« Our Greek ancestors, hey? | Gene Expression Front Page | What kind of diversity? »
April 01, 2003


I recently read four books that are centered around the idea of groups. Unto Others & Darwin's Cathedral were tracts on the neo-group selection thesis. World on Fire & Lords of the Rim deal with rapacious capitalist minorities in developing nations [1]. The Germanization of Early Medieval Christianity is pretty self-explanatory in its subject matter.

Pretty soon (as in days) I will write up something that presents the information that I've gleaned from these books in a semi-coherent fashion. Additionally, I will touch on the topics of human biodiversity and the difficulties of liberal democracy. This is a big enterprise, so I'm going to give a quick pre-review of these books....

First two books, by evolutionary biologist David Sloan Wilson, were pretty good reads. Unto Others was not too technical, but not so squarely aimed at the lay audience that it will bore people with biological backgrounds. I came into the process an individual selectionist of sorts-that's all I really had studied in college. Wilson has convinced me to give group selectionism a second look-and for now I'll say I'm an agnostic. Individual selectionism is parismonious and I find it more elegant. But if the empirical evidence supports group selectionism, well, fuck elegance, no? Wilson does present some data, but his main purpose seems to be to convince others to go into this area of research. Darwin's Cathedral especially goes off on this tangent, sometimes I get the feeling I'm reading a prospectus more than an argument. If you read Unto Others you can skip the technical primer on group selectionism in Darwin's Cathedral, and I didn't pay close attention to the second half of Unto Others since it veered into psychology and I have difficulty enough grasping my own mind let alone generalizing on the human condition.

World on Fire deals with the topic of "Market Dominant Ethnic Minorities." The prelims out of the way first, Amy Chua's jacket photo is far more flattering than the photo on the Yale web site. What's up with that? I suspect that the book photo is about 10 years old or so, she looks to be about 30, near the prime of her sexual peak in terms of physical attractiveness. That out of the way, what about the book? Well, it's a great primer on ethnic conflict driven by economic exploitation and success differentials, but I felt that it went out with a whimper, not really addressing the topic of a practical solution to the coexistence of liberal democracy with multiple lifestyles that correlate with ethnic identification. Lords of the Rim is a more narrow-focus tome that deals with the Overseas Chinese. It doesn't pretend to be public policy or social science, but rather a survey of the Chinese communites of southeast Asia and to a lesser extent North America. The fact that these communities are tied to specific localities and clans between Guangdong and Fujian on the south China coast is highlighted, and the cartel/mafia character is elucidated by multiple examples in various nations. If it was theoretically more ambitious, Lords of the Rim might have become a Sino-version of Culture of Critique. Though neither of these books deal with biology, if the group evolutionary paradigm is going to go anywhere in the human context, these are the "study organisms" that will need to be modelled and examined.

Finally, the last book is more a work of history, but James C. Russell is clearly aware of the work of the group evolutionists (see the article in The Occidental Quarterly and inspect the footnotes). At about 200 pages it is a short but dense read (footnote heavy, something I favor). The basic thesis is pretty straightforward, the German folk religion conquered Christianity and turned the outward form of that faith into a vehicle for their own inner spirit-racial soul so to speak. Russell doesn't present it in such a radical and metaphorical fashion, but that is the gist. From the perspective of a student of history, I have many quibbles, though the work as a whole was very stimulating. Though I don't doubt his background in Northern European history and ethnology, his assertions about the Proto-Indo-Europeans were tenditious at best in my opinion. That field is always disputable, the problem is that a large portion of Russell's work depends on the interpretations of Indo-Europeanists, and so he seems to present the information as more solidly accepted than I believe it is. But the general thrust of the book seems spot on to me-though in a twist of irony, his idea that the Christianity of Christendom was quite pagan in spirit is probably more amenable to Left-Liberals who are intent on forging an alliance with the Christians of color in the Third World that seem to be going through the same sort of acculturation and transmutation that the early Germans did.

To look at the individual as the basic atomic unit of study and organization is very fruitful. Some might assert that this is all we can do, that groups as simply too undefined and amorphous to truly study with any rigor. I'm open-minded on this question-just because one can't study it systematically and scientifically doesn't deny the power of something. Nevertheless, it seems clear to me that the individual is quite often the partial sum of the groups in which they participate. The attempt must be made, and these books are the first in a new wave of scholarship that is ressurecting older intellectual traditions but imbuing them with the sharp edge of modern methodologies.

