#Chapter II: Performance and Prayer

Chapter II: Performance and Prayer

The Role of Difference

The surprise of otherness is that moment when
a new form of ignorance is suddenly activated as
an imperative.

Barbara Johnson, A World of Difference
Performative language, the subject of J. L. Austin’s How To Do Things With Words, appears prominently in the verb structure of the poem. The first verb of the poem is λίσσομαι, “I pray,” which resembles the classic example of a speech act, “I bet” in that its utterance actually performs an action. [1] The fact that all language is in some sense performative and performs an action simply by its utterance (in other words, that language does something) is central to my study; for now, though, I note that λίσσομαι represents a speech act in Austin’s original sense of the term. The explicit performative element of Sappho’s poem is easily identified in that we have a specific and accepted action—prayer—as its referent. My initial presentation of speech act theory is admittedly simplistic, but I use it as a vehicle towards its more complex ramifications; once we recognize the poem’s nature as an act of prayer and the effects of that nature on the poem, we may explore what happens in the relationship of the explicit speech act—which in itself is a relatively simple concept—to the other elements of the poem’s language.
Λίσσομαι gives the poem its initial motivation and its principal force. The verbs which follow λίσσομαι all depend on it in one of several ways. The series of imperatives, μὴ δάμνα, ἔλθε (twice), τέλεσον, λύσον, and ἔσσο, forms a chain of requests which fulfill the initial λίσσομαι. Sappho’s account of the past epiphany of Aphrodite follows ἔλθε and to {11|12} the extent of sense depends upon that imperative, for Aphrodite is aksed to come if (αἴ) she has ever come before. We see, then, a structure of an initial entreaty and a further expression of that entreaty. The imperatives extend λίσσομαι, indicating what Sappho prays for. Since the epiphany is dependent on part of that extension, ἔλθε, the whole poem hangs from the utterance “I pray.”
The words “I pray” have a function beyond describing what the speaker does in saying them. As I indicated above, “I pray” is much like the classic example of a speech act, “I bet.” The speaker of both utterances performs an action by speaking: by speaking “I bet” in the right circumstances, the speaker actually makes a wager; in saying λίσσομαι as a devotee Sappho actually entreats Aphrodite to grant the wishes expressed in the coming imperatives. While the words “she bets” or “she prays” carry a force of description, the first person singular forms of the same verbs carry a weight of accepted verbal action in addition to describing the subject’s betting or praying. In this way, because the poem depends on the first “I pray,” it is entirely a speech act, or to use Austin’s alternative term, a performative utterance. [2] In saying λίσσομαι, Sappho does not primarily describe her action but performs her action. Austin makes the helpful though not unproblematic distinction that a speech act is not true or false but “felicitous” or “infelicitous.” [3] We cannot prove λίσσομαι correct or incorrect as we might λίσσεται if we knew that the subject were not praying; instead, as “I bet” does not have performative force when the speaker has nothing to bet on and the utterance is thus infelicitous, λίσσομαι loses its ritual weight when the speaker is not in a position to pray. The words “I pray” represents a real action because they occur in a context where the speaker is empowered to pray: she is a devotee of the goddess. The prayer thus conforms to the condition Austin gives for “felicity.” We must recognize then that as a felicitous speech act, the utterance is a prayer, and claims the framing context of prayer, the accepted and received traditions {12|13} which themselves imposed the conditions for prayer in the first place. We must now define these traditions beyond the simple “ever previous prayer”; we must ask what exactly a prayer is.
