Again. ScienceBlogling razib discusses some noises various biologists are making about levels of selection (I've touched on this topic before in the context of group selection). Sweet Baby Intelligent Designer, save us from this madness.
I've been through this before--hell, I've even published stuff related to the topic**, and there isn't much there except for intellectual masturbationuninteresting and bad philosophy. If it doesn't provide me with testable hypotheses and the conceptual tools to do so, it's just not useful. That's what happened the last go around with this in the late 80s and early 90s. Do the experiments and I'll be interested, because the last time it was a lot of yak and very little data.
I have a hunch that I'm just going to become depressed watching a bunch of very bright people waste their time over this.
Related note: The popularization of this discussion is going to be even more infuriating, as any philosophical and analytical rigor that is actually applied to the topic will go by the wayside in the popular press.
*Here's the abstract cuz the article is probably behind a firewall (and it's not listed at Medline):
Previous analyses of the selective forces operating on allorecognition systems in colonial marine invertebrates have suggested that advantages to fusion with kin have selected for the ability to recognize and fuse with related colonies. While this explanation is compatible with the observation of aggregated settlement offusible larvae in an ascidian species, it is not compatible with two other prominent features of allorecognition systems-the extensive allorecognition allele polymorphism commonly observed in natural populations and the recently reported instability of chimeric colonies. We suggest that selection for fusion with self rather than fusion with kin, oflets a more parsimonious explanation for the two features listed above. Consequently, self fusion may be a major selective force acting on allorecognition systems 'in colonial invertebrates.
- Log in to post comments
If it doesn't provide me with testable hypotheses and the conceptual tools to do so, it's just not useful. That's what happened the last go around with this in the late 80s and early 90s. Do the experiments and I'll be interested, because the last time it was a lot of yak and very little data.
you sound like my PI did when noting time spent writing papers on wright's shifting balance theory which were hard to apply empirically.
Hmmm. Well experiments aren't just going to pop out of nowhere--people have got to think about this issue a bit before well-designed tests emerge, don't you think?
I would say that this points up a problem with science as an epistemological method, not with group selection. I mean group selection has a certain plausibility to it, and if it could in principle be true and no one can conceive of a test, I think science has some serious problems.
"I mean group selection has a certain plausibility to it, and if it could in principle be true and no one can conceive of a test, I think science has some serious problems."
What would the serious problems be? If you can't find a way to test the proposition or one of its consequences, then it doesn't have many important consequences, does it? If there are other frameworks that do just as well, then scientifically - practically, who should care?
This is vaguely like the stuff that goes on around interpretations of QM in physics like many worlds vs Copenhagen vs Bohm guiding wave and such. Its all fun, but unless the different ideas actually predict different things, it is literally meaningless in science. If you think that meaningful directions for experiments to discriminate things could eventually be gotten to , that is what people should be working for in my opinion.
I mean group selection has a certain plausibility to it, and if it could in principle be true and no one can conceive of a test
well, part of the issue is getting enough money & resources together to attain the level of omniscience that some levels-of-selection models need to be validated.
well, part of the issue is getting enough money & resources together to attain the level of omniscience that some levels-of-selection models need to be validated.
Who said we couldn't match string theory? At least we're also unable to test predictions w/o godlike funding...
Who said we couldn't match string theory? At least we're also unable to test predictions w/o godlike funding...
LOL. well, i'm talking about having a shit load of ecologists track all the individuals in a local deme, do intense DNA fingerprinting, etc. so a few orders of magnitude cheaper.