[1] Yes, Lords of the Rim really needed to be titled something that didn't make it sound like a gay porn flick....

Posted by razib at 02:52 AM

So...Whats up with the babes?

Posted by: Jason Malloy at April 1, 2003 06:52 AM

Richard Fletcher: The Barbarian Conversion is a summary of the whole period from ca. 400 AD to 1386 when the Lithuanians became Catholic. The pagan practice of horse sacrifice apparently survived in Lithuania until the sixteenth century. (More at my URL: the description in Spanish of the last pagan/ first Catholic Jogaila is hilarious -- very bad table manners, and much worse).

The "our Greek ancestors" phrase is a slippery one to say the least. Few Westerners have any biological ancestors at all from ancient Greece, and except for modern Greeks and other inhabitants of the former Greek empires the ancient-Greek genetic inheritance is extremely dilute. (Given Ottoman/Byzantine history, the Anatolian Turks may have the highest proportion of Greek genes of any non-Greek people).

From the "linguistic determinist" point of view, it could be said that all Indo-Europeans have something in common, but this kind of thing doesn't stand up well. Theoretically then the Finns, and Hungarians would be quite different from other Europeans (but similiar to Mordvins), and all Europeans would be significantly closer to the peoples of Northern India than to, for example, Jews, Lebanese, Japanese, or Chinese. I find this hard to argue. (And peoples with unique languages like the Basques, Burushaski[Hunzas] and the Caucasian-speakers should be unique, but they aren't really.)

This leaves "culture". Greek "culture" had a massive effect on the whole Mediterranian area and the Middle East as far as Afghanistan. The Roman Empire was always bicultural and bilingual, and through the Catholic church Greek and Roman culture were passed on to Northern Europe.

I put "culture" in quotes because, as normally used, it seems to exclude formal institutions of a non-family, impersonal, non-aesthetic type. You do not have people saying, for example, that Chinese peasant culture has existed in a cash economy with alienable real property since 200 BC, or that Chinese peasants are acutely aware of the significance of credit, debt, and capital. But these are true, and there are other "peasant societies" where they are not true. Legal institutions would be another example: Hong Kong Chinese are significantly different than other Chinese because they've lived under British legal forms, albeit distorted by colonialism, for 150 years. American Chinese are still more different, even if to all appearances they are not well assimilated.

So anyway, I think that one thing to look for is formal public "institutions", which are the core of culture but usually thought so (anthropologists prefer to study peoples who are outside these kinds of institutions).

I haven't looked at the Germanization book for awhile, but from the Germanic conquests of ca. 500 AD up until the Renaissance the dominant political group was defined according to military values and was very thinly Christianized. Christianity consisted of supporting the church financially, not profaning churches and raping nuns (big problem in the earliest period), and following certain rituals -- fasts, funerals, weddings, prayers, baptisms (and not sacrificing horses). Duelling was a military practice which went counter to all Christian teaching but never died out; Spanish theologians (weakly) condoned duelling violence that reminds you of the Bloods and the Crips (death for a scuffed shoe). In the XIX C. duelling to the death was still a reality in Russia and the US (Andrew Jackson, Aaron Burr, and almost Lincoln).

In short, I think that the most important and most often neglected variables in pop discussions of history and politics are formal institutions of the legal, political, economic/fiscal, legal type, which often vary entirely independently of race, language-group, and "culture" as the word is usually used.

Posted by: zizka at April 1, 2003 07:48 AM

Razib, Amy Chua looked sooper-fine on Booknotes. Or on the tiny video viewer box on my monitor that was playing Booknotes.

Posted by: Justin Slotman at April 1, 2003 08:12 AM

zizka, the last 'snake groves' were burned down in the 1760s from what i have read.

Posted by: razib at April 1, 2003 01:10 PM

"...(filler space to important stuff)...World on Fire deals with the topic of "Market Dominant Ethnic Minorities." The prelims out of the way first, Amy Chua's jacket photo is far more flattering than the photo on the Yale web site. What's up with that? I suspect that the book photo is about 10 years old or so, she looks to be about 30, near the prime of her sexual peak in terms of physical attractiveness... (filler space again)"

A woman reaches her physical 'peak' usually between 18-23 yo. By the time she's 30, she's already begun her slow decline in terms of looks. NOTE, before I get tons of angry hate mail for saying the obvious, I am not saying that a woman at 30 yo cannot look good. In fact, most of the big name Hollywood actresses are already near that age or older by the time they've gotten their big break and major name recognition on the silver screen. I am simply suggesting that someone like Nicole Kidman looked better at 21 yo than her mid to late 30's. She's still beautiful but was even more so when younger.

