1. Veltrusky, Jiri, ‘Text as a Component of Theater’, Semiotics of Art, Ladislav, Matejka and Irwin, R. Titunik, eds. (Cambridge, Mass.: MIT Press, 1976), p. 114.Google Scholar

2. Pavis, Patrice, ‘Semiology of Theatrical Gesture’ Poetics Today 2.3 (1981): 87.Google Scholar

3. Pavis, Patrice, Languages of the Stage (New York: Performing Arts Journal Publications, 1982), p. 147.Google Scholar

4. Veltrusky, p. 115.

5. Ingarden, Roman, The Literary Work of Art, trans. George, Grabowicz (Evanston, Ill.: Northwestern University Press, 1973), pp. 332–55Google Scholar; see especially p. 341.

6. Gadamer, Hans, Truth and Method, trans, Joel, Weinsheimer (New York: The Crossroad Publishing Company, 1988), p. 104Google Scholar, emphasis added.

7. Veltrusky, p. 96.

8. Wolterstorff, Nicholas, Works and Worlds of Art (Oxford: Clarendon Press, 1980), p. 232.Google Scholar

9. Ingarden, p. 318.

10. Wollheim, Richard, Art and its Objects, 2nd ed. (New York: Cambridge University Press, 1980), p. 5.CrossRefGoogle Scholar

11. According to Ingarden's theory, the tokens, or ‘concretizations’, of a novel are not the individual copies of the novel, but the individual readings of the novel (where ‘reading’ does not denote the experience of reading, but the intentional content of that experience; see p. 336). Consequently, whereas Wollheim establishes a parallel between a performance and a book-copy, Ingarden establishes a parallel between a performance and a book-reading. David Cole explores the parallel between acting and reading at length in Acting as Reading: The Place of the Reading Process in the Actor's Work (Ann Arbor: The University of Michigan Press, 1992). Cole demonstrates that the analogy between performance and reading is deep and suggestive. Nevertheless, the logical relationship of a novel to its readings is significantly different from that of a play to its performances. To read a novel produces a ‘reading of a novel’, but to see a play does not produce a ‘performance of a play’. Rather, one might say that to see a play produces a ‘reading of the performance of the play’. The reading of a novel, then, seems to correspond more closely to the experience of watching a performance than to the performance itself.

12. I should stress that in adopting Wollheim's application of the type/token relation to theatre, I am doing no more than registering the fact that performances of plays are instances of those plays. This caveat is important because for Wollheim the type/token distinction carries along with it additional ontological baggage, rendering many of his specific claims about performance problematic. See Margolis, Joseph, Art and Philosopy (Atlantic Highlands, NJ: Humanities Press, 1980), pp. 23 and 70–1.Google Scholar

13. Pavis, Languages of the Stage, p. 143.

14. Fischer-Lichte, Erika, ‘Theatre in Cultural Contexts', Semiotica 86–1/2 (1991): 132.Google Scholar

15. Fischer-Lichte, Erika, The Semiotics of Theater, trans. Jeremy, Gaines and Doris, L. Jones (Bloomington: Indiana University Press, 1992), p. 206.Google Scholar Though she ends up advancing a translation theory in spite of herself, Fischer-Lichte prefers to use the word ‘interpretation’ in describing the relationship between text and performance. Note, however, that the proposal that ‘performance is interpretation’, even if true, does very little to help us define the play/performance relationship. Performances do, of course, entail or implement interpretations of plays. (Wollheim himself makes this point; see pp. 84–7.) But something that is a performance and that interprets a play is not thereby a performance of the play. If performance is a species of interpretation, it cannot be just a performance and an interpretation, but must be a special kind of interpretation, a ‘performative interpretation’. The performance interprets the play by performing it the way that it does. For example, we might say that Orson Welles's 1937 production of Julius Caesar presented a particular interpretation of the play—in this case, a reading of the play as a statement about fascism—by having the cast perform the play wearing Fascist uniforms. To flesh out this theory, however, we would still need to explain what it means to ‘perform the play’. Hence, the proposal that ‘performances interpret plays’ cannot explain what constitutes the relationship between a play and its performances, since it presupposes that relationship. While ‘interpretation’ may be one of the things that a performance does, it is certainly not what performance is. For an interesting discussion of the disanalogy between performance and interpretation in the case of music, see Levinson, Jerrold, ‘Performative vs. Critical Interpretation in Music’, in The Analytic Aesthetics and Dramatic Theory 275 Interpretation of Music, ed. M., Krausz (London: Oxford University Press, 1992), pp. 33–60.Google Scholar For opposing arguments in defence of the view that performance is a species of interpretation, see Margolis, p. 117; and Scruton, Roger, Aesthetic Understanding (Manchester: Carcanet Press, 1983), p. 30.Google Scholar

