American Council of Learned Societies
Occasional Paper No. 22

The Limits of Expression
in American Intellectual Life

Kathryn Abrams

Art, Transgression, Shock, and the First Amendment
W.B. Carnochan

Truth or Consequences: Putting Limits on Limits
Henry Louis Gates, Jr.

Free Expression in the Academy:
Three Hard Cases That Test Easy Assumptions

Robert M. O’Neil

Creeping Absolutism and Moral Impoverishment:
The Case for Limits on Free Expression

Kathryn Abrams
Cornell University

In my comments today, I want to do two things: First, explain why we need limits on free expression in intellectual life; and second, talk about how we should argue for them, both as a general matter and in at least one individual case. I offer my comments from the perspective of a legal scholar looking at the law, and seeing how its principles are mirrored and amplified in social institutions, such as institutions of higher learning. My argument is that the absolutist tendencies of the First Amendment legal regime have contributed to a climate where expression is overprotected, and members of the intellectual community are deterred from thinking systematically about how to reconcile expression with other norms — for example, respect for and recognition of politically marginalized groups. Our task should be to identify and transform the assumptions that have produced this tendency in law, and its counterpart in liberal theory. This transformation should produce forms of decision-making that are contextual and multi-factoral, but it won’t entail the descent into analytic and normative chaos that many of its opponents suggest.

I start with the proposition that the current First Amendment regime has absolutist tendencies. By this I don’t mean to say that First Amendment doctrine requires absolute protection for speech. As a practical matter, several categories of limitations on expression have been approved by the courts. There are certain types of expression — commercial speech — which receive more limited protection. There are areas — obscenity and defamation — where expression is limited to satisfy other values or interests. There are even circumstances where it is possible to argue — as my colleague Steve Shiffrin, for example, has argued1 — that the courts have implemented restrictions on expression that are not content neutral. But I would argue that the current First Amendment regime tends in the direction of absolute protection for expression. There are two reasons for this. First, the rhetoric, if not the practice of the courts tends toward absolutism. Departures from absolutism are treated as sui generis exceptions or hidden beneath subterfuges that preserve the more absolute statements of principle. What is left on the face of the courts’ pronouncements are a series of descriptive and normative principles that make arguing in favor of limited, contextually-based restrictions very difficult. Second, because of their resonance with certain precepts of liberalism, these principles have been taken up by broad segments of the public in ways that extend the scope of their application beyond that required by legal doctrine. The result, I would argue, has been overprotection of expression and an impoverishment of moral discourse about how to respect and recognize particularity in intellectual life.

In what ways do the assumptions or principles of First Amendment doctrine tend toward absolute protection of expression? First, the framework for defining speech and assessing the harms that flow from it is rigid and dichotomous. What we have in a given case is either speech or (in a very few contexts) action; this speech either produces imminent, specifically targeted harm (in which case the law can intervene) or it produces no harm of which the law can take cognizance. This dichotomizing has proved remarkably resistant to the ambiguities presented by expressive practices such as pornography, or to harms distinct from “yelling fire in a crowded theatre” — harms that go to dignity or the formation of self-conception.

Second, First Amendment law declares that governmental decisions regarding expression must be neutral as between groups. This is embodied in the general ban on content-based restrictions on speech. As the recent decision on the St. Paul “hate speech” ordinance makes clear, this principle is subject to numerous exceptions. But the face that the First Amendment has turned to marginalized groups — both the varied groups covered in the St. Paul case and the women protected by Indianapolis’s anti-pornography ordinance — has reflected agnosticism about the content of speech, and studied disinterest in the political and material circumstances of the contending parties.

The most frequently cited reason for this position goes beyond content agnosticism, or the tendency of the courts to view rights-bearers as individuals without social location; it constitutes the third factor pointing First Amendment doctrine in the direction of absolutism. The courts suggest that any break with this posture of governmental neutrality towards groups threatens uncabinable chaos. In the legal realm this argument is known as the “slippery slope” — available in a range of contexts but never more vigorous in its hold than here. The slippery slope arises from radical doubt about the possibility of distinguishing one group’s claim from the next. If we prevent the Nazis from marching on Skokie, there’ll be no way to distinguish them from the next group that comes along, and before we know it there’ll be no protection of speech. This form of reasoning is strongly anti-contextual. It arises from an inexperience with — and therefore an anxiety about — claims about the specificity of context or group affiliation, which is born of the constitutional habit of considering rights-bearers as unaffiliated individuals. I should add that group-related neutrality and the invocation of the slippery slope are not unique to First Amendment law. They have become a depressing staple in areas that have nothing to do with expression, and a lot to do with material inequalities relating to, for example, race. Both of these premises play a prominent role in the Court’s rejection of race-conscious remedies — both in the academic area and in the area of government set-asides.

