No Wonder So Many People are Depressed

[Mugwump]

[K]

It would be nice to see a list of the supposed gains and losses side-by-side, fairly considered, and fairly acknowledged.
...
If the intention is to compare historical periods, someone needs to run these things alongside classical proxies such as the murder rate for better or worse.

We can consider anything. We don't have space or time restrictions here. But each should be looked at in a way that isn't entirely superficial. I mean, the important first step is to establish what actually may be true and how confident of that we are.

I also don't think homicide is just a "proxy"; surely, it's important in its own right. Homicide, genocide, war... The importance of these things is that they relate to extremes of both human suffering and human evil. Technology has made those of us fortunate enough to benefit from it much more comfortable than in the past. That's really not the heart of the matter.

Then there's the extreme end of largely irrelevant supposed gains... [I'm now departing from anything ptid was talking about.]

Retirement (Pinker's EN, Fig. 17-2)? International tourism (EN, Fig. 17-8 )? We'll have to let David report what these figures look like and are supposed to signify. But they are really at best first-world luxuries, at worst just trivia. And death is bad, but since the author cares only about aggregate numbers, why talk about deaths by lightning strikes (EN, Fig. 12-9)?? Surely, that's always been a minuscule fraction of total deaths? Not to mention his focusing on (presumed) reductions in fatality rates in modes of death that simply did not exist a century ago.

And then there are proxies (to borrow ptid's term) Pinker uses that just reflect his own set of political ideologies and prejudices (e.g. vegetarianism, which certainly does not have near-universal support as being "better"). But let me leave that topic for another post.

[pietillidie]

^We can look up the murder rate in a couple of minutes, but that's plainly not why we're here. Instead, we're variously making some grander claim. And given there are no units of 'civilisational decline', or whatever is being discussed, it's a proxy variable.

(You might just be interested in the murder rate, to be fair, but that's not the only discussion).

[K]

Homicide is not the only thing I think is important, but I think it is important in and of itself, just as other things are also important in and of themselves. I don't want to be murdered and I don't want innocent people to be murdered. I was, at least initially, wanting to discuss things one at a time (though now I think interleaving subtopics may not be so bad and may even be necessary).

We probably shouldn't dwell too much on semantics, but "proxy" sounds too much to me like (say, in the context of medicine) "marker", as in "marker for cancer". Homicide, to me, is like the cancer, not like the marker. You can say "homicide is a proxy for civilisational decline", but that is like saying "cancer is a proxy for health decline" (which I personally would find a rather odd thing to say): "xxx decline" is an overly general term.

There actually are data for things like child abuse, domestic abuse, health coverage, societal attitudes, etc. (from WHO, Pew Research Center, etc.), which posters can look up and attach here for discussion, if that's what they wish to discuss. (But it most definitely does not take just "a couple of minutes" to look up reliable homicide rates we want: we still don't have all that we want, and the reliability of what we have has to be determined.) I would actually welcome people engaging more in such actions. Before making grand claims about why something is so, one should make sure it actually is so first, which does require some sort of empirical evidence.

[pietillidie]

Homicide is not the only thing I think is important, but I think it is important in and of itself, just as other things are also important in and of themselves. I don't want to be murdered and I don't want innocent people to be murdered. I was, at least initially, wanting to discuss things one at a time (though now I think interleaving subtopics may not be so bad and may even be necessary).

We probably shouldn't dwell too much on semantics, but "proxy" sounds too much to me like (say, in the context of medicine) "marker", as in "marker for cancer". Homicide, to me, is like the cancer, not like the marker. You can say "homicide is a proxy for civilisational decline", but that is like saying "cancer is a proxy for health decline" (which I personally would find a rather odd thing to say): "xxx decline" is an overly general term.

The problem is that people are making grand claims, and your untheorised focus on the murder rate isn't addressing them.

'Proxy' is an abbreviation for 'proxy variable', a technical term which forces us to recognise that we are theorising. The validity of the proxy variable in any complex claim is always subject to scrutiny. A great example of this is Mugwump's argument that medical technology might be causing the murder rate to be understated, with GBH possibly being a more valid proxy. So, not even homicide rates matter at face value. (So much so that a worsening of the homicide rate might even be considered a good outcome in cases where even grimmer heights were likely).

The point I was trying to make is that the murder rate is the simplest thing being discussed in this thread. I entirely agree that each data point needs examining; we're on the same page on this. However, the validity of the use of the murder rate as a proxy variable in any number of wild theories needs examining at least before 2050. One way of contextualising it would be to run a 'units of suffering' thought exercise.

