Spatial ability is a bit of a fuzzy concept. There are lots of different tests that can be used to assess it, like mentally rotating a 3D shape or reasoning out how a mechanical object would work. Are all of these tests really looking at the same thing, or are they probing unrelated capacities that we’re artificially combining under the heading of “spatial ability?”

A paper in PNAS this week suggests that different spatial tests are all basically testing the same underlying ability—and that this ability is only partly explained by general intelligence. This means that spatial ability is, to some extent, independent: you can have better (or worse) spatial ability than your general intelligence might suggest. The results also suggest that about a third of the differences in people’s spatial scores can be explained by genetics.

Your genes don’t have the last word

For any given trait that people have, both genes and environment will play a role. The tricky part is working out how much of a role each can claim. Take height, for instance: if everyone has enough food when they’re growing up, the role of environment will be limited, and so people’s heights will be determined mostly by their genetics—maybe as much as 80 percent. In a more unequal society, someone may have the genes to be very tall but not get enough food as a child, and this person will end up much shorter than they might have been. In this more unequal environment, genes can explain much less of the differences between people.