Monday, September 3, 2007

A Fatal Argument Against Intelligent Design Theory

Intelligent Design (ID) theory states that certain features of the universe, including the complex machinery of living things, are best explained by intelligent causation. ID proponents claim to have made this ancient theological argument into a rigorous and empirically grounded scientific theory.

Even if we pay close attention to how ID's arguments are framed, and give the benefit of the doubt on a number of issues, there are still no grounds for claiming that scientific evidence can point to an intelligent designer of life. By looking at how we actually investigate intelligent causation, and what we have found, we can see that the way ID tries to infer the mental abilities of unknown beings is not possible, and has no counterpart in any sort of scientific endeavor.

Complex Form and Function

In order to transform their arguments from natural theology into scientific theory, the promoters of ID have attempted to formalize a method for recognizing designed artifacts. According to ID, things can be analyzed to discover if they contain "complex specified information" (CSI). Objects with high levels of CSI are extremely unlikely to have occurred by natural processes, and correspond to a pattern that is independently recognizable; typical examples of things that contain CSI are the faces carved into Mt. Rushmore or the pyramids of Egypt. A fundamental claim of ID theory is that the presence of CSI must always result from the action of an intelligent agency, so CSI is a reliable indicator of intelligent causation.

ID critics have pointed out a number of problems regarding how CSI is defined. In general, it appears that nobody can actually compute the complexity (or improbability) of things we find in nature, or of many things that humans build either. How does one judge the likelihood, for example, that pyramids of stone blocks with tunnels and chambers - like those we see in Egypt - would form by means of natural processes? Still, these structures that humans build do not arise by any other means, and for the sake of argument I will grant that CSI simply means "complex form and function" that independent researchers can agree would not arise via known processes (fixed law and chance).

What will become clear, however, is that CSI is not necessarily a sign of intelligence at work.

Intelligent Causation

Here, ID theorist William Dembski explains how ID defines intelligence and its relationship to CSI:

To see why CSI is a reliable indicator of design, we need to examine the nature of intelligent causation. The principal characteristic of intelligent causation is directed contingency, or what we call choice. Whenever an intelligent cause acts, it chooses from a range of competing possibilities. This is true not just of humans, but of animals as well as extra-terrestrial intelligences. A rat navigating a maze must choose whether to go right or left at various points in the maze. [1]

According to Dembski, then, CSI can only be created by intelligent causation, which is characterized by "choice" - or, more accurately, free choice. One might say a river "chooses" a path to the sea, but this isn't the sort of choice Dembski is talking about, since the river does not have the capacity to act in any other way. Free choice implies the ability to do otherwise, in a way that is not determined by physical cause.

Philosophers call this ability "libertarian free will", and its existence has been the subject of debate for thousands of years. There is nothing resembling a consensus on either side of the issue: Nobody knows if human beings are capable of transcending physical causality when they make decisions, and science is unable to answer the question. [2]

Evolutionary theory, like every other well-accepted scientific theory, is not predicated on the truth of any particular proposition about metaphysics. The ultimate nature of mind, soul, free will, or God has no bearing on our scientific explanations. In contrast, ID theory critically depends on one particular metaphysical stance, viz that mental and physical causes are ontologically distinct. If mind does not operate in a way that transcends natural law and chance, then ID's entire notion of "intelligence" as a cause distinct from natural causes becomes incoherent. We can of course discern the activity of human beings or other animals, because we know the sorts of things we do. But if intelligence is just another name for certain types of natural cause - as most scientists who study intelligence believe - then ID's claim that we can distinguish intelligent cause in general from natural cause is nonsense.

But again let's give ID the benefit of the doubt, just for the sake of argument. Even though ID's conception of intelligent causation as free will is purely a metaphysical claim, without a shred of scientific evidence, let us accept this assumption that human beings do in fact have libertarian free will. Furthermore, let us grant that other things might have libertarian free will too, like the rat in Dembski's example. This class of things that transcend fixed law and chance - called intelligent agents in ID theory - is described like this:

Agents can arrange matter with distant goals in mind. In their use of language, they routinely ‘find’ highly isolated and improbable functional sequences amid vast spaces of combinatorial possibilities. … Intelligent agents have foresight. Agents can select functional goals before they exist. They can devise or select material means to accomplish those ends from among an array of possibilities and then actualize those goals in accord with a preconceived design and/or independent set of functional requirements. [3]

Once we accept arguendo that CSI can be reliably detected, and that agents which can plan with foresight do so with libertarian free will, an even more clearly delineated problem with ID becomes apparent: It turns out that CSI routinely arises by causes that demonstrably lack the ability to respond flexibly, invent novel solutions, or create plans by using foresight. In other words, CSI is not a reliable indicator of intelligence.

Design Without Free Choice

Consider the following artifacts: a cave dug into the earth with a door on swinging hinges fitted to the entrance, or rope that has been tied into lassos and woven into nets for use in hunting, or a highly structured complex of chambers and tunnels used for living quarters, with specialized, irrigated compartments for agriculture, and highly efficient ventilation systems. By the same reasoning that ID would apply to explain how the pyramids or Mt. Rushmore came to exist, ID would infer that intelligent agents were responsible for designing these examples of complex form and function.

