Dec 01 2015

Male and Female Brains

Is it more accurate to say that male and female brains are generally the same or categorically different? This question has been somewhat of a controversy, both scientifically and culturally. A new extensive comparison of male and female brains with fMRI scans hopes to provide a definitive answer.

First for some background, we need to address the basic question of how we even approach or address the issue of categorization. Nature is fuzzy and complex, but humans tend to prefer neat and tidy categories to simplify the task of keeping track of everything, and even to help our understanding. There is therefore frequently a conflict between our desires and reality when it comes to creating categories.

The Pluto controversy is a good example of this, one which was surprisingly heated despite the fact that there are no real social or political issues at stake. There is no objective and definitive line between planets as solar system objects and other planet-like objects. Astronomers had to come up with some rules, rules that are unambiguous to apply. Ideally, such rules of categorization will reflect some underlying phenomenon, in this case, for example, how planets form.

Categorizing life on Earth has been very challenging, and our systems have changed over time as our understanding of biology has changed. The latest system, called cladistics, classifies creatures entirely based on their evolutionary relationships. Some biologist disagree that this is the best system, arguing that morphological similarity should count also. Birds, for example, cladistically may just be one small subgroup of dinosaurs, but some would argue they are different enough to warrant their own category.

When we get to categories that have huge social and political implications, fighting over categories spreads beyond the scientific journals and meetings. The three that most readily come to mind are sex, gender, and race. I often find that such questions are approached in black and white terms – does race exist, for example. I don’t think, however, that phrasing the question that way is helpful.

The problem with trying to argue that men and women are the same, or that race is just a social construct with no biological reality, is that such absolute positions are difficult to rectify with our common experience. This can lead to rejection by some of the underlying point because it sounds like political correctness rather than a scientific conclusion.

A better approach is to ask several more specific questions. The first is, are there objective categorical differences between two or among three or more alleged groups? A categorical difference is a characteristic that is present in all of one group and none of the other, without any overlap. You can also ask, how frequent are exceptions to apparent categorical differences?

For other traits you can ask if there is a statistical difference, is there any overlap, and how likely are members of various alleged groups to have specific traits. This is where everything gets fuzzy, and it essentially becomes impossible to answer the category question with absolutes.

For example, we can look at biological sex in general. Humans mostly are sexually dimorphic, with two distinct categorical sets of genitalia. The genitalia take one of two developmental pathways, with no overlap. There are a minority of people, however, with ambiguous genitalia, usually associated with known hormonal, genetic, or developmental anomalies. Genitalia are not a continuum – there are two distinct groups with some exceptions in the middle.

If we look at a trait like height, however, we see a different picture. Men on average are taller than women, but there are tall women and short men. There is tremendous overlap along a continuum. This is not a categorical difference but a statistical difference.

Another way to look at the difference between categorical and relative difference is to ask this question: If you know someone is male, can you predict their genitalia? Can you predict their height? The answer to the former question is mostly yes, and to the latter mostly no.

The question the researchers of the current study were asking is this – are male and female brains categorically different like genitalia, or perhaps only statistically different like height, or perhaps not different at all?

They looked at four data sets of MRI scans and fMRI scans of brains, including over 1,400 samples. They looked at specific anatomical structures in the brain in which size was measured, and at connections or pathways in the brain for robustness. Lead author Daphne Joel is quoted as saying:

“We show there are differences, but brains do not come in male and female forms. The differences you see are differences between averages. Each one of us is a unique mosaic.”

They found that the differences were statistical, and not categorical. If you look at any one region or pathway in the brain, there were statistical differences between male and female. However, there was a tremendous amount of overlap. Further, as with height, knowing a person’s sex does not allow you to predict any one trait.

Further, individuals rarely had all male or all female traits. Across the four data sets they found that 0-8% had traits consistently of one sex, while 23-53% had a combination of male-end and female-end traits. Individuals are mosaics, with only statistical differences between males and females.

This does not mean that males and females are the same, or that there are no differences. It does mean that individuals are individuals. People are not mentally defined by their sex.

The paper did not address the issue of whether the statistical differences seen were due to inherent or cultural causes. I suspect the answer is both, for various traits. It seems, for example, that testosterone makes males more aggressive. However, a preference for the color pink appears to be a minor and entirely cultural difference.


