1. Austin, J.L., How to do Things with Words (2nd ed. 1962).Google Scholar

2. Writers including MacKinnon, Langton, and Hornsby (see infra notes 3–5). The claim that “free speech includes free illocution” is our own summary of one of the ideas common to these three authors (though we do not say that MacKinnon would wish to formulate the claim in this way).

3. This way of putting the points is in Langton, R., Speech Acts and Unspeakable Acts, 22 Phil. & Pub. Aff. 293 (1993).Google Scholar Parts are reprinted in The Problem of Pornography (Dwyer, S. ed., 1995).Google Scholar

4. This way of putting the point is in Hornsby, J., Speech Acts and Pornography Women's Phil. Rev., 11 1993, at 38.Google Scholar Reprinted with a “Postscript” in Dwyer, S. ed., supra note 3.Google Scholar

5. See MacKinnon, C., Francis Biddle's Sister, in Feminism Unmodified (1987)Google Scholar and MacKinnon, C., Only Words (1993).Google Scholar

6. Jacobson, D., Freedom of Speech Acts? A Response to Langton, 24 Phil. & Pub. Aff. 64 (1995).CrossRefGoogle Scholar

7. Dworkin, R., Liberty and Pornography, N.Y. Review of Books, 08 15, 1991.Google Scholar (A nearly indistinguishable paper, on which this is “based,” is Two Concepts of Liberty, in Isaiah Berlin: A Celebration (, E. & Margalit, A. eds., 1991Google Scholar).) The assumption was exposed in Hornsby, , supraGoogle Scholar note 4. It is present in Dworkin's later work, which we consider briefly in part 4 below.

8. Whereas the whole of Mill, J.S., On Liberty (1859)Google Scholar can be read as a defense (inter alia) of free expression, Chapter 2, Of the Liberty of Thought and Expression, can only be read as a defence of free speech proper. See Schauer, F., Free Speech: A Philosophical Enquiry (1982).Google Scholar

9. Note that the shape of various debates is affected by whether or not speech is taken to encompass expression of all sorts. If pornographers' “expressions” are on a par with the speech of women, which some feminists want to defend, then the feminists' arguments will suggest that pornography provides an example of a clash of a certain right: the right of pornographers to free speech-and-expression is set against the right of women to free speech-and-expression. If, on the other hand, there is a defensible right to free speech as such, and pornographic material by and large is not speech, then the feminist who contemplates empowering women to put a halt to the publication or distribution of material whose dissemination interferes with women's freedom of speech will have to set one's group right—i.e., women's right to free speech—against another group's different right—i.e., pornographers' right to free expression. (We describe the feminist here as “contemplating empowering women…” in order to draw attention to the fact that those who question pornographers' license do not necessarily advocate pornography's censorship. If Catharine MacKinnon and Andrea Dworkin had eventually been successful in the courts, the legislation their campaigns would have introduced would have enabled individual citizens [of States governed by the ordinances to which they brought amendments] to sue in civil court to put a future ban on the publication of specified sexually explicit material and to collect damages for the harm that its publication had done Their underlying idea was not that pornographers have no right to publish necessarily, but that women need to be empowered to defend themselves against being silenced.)

10. Notice that any actual piece of speech can typically be classified under all three heads: one can ask about a single piece of speech what locutionary kind it is of and what illocutionary kind it is of and what perlocutionary kind it is of. Whether a piece of speech is protected depends upon which acts it is used to perform. As our concern here is with only a limited range of speech acts (and there would be very much more to be said, if, for instance, we were to consider the use of nonindicative sentences), our overall claim that free speech includes free illocution might be put (allowing for our limited concerns) by saying that illocutionary acts may be such that pieces of speech merit protection in virtue of being performances of them.

11. Much confusion has resulted from Austin starting from a particular notion of a performative— “Mark I performatives” as they are called by Warnock, G.J., Some Types of Performative UtterancesGoogle Scholar, in Essays on J.L. Austin (Berlin, I. et al. eds., 1977Google Scholar). Mark I examples are all of formulae used in institutionalized procedures. This starting point led Austin to associate the idea of convention with the idea of illocution generally. See further Hornsby, J., Things Done with WordsGoogle Scholar, in Language Duty and Value (Moravcsik, J. et al. eds., 1988).Google Scholar

12. We believe that in the order of definition the illocutionary should come first (and that the definition of illocutionary must not be focused exclusively on Mark I performatives, see note 11). Arguably, “repercussions of communicating” cannot be used on its own to define the perlocutionary, because the immediate upshot of the use of a Mark I performative is not a merely communicative one. But the upshots of Mark I performatives, although they may be thought to be further consequences of speaking, can nonetheless be distinguished from perlocutionary consequences: they are so distinguished when they are seen as consequences specifically for institutions within which there are performances working through reciprocity.