The conditions for felicity for a speech act involve invoking a tradition of similar speech acts which have gone before. Both “I bet” and “I pray” rely for their effect on the speaker’s and the hearer’s knowledge that the actions of betting and praying have occurred felicitously many times in the past. In other words, these speech acts look to a tradition of performative utterance for their force. When Sappho says λίσσομαι, she looks to the prayers of her own and others’ past devotions, who themselves said λίσσομαι (or εὔχομαι, or any other word of entreaty). Without a tradition of betting taken for granted, the bettor could not gamble, for the action would have no meaning; without a tradition of prayer, a devotee could not make her supplication. In this respect, performative language, the element of the speech act which actually performs the action, resembles ritual; in fact, in a wider sense, performative language is ritual. [4] The use of the proper ritualistic cult forms—the hymnal elements of the invocation (ποικιλόθρονε, ἀθάνατε, παῖ Δίος, δολόπλοκε), [5] the hypomnesia of the epiphany, [6] the imperatives—reinforce the felicity of the speech act. Since each of these elements depends, as I noted above, on the initial λίσσομαι, each carries some of the performative force of the poem. All of the ritual elements of Sappho’s prayer, the cult epithets and hypomnesia, look to past uses of the same kind of language for their force. Without the force of tradition, this language has no ritual significance.
I thus treat the problem of defining prayer and its traditions from two perspectives, that of ritual language and that of non-ritual language (I deal later with the non-ritual). The presence of performative language as the motivation of Sappho’s prayer, recalling past instances of the same kind of speech act, leads us to wonder what part ritual itself plays in the language of a prayer. Ritual prayer invokes in the performative element of its {13|14} language and its ritual forms
(the cult epithets and epiphany) the tradition of previous prayers employing the same ritual performative language and forms. I follow Burkert’s exhaustive treatment of ritual in Greek Religion and Structure and History in Greek Mythology and Ritual for the following discussion of the forces at work in the ritual aspect of Sappho 1. To begin, I choose as a definition for ritual the following from Burkert: “Ritual, in its outward aspect, is a programme of demonstrative acts to be performed in set sequence and often at a set place and time—sacred insofar as every omission or deviation arouses deep anxiety and calls forth sanctions.” [7] I add to this definition the observation culled from Structure and History that ritual derives its authority from repetition. [8] The performance of a ritual is a repetition of the one before that, a reference to it, the last one to the one before that, and all are references to the founding ritual, which may “exist” only within the context of the ritual. In other words the authority of a ritual relies on series of performances which extends back to the founding and forward forever. As the definition from Burkert says, the ritual must be thought of as unchanging in order to retain its sacred quality. This unchanging series finds its ultimate authority from its perceived institution by the founder. In the case of religious ritual, of which Sappho 1 is an example, the participants conceive of a ritual as a programme prescribed for them by the object of their worship—a power infinitely greater than they. For them, the keeping of the ritual creates an infinite series of unchanging events which in its constancy and eternity identifies itself with the absolute which they worship—let us say, the god.
Burkert says of the god, “Speaking of theos or theoi, one posits an absolute and insurmountable point of reference for everything that has impact, validity, and permanence”; [9] and of ritual, “[ritual] lays claim to the highest seriousness, to the {14|15} absolute.” [10] The reference which the repetition of the ritual creates thus calls into being a resonance with previous and future ritual performances which lends ritual authority and validity to the prayer. The repetition inherent in ritual creates and perpetuates an absolute, undifferentiated series of events (at least within the context of the ritual), identifying for its validity with the god. As Burkert says about ritual [11] and Snell says about the poetry of Sappho, [12] ritual or ritual poetry [13] maintains the fabric of the closed group—in Sappho’s case the thíasos. The closed group, then, which posits the absolute theos as the insurmountable point of reference, the other by which they, like an individual person, necessarily define themselves and maintain themselves as closed and indissoluble, [14] must view the ritual as unchanging, just as the other, the god, is unchanging.