What are we to make of the results of this study? To summarize my own approach, I think it is counterproductive and not scientifically accurate to deny that there are real differences between identifiable categories of people.

At the same time it is important to recognize when those differences are only statistical with large overlap. What this means is that in such cases it is not scientifically justifiable to treat individuals as members of a group. Membership in a group does not predict what traits the individual will have. It is therefore best to treat people as individuals.

From an ethical point of view this also works. The basic principle of respect for everyone’s individual dignity demands that people be generally treated as individuals. It just so happens that science supports that position also.

I think the medical profession has struck an appropriate balance. For categorical differences, like male-female biological difference, we have no problem treating people generally as members of a category. There is an entire medical specialty dedicated to female medicine (OBGyn). This is not an ethical or scientific problem.

Otherwise we treat people as individuals, but we may use statistical information to inform our decisions. This is because medicine is the practice of taking statistical group data and then applying it to individuals, knowing that group data does not always predict individual response and we have to individualize treatment as we go.

Genetic heritage, for example, is used to predict the probability of certain diseases or even the response to certain treatments. We don’t ignore race or sex in medicine, because these categories have a statistical reality that informs our very important practical decisions. But we recognize that these categories don’t always predict individual traits. Patients – and all people – still need to be treated like individuals.

18 responses so far

18 thoughts on “Male and Female Brains”

  1. carbonUnit says:

    I think your conclusion is well stated. Vive la fuzzy!

    This sentence is broken though: ” I often find that such questions are approach is black and white terms – does race exist, for example. I don’t think, however, that phrasing the question that way is helpful.”

  2. mumadadd says:


    Is there much in the way of structural difference? I had an argument about this not so long ago, and did some cursory research. I found that the study of sex differences is concerned with structure, function and chemistry. Just from a plausibility perspective, I would think that differential hormone levels might differentially affect the development of specific structures in the brain — again along statistical lines.

  3. mumadadd – the current study looked at structural and functional difference – the size of specific areas of the brain and the strength of specific connections or networks.

    CU – thanks, I did not have a chance to edit yet, but now I have. All fixed.

  4. Kawarthajon says:

    Another great article. I think the comparison of gender to race is a good one – definitely a lot of fuzz. You’ve said a few times on the SGU that you’d like to have a more indepth discussion of race and what science has to offer the discussion – no time like the present with all of the turmoil going on in the US and the world.

  5. hammyrex says:

    One amusing historical anecdote was the African-American Heart Failure Trial which showed that BiDil (isosorbide nitrate with hydralazine) worked well as a blood pressure medication in black heart failure patients, even though previous studies in a more heterogeneous population didn’t show a benefit. Despite the criticisms that medicine is practiced on too much of a population level and that “personalized medicine” will be the future, the study was heavily criticized as being the first ‘race-based’ prescribed drug. It seems incorporating race into medicine is often a damned if you do, damned if you don’t situation.

  6. pandadeath says:

    “The paper did not address the issue of whether the statistical differences seen were due to inherent or cultural causes. I suspect the answer is both, for various traits. It seems, for example, that testosterone makes males more aggressive. However, a preference for the color pink appears to be a minor and entirely cultural difference.”

    I am so glad you brought this up. Every time I read an article on this issue, I want to read a study that compares the sexes as well as different cultures. Since the sex-stereotypes are not the same across different cultures. It would also be interesting for race, but political correctness for race seems to set in much more firmly than for sex (ie, no one seems to be saying out loud that perhaps members of a certain race may just not be good at or interested in math, whereas it seems to be OK to say that about women). Personally I think it would just show that most differences between brains of sexes and races are cultural, but it would be useful to help settle the issues of sexism/racism in education.

  7. daedalus2u says:

    There are readily apparent statistical health differences between white Americans and African Americans. The source of those health differences has not been identified. GWAS can’t seem to find any genes that would cause this.

    My hypothesis is that the disparity is due to the adverse effects of of the stress of poverty, discrimination and bigotry.

    The data that falsifies most (essentially all?) alternative hypotheses for that health disparity is figure 3 of this.

    Gender ratio at birth does depend on maternal stress exposure during pregnancy. Many species will spontaneously abort or resorb male fetuses if there is exposure to high stress. The explanation is that when times are “hard”, a female is a more reliable bearer of grand children because a male deprived in childhood will be unable to “compete” successfully for mates with males who were not deprived. A female can/will still get pregnant without competing.