Posted by: -R at April 1, 2003 06:24 PM

naahh...when they're 21 most women are still awkward and haven't really discovered themselves..its only when they get to their mid to late 20's that they really blossom...American women are an exception unless they manage to avoid growing to size 10 and above

Posted by: Pawan at April 1, 2003 07:19 PM

"naahh...when they're 21 most women are still awkward and haven't really discovered themselves..its only when they get to their mid to late 20's that they really blossom...American women are an exception unless they manage to avoid growing to size 10 and above"

Assuming proper nutrition, a woman has fully developed (I am using Tanner stage criteria, btw) by the time she's 18yo. I can maybe see some women who might develop fully a few years earlier or take a few more years to fully develop. However, generally speaking, suggesting women are physically at their peak by 18-23yo is valid. If this is the case, then there is only a slow, progressive decline from then on. Attractive women will still be attractive if they take care of themselves by 25yo, 30yo maybe even 35yo, but there's no way they are going to look better than when they were 18yo or 21yo- from a physical standpoint.

If you want to argue that they are going to 'discover themselves' or 'blossom' from an emotional standpoint later in life, then that may be true. If you believe that is attractive to you, that is fine too, but I am talking purely about looks.

Posted by: -R at April 1, 2003 07:50 PM

oh-about the babes-tribute to the oscars :)

oh, and "R"-when i say amy chua peaked around 30, she probably hit a plateau between 20 & 30, she looks like the type that ages well.

guys who look a baby-fat face prolly prefer early 20s, guys who like chisel & cheekbones probably dig lates 20s more....

Posted by: razib at April 1, 2003 11:42 PM

Jeez, Catherine Zeta Jones looks so f*cking hot! What the hell was she thinking marrying that old geezer? It's not like she didn't have money of her own.

Posted by: -R at April 2, 2003 01:23 AM

For a non-biologist, the idea of group selection seems pretty obvious. It’s just an extension of kin selection. I’m sure it’s not easy to mathematically model group selection as opposed to individual selection or gene selection, but just because it lacks rigor and elegance doesn’t mean it doesn’t exist. The real world after all is quite messy. But the logic seems clear. From the genes perspective, if it’s beneficial to create individuals that pass down their genes to the next generation, it would be useful to create individuals that create groups through which genes propagate themselves.

Besides if group selection didn’t exist, we’d have to invent it to understand history.

Posted by: justapolak at April 2, 2003 05:55 AM


i'm sure you know this-but the common response from the individual selectionists is that "group selection" implies that it might be viable to reduce individual fitness "for the sake of the group." the problem is that over time, all the non-altruists outreproduce the altruists, so you left with a self "group." the math is pretty elegant. david sloan wilson has a response, which i find pretty interesting and has made me pull back from that strict position-but you are right, it feels very messy-like a cosmological constant or something :)

Posted by: razib at April 2, 2003 01:15 PM

Razib, all mathematical modeling relies on simplifying assumptions. That’s why some models can be as elegant as they are useless. It seems to me that group selection works in at least two ways:

First, if groups compete, which in humans they do, and the individuals within the more successful group are more likely to reproduce, then the individual behaviors that made that group more successful are likely to spread. Second, the particular cultural practices that make one group more successful than another, is likely to be adopted by the less successful group. This is true from tribes to corporations.

It seems to me that the model you’re outlining where “the non-altruists outreproduce the altruists, so you left with a self ‘group’”, doesn’t take into account competition between groups. If sacrificing yourself for the good of your genetically related group makes that group more competitive than a group of non-sacrificers, then altruistic behavior, at least pertaining to the in-group, will spread. I’m not sure how this can be modeled from the ground up, but I’m sure if you make the simplifying assumption that cooperative groups are more competitive than non-cooperative groups, you can show that the genes of the individuals of cooperative groups will spread, even if individuals are willing to forego reproducing for the sake of their kin, clan or tribe.