So these are the principles or premises that move First Amendment doctrine in the direction of absolutism. Now, however, the plot thickens: This absolutism has been amplified by non-legal actors making decisions in academic settings. The basic problem is that the principles outlined above are applied in settings where the First Amendment does not require their application. I don’t simply mean they’ve been applied by private universities: the public/private distinction is not something I want to emphasize in any unqualified way. First Amendment absolutism is being applied in contexts where the object of protection is not expression per se, but rather choice in the establishment of a curriculum, or of a topic for a university forum. More importantly, First Amendment rhetoric and principles are being applied in contexts where incursions on expression are accomplished not by legal restrictions or sanctions, but by protests, condemnations or requests for inclusion. Last spring, while I was teaching at Harvard Law School, some members of the Law Review published a vicious parody of an article by a recently-murdered feminist scholar. When the campus erupted in protests and condemnations, one well-known law faculty member derided this outcome by describing it as a left-wing or “PC” attempt to restrict the free expression of unpopular sentiments by other students.

As my reference to the comments of a law professor should make clear, my point is not that laypeople are misunderstanding the scope of the First Amendment. The rhetoric or principles of First Amendment doctrine are easily amenable to expansion beyond governmental decision-makers, conventional expression, or official sanctions imposed on speech, because they resonate so strongly with certain precepts of liberalism. The principle that rules should not treat one group differently from another resonates with the liberal tenet that our claim to equality and recognition arises from our shared human nature and our potential, rather than the qualities that differentiate us or our current position in the social structure. The slippery slope resonates with liberal theory as well, particularly as relates to questions of power. The habit of considering people as undifferentiated residuaries of potential, and seeking equality of opportunity rather than equality of result, has made us uneasy about the possibility of containing arguments that invoke inequalities in power. A good example of this is the discomfort even many progressives feel about using arguments related to power inequalities in shaping the curriculum. The frequently advanced argument that groups advocating multiculturalism seek to reduce all curricular discussions to disputes about power (a claim that seems facially incorrect) reflects, instead, a fear that this reduction will be their inevitable effect: that it will be impossible to contain power arguments, and to reconcile them with aesthetic and other claims once they are admitted to the discussion.

So the subtle expansion of the domain of free expression beyond the requirements of the First Amendment is one problem, and it is a serious one. When protests and condemnation are grouped with prior restraints and penalties on speech, people may be deterred from thinking more systematically about how to object to speech short of penalties — how to signal through words what Amy Gutmann has called the difference between speech “that is worthy of respect” and speech that “must be tolerated.” There is a second, related problem, which is potentially more serious. As the requirements for protecting free speech interests become more and more extensive, they begin to occlude consideration of other values that should shape decisions about potentially injurious expression. This is certainly true for those within universities, for example, who seek to discourage hate speech or sexual harassment. The time that must be spent satisfying real and imagined First Amendment difficulties often restricts the opportunity to explore equally important questions. In January I met with a subcommittee of the American Association of Law Schools to recommend a harassment code to its member schools. I can report that we spent one quarter of our time talking about how to satisfy potential free speech objections — twice as much time as we spent talking about the nature of the violation or what groups should be covered. How to weigh the claims of different groups seeking coverage, for example, is rarely a straightforward matter, and we are unlikely to develop the criteria for doing it when we have neither the time nor the incentive to consider the circumstances of particular groups.

Perhaps more importantly, potential speakers may begin to develop a similarly narrow habit of mind, permitting an inquiry into the likelihood of their First Amendment protection to constitute the sum total of their deliberation on whether to undertake potentially injurious speech. In this context, it’s worth mentioning an incident at another law school, where a student circulated a list of mock law review articles, one of which was ascribed to a student known to be feminist, and concerned her enthusiasm for, let us say, emasculatory sexual practices. I mention this not as the ultimate example of man’s inhumanity to woman; it was hurtful and it shouldn’t have happened, but I have certainly heard of more injurious incidents. I mention it because of what it reveals about the judgment of the perpetrator. When questioned by the Dean of Students about why he had circulated such a thing, the student quickly dropped his half-hearted suggestion that he didn’t think the woman would mind it, and asserted that circulating it was within his First Amendment rights. When our legal and cultural reluctance to place other values in the balance with expression begins replicating itself in the moral deliberations of potential speakers, I think we have a real problem.

How can we reintegrate other norms into deliberations about expression? We should focus on the absolutist norms of the First Amendment regime, and demonstrate that their analytic hold, both in law and in liberal theory, is not nearly so secure as the face of doctrine suggests.

