Tuesday, June 13, 2006

Are Adaptations "Real?"

AUTHOR: Allen MacNeill

SOURCE: Original essay

COMMENTARY: That's up to you...

In an ongoing thread at Design Paradigm, Salvador Cordova wrote:

“There are many designed features in biology that make no sense in terms of natural selection but make complete sense in terms of design.”

This statement demonstrates a profound misunderstanding of both the concept of “design” and of “natural selection,” a misunderstanding which lies at the heart of the evolution/design debate. What is “design” anyway? Note that I’m not asking the question that Dr. Dembski thought he was answering, i.e. how can we tell if something has been designed. Before one can even ask that question (much less attempt to answer it), one must first agree on what “design” is.

This is not a trivial problem. Michael Ruse, in Darwin and Design: Does Evolution Have a Purpose?, asserts that one of the most important contributions of Darwin’s theory was that it put “design” back into nature (from which it had been removed by the “Newtonians”). To Ruse, “design” is essentially equivalent to “adaptation,” in that adaptations “solve” problems of biological function.

But the problem here is one that Lewontin and Gould addressed almost 30 years ago in their landmark paper “The Spandrels of San Marco...”. Lewontin and Gould pointed out two things: (1) not all of the characteristics of living organisms are adaptations (i.e. some of them are the result of pure “chance,” not necessity), and (2) even the characteristics that are clearly adaptive don’t have to have arisen because they are adaptive, nor will they continue to exist for the same reason. They coined the term “exaptation” to refer to characteristics of organisms that are not necessarily adaptive, but which nonetheless are biologically significant.

I would go much further than Lewontin and Gould: just as Darwin suggested (but did not come right out and say) that there are no such things as “species” (see "Origin of the Specious" in this blog), I believe that in nature there are no such things as “adaptations,” at least not insofar as such "adaptations" are "solutions" to biological "problems." That is, although there are characteristics of organisms that are correlated with relatively high reproductive success (and would therefore be considered by most evolutionary biologists to qualify as “adaptations”), it becomes problematic to decide exactly which of those characteristics are the “real” adaptations and which are merely “accidental.” Indeed, if one is serious about the variation/inheritance/fecundity/differential reproductive success model of evolution (i.e. the genuine article, not the RM+NS straw-man attacked by most IDers), then all of the characteristics of living organisms are “accidental” insofar as their origin cannot be shown to have been “intended” or “pre-destined” ahead of time.

Here is the real crux of the disagreement, as PvM has pointed out: what qualifies as an “adaptation” in biology can only be determined retrospectively, insofar as it has the practical result of causing increased relative survival and reproduction. No characteristics of living organisms can be shown to have come into being because they would eventually have that result; indeed, I would assert that to even make this claim is non-sensical in the extreme. What characteristics of living organisms currently alive will eventually result in their assendance or demise? We have absolutely no way of knowing, nor even of imagining a way of knowing. At some point in the future, we can look back and say “son-of-a-gun, those funny looking scales are correlated with increased survival and reproduction because they allow the animals that have them to fly, and therefore escape predators and capture prey more effectively,” but until this actually happens (and absolutely nothing in nature guarantees that it will), we can’t make any statements about the “value” of any of the characteristics of organisms now living.

This, rather than the rather vapid speculations Salvador cited for the future of genetic engineering, is the real value of genetic engineering to evolutionary biology (and vise versa). We now have the ability to selectively delete individual characteristics from many different organisms. This makes possible something that natural selection does not: the precise determination of the selective “value” of particular characteristics. This has already been done, and the surprising outcome has been that even some gene sequences that were thought to have been very important in selection (due to having been “conserved” over deep evolutionary time) are apparently insiginificant or even useless. We know this because knocking them out of the genome has no discernible effect on the survival or reproduction of the “knock-out” progeny. If one is the kind of “pan-adaptationist” that Lewontin and Gould criticized, this outcome should come as a severe shock, as it should to every IDer. But, if one is a true “Darwinian” (i.e. a devotee to that tradition which questions absolutely all assumptions, including the very existence of “adaptations” and “species”), it should come as no surprise at all.



Gould, S. J. and Lewontin, R. C. (1974) "The Spandrels of San Marco and the Panglossian Paradigm: A Critique Of The Adaptationist Programme" Proceedings Of The Royal Society of London, Series B, Vol. 205, No. 1161, pp. 581-598.

Labels: , , , , ,


At 6/14/2006 03:02:00 PM, Anonymous Anonymous said...

Excellent points Allen. ID proponents seem to be quick to claim that science is using ID's approach to detect design but on closer scrutiny these claims fall apart quickly.

ID is inherently a claim based on ignorance (elimination) and while it uses some 'fancy sounding' terms like complex specified information, the terms are used in a manner which conflates ID's terminology with how science uses such terminology.

For instance, it should be clear by now that ID does not present any "non begging the question" explanations, pathways, methods etc for items they claim have been designed.

