Papers I'm reading: Science is not always "self-correcting"

by Rudd-O published 2019/07/08 17:54:00 GMT+0, last modified 2019-07-09T10:56:25+00:00
On the widespread tendency of scientists to believe falsehoods if done for "morally righteous" reasons.

Link to the paperHere is a mirror in case the original goes away.

Abstract: Some prominent scientists and philosophers have stated openly that moral and political considerations should influence whether we accept or promulgate scientific theories.This widespread view has significantly influenced the development, and public perception, of intelligence research. Theories related to group differences in intelligence are often rejected a priori on explicitly moral grounds. Thus the idea, frequently expressed by commentators on science, that science is “self-correcting”—that hypotheses are simply abandoned when they are undermined by empirical evidence—may not be correct in all contexts. In this paper,documentation spanning from the early 1970s to the present is collected, which reveals the influence of scientists’ moral and political commitments on the study of intelligence. It is suggested that misrepresenting findings in science to achieve desirable social goals will ultimately harm both science and society.

Choice quotes:

There is widespread acceptance of the idea -- among academics -- that either (a) morality requires people to hold certain beliefs about empirical matters, and that scientists should not conduct research that threatens to uncover facts that contradict these morally required beliefs, or (b) morality requires people to hold certain beliefs regardless of the evidence.

From a philosophical perspective, scientific practice ought to involve simply abandoning hypotheses when they are disconfirmed. However, from a sociological perspective, when hypotheses are regarded as supporting certain moral values or desirable political goals, scientists often refuse to abandon them in the light of empirical evidence. We may not have freed ourselves of the tendency to conflate morality and science nearly as much as is usually supposed. The tendency to conflate morality and science is so strong that, as shall be documented, even highly sophisticated people trained in science and the philosophy of science frequently incorporate explicitly moral considerations into their scientific reasoning.  Some philosophers have even rationalized this practice, and their arguments, though fallacious, have been surprisingly uncritically accepted. There is widespread acceptance of the idea — among academics — that either (a) morality requires people to hold certain beliefs about empirical matters, and that scientists should not conduct research that threatens to uncover facts that contradict these morally required beliefs, or (b) morality requires people to hold certain beliefs regardless of the evidence.

This one appalled me — I used to hold Daniel Dennett in high regard, but he appears corrupt and hypocritical now:

In Freedom Evolves Dennett (2003) says regarding critics of hereditarianism: "I don’t challenge the critics’ motives or even their tactics; if I encountered people conveying a message I thought was so dangerous that I could not risk giving it a fair hearing, I would be at least strongly tempted to misrepresent it, to caricature it for the public good. I’d want to make up some good epithets, such as genetic determinist or reductionist or Darwinian Fundamentalist, and then flail those straw men as hard as I could. As the saying goes, it’s a dirty job, but somebody’s got to do it." 4(pp. 19–20)

Dennett scorns the Marxists for believing it “permissible to lie and deceive in order to further”their cause. But he said openly that it is meritorious to “misrepresent” and “caricature”true theories that conflict with his values. Indeed, he said this is not merely meritorious but “a dirty job” that “somebody’s got to do.”  Lying to promote his values is good, lying to promote values he disagrees with is “dangerous.”

The discipline is totally rotten and nobody seems to care:

In 1984, Snyderman and Rothman anonymously surveyed 1,020 social scientists and educators about their views on intelligence research. (All those surveyed were members of mainstream professional organizations.)  Famously, a majority of respondents indicated that they agreed with Jensen’s (1969,1980) most controversial claims: 94% agreed that at least one-out-of-five possible sources of evidence supported a significant nonzero heritability of IQ in the American white population. On a 4-point scale subjects were asked to indicate whether they thought IQ tests were biased against American Blacks, with 1 being not at all or insignificantly biased, 2, somewhat biased, 3, moderately biased, and 4, extremely biased.  The mean rating was 2.12—revealing that experts thought there is less than a moderate degree of bias (Snyderman and Rothman 1987).

What is less well known about Snyderman and Rothman’s survey is that respondents were also asked to indicate their regard for 14 social scientists, considering only their work on intelligence research. They rated each scientist on a 7-point scale, with 1 indicating “Very low regard” and 7 “Very high regard.” The mean rating for Steven Jay Gould was 4.45. For Leon Kamin, 4.36. For Arthur Jensen, 3.68 (Snyderman and Rothman 1988, Table 4.1). Kamin was known for his extreme position that IQ has a heritability of zero: “Patriotism, we have been told, is the last refuge of scoundrels. Psychologists and biologists might consider the possibility that heritability is the first” (Kamin 1974, 3; quoted in Sesardic 2005, 190).  Gould relied heavily on Kamin’s work in his denunciation of intelligence research. So although most survey respondents agreed with Jensen's scientific claims against Gould and Kamin's,they indicated that they held Jensen as a scientist in lower regard.

Prior to his controversial 1969 article on race differences in intelligence, Jensen received a Guggenheim Fellowship and a fellowship at Stanford’s Center for Advanced Study in the Behavioral Sciences. After 1969 he received no honor from any major psychological organization in the United States, despite having written a number of “citation classics” (Gottfredson2005, 160–161).  Not only has he written citation classics, but his once-controversial emphasis on general intelligence(g) spawned what all intelligence researchers acknowledge was an enormously fruitful research program. Due in part to his work, Sternberg and Kaufman(2012, 235) report that “[i]t is now as well an established fact as exists in psychology that g correlates with many forms of human behavior and their outcomes (see, e.g.,Hunt 2011;Jensen 1998;Mackintosh 2011)."  By contrast, Howard Gardner’s theory of multiple intelligences has never been empirically supported, and the assumptions behind it have been undermined by findings in cognitive science.

This one guy (Gardner) openly admits he peddles lies, implies it's "for the moral good" / "against the 'bad guys'", and also brags about sabotaging research:

[E]ven if at the end of the day, the bad guys [such as Jensen, who emphasize the importance of g,] turn out to be more correct scientifically than I am, life is short, and we have to make choices about how we spend our time. And that’s where I think the multiple intelligences way of thinking about things will continue to be useful even if the scientific evidence doesn’t support it. (at 45:11–31)

Gardner’s use of the term “correct scientifically” seems to reflect a notion that there is another sort of “correctness” besides scientific that can apply to empirical claims. IQ theorists are“the bad guys,” he says. Based on his other speeches and his writings, it is clear that the nonscientific “correctness” he alludes to is moral: it is possible to be correct scientifically but incorrect morally—or incorrect scientifically but correct morally. [...] In the same paper he recounts how he was once personally responsible for preventing research on race differences in intelligence.

A particularly brilliant conclusion among many of this paper.  This is what I have always believed:

Having true beliefs about physics and chemistry is necessary to design airplanes that fly or medicines that cure diseases. In the same way, having true beliefs about human psychology is necessary to design social policies that work. Preventing cumulative progress in psychological science for the sake of social welfare—as advocated by those scientists and philosophers mentioned in this paper—will lead policy makers to design ineffective social programs based on incorrect theories.  Even though it may be painful to think that certain disturbing scientific hypotheses are true or could be true, designing social policies based on comforting but inaccurate theories will lead to even more pain in the long run.