While it isn't clear how one could actually compute the CSI of these artifacts (or the CSI of the pyramids or Mt. Rushmore) it does seem obvious that objects like these do not simply spring into existence by chance, nor from the action of wind, rain, erosion, geological events, or other known natural causes. It would seem that only the intervention of "agents" could explain the existence of these things. As Michael Behe, a principle architect of ID theory, explains:

If you and a friend walked by Mount Rushmore, even if you had never heard of it before, you would immediately realize that the faces on the mountain were designed. Not for a moment would you think they were the result of random forces such as wind and erosion. Your conclusion of design would be certain, because you would see how well the pieces of the mountain fit the purpose of portraying an image. [4]

Certainly this is true for the examples described above as well. We do not expect that hinged doors, woven nets and lassos used for hunting, and large, complex living compounds with built-in ventilation and irrigation systems could result from random forces. But for these particular artifacts the agents responsible might have been spiders or termites, which design these and other amazing artifacts every day.

Do these animals really design these things? Yes, in the sense that they cause these designs to exist. But do they reason, deliberate, plan, and reflect on the devices they build? Do they use imagination and foresight to satisfy their goals and desires? Do they use their conscious free will to choose where to deploy their webs and how to build their trap-door dens and architect their complex structures?

Scientific research indicates otherwise. Web-spinning spiders are born with fixed behavioral repertoires that do not change over their lifetimes and do not vary among individuals of the same species. They cannot learn how to create new types of artifacts, nor can they figure out solutions to novel problems. For example:

There is an automatic, almost robotlike character to many an example of instinctive behavior... a spider preparing to lay its eggs spins a silk cocoon in a particular way, always the same and without any regard for outside factors. She begins by building a base plate, then constructs the walls of her cocoon before laying her eggs within it and sealing it with a lid. So rigid are these behavioral patterns that they cannot be altered, even if the spider needs to do so....

If the spider is moved physically after she has built the base plate, she nonetheless will set about spinning walls and depositing her eggs, even though there is no base plate to hold them. The eggs, therefore, will fall out of the bottom of the incomplete web, but the spider will continue working as before, building the lid for the top. Fortunately for the spider, she has several cocoons. If she is returned to her original completed base plate as she prepares to spin the next cocoon, she will not use the base plate she already has spun; rather, like a robot, she will start from the very beginning, spinning a new base plate as though the original were not there. [5] (emphasis added)

It is always possible that despite their robot-like behavior, spiders could make free choices if they wanted to. Again, there is no way to demonstrate that anything actually has - or lacks - the power of libertarian free will. But if our criterion for attributing free will to something is the ability to generate novel solutions by planning with foresight, we must agree that the spider is a very good example of something that does not have intelligence in the sense that ID defines it.

Here is another well-known example of an animal performing complex tasks without the ability to make free choices:

When the time comes for egg laying, the wasp Sphex builds a burrow for the purpose and seeks out a cricket which she stings in such a way as to paralyze but not kill it. She drags the cricket into the burrow, lays her eggs alongside, closes the burrow, then flies away, never to return. In due course, the eggs hatch and the wasp grubs feed off the paralyzed cricket, which has not decayed, having been kept in the wasp equivalent of deep freeze. To the human mind, such an elaborately organized and seemingly purposeful routine conveys a convincing flavor of logic and thoughtfulness--until more details are examined. For example, the Wasp's routine is to bring the paralyzed cricket to the burrow, leave it on the threshold, go inside to see that all is well, emerge, and then drag the cricket in. If the cricket is moved a few inches away while the wasp is inside making her preliminary inspection, the wasp, on emerging from the burrow, will bring the cricket back to the threshold, but not inside, and will then repeat the preparatory procedure of entering the burrow to see that everything is all right. If again the cricket is removed a few inches while the wasp is inside, once again she will move the cricket up to the threshold and re-enter the burrow for a final check. The wasp never thinks of pulling the cricket straight in. On one occasion this procedure was repeated forty times, always with the same result. [6] (emphasis added)

We must assume that if the wasp was capable of choosing to bring the cricket straight in, it would surely make that choice, rather than endlessly repeating the same futile, automatic behaviors. The only reasonable conclusion we can draw from studies like these (and there have been many of them) is that these animals are incapable of free choice, i.e. they are not intelligent by ID's standards.

Apparently, then, Dembski is mistaken when he asserts that animals use intelligence to make choices, if he means all animals. And he is mistaken about something else as well: Despite the fact that animals like spiders and termites do not manifest free choice, they are still capable of producing complex designs that appear to be just as "specified" as the things that humans or other intelligent animals create.

This seems to be a stark and fatal counter-example to the central claims of ID theory. CSI is supposed to be a reliable indicator that something was created by intelligent agency. But these examples show that artifacts containing CSI can also be produced by automatons following blind instinct. If these complex artifacts can arise by means other than intelligent agency, then Dembski's theory of how CSI arises must be wrong, and CSI cannot be used to infer intelligent causation. And if this inference fails, then all of Intelligent Design Theory fails with it.

How do ID theorists respond to this apparent disconfirmation of their theory? Following are a number of objections I've received from ID proponents when presented with this argument.

Objection #1: Intelligence as ultimate cause

The first objection that an ID theorist will raise is that even if the immediate cause of this CSI is not intelligent agency, intelligence must have played a part at some point in the chain of causes leading up to these artifacts. This is a facile dodge, of course: Whenever the cause of CSI is observably intelligent, then IDists say their claim (that only intelligent agents can generate CSI) is confirmed. But if the cause of CSI is not intelligent, then IDists refuse to admit their claim is refuted, and instead they simply insist there must be some unseen intelligence who designed the unintelligent designer that we observe!

