Theodore Dalrymple in City Journal reviewing Paul Hollander’s From Benito Mussolini to Hugo Chavez: Intellectuals and a Century of Political Hero Worship said (a long but necessary quotation; the emphasis in the third paragraph is mine):
Imprisoned serial killers of women are often the object of marriage proposals from women who know nothing of them except their criminal record. This curious phenomenon indicates the depths to which self-deception can sink in determining human action. The women making such offers presumably believe that an essential core of goodness subsists in the killers and that they are uniquely the ones to bring it to the surface. They thereby also distinguish themselves from other women, whose attitude to serial killers is more conventional and unthinkingly condemnatory. They thus see further and deeper, and feel more strongly, than their conventional sisters. By contrast, they show no particular interest in petty, or pettier, criminals.
Something similar can be noted in the attitude of at least some intellectuals toward dictators, especially if those dictators claim to be in pursuit of a utopian vision. Paul Hollander, professor emeritus of sociology at the University of Massachusetts, Amherst, has long had an interest in political deception and self-deception—not surprising in someone with first-hand experience of both the Nazis and the Communists in his native Hungary. In 1981, he published his classic study of Western intellectuals who traveled, mainly on severely guided tours, to Communist countries, principally Stalin’s Russia, Mao’s China, and Castro;s Cuba, and returned with glowing accounts of the new (and better) worlds under construction there. The contrast between their accounts and reality would have been funny had reality itself not been so terrible.
In From Benito Mussolini to Hugo Chavez, Hollander turns his attention to intellectuals’ views of a wider range of dictators and authoritarian leaders. His study makes no pretension to scientific, or rather pseudoscientific, quantification, for example by first defining random groups of dictators and intellectuals and then administering structured questionnaires to the intellectuals about their attitude to the dictators. This kind of precision is often mistaken for rigor, but measurement is not meaning, and humans inhabit a world of meaning. Hollander’s study is therefore qualitative: none the worse, and a lot more interesting, for that. Even if only 10 percent of Western intellectuals, however defined, were apologists for, or admirers and supporters of dictators—sometimes serially, so that when one dictator finally dies or disappoints, another is adopted as a political hero—the phenomenon would still be significant and important. The list of influential intellectuals who have given their blessing to the most obviously terrible regimes is impressive: H. G. Wells, George Bernard Shaw, Romain Rolland, Jean-Paul Sartre (a serial offender), Norman Mailer, C. Wright Mills, Michel Foucault, and scores of others.
Pseudoscientific quantification. Amen and amen.
Two problems. Statistics, which is to say classical statistical modelinig, is now everywhere accepted as replacement for reality. We can no longer just say “Here is what happened.” No, we have to say what happened might not of happened because what happened wasn’t consonant with some ad hoc model. That’s called hypothesis testing, and is the first leg of pseudo-quantification.
Second, putting numbers where numbers don’t belong. I recently wrote this to a client about analyzing a set of survey questions:
Remember: do not over interpret. Most times correlations on surveys are only because the questions are near re-wordings of each other. Q1 = “Do you like corn?” Q2 = “Do you eat a lot of corn?” would have high correlation. But if you called Q1 “Maize Appreciation” and called Q2 “Oleic acid ingestion” (because corn oil has that substance), then you’d be foolish to write a paper which claims, “Oleic acid linked to increased maize appreciation.”
But it happens ALL the time.
It does. All the time.