The recent studies of ritual as a basic form of communication, a sort of language, [15] lead me now to ask how language functions within ritual. The best place to investigate this problem seems to be within prayer, a meeting place of language and ritual. I propose to look at the problem in light of Sappho 1, which, as I began to show through my initial discussion of speech act theory, is motivated fundamentally by the speech act of prayer. Prayer, however, is not simply ritual in the sense of Burkert’s definition. There exists integrally within the concept of prayer the necessity of verbal (unlike the biological concept of ritual-as-language) ritual language—like the word λίσσομαι—which has the performative element I detailed above. It is the language of a prayer which refers to the ritual tradition of prayer. The word λίσσομαι and the rest of the ritual elements of the poem enact the prayer-ritual which takes as its referent the unchanging series of prayers {15|16} which have gone before and by extension Aphrodite herself. Sappho’s prayer thus looks to an authority embodied in the performative and ritual language of the prayer itself (λίσσομαι, ποικιλόθρονε, μὴ δάμνα). Each of these ritual elements in the context of the ritual carries the force of the tradition of which it is simply the latest instance. Each ritual element of the language therefore has a quality which is thought of within the ritual as unchanging, because the ritual itself is so perceived. To vary my examples, the imperative ἔλθε, in extending the meaning of λίσσομαι, carries a force we might express approximately thus: “come (as you have come to the aid of everyone else who has prayed to you in the past).” The epithet δολόπλοκε carries a force we could paraphrase thus: “snare-woven (for when addressed by such epithets you have always answered prayers).” These paraphrases really do not capture the sense of the unchanging-in-one-word which exists within Sappho’s prayer, but they point in the direction by illustrating that each ritual element of language carries the weight of past prayer-ritual. If non-verbal ritual itself is a sort of language, the verbal language of prayer has a corollary for its perceived unchanging element. The non-verbal language of ritual, because it is the ritual itself, is thought of as unchanging, as Burkert’s definition said. Just so the language of prayer, which, as we shall see, in many ways does change from prayer to prayer, must yet contain the same unchanging linguistic element. As I noted above, this unchanging ritual element appears in the performative λίσσομαι, the imperatives μὴ δάμνα and ἔλθε, and the cult epithets ποικιλόθρονε, ἀθανάτα, and δολόπλοκε.
To repeat the concept of the unchanging tradition of prayer links the ritual element of the prayer’s language and its performative element, because each relies on the accepted series of similar—to the extent of ritual perception identical—prayers for its felicity. The two examples above, ἔλθε and δολόπλοκε, with their ritual force also have the performative force of the speech act begun in λίσσομαι. The same paraphrases which indicate ritual reference to past use of imperatives and epithets also indicate reference to the tradition which supports the performative force of λίσσομαι; a tradition which extends {16|17} itself through the whole poem. The ritual performative aspect of the language of the poem, in that it is perceived as invariable, identifies on the one hand with the unchanging traditional element of the performative, the speech act, and on the other hand with the timeless, absolute other, the god.
We begin to see why Burnett’s emphasis on memory is so telling. The tradition of ritual prayer invoked by the performative represents in a way the “memory” of the performative language of the prayer. Simply by saying ποικιλόθρον᾽ ἀθανάτ᾽ Ἀφρόδιτη / παῖ Δίος δολόπλοκε, λίσσομαι σε Sappho calls up the infinite series of past and future prayer. Her ritual, performative words carry an undeniable force which seems to me crucial for understanding the poem. The speech act of prayer represents an essential element of the motivating force of the poem. The action performs a ritual, calling on the absolute. This aspect of Sappho’s language informs the entire poem; the performative element of her words creates in the poem a ritual, and the presence of a ritual explicitly and inescapably imbues the song with not only the sort of mythology (for myth is an integral part of the ritual memory invoked by the performative [16] ) represented in the epiphany of Aphrodite taken alone—second-hand, descriptive experience—but also a far more immediate sort of mythology, first-hand, performative experience. Sappho 1 is first and foremost a prayer. It is not a parody of a prayer; [17] it is not an imitation of a prayer. A true ritual is enacted in the poem. The performative force of the word λίσσομαι, the imperatives which extend that