Thursday, May 31, 2007

Darwin and Design by Michael Ruse

 
In Darwin and Design Michael Ruse tackles a tough problem; namely "Does evolution have a purpose?" Unfortunately the correct answer is "no" but Ruse muddles, misdirects, and misunderstands so thoroughly that by the time you reach the end of the book you just want to throw it against the wall.

The main theme of the book is teleological thinking or the idea that things happen in order to achieve a goal. We are familiar with this way of thinking in religion. Ruse spends some time describing the history, culminating in the natural theology of William Paley.

Paley and others argued that the presence of design in nature demanded a God who was the designer. The teleological part of this argument is the recognition that designed species, especially humans, represent a clear goal that needs an explanation. Life has meaning and purpose, according to believers, and it is God who gave it to us.

A teleological argument, or argument from design, is an argument for the existence of God or a creator based on perceived evidence of order, purpose, design and/or direction in nature. The word "teleological" is derived from the Greek word telos, meaning end or purpose. Teleology is the supposition that there is purpose or directive principle in the works and processes of nature.
"Teleological Argument" Wikipedia
Charles Darwin explained how life could appear to be designed by invoking natural selection, thus removing God from the equation. Nevertheless, teleology remains an important part of science, according to Ruse, because nature is designed by natural selection. It is quite appropriate, he says, to argue from design (the eye for example) to cause (adaptation).
This then is the paradox to which Darwin and Design is directed. Darwin seems to have expelled design from biology, and yet we still go on using and seemingly needing this way of thinking. We still talk in terms appropriate to conscious intention, whether or not we believe in God. In biology we still use forward-looking language of a kind that would not be deemed appropriate in physics or chemistry. Why is this?
Ruse seems to be at his best when describing the history of philosophy—as long as that history pre-dates Charles Darwin. His book is worth reading if you want a good summary of the design argument up to 1859. From that point on things begin to fall apart because Ruse does not understand modern evolution and he does not understand the controversies over evolutionary theory that persist to this day. Consequently, all of his history from Darwin on is biased and wrong.

The essence of Ruse's argument is as follows. Life evolves by natural selection. This leads to species and characteristics that are well-adapted. These characteristics have the appearance of design because they are, in fact, designed by natural selection. Because we know that everything is an adaptation it's perfectly legitimate to look at a species or an organ and assume that it as been designed by natural selection. While this adaptationist program may seem teleological because it assumes a purpose, it is, in fact a very legitimate way to do biology because design is a fundamental part of biology.

There are times when one thinks that Michael Ruse must have slept through the last half of the twentieth century. Has he never heard of Gould & Lewontin and The spandrels of San Marco? Is he unaware of the controversy over the validity of the adaptationist program?

Yes and no. He's heard of the controversy but he just wasn't listening. Everyone else who has addressed this question recognizes that the Gould & Lewontin challenge is not going to go away. They attempt to deal with it—usually not very successfully.

To his credit, Ruse seems to have picked up on the rumors that something important was going on so he does mention the spandrels paper and the attack on the adaptationist program. It's right there on pages 234-239. Five pages on structural constraints as introduced by Gould & Lewontin in their famous 1979 paper. Structural constraints? Surely there's more to the argument than that? Yes, there is but Ruse can easily dismiss it,
The point is whether they [Gould & Lewontin] introduce a whole new dimension into the discussion, by showing that much in the organic world is fundamentally nonadaptive. Darwinians have failed to see this and still continue not to see it.
That's it. Ruse is blind to modern evolutionary theory and quite proud of it. According to Ruse everything is an adaptation and "Darwinism" and "evolution" are synonyms.

