Wow. Talk about major failure. A new study out correlates levels of Foxp2 with levels of vocalization in rats: basically, male rats squeak more than female pups when they’re stressed by separation from their mothers, and mothers tend to rescue the rat who squeaks the loudest. They then found higher levels of Foxp2 in males, and also found that reducing male Foxp2 levels in male pups with siRNA also reduced vocalizations. So far, so good; looks like a reasonable and interesting experiment. Then they extended it to humans half-assedly, finding that 4 year old boys have lower levels of Foxp2 in their left hemisphere (the side that largely controls speech) than 4 year old girls; they could not do the siRNA experiment in human children, obviously.
And that’s where it goes so, so wrong. The researchers assumed that women talk more than men, four-year old girls have more Foxp2, therefore their results conform neatly to what they observed in rats.
You know all the times that men complain about women talking too much? Apparently there’s a biological explanation for the reason why women are chattier than men. Scientists have discovered that women possess higher levels of a "language protein" in their brains, which could explain why females are so talkative.
Previous research has shown that women talk almost three times as much as men. In fact, an average woman notches up 20,000 words in a day, which is about 13,000 more than the average man. In addition, women generally speak more quickly and devote more brainpower to speaking. Yet before now, researchers haven’t been able to biologically explain why this is the case.
The only problem here is that that statistic is false, and men and women talk at about the same rate. Oops.
This is so unfortunate. There is evidence that girls on average learn to talk a little earlier than boys, and that would have been a safer correlation, especially since they’re describing different levels in young children. That interpretation is still fraught, though: they haven’t worked out cause and effect, they know nothing about how Foxp2 mediates this vocalization difference (and we don’t even know if it’s a direct effect on vocalization; it could also modulate a stress response), and in humans, we don’t know if different socialization pressures could be causing differences in the expression of this gene.
This is why scientists are supposed to be cautious in their interpretations. Especially when trying to explain human behavior, there’s far too much temptation to push the results to fit stereotypes as a kind of unconscious validation.
Lynna, OM says
Women don’t talk more. Men think they talk more.
chigau (違う) says
p2News = male rats squeaking.
That graph says that men actually talk more than women. The difference is only statistical and very slight, but also very real. It doesn’t mean that all men talk more than all women.
Josh, Official SpokesGay says
Everything is just so fucked up. Everyone is just so thoroughly fucked up in this soup of gender crap we brew in. I wish there were a way to make everyone take the red pill.
Chris Clarke says
They’re confusing loquaciousness with vocabulary.
You cannot judge that by just looking at the graph. No statistics are reported, not even sample sizes.
But I can tell you what that graph does say, and it has virtually nothing to do with the question of sex differences in total word use (i.e. words/d). First clue: the word-count range is only up to 2K words, whereas the daily wordcounts in the (yes, almost certainly bullshit) stats in question were 7K-20K/d. Then, it sez “all conversations,” which, what?
If you track down the source, it turns out these are word counts for staged conversations in which people are handed a telephone and told to ‘talk to this stranger for up to 10 minutes about topic X’.
It’s clearly a special, unusual, artifical circumstance and it isn’t evidence of any kind wrt the actual question at hand.
It’s data-bluffing. And it’s lazy.
The graph says that men and women use about the same number of words during a 10min phone call – not very surprising. It does not take into account how many phone calls men or women make and how long they last.
Er… I’d encourage everyone to read the original study, if you have access. It’s not nearly as bad as the pop reports, promise!
Researchers are often put in the unfortunate position of having to exaggerate their position to get and maintain funding, especially when applying for funding from “human health” sources for what amounts to basic (i.e. not “applied”) biomedical research. This can and sometimes does become a habit, and leak through to their published articles. But this particular paper (and I’ll say from knowing the people involved that this is true of that particular lab in general) is not nearly as horrendously sexist as the media makes it sound. They look at biological correlates of behaviour, have cited early developmental research showing trends for developmental differences between girls and boys, and don’t discuss anything about adults beyond than citing a review discussing the problems with the information we do have.
I’ll also leave this here; it’s the last paragraph from the original article:
Males squeak MORE than females when in distress. Mothers rescue the loudest rat. Therefore…men talk more than women???
