PSEUDO-EGALITARIANISM, CLASS STRATIFICATION, AND THE APTITUDE/IQ TESTING CONTROVERSY Morton G. Wenger Department of Sociology University of Louisville Louisville, Kentucky 40292 MGWENG01@ULKYVM.LOUISVILLE.EDU Colloquium paper presented at the Annual Meetings of the Society for the Study of Social Problems, San Francisco, California, August, 1989 ABSTRACT For all the vast critical literature on the role of intelligence and/or aptitude testing in the acquisition of schooling, and the presumed consequent effects on educational attainment and life-chances, certain key issues remain relatively unexplored. Although much critique has evolved out of the arguments of Michael Young (1958) and Jencks (1972), the form taken by more recent and popular attacks on "ability" testing as the engine of meritocracy and mobility has tended to lose the main thrust of the original expressions; i.e., that "equality" and "merit" are contradictory social moments in a variety of ways, and not merely obvious ones. The essence of the current and widely-disseminated cant on A/IQ testing is that these procedures are solely reproductive mechanisms for the existing class order, rather than the mechanism of social justice that they claim. This position has been transformed into a politico-ideological practice which is now having effects on the major college admission tests, and takes the overt configuration of claims that written tests of verbal and mathematical "ability" are, in reality, little more than obfuscatory measures of the possession by a child of the intellectual formation of a particular class culture. It is then argued that the assignment of educational opportunity and, it is presumed, later social roles on this basis, is nothing more than a deceptive and pernicious replicator of the status quo. The usual counterfoil to this position is a claim that the tests measure something real and "innate" or "genetic," and this dialectic usually spins down into a tired nature vs. nurture polemic. However, several other possible theses exist which are more "radical" than those presented by the adversaries of ability testing. One assumes that these tests measure something valuable, but which is acquired socio-culturally and which comes to be differentially possessed by different classes, much as do prenatal care and beach-houses. This position is suggested by the work of Bourdieu and Bernstein, although not in unambiguous form. While not at variance with the position that ability testing serves in general to reproduce the class order, this view implies that this result is not inherent in such testing, but is a result of the socio-political processes in which it is embedded and of which it is a part. If further implies that the elimination of such tests could result in an educational "spoils" system which would have no less, or perhaps an even greater, tendency to reproduce the existing class order, and that the radical "abolitionist" stance on such testing thus presents no threat to that order, which may explain its growing support in hegemonic organs such as the press and courts. Introduction It is much easier for this author to do social science by virtue of having read Matthew Arnold's poem Dover Beach. The image of "...ignorant armies clashing by night..." cannot but have appeal to some who watch the development of "debates" and controversies within sociology, psychology, anthropology, and their academic siblings. For all the high fallutin' "sociology of sociology" and "philosophy of the social sciences" which attempt a reconstruction of social scientific practice, as seen through the lens of a perhaps equally distorted reconstruction of the "natural" sciences, there is never the clarity of the confrontation between heliocentric and geocentric theses. Why this should be is beyond the scope of this paper, and perhaps beyond the ken of this author. However, it is the main argument of this paper that nowhere is there more to be found a conflation of ideologically confused and intellectually imprecise theorization than in the consideration of the social role, nature, and consequences of aptitude/IQ testing. This is true despite, or perhaps because of, the antiquity of this controversy and the torrent of words it unabatedly engenders. It would seem almost unbelievable that there could be anything more to say about these matters than the deployment of additional "empirical" findings to bolster one or another already well-defined conceptual stance. Nonetheless, this paper argues otherwise. Specifically, the stance taken here is that there are at least two radically incompatible positions on IQ testing which operate from an ostensibly "critical" perspective -- in addition to the more commonly encountered "nature vs. nurture" antinomy -- and that the two "left" conceptualizations differ dramatically in their socio-political or "policy" implications as well as in their analyses. Further, it will be argued that one of these ostensibly critical perspectives is in fact supportive of the larger social status quo and consistent with its hegemonic superstructure, even while claiming to oppose the existing class order. Genetic Determinism vs. Cultural Determinism Although the parameters of the debate were loosely established many decades ago, in the past two decades, the popular media, as well as academic journals, have remained