seanie

I saw Dawkins and Pinker at a debate a couple of years ago. Dawkins was impressive. He was also impressively grumpy.

At one point a member of the audience accused him of reductionism.

You should've seen his face.

He looked like he could kill.

Very entertaining.

Doubting Didymus

Okay, I think we are getting somewhere.

But the change is not having any notable effect on evolution. The change can not mutate, and although it could technically be selected for, the 'improved' sheep is NOT going to have an advantage over a white sheep in a non-inky area, as to move territory is to lose the advantage of blue rain. The blue sheep can not eventually fix the selection advantage permenantly in the population. As soon as the environment changes, the 'evolutionary' change pops back to the original state.

This form of 'evolution' may explain some changes in populations, but it is not going to give you complexity. Environment-instigated population changes can not accumulate, nor progress. Therefore, would you agree that the part it plays in evolution (when separate from genes) is miniscule, temporary, and nearly irrelevant compared to heritable change?

No one suggests that the environment can be ignored in evolution, that would be just silly. Dawkins' point is that the important environmental pressures on evolution can best be measured by its effects on the heritable factors in the organism, heritable factors being the only ones likely to stick around for the hundreds of generations that evolution spreads itself over.

Agreed. There are obviously a few examples here on earth that are non-genetic heritable factors. Seeing as the definition should theoretically encompass evolution on other planets, specifying 'genes' is too big an assumption.

Why is that a fault with the definition? surely determining whether a change is heritable or not is a separate question, to be determined by experiment?

Ah! but although the non heritable change is significant, it is only significant because of its effects on heritable evolution. I don't understand why, if non-heritable changes are evolution, your third definition should specify generations.

I prefer Dawkins' definition: "the non random selection of randomly varying replicators". This definition restricts biological evolution to replicating things, but DOES NOT exclude any other factors acting on that evolution, such as environmental factors.

If non heritable change is an evolutionary change, why focus on generations at all? A lifeless earth could be said to have evolving rocks. Whenever it rains, the population evolves toward wetness, and when they dry out, they have evolved toward dryness again. The only reason evolution focuses on multiple generations at all is because when evolution is doing anything interesting at all, it is doing it using heritable traits.

[ October 17, 2002: Message edited by: Doubting Didymus ]</p>

RufusAtticus

Doubting Didymus,

Your problem is that you are thinking of adaptation by natural selection as "true evolution" and every other type of change as "semi-evolution." This is simply not true, as I think you realize, but you are still stuck in that trap.

Remember these aren't definitions, but rather descriptions. Evolution is an observed process, thus we must describe it, not define it. (I just have a personal adversion of saying that I can define evolution in any absolutive sense.)

How do you test fossils for heritable change?

Because in the third example "generations" is given as an example how one my detect change. Plus, descent with modification needs descent.

Rocks do not form a population in any biological sense. Description 3 is using "population" in that way.

Even the "uninteresting" things are still evolution. You should be beware of determing things on personally biased qualities like "interesting" and "uninteresting."

JB01

I think I understand Rufus's point.

Since the phenotype is always the result of the interaction between the genotype and the local environment, any environmental change that results in a population-wide change in phenotype has effectively changed what certain genes are "for", and should therefore be considered evolution.

Does that make sense?

pz

Not really.

There are a couple of issues here. One is what constitutes evolution: many people who are otherwise good 'evolutionists' are uncomfortable with the idea that non- or mal-adaptive change can be considered evolution. These are the same people who often argue that human evolution has "stopped" because we've eliminated some selective pressures. This is, of course, false. Random genetic drift or founder effects are perfectly good examples of evolution in action.

Another issue that I alluded to earlier was the involvement of the environment in evolution. This is actually a similar misconception: many people hear the word "environment" and think of rocks and annual rainfall and seasonal variations in temperature and prey availability. These are all parts of the story, but perhaps the largest and most significant component of the environment is the population itself! The two are in many ways indistinguishable. Change the frequency of an allele in a population, and you have effectively changed the environment.

Peez

I do not follow you here. There is an observed process that we call "evolution", but the definition of "evolution" is certainly relevant. By one definition a process would be "evolution", by another it would not. We certainly can define evolution in an absolute sense, though you might not happen to agree with our definition. Biologists define evolution as a change in allele frequency over time. That doesn't make it the "true" definition, of course, but if you are using a different definition it would be useful to make this clear to avoid confusion. The theory of evolution, as biologists understand it, is concerned with "change in allele frequency" evolution.

Peez

pz

I would suggest qualifying that as "as SOME biologists understand it". There's more to evolution than that, and many biologists approach it without ever even considering allele frequencies. Some are more interested in mechanisms of morphological change, or the history of change, or the geographic distribution of taxa...

RufusAtticus

Remember, I said I had a personal adversion to saying that I am "defining" evolution, when I feel I am describing it. If I were to define "evolution" it would go something like this.

• Evolution (noun): A term used in the biological sciences that refers to the natural processes that account for the diversity of life.
  

Some biologists do, and that is a good defination to use in High School textbooks. However, I it is plainly obvious that such a defination is very narrow and revisionistic. (In the history of science "evolution" preceeds "allele.")

Sure, population geneticists, like me, work with evolution along those lines. However, other fields which study evolution, like paleontology, don't work with alleles. Furthermore, change in allele frequency can't account for evolution by endosymbiosis.

[ October 19, 2002: Message edited by: RufusAtticus ]</p>

pz

Yeah, like those of us who are more interested in developmental aspects of evolution. Frequency of alleles in populations just don't matter when you're interested in macroevolutionary differences between species or phyla.

Although one virtue of the 'allele frequency' definition is that it is complete and inarguable.

JB01

I understand the points you're making. I was just trying to reconcile what I had always understood to be the definition of biological evolution -- the change-in-allele-frequencies definition -- with Rufus's statement.

What I had in mind was something like the following hypothetical example: suppose an isolated population of ammonites lives in a body of water in which the concentration of calcium salts is gradually increasing over many ammonite generations. Then suppose that this leads, without any particular genetic change, to later generations of ammonites having thicker shells than earlier ones.

Millions of years later, a paleontologist digs up the fossils, notices an unmistakeable trend in increasing shell thickness, and calls it evolution. By the standard definition, the paleontologist would simply be wrong. But there are good reasons not to want to say that.

1. It would make it almost impossible to state with reasonable certainty whether many types of morphological change in the fossil record are evolutionary or not, when satisfying the technical definition really doesn't seem to matter.

2. More significantly, in the ammonite population in my example, genes for 1 mm-thick shells are gradually being replaced over time with genes for, say, 2 mm-thick shells, even though the genes themselves are not changing. This is taking place under the usual conditions of natural selection. If thicker-shelled ammonites don't thrive at least as well as thinner-shelled ones, we would expect the population to either evolve new genes that make the shells thinner again (say be decreasing calcium uptake); migrate down the calcium concentration gradient; or die out. Since I'm stipulating that they haven't done this, it doesn't seem too far-fetched to claim that traditional evolution actually is going on.

What do the biologists here think?