But once again let's give ID a pass for the sake of argument, and see why IDers insist that some intelligent agent must have designed the spider to design the web. They argue that the spider is analogous to a robot that manufactures cars: The robot itself is not intelligent, even though it is capable of assembling complex machinery. Instead, the intelligence comes from the robot's programmer. In the same way, while the spider is not intelligent per se, whatever created the spider must have been. ID theorists would call the spider, like the car-building robot, a "CSI conduit".

A conduit can transmit CSI from an intelligent agency into an artifact. ID places no limits on what CSI conduits can produce, except that conduits cannot produce an increase in the total amount of CSI. This follows from what Dembski calls the "Law of Conservation of Information", which he places among the other conservation laws established by physics, like the conservation of mass/energy. Information can be transformed by a conduit, but whatever CSI comes out of a conduit must have already existed, created by intelligence.

This all sounds very scientific, until one realizes that what is being said can't possibly be evaluated by scientific means. Here's why: Obviously, nobody has ever observed this hypothetical intelligent being creating various species of spiders and programming them to build webs. So for all ID can tell us, the spider itself could have been the product of yet another CSI conduit. Some mindless process could have been responsible for producing the spider, just as the spider's mindless behaviors produced the spider web. And how might that mindless process have come into existence? Again, for all ID can tell us, that process could have in turn been the result of yet another mindless process - and so on.

ID can't specify where in this causal regress intelligence may have intervened to create the CSI that ends up in the spider web. Still, ID insists that it must have intervened somewhere, even if it was at the very first step in a chain of mindless causes (ID calls this possibility "front-loading"). If there was never any intelligence involved, says ID theory, then the CSI could never have come to exist, lest the law of conservation of information be violated. ID holds that no chain of CSI conduits could ever have gotten started without a "first cause" that was intelligent.

But can ID validly claim that this is a scientific inference, supported by the evidence? The answer is "no" - and it is none other than the ID theorists themselves who point out why! In order to escape the difficulty that ID faces in explaining "Who designed the Designer?", ID proponents contend that questions relating to ultimate cause are not amenable to scientific inquiry:

[E]vents like the origin of the first cause are so far in the past that they are not accessible to science. They can only really discuss them using philosophy and religion. Here, essentially, this question is about the "first cause"--and physicists have discussed that when we are dealing with things very early in the universe, or predating the universe, like the origin of the first cause, the normal laws of physics and reality can break down. [7]

So on one hand, ID theorists deny that ultimate causes are accessible to science. Our understanding of how reality operates, including fundamental laws such as the conservation laws, are not applicable when speaking about what happened at the moment the universe began (or "before" that - whatever that might mean). But on the other hand, when it suits them, IDists want to say that scientific principles can be applied to ultimate causes, and insist that their law of conservation of information must necessarily hold even at the beginning of time, to show that an original intelligence was required for CSI to exist in the universe.

Obviously ID wants to have it both ways, but something has to give. Either ultimate causes are scientifically accessible or they are not. If not, then ID can hardly turn around and insist it demonstrates something about ultimate causes anyway. But if ultimate causes do fall within the realm of scientific inquiry, then IDists are faced with the question that they very much want to avoid: How could some intelligent agent simply pop into existence without some prior cause?

It is no more plausible to imagine that some intelligent agency somehow popped into existence and proceeded to design the various natural mechanisms that resulted in spiders and spider webs than it is to imagine that these CSI-loaded mechanisms popped into existence all by themselves. If anything, the latter is at least a simpler hypothesis. The original cause of CSI may have operated like a "super-human", using its built-in powers of thought and foresight, or it may have operated like a "super-termite", blindly following its built-in instincts without any awareness of what it is doing. In any case, science is unable to adjudicate the matter, as ID theorists readily admit.

Dembski's notion of conservation of information does not entail that an intelligent entity created the original CSI any more than conservation of mass/energy implies that such an entity created matter, so this appeal to prior cause (up to and including ultimate cause) fails. ID can provide no support for the supposition that the CSI in the spider web (or the spider) required intelligent intervention at any step in the causal regress. The idea of an intelligent "first cause" remains in the realm of theological speculation, just as it has been for thousands of years.

Objection #2: More complex designs imply more sophisticated intelligence

Some ID proponents have argued that while non-intelligent animals might actually produce fairly complex designs such as termite mounds, the designs we see in biological systems themselves are far more complex. In order to be able to design something as complex as a living thing, surely a mind with free choice and great powers of foresight and planning would be required.

But this argument does not stand up to scientific scrutiny either. The fact of the matter is that our study of intelligent agents has revealed that intelligence (as described by ID) does not correlate with the ability to create complex designs.

Consider apes such as chimpanzees, gorillas, orangutans and bonobos. These animals display many signs of intelligence in both controlled experiments and field observations. They can be seen to invent novel solutions to problems, and make plans in advance to reach their goals. Yet the most advanced artifact they produce is a twig that has been stripped of leaves, used to fish termites out of holes. These tiny termites, however, produce very complex and functional structures as we've seen, while displaying virtually no intelligence (as ID defines it) at all. Other animals displaying relatively high intelligence, such as crows, dolphins and whales are also incapable of designing complex artifacts, while bees and wasps build things of complex form and function in a mindless, robotic manner.

