There is a new study, published in the journal Nature, entitled "The Phylogenetic Roots of Human Lethal Violence." The study argues two points: (1) along with many other mammals and particularly primates, human lethal violence is innate because it is part of a long "evolutionary history"; and (2) However, for humans, it is also a behavior that is responsive to our cultural environment. So, over time, "culture modulates our bloodthirsty tendencies."
What is particularly original about this study is that it places human violence against the backdrop of general mammalian and primate lethal behavior. The researchers found that there is a correlation between the level of intra-group violence of those species that lie close to each other on the evolutionary tree.
Just for the reader's information, it seems that killer whales almost never hurt each other, and bats and anteaters are quite peaceable to others of their kind. On the other hand, if you're a cougar, chinchilla or marmot, things can get very dangerous and one has to stay wary of the neighbors.
Getting back to humans, almost every serious historian knows that our propensity for lethal violence has been around for as far back as we can go. Thus the proposition that this behavior is inherited from our pre-human ancestors seems reasonable. However, there is an effort on the part of some researchers in this field, including those who wrote the Nature article, to make the argument that humans are getting less violent. For instance, this study claims that among Paleolithic hunter-gather groups, roughly 2% of deaths were the result of lethal violence. Later, in medieval times, this allegedly jumps to 12%. But in the modern age, with "industrialized states exerting the rule of law," the rate appears to have fallen to 1.3%. Is all of this really accurate?
The authors are not the first to make this claim. The Harvard psychologist Steven Pinker, in a 2011 book entitled The Better Angels of Our Nature, argues that humans can and have lowered their level of interpersonal violence through creating institutions and laws that discourage such behavior.
As a general rule we should be wary of such sweeping claims about behavior over such large expanses of time. As one observer of the Nature study commented, much of the data [sources range from archeological digs to modern crime statistics] is "imprecise." The same is true of Pinker's evidence. It is due to just such challenges that such studies present these claims in terms of statistical models.