Helping others at no cost to oneself is generosity, but altruism implies sacrifice, yes? We can happily disagree on that - I was merely pointing that it is a very clear and usable definition (from Sober/Wilson) and addresses all the criticisms raised by sceptical biologists.
> CB: I gotta go , but I don't think differentiation between individuals
> comes out of the fitness struggle.
I know we are speaking colloquially, but it is worth noting explicitly that the struggle for survival itself does not produce traits (in the classical model; newer more speculative research seems to suggest development pathways that might be triggered or guided by external conditions, but I won’t get into that here to avoid muddying matters further). Traits i.e., the underlying genetic mutations, occur at random (within constraints imposed by molecular biology/chemistry). The vast majority are maladaptive - they decrease the fitness of the individual and are typically wiped out because they cause the individual to be wiped out before s/he can reproduce to pass it on. The ones that happen to have - entirely by chance - a fitness advantage lead to an individual with a slightly differentiated genetic makeup (of course we all have different genetic makeup already and those are fitness neutral or negligible).
Genetic variations - which manifest as phenotypic variation - occurs within/between individuals. I assume you agree with that part. With that in hand...
> The result of a "fitness" struggle
> between individuals would be one individual survives to reproduce and
> the other individual doesn't. That wouldn't result in differentiated
> individuals co-existing and going on to become sexually incompatible.
> It would lead to only one individual and offspring continuing to
> exist. In Darwin's model, it is whole species that go extinct.
> Becoming sexually incompatible arises from geographic separation of
> populations, not some "fitness" struggle between individuals.
Species extinction is a somewhat separate matter: it occurs because species exhaust their resources, are unable to adapt to a changing environment (or adapt out of their current form), suffer catastrophic environment collapse (e.g: K-T extinction) or lose the evolutionary arms race to a predator.
In fact, what you write in the first two sentences above, taken together, is roughly what is posited by Darwin’s theory (and continued in the New Synthesis). We left off in the earlier section on a hopeful note of agreement that new traits emerge in individuals as a result of random genetic variation. Assuming that hope is justified i.e., you agree with that: a small number of these traits have a selective advantage: they stretch one (or a few) giraffe’s neck a bit longer, they make one gazelle run a bit faster, or in trickier cases when present as a recessive allele in red blood cells they provide resistance to malaria. The “fitter” individuals (the ones with these new traits) have greater reproductive success and leave more offspring sporting their traits. They in turn reproduce at a higher rate, so on … these multiple variations coexist for a [possibly indefinite] period. This is the evolution of a species based on fitness measured by natural selection.
On sexual incompatibility or speciation, as you note geographic separation/isolation and similar factors can produce a split i.e., cladogenesis. I agree with that point. What I meant - stated poorly (by mixing a particular kind of speciation - anagenesis - with fitness/selection) - was that this is not a matter of or evidence for *selection* occurring at the species level (see previous paragraph).
To see why speciation does not give evidence for selection/fitness across species, consider the case of geographic or other isolation leading to the formation of a new species. Why is it that species X1 and species X2, now separated over generations, are unable to interbreed? It is because they are too dissimilar in their genetic signature. But they had the same signature when they cohabited as species X. What caused their genetic makeup to differentiate and fixate?
In summation, natural selection for fitness, in the classical/standard model, addresses the question of how and why traits arise within species, not how species compete against each other (*). In other words: species are the unit of evolution but individuals are (in the traditional/accepted case) the unit of selection (*).
—ravi
(*) This is not to say that species do not compete for the same resources or that they do not prey on each other. In that sense, they are part of the “natural” selectors that influence the traits that are selected in the target species. Powerful lions and leopards are part of the environment that “selects” faster gazelles. But it is the gazelles that have evolved a “faster” phenotype that survive, and ultimately reproduce more and up the top speed of gazelles.
Of course this is a simplification:
1. Summarising Maynard Smith, fitness is not a property of an individual but of a class of individuals - the class in which the new phenotype/trait is expressed.
2. Phenotypes coexist with others, and it is the totality of these expressed in a class of individuals that is subjected to test by natural selection.
3. Fitness can certainly be measured at any group level: genes are coded in pairs of alleles. A trait that decreases reproductive success of the individual in whom it is *expressed* might still provide a species-level benefit since the trait might be suicidal to the individual when expressed in a few individuals but beneficial to the rest of the group/species in whom it is present only as a recessive allele.
One example might be Wilson/Sober’s brain worm, a parasite that lives inside ants and attempts to modify the behaviour of the ants so that the ants are consumed by cattle (in whose gut the worms want to end up). They achieve this behaviour modification in an amazing way: one “altruistic” individual burrows his way up into the host ant’s brain and “rewires” (there’s that word again) it to attain the desired behaviour. His/her work done, this individual perishes in the brain of the ant, giving up his/her life for the benefit of the rest, who carry this suicidal trait in as a recessive allele, to be passed on to a select few in expressed form in future generations.