replete with argument derived from the following: scores on IQ tests and similar aptitude tests, such as the SAT, differ substantially between "racial" and/or ethnic groups, between the genders, and -- where the term has dared raise its head -- between the classes. In its most recent forms, as students of the voluminous popular and scholarly literature are aware, the differences that have drawn the most attention are overall score differences between Euro- and Afro-Americans, to the detriment of the latter; differences between males and females, particularly in terms of differential "success" on verbal and mathematical test components; and based on a more extreme and inverted variation of the gender phenomenon, differences between Asian-Americans and all others. More specifically, the origins of the controversies lie in the following findings: Euro-Americans score higher overall on such tests than do Afro-Americans; boys tend to score higher than girls on mathematical than verbal components, with the obverse also being true; and Asian-Americans tend to score higher than others on the mathematical components, and score less impressively on the verbal components. A recurring sub-theme which has been emphasized by feminist critics of the tests seems to have been that the mathematical component is somehow evaluated more positively than the verbal, to the detriment of females. In toto, these findings, the validity of which go generally undisputed and which will remain so here, form the core of a debate between those identified here as genetic and cultural determinists. This conflict has not been confined to the academy or the editorial pages, but after more than a century of dialectic around the theses of Malthus, Galton, and Binet, has even found its way into state legislatures and the courtrooms, where the test abolitionists have had considerable recent success. As is well known, the dispute between the two positions now at issue distills down to the question of whether the test differences are reflections of some underlying, genetically preordained difference between the categories enumerated above, or whether they are somehow "culturally" derived. The genetic determinists reply in the affirmative, and spring-board from this to the position that these differences are somehow socially relevant, and that these differences justify the equally undisputed skewed distributions of educational resources and social benefits which adhere to the test-identified social categories. The first wrinkle in this which needs attention is that even if the GD position were true as to the origins of test score differences, the subsequent syllogistic steps are not concurrently validated. That is, there is no convincing evidence presented by the proponents of the GD position, such as Herrnstein (1973), that these scores are "performance-related" in any socially significant way. This can be taken in the sense of little evidence that these scores are related to specific performance in any but a very narrow range of occupations, and a total lack of evidence that a society which rewards high test scorers, disproportionately expends resources on their social reproduction, and gives them additional responsibility and power, is any more effective in any measurable way than one that does not. Indeed, the justification for such a position seems to be an almost wholly unreflexive acceptance of "functionalist" theses of inequality as axiomatic. The wide-spread diffusion of these latter ideological principles underpins the general ease which the GD position has in establishing hegemony when its premises about testing remain unchallenged. Tumin (1953) may have triumphed over Davis and Moore (1945) in the graduate seminars of the last forty years, but clearly not in the common-sense world of everyday life. This is also notwithstanding the works of Jencks (1972) and Young (1958), about which more later. The proposition that rewarding the "more able" in a more favorable way may be extra-functional or even dysfunctional still appears as bizarre or even perverse to many in this society. The main point here, however, is that the leap from the GD position to a functionalist stance is internally consistent and ideologically sustainable once it is accepted that ability differences are genetic in origin and measured by A/IQ testing. Even though this is no small syllogistic hurdle, it is nonetheless a common one for a legitimating ideology, given that the majority of those who accept it (and these are perforce the majority of social actors in a given society) must also accept their consignment to permanent subordination, and perhaps more poignantly, must accommodate themselves to the thought of a similar fate for their near and distant descendants. Indeed, it is out of this social and political contradiction that a particular variant of the CD position emanates, that which will be labeled here the "populist-abolitionist" perspective, and it is from the consignment of some to a status of allegedly permanent subordination that it draws what popular support it may have. The essentials of the CD position are simple and widely disseminated, and are largely the opposite of the GD position: rather than measuring an underlying genetic trait or traits, A/IQ testing taps into "cultural" differences which exist between social categories. In