The Decline of Violence

Forensic archeology—a kind of “CSI: Paleolithic”—can estimate rates of violence from the proportion of skeletons in ancient sites with bashed-in skulls, decapitations or arrowheads embedded in bones. And ethnographers can tally the causes of death in tribal peoples that have recently lived outside of state control.

These investigations show that, on average, about 15% of people in prestate eras died violently, compared to about 3% of the citizens of the earliest states. Tribal violence commonly subsides when a state or empire imposes control over a territory, leading to the various “paxes” (Romana, Islamica, Brittanica and so on) that are familiar to readers of history.

Full post on the decline of brutality.

 

Comments (6)

Trackback URL | Comments RSS Feed

  1. Ken says:

    Very interesting post.

  2. Devon Herrick says:

    . The notion that in “pre-state” days people more often died violent deaths makes sense. I wonder whether or not the violence was precipitated by a gang making its attempt to grasp the power of state or if violence was dealt out randomly by individuals at enemies or used as a way of security resources?

  3. Virginia says:

    How does this fit with the increase in violence in video games and on TV? Maybe we humans have a minimum level of violence that we need to be happy? Or perhaps pseudo violence has to do more with boredom and novelty than actual violence?

  4. Bruce says:

    The state reduces violience? I would say the state IS organized violence.

  5. Brian says:

    I wonder how these archeological investigators/ethnographers would classify violence that happens within the territories of states that do not have full control of the people within their borders. There are obviously few exceptions in the world today (I can only think of the Amazonian tribes).

  6. Linda Gorman says:

    I wonder how they can make claims like this when they have no idea what kind of sample they are looking at. What makes them think that extant skeletons are a random sample?