Gene/culture co-evolution
« previous post | next post »
Recommended reading: Simon Fisher and Matt Ridley, "Culture, Genes, and the Human Revolution", Science 24 May 2013. (Another version is here.)
A common assumption is that the emergence of behaviorally modern humans after 200,000 years ago required—and followed—a specific biological change triggered by one or more genetic mutations. […]
This prevailing logic in the field may put the cart before the horse. The discovery of any genetic mutation that coincided with the “human revolution” (6) must take care to distinguish cause from effect. Supposedly momentous changes in our genome may sometimes be a consequence of cultural innovation.
In certain cases this is obvious. Lactase-persistence mutations did not trigger dairy farming; they spread as an evolutionary response to dairy consumption. The higher alcohol tolerance of Europeans relative to Asians did not prompt, but followed, greater alcohol consumption in Europe. […]
Under the culture-driven view, many critical genomic alterations that facilitated spoken language, for example, might have spread through our ancestors after this trait emerged. That is, prior behavioral changes of the species provide a permissive environment in which the functionally relevant genomic changes accumulate. The selective advantage of a genetic change that increased language proficiency would likely be greatest in a population that was already using language.
This is an entirely traditional and common-sensical view, corresponding to the position advanced more than twenty years ago in Steven Pinker and Paul Bloom, "Natural language and natural selection", Behavioral and Brain Sciences, December 1990.
For universal grammar to have evolved by Darwinian natural selection it is not enough that it be useful in some general sense. There must have been genetic variation among individuals in their grammatical competence. There must have been a series of steps leading from no language at all to language as we now find it, each step small enough to have been produced by a random mutation or recombination, and each intermediate grammar useful to its possessor. Every detail of grammatical competence that we wish to ascribe to selection must have conferred a reproductive advantage on its speakers, and this advantage must be large enough to have become fixed in the ancestral population. And there must be enough evolutionary time and genomic space separating our species from nonlinguistic primate ancestors.
Eric Lenneberg made the same set of assumptions in his work in the 1960s; and for that matter, it's how Charles Darwin saw the question.
But the old and impeccable pedigree of this classical account of gene-culture co-evolution, in the origins of language and elsewhere, has not prevented the spread of "gene for X" bizarreness, or other forms of anti-evolutionary mysticism. So it's nice to see a couple of eminent scientists expressing these ideas, clearly and forcefully, in the pages of a widely-read journal.
Joe said,
June 13, 2013 @ 7:46 am
It seems to me that, without specifically detailing what these "series of steps leading from no language at all to language as we now find it", we are opening ourselves up for some more "anti-evolutionary mysticism" where non-evolutionary events may have contributed to the current state of language. I think folks find it difficult to comprehend how these "series of steps" could produce sophisticated mechanisms such as human consciousness or language in a span of 200,000 years. This is the difficulty, I think, faced by folks like Chomsky and Thomas Nagel. The explanation of "series of steps" is not enough, I think, until we can show specifically how a complex system could develop evolutionarily – and that 200,000 years is sufficient time to make that happen.
And, no, pointing to the current design and identifying that as proof of evolution won't work – that's exactly how mysticism explains phenomena. Until we come up with convincing explanations, I think, we'll surely be open to more mysticism.
Mark P said,
June 13, 2013 @ 8:10 am
I don't think it's necessary for specific changes for language to have occurred in 200,000 years. The capabilities may well have taken much longer to appear in the population, but only became a selected trait over a shorter period. It's clear that existing traits can be emphasized and bred for over much, much shorter times than 200,000 years. For example, consider the characteristic behavior of various dog breeds. The traits of domestic dogs vs wolves existed in the wolf population but were bred for over a period of maybe a few thousand years as dogs were domesticated. Once dogs were domesticated, traits like herding, fetching, sight hunting, scent hunting and others were selected and emphasized over even shorter periods.
Todd said,
June 13, 2013 @ 8:13 am
So… do we find differences in language proficiency between modern populations, in the same way there're differences in alcohol or lactose tolerance?
[(myl) There are individual differences in many aspects of speech and language skills, as in everything else — but we wouldn't expect population differences, unless there were different environmental circumstances maintained over a long enough time, as there were in adult milk consumption or common use of alcoholic beverages.]
Jonathan Gress-Wright said,
June 13, 2013 @ 8:15 am
It seems to me that as long as our theory of grammar is "irreducibly complex", we cannot conceive of any kind of variation in grammatical competence, and so we have no way to model an evolutionary pathway from non-human communication systems to human language. I believe Jackendoff 2002 attempted to devise a theory that could be broken down into simpler versions (e.g. complex phrase structures as extensions of simple paratactic operations like compounding), which could in turn be traced back to primate communication systems through a series of small mutational steps, but my impression is that this has not been widely accepted by linguists, whether for scientific or ideological reasons I'm not sure.