Posted by: justapolak at April 2, 2003 01:59 PM

well-you are on the right track "polack"-i'm going to post a summary of d.s. wilson's argument soon. a lot of it is almost semantically simple-what is a "gene," "individual" or "group" after all? kin selection btw isn't really what i mean-that is selfish individualistic behavior-defined the proper way :)

Posted by: razib at April 2, 2003 09:14 PM

I haven't (yet) read D S Wilson's books, just some of his papers. I don't think he has ever answered the criticism, put most forcefully by John Maynard Smith, that his (Wilson's) concept of 'group selection' overlaps with what is usually regarded as kin selection, and therefore just causes (deliberate?) confusion in an already-confused debate.

Posted by: David Burbridge at April 3, 2003 02:50 AM

PS to my previous post: I have just located John Maynard Smith's review of Sober and Wilson's 'Unto Others', in Nature, vol 393, 18 June 1998, pp. 639-40. This puts the objections to D S Wilson's approach far better than I could.

Posted by: David Burbridge at April 3, 2003 10:08 AM

Considering the fact that culturally-transmitted information is orders of magnitude greater than genetically-transmitted information in humans, I would be very surprised to see selection not operating on the cultural as well as the biological level -- after all, culture appears to affect our behavior more than biology does. And note that "culture" in the anthropological sense is not limited to humans -- whales and nonhominid primates have definitely displayed cultural transmission of behavior, and it's not obvious where the lower bound of "culture" might be set (look at the transmission of hunting behavior in cats, for instance).

Feedback of culture into biological change is slow, of course, which makes it harder to see. But evolution very plainly operates simultaneously on culture and genome, as demonstrated by the Ashkenazi Jews (among many others). I see this as one endpoint of a spectrum; it's not clear where the other endpoint is, however -- what organisms have no interaction at all with others of their species?

I see the denial of group selection as possible only when evolution is defined not to include "cultural" behavior.

Posted by: Troy at April 3, 2003 11:24 PM

“I see the denial of group selection as possible only when evolution is defined not to include "cultural" behavior.”

I agree. The more social an animal is the more it evolves in relation to its social environment and the more the group as a whole, and less so the individuals of the group, evolves in relation to its physical environment. The difficulty with group selection theory is not its broad argument, which I find compelling enough to conclude it’s obvious, but in modeling group selection.

As a thought experiment, imagine a game of chess where an individual piece can veto a move based on danger to itself. It can flat-out refuse to sacrifice itself and another move will need to be made. The trick is to work both group and individual selection into the process. One way would be to assign initial altruism variables to the pieces, so that pieces would be willing to sacrifice themselves to varying degrees. Those that sacrifice themselves are not reproduced for the next iteration. Their places are taken from a composite of the remaining pieces (to ensure genetic relatedness of the group). In this way altruism has a reproductive cost. The benefit comes in if the color wins. In this case, the winning color benefits reproductively at the expense of the losing color. Though to not grind things to a quick stop, the individuals on the losing color still reproduce, just at a lower frequency than the winning color.

This sets up a conflict for the individual piece. Sacrifice yourself and not pass down your superior “altruism gene”, but allow your genetically related other pieces to pass down theirs at a greater frequency than if you were not to sacrifice yourself. Or succeed in passing down your selfish genes, but at lower frequency than if your color had won.

This model, though complicated, isn’t nearly as complicated as real life. But I think if such an experiment were set up, the results would indicate that the altruism tendency would spread if it had benefits for the group as a whole.

Posted by: justapolak at April 4, 2003 08:05 AM

Justapolak writes as if he were the first person to try to devise models of group selection. In fact, biologists have been devising such models for over twenty years.

The model he describes appears to be one of 'kin selection' (as defined by Maynard Smith in 1964), since it depends on 'altruists' conferring benefits on their genetic relatives.

The dispute between Maynard Smith and D S Wilson is largely about whether the latter's concept of 'trait group selection' should also be regarded as a case of kin selection. Wilson's model only works if altruists associate with other altruists more often than would be expected by chance in the total population. Maynard Smith points out that the main, if not the only, reason for this occurring would be because organisms living in the same locality tend to be genetically related. Therefore the effect can be explained by kin selection without any new concept. Wilson's answers to this point seem largely rhetorical.

Posted by: David Burbridge at April 5, 2003 12:56 AM

David, good to know I'm only 40 years behind the times!;-) But I'd like to ask, has such a model been programmed? It seems to me a nice programming trick, one that can become progressively more complicated to more accurately describe the real world - unitl each piece (or cell) is it's own AI unit. A long way off. Without it, I think the idea of group selection will remain vague and really more a philosophical argument than a biological argument.

