George Orwell had an interesting perspective on science, as opposed to “science!(tm)” that he elaborated at length in 1984. Unfortunately, I can’t find my copy of the novel which is hidden somewhere and the appropriate passage, spoken through Emmanuel Goldstein in the Theory and Practice of Oligarchical Collectivism (the book within book in 1984) so I can only summarize what I remember roughly: there is plenty of “scientific and technological progress” taking place in Oceania, as long as they are in support of the Party’s goals, in form of better “rocket bombs,” etc., but “science” in the old sense of critical and empirical thinking have ceased to exist, to the point that there is no longer even a word for “science” in newspeak. A much more succinct version, in form of a letter from Orwell to the editor of a newspaper, can be found online, fortunately or unfortunately.
Orwell’s critique of the kind of proposal made by this Mr. J. Stewart Cook might as well apply to the attitude towards “science” and “knowledge” among today’s population, indeed, a whole new kind of thinking that undergirds elite thinking about politics, society, and the economy: that society should be run by the elites who “know better” to the exclusion of the ignorant masses, and so doing, the unstable effect of “democracy,” or as they term it, “vox populi risk,” should be minimized. This is, of course, hardly a new attitude: this type of thinking shows up in Plato. People like Richard Dawkins are practically dripping in this attitude. Stephen Jay Gould argued that this type of attitude, prevalent in early 20th century, is what led William Jennings Bryan to become a militant creationist.
While Gould, as a scientist, has a lot of issues with Bryan’s science, he fundamentally agrees with the morality of Bryan’s campaign. Indeed, he has written elsewhere at length the kind of abuse that came out of this type of thinking. In this dimension, Gould and Orwell share the same viewpoint, as did Feynman: The basic premise of any science is that, well, science is “wrong,” or at least, is an incomplete and potentially erroneous characterization of the universe. In order to understand the universe, we need to focus on when and where our theories mislead us. Whenever we find them, they offer opportunities to refine and develop our understanding further. It would be silly to pronounce those irregularities and incongruities “wrong” because they do not fit our sense of what the universe “should and must be.”
Yet, using the theory to dismiss inconvenient fact is what always happens, or, even worse, using a mass of facts to dismiss a handful of facts that don’t fit a given pattern. If the Party, the Doctrine, or the Cargo Cult must always be right by definition, the inconvenient observations must be wrong. Paul Feyerebend argued that this is, indeed, a rather common pathology among scientists. Taking the pattern and applying to the whole, including when it is not especially applicable, is a common practice among those who deem themselves “politically knowledgeable.” Dismissing the inconvenient facts, of course, is easy because many facts are, in fact, not facts at all, but observations with varying degrees of uncertainty. Against the whole that is mostly true, a handful of odd observations could simply be mistakes, mirages, or other oddities of no significance. If X is true 99% of the time, why bother checking the exceptions that are rare and unusual?
The attitude that winds up prevailing, then, is not an inquiry based on skepticism and open mind, but a systematic and coordinated attempt to justify the usual and the common place, why the things that we “know” to be true are true and must be true–if you will, a cargo cult of the mundane and the usual. The original cargo cult seemed strange because the sudden prevalence of cargo, delivered and gifted by US military personnel in far off Pacific Islands, was but a transient event and it seems odd that a whole cult should have sprung up so quickly around it and persisted for decades afterwards. Yet, history of mankind is full of other cargo cults of far longer standing and even longer persistence: “why earth is flat,” “why earth is the center of universe,” “why things always must fall down,” and so forth. All these are “true” in common observations–that these are, in fact, not really true at all requires observing the kind of subtleties that are not readily apparent to the usual observer. When everyone “knows” that earth is the center of the universe, and when every person of import has spent time explaining why it must be so, those who deign to ask, “is earth the center of the universe” must be insane, idiotic, or worse and should be dismissed accordingly–unless he is such a recognized genius with friends in high places as Galileo, in which case a lot of political controversy would ensue, to embarrass his friends and allies mostly.