ID starts with an unfounded assertion that design is that which remains once natural processes of regularity and chance have been eliminated. Why should we accept this when available empirical evidence and logic suggests that there is nothing necessarily supernatural about intelligence. In fact, intelligent behavior is quite well reducible to regularities and chance as polling, profiling, advertising and many other arenas show. Intelligence is in other words predictable and since intelligence has the ability to make choices given multiple options, there will be a certain level of variation or uncertainty present.
Amazon for instance uses this to propose to its users, items of interest based on their own past interests as well as based on the interests of those who have bought similar items.
But let's for the moment accept ID's Explanatory Filter approach. What is it all about in biology? Well, specification and complex information. Specification is trivial in biology as it refers to function and information refers to the negative log(2) of the probability. Now we get into some interesting territory. Dembski argues that if something can be explain as a regularity, it's probability becomes close to 1 and the information goes to 0. But the same applies then to intelligent design. If something can be explained as intelligently designed, the amount of information is zero.
So that does not really work well. So perhaps we can define the amount of information as the probability that the item arose under uniform probability? Under that scenario, something is 'designed' if it has a function and it's pure chance probability is too low. But then we still do not know if designed means 'designed by regularity/chance' or 'designed by an intelligence' (remember I am for the moment accepting the distinction between the two and I am showing how even accepting the distinction the filter suggests that the two explanations are nothing different). So how does the filter work? Well, it argues that if chance alone does not explain it and if regularities cannot explain it (yet) then we have to accept 'design' as the default explanation. So 'design' includes anything from 'intelligent designer' to 'an unknown regular process'. Once again ID fails to explain how to distinguish between actual and apparant design.
And now the best one. Even if we accept 'design', Dembski has shown that this does not necessarily need to involve an intelligent designer. Confused?... I bet... Ryan Nichols points out that:

“Before I proceed, however, I note that Dembski makes an important concession to his critics. He refuses to make the second assumption noted above. When the EF implies that certain systems are intelligently designed, Dembski does not think it follows that there is some intelligent designer or other. He says that, “even though in practice inferring design is the first step in identifying an intelligent agent, taken by itself _design does not require that such an agent be posited. The notion of design that emerges from the design inference must not be confused with intelligent agency_” (TDI, 227, my emphasis).

Source: Ryan Nichols, Scientific content, testability, and the vacuity of Intelligent Design theory, The American Catholic philosophical quarterly, 2003 ,vol. 77 ,no 4 ,pp. 591 - 611

As Elsberry has shown, given Dembski's logic, natural selection matches his definition of an intelligent designer. Once again we notice how ID fails to distinguish between apparant and actual design.

And since ID refuses to propose positive hypotheses, it is thus doomed to be unable to deal with the issue of apparant versus actual design in any scientifically relevant manner.

And that is why Intelligent Design is scientifically vacuous.

At 6/14/2006 05:57:00 PM, Anonymous Anonymous said...

What characteristics of living organisms currently alive will eventually result in their assendance or demise? We have absolutely no way of knowing, nor even of imagining a way of knowing.

Do you mean that we can't know in practice, or even in principle?

I can imagine looking at a population of animals, noting some of the variation that exists, and predicting that the faster runners, or the ones with the most resilient immune systems, or the ones with the sharpest teeth will have an advantage over the rest of the population.

Obviously, these predictions come with no guarantee; and evolution is far cleverer than I am -- I certainly would never have expected funny-looking scales to give an advantage -- but it seems to me that one can make some short-term predictions.

Or did you mean that genetic engineering gives us the ability to measure, rather than estimate, the advantage given by a particular gene?

At 6/14/2006 11:36:00 PM, Blogger Allen MacNeill said...

Both: we can't know in practice, or even in principle.
The reason we can't is that we can't know what kinds of environmental changes will occur in the future, nor what kinds of effects such changes will have on the organisms living at that time.

Two examples: it would have been literally impossible to predict the assemblage of vertebrates following (1) the K-T bolide impact and (2) the end-Pleistocene human migration into North America. Retrospectively, of course, we can see tremendous changes and attribute all of them to evolutionary processes, but prospectively we literally cannot predict what kinds of major (and in many cases minor) events may occur that will have significant impacts on the further evolution of life on Earth.

Another example: it would have been literally impossible to predict the evolutionary impact of the importation of African frogs for research (combined with global warming) and the subsequent spread of the chytrid fungus that is driving New World frog species to extinction at an accelerating rate.

In other words, our understanding of how evolution works and its actual effects are, like evolution itself, relentlessly retrospective. Natural selection cannot plan for the future; it is driven entirely by events happening now and in the past. In the same way, we cannot predict where evolution will go, we can only infer where it has been.

As to the genetic engineering example, since we now have the ability to "knock out" particular genes (or gene complexes), it may now be possible to determine what genes are necessary for survival and reproduction aren't. The former would qualify as adaptations, whereas the latter would not.

At 6/15/2006 02:01:00 AM, Anonymous Anonymous said...

Allen, have you noticed any improvement in the ability of your conservative Christian students to admit error?

From reading that "Design Paradigm" blog you refer to, it seems as if these folks ("Wiglaf," "Freawaru") remain entrenched in their willful ignorance but eager to learn more about the "tools" of effective argumentation.

It goes without saying, of course, that Sal Cordova will never change.

But I'm interested in your perspective: are your ID promoting students learning to admit error and admit that nearly every one of the professional "ID researchers" is an easily demonstrated charlatan and/or professional obfuscator?

I look forward to hearing your perspective on this issue, if not now then later.

At 6/15/2006 12:26:00 PM, Blogger Allen MacNeill said...

My own personal experience is that no amount of argument or evidence can ever sway the mind of a committed Christian: their committment is not based on argument nor evidence, after all.

But I keep engaging with them because it helps me clarify my own positions in my own mind, and perhaps sways non-committed readers, whose minds are not necessarily completely closed.

At 6/19/2006 10:21:00 AM, Anonymous Anonymous said...

Allen: "As to the genetic engineering example, since we now have the ability to "knock out" particular genes (or gene complexes), it may now be possible to determine what genes are necessary for survival and reproduction aren't. The former would qualify as adaptations, whereas the latter would not."

"Adaptive" traits are *always* context dependent and "context" is a word that covers a lot of changeable things.

Prediction is possible, if limited. See work by Dan Hartl, Daniel Dykhuizen and Antony Dean. Lots of interesting work has been done with chemostats and other forms of continuous culture (Lenski, too).


Post a Comment

<< Home