Professor X teaches the non-required "Sociobiology and Theology," in which he repeatedly asserts and presents an extended defense of the thesis that all women are vile whores by nature (all Ws are H), which he takes to be a sociobiological update of some medieval theological doctrines. His behavior in class and outside of class is (otherwise) remarkably nonsexist. Some, perhaps all, of the thirty students are women, some (perhaps all) of whom know in advance that Professor X holds controversial views. (Will class size, gender ratio or degree of foreknowledge matter? Let's see.)
If X were to say to Ms. A, "you are a vile whore by nature," and to say this several times during every class, he'd be targeting her and harassing her; if he also says the same thing to Ms. B as often, the same holds. But he does not say, "you are a vile whore by nature" to anyone. (Nor does he use names or individuating descriptions in place of the demonstrative "you" - "Ms. A is H," "the youngest woman in this room now is H.") But he does assert repeatedly, with enthusiasm, "all Ws are H." Indeed, the whole point of the course is to establish this claim. Not every student is skilled at deductive inference, but some correct inferences are so obvious that the disposition to make them readily is part of any legal definition of 'reasonable person'. Here's one such inference so salient that its obviousness is not likely to be missed:
X believes that all Ws are H
X knows that I'm a W
Therefore X believes that I'm H
-thought by a student in the class who knows that she herself is a W. So, any reasonable W in the class will be vividly aware of the fact that X holds her to be H. There is no significant difference between X's saying to A (and to B, and to ...) "you're H" and his asserting in class, "all Ws are H." If the former constitutes harassment, then so does the latter.
This line of reasoning loses little force if we replace "all Ws are H" with a variety of weakenings, e.g., "all Ws are whores," "all white Ws are H," "more likely than not, every W is H" or "it's highly likely that almost every W is H." Nor does it make much difference if X introduces the term, "blarg" as an abbreviation for his long, complex characterization of what it is to be a vile whore by nature, never uses the words "vile" or "whore" or similarly obviously offensive terms, and asserts instead, "all Ws are blarg."
It is a psycholinguistic fact that demonstratives (e.g. "you") tend to target more sharply than proper names ("David Austin"), individuating (definite) descriptions ("the one and only person now in the last seat in row 17"), or universal statements ("every person now sitting in the last seat in row 17"). But in the case described, these typical differences are effectively overridden. "No, X didn't say the words 'you are H' to me, but he might as well have." Given the extreme obviousness of the required deductive inference, and the reasoner's knowledge, the latter claim seems correct. (More cuttingly: denying the latter claim is 'legalistic' in the worst sense.) The situation as described is one in which the usual difference between speaking to someone and speaking about them is effectively erased. Why should it matter, for purposes of determining if harassment has occurred, whether the instructor asserts a universal or its instances, given the painfully obvious deductive connection in the sort of case described? Since the repeated assertion of an instance constitutes harassment of the instanced individual, the speaker can't be let off the hook just because he takes care to 'target everyone at once' (exculpatory points for linguistic efficiency?!). Indeed, why isn't it worse to hit everyone with the speech act?
One of the reasons I used symbolic predicates in the example was to encourage thinking of them as schematic: they could stand in for many others. So the very same dilemma arises for any possible form of harassment; current law provides characterizations of at least six other sorts of harassment, corresponding to the broader categories of discrimination based on race, color, national or ethnic origin, religion, age or disability. If those in a powerful group can be harassed ["all men are (by nature) rapists," "all Christians are (essentially) sexist," "all those who are not baptized are destined for Hell," etc.] - and the possibility is granted by the courts - then the extent of the dilemma is even greater. And, of course, there may be types of harassment that are not yet specified in law (e.g., types based on sexual orientation). This line of reasoning therefore seems to present a dilemma of considerable extent.
The natural reaction is to say that a line must be drawn somewhere to prevent slipping all the way down the slope, but the argument suggests that we've already slid to the bottom once we've adopted a targeting standard. To add grease to an already well-lubricated incline, consider the following series and question:
youa are H
youa are H and youb are H
youa and youb are H
all of you are H
all Ws are H
Where and for what reasons consistent with the motivations for a targeting standard does one draw the line in this series?