16. Wollheim, p. 5.

17. A general theory about the conditions necessary for the existence of a type is beyond my purview. Certainly, I do not mean to suggest that the ability to generate new tokens is a necessary condition for the existence of any type, though it may well be in the case of plays. It is impossible to produce a new, genuine, Model T (though one might produce replicas), or a new print of a Rembrandt etching, but the types (i.e. the model of car and the etching) still exist at least as long as there are still tokens. On the other hand, the existence of tokens is not a necessary condition either, as our consideration of plays, and the example of the 1995 Ford Taurus, has shown us. It may be that either the existence of tokens or the ability to produce new tokens is necessary to the type's existence. Of course, some theorists have argued that type entities such as literary and musical works are timeless. Though I find this view implausible and unappealing, there is no need to settle the issue here.

18. The term ‘sortal concept’ is David Wiggins's. Strawson uses the term ‘sortal universal’ to express the same notion, which derives from Locke and ultimately has its roots in Aristotle's distinction between substance and quality. See Strawson, P. F., Individuals (London: Methuen, 1959), pp. 208–14CrossRefGoogle Scholar; Wiggins, David, Sameness and Substance (Cambridge, Mass.: Harvard UP, 1980), pp. 7–8, 77–101.Google Scholar See also Wolterstorff's painstaking analysis of ‘kinds’ (pp. 45–58).

19. The underlying problem here concerning the ‘existence’ of type-concepts is closely related to the scholastic debate between realists and nominalists concerning the ‘existence’ of universals, which itself has its origins in Aristotle's critique of Plato's theory of Forms. From a nominalist perspective, Gadamer's suggestion that plays ‘exist really’ only in performance is precisely right.

20. Ingarden, Roman, The Work of Music and the Problem of Its Identity (London: Macmillan, 1986).CrossRefGoogle Scholar

21. Cavell, Stanley, Must We Mean What We Say (Cambridge: Cambridge University Press, 1979), p. 160ff.Google Scholar Dufrenne promotes a similar conception of plays when he describes what he calls a play's ‘inexhaustible depth’. Dufrenne, Mikel, The Phenomenology of Aesthetic Experience, trans. Edward, Casey, et al. (Evanston: Northwestern UP, 1973), pp. 26–7.Google Scholar

22. See Ubersfeld, op. cit.; and also Issacharoff, Michael, ‘Stage Codes', Performing Texts, ed. Michael, Issacharoff and Robin, F. Jones (Philadelphia: University of Pennsylvania Press, 1988), pp. 59–74CrossRefGoogle Scholar, and, for an interesting twist on the issue of stage directions, Suchy, Patricia A., ‘When Words Collide: The Stage Direction as Utterance’, Journal of Dramatic Theory and Criticism 6.1 (1991), pp. 69–82.Google Scholar

23. Playwrights such as Richard Foreman and Heiner Müller often do not divide their text into sections of dialogue, leaving actors and directors free to do with the text what they please. Such texts are playscripts in only the most minimal sense. Karl Toepfer has aptly described Miiller's texts as provoking performances rather than prescribing them. ‘Strategies of Temporal-Spatial Appropriation in Postmodern Aesthetic Performance: Part I’ Theater Three (Spring 1989): 80.Google Scholar Nonetheless, to the extent that they are playscripts at all, it is only because they have been published as playscripts. The act of labelling a text as ‘play’ provides a bare minimum of didascalia; it implies the instruction to use the text in a performance. Ostensibly, any performance that follows this minimal requirement will constitute a token of the play.

24. Ubersfeld, pp. 255–258.

25. De Marinis, Marco, Semiotics of Performance, trans. Aine, O'healy (Bloomington: Indiana University Press, 1993), p. 43.Google Scholar Marco De Marinis chooses to base his analysis on Alf Ross's taxonomy of speech acts instead of Searle's more widely accepted taxonomy. All in all, De Marinis offers the most subtle and plausible attempt so far to define plays in terms of speech acts.