Does this mean that termites are more intelligent than gorillas? Of course not. We humans happen to be capable of producing the most complex designs on Earth, and we happen to be intelligent too. But one data point does not establish a correlation, and ID offers no reason why such a correlation should exist. Indeed, the essential error that ID proponents commit is anthropomorphism: When there is something about nature we don't understand, they assume that it must have come about by means of applying reasoning and knowledge, simply because that is the way a human being would approach a problem.

For example, think of how much intelligence would be required for a human being to create a large, high-voltage arc of electricity. A person would need to be quite proficient in electrical engineering in order to produce such a phenomenon. But thunderclouds produce high-voltage arcs (lightning) on Earth every day, and nothing in the clouds understands the first thing about electrical engineering [8]. Similarly, if human scientists set out to create a spider web, a good deal of knowledge and planning would be required to develop it; spiders, however, build them without a thought.

Over and over we see that human beings using foresight and planning to achieve some result does not imply that the same phenomenon requires foresight and planning when humans are not involved. Just because a human being would have to be very intelligent to create a living organism doesn't mean that intelligence was required in nature.

Objection #3: Show me the mechanisms!

A particularly ironic objection comes from those who demand that we must be able to explain how biological complexity comes to exist in order to discount ID theory. Obviously, the vast majority of scientists believe that we already have an explanation of how complex form and function arises in biology. But even if we assume that evolutionary biology cannot account for biological complexity, we still have no reason to believe that intelligence was involved. None of my arguments here have anything to do with evolution, but they still make clear why intelligent design is unwarranted by the evidence.

Moreover, those who enjoin scientists to provide details regarding how natural mechanisms manage to produce spiders fail to appreciate that invoking the words "intelligent causation" is no more a detailed explanation than appealing simply to "unspecified natural processes". Nobody can explain how human beings manage to solve problems, and of course nobody can begin to explain how anything can exert free will. If we were to consider "intelligent causation" as a scientific explanation, without any further accounting of what that entails, then we would have to consider "fixed law and chance" as an equally informative alternative theory, without even trying to explain why. Of course neither "intelligence" nor "fixed law and chance" constitute any sort of helpful explanation at all. But for ID enthusiasts to require detailed models of natural processes while denying that they are obliged to even attempt to explain the processes they propose is an especially clear example of the fallacy of special pleading.

And simply defining intelligent causation as a fundamental, irreducible causal force in the world, as some ID proponents would suggest, doesn't help. It is of no use to posit an irreducible force unless we can characterize that force in some specific manner. All other fundamental, irreducible forces that serve as explanatory constructs in scientific theories are carefully and precisely characterized in ways that allow independent demonstration of their existence.

If, for example, Newton had explained why apples fall to Earth merely by concocting an irreducible "falling force" defined only as that which makes things fall to Earth, his theory would have been worthless. Instead, he precisely characterized the force of gravity: It acts instantaneously between any two masses, with a force proportional to a universal constant and to the inverse of the square of the distance between them, and so on. Armed with these testable characteristics, scientists could actually see if Newton's gravity existed as it was described.

Newton's detailed description of gravity made it possible for the first time to demonstrate that the same thing that causes apples to fall causes the motion of planets and moons as well. In contrast, there is no description of intelligent causation that would allow us to decide if whatever enables humans to design a car or a watch is the same thing that causes biological complexity. ID simply attributes to "intelligent causation" any and all powers that might be needed to create living organisms. Because there is nothing that this hypothetical causal force cannot do, and nothing specified as to how it might operate, we have no way to determine if we see its effects or not.

To be sure, cognitive scientists have made great strides in learning about how humans and other animals perceive and reason about the world, but there is no general theory of intelligence that defines what it is across all possible intelligent agents. ID theory has made no contributions to cognitive psychology, nor any other discipline connected with the study of mind; books on ID typically don't even reference the huge literature on scientific studies of mentality. This is to be expected, however, since what the research has shown is notably inconvenient for ID. It is abundantly clear that our mental abilities depend critically on complex, physical mechanisms capable of processing information: our brains.

While philosophers and theologians might argue whether or not brains are sufficient to produce all of our mental experiences and abilities, there is a vast body of evidence to confirm that a brain of some sort is required for us to be able to design anything. And it's also clear that if intelligent causation cannot proceed without complex physical mechanism, then one cannot logically suggest that intelligence was responsible for the origin of complex physical mechanism in the first place.

Objection #4: If science cannot infer intelligent agency from complex artifacts, then archeology is unscientific

ID proponents often argue that science already performs "design detection", and that ID is simply extending those very same and well-accepted methods to the study of biological designs. But this is simply false, as many others have already pointed out.

When we see the faces carved into Mt. Rushmore, or the pyramids of Egypt, no reasonable person would assume that these structures arose by chance or were caused by something like erosion. When archaeologists find a shard of pottery, or a spearhead, they assume that these artifacts did not arise by wind, rain or geological events. In each case, of course, we conclude that these things were created by human beings.

Contrary to what the proponents of ID theory say, we do not infer that "intelligent agency" was responsible. Nobody thinks that intelligent agents like rats create arrowheads, pottery, pyramids, or carved faces. Nobody would imagine that gorillas or dolphins or parrots would have built these things either. And if someone suggested that these artifacts were constructed by invisible supernatural beings, we might suggest that a psychological evaluation would be in order. So instead of positing that an "intelligent agent" was responsible, by virtue of our knowledge of human beings we can recognize that human activity was the cause.