I agree that group selection and kin selection are essentially the same thing. Group selection is really just a generalization of kin selection, and generalization by their nature are more vague.

I think part of the confusion steps from the idea "object of selection." What is the object of selection?

Posted by: justapolak at April 5, 2003 06:20 AM

"the non-altruists outreproduce the altruists, so you left with a self ‘group’"

This depends on how the balance between individual advantages within the group versus group advantages over competing groups works out. That is, it's no good being the last survivor of your tribe...

Beyond that, many species show behavior patterns that reward "altruistic" behavior. E.g., consider a flock of birds where the male birds will take turns staying on watch while the rest of the flock feeds. The females may be noting which males are up there on watch the most, and those males may thus get the most reproductive chances. In some species, this is a big enough advantage that the males will actually fight each other for the sentry positions. Of course, being able to stand watch at all rather than feeding whenever possible means that the male is pretty good at finding food for himself, suggesting superior genetics...

So there is a self-reinforcing closed loop that rewards all the participants. To some experts, this means that the sentries are not "altruistic" at all, but I think that it means that at some time in the past one or a few flocks evolved the behavior patterns forming this loop, and their genes spread across the species as predators gobbled up the flocks that didn't post sentries.

Of course, there are also half-way versions of this in some species. In baboons for instance, young males will show off by taunting any well-fed lion that wanders nearby in what seems to be a pretense of protecting the pack from a predator. It's a pretty safe pretense, since lions do not normally prey on monkeys. But when a leopard comes into sight, the same young males will rush to the tree-tops, leaving females with children behind. Looking like a hero can't pay off if it puts you into a leopard's belly...

In humans, the analogy to the sentry birds is soldiers, policemen, and firemen. There's an old saying that women love a uniform. (My wife certainly did, long ago when I was in the Air Force.) On the flip side, in England in WWI, it was common for women to taunt young men who avoided military service with white feathers - there are records of men who couldn't pass the physical committing suicide due to all the taunts. Since human behavior is largely modified by culture, it's hard to say if there is any genetic element to this. It's often not true in the USA since the Vietnam War (my wife, from a conservative rural family, being an exception). I doubt that soldiers get much respect in western European countries anymore.

But the question is, once a society has become so "civilized" that soldiers are not valued, how long will it be before that culture is wiped out by a somewhat more barbarous one that encourages it's best and brightest to join the military?

Posted by: markm at April 5, 2003 09:33 AM

Justapolak: I'm sure that various models of group selection have been run as computer simulations. One example mentioned in the literature is quite early: Levin & Kilmer, 'Interdemic selection and the evolution of altruism: a computer simulation study', in the journal 'Evolution', 1974, vol. 28, pp. 527-45.

I wouldn't go so far as to say that group selection and kin selection are the same thing, but the two concepts do overlap and intertwine. In some models of group selection, groups are formed out of unrelated individuals, with varying proportions of 'altruists', and it can be shown that under certain circumstances altruism can spread through the differential extinction of groups despite being disadvantageous to altruistic individuals within the group. But for this to work the conditions are fairly stringent, especially that there must be very little migration between groups. Also, I think it is agreed that if models of this kind are to work the fitness effect of 'genes for altruism' must be more than simply additive.

For useful introductions to the issues I would recommend E O Wilson's 'Sociobiology' (the full text edition), John Maynard Smith's 'Evolutionary Genetics', or Krebs and Davies (ed.) 'Behavioural Ecology: an evolutionary approach' (chapter by Grafen).

Posted by: David Burbridge at April 5, 2003 10:12 AM

I forgot to mention the essay on 'Innate Social Aptitudes of Man' in W D Hamilton's 'Narrow Roads of Gene Land', Vol. 1, which is facinating all round (but don't try following the algebra unless you are a much better mathematician than I am - not that that is difficult!)

Posted by: David Burbridge at April 5, 2003 10:19 AM

David, I didnít mean to leave the question out there without attempting to answer it, but real life interfered. And Iím also pretty familiar with the literature involved so I donít want to sound like I think Iím coming up with anything new, I know there is plenty of research work going on in the kind of system I tried to outline. I also recognize that nothing short of AI will go far in modeling evolutionary processes, so really satisfactory models of group selection is not going to be possible anytime soon. Interesting sites Iíve found that do a good job of exploring the philosophy of evolution is the edge and kurzweilís site and the many associated links and many of these notions are cobbled together from ideas found there.

