Desperately seeking the secret of FOXP2


Since the early 2000s FOXP2 has shown up again and again in the press and scientific literature. Dubbed the “language gene” it exhibits evidence of accelerated evolution in the human lineage after it split from other apes. Additionally, a homolog of the same gene shows evidence of evolutionary change distinctive to songbirds and whales. Obviously this locus is involved in vocalization. Mutated mice on FOXP2 can even sing.

It isn’t difficult to connect the dots here. From 2002:

Dr. Paabo says this date fits with the theory advanced by Dr. Klein to account for the sudden appearance of novel behaviors 50,000 years ago, including art, ornamentation and long distance trade. Human remains from this period are physically indistinguishable from those of 100,000 years ago, leading Dr. Klein to propose that some genetically based cognitive change must have prompted the new behaviors. The only change of sufficient magnitude, in his view, is acquisition of language.

Klein’s thesis, advanced in The Dawn of Human Culture, is that a singular genetic change resulted in some sort of developmental cascade that allowed for the emergence of syntactically rich recursive language. And from language comes culture, and from culture comes world domination.

It was a clean and powerful hegemony while it lasted, but genomic and archaeological findings of the last decade have put such a elegant and simple model under a harsh light. With genomic technology even FOXP2 turns out to be much more complex and rich than the earlier reports had suggested. Neanderthals exhibited all the same mutations as modern humans to make them distinctive from chimpanzees. In other words, the changes on FOXP2 by and large predate the emergence of modern humanity, and go back closer to the root of the hominin lineage (Neanderthals and modern humans diverged ~600,000 years ago).

But FOXP2 keeps coming back. Why? It is an important gene. But another issue is that researchers still perceive in it the key to the holy grail of finding out what makes us distinctively human.

A new preprint (which is somewhat peculiarly formatted), takes another look at FOXP2, Human-specific changes in two functional enhancers of FOXP2:

Two functional enhancers of FOXP2, a gene important for language development and evolution, exhibit several human-specific changes compared to extinct hominins that are located within the binding site for different transcription factors. Specifically, Neanderthals and Denisovans bear the ancestral allele in one position within the binding site for SMARCC1, involved in brain development and vitamin D metabolism. This change might have resulted in a different pattern of FOXP2 expression in our species compared to extinct hominins.

The big picture is now the authors are focusing on gene expression levels as what might allow for modern human traits to be distinctive. Basically the DNA does not magically turn into protein. Biological machinery has to transcribe the sequence, and to transcribe it it has to bind to a particular region, a transcription factor binding site.

Most of the analysis involves comparing genomes of Neanderthals, Denisovans, and the human reference. I would be curious if they looked across lots of whole genomes to check if there was polymorphism in modern human populations. If modern humans with Neanderthal and Denisovan mutations had perfectly fine speech, that would be interesting.

Also, they spend a lot of time talking about how other genes interact and express with FOXP2, and all the other functions that are implicated. This is important, because of course selection may have nothing to do with speech, though perhaps speech changes are a side effect? Remember to that the Altai Neanderthal had some modern human admixture, and that one of the introgressed regions turns out to be FOXP2.

This sort of comparative genomic style research is interesting and suggestive. But we need more population wide analysis.

But, the authors do allude other work using genetic engineering where cell lines did show radically difference gene expression based on the mutation above. I do believe that CRISPR/Cas9 technology is cheap enough and going to be widespread enough that someone’s going to play around with splicing in “human” variants into primate models. Meanwhile, bioethicists will furrowing their brows about sequencing humans….

9 thoughts on “Desperately seeking the secret of FOXP2

  1. The Klein/hopeful monster thesis has always struck me as profoundly and deeply backwards. The way pop gen works is you first ask what drove selection pressure. You don’t first ask what random genes caused a change. Without selection pressure, those mutant genes won’t go anywhere.

    So one argument for where the selective pressure came from is what I’d roughly call the Michael Tomasello and Joeseph Heinrich school of thought. Where our genus homo crossed some kind of threshold where cultural learning became possible. This turned on a gene-culture ratchet that drove all homo lineages towards greater cognition. With Homo erectus of course being the first truly successful lineage to go widely global. So the encephalization growth across homo lineages reflect the fact that they all had a gene-culture ratchet putting selective pressure on them.

    In this view, FOXP2 was in some sense inevitable. In that the selective pressure continued for millions of years, and eventually linguistic capability would get selected for, however randomly it got enabled through preadaptation. FOXP2 now clearly being one key pathway to get to that goal. To be clear, the exact genetic pathways of how this happened is fascinating, and worth deep research. So the genomics research and aDNA work is super cool.