Of course, tracking probable psychological effect of speech will help with cases where the relevant inferences are less likely to be made than in the case described - so while X is presenting a complex argument to show that all Ws are H, but has not yet asserted the conclusion, he's not harassing; but once he gets to (or perhaps close enough to) the conclusion, he's harassed students. ["Oh no! X is talking about/to me!" the (reasonable) female students think, once X actually gets (close) to his conclusion.]
If the argument has its intended force, then it offers those who accept a targeting standard one of two alternatives concerning the assertion of objectionable views: be unintelligible or silent, on the one hand, or, abandon the standard and speak very freely indeed, on the other.
There are many possible responses to this line of reasoning, but one that seems promising to me admits its force and seeks to compensate through informed consent. A student might enroll in X's course for many reasons (and with different sorts of foreknowledge); because, e.g., (a) the course fits her schedule; (b) the course fulfills a distribution requirement; (c) she's heard that the instructor has controversial views; (d) she's heard that the instructor has controversial views about women; (e) she's heard that the instructor argues that all Ws are H. Students can meet some or all of (a) - (d) without meeting (e), and even those who meet (e) could be surprised by powerful psychological effects of X's repeated, argument-backed assertion that all Ws are H. How much forewarning is necessary for informed consent that reduces harm and prevents harassment? ("You really did know what you were getting into when you signed up for the course.") Without a clear answer, this line of response remains no more than promising.
My 'oversimplification' warning light is flashing, but I don't yet see how to shut it off.
See Andrea Dworkin and Catharine A. MacKinnon, Pornography and Civil Rights: A New Day for Women's Equality, 36, model ordinance definition of "pornography," clause (vii) "women are presented as whores by nature." The qualifier "vile" is used by Dworkin.
For a pertinent reading of some Catholic theology, see: Jeanne L. Schroeder, "Feminism Historicized: Medieval Misogynist Stereotypes in Contemporary Feminist Jurisprudence," 75 Iowa Law Review 1135, 1136-47, 1189 (1990). There is a widespread conviction that protection for freedom of speech should be at its strongest in the classroom, so the example involves the classroom to address that conviction at its strongest. (I've been told that a member of the Anthropology faculty at UC/Berkeley teaches a large lecture course, without discussion sections and with very little opportunity for in-class discussion, in which he argues that African Americans are in some ways inferior to whites.) Examples similar to mine could, of course, be constructed for the workplace.
"After all," he remarks years later, "some people can rise above their natures." He makes no such remark during the time when the course is offered. X does comment on Ms. A's highly critical course paper, "I do, of course, disagree with you and find your reasoning subtly flawed, but your counterarguments are ingenious and insightful and your research extremely thorough - clearly an 'A' paper. Excellent work! And, may I add, it's been a pleasure 'sparring' with you in class. I've learned from you about how to clarify and fortify my position." On Mr. Z's course paper, X comments, "While I agree with your conclusion that all Ws are H, the arguments that you give to support it are not adequate to the task, and your research went no further than course texts and lectures. It even appears that you did not understand some obvious objections to my thesis, which I explained how to overcome in lecture, and your arguments are open to those very objections. Grade: 'D'." Students A through Z and beyond agree that Professor X has been very fair and respectful, except for his repeated assertion that all Ws are H.
It seems to be a fundamental presupposition of harassment law that people typically know when they are of a kind covered by that law - so women typically know that they themselves are women, the disabled typically know that they themselves are disabled, etc. Since not all relevant properties are as readily detectable as W is, the story would need to be augmented in other cases to secure the truth of the second premise. The legal standard would presumably be that a reasonable speaker would be expected to know that audience members have the relevant property.
There are, of course, many differences between serial, parallel and universal assertive speech acts. But how much time these acts take does not seem to matter here in making the relevant legal determinations. Would asserting, "All judges are idiots," in court be likely to result in a citation for contempt of court? ("But your honor, I didn't mention you!")