For now the group selection remains as Razib said, mostly a matter of semantics. And one of the ideas that from what Iíve read still remains murky is the ďobject of selection.Ē Biologists seem to be split between the gene and the individual as the object of selection. From a reductionistís (Dawkins, et.al.) point of view, the gene as the object of selection is far more coherent. It is, after all, the gene which replicate, and a gene is useful as the object of selection because it could be reasonably approximated as a discrete variable. But in reality a gene is a snippet of DNA that codes for a particular protein. So itís probably better approximated as a pattern of information or a snippet of code. So in this sense it is the snippet of code that can be more accurately described as the object of selection.

The other idea stretching back to Darwin and on thru Meyer is that the individual is the object of selection, because it is thru the individual that genes get passed on. To properly model an individual requires a genome, though for the sake a practicality, an artificial genome will do. Also needed are certain assumptions and set certain rules (laws of the artificial nature) by which the artificial genome operates. Ultimately the genome needs to run a program whose purpose is to woo another program of the same species and create a bunch of variations on the two original programs and then die, ad infinitum. In this scenario, I agree with the reductionists that using the individual as the object of selection creates modeling difficulty and lacks some coherence because the individual does not persist, nor does itís genome, but itís the snippets of code that persist.

Though I agree with this reductionistís logic, as a practical matter, itís not especially useful. I think in studying human behavior, group selection is an especially useful tool in at least the two ways I mentioned: Individuals evolve in relation to the group as well as the physical environment and the group evolves in relation to other groups and again, also in relation to their environments. You mention that group selection can make sense, but that ďthe conditions are fairly stringent, especially that there must be very little migration between groups.Ē When there are language and cultural barriers between peoples, it strikes me that outmigration was not especially common in our ancestral environment. Just as a gene is not free to wander from genome to genome, most animals within social species are not especially free to wander from group to group. It can be done, but thereís a good deal of risk.

To markm, good point. Altruistic behavior is probably awarded thru sexual selection, so individuals who display altruistic behavior are rewarded, before they go off and get themselves killed, with more mating opportunities. And the risk of getting yourself killed, is offset by the booty at the end of the battle.

Real life rears itís ugly head againÖ

Posted by: justapolak at April 5, 2003 01:50 PM

Steve (justapolak) points out that the "snippet of code" contained in the gene may be the object of selection rather than the gene, noting that "the individual does not persist, nor does itís genome, but itís the snippets of code that persist." Two thoughts:

First, the genotype obviously doesn't determine the phenotype with a probability of 1.0 -- the environment (beginning with other genes in the same conceptus, and including the uterus in mammals, which itself is subject to influence from environment) modulates the genetically-programmed outcome and therefore affects selection forces; not only are some of these effects beyond the control of the gene, but some are actually "culturally" controlled. (And of course I recognize that culture itself is influenced by genetics; the feedback is too complex to model adequately, which is part of the problem.)

This raises the second point: there is "cultural code" at work in addition to the "snippet of code" in the gene, and both sets of code are subject to selection forces; furthermore, they both tend to persist. I would posit that it is the information itself which is the object of selection.

Posted by: Troy at April 5, 2003 04:51 PM

I am not adamantly opposed to group selection, but I do think it should be treated with great suspicion, because:

(a) historically, there is no doubt that loose 'good of the species' thinking was a serious obstacle to progress in evolutionary theory

(b) it is often less parsimonious than an approach based on individual selection plus deme structure (which needs to be built into any full population genetics model in any case), and

(c) some people seem to be biased in favour of group selection for political/ideological reasons, because they think it entails co-operation rather than competition.

Justapolak: sorry if my earlier posts sounded too confrontational!

On a factual point, most animals do not live indefinitely in small reproductively-isolated groups. If they did, the groups would go extinct due to chance fluctuations in conditions. They would also become highly inbred, which usually reduces fitness. In fact, most social species have a dispersal phase during their life cycle, during which males, females, or both, disperse outside their original group. E.g., among chimpanzees the females disperse, while among elephants the young males are driven out of the herd when they reach sexual maturity.

Among primitive humans (including modern hunter gatherers), many tribes systematically practise exogamy, exchanging females with other tribes or clans. There is also a great deal of informal exogamy as a result of capturing females during warfare.

Posted by: David Burbridge at April 6, 2003 04:25 AM

I don't have any disagreement.

Posted by: zipcode at October 12, 2003 05:54 AM