Friday, October 28, 2011

Triumph of the Darwinian method

In our recent spoof posts about Australopithecus erotimanis, we tried to make the point of how challenging it is to infer why traits found today evolved or, even more problematic, to infer this from fossils.  This raises several fundamental issues, and they are ones that Charles Darwin addressed even before Origin of Species:  explaining the nature of function from observed traits, explaining the fitness effects (for natural selection) of traits past and present, and explaining relationships among species (taxonomy).

Darwin didn't have all the answers (and neither do we), but he did introduce a method into biology that he applied to all of these questions (and so do we).  A very fine book, now about 50 years old, by Michael Ghiselin, called The Triumph of the Darwinian Method, goes in great detail using Darwin's own work, to show how transformative and powerful this was, and how Darwin used it in all the many contexts of his writing and thinking. Of course, legions of books and articles have adopted and reflected this approach to the present day.

Darwin didn't proclaim a method and then use it, but adopted and adapted the 'hypothetico-deductive' scientific method that had been developed over the previous century or so, for sciences based on concepts of 'laws of nature', as a way of identifying those laws and testing data in connection with them.  It was a framework for evaluating the empirical world.

To do this, to a considerable extent Darwin tried to show that life followed 'laws'.  So he didn't invent the method, but he did show more definitively than his predecessors, how it could be applied to life as well as planets, ideal gases, and so on.  The key was the assumption that life was a law-like--he compared natural selection to gravity, for example--descent from a common origin.

Given those assumptions, he interpreted both species formation and variation, adaptive functions of traits in fossils and in comparative morphology, taxonomy, biogeography, ecology, and so on.  From barnacles to earthworms, orchids and facial expression, he applied the selective, divergence framework.

There were mistakes, and one could go on about them, because some were serious.  To us, the most serious is that he took too strongly the assumed 'laws', so that his explanations are sometimes too rigid and in some instances too teleological or even near-Lamarckian--directional evolution, for  example, by which a fossil or barnacle was explained as being on the way to some state.

When we (that is, anthropologists who know something) try to explain fossils, like the Australopithecines we have just had some fun being satirical with, we have to try to understand all the central issues--taxonomy, function, adaptation.  Given the many problems involved in reconstructions, the conplexity of genetic control, the elusiveness and non law-like aspects of natural selection (versus genetic drift, for example), the incompleteness of data, and much more, it is no surprise that conclusions have to be tentative and assertions circumspect.

One of the main axes that we grind on MT is that we feel that there are many pressures of diverse kinds that encourage us not to be sufficiently circumspect, and to claim far too much.  This gets attention, and grants and promotion, but it can mislead science, when people take distorted claims as true and try to build on them.

Still, while it's fun to have fun with fossil claims, and fun to lampoon the hypersphere of our media-driven science circus, the problems are real and respectably challenging.  Perhaps much more than the public--or even many  professionals--actually realize.

Darwin, by the way, was often far more circumspect than is the usual case today.  He wanted to be right, of course, and didn't always exercise or express enough caution.  But he did frequently acknowledged where he was speculating, and so on.

8 comments:

Holly Dunsworth said...

I just gave an informal talk yesterday about my obstetrical dilemma project. And these discussions on the MT have been timely. With this project I'm asking people to join me in questioning something many of us (pros and pops) have taken for granted -- that the pelvis has uniquely influenced human gestation length and neonatal development. And many things are percolating out of this experience.

One thing ... I'm understanding better than ever, from an insider's perspective, how frustrating human exceptionalism is. It's difficult to test! Yet so often human uniqueness becomes the accepted wisdom when,instead, we should default to a null hypothesis, that humans aren't any different, until strong evidence shows it to be otherwise. Relative to most of the discipline's history, this is a totally flipped, upside-down way of doing biological anthropology.

And for OE and Ken and Anne, especially...
During my presentation I offered a hypothesis of "just how it is" as an alternative to "too much of a coincidence to ignore" for the tight fit at childbirth. I worried that this would stir some negative reactions, so I asked, "Is it a coincidence that my finger fits nicely in my nostril?"

Ken Weiss said...

All very perceptive. Even here at Penn State,the birth-canal explanation for the pelvic bone shape was presented in a journal club on erotimanis....I mean sabedi, which may or may not be something the Science papers mentioned (I didn't read them carefully enough in that context).

But nobody mentioned your talk about that unquestioned assumption, even though I imagine that at least some at the journal club were at your talk.

If they were, then it shows how easily such things can be forgotten, even when clearly stated, in the face of the momentum of the accepted-wisdom.

So, one need not apologize for being snotty (so to speak) about the way business-as-usual is usually conducted, and trying to push people towards less pre-constrained thinking. At least, right or wrong, that's a major thing we at least try to do on MT

Holly Dunsworth said...

I'm not sure anyone at PSU was at my AAPA talk about this and it predated the A. sediba pelvis paper anyway. This talk yesterday became a discussion at times, which was fantastic!

Some clever folks here (Harvard) who know more about the pelvis than I do, agree with me that the sediba paper only highlights how so much hominin fossil (pelvic) anatomy is assumed to be adaptive for this (bipedalism) or that (childbirth) without much evidence (if any at all) to back it up. When finds like sediba come out, that expose a lot of those misleading assumptions then it's a bit embarrassing.

I think my obstetrical dilemma project embarrasses people and I hate that because I'm not out to make this personal or to play gotcha. Besides, it could be that in humans the pelvis does actually influence our life history uniquely... my point is only that there's no support for that idea with current evidence.

Ken Weiss said...

You are a respectful and conscientious person. Nobody should be personally offended by things you say, but you should not feel sheepish about asserting a point of view.

It is not just anthropology, but all of science, indeed all aspects of human culture perhaps, that are vulnerable to accepting accepted wisdom. Unlike religion and other ideologies about society, science is supposed to be, specifically, the field about understanding the world as it is,rather than how we want it to be.

Today in the news, even the limiting nature of the speed of light is being questioned. Dogma should always be questioned.

And this is nowhere more true than when the public is paying for the work, based on various kinds of expectations of results, and trust in the integrity of the scientists--and of the science.

Many have argued that dogmas can be defended as the best we know so far, and not dropped lightly, since they have usually developed out of knowledge at some previous stage. But they must be challenged!

occamseraser said...

Revelations from the hominin fossil record continue to erode the dogma of human exceptionalism, but there will be a time lag before this is fully appreciated (per Holly's comment). And deniers will persist because careers have been built on the notion of human evolution as one big singularity.


I think that many human paleontologists interested in function and biomechanics realize that when they speak of "adaptation", they know full well that they have not and probably cannot ever provide real evidence for differential reproductive success. More often, we're attempting to evaluate biological form relative to some model of optimization, and the A-word tends to be shorthand for goodness-of-fit to model expectations. We should probably use a different word lest we be spandralized!

Ken Weiss said...

Good points, but ideas constrain thinking the way cages constrain motion, so in science we should avoid misleading terms. That means in classes, in our papers, and in our textbooks. But ruts are hard to get out of: the road ahead may seem too threatening! People want comfy security.

Torbjörn Larsson said...

I'm late (by way of the Carnival of Evolution), but this jumped out to me:

Analogously to how we have problems to resolve biological history, we have problems of resolving cultural history. Darwin may have been too embracing of lawfulness and teleology, but he was breaking with old dogma and may have been forced to be content to lay the foundations.

The hypothesis is based in what I think may have happened to Newton. Already in classical mechanics there is nothing useful by hypothesizing an 'absolute' room, but Lagrange et al retracted from Newton's claim to a relative room way before relativity. (Which 'merely' brings in time on the same level as space.)

My early understanding is that Newton made a bad "common sense" decision, but that turned out wrong. According to accounts he had to get away from the earlier cultural precept as room consisting of the objects inhabiting it. (Here a space is between a chair and a table, for example, and is different in the kitchen and the living room.) Hence he introduced a separate, "absolute" room concept for that reason alone. He was probably very well aware of the overreach he made by that, but he fought millenniums of cultural inertia while trying to be accepted.

Since we have access to much more letters from Darwin than from Newton (I think), it would be interesting to try to resolve such points. It is very probably wrong, but it could have happened.

As for using "adaptation" straight up, I (a layman) cannot but see it as a productive research strategy. It is, I believe, what most likely makes new traits, and you would start there while waiting for testing either way. Isn't it as simple as that?

Ken Weiss said...

There is an interesting article on cultural evolution in the current edition of 'Evolutionary Biology', and I have a commentary on that that will be on their website soon. I think you are totally right and that cultural evolution should not be built on Darwin's ideas about biology, but on its own terms.

I don't know anything about that aspect of Newton's thinking, or about Lagrange in this regard. But even Einstein was essentially searching for universal 'laws' (e.g.,'law of special relativity'). I would like to know more so I could understand what you mean.

As to natural selection and adaptation, anything functioning is 'adaptive' (or was until the previous generation, at least), but chance and various other phenomena can generate orderly acceptable traits. Selection certainly is part of that, and very important in some instances. But even chance can be important. The problem besides just assuming adaptation is that then you feel licensed to decide what a trait was adapted 'for', meaning reproductive success of something genetic, and that's where the story-telling begins, because what we know of traits, their development, and their genetic basis does not suggest that such simple stories are usually very accurate.