Trending down since the first states appeared (life in traditional bands and tribes was brutally violent), and then trending down even more steeply since the Enlightenment. London in the Middle Ages was about 25 times more violent than today.
Source:
The Better Angels of Our Nature by Stephen Pinker
Looks like an interesting book, but I'm not surprised it has caused so much controversy. It is being attacked from both the left and right, and also by people who claim his nurture over nature claims are ill founded. I'm not quite sure what to make of that.
For individual violence, an overall trend downward would seem the logical thing to expect. For state sponsored violence it would be more surprising and much more contrary to what most historians have claimed (not that you'll find me arguing that a high degree of faith is to be placed in "consensus" historiography). It is a very slippery subject, since you can get about any answer you want by choosing what periods to sum for (hourly, daily, annual, or decadal deaths for example) and what period to look at (from 100,000 BP to now or 1950 to now, for example), what parts of the world to look at, and whether to consider individual violence, group violence, or totals, the devilish details of how to define all those things, and what data to consider reliable. That is, when there IS any real data. And all those choices have their honest uses, although they can be used misleadingly as well.
Regarding group violence in the distant past Pinker's claims have been severely criticized by Prof. Douglas Fry. Fry is an anthropologist whose career has been largely focused on this sort of thing. Here is an abstract of a paper he pub'd in Science:
Lethal Aggression in Mobile Forager Bands and Implications for the Origins of War | Science
At the other end of the scale, if you look at the last cent, trends look really rosy, but you're starting misleadingly with an historically abnormal (but what's "normal" depends on scale also) baseline since the first half of the period includes WWs I & II.
This graph, covering 1400 to 2000:
https://ourworldindata.org/wp-conte...military-civilian-fatalities-from-brecke1.png
from here:
War and Peace before 1945 - Our World In Data
seems to be made by people who are at least making some attempt to think logically about this sort of thing, and does at least make clear that the data is still pretty noisy, even at this larger scale. It SEEMS to show:
- pretty quiet from 1400 to about 1625
- pretty violent from about 1625 until roughly 1960
- returning to the earlier norm for 30 years or so
- a radical drop over the last decade or so covered
This isn't what I'd call a clear long term trend. The graph seems to me a little suspect though, because despite including circles labeled "Armenian Genocide", "Genocide of the Jews", and "Cambodian Genocide", all of which were largely mass murders conducted by states against their own subjects, with the third not even associated with a "war" as normally defined at all AFAIK, it doesn't have circles for either of Stalin's or Mao's massacres which would seem to qualify under any criteria that covered the other 3, and are virtually universally conceded to be of magnitudes at least comparable to (most authorities say larger than) Hitler's and certainly greater than Pol Pot's.
I can't find anything similar for earlier times. I've always read that earlier wars involved a much smaller fraction of the population in actual combat and that a great deal of prisoner taking for ransom occurred, both resulting in lower fatalities per capita but though I've seen the claim made many times, I've never seen anybody put real numbers on it, and I suspect the data may not exist.
This guy may be a kook, but his remarks on this subject seem pretty cogent and he does address the earlier period, which by trad accounts should have been (and he claims were) pretty peaceful as far as state sponsored violence goes:
Steve Pinker’s bogus statistics: A critique of The Better Angels of Our Nature (Part Two) | Uncommon Descent
So, definitely maybe I dunno. If anybody will loan me a time machine, I'll be glad to take a look and let you know.
Thank you for resurrecting this one, an interesting read. The greatest necro I have seen recently was on a gaming forum. It was a fifteen year old thread, revived earlier this year by somebody who had to do some impressively deep forum diving to dig it up. I'm thinking of it like a form of time travel. We have all these snap shots of moments in time, scattered across the net. Some of them fade away, as sites are closed etc, but many are still there. Think about it. In fifty years time, we will be looking at those old threads like we look at grainy tapes of the Ed Sullivan show. That gives us a decision to make. Do we preserve those threads as they were? or do we keep them alive, for our future selves to revisit and add comment to? Would that sully the historical significance? or would it be our new form of time travelling virtual tourism, skipping around the virtual landscape digging for golden nuggets of history? And what form will that take? Virtual Reality representations of the world wide web, with avatars running around and digging into those old repositories of data? Finding out what your great grandfather posted about the prime minister of Japan on November 3rd, 1999?