Likewise, when ethologists find a termite mound deep in the jungle, they do not infer that human beings built it, but because of what they know about termites, they can comfortably identify the cause. This again makes clear that the inference is not to "intelligent agency", but rather to some specific type of agent (animal) as the cause of these objects. The pyramids were built by human beings, who happen to be intelligent agents, but in the case of the termite mound, the agents responsible do not happen to be intelligent in ID's sense of the word.

Scientists never infer mind from complexity; they only infer specific agents likely to be responsible for specific artifacts. The artifacts may be simple or complex, and the agents may or may not be intelligent. So, quite unlike ID, archeology, forensics, and cryptanalysis do not attempt to infer intelligent causation from CSI in artifacts. The only way that science ever infers that something has the power of foresight is to observe actual behaviors in different situations. It is by observing behaviors that we know that humans and gorillas can solve novel problems, and that termites and spiders cannot.

At this point, ID enthusiasts invariably bring up SETI, the search for extra-terrestrial intelligence, as an example of a scientific discipline that attempts to infer intelligence from complexity (i.e. complex signals received from space). But SETI researchers explain that their approach is completely different from what ID tries to do. While ID looks at living things in nature and tries to infer intelligence, SETI looks for things that are not found in nature and tries to infer life.

SETI does not look for complex signals bearing intelligent messages; instead, they make the assumption that human-like life forms may exist elsewhere in the universe, and look in places where life might exist to find signals similar to those human beings would transmit for the purpose of announcing our presence. In order to rule out other causes, SETI picks a particular type of signal that we do not observe arising from astrophysical phenomena. If we ever encounter a signal like this (likely to be a simple, narrow-band transmission, perhaps with no encoded symbolic information at all) and if the signal arose from a planetary system thought to be hospitable to life, then SETI might suggest that a human-like life form was actually responsible.

Since we already know that human beings happen to be capable of foresight and planning, it would be reasonable to hypothesize these extra-terrestrial life forms might be intelligent too. If, on the other hand, the signal arose from within a neutron star, we would probably not assume that any life form was responsible. We would simply say that the signal was the result of some unknown process, and there would be no reason to think that this process was intelligent. [9] Obviously, the latter situation applies to the context of ID, since life forms are not candidates for the cause of original life. [10]

Objection #5: Teach the controversy!

When all other arguments are refuted, the committed ID proponent will remind us that their theory is not intended to be a deductive argument proving intelligent design - it is an abductive argument, merely reasoning to the best explanation supported by the evidence. Even if the science can't confirm that an intelligent designer was involved in the creation of life, neither can it prove that it was not. Both hypotheses ought be considered along with their strengths and weaknesses, and each of us should decide on our own what to believe.

The problem with this is that ID fails to show any evidence that could confirm the idea that living things were designed by an intelligent being; all that is presented is attacks on other theories, and specious arguments. As we've seen, the only evidence that ID presents - complexity in living things - is not evidence for intelligent causation at all. The analogies to archeology and SETI are invalid, and all of the mathematical exercises with CSI intending to show that intelligence is the likely cause are moot. And needless to say, there is no direct evidence whatsoever, as there are no observations of any candidate intelligent entity (at least no replicable observations).

Moreover, there are no predictions that arise from the bare assertion that "something intelligent" created life. Without saying something about the nature of the Designer (something that ID is careful to avoid), then obviously we have no way to generate a set of expectations of what He would or wouldn't design, and so we can't evaluate what we find against our expectations. Intelligent agents act upon their wants, needs, beliefs, desires, likes and dislikes, and without positing some explicit model of what these might be in the context of ID, we can't predict what the Designer would have done.

Surprisingly, some have suggested that ID can make predictions after all. ID is sometimes said to predict, for example, that function will be found for "junk DNA", on the basis that intelligent designers wouldn't leave non-functional codes in their creations [11]. But this does not follow from merely asserting that a designer is intelligent; again, it is an anthropomorphic projection to suggest that the Designer of Life wouldn't want non-coding DNA cluttering up genomes just because human beings wouldn't. And worse yet, everybody knows that human designers often leave non-functional code in their designs - ask any software engineer who has worked on a complex project.[12]

Of course IDists could bite the bullet and admit that one of their assumptions is that the Designer is truly human-like (or that humans were created in His image). That might arguably lead to a number of predictions that could be tested. Unfortunately for ID, however, these predictions fail miserably, because the machinery we see in living systems look nothing like the creations a human would engineer. Once ID allows us to make predictions based on what a human would do, we can predict that we would never find, for example, the vagus nerve looped around the aorta in mammalian anatomy - a decidedly inefficient arrangement. [13]

Traditionally, ID has sidestepped these "arguments from sub-optimality" by declaring that we cannot possibly know what the Designer had in mind when He jury-rigged these odd solutions. But once again, we see that ID is trying to have it both ways: Either we can make predictions about what the Designer would do, or we can't. If we can infer what we expect of the Designer based on human values and proclivities, then it is perfectly reasonable to consider the often baroque designs of living things to count against the idea that any human-like intelligence was involved. On the other hand, if ID denies that we can discern the motivations of the Designer, then we can't possibly test our predictions.

