Can you define the terms that you’re using? The PDF you linked to, for example, takes a rather broad view of “violence” as encompassing, say, corporal punishment of children by parents. Or when you say that the trend “has always been downwards”, how large a perturbation are you willing to ignore?
Under a sufficiently wide definition of violence, every person engages in it (some more frequently than others).
Homicide and assault are what I think most people are referring to. The harming of the physical body through force. Additionally these numbers are referenced per capita and not as raw numbers. If we look at raw numbers, modern times certainly have more. If we look at per capita, the trend is downward.
As far as size of perturbation? That’s difficult to really answer. My rough opinion would be to ignore any perturbations that span less than 100 years. So while WW1 and WW2 might cause a spike to the graph, post-WW is still lower than pre-WW and so the trend is still continuing and valid. Could also try reducing it down to 50 year averages as well to help smooth out the variation.
Can you define the terms that you’re using? The PDF you linked to, for example, takes a rather broad view of “violence” as encompassing, say, corporal punishment of children by parents. Or when you say that the trend “has always been downwards”, how large a perturbation are you willing to ignore?
Under a sufficiently wide definition of violence, every person engages in it (some more frequently than others).
Homicide and assault are what I think most people are referring to. The harming of the physical body through force. Additionally these numbers are referenced per capita and not as raw numbers. If we look at raw numbers, modern times certainly have more. If we look at per capita, the trend is downward.
As far as size of perturbation? That’s difficult to really answer. My rough opinion would be to ignore any perturbations that span less than 100 years. So while WW1 and WW2 might cause a spike to the graph, post-WW is still lower than pre-WW and so the trend is still continuing and valid. Could also try reducing it down to 50 year averages as well to help smooth out the variation.