The rest of the five pages on Gould & Lewontin are no more enlightening. Lest you think I'm being too harsh on Ruse, I assure you I'm not. He really doesn't get it. There are two pages devoted to random genetic drift. Two pages! After acknowledging that drift can sometimes cause evolution he dismisses it out of hand with,
Over time, however, random drift would be expected to average out more smoothly than differences due to the ever-changing forces of selection. For this reason the hypothesis that most molecular difference is due to drift has not been well received. Time and time again, measurements have shown that molecular differences are not what we would expect were drift the sole or main cause of change. In fruit flies, we see how random drift was ruled out as a significant factor in changing levels of the Adh gene.(p. 201)
Having summarily dismissed all objections to the ubiquity of adaption, Ruse can defend the argument from design by invoking adaptation as the sole driving force of evolution. In a chapter on "Design as Metaphor" he outlines his version of the adaptationist program. It's not only appropriate to attribute design to living things but it's a very productive way of advancing scientific knowledge.
Organisms produced by natural selection, have adaptations, and these give the appearance of being designed. This is not a chance thing or a miracle. If organisms did not seem to be designed, they would not work and hence would not survive and reproduce. But organisms do work, they do seem to be designed, and hence the design metaphor, with all the values and forward-looking, causal perspective it entails, seems appropriate.(p. 276)
Critics of the adaptationist program—I am one—argue that it begs the question. When you see something in nature it is reasonable to assume that it arose by evolution. The question we want to answer is what kind of evolution gave rise to that particular characteristic?

Take the fact that some people can roll their tongue as a simple example. We know there is a genetic basis to tongue-rolling. Some people have the allele that allows it, and some don't. We want to know why tongue-rolling exists.

     Once you have the metaphor of design in play, then of course you can ask questions about borderline instances and extensions and so forth. The real question, though, is whether, in the first place, the metaphor itself is an appropriate one. The question is not whether metaphors should be used at all but whether the specific metaphor of design should be used to explain evolution.

     Darwinians argue strenuously that it must be used. Richard Dawkins speaks to precisely this issue, asking what job we expect an evolutionary theory to perform. ... Dawkins agrees with John Maynard Smith that the "main task of any theory of evolution is to explain adaptive complexity, i.e. to explain the same set of facts which Paley used as evidence of a Creator."

Michael Ruse p. 278
If you are a modern evolutionary biologist then you are aware of several possibilities. It could be just an accident that has no great significance at all. Maybe tongue-rollers and non-tongue-rollers have an equal chance of leaving offspring and the alleles will be fixed or eliminated by random genetic drift. Or maybe one of these groups has a selective advantage. Maybe tongue-rollers are more successful at having children and that's why the allele persists in the population. Eventually everyone will be a tongue-roller because natural selection is operating.

If you are a committed adaptationist then you begin by assuming that the ability to roll your tongue is designed. Your task is then to explain how this design arose and you have only one choice—evolution by natural selection. Thus, your choice of the design metaphor has blinded you to the possibility that tongue-rolling may not be an adaptation at all. This is a very restrictive research program because the question pre-supposes the answer. In other words, by imposing design and purpose on the natural world—albeit natural and not divine purpose—Ruse and his colleagues are avoiding the very question they should be asking; namely, "is this an adaptation?" This bias leads to fanciful just-so stories as the adaptationists struggle to come up with imaginary ways of explaining the design that they think they see in nature.

Does Ruse have an answer to this objection? Yes he does,
The critic might respond that one has here a circular situation: Darwinians make searching for adaptation central to their program, and then when they find the adaptations they so fervently seek, they use them as support for Darwinism. But a better term than "circularity" might be "self-reinforcement." Darwinism is a successful theory—our scientific examples show that—and at the moment (and for the foreseeable future, whatever the qualifications) it is the only game in town, on its own merits. Fruit flies, dunnocks, dinosaurs, fig wasps—this is a theory on a roll. It has earned the right to set the agenda. (p. 280)
As far as I'm concerned this is dead wrong. Darwinism is not the only game in town and we've known that for almost fifty years. At the very least you have to consider fixation of alleles by random genetic drift. If this is how a character actually evolved then there is no design. The metaphor is inappropriate. The program is useless. (There are other non-Darwinian processes.)

The entire thrust of Ruse's argument for design and purpose in evolution is absolutely dependent on one critical assumption: that natural selection is the only significant mechanism of evolution. If this isn't true then his whole argument falls apart. It isn't true.

I accept Ruse's challenge when he says,
Of course, Lewontin and his school do not care for many of the findings of the adaptationists. But to say that we should not play the game at all, or that we should count all as equal, requires some persuasive arguments. Better than arguments would be examples. Let those who worry about explanatory adaptationism show their dunnnocks and dinosaurs and fig wasps. When they demonstrate that they can do science which explains and predicts without invoking adaptation even implicitly, then we can start taking their position seriously. (p. 281)
There are literally dozens of examples of non-adaptive evolution that have been widely discussed in the scientific literature. It is more than "silly" of Ruse to issue a challenge like this. It's just plain ignorant.

Scientists who study junk DNA, for example, are doing very legitimate science when they predict that junk DNA sequences will not be conserved between species. Scientists who study blood type in humans are doing real science when they test the null hypothesis by asking whether the alleles conform to the Hardy-Weinberg distribution. (They do, suggesting strongly that they are not under selection.) Scientists who study speciation in birds ask whether the founder effect is real. (It is, and this shows that morphological changes during speciation are not due to adaptation.) The late Stephen Jay Gould and his colleagues have done good science by developing theories of punctuated equilibria and species sorting without assuming that natural selection and adaptation are essential. Ruse needs to take their position seriously. Meanwhile Ruse has demonstrated that we don't need to take him seriously.

The entire field of molecular evolution is based largely on explanations and predictions that rely on random genetic drift of neutral alleles. As far as I know, the people who work in that field are good evolutionary biologists even though they don't assume design when constructing their phylogenies.

And lets not forget about one of Lewontin's favorite examples. The African rhinoceros has two horns while the Indian rhinoceros has only one. Why? If you accept the modern theory of evolution then your choices of explanation can range from adaptive to accidental. If you restrict yourself to Darwinism then you must assume design and your explanation has to invoke natural selection. Somehow you have to come up with ways to explain why African rhinos were better off with two horns while Indian rhinos were better off with only one.

Using the metaphor of design and purpose forces you to assume the answer to the very question you are asking. It forces you to reject known evolutionary mechanisms such as random genetic drift. This may be good philosophy but it's not good science.

Getting back to the title of the book. Is nature designed? Partly, but there are lots of things that don't look designed and are not the end product of natural selection. Our genome is a good example. It's more like a Rube Goldberg apparatus than a well-tuned machine. It is not particularly helpful to say that living things are designed, or even that they have the appearance of design. If we stop saying that everything is designed then we will be better prepared to consider other possibilities, like evolution by accident.

10 comments :

  1. Thanks for putting this up. Sooner or later I'm going to run out of Gould, Diamond, and Sagan books to read, and I'll need something else. I'll spend my money on something other than this book.

    ReplyDelete
  2. I have always found eschatology ('study of the Last Things') such as Ruse's as applied to design more like scatology, trying to determine biology by studying its outward arrangements and even obsessing about it.

    A brief review of some causal descriptions in a biological and cosmological setting is Wilkins ( http://www.talkorigins.org/faqs/evolphil/teleology.html ).

    The at times philosophically inspired perspective may feel somewhat odd for physicists though. We don't especially care if boundary conditions are initial constraints or not except when the physics force us to. But it is nice to have a reference for some of the historical baggage of human views.

    ReplyDelete
  3. Just as many philosophers and non-biologists were swayed by the rhetorical style of Gould & Lewontin into thinking that studying adaptation was somehow broken and that they accurately represented the state of evolutionary biology then and still do, Ruse seems to have felt that Gould & Lewontin were all rhetoric and weren't actually making very good sense.

    Ruse needs to give Mike Lynch's new book a pretty close read.

    ReplyDelete
  4. "Organisms produced by natural selection, have adaptations, and these give the appearance of being designed. This is not a chance thing or a miracle. If organisms did not seem to be designed, they would not work and hence would not survive and reproduce. But organisms do work, they do seem to be designed, and hence the design metaphor, with all the values and forward-looking, causal perspective it entails, seems appropriate." (Ruse p. 276)

    Is it just me or is this bollocks?

    Surely, to introduce design, actual or metaphoric, one need to to assume that the product was a desired outcome. The process of design infers that there is a desired outcome.

