P&LS Darksiders Lab Project (Part 2 of ?)

Pardon the loose way data are presented. Typepad is terrible for that sort of thing. Or at least, that's what I've found.

As I mentioned in a previous post, my lab ran down the 285 articles citing CBT 1997 and WG 1998 through 2009. Those articles were scored as follows: 0 - cites to CBT or WG as mere general historical context (or did not make a connection at all); 1 - the area of contention in the article is the exchange between CBT/WG; 2 - the article uses the controversy to contextualize the paper's problem and/or uses Fisher's or Wright's work directly to support the article's central claims. We used only articles scored as a 1 or a 2.

223 articles were scored either as a 1 or 2. The group then studied the papers to determine whether they engaged one of the following nine issues:

1. Fisher's Fundamental Theory of Natural Selection (FTNS)

2. Wright's adaptive landscape metaphor

3. Wright's shifting balance process/theory

4. Importance of epistasis in evolution

5. Population structure

6. Speciation (on the theories)

7. Evolution of dominance

8. evolution in Panaxia dominula

9. evolution in Cepaea nemoralis

These nine issues are either the focus of the original exchanges between Wright and Fisher, or Wade and Coyne, or both. The table shows which topics were discussed most often.

168 of 223 articles focused on one or more of the adaptive landscape, the shifting balance process (or its phases), the role of epistasis in evolution, and the nature/extent of population structure. And here, really, is where we thought things got interesting. Let's leave out the adaptive landscape for now and just consider the shifting balance process, epistasis, and population structure.

The pie chart below shows that of the 85 articles scored on the shifting balance process, most authors thought that the process at least sometimes occurs in nature. (It's hard to say what "sometimes" means, but that's not something the lab can say anything about.) Obviously, this contradicts CBT's claim that the SBT is evolutionarily unimportant, i.e., that, as they say, no case is explained better by the shifting balance process than by the mass selection. When I wrote about CBT/WG back in 2002, my view was that WG were right: there's plenty of room in the evolutionary domain for Wright, Fisher, and no doubt other evolutionary theories. It looks to us like WG are vindicated on this score. In fact, only a few authors thought the shifting balance process never occurred.

And, actually, I think things get more interesting turning to epistasis and population structure. Of the 71 articles taking up the issue of the role of epistasis in evolution, about half believe that epistasis is important in evolution, followed by close to half who believe that it's sometimes important --see the pie below. Now, CBT worked hard to claim that they understand the importance of epistasis in evolution and, in fact, that Fisher himself included it in his natural selection theory. I think the claim about Fisher is problematic: epistasis is relegated to the environmental component of fitness and is virtually ignored. Certainly, Fisher was not thinking about epistasis along the lines of the way that 51% of the biologists engaged in the CBT-WG disagreement were.

Lastly, consider population structure.

Of the 64 articles taking up this issue, most think population structure is more Wrightian than Fisherian. That is, populations are more likely to be smallish and semi-isolated within the global population than with sufficient gene flow to treat them as large and basically panmictic. It struck me as interesting that, in fact, only a few biologists held to the claim that population structure is more Fisherian.

With all that said, take the results together: the shifting balance process sometimes accounts for evolutionary phenomena, epistasis is mostly evolutionarily important, and populations are structured.

A small point can be drawn from this: Coyne and crew are just wrong. In fact,  Wright seems to have had much more influence on evolutionary thought than Fisher, if we think of influence as teaching us the right kinds of lessons. After all, it was Wright who emphasized things like epistasis and population structure. (There are other small points, but let's leave them aside for the moment.)

A larger conclusion: It appears that epistasis and/or structured populations are more evolutionarily significant than the shifting balance process. This, I think, is unsurprising since the shifting balance process requires fairly restrictive parameters to work. But if that's true, then what can be said more generally about how evolution works? Is a strong adaptationist view compatible with epistasis singly or in combination? Or do these properties of biological systems require another somewhat general account of evolution, taking into account a shifting balance of evolutionary causes in a way different from Wright (and obviously Fisher)? Or, more difficult, are biologists in a "hunt and peck" situation, looking for whatever combinations of evolutionary processes apply in concert to any given population? That is, sometimes mass selection, sometimes selection and drift, sometimes selection and migration, and so on?

In the lab's view, we might want to go out on a limb and say that it's time to drop Fisher and Wright in our studies of evolution. That is, it's time to drop big frameworks for thinking about how evolution works and trying to apply such frameworks --theories-- to the evolution of populations. Perhaps a bottom-up approach in terms of discovering evolutionary causes of evolution in populations is a better strategy in evolutionary work. Wright said, in 1988, that himself, Fisher, Haldane, and Kimura each proposed evolutionary theories and that all are correct. I would say, now, that such a view is not strong enough --abandon "theories" of evolution altogether because there just aren't any to be had. Instead, search for the combination of evolutionary causes that underwrites particular evolutionary explanations of particular cases of evolution.

