The Failures of Mathematical Anti-Evolutionism
Download 0.99 Mb. Pdf ko'rish
|
The Failures of Mathematical Anti-Evolutionism (Jason Rosenhouse) (z-lib.org)
(Dembski 2004, 87)
In this scheme, “complex” indicates that the event occurs with low probability, while “specified” indicates that the event conforms to some independently describable pattern. Dembski’s basic idea is readily explained through a few examples. Suppose we flip a coin 20 times and obtain this sequence of heads and tails: H H T T T H T H T T T H H H T H T H T T How likely were we to get this exact sequence? Using what we learned in Section 5.2, we would say that each toss is independent of the other tosses, meaning we can multiply the individual probabilities. Since each toss comes up H or T with probability 1/2, the probability of this sequence will be 1/2 multiplied by itself 20 times. That works out to less than one in a million, which for illustrative purposes we can take to represent a low probability. However, this low probability by itself does not make us suspect trickery, since events of low probability occur all the time. This exact sequence was extremely unlikely, but something had to happen. Now suppose that we got this sequence instead: H H H H H H H H H H H H H H H H H H H H This exact sequence has precisely the same probability as before: less than one in a million. But now we have an obvious pattern as opposed to some random jumble of heads and tails. In Dembski’s terminology, we now have both complexity and specificity, and this makes us suspect some sort of trickery. Perhaps the coin was tossed by a skillful sleight of hand artist who was subtly manipulating how the coin would land, or perhaps the coin was weighted in a way that made it almost inevitable that it would land heads. We would suspect 134 5 probability theory an explanation of that nature, rather than think the event occurred from a fair coin flipped in a fair way. We could make the same point with a deck of cards. Shuffle the cards and deal them out on a table. Reasoning as with our coin tosses, the probability of getting that exact sequence of cards was very small, 1/52! to be exact, but we are not yet suspicious, because some sequence of cards had to appear. Looking now at the entire 52-card sequence, we might find, say, two consecutive aces or three consecutive hearts. Those short runs have specificity, but we do not have complexity, since we expect short patterns like that to happen by chance with high probability. But if the 52 cards have sorted themselves by suit, with each suit running in sequence from ace to king, then we have both complexity and specificity, and again we suspect some sort of trickery. We could also imagine throwing a large number of lettered tiles onto a table. We would not be surprised if simple two-letter or three- letter words appeared just by chance. They are specified, in the sense that they are recognizable as English words, but they are not complex, since a few short strings of that kind will occur by chance with high probability. A lengthy, nonsensical string of letters might be complex, but since it is not specified we do not yet suspect design. But if the tiles form a long English sentence, then we have both complexity and specification, and we once again suspect some sort of trickery. If a friend told us he did this experiment and that the tiles just happened to spell out “It was a dark and stormy night,” then we would suspect he was not telling us the whole story. This sort of reasoning arises in many situations. Any pattern of crags and grooves on a mountain is extremely unlikely, but the faces on Mt. Rushmore are also specified, and therefore we quickly identify them as having arisen from design. An archaeologist can quickly distinguish stone tools from random rock formations, since the former exhibit arrangements of parts that are unlikely to occur by chance, and since the arrangement is specified by virtue of being 5.6 complex specified information 135 useful for a clear purpose. A large number of metal gears and springs can be arranged in many ways, but if we are confronted with an arrangement that makes a functioning pocket watch, then we know we have an instance of intelligent design. We could multiply such examples endlessly. Dembski first presented these ideas to the public in his 1998 book The Design Inference (Dembski 1998). To this point, the argu- ment has nothing to do specifically with biology. Dembski’s intention was to provide a rigorous and general method for detecting whether design was involved in the causal history of any event, whether drawn from nature or from human history. After presenting his framework, he goes on to argue that he can apply it to biology. He claims that we can take for our event the appearance of some complex biological system and apply his method to it. When he does so, he continues, the result amounts to a mathematical proof that the system is in some way the result of intelligent design. For example, we might take as our event a complex structure like the flagellum used by some bacteria to propel themselves. It is a complex biomolecular machine that involves numerous proteins working in concert. Following Dembski’s lead, we might say, “The probability of that specific arrangement of parts arising by chance is extremely small, and the flagellum is specified by virtue of the fact that it is a functional machine that operates in much the same way as an outboard motor on a boat. The flagellum is therefore both complex and specified, and this implies that it cannot be explained without some recourse to intelligent design.” As it happens, the flagellum is the main example Dembski uses to make his case for ID, as we shall discuss in Section 5.7. Though it will not be relevant to the ensuing discussion, we should note for completeness that Dembski puts forth 10 −150 as the probabilistic threshold for complexity. In other words, he argues, based on various criteria drawn from physics, that an event with a probability smaller than one in 10 150 should be considered complex for the purposes of his framework. 136 5 probability theory At this point we should recall once more the parallel tracks of mathematical reasoning. Taken at a track one level, there is a superficial plausibility to Dembski’s argument. You were probably nodding along with the foregoing examples, thinking, “Why, yes, come to think of it, improbable things that fit a pattern really do suggest intelligent design.” However, if we are to accept this as a serious argument, then we also need the second track, where we pass from general intuitions to rigorous mathematics. Dembski claims his methods allow him to prove mathematically that evolution has been refuted, but there are several issues to resolve before we can analyze that claim. It is one thing to apply Dembski’s method to simplistic exam- ples involving coins and playing cards, where it is easy to carry out probability calculations, but everything becomes murky when we consider nontrivial examples, especially those drawn from biology. If the causal history of an event is entirely unknown to us, then how can we carry out a meaningful probability calculation for its occurrence? As we have seen, any such calculation requires that we define a probability space. Doing so entails knowing the full set of alternatives in which the event is embedded and the probability distribution appropriate to those alternatives. However, possessing that knowledge would seem to require some familiarity with the causal history of the event. “Specification” is likewise problematic, since there is a danger of doing the equivalent of looking at a fluffy cumulus cloud and saying, “Gosh, that looks sort of like a dragon.” We have enough expe- rience with coins and playing cards to distinguish design-suggesting patterns from more mundane arrangements. We know what moun- tains look like when we do not carve faces into them, and that makes it easy to recognize Mt. Rushmore as something designed. This background knowledge is precisely what we lack in the biological context. No one has intuitions about what will arise after billions of years of evolution starting from a relatively simple sort of life. We have no base of experience allowing us to say, “This structure is not 5.7 is the flagellum complex and specified? 137 the sort of thing evolution can produce, so we will have to explain it by recourse to intelligent design.” Thus, we need to be rigorous about how we distinguish the design-suggesting patterns from the ones we impose on nature through excessive imagination, and it is very unclear that this can be done in general. There are also difficulties with how complexity and specifica- tion relate. Dembski’s method tells us first to carry out a calculation to establish complexity and only then to consider the question of specification. It is not clear that this is workable, since the manner in which we specify the event will influence the probability space we use to carry out the calculation. In light of these and many other issues, we have reason to be skeptical of Dembski’s framework as a general method for detecting design. At a minimum, when Dembski attempts to apply his method to biology, we will need to pay close attention to how he tries to circumvent the difficult questions we have raised. We shall see that when the object in question is a biological adaptation, Dembski has no sound way of establishing either com- plexity or specification in the precise technical sense that his theory requires. 5.7 is the flagellum complex and specified? In all of Dembski’s voluminous writings defending both the theoreti- cal rigor and practical utility of his ideas, he has only once tried to apply his method to an actual biological system. This occurred in his 2002 book No Free Lunch: Why Specified Complexity Cannot Be Download 0.99 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling