I Fear Maladaptive Culture – by Robin Hanson – Beragampengetahuan

I’ve been reading, thinking, and talking, trying to get clearer on what exactly are the culture problems I’m worried about, and how best to describe them. I seek descriptions not only easy for an outsider public to understand, but also for prestigious insider specialists to embrace. 

It seems “maladaptive culture” might be a better name for the problem. So that is the title of this post. Also, I tentatively see four key ways to distinguish more from less problematic cases; the big problems that I fear sit mainly in one corner of that 16-cornered 4D cube of possibilities. Here are the four dimensions:

1. First, culture can work great when tied to particular relevant observable outcomes and inputs. If you want to catch more fish, it can make sense to copy the fishing-related behaviors of the people around you who catch the most fish. You need to be able to tell who gets more fish value (e.g., quantity and size) per effort invested (e.g., time and harms) and you also need to be able to tell which of these folks’ many features and behaviors are plausibly oriented to catching fish. So you can’t do this sort of cultural change until you’ve developed sufficient “cultural gadgets” to see these things. But once you do, things can work great.

The problem I’m worried about arises when it gets harder to apply such conditional strategies, because it gets harder to identify particular relevant observable outcomes and inputs. In this case, you can instead try the relatively unconditional strategy of copying most all behaviors of people (of your species) who are overall successful, without much distinguishing which of their behaviors lead to which success outcomes. As this is sensitive to how exactly one judges “success”, it can go wrong in big ways. But this sort of unconditional copying was probably our main cultural engine when human culture first got going. 

2. Second, culture works great when deviant trials can be small, and there are only weak conformity pressures and penalties for deviating from typical behavior. In this case many different things can be tried, and the best versions can slowly come to dominate practice. The problem I’m worried about arises when trials require huge groups to coordinate to create them, when trials need very long periods to test them, or when the larger culture coordinates to punish and repress deviant behavior.

For example, murder habits are greatly suppressed by norms and laws against murder. In the US today, so are habits of racist verbal expressions. In the face of strong conformity pressures, deviations are unlikely to become common, even if they would be individually beneficial in the absence of these pressures. 

When cultures coordinate via conformity pressures to induce similar behavior in all members, effective cultural evolution of that sort of behavior requires selection of cultures, not just selection within cultures. And effective selection of cultures requires there to be enough sufficiently-insulated-from-each-other cultures with different conformity pressures, and also a sufficient rate at which new cultures are born, and others die, due to famine, pandemic, invasion, and emigration. I worry that our world today lacks sufficient variety and selection of norm-setting macro cultures.

3. Third, the potentially-problematic combination of unconditional copying of culture-shared norms, status markers, and other conformity coordination points, in a context of low cultural variety and selection, would still not actually be that much of a problem if such norms, etc. were stable, and not very sensitive to context. After all, our ancestors have already long experienced great variety and selection pressures, so if the solutions that they found before were robust and likely to continue to work in our new technical and social environments, we could just continue to use them now. The lack of much variety or selection of cultures now would prevent us from improving on those solutions, but not from continuing to use them profitably. 

But in fact, most world cultures have long seemed to experience high and increasing rates of change in their norms and status markers, changes driven mostly from within those cultures. In fact, great energies are devoted to inducing such changes, and to fights over their directions, with great prestige accruing to fight winners. As these changes are only moderately predictable, I find it hard to see them as adaptive. And the world’s long fertility fall suggests to me that they are in fact maladaptive. If so, the world’s main cultures are plausibly going “off the rails.” 

4. Fourth, even fast cultural change of unconditionally copied norms in a context of low cultural variety and weak selection might still be okay if, as some suggest, human cultural change were actually mainly caused by reason. If humans first evolved a powerful ability to reason about which behaviors of whom are useful to copy how, and this reasoning ability induced people to actually copy particular behaviors, then the rapid culture changes we see now would plausibly then mostly be reasonable. They would on average make sense, even if errors resulted in particular cases. The fact that we are unaware of this reasoning, and can’t understand or predict its results, need not imply that they don’t make sense.

However, the consensus of the field of cultural evolution, as best I can read it, agrees with my estimation that cultural evolution came first, and reason second. We quite commonly seem to use reason only to justify decisions we came to in other ways. And only a few small DNA changes were needed to change something like chimps into creatures that could habitually accumulate culture. With that ability they could much more rapidly acquire many “cultural gadgets”, including powerful abilities to reason, including about when to copy who.

But in this scenario it seems implausible that cultural evolution would let each new gadget to totally overrule the choices it would otherwise make. Culture would more plausibly authorize, for each gadget, some limited range of choices where that gadget rules, while leaving other choices unchanged. And the choices most likely to be left to basic cultural evolution are the ones other gadgets find it hardest to judge. And changes to unconditional norms seem especially hard to judge. Thus such choices are not plausibly made using powerful unconscious but reliable reasoning. 

So this remains my worry: our rapid rates of change in unconditional choices of cultural norms and other culture conformity points are not mostly driven by reason, but instead by a crude ancient cultural evolution process that has lately gone “off the rails”, due to a world of insufficient cultural variety and weak cultural selection.