[(myl) Aside from Ray Jackendoff's ideas about how to break "grammar" down into a plausible series of steps, let's not forget the large number of features of vocal-tract anatomy and physiology, acoustic perception, etc., that are plausibly adapted for spoken language.]
J.W. Brewer said,
June 13, 2013 @ 10:44 am
So is Mark P positing some sort of Chariots-of-the-Gods aliens who were selectively breeding hominids in order to develop their inchoate capability for language?
J.W. Brewer said,
June 13, 2013 @ 10:55 am
I'm not sure the antidote for anti-evolutionary mysticism is pro-evolutionary mysticism. What's wrong with having the humility to say "we don't have a goddam idea where human language came from and, because both science and history are evidence-based enterprises and we're not sure at present if we will ever be able to obtain the evidence necessary to offer a non-speculative answer, we're just going to leave it at that for now"?
Coby Lubliner said,
June 13, 2013 @ 12:01 pm
@J.W. Brewer: What is wrong with that kind of humility is that it doesn't get you (mis)quoted in the media, something which nowadays seems to be essential for an academic career.
Joe said,
June 13, 2013 @ 2:01 pm
@J.W. Brewer: The aliens are us. Our domestication of dogs selected specific traits – our culture may function the same way with regards to language and consciousness.
D.O. said,
June 13, 2013 @ 2:22 pm
@J.W. Brewer: accumulation of evidence is only part of the story of how science works. Another part is putting ahead (sometimes baseless or far-fetched) hypotheses and then trying to prove or disprove them, as opposed to collecting whatever data is ready for collection and keeping an open mind.
D.O. said,
June 13, 2013 @ 2:25 pm
@Todd: some people speculate that Ashkenazi Jews were selected for reading skills.
Charles in Vancouver said,
June 13, 2013 @ 2:31 pm
This reminded me of the "Inuit Paradox" – that in a northern climate, the Inuit have much darker skin than other indigenous residents of similar latitudes in Europe. The theory to resolve this is that vitamin D was high enough in their diet (e.g. lots of fatty fish), and as a result there was no evolutionary pressure that favoured mutations to lighter skin. Whereas Europeans took up agriculture and their vitamin D intake dropped enough that there was an evolutionary advantage to preserving lighter-skinned mutations. So it's a case where behaviour steered evolution.
Certainly there are today genetic variations that reduce language competence. It's just that we view language competence as such a fundamental measure of intelligence, that a severe reduction in such competence is labelled as "developmentally disabled".
Some people with autism have trouble with language because they experience the world in such a different way. Temple Grandin (who has autism but is high-functioning and does *not* have any language competence problems, just to be clear) has spoken about how she feels that the autistic brain has some similarities to animals.
What if some of the people we view today as intellectually disabled are in fact the bearers of genetic variations that hearken back to our evolutionary past – and were marginalized by evolutionary pressures that gave an advantage to high language competence in terms of both survival and reproduction?
Jonathan Gress-Wright said,
June 13, 2013 @ 2:42 pm
I'd be interested to know what work there is out there that deals with cases of variation in grammatical competence. If it varies at the individual level I wouldn't be surprised to see it vary at the population level. It is certainly a foundational assumption of generative linguistics that grammatical competence does not vary within the species, though I'm not sure how essential this assumption is to the actual description of language.
Mark P said,
June 13, 2013 @ 3:39 pm
@J.W. Brewer: I don't think I remotely implied that aliens or any other agency bred humans for any traits. I simply pointed out that traits may exist in a sort of diffuse way in a population and then be selected through some mechanism. In the case of dogs, it was an active selection by humans. In the case of human language, I assume it was a survival trait and was selected for by no particular mechanism other than breeding success.
J.W. Brewer said,
June 13, 2013 @ 4:07 pm
It actually seems moderately surprising when one thinks about it that none of the dizzying array of different environments prehistoric human population groups lived in over the millenia when they were differentiating themselves with regard to lactose tolerance, ability to absord vitamin D via skin, etc etc etc would have created any selective pressure on language capability. Was a given level of innate language capacity really equally advantageous (in the sense of survival/reproduction/propagation-of-relevant-genes) in every single environment from the equator to the arctic circle, whether you were agriculturalists, pastoralists, or hunter-gatherers? Never materially more advantageous, and never materially less? Despite the many well-documented difficulties our poor monkey brains have with overnegation and misnegation, did the species reach some sort of Goldilocks/Panglossian equilibrium state where pretty much every subpopulation group throughout the world had just the right average level (and just the same average level) of language capacity, in the sense that there was no further incremental evo-bio payoff to the above-average?
Chris C. said,
June 13, 2013 @ 5:04 pm
This comports well with Bickerton's application of niche construction, doesn't it?
