STEPHEN JAY GOULD's new book is a provocative

study of the susceptibility of basic science to social bias. It is also a superbly told tale. Yet, despite these impressive scholarly and stylistic virtues, The Mismeasure of Man leaves me with a sense of incompleteness and disquiet about the scope and judiciousness of the scietific indictment that the author makes.

Here, in his third book written for the general public, Gould, an evolutionary biologist at Harvard, traces the evolution, over the past 200 years, of scientific error and scientific fraud in studies purporting to prove that human abilities and character traits are biologically predetermined. Gould's penetrating analyses reveal a continuous and convincing record of the power of unscientific convictions in leading investigators of the highest repute to accept inadequate data, overlook contradictory evidence and misinterpret--or even falsify-- scientific results. Professor Gould has a capacity, rare among his academic peers, of translating scientific inquiry into the captivating detective story that it truly is. In this instance, there are two plots being developed at the same time, each with a special drama of its own. In the first, the super-sleuth is Gould himself, who with the meticulous care and brilliance of a Poirot, searches through the annals of science, published and private, to discover incriminating evidence of scientific misfeasance and malfeasance. Next is the tale told by the evidence itself. Here the dramatic twist is a turnabout; scientific detectives of the highest rank are unmasked as malefactors who disregard and distort evidence in order to justify their prior suspicions.

Gould is quick to point out that, except in a few instances, the disregard and distortion are not deliberate; rather, they represent well-meaning accommodations and corrections of data to conform with beliefs prevailing in the scientific culture of the time. In the author's words, "Science must be understood as a social phenomenon."

To set this phenomenon in historical context, Gould begins by citing patently racist statements by such apostles of American egalitarianism as Jefferson, Franklin and Lincoln. These quotations are followed by similar pronouncements, now given scientific legitimacy by the leading naturalists of the mid-19th century--Linnaeus, Cuvier, Lyell, Darwin and Agassiz. During the same historical period, there appeared the first empirical investigations of race differences in mental capacity based on measurements of the skull. It is here that Gould steps into his own as scientific detective. Through painstaking reexamination, and--on occasion--recalculation, of original data, he discovers arbitrary elimination of cases that contradicted the investigator's thesis, exaggerated claims resting on negligible differences in measurement, outright fabrications of scientific material, and other instances of what Gould aptly dubs "scientific skullduggery."

An especially flagrant example was brought to light by the author's Holmsean ingenuity. Enlisting the aid of a colleague who was an expert in modern photographic technology, Gould was able to confirm his own suspicion that the portraits included in H.H. Goddard's widely cited 1912 study of inherited traits in successive generations of the Kallikak family had been deliberately touched-up to convey an impression of imbecility and depravity.

At the beginning of this century, the cause of scientific hereditarianism gained new momentum from three events originating in Europe. The first was the new tide of immigration to the United States from countries in southern and eastern Europe; the second was America's entry into World War I; and the third was the invention by Alfred Binet of the standardized intelligence test. The instrument was subsequently adapted by influential American psychologists and used widely and indiscriminately for screening immigrants and evaluating the mental ability of draftees. The biased results were then presented to the public as scientific proof of the genetic inferiorty of ethnic groups coming from such areas as Africa, and southern and eastern Europe. Gould marshalls damning evidence of the inappropriateness of the test content for newly arrived immigrants, the stressful conditions under which the examinations often were administered, the inadequacy of the instructions given, and, especially, the shameful way in which prominent scientists misrepresented findings in their successful effort to establish the discriminatory immigration quotas that were embodied in the Immigration Act of 1924.

This was by no means an aberrant episode. The author proceeds to document a continuing heredtary bias in the work of major figures in the field of mental testing from Goddard and Charles Spearman in the early 1900s, through Robert Yerkes, Carl Brigham, Lewis Terman, and L.L. Thurstone in succeeding decades, to the discredited findings of Cyril Burt and the ponderous but ultimately invalid statistical arguments of Arthur Jensen.

Gould points his detective's revolver not only at the misinterpretation of mental test data, but also at the tests themselves. What emerges in the second half of the book as the ultimate "mismeasure of man" is the summary statistic derived from tests of intelligence--the IQ, or intelligence quotient. For Gould, the IQ embodies two basic fallacies inherent in all scientific efforts to quantify human abilities and traits. The first is reification, "our tendency to convert abstract concepts into entities"; the second is ranking, "our propensity for ordering complex variations in a gradual ascending scale." What is the effect of these scientific operations? In a revealing passage, Gould acknowledges that IQ tests can be useful for diagnosis of particular difficulties in mental functioning; it is the total score that he regards as meaningless: "Speaking personally, I feel that tests of the IQ type were helpful in the proper diagnosis of my own learning-disabled son. His average score, the IQ itself, meant nothing."

It is noteworthy that in contrast to his extensive use of systematic studies as bases for exposing the fatal flaws of IQ tests, Gould's only allusion to their possible virtues draws on personal experience. One would never guess from the broad review of research on human abilities presented in this volume that the most telling evidence against what Gould called "the hereditary fallacy" comes from studies principally employing the IQ test to measure their outcome. The most compelling data come from a widely used research paradigm curiously never mentioned by Gould. He identifies the comparison of identical twins reared in widely different environments as "the only really adequate natural experiment for separating genetic from environmental effects in humans." But the interpretation of twin studies is complicated by the fact that twins inevitably share a common prenatal environment and, because they look alike, also tend to evoke similar reactions from the external world. For this reason, a somewhat more elegant and powerful design is provided by comparing children of adoptive and natural parents. Here the evidence shows, for example, that youngsters born in low-income black families but placed for adoption during the first year of life in well-to-do homes, when they were tested in school several years later, scored about 20 points above comparable children raised in low-income environments.

The coup de grace to the hereditary fallacy was delivered in a recent study in which the hypothesis of racial inferiority was tested directly and yielded results that can reasonably be called unequivocal. To provide a precise measure of racial variation within a black population, blood groups were used to estimate the proportion of each person's African and European ancestry. Holding socioeconomic conditions constant, the degree of white ancestry had little or no effect on mental ability as measured by IQ tests.

Would Professor Gould have us draw no scientific implication, let alone a bit of nonscientific joy, from such findings on the ground that a "reified" IQ score can have no meaning? Hence my sense of incompleteness and disquiet about this very fine book. Studies demonstrating an environmental basis for racial differences in mental functioning as measured by IQ scores go back more than half a century. Their results have powerful implications both for science and for social policy. Yet, such investigations are not mentioned in Gould's otherwise comprehensive and penetrating analysis. The IQ, it would seem, must remain a meaningless, scientifically fabricated abstraction--"mismeasure of man."

The fact that it comes out this way, however, can also be viewed as a validation of Gould's major thesis. As he tells us, "My message is not that biological determinists were bad scientists or even that they were always wrong. Rather, I believe that science must be understood as a social phenomenon, a gutsy, human enterprise, not the work of robots programmed to collect pure information. I also present this view as an upbeat for science."

In the last analysis, that's what this book is.