Featured image is Saint Jerome in His Study, by Marinus van Reymerswaele.
It is easy to think that a paper or a conversation, or even an academic convention, happen in relative isolation. After all, if a paper gets hundreds of citations, it is considered wildly successful. Meanwhile, the biggest websites measure their link counts in the tens or hundreds of millions. An academic field provides a relatively small audience even for its rockstars, and specialized sub-fields are yet smaller.
But the consensus of these fields has implications for all of us. We have nuclear power but also nuclear weapons, vaccines and antibiotics but also chemical and biological weapons, sophisticated tools for financing businesses and college educations and home buying…but also the volatile and bewilderingly complex financial system. Academic conversation played a role, to varying degrees, in each of these cases.
In the human sciences, the academic conversation concerns the very question of what it means to be human. And the conclusions which are drawn there, again, do not occur in a vacuum. The approach that a legislator or an agency rule-maker takes depends to no small degree on just what sort of animal they—or their social scientist advisers—believe human beings to be.
The architects of the American administrative state believed that individual humans were nothing more than organs or cells which added up to a body politic, with the state as the head. They believed that individual rights were a relic of historic superstitions, and in any case were articulated before the Industrial Revolution, which changed everything. All that mattered was the health of the body politic—and so undesirable individuals became mere polluting elements, threatening the short or long term health of the body. These illiberal reformers thus set to work drafting immigration restrictions, as well as forcible sterilization laws, to name but two of the legacy of their honored stewardship of the American project.
I belabor these examples in order to emphasize that ideas matter, and what ideas are cherished in the human sciences matters a great deal.
In both neoclassical economics, and the favored corrective for it—behavioral economics—the relationship of practitioner to subject matter is not sufficiently appreciated. And so pernicious models of human nature are allowed to thrive.
Knowledge and persuasion, as experienced in human life, have no place in standard economic models. It is all, as Dan Klein puts it, flattened information—a thing to either have or not have. But does that describe the experience of learning economics? Perhaps textbooks provide this illusion to a certain extent, but any decent textbook will acknowledge areas of controversy. And in any case, a textbook is a poor starting point for an academic paper—as any economist knows!
An economist is typically exposed to multiple models making conflicting claims about the same phenomena. Anyone worth their salt will probe deeper still, and see that even the array that are currently presented represent a consensus; the consensus that these are the differences worth discussing. This consensus is itself a contested matter. Figuring out where one stands in the scheme of this contestation is not a matter of having or not having information, but reading a lot, thinking, talking with peers and mentors, and then making judgment calls. This is not like accumulating objects into a pile. The gap between the process of learning and arriving at a particular understanding requires a big leap in the dark to reach the other side. Such a leap requires courage—the courage of our convictions.
If conventional economics provides us with a mechanistic information processor, behavioral economics provides us with an unreflective, biased, instinct-driven animal. But again, behavioral economists and psychologists appear to fail to account for themselves. If we can demonstrate that even statisticians fail at basic statistical inferences when put on the spot, one has to ask how and in what context they do succeed in making statistical inferences. Would any of us claim that a statistician working on an important paper, or a data scientist at Google, is incapable of making highly sophisticated, never mind basic, statistical inferences?
And how are we to account for these scientists who carefully design experimental conditions in which biases can be isolated and observed?
What is presupposed in the very pursuit of insights about human beings is that human beings are capable of finding such insights. The so-called “credibility revolution” in economics presupposes that statistics, probability, and mathematics are valid fields of knowledge, which itself sets boundaries for what we can assume about a species capable of understanding such things. There are similar limits set on what we can assume about a species capable of designing experiments and treating observations as generalized data. Not to mention a species capable of theorizing, period!
It is one thing to say, as many do, that models that are wrong on their face are still useful for arriving at correct judgments. Milton Friedman famously said as much about economic models. But in practice social scientists seem to treat their models as literally true, especially when making policy recommendations. This will not do. It is of no help if one will concede the false-but-useful nature of these models when in a philosophical mood, if you then go forth and design policy based on models that are decidedly inhuman.
I used to be of two minds about specialization in the human sciences, given the value of specialization in general. I am increasingly of one mind. The current specialization of professional social science is practically designed to arrive at dehumanizing conclusions.
What is the point of students of human nature who are ignorant of human history, philosophy, poetry, and law?
Previous Posts in This Thread:
Sweet Talk 
I’m not totally clear on what you’re criticizing here other than a vague overconfidence social scientists have in the ontological and epistemic status of their models. Is this essay against a lack of humility? Can you give examples that illustrate your point?
Well, I think the essay probably _exhibits_ a lack of humility on the part of its author 😀
I would put it like this: against models of what it means to be human that are ruled out by the very process of constructing the model.
An example would be Russ Roberts, who has taken the behavioral economics critique of mainstream economics to such an extreme that he appears to have ruled out the ability for human beings to successfully theorize about human matters at all—which must not be true, because he’s drawing this conclusion on the basis of psychology, a thing that wouldn’t be possible in that framework!
I’m probably being unfair to Russ there, but that’s something I’ve seen a lot of. Mainstream and behavioral econ used as a dichotomy when, even combined, they do not add up to a full human being.
Stephen Landsburg, for instance, exhibits the dehumanizing tendencies from mainstream econ: http://www.thebigquestions.com/2016/02/01/how-to-be-reasonable/
So when you say “human” and “humanizing” you mean “highly imperfect?” I’m reading your post again. Would it be fair to say that we now know that no model makes predictions with 100% certainty and that the remainder must be filled in with human faith and courage? If that is a fair read, then, are you claiming that social scientists themselves have not taken that point?
John Levi Martin discusses some of this kind of chauvinism, and how social scientists often unselfconsciously construct the social situation respondents are responding to, without accounting for the “endogeneity” of their own influence. It is in the first few chapters of Explanations of Social Action.
I mean largely that broader, richer, and for some aspects ineffable aspects of what it means to be human. The stuff that poets and philosophers grapple with just as legitimately as social scientists.
I’ll have to check out John Levi Martin’s stuff, thanks for the pointer!
To give more concrete examples: the difficult to understand nature of abstract concepts, which we rely on without giving much though. A non-utilitarian theory of love. The relationship between theory and practice. The nature of language. Etc. The answers to these things (which are rarely conclusive) point to something much bigger than utility or behaviorist type models are capable of explaining.
This is perhaps your most Ryan-ish post yet. Consequently, I love it! Not only did you quote Ayn Rand, but you argued for the supremacy of results over theory. Three cheers.
Glad you liked it!
Gadamer argues that all understanding just is successful application, even when the goal is supposedly theoretical (as in a law professor assessing the meaning of a law vs a judge ruling on a case where the law applies). I wonder if perhaps that is why there’s a confluence between what I’m driving at and your perspective concerning results?