Beware Moral Fashion

Many things change. Outside of humans: cosmology, geology, climate, biology. Among humans: cities, machines, housing, language, clothes, morals, conversation topics. Dead things change due to interactions, decay, cycles, collisions. Life also changes via evolution, where many pretty random changes appear, from which selection chooses the best, leading to progress over time. Human culture also evolves via cultural evolution. 

Among humans, we also see changes we call fashion. In the short term, fashion changes look much like evolution changes, with some initially-random variations growing in popularity, as if they had superior lasting value. But then these changes go away, replaced by new fashions, without net gains resulting. Often such changes go through recognizable cycles, such as hair being short then long then short, etc. These changes don’t seem explainable via simple randomness, lasting innovation, or changes matched to relevant context (like climate). (Typically such fashions wander in a limited space, causing no net long term harm, but without such limits more harm can result.)

The usual explanation for fashion, which makes sense to me, is that humans assign higher status not only to people who help initiate lasting innovations, but also to those who help initiate more temporary fashion changes. After all, those who start both kinds of changes, or join them earlier than others, show impressive cleverness, perception, info sources, and social influence. This creates incentives to try to make and join new fashions, and to drop them when enough low status folks embrace them. To serve this purpose, new fashions only have to be not clearly worse than prior ones. 

Relative to efforts to induce lasting innovations, efforts to induce fashion changes seem more of a social waste. But compared to impressing others via war, they seem less of a waste. Also, most ancient societies had little of innovation or fashion, and Europe started to have lots of fashion ~1400, well before it started to have lots of innovation. So a love of fashion might have led to a love of exploration and innovation; we may have fashion to thank for modern innovation rates.

We tend to see more fashion changes in things that are more socially visible, that are more a matter of individual than collective choice, and that are capable of changing faster and cheaper. 

I also think we also see more fashion where we less agree on evaluation criteria. For example, while many costs and revenues contribute to firm profitability, we agree to measure profit as total revenue minus total cost, using exactly unit weights on each different kind of cost and revenue. We could imagine a world where such weights varied from unity over time according to fashions, but that isn’t our world. 

Similarly, we consistently measure the height of a person or building as the min distance, when they are stable, from their highest peak to the ground. There are thousands of other possible ways we could define “height”, and we could imagine a world where they often change how they define height according to fashion. But that is not our world. 

We also don’t see much fashion in engineering designs for behind-the-scenes structures like nails, screws, tapes, and trusses. Plausibly because we mostly agree on the engineering criteria by which to evaluate such designs. Yet we see greatly changing fashions in many areas where diverse factors contribute to value, and we disagree on relative factor contributions. Such as in clothing, furniture, music, literature, and politics. 

The explanation here, plausibly, is that we are reluctant to admit to supporting fashions, as opposed to innovations. We instead want to presume that any change we support has a decent chance of becoming a lasting innovation; it seems actually better. And we can only reasonably presume this while pushing fashions in areas where relative priorities are in substantial doubt. For example, we can find many reasons why the new striped socks fashion seems more aesthetic, at least in our current context, and no one can prove otherwise. This seems harder re screw design. 

This view of fashion helps us to see just how much trouble we are in from cultural drift, and also how we might perhaps fix it. While we have little to fear from cultural evolution tied to measurable outcomes (e.g., copying fishers who catch more fish), we have more to fear from cultural evolution not tied to such outcomes, and with features shared within cultures. Such as the evolution of morals, norms, and status markers. 

Not only do we now have historically low levels of variety and selection pressures re such shared cultural features, but also historically high rates of change. And these changes just are not plausibly mostly adaptive. Yes, some value changes are plausibly due to ancient evolved conditional value strategies, by which our values change with context like density, relatedness, sex ratios, mortality rates, pathogen prevalence, and resource availability, patchiness, and predictability. But today, most seems due to moral/norm fashion change. 

The only discussion I’ve found of moral fashion is by Paul Graham: 

Did we actually dress like that? We did. And we had no idea how silly we looked. It's the nature of fashion to be invisible. … What scares me is that there are moral fashions too. They're just as arbitrary, and just as invisible to most people. But they're much more dangerous. Fashion is mistaken for good design; moral fashion is mistaken for good. … What would someone coming back to visit us in a time machine have to be careful not to say? That's what I want to study here. But I … want to find general recipes for discovering what you can't say, in any era. (More)

Alas, Graham treats fashion change as clothing-fashion-like white noise, randomly sampling from a limited space, and at each time distorting a limited fraction of morals. In which case, his strategy should work: find the few most distorted morals of your time, and substitute for them an average on those topics from other times.

But moral fashion actually seems to be more of a (not necessarily zero mean) random walk in a not so limited space. Centuries ago, when the world had 100Ks of small poor cultures amid frequent famines, pandemics, and wars, cultural selection was strong, assuring that morals were adaptive. But then we cut variety, weakened selection, and elevated the status of moral fashion changing “activists”. After which morals have rapidly wandered away from their initial adaptive places, without obvious limits, and plausibly mostly due to fashion processes that don’t on average promote adaptiveness.

In contrast, clothing fashion changes may be wasteful, but clothes are constrained to wander in a space limited by our limited willingness to sacrifice comfort, convenience, or cost for clothing fashion. More threatening would be cultural changes that made us much more willing to make such sacrifices, and moral fashion is exactly the sort of thing which might do that. 

A possible fix to this problem is suggested by the observation above that fashions arise more where it is harder to judge relative contributions of various evaluation factors. So if we could instead agree on a way to precisely and canonically evaluate cultural packages of morals, norms, values, and status markers, then it should become harder for moral fashions to change them. But yes, I get that this is a huge ask. 

I’ve previously suggested that a futarchy-based government tied to a constant long term goal like when we get to the stars might work here. Futarchy speculators would see civ collapse as an obstacle to this goal, and approve policies to limit whatever contributes to culture drift, including moral fashion. Though it is an open question if people could feel committed enough to that goal to suffer the pains needed to reach it.