An Engineering-Grade Science of Social Technology

I think one of the highest leverage things we can work on is research towards an engineering-grade science of social technology.

Let me clarify what I mean by that:

First of all, social technology is an idea that has taken off in the discourse as people strive to understand society through fresh conceptual lenses, examining the functional purpose of culture. The idea is fairly simple and self-explanatory:

What if we imagined cultural norms and beliefs, institutions, not as inviolable Tradition or moral absolutes, meaningless personal aesthetics, or problematic systems of oppression, but as technology: sometimes-consciously designed, optional but useful, ways of organizing our means. This view isn’t unheard-of, but it’s notably absent especially in ideological discussion.

Social Technology is the technologies built out of human behavior. Institutions. When we get married, have friends, decide who was wrong in some dispute, walk on the right side of the sidewalk, start a company, or vote on a motion, we are using social technology, often embedded in larger institutions, to solve real problems that would be otherwise harder to solve.

The idea of social technology has profound implications for how we see society and institutions.

Next, the idea of sceince. As I see it, something is scientific when it has a sharp enough core theory, a paradigm, to enable it to be integrated into the rest of our scientific-mathematical knowledgebase.

A science of social technology would have a sharp enough theory of anthropology and social organization to make it possible to import and crosspolinate knowledge from network theory, complexity theory, thermodynamics, evolutionary biology, economics, and so on. Can you do sharply predictive derivations of theorems in the field? If not, your knowledge is not scientific.

Biology is an edge case, because it’s a lot of cataloguing rather than simple models. But it does have its simple models and overarching paradigms integrated into the rest of science, and people do math in it.

People often object that a science of humans is impossible, because the existence of free will invalidates the mechanistic assumptions of science, or because humans are inherently too squishy and vague to be predicted. This is of course not true: mechanistic assumptions are not necessary to science, only to those fields where reality is in fact mechanistic. And humans are actually pretty reliable: what’s the reliability, for example, of people staying on the sidewalk, and not wandering into the street. Better in some neighborhoods than others, to be sure, but everywhere we see people following rules orders of magnitude more reliably than chance. Vagueness and squishyness is always a property of our concepts, not reality itself. We don’t have a good paradigm for human sciences yet, but given better concepts, these problems would resolve.

A science of social technology would enable us to make sharply predictive statements about social technology and institutions: This system does these things, and has that failure mode. That phenomenon is central to this social structure.

An engineering-grade science is one where the paradigm is mature enough to actually enable engineering, which is the conscious construction and analysis of technology using scientific knowledge of the phenomena in question to plan and design.

So an engineering-grade science of social technology would be a teachable formal discipline of building and implementing social technology and social fabric. Note the implied assumption that institutions are built. They don’t come from nowhere.

My bet is that such a discipline, if better developed, larger, and more deliberately applied than what we currently have, would lay a very powerful groundwork for a positive restoration of the many rough spots in our institions.

It’s easy to misinterpret this kind of approach for the cavalier top-down social engineering fetishized by the 20th century modernist regimes. As such, it is important to note that societies are complex organic systems, as much like an ecosystem as like an industrial plant. Total central direction is not desirable or effective.

That said, neither should we fall into the anti-rationalist social mysticism and primitivism which rejects any central organization or conscious design.

The social technologist wields power, at the largest scales the power of the state, but must do so organically, embedded in the system he is shaping, more like an organic permaculture gardener than like either an industrial farmer or primitive gatherer.

Getting to an engineering-grade science of social technology from here would also force us to fix a bunch of our existing social infrastructure and paradigms, and that’s half the fun. If we can get this right, the implications for human society are vast and immensely positive.