[Mugwump]

Homicide is not the only thing I think is important, but I think it is important in and of itself, just as other things are also important in and of themselves. I don't want to be murdered and I don't want innocent people to be murdered. I was, at least initially, wanting to discuss things one at a time (though now I think interleaving subtopics may not be so bad and may even be necessary).

We probably shouldn't dwell too much on semantics, but "proxy" sounds too much to me like (say, in the context of medicine) "marker", as in "marker for cancer". Homicide, to me, is like the cancer, not like the marker. You can say "homicide is a proxy for civilisational decline", but that is like saying "cancer is a proxy for health decline" (which I personally would find a rather odd thing to say): "xxx decline" is an overly general term.

The problem is that people are making grand claims, and your untheorised focus on the murder rate isn't addressing them.

'Proxy' is an abbreviation for 'proxy variable', a technical term which forces us to recognise that we are theorising. The validity of the proxy variable in any complex claim is always subject to scrutiny. A great example of this is Mugwump's argument that medical technology might be causing the murder rate to be understated, with GBH possibly being a more valid proxy. So, not even homicide rates matter at face value. (So much so that a worsening of the homicide rate might even be considered a good outcome in cases where even grimmer heights were likely).

The point I was trying to make is that the murder rate is the simplest thing being discussed in this thread. I entirely agree that each data point needs examining; we're on the same page on this. However, the validity of the use of the murder rate as a proxy variable in any number of wild theories needs examining at least before 2050. One way of contextualising it would be to run a 'units of suffering' thought exercise.

You seem to misunderstand. It is not about "units of suffering", though one day it may be. It is about the fabric which holds society together and allows us to continue to develop, rather than regressing - as many societies have throughout history - into civil disunion, terror and chaos. You seem desperate to defend the status quo (underneath all supposed radicals one finds a conservative for a new order) and so you suggest that technological progress can be set against this, but these things are different categories, as you surely know.

Am I making "grand claims" ? I am showing that the ultimate disrespect for our fellow citizens, murder and GBH, seems to be far more prevalent than it used to be. If this does not strike you as a powerful sociological statistic, and a probable leading indicator of great social dysfunction now and in time to come, perhaps you are immune to certain tendencies in the evidence.

Old-fashioned conservatives knew that they too were sinners trying to make the best of a wicked world. The libertarian pluralists, much more conservative in their way, feel that they are harbingers of a better humanity and the world they have brought into being must - simply must - be better, regardless of the evidence. Look at all our new technology ! It was the same mentality that doomed the Soviet debacle.

[Mugwump]

[K]

^ Try the Option "Add an Attachment" (then "Choose File", etc.). Image files (png, jpg, gif, etc.) have a 50kB limit. Document files (pdf, ps, xls, etc.) can be 3MB.

[K]

...
The problem is that people are making grand claims, and your untheorised focus on the murder rate isn't addressing them.

.... A great example of this is Mugwump's argument that medical technology might be causing the murder rate to be understated, with GBH possibly being a more valid proxy. So, not even homicide rates matter at face value. (So much so that a worsening of the homicide rate might even be considered a good outcome in cases where even grimmer heights were likely).
...

Can you elaborate on what you mean by "untheorised focus"? I want to see and evaluate data before making claims. And I guess one can try either a "top-down" or "bottom-up" approach, or both simultaneously (especially when there are multiple people to do the trying). In the bottom-up approach, we first (say) examine the hypothesis that medical advances have damped the homicide rates and understated the degree of societal violence, and we look to quantify that. To the extent that we are successful, we can then address more with that same variable.

[Mugwump]

^ Try the Option "Add an Attachment" (then "Choose File", etc.). Image files (png, jpg, gif, etc.) have a 50kB limit. Document files (pdf, ps, xls, etc.) can be 3MB.

Ok, thanks, will try that.

[Mugwump]

The chart I mentioned - thanks K for the instructions.

The 1950s rate is pretty much a continuation of the first half of the 20th century. The gradient shift really happens in the mid to late 1960s. Oddly, the 1970s show a plateau for this type of crime. This has never been satisfactorily explained, as the wider violent crime rate (which includes minor injuries and no injury) continued to escalate in a more or less exponential fashion after its 1960s lift-off.

One can argue about the fine details, but the order of magnitude is staggering. It is true that the rate has declined in the last few years, though it seems to have leveled off far above the early 20th century. The reasons for the recent falls have been widely theorized (including above) but remain hard to prove.

[pietillidie]

...
The problem is that people are making grand claims, and your untheorised focus on the murder rate isn't addressing them.