    One aspect of this ratchet frame is that we should expect that many of the homo lineages could exhibit “modern” human behavior. There was no sudden step up. Technology of course can jump. Maybe at certain population densities. So things like fire, and then farming, brought new selective cognitive pressure to bear. But the ratchet itself was running across all the homo lineages, so we should expect that language could go very deep into the past.

    Anyway, not sure how much this is a common view. Versus maybe my over indexing on certain pop science writers. But given what I’ve read this makes sense to me. What made our lineage human? Crossing the threshold to where the gene-culture selective ratchet lifts off. Which to me is clearly Homo erectus, given Homo erectus global success. So this frame means I’m a lumper, not a splitter.

  2. The Instruction of Imagination: Language as a Social Communication Technology by Daniel Dor:

    What separates us from the apes is a sequence of social and technological revolutions—one major change after the other in the life experiences of human communities. The emergence of language was one such revolution, not the last and definitely not the first. It was preceded by an entire history of revolutions, all of which brought human communities to the point where they could invent language—as yet another feat of collective genius. As language gradually established itself in human communities, individuals began to be selected for the capacity to meet the growing challenges of the new technology. It was the collective invention that eventually shaped the cognitions of its users, not the other way around.

  3. Because I was trained on Chomsky, and because most geneticists probably don’t read Language, I’ll play devil’s advocate here: the notion of a gene-culture ratchet giving rise to language seems to suggest that natural language itself should have gone through some kind of ratcheting process, and that its own developmental evolution should be recoverable. But it’s not. Human language, in its current form, is either complete or it’s not language; call it merge, or recursion, or embedding, or whatever—the only way language seems to work is if it’s always already human language (to use a postmodernism). Sure, you can argue that language’s developmental evolution is too distant in the past to be recoverable from language’s current form, but that strikes me as a rather convenient dodge of generative grammar’s arguments.

    fwiw, I’m only playing devil’s advocate here. I don’t count myself a Chomsky disciple, but going through a Chomsky-heavy linguistics grad program, I was indeed challenged with good evidence that syntactic structure doesn’t seem like something that would emerge from simpler vocalizing behaviors. I think it did, but how it did, we’re not even close to understanding how to go about answering, much less answering.

  4. Human language, in its current form, is either complete or it’s not language.

    I am quickly above my paygrade, but couldn’t you make the same argument for human intelligence?

    Rudimentary language could have developed along the same lines as human intelligence.

  5. Nathan,

    The view Kevin Laland’s *Darwin’s Unfinished Symphony* is that language was a ratchet that increased the fidelity of cultural transmission, which has an exponential relationship with the generations it persists (according to the models he describes). But where it differs from Secrets of Success (I think, since I’ve only read the reviews and it doesn’t come up) is that language was a ratchet for teaching kin, allowing higher fidelity of cultural transmission and therefore a massive accumulation of culture.

    Definitely worth checking out. Lots of pages about stickleback learning, though.

  6. @iffen

    Rudimentary language could have developed along the same lines as human intelligence.

    So what would “rudimentary language” look like? I’ve yet to see a satisfactory answer to that question, in part because language is a phenotype without a genotype, so to speak.

    If you imagine cavemen mumbling nouns and base verbs at each other, then you’re already assuming a complex system of phonology and morphology at work, which just raises the whole question all over again. If you imagine cavemen vocalizing meaning-laden grunts or whistles at each other, well, birds and pigs can do that, so you’re not saying anything about how humans went from grunts and whistles to, e.g., anaphora, agreement, recursion, and phrase embedding.

    couldn’t you make the same argument for human intelligence?

    Actually, I’m thinking that you could probably make the same argument about human “culture.”

  7. From Resurrecting Surviving Neandertal Lineages from Modern Human Genomes, on the lack of Neanderthal FOXP2 in modern humans:

    A strong depletion of Neandertal lineages spanning ~17 Mb on 7q encompasses the FOXP2 locus (Fig. 2A), a transcription factor that plays an important role in human speech and language (13). The observed negative correlation between odds ratio and divergence remained significant when East Asians and Europeans were analyzed separately (fig. S11) and when explicitly controlling for the presence of Neandertal lineages in modern humans (10) (figs. S12 and S13). These results suggest that sequence divergence between modern humans and Neandertals was a barrier to gene flow in some regions of the genome and associated with deleterious fitness consequences (14).

Comments are closed.