Item: Last year, after the Board of Regents of the University of California System voted to forbid racial preferences in admissions, Charles Young, the chancellor of UCLA, remarked that "UCLA would not have achieved its current level of diversity without affirmative action." He observed that more than two-thirds of entering students in 1996 belonged to ethnic minorities, in contrast to 1980, when two-thirds of the freshmen were Caucasian. "We are a much greater university today," he concluded, "in large measure because we are more diverse."
Item: Also last year, after a federal court struck down the policy of the University of Texas Law School that reserved a portion of its entering class for blacks and Mexican-Americans, the law school petitioned the Supreme Court for review. It urged the Court to reassert the right of colleges and universities to give racial and ethnic preferences in order to promote diversity on their campuses.
Two Kinds of Diversity
The word "diversity," which echoes in every campus debate about affirmative action nowadays, joins ambiguity to ubiquity. On the one hand, the word has become simply a term of art that means the same thing as "minority and/or gender representation." When Chancellor Young spoke of UCLA's "current level of diversity," what he referred to is the two-thirds ethnic minority representation on his campus. When universities list their diversity policies, set up offices of diversity affairs, and measure their progress in achieving diversity, the word in every case is a synonym for minority/gender representation.
On the other hand, when the University of Texas Law School asked the Supreme Court to allow colleges and universities to take race and ethnicity into account in selecting students, it invoked a second sense of diversity as a justifying reason. It appealed to the idea that a university, given the kind of institution it is, needs a diverse faculty and student body. This second sense of diversity refers to the mix of viewpoints, opinions, talents, and experiences that enrich the university and facilitate its mission.
In a widely circulated report in 1996, Neil Rudenstine, president of Harvard University, justified Harvard's commitment to diversity in this second sense by invoking John Stuart Mill, who stressed the value of bringing "human beings in contact with persons dissimilar to themselves, and with modes of thought and action unlike those with which they are familiar." A diverse student body, argued Rudenstine, is as much an "educational resource" as a university's faculty, library, and laboratories. Consequently, Harvard takes great pains to assure that its admissions process results in such a student body.
Elizabeth Anderson, a philosopher at the University of Michigan, makes a complementary argument. It parallels Rudenstine's, but emphasizes epistemic rather than educational considerations. "A knowledge claim gains objectivity and warrant," Anderson insists, "to the degree that it is the product of exposure to the fullest range of criticisms and perspectives. . . . Universities [should] recruit students and faculty to ensure broad representation of people from all walks of life, so that the products of inquiry are open to critical scrutiny and influence from the widest range of viewpoints, and so that the subjects and direction of inquiry are responsive to the widest range of interests." But Anderson goes further than simply commending broad representation. She maintains that
[t]he internal knowledge-promoting aims of the university call for measures to promote equality of access by all groups in society to membership in its ranks. This is an argument for affirmative action in university admissions and faculty hiring that recognizes the positive contributions that members of oppressed groups can and do make to enhancing the objectivity of research. Equality of access through affirmative action policies is not, therefore, an external political goal that threatens to compromise the quality of research. It is a means to promote the objectivity of that research.
Here we have an explicit linkage drawn between diversity of viewpoints, opinions, talents, and experiences, on the one hand, and diversity of race and gender, on the other. The latter diversity, in turn, gets parsed by Anderson as "equality of access," and equality of access gets equated with affirmative action.
Is there such a tight linkage between the two kinds of diversity? Does pursuit of perspective-diversity provide the right sort of reason for pursuit of the racial/gender proportionality that preoccupies affirmative action? For all its facial plausibility, I think the way Anderson links diversity and affirmative action collapses under scrutiny. In order to see why, let's explore a bit further the educational/epistemic value of perspective-diversity. What sorts of considerations augment such diversity on campus and what are we prepared to do about them?
Valuable Perspectives
Consider, first, some of the different perspectives that may or may not be "broadly represented" in the university.
· Age. The way we look at the world varies considerably by age, so a broad representation of views should be sensitive to this dimension.
· Region. This was one of the first dimensions of "diversification" embraced at Harvard, as Rudenstine notes in his report. Though more so earlier in our history, even now people from different regions of the country possess somewhat different values and perspectives.
· Political affiliation. People divide deeply and sharply on matters of politics. Political views play a very important identity-defining role in individual lives. And taken together, political views profoundly affect the direction of collective life.
· Nation. This factor was prominent on the list of "diversities" of an earlier Harvard president. Certainly, foreigners bring to our universities customs, experiences, and viewpoints very far removed from our own.
· Occupation. Whether we labor with our hands or minds, with tools or concepts, on teams or individually, occupation affects our values and outlooks.
· Urban vs. rural upbringing. Life in big cities is very different from life on a farm or in a small town remote from large urban areas.
· Historical experience. People who have lived through economic collapse, war, natural disaster, mass migration, or political convulsion are marked by their experiences and often possess different values and perspectives than people who have not had such experiences.
· Religion. People's religious (and philosophical) views support their attitudes toward politics, education, community, justice, war, family, work, and the like. Furthermore, their religious (and philosophical) views underwrite the very meaning of their lives.
· Military service. The experience of being a soldier shapes people's outlooks in both predictable and unpredictable ways.
· Special aptitudes and skills. Being a pianist, painter, cook, chess master, competition swimmer, or skydiver counts in your favor from the point of view of diversity, since you exemplify a particular excellence and model a particular vocation that can inform and inspire others.
There are many more items we could add to this list, but let's stop here to make a few crucial observations. First, the items are not strictly ranked in importance. Second, they are desiderata, not imperatives. Third, they admit to trade-offs. Let me explain.
Trade-offs
Take the example of age. As it turns out, two very important age groups are not represented at all in the university -- the very young and the very old. There is virtually no one in the campus community under 16 or over 75. The student body, in particular, is heavily skewed toward the 18-30 range of ages. Now, these failures of representation derive from structural facts about the university: university studies demand a prior preparation unlikely to be possessed by anyone under 16; and faculty past 70 have retired and left the university. We are willing to live with these facts. Although two very important age-perspectives don't get represented on campus, we don't consider this failure of representation important enough to take special steps to cure it. Indeed, we even exacerbate it with some of our policies, such as encouraging early retirements among faculty and formally or informally kicking them all out the door at 70. We do this to make room for younger faculty.
This last observation points up the fact that we make trade-offs in realizing the desiderata on the list above -- both internal and external trade-offs. With respect to faculty age, for example, we trade off the gains of having faculty in their 70s and early 80s for the gains of having more faculty in their 30s and early 40s. Our exclusion of very young people from higher education involves a trade-off, as well. After all, we could get 12-year-olds in the university if we wanted, but to compensate for their greater immaturity we would have to change the university in many ways, such as lowering the intellectual level of many courses. We're not willing to do this. Moreover, we may not think that learning and research suffer all that much from the exclusion of 12-year-olds. Or 75-year-olds.
But, then, if research and education don't suffer much from these exclusions, perhaps age-perspectivity might be sacrificed in many other ways as well without degrading the overall quality of education or research. And if age-perspectivity can be sacrificed in many ways, perhaps other perspectivities can be sacrificed as well.
The "broad representation" of people and views demanded by good education and research may allow a lot of variation as to how the representation is composed and may even allow considerable omissions. This is certainly suggested by the way Rudenstine describes the admissions process at Harvard. What seems crucial is that each entering class be richly diverse, not that its diversity always reflect the same pattern. Thus, one year there may be more concert pianists and fewer rugby players in the class, the next year more rugby players and fewer pianists. Or perhaps both get shorted some years for more student government leaders or an unusually rich crop of Zen Buddhists. The point is, each of these mixes would be roughly as good as any other.
We find, then, that although a good student body or a good faculty will be diverse along many of the dimensions on the list above, exact mixes will vary. Nor can we expect "proportional representation" of diversities to be a useful standard. With respect to some items on the list, such a standard would be meaningless. (What would it mean to take in foreign students "proportionately"?) In other cases, we don't expect universities to undertake the efforts that would be required to attain proportional representation. We may regret that there are hardly any farm girls and farm boys among the professoriat at large, but no one proposes taking vigorous measures to alter this fact.
We do expect a university to be more concerned about some of the dimensions on the list than others, especially those -- like religion and politics -- most closely correlated with vital differences in value and opinion. A university probably will feel more concerned about a faculty overwhelmingly Protestant or Democratic than one overwhelmingly urban. It will regret the absence of certain minority political voices more than the absence of students from the Rocky Mountains. To make sure that certain political views get a hearing on campus, admissions officers might even admit young socialists or anarchists in greater proportion than they occur in the general population. This suggests that the critical factor in university admissions is not demographic proportionality but, instead, what Rudenstine refers to as "critical mass." There needs to be enough of a group, he says, so that its voice gets some attention on campus, its views get taken account of. In some instances, critical mass may require overrepresentation if the group to be represented is a very small, though intellectually important, fraction of society. In other instances, critical mass will be achieved even if the group is underrepresented in relation to its numbers in the larger society. Even if only 25 percent of the students on campus are Republicans rather than 35 percent, we don't worry very much that the Republican voice won't get an adequate hearing.