The complexity of living systems tells us nothing about how they came to exist. If we discount the explanations that biologists believe are supported by the evidence - that is, evolutionary mechanisms - then the only answer we have is "we do not know". Humans have historically offered intelligent agency as an explanation for anything we do not understand, but it has been shown over and over again that the strategy of filling the gaps in our knowledge with intelligent agency does not lead to increased understanding. It turned out that planets are not guided through the heavens by demi-gods after all, and that no god of thunder aims lightning bolts at church steeples. As a result, science instituted the eminently reasonable policy that these explanations - always available to explain everything, but never independently testable - are not something that we can justify as empirically grounded theories.

Even if current evolutionary theories were wrong or radically incomplete, we still would have no way to evaluate the truth of the proposition of intelligent design. There is no observable evidence of the existence of a primordial intelligence capable of designing life, just as is there is no evidence that human intelligence operates by means of anything but physical law. These are beliefs that cannot be evaluated against our shared experience, which is why after thousands of years of speculation about these issues they remain purely matters of opinion, or faith, and why no consensus has ever been - or will ever be - reached. In contrast, scientific theories have been generated and confirmed to explain a huge number of natural phenomena, and people all over the world accept these scientifically justified explanations despite radical differences in religious and metaphysical views.

People who wish to confer scientific status upon the argument from design covet the authority of science; otherwise, IDists would be content with the theological arguments as they stand. But the authority of science must be earned by providing objective evidence that can be confirmed by independent researchers, and providing means to test our inferences based on our observations.

Objection #6: The tu coque fallacy

Finally, faced with the prospect of having to once again abandon their attempt to position theism (or something quite like it) as science, supporters of ID and creationism fall back on the most wrong-headed argument of all: Evolutionary theory, they say, isn't supported by the scientific evidence either! But whether or not there is any scientific support for evolutionary theory is completely irrelevant to the question of ID's status as a scientifically supported, testable theory.

Proving one theory wrong, or inadequate, does not confer scientific validity to any other idea, for the simple reason that they both could be wrong, or incoherent, or untestable. Scientists are constantly challenging various aspects of evolutionary theory, and they always have, so ID brings nothing new to the table in this regard. If ID enthusiasts want to try their hand at attacking evolutionary theory, they can get in line behind all of the qualified researchers and theorists who test, evaluate, and revise our theories every day. But if they want to propose an alternative theory, they are obliged to describe what it is they think might be responsible. Alluding to the metaphysical notion of free will, and providing no empirical evidence to suggest that it exists - much less that it was involved in the origin of the species - does not qualify as a testable alternative theory. It is a non-starter.


ID proponents like to play victim, and pretend that their lack of actual research is due to an evil conspiracy of elite scientists who are afraid of losing power to a new paradigm. But this same gambit is used by every peddler of crackpot theories, perpetual motion machines, and snake oil remedies.

There is nothing to ensure that our current theories are correct, and there is nothing to ensure no new theory could unseat even our most well-established scientific ideas. Of course there is resistance to new ideas, but good ideas win out despite human foibles by virtue of the institutions and protocols that keep science honest in the long run. Complaining that mainstream science censors ID appeals to our sense of injustice, but it does not mean that ID actually has anything commending it for inclusion in the first place.

Science is a human endeavor, and scientists make mistakes, come to erroneous conclusions, and even exhibit bad behavior. But none of these obvious points argue against the effectiveness of science as our best method for generating knowledge. Nobody has described any better system of vetting our beliefs, and ID supporters come dangerously close to advocating relativism, where each individual's opinion or religious belief should count as strongly as the consensus of the worldwide scientific community. Everyone is free (in this country, at least) to believe whatever they would like to about how the universe was created. Nobody is free to co-opt our scientific institutions to lend authority to their particular religious beliefs.

Evolutionary theory might be right, or it might be wrong, but despite the rhetoric of the ID movement, it is not a religion. The mechanisms proposed to account for biological complexity (mutations, differential reproduction, and so on) actually do exist and can be observed, whether or not one believes they fully explain speciation. Neither God nor libertarian free will can be observed to do anything at all, no matter how many people fervently believe otherwise.

1. Intelligent Design as a Theory of Information, William Dembski
2. Experiments designed to test the idea that humans' conscious decisions somehow control their actions without prior physical cause have actually suggested (but by no means proven) the opposite: It appears that we become conscious of our "voluntary" behaviors only after our brains initiate the actions. See, for example, the work of Benjamin Libet, or recent work by Daniel Wegner.
3. Debating Design - From Darwin to DNA, pg. 388. Note that ascribing these abilities to "free will", rather than attempting to understand and explain them by conducting research in the cognitive sciences, is yet another example of how ID surrenders to ignorance.
4. See here for Behe's comments
5. See here for this and other examples of blind instinctive behaviors
Woodridge, D. (1963). The machinery of the brain. New York: McGraw Hill. p. 82
ID Center FAQ
8. At least we know
now that lightning does not require thought; it was not very long ago that people were convinced that the intervention of intelligent agency was necessary for lightning as well (see here for details). How else, they reasoned, could lightning be so effectively aimed at church steeples, if not for some agent who could look around from the heavens, purposefully choose a target, and aim the lightning bolt at its intended destination? This is only one of many failures of ID theory historically applied to natural phenomena.
9. See here for SETI researchers' own refutation of any analogy between SETI and ID's "design detection".
10. Of course, alien life forms could be suggested as the source of life on Earth; this is an old idea (promoted at one point by Francis Crick) that suffers from none of the same problems discussed here regarding "intelligent agency". The hypothesis that life was transplanted to Earth from elsewhere has simply languished for lack of evidence, and because it fails to answer the question of how life originated in the first place.
11. Such predictions are explained here. This is a particularly safe "prediction" of course: If function is found for some non-coding region of DNA, ID can claim a confirmation, but if none is found, it can always be said that we simply aren't clever enough to have deciphered it yet.
12. Quite embarrassingly, ID proponents have made this very same observation when arguing against the charge of sub-optimality in biological mechanisms. On one hand, the Designer can be predicted to eliminate junk in the genome, but on the other hand, junk in the genome does not discount the hypothesis of intelligent design!
13. Here is a list of how some obvious predictions based on human design principles are refuted by the biological evidence of (sometimes comically) sub-optimal design

