The Tails of the Cultural Bell Curve
If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:
I.
As you probably all know, a significant number of inborn human traits, when plotted, turn out to form a bell curve, or something very close to that. This distribution is so common that it’s just called the “normal” distribution. Height is often used as the classic example, but it goes well beyond that to things like strength, shoe size and blood pressure. Nor is it limited to physical attributes. IQ and the big five personality traits can also be graphed as a bell curve. This list is by no means exhaustive, and in addition to things which are easy to plot, like height, there are probably things which are less amenable to measurement which are nevertheless normally distributed.
We’ve been talking a fair amount recently about culture, is there some sense in which it might also follow a normal distribution? Perhaps, but if so what would that mean? Or if it doesn’t follow a normal distribution, what sort of distribution does it follow? The reason I’m curious doesn’t have much to do with the peak (or peaks) of the distribution. The tails of a distribution is generally where all the action is, and you can see the beginnings of this idea scattered among my last few posts: What do these cultural “tails” look like for a culture that’s deep into involution? What does it look like if tail behavior is forbidden by super strong taboos? Finally, which behavior is in the middle of a distribution and which is tail behavior, dads or cads? Or are they on opposite ends?
But of course, unlike the genetic traits we mentioned at the beginning, where the graphs move very slowly, culture can evolve quite quickly. What would that movement look like if you graphed it? Should we want it to move in some directions rather than others? If there is a peak how is it determined? Furthermore, inward behavioral preferences would seem to have different graphs and peaks than outward behavioral reality. The broader society would seem to have a reasonable chance of changing outward behavior through incentives, but a more difficult time altering preferences. Though, as a further complication, much of the focus of modernity has been on eliminating the differences between preferences and reality. Is that bad or good?
In order to get our head around these questions and the topic more generally let’s move on to an example:
II.
I’ve already alluded to my review of The Case Against the Sexual Revolution: A New Guide to Sex in the 21st Century, by Louise Perry, so let’s start there. In 1970, when the sexual revolution was just getting started, 80% of people aged 25-50 were married. Now it’s around 50%. I couldn’t find any post-pandemic figures, but the consensus appears to be that the pandemic would have driven it even lower. Clearly in 1970 if you drew a graph depicting the culture of marriage, you would have a large peak which represented monogamous marriage. But that’s a map of outward behavior. It’s unclear what a graph of inward preferences would look like, though one assumes it would be different from their outward behaviors, but that societal expectations and incentives caused people to act in ways that weren’t directly aligned with their individual preferences.
As we can already see, it’s obvious that some graphs are easy to draw because we have the data, while some are far more difficult to plot because all we have are anecdotes. Unfortunately those are the ones we’re most interested in, and one of the goals of this post is to try to infer what those graphs might look like and how they might have changed — perhaps gaining some insight in the process. Such an endeavor is necessarily very speculative, but in spite of this I hope it will be useful.
In any case, while it’s straightforward to chart how many people were married, what’s less straightforward is charting inward preferences. And then there’s the question of whose preferences are we talking about? One of the central claims of Perry’s book is that men and women have different preferences. So when the sexual revolution came along and started eliminating the gap between behaviors and preferences, was it eliminating the gap between male behaviors and male preferences or between female behaviors and female preferences? Sex positive feminists argue that it was the latter, while Perry argues that it was the former, that in fact the 1970’s graph was closer to the preference of the majority of females than the current graph.
There’s obviously a huge debate to be had there, and rather than go back down that rabbit hole, let’s examine something which is hopefully more straightforward: marital contentment. Yes, there are going to be some marriages where one spouse is far more or less content than the other. But my assumption is that it’s hard for one spouse to remain happy if the other is miserable and vice versa (though to a lesser extent), and as such we can meaningfully talk about the contentment of the couple rather than just the individuals. I’m going to make the further assumption that historically marital contentment was normally distributed, that most marriages clustered around the middle, with a few that were much less happy and a few that were much more happy than normal. In other words that previous to the sexual revolution, and indeed for most of history, marital contentment resembled a bell curve.