this context, the term "cultural difference" usually denotes the idea that there exists a body of symbols particular to a social category which is sufficiently distinctive and exclusive so as to be amenable to reliable and consistent measurement. Thus, the argument proceeds, A/IQ tests are not a measure of any innate characteristic, but rather are sensitive to the degree to which the test-taker possesses a body of symbols to which the test is attuned. Further mechanisms are posited, sometimes explicitly, to account for how this works in practice; e.g., the test-makers are of the same categories as the test-takers who do well on them. Thus, in one form of this argument, the similar mental formations of the evaluators and the evaluated account for the differential outcomes between social categories. In other instances, this process is seen as having a more intentional character, as in those analyses which go back to the origins of IQ testing in French political controversies over mass public education in the time of Binet. In any event, the common themes in all variants of this position are that IQ tests are not measuring an innate characteristic, but rather an acquired one, that the trait(s) being measured are group-specific rather than individually distributed, that the tests generate higher scores for those showing the traits distinctive of those already socially privileged and, as a results, A/IQ tests are fundamentally a means of reproducing the existing class order, rather than being "neutral" and functionally-oriented measures of "ability." This is less often framed in terms of class differences than it is in terms of gender and "racial" differences, particularly when they are seen as being to the detriment of Afro-Americans or females. While there is somewhat less unity over the question of whether the cultural "biases" of the A/IQ tests reside in their form and/or structure, their content, and/or their paradigm of "correctness," these are not mutually exclusive assertions and are also not inconsistent with the overall stance. However, there are other, generally unnoticed nuances in the CD position which make this orientation somewhat less cohesive and easy to characterize than the GD position, and these subtleties are particularly fruitful for an understanding of the ideological deep structure of the entire controversy. Revision and Abolitionism There are two, apparently modal, proactive manifestations of the CD position, each with sub-variations. One, which might be called the "revisionist" position, seems to find the idea of mass paper and pencil testing palatable, but would like to change the content of the tests to diminish their biases. This position assumes that the biases are obvious enough to be identified, before or after the fact of test construction. Another sub-species, in the middle of another CD road, would see the various parts of the test emphasized differently by end-users. One form that this takes is the idea that math components are overemphasized in various selection protocols, particularly to the detriment of women. An interesting wrinkle to this, of some significance later, is the question of the effect this change would have on Asian-Americans, who tend to score highest of all categories on mathematical sections of A/IQ tests. A slightly more radical position within the revisionist camp is that which wants to reemphasize the form which testing takes. This is making real inroads in testing practice, with movement toward essay-type responses on the MCAT already underway, and with ETS now undertaking a long-term study of the desirability of moving toward show-your-work rather than answer-alone testing. These developments are extremely important to the current analysis, and will be discussed later. For now, it is sufficient to note that several positions classified together here as "revisionist" all accept aptitude/IQ testing as a flawed, but ultimately reparable and useful device for the allocation of social resources and roles. The more radical of the two CD variants mentioned above is the "abolitionist" stance. It argues, albeit with some diversity, that testing is essentially flawed, and that whatever format it takes, it will ultimately result in outcomes which are reproductive of the existing class/race/gender inequality order. This position recognizes the role of testing in a society with the inconsistent characteristics of gross inequality and a democratic legitimating ideology, and wishes to dispense with the entire phenomenon. It arises most frequently from CD discussions of differences between the Afro- and Euro-American descent groups, and seldom goes further than to recommend the abolition which leads to its label here. It is often associated with the political agenda of the more "radical" activists among Afro- Americans and their supporters. To the extent that it is developed, this position is logical, rests on some sound, albeit limited empirical findings, and generates a practical solution which is consistent with its analysis: A/IQ tests are socially deceptive, morally wrong and, less frequently, even socially dysfunctional, in that they fail to identify "talent" to which they are insensitive. However, this position is almost wholly critical in its stance: it seldom goes beyond recommending the abolition of the tests; it tends to be unattached