The Development of the Family and Marriage in Europe makes a comparable argument.

And as Razib has noted, group selection is simply a weaker force than family-level selection.

just to be clear: when i say this i generally mean that higher level selection is weaker relative to lower level selection. i think group processes occur, but they're hard to get a finger on, and what people perceive as 'group level' dynamics can often be reduced to aggregates of individuals. i haven't read the paper, so i won't comment in detail, but i think there are some truths here....

If private property-owners as individuals have a selection advantage over (say) tribal members, it means that there is a group selection advantage on the average for members of (social) groups within which there is private property.

These social groups may or may not be kin groups to some degree; but for example, West Germans are not very different from East Germans genetically, but members of the W. German group had an advantage over members of the E. German group.

This is a different thing than pure biological group selection, but in some ways it functions that way.

Within a libertarian individualism ontology or ideology, economic competition seems like natural biological competition, but it is done within a non-natural (cultural) institutional framework allowing and encouraging certain kinds of competition and not other. (Competitive capitalist culture forbid, for example, polygyny, rape, and the murder of male rivals for sexual favors.)

As I understand it, two-level biological + cultural evolution of this type is a kind of group selection, but not exclusive of individual selection, and is not the same as the kinds of biological group selection people were talking about in the past.

One thing I read suggested that members of different cultural groups live in different environments produced by the group to which they belong. (This is self-referential, since the group I belong to is formed in part by my habitual behaviors; it's not something I respond to, or an external influence on me.)

For example, people in the developed world live in an environment in which Type I diabetes does not reduce evolutionary fitness much, whereas in a third-world environment untreated diabetes tends to be weeded out. From this point of view, you could say that there was just individual competition within different environments, but the different environments are created by culture.

Cultural + biological evolution (linked to one another) is an enormous question which I'm only gradually getting some grip on. I don't think that anyone has it quite figured out yet.)

If private property-owners as individuals have a selection advantage over (say) tribal members, it means that there is a group selection advantage on the average for members of (social) groups within which there is private property.

a semantic note: if members of group x have average fitness 1.1 and those of group y have average fitness 1.0, by definition group x is, on average, fitter than group y. but, that does not necessarily imply group selection. the reason is that group selection implies selection upon group level traits, and the difference in fitness of individuals averaged together is not a group trait. the group of growth x vs. group y can be fully reduced to the individual differences of group x vs. group y (i.e., if you tossed the two groups together the dynamics might be no different, more future descendents in the panmictic group might be descended from the orignal group x than group y). group selection would imply selection upon group traits, which are not reducible purely to individual level differences. i think what herrick is implying here is that

1) there is more within group variation than between group variation

2) in pre-ag groups selection was between groups, in post-ag groups selection was within groups. since there is more within group than between group variation, and the response to selection is proportional to extant heritable genetic variation

3) the evolution of post-ag groups would then be faster than pre-ag groups because there is more response when you are selecting within group traits

an explicit illustration is the idea of a 'winner take most' game in reproductive fitness. in a post-ag group a 'big man' might reproduce a lot, and so his individual selectable traits are being boosted. in a pre-ag context meta-population dynamics are at work, and no one individual can translate their own genetic toolkit into a fitness bonanza, groups live or die together. so the differences between groups are what is crucial now ...the relevance of within group differences is not important because superior individual members of the group don't get a proportionate increase in fitness vis-a-vis inferior members of their group. similarly, superior members of groups which don't succeed go extinct just like inferior members. in short, in this reformulation group selection is pretty wasteful of genetic variation....

Indeed, Razib is correct: His 2) and 3) clarify my key point:

2) in pre-ag groups selection was between groups, in post-ag groups selection was within groups. since there is more within group than between group variation, and the response to selection is proportional to extant heritable genetic variation

3) the evolution of post-ag groups would then be faster than pre-ag groups because there is more response when you selecting within group traits

In sum, individual observations are noisier than group averages.

N.B. There are many more good ideas in Galor and Moav's paper, and I'll surely return to that well a number of times....

Presumably, though, the higher rate of evolution in agricultural societies would itself be a group-level trait. And if peoples who discovered agriculture early have displaced those that did not (as they have), then that would be group-selection?

anyone posting here should read the wiki summary of group selection.

And if peoples who discovered agriculture early have displaced those that did not (as they have), then that would be group-selection?

this is a complicated question/answer. it seems that in north-west europe agriculture was mostly a cultural adaptation...insofar as the 'wave of advance' from the middle-east (anatolia) was very diffuse by that point. but the reason that i am cautious about using the term 'group selection' in a genetic sense is that demographic expansions are often characterized by genetic assimilation and absorption, not deme vs. deme competition and exclusion. to be exactly clear: group selection means selection which operates on the level of groups and can not be reduced to a lower level. if adoption of agriculture by group A results in increase of individual fitness, that is increase of individual fitness which sums up to an average increase in group fitness. but, it is not something fundamental on the group level. individuals in neighboring groups may adopt agriculture as a direct function of distance from the point of innovation. because of the demographic 'wave of advance' the initial group which totally adopted agriculture would be a more successful group, but only because the individuals at that locale were lucky to be near the center cultural innovation. in contrast, as one expands outward there would be different rates of adaptation/adoption within groups, and also intermarriage with the expanding initial population. modeling it as deme vs. deme extinction, that is, a winner take all competition between two teams, just isn't really illuminating in my opinion. rather, we're talking individual players on notional teams who act as self-interested players (e.g., most status and competition is within group). warfare and meta-population dynamics, where one group exterminates a neighboring group, would be a sort of group level selection since it makes irrelevant individual fitness in the group which is eliminated. but, in these cases women of conquered people are usually enslaved and used as concubines, resulting in genetic assimilation.

to use a sports analogy, you can create an all star team, or, you can make a team which operates well together. a lot of the things that people classify as group selection are really the former (e.g., cultural innovation which increases fitness within the group). you're just cranking up individual stats. in contrast, group level effects are more than the sum of their parts, and can't be reduced to just better individual traits.

The thesis about agricultural societies is plausible for the Chinese example. China is often used as an example of a place where the group dominates the individual, but China has had private property in land for two millennia, and China has been merciless toward the unlucky. There's charity only for near kin, and Leon Stover estimates that historically, even during good years 10% of Chinese were below subsistence.

I'm going to bow out of group selection discussions until I have a better grip on the concept.

The one thing I'd say again is that cultural/political/economic evolution and biological evolution are two different historical developmental paths which are complexly inter-connected. And a cultural group can have essentially zero genetic connection (e.g. an international finance corporation working out of New York).

Striking examples of the other case (genetically close but culturally different) include Poles and Polish-Americans, Sicilians and Sicilian-Americans, Lebanese and Lebanese-Americans, and (still) Chinese and Chinese-Americans.

A complication is that both economic / cultural success (W. Europe) and economic / cultural failure (Poland) can be associated with low fertility.

Based on Razib's private property argument, there would be different forces at work between areas which had slaves, serfs, hired labor, and free (landowning) peasants working the land. these forms are fairly randomly scattered, even in Europe, and don't immediately seem to be correlated with much of anything.