Yeah, I’m totally lost.
rubymoon, to (try to) clarify:
Male rat pups (2 and 4 days old) vocalize more when distressed, and have higher levels of Foxp2 than female rat pups (2 and 4 days old).
Young girls (4-5 years old) have higher levels of Foxp2 than young boys 4-5 years old).
The sex difference in Foxp2 levels is reversed in humans vs. rats.
(And just to confuse things further, interfering with Foxp2 expression in male rat pups reduces their vocalizations and increases the frequency (as in ‘pitch’) of vocalizations, but in female rat pups increases vocalizations and reduces their frequency. Um… stuff is complicated? :P )
Yeah, that graph probably wasn’t the best one to back up the ‘no sex difference’ thing. IMO, the best data we have are from this brief article:
…but the data are presented there in a table and not a graph. Not hard to make a graph out of the data, but I can understand it being easier to copy/paste ;)
Mattir, Another One With Boltcutters says
It would be very very very nice if some of these evo-psych people would actually learn something about attachment theory and how it works. Then we could move on to why there might be an evolutionary benefit to having female offspring able to attract more maternal care via separation squeaks and the like, and what the mechanism for that is (hint – it’s extremely unlikely to be JUST foxp2 expression) and then, having done a bunch of studies in non-human animals that practice parental care, we could move on to how these mechanisms operate in humans and how the mechanisms themselves are magnified or surpressed by cultural and other environmental factors. And the role of (get this) caregiver attachment organization, which is shaped by the care a particular human or animal mother received as an infant, on their own infants’ vocalizations.
Shorter – this is a really interesting study, ruined by blathering in the discussion section. Researchers who care about (ZOMG) science should know better than to attract FoxNews and their ilk to their study via blather in the discussion section. (Wait, I think I sense a bit of a correlation here…)
Um… the study’s not perfect, but AFAIK (having spoken to her) McCarthy isn’t an “evo-psych” person, other than to say that “the human brain evolved”.
I may just be blinded by my own biases (a big part of my own research is studying sex differences in animal behaviour and physiology, and I have met with Dr. McCarthy on several occasions), but can someone point out specifically what “blathering” is being done in the discussion section of that paper? Or where it gives the impression that it repeats the errors typical in modern evo-psych?
Mattir, Another One With Boltcutters says
OK, I take the word “blather” back, but the “rat attachment behavior explains why women talk more than men” is a huge huge overreach that attracts media attention but should have been phrased (if at all) in very conditional, “areas for further research” language.
I’ll go hunt up the paper later, since it does sound very interesting. (Plus I think attachment behavior is a really interesting field.)
Okay, those women might not talk more than we men, but they sure do talk more than we men want them to, amirite? Up top! C’mon! C’mon, you were all thinking it; I just said it! *BELCH*
Giliell, professional cynic says
And even there we run into difficulties as children are not raised in laboratories.
AFAIK there’s evidence that caregivers and family do in fact talk more to baby girls than baby boys. Self-fullfilling prophecy and everything with people not expecting boys to talk earlier (and girls to walk earlier) and therefore not givig them the same input and reinforcement they give girls.
Uh, as you point out the positive results from the rat experiments, this is not a problem with reductionism, this is an extrapolation issue.
as in, extrapolation with conclusions instead of extrapolation with hypothesis.
the researchers, in getting a supported result for their hypothesis regarding Foxp2 in rats, SHOULD have made a NEW hypothesis about its applicability to human behavior, and then tested that. Instead, they simply concluded it had applicability.
this is a failure, not of reductionism, but simply of applying standard scientific method.
“Male rats squeak more than female pups when they’re stressed by separation from their mothers, and mothers tend to rescue the rat who squeaks the loudest.”
Atta boy P Zed. Keep torturing sentient animals in the name of profit … er, I mean … science. Glad to know rats are involuntarily drafted into studies like this. Oh so important to know these things, like why male and female humans learn to talk at certain rates. Wow. Riveting stuff. I’m sure the rats are pleased they could suffer for your gain.
Not to mention the chimpanzees that are involuntarily drafted into smoking studies, whereby they are surgically implanted with a smoking device, strapped down and forced to smoke until their lungs are riddled with cancer. To mention just one of thousands of studies so called “scientists” perform on unwilling subjects.