13. In Hornsby, J., Illocution and Its SignificanceGoogle Scholar, in Foundations of Speech Act Theory (Tsohatzidis, S.L. ed., 1994Google Scholar). This provides a fuller account of the speech-act theoretical material relevant here.

14. The idea that to “fully succeed” in certain acts, recognition on a hearer's part is necessary is in Searle, J.R., Speech Acts: An Essay in the Philosophy of Language 47 (1969).Google Scholar In the present paper, we use “to do the illocutionary act of X-ing” (and variants) to mean what Searle meant by “fully succeed in X-ing.” Langton herself sometimes used just “to X” (see infra, note 18).

15. These examples are discussed in Langton, supra note 3, where the argument of this and the following part is given in more detail. The example of the “low type” revolutionary is from Austin, , supra note 1, at 239–40.Google Scholar The example of the actor who shouts “Fire” is from Davidson, D., Communication and Convention, in Inquiries into Truth and Interpretation 269 (1984).Google Scholar

16. See MacKinnon, 1993, supra note 5, at 63–8.Google Scholar

17. Supra note 6, at 78.

18. Austin, , supra note 1, at 115.Google ScholarJacobson, , supra note 6, at 73.Google Scholar Jacobson thinks that Austin “overstates things.” Well, we noted supra (where note 14 is flagged) that verbs for illocutionary acts, like “warn,” may be thought to have application even where the hearer does not take the speaker to be warning. We advocated putting Austin's point by saying that no illocutionary act of warning is performed unless there is uptake. Put that way, the danger of overstatement goes away. Interpreting her remarks in their context, Langton, when she says that the speaker does not refuse, can easily be read as meaning that there is no illocutionary act of refusal (in the sense we give that here).

19. Jacobson, , supra note 6, at 72 and 73.Google Scholar

20. Id. at 76.

21. Id. at 74.

22. Id. In Strawson, P.F., Intention and Convention in Speech Acts, 73 Phil. Rev. 439 (1964)CrossRefGoogle Scholar, it is said that the aim, not the achievement, of securing uptake is essential to the illocutionary act (We think that Strawson was registering appreciation of a point we acknowledged in the text supra—about what appear to be necessary conditions for the application of words like “warn” or “refuse.”) Jacobson says that he will use only a weaker claim than Strawson's, that “at most the aim of securing uptake… is required” (id. at 74). Because the issue is whether there is any further requirement, it is not clear to us how Jacobson's claim can be relevantly weaker than Strawson's.

23. Id. at 76.

24. What Jacobson needs is a case where Bill does not recognize that Sally intends to invite him. There could be a story in which Bill believes that the invitation was sent to him by mistake, and Austin then might well say that the invitation misfired and she failed to invite him. (More cautiously: one might well say that she failed to perform the illocutionary act of inviting; see notes 14 and 18.)

25. Jacobson wants to preserve something of the Austinian view of uptake as required for illocution, but what he himself requires is counterfactual uptake by a “competent hearer”: “[W] hat a competent auditor would take as an unambiguous warning is one” (id. at 78). The most obvious way to understand “competent hearer” here is as someone who would recognize the intended illocution. But so understood, Jacobson's apparent innovation appears to collapse into the Strawsonian view on which he claimed not to rely (see supra note 22).

26. Id. at 77: the italics are Jacobson's.

27. See supra note 18.

28. Within the mens rea component of many legal definitions of rape, absence of consent on the woman's part is not sufficient for rape; there is a requirement also, e.g., that the man not be reckless in his belief that the woman consented. This ensures that, in a legal context, the fact of a man's taking a woman not to refuse could be relevant to the question whether he is to be found guilty of raping her. But this legal point is independent of questions about the conditions for illocutionary acts. Even the most satisfactory legal definitions of rape may fail to accord with an intuitive understanding of it, because the legal definitions will always be constructed to register burden of proof.