The insistence on a temporally imminent, physically oriented and specifically targeted harm hangs on in First Amendment law, but it hangs by a thread. Increasingly persuasive within legal academic circles are notions of temporally extended, group-based oppressive harms, such as those described by Catherine MacKinnon in the area of pornography,2 or by Charles Lawrence in the area of racial hate speech. Some elements of this vision, including a more extended temporal frame for “imminence,” have already been accepted in other areas of law such as battered women’s self-defense. Developments in political and social theory give an assist here as well. Liberal communitarian and post-structural social theories which emphasize the social construction of the self-help explain why expression which distorts or undermines one’s self-conception can be a serious social problem. Continuing efforts to inject these notions into First Amendment calculus should ultimately be successful at the legal level. But elaborating them outside the context of any legal restrictions on speech can also inform the moral imagination of potential speakers and offer them an alternate way to reflect on injurious speech.

The next challenge lies in contesting the requirement of governmental neutrality among groups. This effort has the largest potential pay-off, because it could affect areas such as race-conscious remedies as well. The strategy here should be to demonstrate that it is plausible, even in the realm of constitutional law, to make government take account of group particularity or inequality. The critique of objectivity, increasingly prominent in law and other fields, has helped to demonstrate that “neutrality” is not always neutral, and may be a means of perpetuating current power inequalities. There have also been more specific applications. Jack Balkin, for example, has pointed out that the legal realists in the twenties and thirties were able successfully to challenge governmental neutrality with respect to another critical constitutional right: the right to contract.3 By showing that terms like “liberty,” “will,” and “contract” didn’t represent some determinate state of affairs, and that governmental action contributed to inequalities in bargaining power that defined the extent of “will” or “liberty,” the legal realists ultimately persuaded the courts that insistence on formal equality in contracting should give way to attention to context, including economic inequalities. A similar emphasis on the role of the government in effecting political power or marginality, in creating the captivity of certain audiences or facilitating their ability to respond to harmful speech with more speech may succeed here are well. In addition, new arguments in the realm of political theory may problematize the conventional liberal insistence that democratic governments must recognize their citizens only as equals in human nature and potential. Charles Taylor has argued, for example, that a dissonant demand that governments recognize citizens for their cultural or material particularity has important roots in many of the same texts that ground our usual emphasis on formal equality.

Finally, we have to challenge the premise that breaching neutrality, by permitting decision-makers to notice particularity or acknowledge power inequalities, will lead to chaos. Part of this task involves demonstrating that multi-factoral, contextual decision-making has been part of legal decision-making since English common law. But since this slippery slope argument applies to moral and political decision-making as well, transformation requires not just legal precedent but practical demonstration. If we can develop flexible yet articulable criteria for more contextual decision-making, we can formulate new answers and combat old assumptions at the same time. To conclude my remarks I will examine one set of criteria for balancing interests in expression against harm to politically oppressed groups, one that might be used as a prototype for other contexts as well.

These criteria come from the area of sexual harassment — an issue that has emerged in, but is hardly specific to, academic contexts. Sexual harassment is an interesting area in the discussion of expressive interests, because, unlike the area of hate speech, the understanding of the harm emerged and was legally recognized before anyone realized that enforcement might implicate the speech interests of harassers. By that time the harms to women were sufficiently well established that only a few unqualified First Amendment advocates argued that free speech claims should prevail in an unqualified way. Instead most lawyers began working toward a system of balancing that provided for restriction of expressive interests in a number of delineated contexts. These efforts to strike a balance have focused on a number of factors, including: 1) the nature of the environment; 2) the captivity of the victims — both within those areas of the workplace where harassing literature was posted, and within this particular workplace in general; and 3) the nature of the harassment — including its severity, its pervasiveness, and the extent to which it was addressed to, or targeted, the particular victims. The social or political characteristics of the victim were not a criterion in this context, as protected groups in the area of sexual harassment are established by statute. But they might be in other contexts.

These criteria do three things. First, they permit decision-makers to recognize contexts, such as educational institutions, where expression has a particularly important status. Second, they focus on characteristics such as “captivity” (the “captive audience”) or “targeting” (used in “fighting words”), which exist as narrow exceptions within First Amendment doctrine, yet they offer them more expansive interpretation. Third, they permit decision-makers to focus on and develop distinctions regarding important issues such as the circumstances of the victim and the nature of the harm, which can’t be fully analyzed under the current regime.

Such criteria pave the way to a system where a speech interest will be neither an icon nor a ground for avoiding moral judgment, but one factor to be placed in the balance with other, socially valued goals.


1. See Shiffrin, “Comment on RAV v. City of Saint Paul ” (unpublished draft on file with author); Shiffrin, “The First Amendment and the Meaning of America” (unpublished draft on file with author). [Return to text]

2. See generally, C. MacKinnon, Feminism Unmodified: Discourses on Life and Law (1987). [Return to text]

3. Balkin, “Some Realism About Pluralism: Legal Realist Approaches to the First Amendment,” 1990 Duke Law Journal 375. [Return to text]