So be proud P Zed. Your profession will be duly noted as villains of history. You can count on it.
(Cue up the ivory tower dittoheads, the hive mind, to tell us about all the greeeaeeeeeeat things science does in the name of humanity, blah blah blah. Tiresome. As is this A+ nonsense, whereby the only requirement is to not believe in god. The lowest bar possible).
wow, I can’t figure out how a person could be more off base, right from the start of their inane missive.
only to the ignorant and inane.
that’s YOU , fuckwit.
Thank you. I somehow mis-read the post as saying human boys had the same higher amount of Foxp2 as the male rats, and was utterly baffled as to how that could lead to a conclusion that girls talk more than boys.
chigau (違う) says
Did you know that Harvard grants MBA degrees?
Try reading the posting again.
harvardparkingticket is back again I see. Sneering at people who want to know facts and stuff like that.
Mattir, individual researchers often have an unfortunately small amount of control over what their university’s PR department leads with. That being said, Dr. Bowers has publically made at least one stupid comment. I certainly hope that Dr. McCarthy has spoken to him about being so incredibly careless (Bowers is a postdoc in her lab.)
Ichthyic, ‘observation’ is still a big part of the scientific method, basic research still has its place. What would you have wanted them to do differently, considering what we already knew about Foxp2 expression in human brains? What ‘new hypothesis’ do you think they should have generated instead?
I don’t want to argue about whether the media portrayal of this article is horrendously sexist and stupid and wrong – it is. And I’m not saying the study is perfect – few (none?) are, and this one is certainly no exception. But I really don’t think that the researchers involved deserve the short shrift they’re being given here.
Yes, scientists do need to be cautious in their interpretations. But AFAIK, most of the scientists involved in this case ARE cautious, and choose their words very carefully. Nearly all the damage is being done by the media here, not the scientists.
chigau (違う) says
harvardmba never actually reads the post.
I figured I would be charitable.
Look how well that worked out. I guess I should have known better when the first part was the least wrong part.
Incidentally, for those looking for the original study, it’s published in the Journal of Neuroscience (33(8): 3276-3283):
The abstract is freely available if you don’t have full access to the journal content:
I’d like to point out the final sentence in the abstract, which reflects the nature of the discussion in the paper itself:
“We extended these observations to humans, a species reported to have gender differences in language acquisition, and found the amount of FOXP2 protein in the left hemisphere cortex of 4-year-old boys was significantly lower than in age-matched girls.” (emphasis mine)
Crap, my bad. I read the paper through, and there is indeed one specific part in the discussion that could be noted as making an unfounded claim:
Now they do speak throughout the paper about ‘girls’ and ‘boys’, not adults, and they do talk about the “linguistic advantage of girls” while studying 4-year-olds specifically, so a charitable reading would be that they’re only talking about the “more communicative sex” at a certain developmental age. But yes, that clarification is lacking and that sentence (especially out of context) could be very easily interpreted as referencing women of all ages.
In all fairness, the way PZ writes it makes it sound as if the researchers, the scientists who wrote the article spew that nonsense. But it’s not. It’s a couple of pop science “reports”, horrible science journalism, that produced those ridiculous, inaccurate statements. From Language Log:
The last sentence of the abstract is overstretching, of course, but it’s not anywhere near as bad as the mindless inaccuracy and sensationalism of the pop sci reports:
Nick Gotts (formerly KG) says
This article, linked form PZ’s first link is a better reference than the Fisher 2003: an experiment recording samples from subjects’ everyday lives, which showed no significant difference between women and men (about 16,000 words/day in both cases).
Annnnnd… screwed up again. The problematic comment from Dr. Bowers isn’t really the one I linked to in #28, it’s this one (including the text this time so people don’t have to click the link):
“We can’t say that this is the end-all-be-all reasoning, but it is one of the first avenues with which we can start to explore why women tend to be more verbal than men.”
Honestly, I could understand such an unfounded assumption coming from an undergrad or even a junior graduate student who is often relying on what others tell them for information, but a postdoc actively studying sex differences really should know the literature better.