    In my field area when I was doing fieldwork, there was a rock outcrop that worked brilliantly as a bench. It was the right hight so that I could sit on it with my feet on the ground, It was deep enough to accomodate me as I could leand back and rest against the rest of the outcrop that continued upwards. However, no-one would claim that that the natural forces that shaped the outcrop intended the outcrop to function as a bench, or that it was designed as a bench, even though it had many of the characteriestics we associate with benches.

    Ruse seems to be arguing that because something works, then we can infer some design process.

    I would argue that we can only infer design if the finished product works in a way that was envisaged before the product was produced.

    Evolution (whatever the process) did not have in mind an organ with the functions of the human liver when the first metazoan body plans incorporated specific cells to filter circulating fluids. The human liver works, and work in the way it does, not because it was intended to work that way, but because evolutionary processes shaped it that way, just as the natural weathering processes shaped my bench. The fact that the liver works in the way it does is testament to the power of evoutionary processes, not their forward planning.

    "We want to know why tongue-rolling exists."

    That's easy. Tongue-rolling exists to show the inate superiority of tongue-rollers over dirtystinkingrottennon-tongue-rollers.

    Chris (not designed, and proud tongue-roller)

    ReplyDelete
  5. "I would argue that we can only infer design if the finished product works in a way that was envisaged before the product was produced."

    The problem with that is that humans can (and do) use evolutionary methods to design things. If the typical overall design process has only a facile similarity to evolution, then genetic algorithms are used more or less routinely to develop for example some specific efficient and space-saving antenna designs in mobile communication.

    So I can accept Moran's description of adaptationism as design. The difference natural vs artificial design would then be that the constraints that controls what the process selects are contingent (i.e. 'goals') vs predetermined (i.e. goals).

    Torbjörn (not only a dirty-et-cetera-non-tongue-roller, but also a stinking-et-cetera-non-whistler; wonder if that is a trait too, or if I just need to practice more)

    ReplyDelete
  6. Good job, Larry

    I would say , though, as an excuse for Ruse, that there are plenty of biologists responsible of sending someone like Ruse astray in his teleological adaptationism. Nobel prize Jacques Monod sais teleology is intrinsic to life.
    Mayr was more scrupulous and replaced teleology for the similar "teleonomy": the existence of adaptive programs (physiological, developmental, etc) that are the result of natural selection and that seem "purpose-driven". The execution of these programs was to him a "proximal", non-evolutionary mechanism with proximal, non-evolutionary explanations (developmental, physiological), whereas the origin of that program, was the result of evolutionary mechanisms that provided an "ultimate" explanation.

    But, natural selection has been shown to be isufficient to explain the origin of complex adaptations. Exaptation (and spandrels, structural possibilities and contraints), are an essential part of understanding the evolution of ANY complex adàptation. Yet this is never acknowledged enough and the cultural bias persist to simply give all credit to natural selection, always. It's an easy one to pull off.

    ReplyDelete
  7. and then we have Maturana and Varela, who say that we should not concede toward any concept of finalism, since finalism has no mechanistic detail. It is indeed scientifically vacuous...
    They argue that teleonomy is a way of describing organisms that reveals nothing about their organization, only that the organism is consistent with the environment in which it is observed.

    ReplyDelete
  8. alipio says,

    Nobel prize Jacques Monod sais teleology is intrinsic to life.
    Mayr was more scrupulous and replaced teleology for the similar "teleonomy": the existence of adaptive programs (physiological, developmental, etc) that are the result of natural selection and that seem "purpose-driven".


    Actually Monod used the term "teleonomy." Here's a quotation from page 9 of Chance and Necessity.

    Every artifact is a product made by a living being which through it expresses, in a particularly conspicuous manner, one of the fundamental characteristics common to all living beings without exception: that of objects endowed with a purpose or project, which at the same time they exhibit in their structure and carry out through their performances ...

    Rather than reject this idea (as certain biologists have tried to do) it is indispensable to recognize that it is essential to to the very definition of living beings. We shall maintain that the latter are distinct from all other structures or systems present in the universe through this characteristic property, which we shall call teleonomy.


    Mayr claims he invented the term "teleonomy" back in 1961. Julian Huxley published a paper about teleonomy in 1960.

    ReplyDelete
  9. Teleonomy is clearly meant to be a surrogate for teleology (BTW, I'm not sure Monod's teleonomy is the same as Mayr's...or Huxley's, for that.)
    Mayr, in replacement of a final cause, introduces the notion of "ultimate causation" of adaptive programs, driven by natural selection. It is legitimate to say, for instance, that the lung is "for" breathing, if it evolved through natural selection for that function (the sad truth is , natural selection is usually just one factor. Quite plainly, evolution is not all about adapting to the environment. Period).

    D'Arcy thompson, a critic of darwinism, said that the mere sifting out of the good from the bad had created a teleology without a telos...

    ReplyDelete
  10. Torbjörn
    "So I can accept Moran's description of adaptationism as design. The difference natural vs artificial design would then be that the constraints that controls what the process selects are contingent (i.e. 'goals') vs predetermined (i.e. goals)."

    I think both you and Larry are wrong here. Look at what Ruse says:

    "Organisms produced by natural selection, have adaptations, and these give the appearance of being designed. This is not a chance thing or a miracle. If organisms did not seem to be designed, they would not work and hence would not survive and reproduce. But organisms do work, they do seem to be designed, and hence the design metaphor, with all the values and forward-looking, causal perspective it entails, seems appropriate."

    It is trivially true that organisms must 'work' to survive, so yes all life on Earth has worked. But it is also true that 99% of all life on Earth no longer works. How does natural selection as designer account for such extinction? An inbuilt redundancy factor?

    My argument is the term 'design' is not useful. It is a loaded term. Look up the work in any dictionary and you get things like, "to plan", "to make" "to intend for a definite purpose", "a scheme", "intention". To most people design means to intentionally plan and make something for a purpose. But there is no intent in evolution, there is no purpose. Evolution is the result of the interaction of natural processes.

    In fact I'd go further and say that design and evolution as completely separate processes. Design is, by and large, a constructive process. Evolution isn't, it's an incrementally constrictive process.

    One can envisage the evolution of major groups as operating as a random walk through morphospace (or, as a member of the U of E, I should use the term drunken walk). However, this drunken walk is not truly random because each step constricts the amount of morphospace available for the next step. Thus major groups today are the result of a long drunken walk, each step of which constricted the available morphological and hence the diversity they can show (no 7 legged mammals, no functioning gills on reptiles). This isn't design as we understand it, it's the interaction of natural processes constrained by morphospace and developmental constraints.

    Introducing a sub-division, as you have done, of artificial v. natural design simply goes to show how bad the term is that we have to introduce categories to enable the description to work. This is like Behe et al saying 'everyone knows what science means, but we're introducing a slightly different meaning to allow our pet religion to be classified as science'. This is 'we are calling it design but we mean it differently to the usual meaning, of a designer intentionally making something for an specified or intended purpose, and trust that you will be able to tell the difference.'

    What is 'natural design' anyway? No one would call my bench, mentioned previously, a natural design, because there was no purposeful activity, just natural processes (weathering and erosion) constrained by the differential susceptibility of the different rock layers. Similarly is a rust pattern on a fence 'natural design'? I would say there is no design, just the application or redox potentials on random atomic structure.

    We should fight against the use of the term 'design' to describe evolution. Not because the term has been corrupted by the Idiots, not even because it carries it's own readily identifiable meaning, but because it simply does not describe the processes of evolution.

    One last thing, Larry says

    The entire thrust of Ruse's argument for design and purpose in evolution is absolutely dependent on one critical assumption: that natural selection is the only significant mechanism of evolution. If this isn't true then his whole argument falls apart. It isn't true.

    I think there are two arguments Ruse is putting forward here (and has been for some time). One is that natural selection is the only significant mechanism, here I agree with Larry (having disagreed before it's only polite . . .). However, the second argument is that evolution is not random, hence . . .

    the design metaphor, with all the values and forward-looking, causal perspective it entails, seems appropriate."

    We don't need a design metaphor to understand that evolution is not a random process. There are random elements to it, mutation, fixation, climate catastrophes etc., cannot be predicted. But the drunken walk process constricts the available morphoshere, as does the organisms developmental process. These still allow a great deal of diversity, but rule out a great deal more.

    ReplyDelete