Saturday, July 21, 2007

A Discussion of Origins

aiguy: I think that life was created by an intelligent computer.
IDguy: How do you figure?
aiguy: Intelligence was needed to create the first life. And it obviously couldn’t have been biological intelligence. The only other kind of intelligence we have any experience with is computer intelligence.
IDguy: But computers need to be programmed by a real intelligence in order to do anything intelligent. Who do you suppose programmed this intelligent computer?
aiguy: Hmm, good point. Ok, what’s your theory?
IDguy: I believe in ID theory. An Intelligent Designer created life.
aiguy: But… the designer must have itself been designed, right? Who do you supposed designed this intelligent designer?
IDguy: You can’t ask that of ID - is not in the scope of the theory! ID is strictly scientific. We leave questions of ultimate cause to religion.
aiguy: Wait a minute. How come it’s OK for you to ask “Who programmed the intelligent computer?” but it’s not OK for me to ask the “Who designed the intelligent designer”?
IDguy: This is absurd. Only people build computers. Before there were people around, there couldn’t have been any computers to design life.
aiguy: That’s true. But only living things are intelligent, so before there were living things around, there couldn’t be any intelligence that could design life.
IDguy: How do you know only living things are intelligent? Maybe something could be intelligent that isn’t a living being.
aiguy: How do you know only people make computers? I think it’s more likely that some sort of computer was around before life started than it is for some disembodied thing to be able to design life. How can the Designer be so brainy when it has no brain? How does it do its handiwork with no hands?
IDguy: Who said anything about being disembodied?
aiguy: If it didn’t have a biological body, then what sort of body did it have?
IDguy: Ok, it was disembodied. But it is no more unlikely for a disembodied intelligent agent to have designed life than for a computer to have done it.
aiguy: In other words, Intelligent Design Theory is no better than Intelligent Computer Theory?
IDguy: Well… Let’s teach the controversy! I say an Intelligent Designer designed life, and you say a computer did it. Who do you think most people will believe?

* * * * *

aiguy: Neither of these theories is any good at all. Just because people already believe in God doesn’t mean this ID theory stands up to scientific scrutiny.
IDguy: Who said anything about God?
aiguy: If it isn’t God, who is this Intelligent Designer?
IDguy: Ok, it’s God. Once you believe in God, ID theory is very appealing.
aiguy: What if I don’t believe in God?
IDguy: Then you must be a moral relativist, and you worship only money. You have no purpose in your life. Oh, and you probably won’t believe in ID theory either.
aiguy: No, I also worship Barack Obama and the liberal media... but that’s beside the point. I thought ID was supposed to be scientific. How can you test this Intelligent Designer hypothesis?
IDguy: I can scientifically demonstrate that random mutation and natural selection is incapable of producing the complex specified information and irreducibly complex structures that we observe in biological systems.
aiguy: But my theory is about an intelligent computer, not evolutionary processes.
IDguy: Well, who programmed the computer?
aiguy: You already tried that one.
IDguy: Here’s what you don’t understand. Intelligence is a fundamental, irreducible force in the universe, like gravity. You can’t detect it directly; you can only detect its effects. That is what is responsible for the complexity of living things.
aiguy: Anybody can make up a force and say it explains whatever they want it to explain. My force is called “orbis” and it explains crop circles.
IDguy: How does “orbis” explain crop circles?
aiguy: Orbis is a fundamental, irreducible force of the universe, like gravity. You can’t detect it directly; you can only detect the crop circles.
IDguy: You don’t really believe all of this stuff you’re saying. There’s no such thing as the “orbis” force. You’re just playing word games.
aiguy: I don’t believe in the orbis force, no. I was just trying to show that inventing a force to explain something isn’t helpful unless there are some specifics to back it up.
IDguy: Maybe, but evolution can’t back up what it says either. Nobody can show macro-evolutionary changes result from mere chance. That’s so dumb, it makes you wonder why these intellectual elite scientists think they’re so smart.
aiguy: But my theory wasn’t about evolution. It was about…
IDguy: Cut it out! You don’t believe an intelligent computer designed life either!
aiguy: You’re right, I don’t. But just because you think you’ve shown that one theory is wrong doesn’t mean some other theory has to be right. Even if evolution fails to account for life, it doesn’t make your ID theory any better. It’s not one bit better than my ridiculous Intelligent Computer theory. And it’s not any more helpful than my “orbis” theory, either.
IDguy: If evolutionary theory is wrong, and ID theory isn’t any better than these silly things you made up, what do you think we should teach our children?
aiguy: Who cares? I’m an atheist – I ate my kids.