See, here’s what gives me A Sad about this particular post: unfortunately, PZ didn’t just ‘make it sound’ that way, he outright said:
…without justification outside of claims made by reporters who couldn’t quote the researchers themselves saying anything incriminating. I’ll happily eat my defense of the lab, the study, and the article if I’m wrong, here, but in this case it seems pretty clear that PZ means to criticize the researchers and the study when the faults he’s picking out all lie in the media reports of the work.
If anyone has a suggestion for how to better approach the issues addressed by this study, I would love to hear it. I’m not being sarcastic here, btw, I have a vested interest in such discussions given the research that I do myself! How were the methods of this study biased by any a priori assumptions the researchers may have had? How would you determine whether your animal research might be applicable to humans? How would your interpretation of the published results differ from the interpretation(s) presented in the published article?
That pop media present biased and overblown versions of scientific reports I hope is well-known by the commentariat here. Less well-known perhaps but no less evident when you look for it are the biases and overblowing introduced by university PR reps/departments in official press releases. If you’re going to criticize (or report on!) a study, please at least actually read the thing you’re criticizing :(
Profs. Liberman, Pullum and Zwicky really should be invited to far more skeptical meeting/pubs. As evidenced here, too, even seemingly rational people are often linguistic creationists.
And they’re fun in general. (Even if I’m banned from Zwicky’s personal blog.)
And Pullum juggles seafood for entertainment.
Everyone knows the claims about smoking causing cancer is just Nazi propaganda.
Origins of crap media circuses about science:
1. The researchers themselves (e.g., Kanazawa, Wakefield)
2. The researchers’ junior colleagues who took a role in the research but do not understand the full complexity of it
3. The university’s media department
4. Journalists, especially those with no background in science reporting
5. Interest groups who trawl the literature with a selective filter that makes them see only those papers they think they can use to support their agenda, e.g. the DI, anti-vaxxers, the Heartland Institute.
6. Celebrities with a bad case of Kruger-Dunning Syndrome
In this case, I think the fault lies with #3, #4, and to a lesser extent #2. I don’t think it’s fair to throw so much of it at #1 (except for swallowing the women-talk-more-than-men stereotype uncritically, which the authors should have known was not supported by the literature by the time they wrote the paper — this particular criticism is full warranted).
…Except that uncritical acceptance of this stereotype by the researchers is not actually evidenced by the published article. Not in the stated justification for doing the study, not in the methods used, and not in the presentation or interpretation of the results. There is one sentence in the discussion that might, especially if taken out of the context of a developmental paper, be interpreted as suggesting it. Nothing outside of the article, except for one sentence by one author that “women tend to be more verbal than men”. Now yep that could absolutely mean “women talk more amirite!”, but it also could have been cherry-picked from a discussion with a reporter and mean nothing more than “women developed their verbal skills prior to men”.
(Seriously, ever spoken with a reporter about something you’ve done? Now try correcting the record after your words have been repeatedly misrepresented by all the others who never bothered to try to speak with you directly or even reference the work you’ve actually done. Even minor missteps that you’ve made every effort to correct can haunt you…)
Now sure, there are TONS of perfectly valid criticisms to be levelled at any given work coming out of that lab, at their general approach to a given problem, at the direction they choose to take their work… hell, at the state of the study field of sex differences in general that I’ll gladly jump on board with. But as far as I can tell (and let me say that I’m more than willing to be corrected here), none of the criticisms discussed here actually apply to the researchers involved or the study cited.
having now read the paper, I mostly agree with you, but as to your earlier comment about foxp2 in humans having been studied?
this is correct, but NOT in the context this article implies. Right now, all they have is correlative extrapolation, and the NEW hypothesis I requested, that seemed obvious to me, is still the same:
they still need to actually test that this applies to humans in a similar fashion to rats.
sorry, but that’s really pretty basic.
You are right that their statement, while a minor part of the paper, is being overblown by the discussion of it here. However, that does not mean that they should not have presented it as more “work in progress” than conclusion.
btw, as an aside, I used to work under Howard Bern at UCB. I’d be curious to see your lab’s publication lists, as behavioral endo was always an interest of mine.
you should try it sometime when working with animals like sharks and see what happens.
it teaches you directly to vet the people you grant interviews to.
I have to say, that’s an excellent point.
What kind of conclusions can you draw on human nature based on six samples of university students?