    In terms of brain differences, the are obviously a lot of differences between the brain of someone who speaks English, and someone who speaks French and someone who speaks both. I suspect that there is no technique to reliably image those differences, but there do appear to be structural differences in people who are bilingual.

  8. steve12 says:


    “The data that falsifies most (essentially all?) alternative hypotheses for that health disparity is figure 3 of this.”

    Come on – don’t undersell it now!

  9. steve12 says:

    I think there’s this notion that something isn’t genetical coded if it’s expression relies on a complex, but somewhat reliable, interaction with the environment.

    E.g., visual cortex will connect very differently if there’s no light exposure at sensitive periods of development. The exact layout of the visual cortex is not genetically code per se. Does this mean that our ability to see is not genetically encoded? I say no.

    It’s essentially a compression algorithm where most of the information is stored in the environment – and the algorithm won’t find a solution w/o that information.

  10. Andreas says:

    Two observations:

    1. The difference between categorical and statistical difference is itself statistical! A categorical difference is just one where the distributions of measurable features (e.g., shape of genitalia) are very peaked and/or well-separated.

    2. The perceived fuzziness of observable classes is also dependent on the extent to which a variety of features are correlated. A high degree of correlation among different features enhances the perception of underlying categories that “explain” the features. For example, sex correlates with a large number of morphological as well as behavioral observations, that makes sex and by extension the associated observations robust categories.

    My question regarding the study is: have they found that several brain features they measure are highly correlated and is that correlation explained by sex (in a statistical sense)?

  11. Sherrington says:

    This is interesting, but I wonder if these results apply to structures that are not apparent in brain scans. We know that rats (among other animals) have a sexually dimorphic nucleus (SDN) that is larger in the male and that its size is controlled by early testosterone exposure. I realize that human gender is more complex than rat sex, but LeVay and others have reported a size difference between men and women in a structure in the hypothalamus that may be analogous to the SDN; this is only observable by studying brain tissue.

  12. Iolaire says:

    About the cultural female preference for pink: recently I (female) was in an outdoor shop buying a rucksack and the model I wanted came in a man’s version, black with red straps; and a woman’s version, blue with pink straps. Luckily the man’s version was a better fit because, as I told the shop assistant, I wouldn’t have bought a rucksack with pink straps. ‘No’, he replied, ‘and neither do any other women that come in here. We keep telling the manufacturer not to make the women’s models of high-end kit in pink, but they take no notice’. It’s the same with mountain bikes: women’s versions of the higher-end models don’t sell well if they’ve got pink frames with flowers on. It’s as if manufacturers have this simple idea that if they want to sell X to females it needs to be pink, not realising that women who are seriously interested in strenuous outdoor activities don’t fit into the common cultural stereotype.

  13. ccbowers says:

    A common complication that sex/gender and race have are the difficulty in distinguishing the science from the everyday understanding and cultural and social implications of the science.

    For example with race, when someone says something like “race is just a social construct,” they might be correct in that the way the term “race” is used in everyday discourse is a just social construct (and this may have little correlation to meaningful biological differences). For example, people often group the entire continent of Africa as one group to compare/contrast, when Africa itself has more genetic diversity than the rest of the world combined. Yet the common categorization bares little resemblance to the scientific understanding.

    On the other hand, people may confuse this in a way that biases them from acknowledging that there are some genetic differences between groups of humans. This is an issue of untangling the science from the (apparent) cultural and social implications. Some people seem to have an ideological (or perhaps just reactionary) drive to dismiss any idea suggesting genetic differences between human groups. I understand that the topic carries a lot of baggage, but we still should be intellectually honest on the topic. We should not be afraid of a better scientific understanding of the topic- at worst it has nothing to say about how we should deal with the topic of race culturally or socially. If anything, it could at least help to counter the common misunderstandings of the race among the general population, which are harmful.

    It is pretty clear that between human groups (e.g., haplogroups), the genetic differences we find are not significant enough to qualify as a race (which I think would be ~= subspecies), but humans have evolved enough for some traits to cluster during the relatively short periods of geographic separation. This clustering has occurred to varying degrees due to different selective pressures during the varying periods of separation, and is one reason why people find it intuitive to think in terms of race.