Narmitaj said,
June 13, 2013 @ 5:35 pm
You can hear Matt Ridley discuss this paper with Quentin Cooper on BBC Radio 4's Material World at http://www.bbc.co.uk/programmes/b01snlsw from about 12:55. (The other parts of the programme include Laura Mersini-Houghton on Planck and evidence for the multiverse, and Lee Smolin on time).
Jonathan Gress-Wright said,
June 13, 2013 @ 5:39 pm
If we can show how language evolved, I am pretty confident we would have to explain it in terms of the selective advantage of a discrete combinatorial system for expressing thought in the environment of East Africa approximately 100,000-200,000 years ago. Whatever one may say about other cognitive traits, the capacity for language does not seem to vary significantly across human populations.
If we can "break down" grammar into smaller chunks a la Jackendoff 2002, then at least we have a way to model the incremental evolution of the phenotype in question, i.e. language, though we are almost certainly looking at a polygenic trait here. If we just stick with Chomskyan minimalism, we cannot even model this; we have to suppose grammar just appeared suddenly out of nowhere. Of course, small genetic mutations can have big morphological consequences, so this is not completely absurd, but that suggests the genetic basis of language should be easy to isolate and I suspect that we should have found it by now.
Dave K said,
June 13, 2013 @ 8:58 pm
@J.W. Brewer–
So what's an example of an environment that would make language acquistion either more or less advantageous? I can't come up with even a hypothetical one.
Garrett Wollman said,
June 13, 2013 @ 10:05 pm
The point that J.W. Brewer is making (which I would agree with) is that Mark's post suggests that there are only two possibilities, when in fact there is a third: language ability could be a spandrel of some other mental ability that was undergoing positive selection prior to, or in parallel with, the development of culture. (One should also not forget that cultural evolution proceeds in a Lamarckian mode, whereas our biological capacities evolve much more slowly by the Darwinian route.)
AntC said,
June 13, 2013 @ 10:42 pm
@ Dave K So what's an example of an environment that would make language acquistion either more or less advantageous?
Engaging in rampant speculation, let's note that a key feature of language is displacement.
That's going to be an advantage in planning a mammoth-hunting party: "We'll stalk the herd across the tundra. You guys stand over here and edge them over that cliff. I'll make a bird call when they're about to appear over that hill. Everybody keep calm, no hollering. We don't want to scare them into stampeding all over the place. Remember what went wrong a couple of years ago and half the tribe starved to death that Winter?"
Contrast Everett's claims about Pirahã 's 'concern solely with matters that fall within direct personal experience, and thus there is no history beyond living memory.' [Wikipedia] Perhaps you don't need to talk so much when you can walk into the jungle and point?
(I did say "rampant speculation".)
Bill Benzon said,
June 14, 2013 @ 9:10 am
Since Mark mentioned Eric Lenneberg I'd just like to put in a good word for his Biological Foundations of Language (1967). It was astounding in its time and, when I've looked into it in recent years, it's still worthwhile. I'm thinking particularly of his consideration of the role of rhythm in language, where he talks about neural conduction speed and the time difference between brain-to-jaw and brain-to-lungs and trunk signals. As I recall he also had a sophisticated discussion of differential lags in ontogeny. I wonder how that discussion would hold up in the context of current eve-devo work.
J.W. Brewer said,
June 14, 2013 @ 9:47 am
myl now has another post up discussing academic speculations re how variations in environment might create Darwinian-like pressures selecting for one linguistic feature rather than another out of the repertoire apparently theoretically possible for all humans with standard language capacity.
Re Dave K.'s point, try thinking of it in purely negative terms. Being substantially below-average in ability to master "modern" language and thus in ability to communicate effectively with fellow tribe members is probably never a good thing and always to some extent a Bad Thing. But how Bad a Thing it is (in terms of how dramatically it increases your percentage odds of dying childless, either from dying young or from no one being willing to mate with you) might plausibly vary from place to place and time to time. Maybe over enough time lacking the basic genes associated with normal modern language-capacity was a sufficiently Bad Thing that it got completely weeded out of all population groups in all environments, even if that might have happened more quickly in some envioronments than others. Or maybe that transition (in which those w/o the requisite genes disappeared from the population) occurred way back in Africa when the entire species was in more or less the same environment. But that just gets us back to the question of why there would not have been any subsequent development, with uneven geographical spread, after the species had dispersed. Getting from one of Jackendoff's posited intermediate stages to the current stage was so adaptively awesome that those only capable of the intermediate stage disappeared, but the current stage has left no room for improvement? Of course, maybe we have fundamental difficulty imagining what language might be like if our biological language-capacity had evolved a step further than it thus far has, just as a species that lacks color vision might have difficulty imagining what it was missing.