to any overall analysis of the stratification processes at work elsewhere in the role allocation mechanisms of society; and most telling, it does not consider the consequences of the elimination of testing in an otherwise unchanged social context. What (and Who) Rises and Falls with Testing? There are certain facts which proponents of both the revisionist and abolitionist positions on A/IQ testing ignore: first, their use is not to the detriment of all racial/ethnic minorities, including some which are not now privileged, or which were not privileged at all points in their history. As was mentioned before, this has arisen in the context of the attempt to exclude Asian-Americans from admission to prestigious colleges and universities, where the use of composite SAT scores has been seen to favor them "disproportionately;" i.e., to harm the interests of members of the dominant majority group. When this controversy has arisen, as it has recently in the state university systems of New York and California in terms of access to "flagship" campuses, the argument has proceeded with little deployment of ideas of genetic or cultural causality, but has proceeded simply on the grounds of interest group politics and the "optimal" ethnic mix at such institutions. It appears in this context that when the tests fail to protect the heritability of social position, they are merely reduced in social significance, and protectors of the dominant interest group become rapid converts to de facto abolitionism. The suggestion that verbal components of A/IQ testing be given additional value in selection procedures is baldly discriminatory against minority group members whose facility with English is limited by their nativity, and is unaccompanied by any provision for taking such verbal tests in Vietnamese, Mandarin, or Korean. The silence in the face of this outrageously discriminatory proposal is profound among abolitionists whose focus is Afro-American contra Euro-American inequalities, but this is at least consistent with an overall condemnation of testing. However, the same cannot be said about those whose focus has been male-female differences, which makes feminist- oriented abolition somewhat suspect as to the sincerity of their opposition, given that it is young women of the dominant class and ethnic groups who would benefit most from such a change. As was alluded to above, the past history of minority groups other than Afro- Americans tends to be ignored in the abolitionist syllogism. The use by racist ideologues after WWI of results from the Army "Alpha" test are notorious among students of American minority relations. These "findings," which showed first generation and foreign-born Eastern European Jews and Italians to be profoundly sub-normal in intelligence, were the "scientific" base upon which much of the restrictive immigration legislation of the post-WWI period relied for "non-political" validation. Of course, such practices are seen by many contemporary analysts as primitive and absurd obfuscatory tactics designed to hide an anti- leftist, anti-Semitic, and anti-Catholic agenda behind a pseudo-scientific facade, which penalized second-language English speakers for the lack of an acquired skill rather than measuring anything innate. Yet, this retrospective view of the 1920s seems to have done little for the compatriots of these groups living in the 1980s. Perhaps more importantly to an overall understanding of the deep structure of the A/IQ testing controversy is the history of the groups denigrated on the basis of the A/IQ testing of the post-WWI period, subsequent to that time. Although, interestingly, it is difficult to find hard data on differential test results of that time, it is clear that as the offspring of European migrants acculturated, their test scores increased on the various devices used at that time, as well as their grades in schools which were becoming less hostile to them, possibly in part due to the increased presence of teachers of their own background into the public schools. At that time, as less and less is heard about innate inferiority, the quota system becomes a crucial factor in limiting access to legitimated access to highly rewarded social roles. While the quotas were sometimes overt, they sometimes operated under more subtle guises, such as "regional diversity," as in the case of Columbia University's imposition of different admissions criteria within and outside of a sixty mile radius of New York City. The effect of this was to exclude from fair evaluation over one-third of the Jews living in the United States, and a substantial proportion of Catholic Italians and, to a lesser extent, Catholic Irish. The current controversies at Berkeley and Binghamton about formal or informal quotas for Asian-Americans are only an echo of earlier struggles. In either case, the point is that A/IQ testing- from the perspective of the hegemonic class- is "fairly" applied only when it results in outcomes favorable to dominant class and/or ethnic and/or gender categories. While such an analysis would seem on the surface to bolster the "abolitionist" position on A/IQ testing, given that it asserts that such tests are ultimately directed only to the interests of specific groups, regardless of who ultimately scores what, it is not quite that simple. The issue here goes beyond the