26. Pavis, Languages of the Stage, p. 141.

27. For a discussion of the difference between constitutive and regulatory rules, see Searle, John, Speech Acts: An Essay in the Philosophy of Mind (New York: Cambridge University Press, 1969), pp. 33–41.CrossRefGoogle Scholar

28. Speech act theory can help elucidate the nature of the performance event, not by clarifying the illocutionary status of the plays, but by clarifying the status of the speech acts performed by actors in the course of performing plays. See Saltz, David Z., ‘How to Do Things on Stage', Journal of Aesthetics and Art Criticism 49.1 (1991): 31–45.CrossRefGoogle Scholar

29. See Esslin, Martin, ‘Beckett and the Art of Broadcasting’, Mediations: Essays on Brecht, Beckett and the Media (New York: Grove Press, 1980), pp. 125–54.Google Scholar

30. James O. Halliwell-Phillipps, quoted in Horace Howard Furness, ed., Hamlet: A New Variorum Edition, vol. 1, 13th ed. (Philadelphia: J. B. Lippincott Company, 1877), p. 375.

31. Levinson, Jerrold, Music, Art and Metaphysics (Ithaca: Cornell University Press, 1990), p. 81.Google Scholar

32. Of course, we are free to define our sense of what counts as performing a play in terms of what we believe to be or to have been the playwright's intentions. But ultimately, the playwright's intentions are definitive precisely to the extent that audiences and performers hold them to be definitive. As Kendall Walton has shrewdly suggested, ‘Even if we do understand a thing's function to be linked to the objectives of its creator, this may be so only because there happens to be a tradition or convention or understanding whereby this is so’ (Mimesis as Make-Believe [Cambridge, Mass.: Harvard University Press, 1990], p. 52).Google Scholar

33. Quoted in Helbo, André, Theory of Performing Arts (Philadelphia: John Benjamins Publishing Co., 1987), p. 55.CrossRefGoogle Scholar

34. Goodman, Nelson, Languages of Art (Indianapolis: Hackett Publishing Company, 1976), p. 186.Google Scholar

35. Goodman's proposal that the dialogue portion of a playscript, like a musical score, is ‘in a virtual notational system’ implies that the performance of a play must (to be a performance of the play) conform perfectly with the playscript at least with respect to the dialogue (210). Presumably, on Goodman's view, a failure to comply with stage directions, which he describes as ‘scripts in a language that meets none of the semantic requirements for notationality’ (211), would have no effect on the identity of the performance.

36. Wolterstorff, pp. 62–65. For Wolterstorff's argument against Goodman's position, see pp. 98–105.

37. For Goodman on variations, see Goodman, Nelson and Elgin, Catherine Z., Reconceptions in Philosophy and Other Arts and Sciences (London: Routledge, 1988), pp. 66–82.Google Scholar Note that the only reason we would need to consider Breuer's production a variation according to Goodman is because Breuer altered the text to accommodate the gender reversals. Since for Goodman only the dialogue is constitutive of a play's identity, simply cross-casting a play with respect to gender would not in itself affect the performance's identity.

38. Goodman, Languages of Art, p. 128.

39. Wolterstorff, p. 63.

40. See Savran, David, Breaking the Rules: The Wooster Group (New York: Theatre Communications Group, 1986), pp. 191–193.Google Scholar

41. Strictly speaking, Goodman might consider the performance to be a clear-cut instance of the work only if Cage recorded his score using a notational system. This qualification does not affect my argument, however, since one could perfectly well record Cage's score in this way. Wolterstorff does consider the intention to follow the score to be a necessary condition for a event to count as a performance of a musical work. (See pp. 74–83.) He introduces this condition to rule out, for example, playing a record as a case of performing the work. He does not, however, consider intention to be an aspect of the work itself. On his view, the musical work itself is a sound-structure that conforms to a set of normative requirements. Hence, if we apply the logic of Wolterstorff's theory to the case at hand, we would say that in unintentionally following Cage's score, our pianist did not perform 4′ 33′′, but did produce a correctly formed example of 4′ 33′′.

42. Schechner, Richard, ‘Restoration of Behavior', Between Theatre and Anthropology (Philadelphia: University of Philadelphia Press, 1985).Google Scholar