In fact, I’m just gonna shut the hell up and let you emphasize it.
LOL @ sharks and the media… yeah you’re not kidding…
By “what we already knew” I did mean that to include “…and what we didn’t already know”, and I’m really very curious as to how others would go about seeing if a particular line of animal research might apply to humans. I spend much of my time trying to get grants for basic fact-finding research that just isn’t where the money is going right now.
What new hypothesis you might have proposed? Because for the human part, I took theirs to be as simple as “Foxp2 is involved in gender* differences in the pattern of speech development in humans”, which seemed reasonable to me based on what they found in rats, what had been reported in other species, and the little we do know about the effects of Foxp2 dysfunction in humans, and the fact that the lab has been focussed on early development.
(*I say ‘gender’ here because samples were categorized based on what was presumably the assigned gender of each child that the samples came from. I’m more often working with animals, typically dividing them by genetic or gonadal sex, so someone please correct me if that’s not the correct term/phrase to use there.)
Er… normally I’d be happy to provide someone with citations from the labs I’ve worked in, but since this is a public forum and I’m currently trying to get a faculty position, I’m going to be ridiculously vague and say that I’m a neurobiologist and a “lab grandchild” of McEwen ;)
Just for the record I chose image to illustrate the point (which was close enough in form to the larger data set from Mehl, Vazire, etc.) that nearly all actual non-derived-from-Cosmo-or-televangelist-fluff research suggests that there aren’t big disparities in words used by women or men. I did not intend the graph to to accurately reflect all such research. I realize I might have created that impression and therefore apologize.
Also for the record, one of the lead researchers, Mike Bowers, really did get on The Today Show and say
Also, while Bowers and most of the reporters are clear that he worked with children under development he and almost everyone else spoke as if his findings help explain a phenomenon in adults that can’t be explained because it doesn’t exist: that adult women are more verbal than adult men.
And finally, speaking as a former stay-at-home dad who was quite involved in both my son and daughter’s early education I’m not surprised that 5-year-old girls might be expressing more FOXP2 than 5-year-old boys because in general up till some time around high-school age girl children actually do develop faster than boys.
But here’s the thing: girls in general also tend to enter puberty before boys. And consequently if you measure sex-hormone levels in, say, 7th grade age children you’re going to find a lot more of it in girls than in boys the same age. But! Only a complete and total moron would say something like “research into hormone levels in middle-schoolers help explain why adult women’s sex hormones are three times higher than men’s. Because… it wouldn’t explain any such thing because… by adulthood it’s not true anymore! Well, same for little kids and expression of language-acquisition genes.
You wonder some times if any of these people have actual children and, if so, whether they pay any attention to them.
Sweet mother of pearl! As demonstrated by my poorly-labeled graph, above, I’m not the world’s best science reporter, let alone best actual scientist, let alone best actual Today Show host. Instead I’m just a former stay at home dad with a blog. But at least I don’t just phone it in!
yes, that IS the hypothesis, what I’m saying is that this hasn’t actually been tested yet. I do think they extended the results of their current study into a far more complex system that hasn’t the causal links demonstrated in rats, and I don’t think they made that clear in the discussion section. It should have been formulated indeed as a new hypothesis, since the hypothesis of foxp2 activity in rats was the one tested directly in this paper.
phht. what money?
the reason I got OUT of behavior studies is because the trend was moving quickly away from NIH and NSF funding for direct studies of behavior, both in animals and humans.
I wish you luck, seriously!
Yeah, I thought it was unfortunate that they didn’t provide more justification for why they were looking at human tissue at all. But then again, J Neurosci has strict text length requirements for their intro and discussion sections, so you’re limited to how much you can really say in an article (while a good professional write might be able to say more, most researchers don’t have professional writers on staff in the lab…).
I do agree that far too many papers barely pay lip service to the idea that behaviour and its biological bases is complicated, that state is typically either assumed to be self-evident or is completely ignored, and of course it’s next to impossible to tell the difference in a publication. It’s so ridiculously complicated that one lab isn’t going to sort it out alone, and since we have limited resources allocated to science it’s also a competitive pursuit, so people will typically go for the easy pub rather than work their way through the whole system and work out all the details. No university is going to hire someone on the understanding that they’ll sort out all the details of any system before being “productive” i.e. publishing. Best we can do without somehow forcing a revamping of the whole field (although if anyone has an idea how to do *that*, I’m all ears!) is to convince someone to let us work on our one small corner at a time, and keep working to get that corner noticed as “important!”.
