That might be the same study I was thinking of, or a different analysis of the data. What I'd read was from the Economist back in 2002. Your take seems more in line with this blog post from 2005 talking about perhaps the same study, or perhaps a subsequent one, where other reseachers were dismissing some of the author's conclusions, which seemed to be more about the maternal mother-in-law's role, though that was an aspect in the first article, too. But it definitely wasn't medieval as I had thought. I must've spent too many years studying medieval German violence and now it's sticking in my head.