Scientists have just made a fascinating new discovery about violence, and one that might be totally unexpected. The study, which was published in the journal, found that while murder is innate in human beings, we seem to have driven a lot out of us through civilization. And it may surprise you just how murderous we once were.

The findings are based on an analysis of 1,000 mammal species, in which scientists found that closely related species often kill their own kind at similar rates, so we can tell a lot about our tendency toward violence based on related animal species — and we live in a violent neighborhood of the evolutionary tree, the study found.

Mammals are the most violent of the animal groups at three killings for every 1,000 deaths, and medieval humans definitely set a pretty high mark for violence, topping out at 120 per 1,000 deaths. But since then, it appears that civilization has tamed the beast inside of us. The figure has declined to 13 in 1,000 in the modern age. That’s still quite violent, but a far cry from where we used to be.

The killer whale is actually the least murderous species. It gobbles up penguins and other animals, but it rarely kills its own kind. The same can be said of bats and anteaters. Meanwhile, cougars, baboons, lemurs and chinchillas are all more murderous than humans at more than 100 per 1,000.

Here is the paper’s abstract:

The psychological, sociological and evolutionary roots of conspecific violence in humans are still debated, despite attracting the attention of intellectuals for over two millennia. Here we propose a conceptual approach towards understanding these roots based on the assumption that aggression in mammals, including humans, has a significant phylogenetic component.

By compiling sources of mortality from a comprehensive sample of mammals, we assessed the percentage of deaths due to conspecifics and, using phylogenetic comparative tools, predicted this value for humans. The proportion of human deaths phylogenetically predicted to be caused by interpersonal violence stood at 2%. This value was similar to the one phylogenetically inferred for the evolutionary ancestor of primates and apes, indicating that a certain level of lethal violence arises owing to our position within the phylogeny of mammals.

It was also similar to the percentage seen in prehistoric bands and tribes, indicating that we were as lethally violent then as common mammalian evolutionary history would predict. However, the level of lethal violence has changed through human history and can be associated with changes in the socio-political organization of human populations. Our study provides a detailed phylogenetic and historical context against which to compare levels of lethal violence observed throughout our history.