June 16, 2008
Jerry Hirsch 1922 - 2008. Scientist, Rebel.
Jerry Hirsch has passed away at the age of 85. He will be sorely missed.
During the early stages of his career, roughly from 1955 to 1970, Hirsch rose quickly to prominence as a leading figure of a new discipline, behavior genetics, he helped create and name. For the rest of his life he waged an untiring campaign against all those who threatened to impugn its good name, for whatever motives.
Skipping the details of his pre-graduate school education, we meet Hirsch in the early 50's as a graduate student in the Psychology Department at Berkeley. In retrospect, this was a most auspicious beginning. The department seemed tailor-made for his later career, with Edward C. Tolman, and later Robert C. Tryon, pursuing experimental work with rats on the heritability of what they billed as "intelligence" ("maze bright rats").
"Tolman (1924) reported the results of what was apparently the first selection experiment for maze learning by rats" a preliminary report of a subsequent selection experiment was published a few years later (Tryon, 1929). "A fairly consistent divergence of the strains was noted through generation VII, at which time little overlap of the distribution of the strains occurred" (DeFries, 1967, p.330f).
In most respects Hirsch's research paradigm was similar to Tryon's except that he worked with fruit flies, because (a) their life cycle is shorter, (b) they are more cheaply maintained, and (c) their chromosomal structure was better understood. As dependent variable he chose geotaxis (the tendency to fly either up or down), and later (with Boureau) phototaxis (light seeking vs. avoidance behavior). Both traits were assumed to be caused by more than one gene. The basic idea was to create two genetically pure strains by selectively interbreeding, respectively, extreme performers to study the heritability of polygenetic traits in fruit flies. While still a graduate student, Hirsch ingeniously constructed a lattice-shaped maze out of Y-shaped tubes to automate the data collection task that soon became popular with other researchers in his field.
After obtaining his Ph.D. in 1955, Hirsch moved to Columbia, whose faculty included Theodosias Dobzhansky who, together with Ernst Mayr, had been one of the architects of the so-called "evolutionary synthesis" that in the early 40s had helped heal the rift between genetic naturalists and experimental geneticists (cf. Mayr, 1982, p. 542), thereby facilitating the genetic breakthroughs of the 50s.
Hirsch could hardly have been more judicious in his choice of time and place. The early 60s were the halcyon years of higher education in the United States. Responding to the early successes of the Soviet space program, government funding had been stepped up, academic job opportunities had become plentiful, salaries more attractive, and promotion through the ranks more rapid. After five years at Columbia, Hirsch accepted an associate professorship at the University of Illinois. By this time, barely five years after obtaining his Ph.D., he had already established himself as a leading figure in his chosen discipline. He was sought after as speaker, reviewer, and contributor of book chapters. This provided him with a platform for spreading his new gospel of a radical break with tradition that, up to that time, had been under "...the pernicious influence of American behaviorism" (Hirsch, 1967a, p.420. Note 1).
According to him,
"An important task for our immediate future is to cast off the hand-me-down philosophy, overgeneralized from physics, and to develop one appropriate for biosocial science" (Hirsch, 1967b, p. 120).
Keeping in mind that he was essentially still the new kid on the block, this characteristic bluntness was astounding. Even more astounding, apparently nobody of stature seemed to have challenged him as he, in effect, undermined psychology's claim to be an objective science, which had been the stated objective of behaviorism. Lest the implications of his message be misunderstood, Hirsch spelled them out:
"Misguided by 'opinion leaders' like Clark Hull (Woodger's disciple) and later converts to mathematical theory, psychology entered a cul-de-sac that genetics had prudently avoided" (Hirsch, 1967b, p. 120. Note 2).
Hirsch cited Mayr (1959, see also 1982) to justify his bold challenge:
"In order to gain some perspective in our attempts to understand the uncompromisingly rigid stand taken toward heredity in the behavioral sciences, it is important for us to grasp certain concepts that are now seen to be fundamental. First and foremost we must appreciate the distinction between typological thinking and population thinking, as Ernst Mayr (1959) says 'No two ways of looking at nature could be more different [p. 2]'. The typological mode of thought is the older one. It is pre-Mendelian and pre-Darwinian. It has been ubiquitous in the behavioral sciences and ironically, it has been the dominant outlook throughout the majority of the more successful sciences like physics, chemistry, physiology, and anatomy as well as in medicine. To typologists individual events have only meaning and importance insofar as they are representative of some class of events. Within such a class individual differences can only be error." (Hirsch, 1967b, p. 120).
Mayr (1982) called the typological point of view "essentialist". He characterized the polarity as follows:
"What is population thinking and how does it differ from essentialism? Population thinkers stress the uniqueness of everything in the organic world. What is important for them is the individual, not the type.
[For the essentialist] Variation was nothing but "errors" around mean values. [Yet] Differences in height among a group of people are real and not the result of in accurate measurement. The most interesting parameter in the statistics of natural populations is the actual variation, its amount and its nature." (Mayr, 1982, p.46f. Note 3).
The implications of this polarity are not limited to behavior genetics, although there they may be most readily apparent: A recurrent theme in behavior genetics is "heritability estimation", i.e., attempts to quantify the relative contribution of heredity and environment to overt behavior. For such efforts to make any sense, it is of course necessary to assume, first and foremost, that both influences are separable, so that they can be measured independently of each other. Technically, this assumption is formalized by positing that the interaction variance is zero. While this assumption is often taken for granted, for example, in heritability estimates for twins, Hirsch categorically disagreed:
"Since genotypic diversity and genotype-environment interaction are apparently ubiquitous, attempts to study the laws of environmental influence have been grasping at shadows." (Hirsch, 1967a, p. 421).
It then follows:
"When these considerations are taken into account, we see why it is impossible to generalize about the contribution to a phenotype of either heredity or environment." (Hirsch, 1967a, p. 421).
Thus, he wiped out with one bold stroke the whole heritability estimation lore that had exercised psychologists for decades, not just on technical grounds (Schonemann, 1997), but in principle: As long as genes interact with environment - an assumption that is usually precluded by definition in derivations of heritability estimates - common sense implies that it becomes impossible to assign proportional contributions of heredity and environment to behavioral observations, regardless of the math. Such estimates are further compromised by the fact that, even in models that do contain an interaction term, its significance remains moot, because statistical interaction tests lack power, as Doug Wahlsten (1990) has stressed.
Let us now apply this essentialist/population dichotomy to other areas of psychology. What about "intelligence", surely a central measurement problem in psychology? Psychologists often say they "measure intelligence" with IQ tests. How can they possibly do this as long as they are unable to define "intelligence"? This was the problem Charles Spearman (1904) had wrestled with from an essentialist point of view.
More broadly, the whole "axiomatic measurement" movement of in the 60s and 70s - now largely forgotten - was essentially imported from physics (Holder, 1901; Krantz et al., 1971). As William K. Estes, one of the early promoters of this movement, eventually conceded, it got nowhere:
"One of the reasons for the relative paucity of connections between measurement theory and substantive theory in psychology may arise from the fact that models for measurement have largely been developed as a body of abstract formal theory with empirical interpretations being left to a later stage. The difficulty with this approach is that the later stage often fails to materialize" (Estes, 1975, p. 273, Note 4).
In short, once simplifying assumptions imported from essentialist physics are violated, any conclusions derived from models that require them become pointless as a matter of principle. The price for ignoring complexity for the sake of convenience is sterility.
Turning now specifically to measurement of "intelligence", we learn from Arthur Jensen, taking a leaf from his precursors in the 30s, that "Intelligence, like electricity, is easier to measure than to define" (Jensen, 1969, p.5).
Judging from their deafening silence, the physicalist measurement experts were apparently perfectly comfortable with this patent absurdity for decades. Did they really not know that nobody in physics measures electricity? What physicists do measure is voltage and current, not electricity, and those are perfectly well defined - in contrast to "intelligence".
The only scholar of note who took this fundamental problem seriously was Charles Spearman (1904). Entirely in the spirit of physicalism he postulated a mathematical model intended to ensure that "intelligence" is objectively defined and measurable, provided certain assumptions are met. As it turned out, they were virtually never met. To deal with this problem, Louis L. Thurstone (1949) relaxed the most critical assumption - that there was only one common factor, "g", arriving at 5 to 7 intelligences, which in the 40s and 50s was widely hailed as a breakthrough. Actually, this "generalization" was a step backwards, because it forfeited falsifiability. With the arrival of electronic computers his "multiple factor analysis" therefore quickly degenerated into a toy for dilettantes fancying themselves scientists akin to chemists discovering elements.
A watershed event in Hirsch's career was Jensen's resurrection of Spearman's g in (Jensen, 1969). Hirsch and Jensen knew each other from graduate school (Note 5). Before meeting the staunchly racist Nobel laureate Will Shockley in 1967, Jensen, like most other psychologists, had subscribed to the environmentalist position of behaviorism. Subsequently, he became one of the most prominent advocates of hereditarianism, the contrary view that human behavior is largely predetermined by genes. After WWII, when the results of irresponsible social science had become too obvious to ignore any longer (see, e.g., Allen, 1997; Mehler, 1997; Wahlsten, 1997, and, more recently, Black, 2003), many psychologists temporarily scorned the hereditarian position as politically incorrect.
One might think that Hirsch, who earlier in his career had loudly lamented the neglect of genetics in the heydays of behaviorism, would have welcomed Jensen's conversion. Hirsch could have saved himself a lot of trouble had he just joined Jensen's eugenic revivalist movement, especially since the time was ripe for a step backwards in history. Despite the embarrassing legacy of the 20s and 30s, the tenets of eugenics had become respectable again for discussions in mainstream psychology:
"Who now wishes to resurrect ...Terman's pronouncements in 1916 about eugenic measures to reduce the incidence of mental retardation; the primitive 1916 Army mental tests; or the US Congress's 1924 Immigration Restriction Act, which cited the 1917 Army test data?" (Jensen, 1982, p.131).
No more disparaging talk about "pseudoscience" - except for Hirsch, who stubbornly stuck to his own standards. Far from congratulating Jensen, he took him to task even before his 1969 Harvard Educational Review (Jensen, 1969) article had appeared:
"A most unfortunate speech in 1967 by A.R. Jensen (who came under Shockley's influence while at Stanford during 1966-1967) illustrates the dangers of inappropriate use of both the concept of heritability and that of race by the biometrically unsophisticated" (Hirsch, 1967b, p. 434).
One aspect of Jensenism that bothered Hirsch was its lack of standards. Science was made to look easy: Just administer a prepared list of puzzles to the subjects, count the number of correct answers, then use a computer program to arrive at "heritabilities" far exceeding anything that competent geneticists working laboriously with animals could match. In his skeptical appraisal of the merits of this kind of research, Hirsch was not alone. The doyen of quantitative genetics, Oscar Kempthorne, was equally appalled:
"...the separation of genetic and non-genetic forces with observational rather than experimental data is hopelessly difficult" The obscurity about the nature of IQ tests makes the interpretation of social differences in IQ entirely a matter of very subjective opinon...We have to ignore the writings of Burt, Jensen, and, particularly, Shockley" (Kempthorne, 1997, p. 111).
Similarly, Doug Wahlsten, an internationally recognized geneticist working with mice, has voiced his frustration at the shoddy standards of human behavior geneticists:
"So widespread are errors in this literature that the critical reader now has good reason to doubt every article published on this topic and to check the arithmetic, algebra and original references before seriously considering the 'findings' and conclusions. The pitifully low standards of scholarship of many who write on heredity and IQ are scandalous and unforgivable." (Wahlsten, 1981, p. 33).
Another thing bothering Hirsch was Jensen's advocacy of the racist myths of the 20s which also feature prominently in Herrnstein & Murray's (1994) widely discussed The Bell Curve. Hirsch (1997, p.219) singled out a passage dealing with the politically explosive issue of test bias against minorities that had apparently escaped notice of the numerous commentators. It illustrates what Hirsch (1975) meant by "Jensenism: The bankruptcy of science without scholarship". In their book, the authors claim:
"We have found no modern, empirically based survey of the literature on test bias arguing that tests are predictively biased against blacks although we have looked for them." (Herrnstein & Murray, 1992, p.628).
In corroboration they added in their Notes (p.770):
"...the following sources...unanimously accept the conclusion that no bias against in educational or occupational prediction has been found: ...Crouse and Trusheim, 1988; Hartigan and Wigdor 1989..." (Herrnstein & Murray, 1992, p. 770).
Yet attentive readers will find that, in fact, chapter 5 of Crouse & Trusheim's book, The Case Against the SAT, is explicitly entitled The SAT has an adverse impact on black applicants. And in Hartigan & Wigdor, 1989, p.7, one reads:
"...in the absence of score adjustments, minority applicants who could perform successfully on the job will be screened out of the referral group in greater proportions than are equivalent majority applicants."
I.e., the exact opposite of what Herrnstein and Murray had claimed (Note 6).