Oh boy do I hear *that*. Ugh :/
Oh yeah and wow do I wish Bowers hadn’t said that. I certainly hope that he feels the same and at the very least doesn’t show up online arguing the point!
I can empathize.
The common misconception that women talk more than men is the result of a tolerance level for women talking at right about zero, so anything above no talking is perceived as excessive.
You’re absolutely right about the paper. I finally got around to reading the whole thing and the worst I could find was that rather silly reference to “the more communicative sex” that you already pointed out — but nothing saying that women spoke 20K to men’s 7K words/day or anything similar, and they did refer to Wallentin’s 2009 review of the literature that found there was scant evidence for any significant sex difference in communications in adult humans.
So I hereby retract my previous criticism of the authors, reformulate it to a much lesser criticism of using a poor choice of words in one minor instance, and go back to bemoaning the media’s insatiable appetite for misconstrual.
PZ, you really should correct the post to make it clear that the paper does NOT claim to support the hypothesis that women talk more than men, even if the morning show chatter-bolloxers do.
J Dubb says
…So one study doesn’t prove anything, unless it takes down a political point that you disagree with?
I thought we were better than this.
My cognitive neuro text has an entire chapter on brain asymmetries. There is a nice large section on male/female asymmetries. Do I need to deny all this research because somebody might incorrectly use it to prop up a patriarchical way of life?
Nerd of Redhead, Dances OM Trolls says
Gee, I thought you were better than this too. Why complain? And just because there are asymmetries, doesn’t mean there it that much difference in function. Unless, of course, your presuppose that to be the case…
We hate those negative stereotypes about women, but we don’t mind those negative stereotypes about men at all, do we?
Nerd of Redhead, Dances OM Trolls says
More critical OPINION without anything to back it up…
why? are you here to reinforce them?
Really, that’s all I ask :) Although I do very much appreciate the rest of your comment… honestly, given the OP and some of the comments here I kept thinking I must have missed something huge!
I’ve actually never understood why a “chatty woman” would be a bad thing. Overall I’ve seen more evidence of women taking second place in male-dominated conversations, but I think that’s more social conformity then any innate difference – the study showing identical wordcounts makes perfect sense (especially as they were taking “all conversations”).
I love my woman’s chattiness – us both wagging our mouths off fills up any awkward, boring, empty silences. Talkative people rock.
If it suggested that women talk more than men I have this funny feeling that you wouldn’t be nearly so skeptical.
No one is denying physical similarities between the brains. Why do you think they are?
Oh, you must be confused. You see, to demonstrate that a difference in behavior is caused by a difference in physiology a few supporting arguments must first be made:
1. The difference in behavior must actually exist (e.g. you should actually demonstrate that women talk more than men before trying to demonstrate why.)
2. The differences in physiology must be functionally relevant to the behavioral difference (e.g. asymmetries between parts of the brain connected to speech and/or attention would be worth looking at; other parts of the brain probably less-so).
3. The possibility of a confusion in correlation vs. causation or even direction of causation must be shown (e.g. it’s clear that women and men behave differently but it’s never too clear whether it’s more because of culture or physiology. If it is culture, and if culture is learned, and if learning affects the morphology of the brain — it does, according to neuroscientists — then it is entirely possible that differences originating from the culture are actually causing the differences in physiology).
So you don’t have to deny the research. But you have to be careful how you interpret and report the research.
Is that OK, feefums, or do you still feel like we’re being awful hard on Professor Kanazawa?
can we all agree that this kind of mind-reading/hypothetical/asked-&-answered/faux-prediction comment is pointless and stupid? They certainly constitute no argument.
I think you’re confused…this makes little sense unless you meant ‘dissimilarities’…the original (misleading) term used was ‘asymmetry’.
uh huh. Keep on rooting for the eventual triumph of panculturalism. Time will tell.
Gratuitous rhetorical assholishness for the sake of it.
Are you trying to emulate the discourse style of Ms. Daisy Cutter or something? Bad choice, imo.