Let me try to be a little bit clearer. Some might say that evolutionary biologists are already doing what my lab is suggesting. But I'm not convinced. I think it's plausible to think that some biologists are pursuing a kind of Wrightian strategy: there are plurality of evolutionary theories that explain the evolutionary domain. (Wade and Goodnight said this in 1998. I said it in 2002. No doubt others have said it as well. But not Coyne, Barton, and Turelli. And not others.) We are not suggesting a plurality of theories. We're suggesting that the evolutionary domain requires a (massive) plurality of particular explanations comprised of combinations of evolutionary causal processes to account for the evolution of life. There is no viable general framework like adaptationism or shifting balance or neutralism, or anything else. In this way, evolutionary biology is like, say, molecular biology. There are only causes/mechanisms, not theories. I just don't think that's what evolutionary biologists are doing. Of course, that's an empirical question.

References:

Coyne, J., Barton, N., and Turelli, M. (1997), "Perspective: A Critique of Sewall Wright's Shifting Balance Theory of Evolution", Evolution 51: 643-671.

Skipper, R. (2002), "The Persistence of the R. A. Fisher-Sewall Wright Controversy", Biology and Philosophy 17: 341-367.

Wade, M. and Goodnight, C. (1998), "Perspective: The Theories of Fisher and Wright in the Context of Metapopulations: When Nature Does Many Small Experiments", Evolution 52: 1537-1548.

Wright, S. (1988), "Surfaces of Selective Value Revisited", American Naturalist 31: 115-123.

Comments

[Roberta Millstein]

My view is that biologists are already doing this, but sometimes with biases towards the likelihood of particular mechanisms. Remember that "adaptationism" wasn't a "research programme" until Gould and Lewontin so named it. It isn't really a theory in any strong sense. And indeed, it seems to me that the argument to consider a plurality of mechanisms (not theories) is one of the Spandrels article's main messages.

On another note, in working on my "concept of population" stuff I read a lot of work on models for different types of population structure. Ecologically-oriented folks in particular have become very interested in metapopulations because there are so many fragmented habitats out there. I agree that the general consensus is that Wright was right on that point, even when Wright isn't cited (though now that I think about it, he sometimes is seen as a precursor to the metapopulation idea, even the it was Levins who coined the term).

[Frédéric Bouchard]

Hey Rob,
cool neat stuff! I hate statements like "Biologists think...", but this kind of work lets us say with more confidence "Biologists write...". Good job. I get the same feeling as Roberta concerning the metapopulation bit (as a way of bridging previously-thought-as-incommensurable models). While Adaptationism is not an explicit research programme, it's definitely a heuristic approach, and Rob's conclusion,as Roberta points out as well, fits this I think.

[Michael Dietrich]

Rob, of course there are theories! Claiming a plurality of causes is a claim for a theory. The question is: how useful are these theories, especially when they are posed as polarizing and mutually exclusive alternatives? I'm sympathetic to letting "a thousand flowers bloom" in terms of particular causal explanation, but do not think that eliminating theory is necessary. I think what might be more forceful to say is that the kind of advocacy of a big theory by some biologists simply does not connect with scientific practice and is not a persuasive way for them to build support for their view.

[Robert Skipper]

Mike: What's a theory on your view? I don't really know what a theory is and would be happier without them, whatever they are. Still, I'm confused about your claim that "claiming a plurality of causes" is a claim for a theory.

[Michael Dietrich]

I like the semantic approach to theories, which represents a theory in terms of a set of models. I seem to remember some paper you wrote with some people that tend to post comments on this blog that includes the following:

"According to the semantic approach, a model is a non-linguistic entity that can be represented in terms of variables and rules or laws describing the functions operating on those variables. The state of the model refers to any particular set of values for the variables in that model. The collection of all of the states of the model composes the model’s state space. The rules of a theory describe the possible relationships of coexistence, interaction, and temporal succession of the variables in the model. In doing so they describe trajectories through the state space for that model (Lloyd 1988, 19). One virtue claimed for this approach to understanding models and theories is that it coheres strongly with the ways in which biologists understand models within biology (Lloyd 1988, Lewontin 1963)."

"(Mis)interpreting Mathematical Models: Drift as a Physical Process"

[NiciOf]

Such a smashing text just about this post. It is valuable to notice the writing service that will make the essay writing or custom essays. In fact, you like to buy essays for sale.