So far, my assumptions have been unremarkable, and my conclusions ordinary. But as I said at the beginning all the action is at the tails. What effect did the sexual revolution have on the tails of the contentment distribution? Well, the divorce rate skyrocketed, and given that marriages which end in divorce are almost exclusively located at the bottom of the distribution, it seems fair to say that the sexual revolution put more attention on the lower tail of the distribution. But why? I intend to argue that this shift in focus is one of the attributes of modernity, and it’s not limited to marital contentment.
(I actually think it put more attention on both ends of the distribution. That when more focus was placed on bad marriages, that it automatically generated more focus on trying to create really good marriages as well, articles, books, and behavioral changes. But in this post I’m going to restrict things to the lower tail of the distribution.)
This focus on the lower tail was a big change, I would argue that historically most of the attention was given to the middle of the marital contentment distribution. People expected marriages to be average (as indeed most of them were). If you had an average marriage you talked about it, if you had an unhappy marriage you didn’t. Average marriage was the assumed default. Average marriages were generally what got depicted in fiction. Beyond this societal forces pushed people towards the middle. For much of history people didn’t even consider getting a divorce. Men were expected to be dads not cads, regardless of their preference. Taboos existed against caddish behavior and against other behavior at the lower end of the distribution. If your marriage was unhappy you nevertheless pretended that it was normal or average. And as Kurt Vonnegut pointed out, in his under-rated novel Mother Night. “We are what we pretend to be.”
III.
You may think that the Vonnegut quote is a weird tangent, but in reality it represents the crux of the issue. There’s every reason to believe that the graphs of inner preferences and outward behavior do not match, so for hundreds of years we decided to pretend to be something we’re not. In the “land of pretend” 80% of us are married. The divorce rate is 1 per 1,000 (or 0.3 per 1,000 if you go back far enough). And everyone pretends that their marriage is problem free. And as Vonnegut observed, if you pretend hard enough it becomes difficult to distinguish the pretense from the reality.
But at some point in our journey toward enlightenment and rationality we decided that it was wrong to pretend. That our outward behavior should match our inner preferences. And now we’re in the real world where only 50% of us are married. The divorce rate is 2.9 per 1,000 (down from a peak of over 5, but that doesn’t mean much if the marriage rate doesn’t stay constant.) And no one has to pretend anything about their marriage anymore. Yes, people still do pretend, but it’s not expected. As a concrete example of what I mean, prefacing something with “I’m going to stop pretending,” justifies almost any statement.
If pretending is so bad. If we’re well rid of it, why did we pretend in the first place? Here we return to the graph. Most of our pretenses, and our taboos, and our obfuscations were designed to push the distribution towards the middle, towards the average. And what was so special about this average? Presumably it was a point of stability; a point that the culture had arrived at after a great degree of trial and error, a point of safety, but even if this average represented a point of stability in an ocean of chaos. (I don’t think it’s a coincidence that the bell curve also resembles an island.) And something that was beneficial to the society as a whole, it wasn’t something that all members of that society benefited from. All this pretending in order to coalesce around a mean masked a bunch of unhappiness, shame and guilt.
In the past, when survival was more tenuous, the advantages of having a strong cultural center were probably so great that no one cared about the tails of the distribution. But as abundance of all sorts increased, and cohesion no longer spelled the difference between success and ruin, those on the bottom end, who had to pretend the most of all, demanded that they should no longer have to. At the same time as demands were flowing from the edges to the center, charitable impulses were flowing from the center to the edges. As the need to marry and reproduce lessened (say in order to have kids to work the farms) then being unmarried becomes a harmless choice, not a matter for societal scorn. And surely everyone has friends and family who never managed to get married providing first hand knowledge, which further inclines them to be charitable.
The combination of all these factors led to the focus moving from the center to the edges. And this applied to nearly all aspects of culture. Rather than worrying about whether the society was still centered on longstanding cultural practices it became more important to ensure that people weren’t excluded. That if someone would prefer not to get married, or prefer to only get married should everything go perfectly, then they shouldn’t have to.
At first glance this sounds like a good and charitable change. Why would we not want to allow people to express their preferences? Why would we want to make people pretend to be something they’re not? Also, if our abundance allows us to accommodate all sorts of behavior, why shouldn’t we let people behave in whatever fashion they choose?
