The Atlantic has a thought provoking article by Eric Liu.
The heart of the argument is that, while we have been bickering a great deal during this election cycle, we have not been “arguing” much, at least not in the classical sense of arguing, of presenting the premises for respective sides and explaining how and why each side has arrived at their respective conclusions. Not so much to reach a “conclusion” as to who’s right and who is not, but to understand. As per the following passage:
Imagine forming citizen “talking circles” all across the country, where people of differing world views agree simply to listen to one another. The point would not be persuasion or conversion. The point would be presence. And the method would not be to discuss ideology explicitly. It would be to address a simple universal question—something like “Who influenced you, and how do you pass it on?”
I think this is a noble argument, something analogous to what I had argued about before as well. It is also something whose prospects I am deeply skeptical of.
The premise behind this line of thinking is ultimately that, because the uncertainty about the universe is so great, being “right” or “wrong” does not really matter. Everyone has a sliver of truth, and we are all better knowing all these slivers of the truth. Wonkism, a cousin of “scientism” the unstated ideology that has come to dominate all discussions these days, explicitly rejects that premise. The idea behind scietnism is that “science,” arrived at from a certain specific set of assumptions and premises is necessarily right and the conclusions to the contrary necessarily wrong. The acceptance of these assumptions and premises indicate someone as a member of the “righteous” (or the “brights” or “rationalia” or whatever) while the refusal signals that they belong to the tribe of “uncivilized savages.” Alternate interpretation of the assumptions and premises is not allowed. Furthermore, challenge to the assumptions on data is not usually permitted: the accepted assumptions and premises apply to the interpretation of the data as well. In order to challenge the accepted assumptions and premises, the data analysis must be presented in a manner that does not overtly contravene accepted assumptions and premises and still show a degree of contradiction. Not an easy chore for almost anyone.
One might say that , of course, in a sense, this is how science usually works. Contrary to what people think, Galileo was a crank, whose ideas may have been, in retrospect, more right than his contemporaries’ but who had no means of backing up his claims. What he ran up against was not an obscurantism determined to thwart science and “progress,” but demanded sufficiently credible evidence to justify tossing out hitherto widely accepted assumptions. This made them obstructive and reactionary, but not necessarily because they wanted to be. This becomes more dangerous when “science” is mixed with ideology: Stephen Jay Gould had made this argument about scientific racism in The Mismeasure of Man. Whereas Galileo’s adversaries respected his genius and made reasonable, even if ultimately wrong, suggestions as to how his evidence and accepted conventional wisdom could be reconciled and showed themselves willing to be convinced for the right evidence, advocates of scientific racism were committed to faith in their own righteous as well as maintaining their view as the dominant view that actively shaped policymaking. What was already difficult chore of breaking through the inherently “conservative” bias of “science” was redoubled by the stonewalling of those who had a vested in interest in rejecting contrary views.
Compared to, say, modern physics, social sciences suffer from another problem: fundamental imprecision of its phenomena. Einstein’s theory about gravitation made a very precise prediction, which could be demonstrated by a single experiment involving solar eclipse in 1919. How many propositions in social sciences are similarly elegant and simple? Humans behave in too erratic a fashion to allow for any prediction to be obeyed with sufficient precision. Theories involve too many moving parts to generate a clean enough prediction without too many ceteris paribi to hold. The means with which conduct actual experiments and collect measurements lack sufficient precision. All data we deal with will be approximate, incomplete, biased, or some combination of these and other problems. Very few data analyses will be sufficient to knock any set of assumptions off as unjustified. At best, they can provide support for or against some set of premises, but not necessarily prove them “wrong.”
The problems of “science” and “evidence” in social sciences make for two possible paths: it could encourage the kind of arguments that Liu suggests, by opening up discussion to multitudes of possible scenarios, premises, and arguments that are compatible with the observed data, encourage alternate approaches to thinking about facts, and different logical implications thereof. Or, it can allow for obdurate persistence of closed-mindedness, impervious to contrary evidence because no evidence is “good enough.” If the acceptance of wonkish premises is part of the equilibrium in a signaling game where it indicates membership in a particular tribe, the latter is far more likely, to be complemented by the rejection of the same logic, the good and bad, as the signal for membership in the other tribe. This is, of course, what seems to often take place in United States today, in course of the culture wars: if evolutionism, including scientifically erroneous versions thereof, has become the signal for membership in one tribe, creationism indicates the signal in the other tribe. There is little or no room between them. Likewise, you buy into the wonkish premises, and the notion that that makes you always right, by faith. To the degree that wonkism is based on real evidence, you probably are, on average. But it seems absurd that you should be invariably right–unless it is a cult of cargoish variety that becomes its own self-licking ice cream cone that perpetuates itself through self-justification.
To break down the barriers of wonkism, it seems that two steps are necessary. First, wonks need to secularize themselves: theories are only right within limits. They are words of God, inherently “right,” but are right only as long as the pertinent conditions are met–which is hardly always. If you are standing at the limits of the theory, you need to open yourself up to fresh ideas, even if that means listening to heretics and nonconformists–who might actually know something that you don’t. At the same time, opponents of wonks need to accept that, for many things, perhaps even most things, wonks know what they are talking about. Facts and logic are not controvertible, at least where they are applicable. Opponents of wonks need to come up with something better than they disagree with wonks’ conclusions, but be aware of where their facts and logic come from, where they work, and where they fail, and why. Paul Krugman, before he became too wonkish, had a nifty essay on mapmaking: how the people of Middle Ages and Renaissance, before the age of modern cartography, often had a lot of good information about how the terra incognito looked like, but also had a lot of crappy ideas too, and how the modern cartography wound up throwing away the good with the bad and had to take centuries relearning them. My reaction when I first read the essay was that much of that effort was ultimately wasted–if the cartographers and sailors, travelers, etc. could talk to each other in full recognition of limits to each others’ knowledge, the mapping expeditions and such could have been more precisely targeted, saving lives, time, and expenses while producing better maps and expanding knowledge faster. Krugman, perhaps already being a wonk, does not seem to contemplate this possibility much, though, as he seems to relish how the cartographers had to tear down the knowledge in order to rebuild it in image they could understand even at expense of centuries’ time and lives of many explorers.
An open, honest, and knowledgeable argument, humble in recognition that our knowledge is limited by genuine uncertainty, in a manner suggested by Liu, could be a great thing. Could we actually pull it off? That I am a bit skeptical. Wonkism, even if built on knowledge, perpetuates itself by convincing itself that wonks know more than they do. No one to remind them “Respice post te. Hominem te memento.” Vespasian was joking on his deathbed, as he said “Vae, puto deus fio.” The same lines seem to be uttered internally by many wonks today, without any pretense of joke, I fear.