A gene for human speech?

In yesterday’s New York Times, Nicholas Wade reports new research on the FOXP2 gene (see original Cell paper by Enard et al. here).   If you’ve read WEIT, you’ll remember that I discuss FOXP2 as one of those potential genes that “makes us human.”  In other words, evolutionary change at this gene was supposed to have been involved in one of the key traits — speech –that distinguish us from other primates.   This supposition is based on some genetic and developmental observations:

1.  When mutated, FOXP2 causes both speech and speech comprehension difficulties in humans,

2.  The gene has evolved very quickly on the “human” side of our lineage since we split from the chimp lineage (and very slowly among mammals in general), leading researchers to suppose that natural selection for its effect on speech caused the rapid evolution, and

3.  When FOXP2 is mutated in mice, the young mouse pups have trouble making ultrasonic squeaks (a way of communicating with their mother). These squeaks were taken to be homologous with human speech.

However, I’ve been pretty cautious about considering FOXP2 a “human-ness” gene, for there is a lot of wishful thinking in this are.  We (and especially journalists!) are just so eager to find those genes that differentiate us from our “lower” relatives.  The new paper describes research designed to address this question.

What Enard et al. did was to change the sequence of the FOXP2 gene in mice so that it produced a protein having the same two amino acids that differentiate chimps from humans (there is also one amino acid difference between mice and chimps, so the “human” gene in mice actually was not exactly identical to the real human gene).  They then looked at the behavior, brains, and bodies of mice who had two copies of the “human” allele compared to mice having their own “mouse” allele.   The major effects of this genetic engineering were the following:

1.  The human gene reduced the “exploratory behavior” of mice,

2. The human gene increased the length of brain dendrites (inter-nerve-cell connections) in mice compared to those in carriers of the “mouse” allele, and

3.  Mice “pups” with the human gene produced ultrasonic “squeak calls” that had a significantly lower frequency (i.e., vibrations per second) than those produced by mice with the normal mouse allele.

From this the authors conclude that this transgenic experiment sheds light on the evolution of speech and language in humans.  It’s this conclusion (and the transgenic manipulation itself, which is the first time that an evolutionary hypothesis has been studied by changing a mouse protein to a human protein) that gave the authors such publicity, and got their work promoted in the Times.   But, as in the case of the darwinius primate fossil, I think the results are overblown.

All the authors have shown is that putting the human copy of a gene into a mouse changes the structure of the mouse brain and changes the squeaks of babies (what about the adult squeaks, by the way? Why weren’t they studied?).  While this is consistent with a role of FOXP2 in the evolution of human speech, it’s not very impressive.  FOXP2 is active in many tissues, and these effects could be merely pathologies.

Of course the definitive experiment, swapping a human or chimp gene with the copy from the other species, and observing the result, is unethical.

At any rate, I don’t think we’re much closer to finding “humanness” genes than we were before.  Indeed, there may be many, many genes that differentiate us from chimps (see WEIT for an estimate); and if the effet of each gene is very small, we won’t readily find those elusive “humanness”” genes.

The lesson is caveat emptor, at least when it comes to reading about science in the newspapers.


Enard et al. (2009).  A humanized version of FOXP2 affects cortico-basal ganglia circuits in mice.  Cell 137:961-971.


  1. Posted May 30, 2009 at 10:29 am | Permalink

    I have to ask – why do you think repeating this experiment with chimpanzees would “of course” be unethical? You seem to treat it like a black and white issue, but I would think it’s more complicated then that. I’m asking this from honest curiosity about the ethics – not saying you’re necessarily wrong. I more fully explain my thoughts at my blog: http://blaghag.blogspot.com/2009/05/human-foxp2-in-chimps-ethical-or-not.html

    Unrelated, I love your blog and look forward to meeting you at Evolution 2009 (assuming seeing your name on the list means you’ll be there). I’m going as part of the undergraduate diversity program! =)

  2. pebird
    Posted May 30, 2009 at 10:59 am | Permalink

    What I don’t understand is that if the chimp genome is 99% identical with the human while the human male and female genomes differ by about 3% in base pair size, does this mean that men are more closely related to chimps than they are to women?

    It would explain a lot.

  3. Tom S
    Posted June 5, 2009 at 12:48 pm | Permalink

    Why is it unethical to transplant the FOXP2 gene into chimpanzees? Doesn’t seem obvious to me. The “of course” makes it seem like this is really unexamined taboo rather than ethical reasoning.

    Would it be cruel if the transplant allowed the chimpanzees to speak and understand speech?

    If as a result of receiving the FOXP2 gene, they were able to speak and understand speech, and demonstrated the emotional and intellectual skills of a 3 year old, would that change our perception of how we could treat them? If so, maybe this line of experimentation is an ethical imperative?

  4. Rory Talbot
    Posted July 23, 2011 at 5:09 pm | Permalink

    As a lawyer, I will say that I see nothing unethical, immoral or even illegal about it. The real dilemma is what rights are granted to the new creatures?

One Trackback/Pingback

  1. […] But what exactly does FOXP2 do, and how could such a minute change mean the difference between chimpanzee grunts and Shakespeare? One way to answer that question is to put the human version of the FOXP2 gene into another animal, an experiment that was published earlier this year by a very large team of German researchers. That mouse didn’t suddenly start reciting soliloquies, but it did show differences in “ultrasonic vocalizations,” as well as cellular changes in a part of the brain associated with movement – which makes sense given that FOXP2 is thought to mediate motions related to speaking. Still, changing just one gene to the “human version” in an animal and leaving all the other mouse genes intact would presumably limit the impact of the human FOXP2 gene in changing the mouse brain. (Jerry Coyne wrote about the media reaction to this paper here) […]

%d bloggers like this: