I've noticed that there have been some commentaries (and I've received a few emails recently) connected with some of my statements about gender differences in spatial navigation. It can be incredibly hard to convey what I think is a subtle and nuanced story in a brief sound bite for media, and the fact that there's room for some misconstrual of my meaning is really nobody's fault. It's how all of this works. But here on my blog I have the leisure of laying things out a little and trying to convey the bigger picture as I see it.
I should probably begin by saying that the work I do in my laboratory never focuses on gender differences in the navigation tasks that we use. Yet my students and I know the literature well enough to know that when we look for differences, we are likely to find some. We have many discussions at our lab meetings about this. Sometimes we decide to downplay the differences as much as possible because they are not the focus of our study. At other times though, we (and often that 'we' is just 'me') argue that these differences are an important component of the variability in human behaviour that we want to describe, and so we should muster the bravery to wade in and to try to understand what's going on. When we do compare the genders, the differences that we identify are often nothing like those that might be predicted on a simple-minded dichotomy of gender a=good and gender b=not so good. In fact, the differences often come out in complicated higher-order interactions between our variables (and I'll admit freely that we most often find these interactions very difficult or impossible to explain and, because they are not the prime focus of our research, we don't often pursue them in any kind of determined way). So that's me and how gender influences my own work in spatial navigation. What about the rest of the field?
Here's the canonical view of the field as I understand it. In simulations of real world tasks using either laboratory situations, or virtual reality, or sometimes even in field studies of human beings doing real world tasks, a difference that often seems to arise is that women appear to rely more on a strategy based on keen observation of landmarks and men appear to rely more on a strategy based on an understanding of Euclidean geometry. This distinction has been observed repeatedly and it also has an interesting analog in non-humans. Even laboratory rats show some signs of a similar difference in spatial strategies between the sexes. As with all aspects of human behavior, there are some modifiers that we need to attach to this basic finding. The main one is that there is a tremendous amount of variability in performance on these behavioral tasks and that there is a great deal of overlap between the genders in the distribution of that variability. What this means is that it isn't too hard to find men who do better with landmarks than geometry, nor is it hard to find women who do better with geometry than with landmarks -- I hear from such individuals all the time and I see them in some of my studies. However, the averages suggest a difference between the genders, and a pretty good sized one at that.
I find this difference fascinating. I don't understand entirely what it means. Most accounts try to put together some kind of arm-waving argument that the differences are related to differences in lifestyle of our ancient forebears. Put simply, males were more likely to explore larger ranges while females were more likely to stay closer to home territory but to know that territory and both what it had to offer and the risks it engendered in fine detail. Males had an eye for distance and females for detail. I can see the logic of the argument but it's one of those things that is very hard to bring from the realm of amusing speculation to hard science.
The implications for modern human wayfinding are very interesting as well. I think one of the main implications of such strategic differences, whether they're gender based or not, is that the two strategies that I've described -- landmark-based or Euclidean, are more or less likely to work well in different kinds of situations. In a grid city like Manhattan, Toronto, Calgary, or countless others, a strategy based on Euclidean geometry might work well, but so might a landmark based strategy. In a city with a more tortuous and winding street pattern with lots of irregular intersections, a Euclidean strategy will not work as well because it will be more prone to error, but landmark navigation could work very well indeed. Think of trying to find your way in Venice by keeping track of where North was. You might be able to do it, but I think it would take a tremendous focus of attention to pull it off.
Comments