I take it that this is the point of the story of the 'educated' anti-Semite, in Nicholas Wolfson, "Free Speech Theory and Hateful Words," University of Cincinnati Law Review v60 (1991) 1ff. See also, Nicholas Wolfson, Hate Speech, Sex Speech, Free Speech (Praeger, 1997) 50-51. Condemnation delivered in elite language and backed by sophisticated arguments can be even more devastating to those condemned than a common, 'low language' insult: often, familiarity breeds defense.
For more on these issues, see the discussions of 'overlapping contexts' and 'context-hopping arguments' in: David F. Austin, What's the Meaning of "This"? A Puzzle about Demonstrative Belief (Cornell University Press, 1990), Appendix; and Mark Richard, Propositional attitudes: an essay on thoughts and how we ascribe them (Cambridge University Press, 1990).
"If the pervasiveness of an abuse makes it nonactionable, no inequality sufficiently institutionalized to merit a law against it would be actionable." Catharine MacKinnon, Feminism Unmodified (Harvard University Press, 1987) 115.
In the Brooklyn of my youth, "all youse is H;" in the South of my present, "y'all're H."
A visual metaphor might be helpful. Imagine that in ordinary targeting, the speaker tugs sharply on a string tied to the target's right index finger; in serial targeting, the string is pulled repeatedly; in parallel targeting, more than one string is tugged at a time, perhaps by being tied to one long dowel or held together in the instructor's hand; in universal targeting, many strings are the interwoven strands of a single rope. To remind oneself that harm is the issue, suppose that the strings are replaced by wires and that current flow supplants the tugs. To construct more extreme versions, one might imagine that the wires are attached to more sensitive parts of the body.
This is analogous to quoting hate speech when no endorsement is made or implied by the act of quoting itself. See my "(Sexual) Quotation without (Sexual) Harassment? Educational Use of Pornography in the University Classroom," in James Elias et al, eds., Porn 101: Eroticism, Pornography and the First Amendment (Prometheus Books, 1999).
In correspondence via the CLSPEECH-L list, Eugene Volokh has suggested that saying instructors should not target individuals may be a better way of bringing about the desired classroom behavior than saying instructors should not assert obnoxious universals, since the latter may be more likely to chill properly protected speech. Such pragmatic considerations are, of course, very important in rational policy planning. It is nevertheless important to know what standard we are committed to before deciding how best to bring about the behavior desired.
The latter choice would presumably be coupled with the claim that there are already adequate, even if not often used, barriers to psychologically damaging assertion, and that harassment law is therefore superfluous. Such remedies, through defamation law, for example, would also be available in cases where the particular type of person described is not specified in the law.
Thomas Grey and George Panichas have independently suggested that the instructor must intend to offend if the universally targeted speech acts are to constitute actionable harassment. But, of course, the same ought to apply to ordinary, nonuniversal targeting. Nor does it leave room for negligent speech. It may be that when intent is missing, a 'more speech' remedy is the kind best sought; if students have given informed consent, 'more speech' is more likely to occur. But consent is limited in its protective and enabling potential; on this general matter, see Robin West, "The Harms of Consensual Sex," American Philosophical Association Joint Newsletter on Feminism and Philosophy and Philosophy and Law n94:2 (Spring 1995) 52-55. See also Timothy Shiell, Campus Hate Speech on Trial (University of Kansas Press, 1998), esp. 160 on targeting.
Since actual psychological damage is not legally necessary for harassment to have occurred (Harris v. Forklift Systems, Inc.), reducing harm and cancelling harassment are to some extent independent matters. We must also take care not to permit this remedy to degenerate into a defense offered by Jacksonville Shipyards, Inc. against Lois Robinson's charge (Robinson v. Jacksonville Shipyards Inc.) of sexual harassment: "You knew in advance that shipyards are saturated with sexism, so you were adequately forewarned before you took the job here."
Thanks to George Panichas for asking the question that prompted these thoughts, and to Eugene Volokh, Thomas Grey, Barbara Levenbook, Stanley Fish, Barbara Herrnstein Smith and Randy Carter for helpful comments on earlier versions of this note.
This page was last revised on