“Science” in support of propping up the widely accepted conventional wisdom, of course, is not simply designed to reinforce the “socio-political” status quo. Being able to describe the 99% of the universe with reasonable precision is critical because, most of the time, this is the universe we happen to be living in. We want to design machines, rules, institutions, practices, and such to better take advantage of the patterns that take place most of the time. Questioning whether earth really is flat is irrelevant for this exercise: we want to draw maps on a two-dimensional space, which is about the best we can do much of the time. Creating maps that take advantage of the curvature of the earth, in addition to being difficult, confers very little practical advantage except on very huge scale. In service of the practical, then, open mind and skeptical empiricism are a distraction, something that should be kept out of minds in favor of various “facts” that are true because they are (and they are true empirically 99% of the time–i.e. true enough to be universally useful.) Kuhn argued in Structure of Scientific Revolutions, argued that new scientific ideas emerge when sufficient technological progress has taken place to see the exceptions and oddities that conform to the old conventional wisdom–without the technological means to see where the ideas were off, new scientific ideas have no legs to stand on. It appears to me, however, that Kuhn took too much for granted, that, somehow, technological development that enables scientific progress would naturally emerge by itself. Technological progress that would push the limits of the conventional wisdom emerge when there is a demand for machines, institutions, and practices of the kind that pushes the limits of conventional wisdom, such as building a ridiculously huge domes or a political coalition that spans both the extreme left, extreme right, and/or a bunch of “nonideologues.” And a scientific attention is justified only when these events become sufficiently common enough to demand explanation.
Building huge domes was the craze of the Renaissance, and how to push the limits of the building technology naturally drew attention. Everyone who could afford it wanted to build big domes, and it became painfully clear (literally in many cases) that the old conventional wisdom was no longer applicable as domes grew larger. But political coalitions spanning odd groupings are rare enough to be dismissed as momentary blips and are rarely successful enough to justify attention. Of course, incumbent socio-political institutions are, even more than physical universe, even more hostile to vacuums: creating a vacuum is merely difficult–nature does not “actively” intervene against you. The Democratic Party does intervene actively against attempts at creating an alternate coalition, for example. So the availability of the “facts,” in the form of data, is naturally biased against anything contrary to conventional wisdom. Without an open mind and an inclination for critical thinking, the rare contrary data, rather than used as the means to understand the subtleties where the conventional wisdom is lacking in explanatory power, winds up being rudely dismissed as unnecessary and pointless.
In other words, if the cargo keeps giving, the cargo cult is no longer viewed as a cult by the masses. As long as the cargo keeps showing up, in fact, questioning the cult is itself verboten. Feynman’s argument held sway because people saw it strange that people kept expecting cargo even after it dried up. If they justified the cargo cult on the basis of keeping up some rituals, and if the cargo kept reappearing, the linkage between the rituals and the cargo, however cultish it might be, would have been defended as “science” on the basis of all the “facts” supporting it. This is hardly a new notion: the Aztec religion believed that human sacrifice was necessary to keep the sun reappearing morning after morning. Since they always practiced human sacrifice, all the data supported their belief that human sacrifice is the necessary condition for the sun. To propose that they stop human sacrifice just to see if the sun would still show up would have been, literally, the basis to be the next human sacrifice yourself. Stopping this cargo cult took the appearance of the Spanish conquistadores and a complete genocide.
We now know that the Middle Ages were not, after all, the Dark Ages. In terms of practical technology, the gains were enormous, in metallurgy, construction techniques, shipbuilding, agriculture, animal husbandry, among many other things. They saw patterns in their common experiences and took advantage of them to improve their practices in all walks of life. What they did not do was to ask questions as to what if they were wrong, for they had no need to. Not developing critical thinking skills did not bother them much, and those who did wound up wasting time debating over the number of angels on a pinhead. While some walks of life, e.g. civil engineering, did incorporate a more rigorous scientific approach if only because of collapsed domes and overturned ships, this depiction remained true well into 20th century. Most intellectuals who engaged in exploration into skeptical thinking, asking “what if conventional wisdom is wrong,” were often idle rich, or, at least, were sponsored by the idle rich. There was no expectation that anything “useful” or “practical” should come out of their endeavors and for this reason, their ranks were always rather thin as few could afford to join them.
In a sense, the events of 19th and 20th centuries, where advances in physical and, to a lesser extent, social science helped massively reshape the practical aspects of life, for good or for ill, through the chemical dyes, the atom bomb, polio vaccine, and the New Deal/fascism/communism, brought about an interest in investing in the “sciences,” but these were to be the practical sciences, i.e. how to build better means to enforce the will of the Party, not questioning the Party itself. So “data science,” as practiced, is really to squeeze from the data whatever it is necessary to “make benefit the Great Party,” not to better understand wherever the data is coming from. Is this any different from the “science” of 1984? But, if the Party is paying for the science, why would they want to promote “wasteful” thinking that may not yield profit for itself?
The implications are messy: in a sense, we are entering the new Middle Ages. The promise of “data science,” in the end, is not that we will better understand the universe, but that we will have better plows and bigger domes, because we know “better formulas,” even though we don’t necessarily know why the formulas work (we just know, literally, those are the patterns we find in the data). In a sense, this is smugly antiscientific, incurious, and even offensive, yet, not something that I can find an objection to on practical grounds. Greco-Romans may have had critical thinking, but they also had worse swords and less efficient agriculture, after all.