29. It could be that Langton's discussion of a different example left her open to a misunderstanding like Jacobson's. In discussing Linda Marchiano, Ordeal, Langton suggests that in a situation in which the book is bought and sold as pornography, and readers take it to be pornography, it is pornography (Langton, , supra note 3, at 322Google Scholar). This suggests that something's being taken to be pornography is (in some contexts) sufficient for its being pornography, even if a speaker intends it not to be. Applying this, by analogy, to the example of sexual consent, one might come to suppose that something's being taken to be consent is sufficient for its being consent, even if a speaker intends it not to be. That would give one the “absurd result” of which Jacobson complains, namely that the woman does consent and hence is not raped. Now Jacobson does not in fact argue by analogy, but appears to make the illegitimate conflation we have described. However, if he did argue this way, then Langton's response would be that the illocutions of pornography (if pornography can be described in illocutionary terms) are very different from the illocutions of consent: the latter is a prime example of a communicative illocution, the former is not; and the speaker's intention matters far more in the latter case than the former. This issue certainly deserves more attention than Langton gave it, and than we are able to give it now. Let a couple of sentences suffice. Whereas it is plausible to suppose that intending to consent is necessary for consenting, it is not plausible to suppose that intending to produce pornography is necessary for producing it. There need then be no inconsistency in allowing that a work's being taken to be pornography might (in some contexts) be enough for it actually to be pornography, while denying that a person's being taken to have consented is (ever) enough for her to have consented.

30. Jacobson, , supra note 6, at 71.Google Scholar

31. Jacobson's allusion to the First Amendment here suggests that he may have different objectives from our own in understanding the proponents of free speech (see the end of our introduction supra). If a Bill of Rights is introduced into the United Kingdom—as, with a new government, it is expected to be—then a U.K. citizen might appeal to a right to free speech even while it was granted that (given facts about U.S. legal history) she would stand no chance if she appealed to a First Amendment right

32. See supra where note 11 is flagged.

33. In the version of On Liberty printed in Utilitarianism (Warnock, M. ed., 1962), at 180.Google Scholar

34. Id. at 184.

35. See Brandenburg v. Ohio, 1969Google Scholar: “[T]he constitutional guarantees of free speech and free press do not permit a State to forbid or proscribe advocacy of the use efforce or of law violation except where such advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action.” As this quotation suggests, debate about what may be excluded from First Amendment protection is debate about which perlocutionary acts defeat a free-speech right The debate always assumes (without making the distinction, it is true) that illocutionary, like locutionary, acts are protected. To say that some advocacy may sometimes be proscribed is to say that some pieces of speech intentionally have such consequences that they are then not protected. (Note that there could be no objection to treating “speech,” in this formulation, as illocution.) But the immunity that is then lost, which is an immunity Mill generally advocates, is not immunity merely to do locutionary things.

36. The question whether Dworkin can carry on with his stated defense of pornographers' rights while allowing any genuinely egalitarian considerations to impinge as Dworkin thinks they ought to was answered negatively by Langton, R., Whose Right? Ronald Dwarkin, Women and Pornographers, 19 Phil. & Pub. Aff. 311 (1990).Google Scholar (Parts are reprinted in Dwyer, S. ed., supra note 3.)Google Scholar

37. Women and Pornography, N.Y Review of Books, 10 21, 1993.Google Scholar

38. See pt. II of Hornsby, J., Disempowered Speech, Phil. Topics (Fall 1995).Google Scholar

39. On “right” in “right to free speech,” see the section “Rights and ‘Rights,’” in Hornsby, J., Free and Equal Speech, Imprints 59 (10 1996).Google Scholar

40. Knowing of the opposition (generally speaking) between libertarian and egalitarian versions of liberalism, one may not find it surprising that a conception of free speech like ours, whose repercussions may be egalitarian, cannot rely upon a thorough going libertarian account of the right (Dworkin seems to want to have it both ways: see Hornsby, “Postscript,” supra note 4.)Google Scholar

41. See Dworkin, R., The Coming Battles over Free Speech, N.Y Review of Books, 06 11, 1992.Google Scholar In the present context, of discussions of pornography, two things make Dworkin's rhetoric here seem laughable. The first thing is the mismatch between the material with prepositional content which Dworkin is actually defending (“opinions,” “convictions”) and the pornography producers' products which he would wish to defend (i.e., the mismatch between speech proper and “expression”: see supra note 9). The other thing is the motives imputed to pornography's producers when Dworkin's defense is applied to their case—“respect and concern for others.”

42. Supra note 6, at 76.

43. Of course, Mill defended moral autonomy in On Liberty. But he did so elsewhere than in the chapter concerned with speech (see supra note 8.).

44. Here we can only gesture toward a wider free-speech literature. Democracy, of course, comes in with the famous argument of Meiklejohn, A., Free Speech and Its Relation to Self-Government (1948)Google Scholar, reprinted in Political Freedom: The Constitutional Powers of the People (1965).