There's an old bit of folklore that children tend to more closely resemble their fathers than their mothers. There's a possible evolutionary explanation for this, and one study seemed to confirm it all. Here's why it's all bogus.
The basic problem with evolutionary psychology is that it can be used to explain practically anything if you're willing to think hard enough. Let's look at the belief that children look more like their fathers than their mothers. If we go back to early humans, you might imagine that fathers would be more likely to help raise children if they were certain that they really were the father. Thus, this heightened resemblance would function as a sort of ancient paternity test, and those babies that passed would be more likely to survive to reproductive age than those who were rejected by their fathers.
Not convinced? Well, the idea did get a measure of scientific credibility in 1995, when researchers at the UC San Diego published a study in which people were far more successful matching photos of babies with those of their fathers rather than those of their mothers.
Here's where the theory runs into problems. Every subsequent attempt to replicate these initial findings has instead shown that babies resemble their fathers and mothers equally, and if anything it might actually be the mother that they more closely resemble. Psychologist Robert French, who himself has published studies in conflict with the original paper, explains why the UC San Diego study gained such initial acceptance:
It's a very sexy result, it's seductive, it's what evolutionary psychology would predict-and I think it's wrong. It's kind of hard to distinguish 'just-so' stories from things that are really a product of evolution."
Here's what French means by "just-so stories." There are possible evolutionary explanations for everything we've discussed so far. We already know why babies might resemble their fathers, but at the same time they might not resemble them because this would help force fathers to accept their parental responsibility even in the absence of definitive evidence that the child was theirs. In this reading, the important thing is that every child has a father, even if it's not their biological one, and the lack of an easy way for fathers to know for sure helped kept them in line.
These ideas both have their respective merits, and they're both utterly contradictory. And let's not forget the initial bit of folklore itself. Researchers at Georgia Southern University offered an explanation for where that idea might have come from:
"The bias in how mothers remark resemblance does not reflect actual resemblance and may be an evolved or conditioned response to assure domestic fathers of their paternity."
You can see how fast we've gotten lost here. If everything can be explained using evolutionary psychology, then have we actually explained anything at all? All this certainly isn't meant to completely dismiss the potential merits of evolutionary psychology (although I'd be lying if I said we didn't have our doubts), but it's a useful reminder of why this is a field whose results should be treated with the greatest of caution and skepticism.