    The problem, of course is the most obvious traits are largely superficial ones (skin color, nose shapes, hair color/type, eye shape, body type, etc.), yet these are the ways that people intuitively categorize others. These superficial traits, in turn, may correlate with other traits somewhat, but as a whole these superficial traits are pretty poor representatives of genetic differences. I think we go wrong when we overvalue the obvious trait as being representative because it is most salient (i.e., availability bias).

  14. CrookedTimber says:

    Interesting study , thanks Dr. N

    By coincidence I checked in on Greg Cochrane’s blog today and he also discussed the same study. I only occasionally check that site because the author is too often obnoxious and his regular commenters are worse. He does, however, sometimes make a good point. Needless to say he doesn’t think much of this study. A snippet (after some unnecessarily uncharitable commentary) :

    “If you want to determine a brain’s sex from MRI data (without cheating by looking at the Y chromosome) you build a statistical discriminator – you don’t create a continuum of “femaleness” to “maleness,” for the entire brain and score every individual region-by-region to find out where they fall on that male-to-female continuum.

    Moreover, similarity in gross anatomy does not ensure similar behavioral tendencies. If I compared the brain of a pit bull with that of a similar-sized border collie, I doubt if I could see the behavioral differences in the size of the amygdala or whatever. Those behavioral differences exist, they’re innate, they have a physical/genetic basis – but at the moment I couldn’t tell you what brain differences to look for.”

    He seems to be saying that MRI isn’t the right tool for the job. Not my field, I can’t comment on the specifics but it does seem clear to me that there are *average* differences between male and female humans in certain categories.

    Anyway just thought I would share and would be interested in the groups thoughts on this critique. The full blog post is here:

  15. zorrobandito says:

    Discussing “race” in the United States is deeply problematic because we are badly and irretrievably mixed up about who belongs in what category.

    I have a foster son whose mother was an immigrant from Lebanon and whose father was an American “black.” So just looking at that much, is the son “white” or “black?” But wait, it gets worse. The father, whose family has been on this continent for a very long time, has at least one Native American ancestor and almost certainly some Europeans as well.

    Nevertheless the father has been categorized as “black” all of his life, as has the son, who actually is less than 50% African. Furthermore, most of the rest of us whose families have been here a while have statistically a big chance that they have at least one African ancestor. (Not to mention the Native American wild card.) So when we start in on “racial” differences between “black” and “white” Americans what are we really talking about in the end? Doesn’t it all sort of dissolve?

  16. zorrobandito says:

    The male/female thing. I have been interested in the statements that get thrown around about transgender people, that M. Jenner for example has a “female” brain in a male body. I’ve been deeply skeptical about such claims, and this new study would seem to support my skepticism.

    I am certain that in some cases at least, probably most, the transgender people involved experience a profound disconnect between their interior condition and the social roles assigned to their gender identity. I’m just doubting that we know enough about brain structure to whip out an MRI and say, “see! I have a male [female] brain!” Mostly I doubt this because no one has actually done this in any of these cases.

  17. leonet says:

    @ CrookedTimberon

    Regarding Cochrane’s critique, it seems that he’s essentially proposing exactly what he’s accusing the study of doing which is choosing a statistical technique that’s guaranteed to yield the desired outcome. If there are any statistical differences at all between a set of variables, of course you can come up with an algorithm for binning them into ‘male’ and female’. It doesn’t mean your categories are worth anything (scientifically).

    The second point, that morphology doesn’t determine behavior, is stronger but it undermines his claim that the study’s approach isn’t valuable. If you think about it developmentally, humans don’t have separate “male” and “female” suites of genes for each brain structure. Rather, you get morphological differences in a structure when cells are differently responsive to a sex hormone, for example, estrogen causes certain tissues to develop and others to recede and it does so pretty reliably between individuals. The fact that the brain is a morphological “mosaic” seems to be fairly strong evidence that this is not what’s happening. In fact, I think it may be stronger evidence against innate sex differences than if the study had found even smaller size differences but ones that were more consistent between structures.

  18. mumadadd says:


    “that M. Jenner for example has a “female” brain in a male body.”

    VS Ramachandran has put forward the notion that gender dysphoria could be the result of a mismatched body map, or “homunculus” on the sensory cortex. i.e. trans people may have a homunculus that corresponds to the wrong sex. I read this in one of his books, and I can’t remember if this is well evidenced or tested, but to me this would be a more definitive case of a female brain in a male body, or vice versa.

Leave a Reply