During the latter part of his career Hirsch protested relentlessly against this type of corporate "science" (Bauer, 2004). Just as Charles Brigham, the creator of the SAT, had prophesied 70 years ago (cf. Lemann, 1999, p.40), science cartels guided by commercial interests will go to any lengths to shield false or exaggerated claims from valid criticism:
"...not only have we witnessed the propagation of error but also its confirmation and encouragement through the very mechanisms that had been developed to detect and eliminate error" (Hirsch, 1997, p.214; s.a. Schonemann, unpublished).
Jerry Hirsch represented the highest ideals of scientific inquiry - competence and an unwavering commitment to the truth. Those who still share his ideals will be grieved by his passing. Perhaps his legacy will inspire them to continue To "unfrock the charlatans" (Hirsch, 1981).
Acknowledgement: I have benefitted greatly from discussions with numerous people, including Joy Kaiser, Marge Hirsch, Debbie Dexter, Garland Allen, Barry Mehler, Moritz Heene, Bernie Douglas, and my wife Roberta. I thank them all.
Peter H. Schonemann
Department of Psychology
West Lafayette IN 47907
Allen, G.E. (1997) The social and economic origins of genetic determinism: a case history of the American eugenics movement, 1900-1940 and its lessons today. Genetica, 99, 77-88.
APA (1995). Stalking the Wild Taboo: Intelligence: Knowns and Unknowns. Report of a Task Force established by the Board of Scientific Affairs of the Psychological Association. http://www.Irainc.com/swtaboo/apa_01.html.
Bauer, H.H. (2004) Science of the 21st century: Knowledge monopolies and research cartels. Journal of Scientific Exploration, 18. 643-660.
Black, E. (2003) War Against the Weak. Eugenics and America's Campaign to Create a Master Race. New York: Four Walls Eight Windows.
Boring, E.G. (1957) A History of Experimental Psychology. New York: Appleton-Canturey Crofts, Second Edition.
Crouse, J, & Trusheim, D. (1988) The Case Against the SAT. Chicago: The University Press of Chicago.
deFries, J.C. (1967) Quantitative genetics: An overview and perspective. In Hirsch, J. (ed.) Behavior-Genetic Analysis. 322-339. New York: MacGraw-Hill.
Daniels, V. (undated) Clark Hull's "Hypothetico-deductive Theory".
Einstein, A. (1949) Autobiographical notes. In Schilpp, P.A. (ed.) Albert Einstein: Philosopher-Scientist. London: Cambridge University Press.
Estes, W.K. (1975) Some targets for mathematical psychology. Journal of Mathematical Psychology, 12, 263-268.
Hartigan, J.A. & Wigdor, A.K. (1989) Fairness in Employment Testing: Validity Generalization, Minority Issues, and the General Aptitude Test Battery. Washington, D.C.: National Academy Press.
Herrnstein, R.J & Murray, C.A. (1994) The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.
Hirsch, J. (1967a) (ed.) Behavior-Genetic Analysis. New York: McGraw-Hill.
Hirsch, J. (1967b) Behavior-genetic, or "experimental" analysis: the challenge of science versus the lure of technology. American Psychologist, 22, 118-130.
Hirsch, J. (1975) Jensenism: The bankruptcy of "science" without scholarship. Educational Theory, 25, 3-27.
Hirsch, J. (1981) To "unfrock the charlatans". SAGE Race Relations Abstracts, 6, 1-65.
Hirsch, J. (1997) Some history of heredity-vs-environment, genetic inferiority at Harvard (?), and The (incredible) Bell Curve. Genetica, 99, 207-224.
Holder, O. (1901) Die Axiome der Quantitat und die Lehre vom Mass. Berichte uber die Verhandlungen der Koniglich-Sachsischen Gesellschaft der Wissenschaften, 53, 1-61.
Jensen, A. R. (1982) The debunking of scientific fossils and straw persons. Contemporary Education Review, 1-2, 121-135.
Jensen, A.R. (1969) How much can we boost IQ and scholastic achievement. Harvard Education Review, 39, 1-123.
Kempthorne, O. (1997) Heritability: uses and abuses. Genetica, 99,109-112.
Krantz, D.H., Luce, R.D., Suppes, P. & Tversky, A. (1971) Foundations of Measurement. Volume I. New York: Academic Press.
Lemann, N. (1999) The Big Test. The Secret History of the American Meritocracy. New York, Farrar, Strauss and Giroux.
