No Wonder So Many People are Depressed
Moderator: bbmods
David, be patient. I said I would explain later (more than once). I did, on the other hand, write something not entirely brief (a "few quick points") that did reply to some of your statements. Are you unhappy with something there, within its limited extent?
What exactly do you regard as "the argument"? Do you mean this thread's discussion scrubbed clean of any reference to Pinker? His book is arguing that the world is better and claiming a reason why, so I'm not sure how it is not relevant, especially since you have both books, have read one, and presumably at some point will read the second. And you did actually ask me about my objections to Pinker's book(s).
If someone wants simply to claim the world is "better" now, in the sense that it now matches more closely his own social, political and moral ideologies and prejudices, then that is okay. When he claims that science and data prove his case, when they do not, then that is not okay. (And it's not okay when it seems he has more facts wrong than right.) This is true in general, but even more so when it is science and reason that the author thinks he is championing as the drivers of the supposedly better world. If that sentiment is what you call "having a beef", then yes, I have a beef.
The data displayed in his plots do not make his case at all. I already gave some explanation of why previously (though that was just the tip of the iceberg). That has nothing to do with "other people's criticisms". Other people's criticisms may alert me to the contents of those plots, but I do not need or want them to judge or interpret those plots for me. You seem determined to draw wild conclusions from the fact that I have quoted other people's reviews. I started this part of the thread by quoting Pinker himself, and I will quote Pinker and others again in the future. So? It doesn't mean I necessarily agree, or necessarily disagree; it means I think it is noteworthy and relevant to the discussion.
What exactly do you regard as "the argument"? Do you mean this thread's discussion scrubbed clean of any reference to Pinker? His book is arguing that the world is better and claiming a reason why, so I'm not sure how it is not relevant, especially since you have both books, have read one, and presumably at some point will read the second. And you did actually ask me about my objections to Pinker's book(s).
If someone wants simply to claim the world is "better" now, in the sense that it now matches more closely his own social, political and moral ideologies and prejudices, then that is okay. When he claims that science and data prove his case, when they do not, then that is not okay. (And it's not okay when it seems he has more facts wrong than right.) This is true in general, but even more so when it is science and reason that the author thinks he is championing as the drivers of the supposedly better world. If that sentiment is what you call "having a beef", then yes, I have a beef.
The data displayed in his plots do not make his case at all. I already gave some explanation of why previously (though that was just the tip of the iceberg). That has nothing to do with "other people's criticisms". Other people's criticisms may alert me to the contents of those plots, but I do not need or want them to judge or interpret those plots for me. You seem determined to draw wild conclusions from the fact that I have quoted other people's reviews. I started this part of the thread by quoting Pinker himself, and I will quote Pinker and others again in the future. So? It doesn't mean I necessarily agree, or necessarily disagree; it means I think it is noteworthy and relevant to the discussion.
Given that the graphs are the argument, let's briefly start discussing them.
Here are Pinker's 2015 updated graphs (for clarity, I note that none of the data are actually "his" --- he just collated other people's work, sometimes interpreting them differently from the people who actually produced them):
https://www.theguardian.com/commentisfr ... m-on-trial
The majority of them don't even pass the eye test as supporting his claim of a trend in the direction of improvement. Look at the first one: "Genocide and other mass killings". The second thing that hits you about these bubbles is how little data there is, only spanning 1990-2013. But the first thing that hits you is that it just looks like randomly sized bubbles: there's a huge bubble in 1994, the sizes go up, they go down, they go up again, they go down again... Does your eye even long for a trend here, let alone truly believe in one? If I told you these were my football betting winnings, you'd probably believe it and conclude that I was neither an extraordinarily good nor an extraordinarily bad sports punter.
TBC.
Here are Pinker's 2015 updated graphs (for clarity, I note that none of the data are actually "his" --- he just collated other people's work, sometimes interpreting them differently from the people who actually produced them):
https://www.theguardian.com/commentisfr ... m-on-trial
The majority of them don't even pass the eye test as supporting his claim of a trend in the direction of improvement. Look at the first one: "Genocide and other mass killings". The second thing that hits you about these bubbles is how little data there is, only spanning 1990-2013. But the first thing that hits you is that it just looks like randomly sized bubbles: there's a huge bubble in 1994, the sizes go up, they go down, they go up again, they go down again... Does your eye even long for a trend here, let alone truly believe in one? If I told you these were my football betting winnings, you'd probably believe it and conclude that I was neither an extraordinarily good nor an extraordinarily bad sports punter.
TBC.
It's (presumably) just what data is available (i.e. the data available are insufficient to support any claimed long-term trends). You really want that genocide data going back a lot further in time. But what there is is simply consistent with random fluctuations over time. The fact that there are some 1s at the end rather than the 2s and 3s at the beginning of that short period is not significant at all, especially when you consider that those are obviously not 1s, 2s, 3s, etc., but have been rounded to the nearest whole numbers (so the 1s could be 1.49s and the 2s could be 1.50s, for example). All you're left with is the observation that the rate was high in the period 1994-99 and (relatively) low on either side. It's amazing that he cannot find data earlier than 1990.
You're reading too much into the homicide data. These also do not support a long-term trend. Presumably you're happy to concede the England and Wales rate is not budging. But the US line is basically a few mountain peaks on a horizontal horizon. Sure, the top of the mountain peak is higher than the end of the mountain range to the right. But it looks like that's also true of the missing end of the mountain range to the left. It's amazing that he cannot find data earlier than 1967.
To claim more from these graphs is to go chasing noise. Don't forget either that the claimed ongoing improvement is over recent centuries.
And a couple of the graphs are particularly problematic for many reasons [to be described later --- have a long look at the executions graph for starters].
TBC
You're reading too much into the homicide data. These also do not support a long-term trend. Presumably you're happy to concede the England and Wales rate is not budging. But the US line is basically a few mountain peaks on a horizontal horizon. Sure, the top of the mountain peak is higher than the end of the mountain range to the right. But it looks like that's also true of the missing end of the mountain range to the left. It's amazing that he cannot find data earlier than 1967.
To claim more from these graphs is to go chasing noise. Don't forget either that the claimed ongoing improvement is over recent centuries.
And a couple of the graphs are particularly problematic for many reasons [to be described later --- have a long look at the executions graph for starters].
TBC
Last edited by K on Thu May 31, 2018 8:10 am, edited 1 time in total.
-
stui magpie
- Posts: 54843
- Joined: Tue May 03, 2005 10:10 am
- Location: In flagrante delicto
- Has liked: 132 times
- Been liked: 168 times
I was going to post about the multiple problematic aspects of Pinker's executions graph I mentioned above...
But then, it occurred to me another reason why Pinker's "genocide and other mass killings" graph (discussed above) is bad, so I thought I'd post about that first...
But then, I realized that, with pictures in hand, there was more I should say about the highly dubious reading of the murders graph, comparing it with the war graph, and even the "rape, sexual assault and violence" graph... so maybe I'll post about these first.
Actually, the graph with the most flaws and importance is probably the "battle deaths" graph. Perhaps that should be last, then...
But then, it occurred to me another reason why Pinker's "genocide and other mass killings" graph (discussed above) is bad, so I thought I'd post about that first...
But then, I realized that, with pictures in hand, there was more I should say about the highly dubious reading of the murders graph, comparing it with the war graph, and even the "rape, sexual assault and violence" graph... so maybe I'll post about these first.
Actually, the graph with the most flaws and importance is probably the "battle deaths" graph. Perhaps that should be last, then...
Last edited by K on Fri Jun 01, 2018 8:49 am, edited 1 time in total.