.... A great example of this is Mugwump's argument that medical technology might be causing the murder rate to be understated, with GBH possibly being a more valid proxy. So, not even homicide rates matter at face value. (So much so that a worsening of the homicide rate might even be considered a good outcome in cases where even grimmer heights were likely).
...

Can you elaborate on what you mean by "untheorised focus"? I want to see and evaluate data before making claims. And I guess one can try either a "top-down" or "bottom-up" approach, or both simultaneously (especially when there are multiple people to do the trying). In the bottom-up approach, we first (say) examine the hypothesis that medical advances have damped the homicide rates and understated the degree of societal violence, and we look to quantify that. To the extent that we are successful, we can then address more with that same variable.

What I mean is this:

I also don't think homicide is just a "proxy"; surely, it's important in its own right. Homicide, genocide, war... The importance of these things is that they relate to extremes of both human suffering and human evil. Technology has made those of us fortunate enough to benefit from it much more comfortable than in the past. That's really not the heart of the matter.

Without proper argument, framed as an hypothesis, this is homily. Until I pointed it out, it was locked inside your head; so much so you hadn't considered that you were using proxy variables, and why that matters.

It would be nice if you made your tacit assumptions and theories explicit so we can assess if your proxy variables are valid. Currently, this is a mystery game where only at the end do we discover if we're examining a valid proxy variable.

Merely saying, "The importance of these things is that they relate to extremes of both human suffering and human evil," is insufficient.

Would North Korea be a marvellous place to live if we had reliable data showing its murder rate was zero? To offer more realistic examples, why are Costa Ricans so happy despite their homicide rate? Why are South Koreans so unhappy despite theirs?

We all know the answer is because other things matter, depending on what you're arguing. And if other things matter, what are they, and when will they get a look in?

Again, fair enough; your narrow interest began with Pinker's data selection. But that's not your only interest, clearly. So, it would be nice for you to address your own working thesis so we can assess whether these data sets are worth looking at. For all we know, you might have invited us to watch you mow the wrong lawn.

[pietillidie]

The reasons for the recent falls have been widely theorized (including above) but remain hard to prove.

But they're no harder to identify than the reasons for the increase.

On the rise, I would be looking at demography, geography and economics first. You have to rule out the major social forces before moving onto more exotic theories.

[Mugwump]

The reasons for the recent falls have been widely theorized (including above) but remain hard to prove.

But they're no harder to identify than the reasons for the increase.

On the rise, I would be looking at demography, geography and economics first. You have to rule out the major social forces before moving onto more exotic theories.

"Exotic" theories being the ones you prefer not to believe ?

The hypotheses (for the decline) were discussed above. There are no doubt several. Demographics seems the most plausible (the pros and cons of that explanation were addressed above), economics less so given the discrepancies between nations.

[K]

...
Can you elaborate on what you mean by "untheorised focus"? ...

What I mean is this:

I also don't think homicide is just a "proxy"; surely, it's important in its own right. Homicide, genocide, war... The importance of these things is that they relate to extremes of both human suffering and human evil. Technology has made those of us fortunate enough to benefit from it much more comfortable than in the past. That's really not the heart of the matter.

Without proper argument, framed as an hypothesis, this is homily. Until I pointed it out, it was locked inside your head; so much so you hadn't considered that you were using proxy variables, and why that matters.

It would be nice if you made your tacit assumptions and theories explicit so we can assess if your proxy variables are valid. Currently, this is a mystery game where only at the end do we discover if we're examining a valid proxy variable.
...

Homicide seems to be a proxy variable for you, but, as I alluded to above, it is not a proxy variable for me. Apparently, you find homicide of no interest or consequence in its own right. I think differently, as do criminologists (assuming they find their jobs interesting and consequential). Something can be interesting in itself and at the same time potentially inform us about bigger, or more general, or simply different matters. If it does inform us about broader concerns, then one can regard it as doubly interesting, not simply a proxy variable.*

Football could be claimed to teach us about life. Maybe it does. That does not mean football is merely a proxy variable for life. If one is invested in the on-field fortunes of the Collingwood Football Club, that does not mean that one has unwittingly failed to consider that at the end we'll discover that the CFC is just a proxy variable.

No one is forcing you to be interested in any aspect of criminology or warfare for its own sake. But simply dismissing the topics that you don't care about --- or that suggest conclusions you don't believe --- as proxy variables or homily is less helpful than actually providing concrete data and explicit arguments for the topics you do care about.


* For example, by exploring the effect of advances in trauma care on homicide rates, we can find out more about the relation of homicide to violence more generally. We can then hope to find out about the relation of violence to quality of life. You do not need to frame hypotheses in advance to embark on this plan.