Race and Gender
We haven't yet taken up the "diversities" most talked about these days: gender and race. Add them to the list. What should we say about them? Unquestionably, many of our experiences and views are deeply affected by our race and gender. A university committed to "broadly representing" different views and experiences among its student body would take race and gender into account in its admissions process just in the same way it takes account of region, aptitudes and skills, political affiliation, religious views, and other factors. Or would it? A university concerned to foster the best environment for creating objective knowledge would take account of gender and race in choosing a faculty just in the same way it takes account of other factors. Or would it?
One thing is clear. Universities don't treat race and gender the way they treat the other dimensions of diversity we've been talking about. Race and gender are objects of affirmative action in the university, and affirmative action imposes an acute concern about proportionality. Affirmative action requires the university continually to ask itself, "Are women faculty being hired in proportion to their possession of the Ph.D.?" "Are African-American students being admitted in proportion to their numbers in the applicant pool?" and the like. Further, this concern about proportionality has an imperative quality, unlike the university's concerns about the desiderata on our initial diversity list. The university is rightly willing to make trade-offs among those desiderata. It acknowledges a very loose fit between any particular desideratum and good education and research; consequently, it is willing to forgo some kinds of diversity for others, or for the sake of particular educational missions. The university can decide it would rather have a lot more pianists than rugby players; it can decide to emphasize getting students from foreign countries rather than from different regions in the U.S. And so on. But affirmative action seems incompatible with this sort of approach. Universities can't say: "Well, we have different priorities; we'd rather have a lot of regional diversity than racial diversity," or "We've decided to emphasize political variety over gender proportionality."
Now, this difference in the way universities treat race and gender occurs either because (i) in regard to good education and research, race and gender are different from the other dimensions of diversity; or because (ii) the way universities deal with race and gender under affirmative action is premised upon a different ground altogether than good education and research.
Representation and Objectivity
To see what can be said in support of the first explanation, let's recall the views of Elizabeth Anderson that I set out earlier. Her argument forges a direct link between the university's educational/research purpose and affirmative action. We can reconstruct her argument as follows:
Premise 1: Objective knowledge is a product of the fullest range of perspectives. Broad representation, Anderson tells us, offsets bias. When the community of inquiry is broadly representative, individual biases are less damaging, since then no particular bias will unduly influence the community's acceptance of some theory or finding.
Premise 2: The historic absence from the academy of minorities and women has been particularly damaging to the goal of objective knowledge. Though regional or age bias, for example, may be problems, regional groups and adult age groups have not been excluded from the academy in the way that women and minorities have. Scholarship over many generations has built a huge edifice more or less oblivious to the perspectives women and minorities might bring to the table.
Premise 3: The academy, in furthering the goal of objective knowledge, ought to be especially concerned to include minorities and women. At our historical moment, it is more urgent to offset racial and gender bias than other kinds.
Conclusion: The goal of objective knowledge supports affirmative action in order to guarantee equality of access for minorities and women throughout the university.
Now, clearly, Premises 1-3 have considerable force. We cannot look back over the changes in scholarship in the last thirty years without conceding the significant changes wrought by the growth in the number of minority and women scholars. Still, we might wonder how tight the fit is between Premises 1-3 and Conclusion. Let's look at the places in this argument where dispute might arise.
First, Premises 2 and 3, which underwrite the urgency of including minorities and women in the academy, make a "lumpy" situation seem more uniformly smooth than it is. The premises, for example, don't make any distinctions among fields. Yet, some areas of study seem more affected by the inclusion of minorities and women than others. Moreover, the nature of the effects of inclusion varies.
For example, the impact of gender on writing history seems more profound than the impact of gender on doing astronomy. The writing of history has been transformed in many ways in the last thirty years. Women historians have driven home the fact that although women have always constituted half the human race, 99 percent of written history from time immemorial has recorded the deeds and thoughts of men, not women. Furthermore, "gender" has become an important concept through which to interpret historical events. Thus, even the deeds and thoughts of men can be given new and interesting interpretations when set against the backdrop of "gender." Finally, the new historical research leans less heavily on "official" documentary sources and more on "unofficial" documents as well as material artifacts, bringing not only the past of women more readily into view but that of marginalized classes and groups, as well.