* * * * *

IDguy: I’ve figured it out. ID theory is true, but if you want to think that the Intelligent Designer was a computer, that’s ok. The identity of the designer is outside the scope of ID theory, and if you want to think it’s a computer, you go right ahead. You’re still under the Big Tent of ID. All you have to believe is that intelligence was involved. Maybe it was an extra-terrestrial life form, or maybe it was a computer, or maybe it was God. Whatever.
aiguy: But if it was an extra-terrestrial life form, then ID theory still wouldn’t explain the origin of life. And if it was a computer, then ID theory still wouldn’t explain who programmed the computer.
IDguy: I’m not going to talk about that. You can believe whatever you want to. All I’m saying is that whatever it was, it was intelligent.
aiguy: But no matter what I believe about the Intelligent Designer, ID theory fails to explain how life really got started! That is, unless I believe in God.
IDguy: Well, you might want to rethink your rejection of the Lord.
aiguy: So I should believe in God because then ID theory makes sense, and we can explain the mystery of life?
IDguy: Works for me. Once you understand that only God can have created life, everything makes sense.
aiguy: Even crop circles?
IDguy: No, not crop circles. Those really are by aliens.
aiguy: You’ve given me a lot to think about, IDGuy. God might just fill that void in my heart, and help me to quit abusing small animals.
IDguy: And it’s good science, too!

* * * * * (some time later)

IDguy: So, aiguy, how has your ID research been going?
aiguy: It’s been going very well. Once I accepted the design hypothesis, it has been a powerful heuristic for approaching the most difficult problems in science. I’ve figured out how the universe began, why the physical constants have the values they do, why we have such nice solar eclipses on Earth, and why there are so many different kinds of beetles. It turns out that in each case, the answer was intelligent causation.
IDguy: That’s great. I’m currently working on a theory of how proteins fold. My working hypothesis is that intelligent causation is responsible.
aiguy: Sounds like you’re on the right track.
IDguy: Are you still playing around with those computers?
aiguy: Yeah. I built a system that has figured out how to beat the stock market. I use the profits to fund my ID research.
IDguy: Obviously you mean that you figured out how to play the market - the computer is just a tool, remember? It didn’t figure it out all by itself!
aiguy: I guess so. I don’t really know anything about stocks, though. I just let it randomly develop different strategies until it was doing better than any human expert. It came up with such a complicated method for picking stocks – I still don’t understand how it does it.
IDguy: Obviously it just used a brute-force, mindless exploration of the alternatives – it wasn’t actually thinking about the stock market!
aiguy: Still, if I didn’t know how it worked, I would swear that a financial genius was developing those strategies. It seems so… intelligent…
IDguy: Perhaps, but this is merely apparent intelligence, not real intelligence.
aiguy: You mean like the design of living things is merely apparent design, not real design?
IDguy: No, no – that’s real design, not apparent design.
aiguy: I’m a little confused now. How do we tell real intelligence from apparent intelligence again?
IDguy: You see, real intelligence involves the ability to plan, deliberate, choose from among options, and work toward a goal. Your program just mindlessly tries different things, keeping those aspects of its strategy that work and discarding the rest, until it could predict stock market fluctuations. Undirected processes like that don’t count as real intelligence.
aiguy: But what if we didn’t know that’s what it was doing? What if we only saw the end result – a brilliant strategy for picking stocks that even the smartest financial experts can’t match? How could we tell from just these observable effects that a mindless mechanism was involved, rather than a real intelligence?
IDguy: Your computer didn’t make up its own goal – you put the goal in the program. Your “fitness function” to test for beneficial changes was explicitly programmed by you. The computer didn’t decide all by itself that it wanted to play the stock market.
aiguy: It’s funny you should say that. As it happens, I had built a rich simulation of the economy into the computer, including all sorts of ways to make money – manufacturing, service industries, stocks, bonds, money markets... I didn’t know which one it would pick, but it ended up focusing on the stock market for some reason. It’s like it decided all by itself.
IDguy: Mindless, mechanical processes can’t decide anything by themselves.
aiguy: But I certainly didn’t decide what the program should do, and I couldn’t possibly have predicted what it would choose. If I didn’t decide, and the computer didn’t decide… who decided?
IDguy: Anyway, your simulation of the economy wasn’t realistic – you couldn’t have modeled anything as complicated as the actual economy.
aiguy: True. Every model must necessarily simplify to some extent. Even when simulations are used to test new aircraft designs, it’s not like every molecule of air has been modeled in the simulation. Still, aircraft that “fly” in the simulations do actually fly in the real world. And my program for picking stocks certainly has done well…
IDguy: Well, I’m glad you’ve made a lot money to fund further ID research. What’s your next project?
aiguy: I’ve just built a simulation of Earth, and I’m using my random generate-and-test programs to model the Intelligent Designer. So far, this virtual Intelligent Designer has built some pretty interesting virtual life forms.
IDguy: How can you be modeling the Intelligent Designer using random generation and test algorithms? That is the antithesis of intelligent designing!
aiguy: I know. But it seems to work. There isn’t really any intelligence in my simulation at all – just a few simple rules for coming up with random variation and testing the results against the virtual environment. And these complex, functional, virtual life-forms evolve… Do you think the same sort of the thing might account for real biological systems?
IDguy: Nope.
my space counter