"Important" Evolutionary Causes and Theories

What does it mean to say that some or other evolutionary cause is "important" or "significant?" These attributes are continuously attached to specific causes of evolution,  like natural selection, and to general theories, like Sewall Wright's Shifting Balance Theory (SBT). Typically, significance is measured relatively (Beatty 1995). We see that selection is an important or significant evolutionary cause relative to genetic drift. And we see that Wright's SBT is an unimportant evolutionary theory relative to Fisherian mass selection. I can think of two main ways biologists tag causes and theories with some variant of "important" or "significant." The first, importance as scope, applies to theories. The second, importance as time-course, applies to causes.

1. Importance as scope.

When the "important" tag is assigned to a theory, it seems most often, at least, to have to do with the theory's scope of applicability. Jerry Coyne, Nicholas Barton, and Michael Turelli, hereafter CBT, (2000), ask explicitly whether Wright's SBT is "important" in evolution. Given that they conclude that the SBT fails to explain just about every adaptation it's tried to explain, the theory isn't important; it has almost no scope of applicability. They say, on the first page of their paper,

Our review of the data from nature shows that although there is some evidence for individual phases of the shifting balance process, there are few empirical observations explained better by Wright's three-phase mechanism than by simple mass selection. Similarly, artificial-selection experiments fail to show that selection in subdivided populations produces greater selection response than does mass selection in large populations. The complexity of the shifting balance process and the difficulty of establishing that adaptive valleys have been crossed by genetic drift make it impossible to test Wright's claim that adaptations commonly originate by this process. In view of these problems, it seems unreasonable to consider the shifting balance process as an important explanation for the evolution of adaptations.

Now, I think CBT are unfair to the SBT and, in particular, what biologists have learned about evolution from it. And I argue for this claim in my 2002 paper, "The Persistence of the R. A. Fisher-Sewall Wright Controversy," available here. But never mind who is right or wrong. The question is what's it mean for an evolutionary cause of theory to be "significant."

Ultimately, one evolutionary theory or model is relatively more significant than another just to the extent that the former explains more than the latter. There are examples to go around. Another, just for good measure, is the apparent greater significance or importance of individual-level selection relative to group selection.

2. Importance as time-course.

Biologists tag an evolutionary cause with the "significance" by its time-course. In other words, an evolutionary cause is relatively more significant than another when the former's effect on the evolutionary trajectory of a population occurs in a shorter time than the latter. The idea here is that a cause with a long time-course won't have sufficient time get any traction relative to one with a short time-course and, so, the former will be swamped out by the latter.

In this context, drift is sometimes said to be an insignificant evolutionary cause (e.g., by Fisher 1922). Drift's dependency on the population size, N, has a profound effect on the time required for it to have a significant effect on a population. One of those effects is a decrease in heterozygosity. The time scale for drift to decrease heterozygosity in a population by half is proportional to the population size, N. The time scale for mutation to increase heterozygosity is, similarly, proportional to 1/u. Consequently, when populations are very large, say N ≥ 10⁷, and for a rate of mutation, u, on the order of 10^-6, mutation will increase heterozygosity by half faster than drift can decrease it, i.e., 1/u « N. And when populations are smaller, say N ≤ 10⁵, drift will decrease heterozygosity by half faster than mutation can increase it, i.e., N « 1/u. It's easy to see how the significance tag may be assigned here. Moreover, it's easy to see that causes can be significant or insignificant depending on their context. Here, the ecological context in which drift and mutation are occurring essentially controls the relative significance of each.

These two ways of thinking about the significance of evolutionary causes and models seem to me to be the obvious ones. Actually, I haven't seen any discussion of significance measured by time-course in any philosophical literature. But it's apparent in evolution texts (e.g., Gillespie 2004). I have heard from philosophers, but not read, that the importance tag may be assigned to a cause that explains, for lack of a better phrase, "interesting" evolutionary phenomena. I think this is not a good way to think of importance, since it's then a function of what we think of as interesting. It's not at all clear that any particular evolutionary phenomenon is objectively more important than any other.

There must be other ways of understanding these importance/significance tags. Help is appreciated.

References

Beatty, J. (1995), "The Evolutionary Contingency Thesis", in G. Wolters and J. Lennox (ed.), Concepts, Theories, and Rationality in the Biological Sciences. Pittsburgh, PA: University of Pittsburgh Press. Pp. 45-81.

Coyne, J. A., N. H. Barton, and M. Turelli (2000), “Is Wright’s Shifting Balance Process Important in Evolution?” Evolution 54: 306-317.

Fisher, R. A. (1922), "On the Dominance Ratio", Proceedings of the Royal Society of Edinburgh 42: 321-341.

Gillespie, J. H. (2004), Population Genetics: A Concise Guide. Baltimore, MD: Johns Hopkins University Press.

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

Comments

[Roberta Millstein]

I always thought "importance" had something to do with the strength of the cause, where the strength of the cause is measured in the amount of change that the cause is able to bring about. Mutation is considered to be a "weak" cause because it brings about only very small changes in the distribution of genotypes in the population. Drift is said to be a stronger cause when it is in smaller populations than when it is large; this means that the amount of change is likely to be greater in a smaller population than it is in large. And if arguing over whether drift or selection is more prevalent in a given population, one is arguing over how much of the observed change is due to drift and how much is due to selection (though I agree with John Beatty that this issue is difficult, if not impossible, to settle).

[Matt Dunn]

"Important causes" as those that explain "interesting phenomena" is a common way of cashing out importance in macroevolution studies. But these causes aren't clearly population genetic causes, but rather about the presence of some particular kind of trait, biotic pollination or feeding on angiosperms by phytophagous beetles. But I agree that increased diversity is not an "objectively important" phenomenon.

[Robert Skipper]

Roberta,

What's the measure of strength? Observation of amount of change? If, all else equal, drift is only "stronger" in small populations, then it seems I can go back to time-course as the measure. I can do the same with mutation, probably, though rates of mutation will surely matter. With selection, "pressure" or "magnitude" matters.

So, "importance" may also be cashed out in terms of rates and magnitude in addition to time course. It's not clear that post hoc apportionings is palatable since we want to know why the apportionings are so.

Matt,

I'm happy to entertain non-PG causes. I'm just trolling for good examples of "significance" claims. It's worth thinking about why biologists claim that "interestingness" is "importance." And then to ask them why that matters....

Great comments!

[John Beatty]

Hi all. In addition to “scope” and “strength” (to the extent that strength differs from time course), you might also include the sort of "significance" measured by analysis of variance (to the extent that this differs from "strength").