On the other hand, women haven't had the same impact on astronomy or the other hard sciences. Of course, women haven't gone into these sciences in the same numbers they've gone into history. But what reason is there to think that even larger numbers would affect astronomy the way their numbers have affected history? Sandra Harding, in her book The Science Question in Feminism, argues that modern science is deeply "anthropocentric" (male-centered), suggesting that the entrance of women throughout science would provide a welcome corrective. But all of Harding's examples of "anthropocentrism" are drawn from the biological and social sciences rather than mathematics, physics, chemistry, or astronomy. Although she rightly says we cannot rule out a priori that mathematics and all the hard sciences are gendered, she can't point to any actual problematic, concept, theory, language, or method of mathematics, physics, chemistry, or astronomy as an example of such gendering.
This same observation holds true if we focus instead on racial and ethnic minorities. My point is not that science isn't gendered or racially biased; my point is that, right now, claims about the likely impact of more women and minorities in some fields of knowledge are contentious and far from settled. My point is that objectivity of knowledge as a goal doesn't obviously dictate a decision to get women and minorities uniformly and proportionately in all fields across the board. It may guide us, rather, toward getting more women and minorities -- even disproportionately more -- in some fields. Or would it?
From Perspectives to Groups
This last question arises out of a second problem about Premises 1-3, this time a problem with the move from "perspective" in Premise 1 to "group" in Premise 2. Let me stick with the example of gender and use Harding again. What Harding wants to pit against modern science is an "oppositional consciousness" -- i.e., feminism (or certain feminist theories). It is not women per se that will change science, but oppositional theories. The gendered nature of science won't be modified by adding more women scientists who already buy into the standard masculinist assumptions and research programs. Once we take this point to heart, it becomes harder to insist that the objectivity of knowledge demands we get more women into certain sciences, because we can't equate women and feminism.
The same is true regarding race and ethnicity. Although it is very popular these days to talk about "group perspectives," it is also dubious to talk this way. Women don't share a single perspective, even on matters of gender. Blacks don't share a single perspective, even on matters of race. When someone claims to represent a "group perspective," that perspective is mostly a construction of the claimer (and of like-minded persons), privileging certain propositions about society, justice, and the group's interests.
Now, this is not an objectionable or regrettable process. On the contrary, if knowledge advances through the pitting of views and outlooks against one another, the creation of oppositional perspectives is a vital activity. Moreover, these oppositional "group" perspectives are never created out of whole cloth. They will obviously reflect (and reciprocally influence) the views of many within the respective groups. But it is the perspectives that are crucial, given the goal of objective knowledge. Since there isn't a tight connection between a particular "group perspective" and members of that group, the goal of objective knowledge doesn't speak as unequivocally in favor of proportional representation of group members -- the concern of affirmative action -- as it does in favor of critical representation of perspectives.
The Primary Motive
Suppose, then, we were persuaded that greater inclusion of women and minorities actually wouldn't make a real difference to astronomy. Would we, then, withdraw our support from affirmative action efforts to get more women and minorities into astronomy, including such efforts as creating special fellowships reserved especially for members of these groups?
No, because our primary motive, in the first place, isn't to get women and minorities into the hard sciences for the sake of these sciences but for the sake of women and minorities (and their opportunities). From an affirmative action perspective, the reason we want to lodge a critical mass of women and minorities in the hard sciences is our belief that, absent a history of exclusion, women and minorities would have flourished in these fields, and we want to change the momentum of past exclusion so that women and minorities seek and find opportunities there in the future.
This primary impulse is plain enough, I believe, if we draw the following parallel. From an affirmative action perspective, we are concerned about proportional representation of women and minorities in various academic disciplines in exactly the same way we are concerned about equality of pay for comparable women and men, or comparable minorities and whites. Concern about equality of pay has nothing to do with the objectivity of knowledge or good learning and everything to do with antidiscrimination and fairness. So, too, with the concern about representative numbers of women and minorities in various fields of study and in the university as a whole: it has to do with antidiscrimination and fairness. That is why the concern has an imperative quality, and why trade-offs are out of place.
The educational/epistemic argument for diversity is not wrongheaded or unpersuasive. It certainly might sometimes justify the university taking race and gender into account in selecting students and faculty (to the extent such taking into account is permitted by law). But it doesn't justify taking race and gender into account in the way affirmative action demands they be taken into account. We can't fully and properly justify the affirmative action concerns of universities by starting from premises that talk only about good education and objective knowledge.