Has neuroscience proven that men and women are born different? Or are male brains and female brains mostly similar? Is there even such a thing as a ‘male brain’? And how should scientists approach these questions?
To find out I spoke to Cordelia Fine. Fine studied psychology at Oxford and at University College London, and she’s now a Future Fellow at the Melbourne School of Psychological Sciences and an Associate Professor at the Centre for Ethical Leadership at the University of Melbourne in Australia. Alongside her academic work, Fine has published two books for a general audience, A Mind of Its Own and Delusions of Gender. The second book dealt with the issue of sex and gender in psychology and neuroscience, the theme of this interview.
This is the third in my series of interviews for PLOS Neuro as Contributing Editor! If you missed them, check out my interviews with Srivas Chennu and Michael Corballis. I have a regular blog over at Discover Magazine called Neuroskeptic.
NS: Back in 2010 you published Delusions of Gender (my review), a book in which you suggested that the evidence for innate psychological differences between men and women is much weaker than is generally believed. How was your book received?
CF: Obviously nobody finishes a book on this topic and, laying down their pen, thinks, “Yep, I reckon everyone will agree with this.” But I’m very happy with the attention the book has enjoyed, both in the scientific and popular domains.
In August, you along with three colleagues published a paper called ‘Recommendations for sex/gender neuroimaging research’. In this article you provide a set of guidelines for neuroscientists who are interested in researching sex differences in the structure and function of the brain. What inspired you to write that piece?
We wanted to write a constructive paper offering recommendations intended to be helpful to researchers, editors, reviewers and even science communicators. At the heart of the paper is the argument that researchers (and others) need to pay greater attention to what gender scholarship has revealed about the nature of sex/gender, which is often quite different to the implicit ‘gender essentialist’ assumptions that often seem to guide research design and interpretations in current research.
Gender essentialism is the belief that the psyches of females and males are highly distinct, and the differences between the sexes are natural, fixed and invariant across time and place.
In the ‘Recommendations’ paper, you suggest improvements for sex/gender neuroimaging research. Some of these recommendations are quite general. For example, you say that sample sizes ought to be larger to avoid the problems of underpowered studies, and you also discuss preregistration of protocols to avoid the problem of “fishing” for positive results. These problems have been noted across science – do you think that they’re especially serious in sex/gender neuroscience?
These two issues are certainly general to behavioural science, but there are a few reasons they may be especially acute when it comes to this area of research.
With regards to sample size, different implicit models of sex/gender and the brain will give rise to different intuitions or assumptions about what is an adequate sample size. According to implicit essentialist assumptions, there are there are distinctively different ‘male’ and ‘female’ brains. But non-human animal research has shown that biological sex interacts in complex ways with many different factors (hormones, stress, maternal care, and so on) to influence brain development. Because of the complexity and idiosyncrasy of these sex influences, this doesn’t give rise to distinctive female and male brains, but instead, heterogeneous mosaics of ‘female’ and ‘male’ (statistically defined) characteristics.
Tel-Aviv neuroscientist Daphna Joel has a nice way of describing the implications of this. It means that while certain brain characteristics may be statistically more common in human males than females, knowing someone’s sex doesn’t enable you to predict the particular array of ‘male’ and ‘female’ brain characteristics that person will have.
As for publication bias for positive findings, this has long been argued to be particularly acute when it comes to sex differences. It’s ubiquitous for the sex of participants to be collected and available, and the sexes may be routinely compared with only positive findings reported. As Anelis Kaiser and her colleagues have pointed out, this emphasis on differences over similarities is also institutionalized in databases, that only allow searches for sex/gender differences.
Does academic research on sex and gender itself affect the way that men and women behave and think, by changing public attitudes?
Thinking in a ‘gender essentialist’ way has been linked with a number of negative psychological consequences, including greater endorsement of gender stereotypes both in relation to self and others, stereotype threat effects, greater acceptance of sexism, and increased tolerance for the status quo.
It’s a plausible hypothesis that the conclusions of scientific research will influence cultural beliefs about the sexes. (In fact, faux scientific articles are commonly used in laboratory psychology experiments as a way to temporarily modify these beliefs.)
An interesting recent case study of this process looked at media and social media responses to a study published in PNAS reporting sex differences in brain connectivity (and speculating on their functional implications).
In their content analysis, published in PLOS ONE, Cliodhna O’Connor and Helene Joffe concluded that traditional gender stereotypes were projected onto the scientific findings, which in turn were taken to bring legitimacy to those beliefs.
One of the points we make in our Frontiers article is that brains and behaviour reflect a biosocial ‘entanglement’ – and that the outputs of neuroscientific investigations of female/male differences actually become part of that entanglement. The motivation for the Frontiers paper in part came from an earlier article we wrote for Trends in Cognitive Sciences, where we argued that the outputs of research guided by updated developmental models of sexual differentiation of brain and behaviour bring opportunities to change cultural assumptions about the sexes.
By way of a simple example, neuroimaging research conducted from an implicitly gender essentialist frame tends to take single ‘snap-shot’ comparisons of the sexes – an approach which is guaranteed not to produce any data that can challenge the notion of fixed, universal male versus female neural essences.
In our Frontiers paper we highlight, by contrast, the possibility of drawing on the principles of the contingency of female/male differences, and entanglement, to challenge the stabilities of observed differences and similarities; for example, by experimenting with context, or population.
The reason why it’s important to get the research in this area right is that it influences conversations about all forms of sex inequality. I certainly think behavioural science can inform these conversations, with the caveat that the neural correlates of ‘capability for a successful scientific career’ and ‘inherent drive to work 80 hour weeks in the lab’ won’t be identified any time soon.
You say that “researchers (and others) need to pay greater attention to what gender scholarship has revealed about the nature of sex/gender.” What could be the solution to this problem? Do we need (more) inter-disciplinary collaborations between neuroscientists and other researchers e.g. gender scholars?
I think this is one good option, although of course this would be more likely when investigation of sex/gender influences are motivated a priori, rather than analysed and interpreted post hoc. There are also some great examples of what you might call within-person interdisciplinary collaborations – researchers whose expertise and interests span across neurobiology and gender scholarship. But more simply, the hope was for our article to be part of that solution, by providing a concise and practical overview and recommendations.
You say that “neuroimaging research conducted from an implicitly gender essentialist frame tends to take single ‘snap-shot’ comparisons of the sexes – an approach which is guaranteed not to produce any data that can challenge the notion of fixed, universal male versus female neural essences” Do the same problems exist in neuroscience beyond the question of gender? I mean: do you think that the neuroscience of (let’s say) race, or sexual orientation, are also prone to essentialist assumptions that lead researchers to ask the wrong kind of questions?
I think that’s a very good question. Whenever neuroscientists are comparing categories, it will probably be helpful to consider what “kinds of kinds” (as Nick Haslam has put it) are involved: for instance, are there sharp category boundaries and discontinuities, or a seamless continuum.
On Twitter @Neuro_Skeptic