question of how far Jews, Italians, and Irish, etc., would have "risen" at all in this society had there not been testing and increasingly rationalized and accountable criteria for college admission in the U.S. The answer is probably little, if at all. Indeed, the awareness on the part of Jews regarding this can be seen in the willingness of the organized Jewish community in the U.S. to take terrible political damage over the questions of "affirmative action" in education and employment. Unlike this phenomenon, there are as yet unexplored questions of possible alternatives to the use of A/IQ tests and their consequences. The question is now this: if the abolitionist position on tests were to prevail, what would the reality of the allocation of social roles then be? Assuming that there would still exist a pseudo-rational distribution of social position based on "performance," and that all else remained the same, then the residual and operative criteria would be school grades and recommendations by school functionaries. Have any abolitionists ever stopped to consider the class reproductive consequences of this? Rarely. Is there not a plethora of data suggesting that all forms of evaluation by teachers not based on routinized testing is strongly influenced by the class, ethnic, or gender characteristics of students and teachers, with the latter the less significant variable? Even supposing that general cultural biases toward relatively disadvantaged groups are not shared by members of the disadvantaged groups, what happens then? During the period of runaway grade inflation during the late 1960s and early 1970s, faced with uniformly high grades, those charged with making admissions decisions at "prestigious" learning institutions routinely engaged in the practice of discounting grades from some feeder institutions, and enshrine those from others. Would abolitionism argue that this process is less class-biased that A/IQ testing, both in terms of what institutions get discounted and who is doing the discounting? If there were a shift -- and as was noted earlier, there is -- away from multiple-choice testing toward judgment of narratives, problem-solving techniques, etc., would there be less of a class bias involved in the judgment paradigms of evaluators, and who would the judges be? Indeed, other than a priori quotas, would there by anything in the above to mitigate the unbridled operation of class biases almost impossible to specify? If, as appears to be the case, the answer to all of the above is in the negative, then the appellation of "pseudo-egalitarian" appears to be well-deserved by the abolitionist position. However, there is one form of cultural determinism on A/IQ testing that has yet to be examined here, and there still remains the larger question of the necessity of making any logical or empirical connection at all between the phenomena of "merit" and social equality. A/IQ Testing, Remediation, the Meretricious, and Merit It is possible to accept all of the postulates of the CD position and still not accept the abolitionist posture on A/IQ testing. This position is latent in the above discussion, but requires a more explicit formulation; to wit, that A/IQ tests do not measure anything genetic, that they do measure class-related cognitive skills or styles, that they are utilized socially only for advantaged groups and against the disadvantaged, but that they need not do the latter. Consider that the sphygmomanometer measures blood pressure differences between Afro- and Euro-Americans, revealing considerable disability on the part of the former group. Is it then appropriate to suggest the abandonment of ethnically specific data on the differential incidence of high blood pressure? It is to be hoped that such a suggestion would be universally regarded as absurd. What of the gathering of data which shows lower birth-weights for Afro-American than Euro-American infants? Should children no longer be weighed at birth? Obviously, this is as equally absurd as the previous suggestion. Why is it then not as absurd to suggest the abandonment of A/IQ testing? The only plausible answer is that while hypertension and high infant mortality are unambiguously characteristics to be eschewed, and their opposites unequivocally to be sought, the same cannot be said of that which conventional A/IQ tests measure. However, when the abolitionist discourse is surveyed, it is not at all clear that it is ever maintained that what the tests are measuring is intrinsically valueless. Would the abolitionists maintain that the rise in recent years of SAT scores among Afro-Americans, both absolutely and relative to those of Euro-Americans, is no cause for happiness? This is unlikely. More significantly, were "inter-racial" differences to disappear entirely, would there still be a call for abolition? This is as unclear as it is speculative. That which is clear is that at least at one point in time there was the explicit as well as implicit acceptance of the idea that cognitive skills as measured by A/IQ testing were valuable, not only in terms of their ultimate conversion into salaries, but intrinsically; that lack of them was a negative consequence of oppression no less than lack of access to health care or a nutritious diet, and that something could and should be done about this at the level of social policy. This "remediationist" approach was that which imbued the impetus for the cluster of social programs lumped together in the popular mind as "Head Start," the long-term consequences of which may be the very closing of the A/IQ gap mentioned just above. There is a seeming inconsistency in the above discussion which must be resolved before it can be concluded: if A/IQ testing is, ultimately, only for the already advantaged and against those they oppress, what is the point of remedial action? Won't the rules of the game be changed if the tests become less class-biased in their measurements? Won't, as was suggested above, the tests themselves be changed? While these are questions the answers to which will only emerge politically, and in the context of the current political milieu, lying behind them are deeper social questions which can only be answered in an as yet non-existent political environment. The two most important of these are as follows. First, is the value of a cognitive skill only determined by its exchange value as a commodity? Obviously, in a capitalist society the answer is mostly in the affirmative, but it does not necessarily have to be. As the case of all other advanced capitalist societies shows, the demand for good health and its satisfaction need not await the Millennium (or the millennium). The same may be said for a demand for the full development of human mental faculties, although such a demand is seldom heard and has yet to be fulfilled anywhere. Were such a goal set in a still imperfect society, how would it be possible to ascertain its achievement, sectorally, or for the whole society? Wouldn't it be necessary to deploy some form of universal testing, however its content might be modified and/or enlarged? Second, and implicit in the above, could a society which set full human development as its goal ever be one likely to operate for long on principles of the exchange of labor for its money value? Would it be a society with a false consciousness sufficient to embrace large remunerative differentials among its members? If not, where would be the sting of A/IQ testing? The origins of A/IQ testing as a topic of controversy emanate from the use of such devices as mechanisms of role, and hence wage and status, assignment. Devoid of this use, such tests could be of no more significance than measurements of strength currently are: of interest only as indicators of underdevelopment to ameliorative personnel, or as ego- manifesting devices for a small and idiosyncratic segment of society. This last observation is not really a new one; it is to be found in different form in both the work of M. Young and Jencks that was alluded to before. Their position, expressed in quite different ways, is that there is no necessary link between "equality" and "merit," neither as social goals, nor as ideas; in fact, they suggest, they may very well be mutually exclusive. Rather, the idea in those works and here is that social equality likely cannot be achieved by evaluations of "merit," utilizing A/IQ tests or other devices, but rather recedes as a social goal where role assignment proceeds on that basis. Further, it is suggested here that only where there are conditions of social equality, which ultimately must be achieved by conscious political action, that differences in "ability" will recede to the point where the innatist positions appear as absurd in the common-sense world of everyday life as they do to most sociologists. In conclusion, then, the main problem with conventional criticisms of A/IQ testing is that they attack them as real sorting mechanisms, rather than as legitimations for a class sorting that is logically and empirically prior to testing, and which will dispense with or invert A/IQ testing as it suits the purposes of the dominant group which benefits from that order. The irony is that these tests, or the differential performance thereon, is one more indicator of the desperate need for the replacement of that order. In both obvious and baroque ways, these tests reveal the ugly fact that class position, ethnic membership, and gender identity determine even the most fundamental aspects of human development, as the work of Bernstein (1973 ) and many others suggests. To dispense with the tests only serves to obscure this, and changes nothing, except possibly in a negative way. Jencks (Jencks and Reisman, 1968:150) once said that "America most needs . . . not more mobility but more equality." This could be amended to read that it is not less information about their oppression that the oppressed need; rather it is less oppression. REFERENCES 1973 Bernstein, Basil Class Codes and Control: Vols. 1-3. London: Routledge and Kegan Paul. 1945 Davis, Kingsley and Wilbert E. Moore "Some Principles of Stratification," American Sociological Review, Vol. 10, No. 2, pp. 242-249. 1973 Herrnstein, Richard J. IQ in the Meritocracy. Boston: Little Brown. 1968 Jencks, Christopher and David Riesman The Academic Revolution. New York: Doubleday. 1972 Jencks, Christopher, et al. Inequality: A Reassessment of the Effects of Family and Schooling in America. New York: Basic Books. 1953 Tumin, Melvin M. "Some Principles of Stratification (A Critical Analysis)," American Sociological Review, Vol. 18 (August), pp. 387-393. 1958 Young, Michael. The Rise of the Meritocracy. London: Thames and Hudson.