Mayr, E. (1959) Darwin and the evolutionary theory in biology. In: Meggers, B.J. (ed.) Evolution and Anthropology: A Centennial Appraisal, 3-12. Washington, D.C.: Anthropological Society of Washington.
Mayr, E., (1982). The Growth of Biological Thought. Diversity, Evolution, and Inheritance. Cambridge, Mass: Harvard University Press.
Meehl P.E. & Rosen, A. (1955) Antecedent probability and the efficiency of psychometric signs, patterns and cutting scores. Psychological Bulletin, 52, 194-216.
Mehler, B. (1997) Beyondism: Raymond B. Cattell and the new eugenics. Genetica, 99,153-163.
Schonemann, P.H. (1994) Measurement: The reasonable ineffectiveness of mathematics in the social sciences. In Borg. I. and Mohler, P. Ph. (eds.) Trends and Perspectives in Empirical Social Research. Berlin: DeGruyter, 149-160. http://www.psych.purdue.edu/~phs
Schonemann, P.H. (1997) On models and muddles of heritability. Genetica, 99, 97-108. http://www.psych.purdue.edu/~phs
Schonemann, P.H. (unpublished) The Case Against the GRE.
Spearman, C. (1904) General intelligence, objectively determined and measured. American Journal of Psychology, 15, 201-293.
Suppes, P. (1968) The desirability of formalization in science. Journal of Philosophy, 65, 651-664.
Thurstone, L. L. (1949) Multiple Factor Analysis. Chicago: University of Chicago Press.
Wahlsten, D. (1981) Review of The IQ Game by Howard F. Taylor. Behaviorists for Social Action Journal, 3, 33-34.
Wahlsten, D. (1990) Insensitivity of the analysis of variance to heredity-environment interactions. Behavior and Brain Sciences, 13, 109-120.
Wahlsten, D. (1997) Leilani Muir versus the philosopher king: Eugenics on trial in Alberta. Genetica, 99, 185-198.
Woodger, J.H. (1937) The Axiomatic Method in Biology. London: Cambridge University Press.
Note 1: Behaviorism, logical positivism:
The influence from philosophy arrived via the Vienna Circle...There began a movement which Feigl later named logical positivism, which in science (including psychology) became physicalism, because it reduced all scientific language to the communal language of physics, and which in psychology became behavioristics because the psychological operations are all operation [sic] of behavior. (Boring, 1950, p.655)
Pernicious: referring, presumably, to behaviorism's extreme neglect of genetics.
Note 2: Clark Hull, Woodger's disciple:
For about 20 years Cark Hull held a dominant position in the history of American academic psychology. Vast numbers of experiments by experimental psychologists all over the country were carried out to confirm or challenge various details of his theoretical formulations. Then, during the early 60s, almost as quick as his ideas had come to dominate learning theory, they largely vanished from it ... In retrospect it is incredible how much time and energy went into this endeavor and how little came out of it. (Daniels, undated, p. 1).
Woodger, J. H. was an early promoter of the axiomatic method in biology. See Woodger (1937).
...and later converts to mathematical theory:
Referring, presumably, to mathematical learning theory which had just begun to promote the axiomatic gospel in psychology (see, e.g., Krantz et al., 1971; Suppes, 1968, Estes, 1975).
Note 3: I thank Professor Garland Allen for clarifying this important point for me.
Note 4: This debacle might have been avoided had the axiomatizers paid more attention to Einstein or, for that matter, Hirsch, before foisting axiomatization on a field that was clearly not ready for it:
If it seemed as if knowledge about empirical objects can be obtained through mere thought, then this 'miracle' rested on a fallacy. (Einstein, 1949, p.10).
In practice, the process of theory construction should not begin with axioms. (There should be a body of data. Otherwise, one may be dealing with a mathematical exercise) (Guttman, 1981, p.57).
For more on this see (Schonemann, 1994).
Note 5: To be more precise, Jensen did not really resurrect Spearman's g, but only its name. This was the best he could do at this point because g, in the sense Spearman had defined it, does not exist, as he and others well knew.
An apocryphal story has it that Hirsch had actually kindled Jensen's interest in genetics.
Note 6: Similarly, in their Report of a Task Force, the APA (1995) airbrushed out all information that might conflict with their self-serving fable that mental tests are unbiased against minorities. Among their 200 plus references there is no mention of Meehl & Rosen (1955), Crouse&Trusheim (1988), or Hartigan & Wigdor (1989).