in a universe that clearly exemplifies specified complexity in its physical nature quite aside from living organisms. They can't really reject specified complexity, because they need that to distinguish their designed things from mere randomness, and they can't really claim that there can be eternally-existing beings with specified complexity without giving specified complexity a metaphysical status in the nature of Existence and thus granting, in effect this would allow for specific things (living organisms) to come into existence from a basis of nothing more than the specified complexity inherent in Existence itself, as expressed in the natural ordered complexity of the universe.

I have argued that all instances of specified complexity need some kind of an explanation, even if it is not one in terms of efficient causation, and that, therefore, it does no good to argue for an eternal being who exemplifies specified complexity, because it would either require a non-designer explanation.

True, it would not be an explanation in terms of how it came to be, since that would assume that it did come to be, which could not be true of an eternal being, but that it must have an explanation in terms of such factors as material causes, in terms of factors in the nature of the basic "stuff" of Existence that would make such a being possible.

Neither ID nor conventional creationists have a good answer at this point. That's only partly because they generally manage to evade having this issue raised by various kinds of equivocations and conceptual waffling and ever-shifting definitions modified on an as-needed basis to evade having to deal with such problems (while leaving them cognitively unresolved, of course, since such evasions are intended to obscure facts, not to bring them into the light and clarify them).

Another reason they have no good answer at this point is that there are no good answers for this problem except to give up the stupid premises that led to it to begin with. Arbitrarily resorting to "eternally existence" to evade the argument doesn't help because the problems of eternally existing specified complexity are just as bad as (worse than, actually) the problems of explaining the rise of specified complexity in living things via natural processes.

The final reason they have no good answer is that the evolutionist's answer is so much better. Specified complexity of living things comes from the complexity of the environment to which living things must be adapted if they (or their genes) are to continue to exist. If the environment of a simple replicator could be relied upon to be and remain simple (or, at least, if the survival demands it made could be relied on to remain simple), things living in it would not have to evolve in any direction beyond what they already had, and any modifications that would introduce more complexity would tend to be evolved out of them (as, indeed, it is to some extent in the cases of some parasites, parasites that evolve away the complexities that their remote non-parasite ancestors needed, but which, in their new lifestyle, they no longer need). I cover the relationship between physics and information theory in another essay, so I will not expand on it here.

But, I will add here one final point regarding a claim of some theists that is so stunningly mindless as to temporarily render one speechless. This is the idea that God, despite being omniscient and omnipotent, is absolutely simple, not complex at all, in any way. This, if it existed, would be a God free from specified complexity. But, it would also be a God absolutely devoid of the complexities of a personal God, because it could not have a mind, in any sense, a memory, since it could only store information by introducing changes in something and thus either making it more complex, or thoughts (which are more complex than a lack of thought), etc. It could not be aware of anything, it could not do anything (except in the way a rock can "do" something when it is dropped on someone's head, etc.), and so on. And, it most certainly could not design living organisms, since, by the claims of the ID folks themselves, the designer must be of a "higher" level of complexity than the complexity of what is designed (otherwise, complex living things could be "designed" by brute physical processes, a conclusion their whole argument is designed to avoid).

I have not seen this remarkable claim made by any ID folks, but I discuss it briefly here as a "pre-emptive response" in case someone might think it could be a means of weaseling out of the disastrous implications of ID's own premises while retaining at least the core of ID. It is no such thing.

Summary:

Intelligent design first asserts the requirement for an intelligent designer for things that have "specified complexity," and then denies it for the specified complexity of their own posited designer. They try to resolve this by resorts to eternally existing specified complexity, but that requires at least as much explanation as the specified complexity of an ordinary living organism, explanation that they don't have.

If they assert, as by implication they must, that the specified complexity of their designer is inherent in the nature of Existence, then they face the problem that this should mean that the designer itself is superfluous, since the specified complexity of Existence could explain the origin of life by evolutionary means without the need to resort to a designer (and, since, obviously, it would be an instance of specified complexity that doesn't require a designer). Since they do not allow that this can be the case, they can offer no coherent theory of the nature and existence of their designer because it either is an example of specified complexity that was not designed, proving that such specified complexity can exist without being designed, or it is not needed because the specified complexity of Existence as such (without any inherent intelligence) would then provide an explanatory basis for the evolution of life.

Feedback, discussion, comments, questions: Chris Cogan, ccogan@ou.edu