The whole endeavor promises a better, happier world, but was that promise fulfilled? As time goes on, and we focus more and more on those who were previously shunned and looked down on, what happens? To return to Perry, she admits that in the beginning the sexual revolution and the introduction of the Pill, and all the rest of the changes did go a long way towards making people on the tails happier and more content, but these sorts of situations are never static, there are always trade-offs.
We can see this play out in the cads vs. dads situation. Historically there was always an oversupply of cads, such that we had to pretend that the demand was basically zero to just keep the supply manageable. Since we’ve opened up the floodgates to cads the supply has gotten so great that now women (according to Perry and others) are in the opposite situation. They have to pretend to demand cads (or at least tolerate them) just because the supply of dads is so low. We have clobbered the peak of the distribution and moved a huge amount of behavior to the tails.
As I have already alluded to, this is not just taking place in the marriage market. This focus on the tail is taking place nearly everywhere:
We see it with depression and mental health. This is another area that used to look like a normal distribution with the bulk of people mentally healthy (or pretending to be so) and unmedicated. The situation has flipped, and now rather than hiding illnesses and pretending to be healthy we lean into them. As a result, the number of people in the lower tail of the distribution has skyrocketed. Now to be clear I am not saying that it’s all a matter of societal attitudes. Depression could be increasing for any number of reasons, modernity has brought with it a lot of strangeness. My argument would be that this change in focus from the center to the tail is one of these strange things.
In addition to mental illness we see it with physical illness. We have expanded the number of people who are sick and disabled, and significantly reduced the people who are considered to be healthy. Again I’m not claiming that all the new sick people are pretending. (Though some clearly are, for some definitions of the word “pretend”.) Rather, I’m saying that in the past these people would have pretended to be well. And that the distinction between being well and pretending to be well is not as clear as one would think.
Perhaps the clearest example of the phenomenon is found in the area of gender identification. Up until very recently it’s hard to imagine something more normal, more in the center of the distribution than being cisgender. So much that the word didn’t even exist until 1994 and didn’t appear in any dictionaries until 2015. And yet now a staggering amount of attention is being paid to what was previously an invisible tail in the gender distribution.
If we allow ourselves to get more speculative there are a few other areas which are also worth considering:
Ideology and beliefs are becoming increasingly dominated by the tails. In times past the idea that a presidential election had been stolen would have been out there, but only the outermost fringes would have gotten really worked up about it. Now it’s practically a plank in the Republican platform. On the left I already mentioned gender identification, but things like defunding the police also fall into this category.
Closely related you can also see this in the area of practical politics. And here we actually have a graph. Perhaps you’ve seen it. It always has two peaks, but in 1994 the two parties were so close together that it just about looked like a classic bell curve. It was much the same in 2004, but by 2014 there were clearly two graphs and they were racing to the right and left. I can only imagine what it looks like now. (What’s interesting is that during times of crisis, the graph gets pushed back together, think back to my argument about safety.)
Gambling, and other addictions may initially seem like the opposite of what we’ve been talking about, but I think the dynamic is very similar. The vast majority of people do not develop a gambling addiction, so you might think that up until now we would have allowed everyone to gamble and written off the minority at the tails who ended up harmfully addicted. But recall that the center of the distribution got to be that way because it was an island of safety. A spot of low fragility. To do that societal norms create behaviors which are different from natural preferences. People like to gamble, but at some point we decided it was harmful enough that we placed it under significant restriction. Most people barely even cared, but some people decided that it was an unacceptable restriction on their freedoms. Perhaps in this case the focus is on the other tail of the distribution, those who really wanted to be able to gamble. In any case, there’s been an enormous expansion in the availability of gambling recently, particularly sports gambling.
Having reached the end of this post, I’m not sure how convincing I was. I expect that some people are going to come away thinking. “Well that was a tortured path to saying something that was already entirely obvious.” While others may be thinking the exact opposite, that all I ended up doing was spouting a lot of nonsense. Ideally these reactions represent the two tails of the distribution. That actually most of you form a peak in the middle. The people who think that the idea was both novel and useful.
There’s a lot you could take from this post, but if you were only going to take one thing it should be this. We should all be reading more Vonnegut. So in lieu of a donation check out some Vonnegut from the library. And if you don’t like Vonnegut then you should definitely donate, as penance for your poor taste.