force to the rest of the poem, and the ritual, cult elements of the poem all invoke a tradition which gives the poem an undeniable ritual performative force. While this force does not exclude the non-ritual, even parodic element from the poem (in fact it necessitates it, as I will discuss presently), it is essential to see that the performative language of the poem generates a real action, a real ritual on a footing with non-verbal action. Marcel Mauss in his unfinished essay on prayer calls prayer “un acte. Elle n’est {17|18} ni une pure rêverie sur le mythe, ni une pure spéculation sur le dogme, mais elle implique toujours un effort, une dépense d’énergie physique et morale en vue de produire certains effets.” [18] Mauss’ anthropological perspective emphasizes the fundamentally active, practical nature of prayer. As Mauss indicates, the distinctive characteristic of prayer is that it does something, just as the distinguishing mark of the performative is that it does something. This aspect of prayer, I submit, arrogates for itself a primary role against the claims of the other elements of the language of the poem. In Sappho’s prayer the thread of the performative extends from the ritual significance of the initial epithet ποικιλόθρονε to the ritual supplication of the final ἔσσο. In invoking the goddess with epithets Sappho recalls the hymnic tradition, [19] an act as real as sacrificing a hecatomb. In recalling the goddess’ past epiphany, Sappho executes a deed as ritually significant as carrying the goddess’ image in procession. We must view each of these speech acts, authorized by tradition and part of the overarching performative utterance which turns on λίσσομαι, as essentially active and performative in nature, and only from that essential nature can we differentiate the other forces of the poem’s language.
In recognizing the primacy of the performative element of the poem, we must also immediately see that the ritual performative is not the only force at work in the poem. To say that the only force of the prayer is to ritually pray in the strict sense of Burkert’s definition of ritual is to deny that the prayer is even a prayer, let alone a poem; as I noted above, the presence of verbal language in prayer complicates the matter. Even the crucial word λίσσομαι has a force besides that of its inherent speech act. Besides enacting a ritual of supplication, the word also carries a descriptive force of supplication. To phrase the distinction negatively: how would the force of the words differ if Sappho had used εὔχομαι, the vaunting prayer-word, instead? Both λίσσομαι and εὔχομαι are performative utterances by Austin’s definition, but each carries another force. It is difficult to isolate {18|19} this force in order to grasp it with a linguistic name. Any name like “descriptive” or “constative” (the term Austin himself formulated as an opposite for performative [20] ) implies yet another type of speech act, like prayer or betting. Description and statement are speech acts themselves in that when a speaker says “I describe” or “I state,” their utterance, like “I pray,” does the describing or stating. Description and statement invoke the same kind of traditions as the other performatives: of felicitous descriptions and statements. When we apply the name “constative” to the non-performative force of an utterance we leave further non-performative force unnamed, for the constative is itself a performative speech act (of statement); we must again attempt to isolate the remaining non-performative element. In reading Sappho’s prayer, when we assert that ἀθανάτα “describes” the goddess as “deathless” besides recalling the ritual tradition of epithets in its ritual force, we see that a part of the meaning still escapes. The act of describing carries only the force of the application of the epithet, not the descriptive essence of the epithet. The assertion that the epithet “states” that the goddess is deathless has only the force of the act of statement. No name we apply can encompass the descriptive, constative element of language, for every name is itself a speech act.
In this respect the relationship between the two sides resembles Roman Jakobson’s distinction of marked and unmarked oppositions in language. [21] Like the opposition of “lioness” and “lion,” where “lioness” is informed by a gender marking while “lion” can indicate both genders, ritual, performative language and constative language, its opposite element, do not lend themselves to a simple binary scheme. While the performative remains clearly visible, the constative recedes constantly from sight, escaping marking.
The only place to begin an attempt at a working definition, then, is with defining the element negatively, as the non-performative. I start by indicating the general force of the non-performative. The easiest force to pick out from the non-performative is the {19|20} constative, which I name thus simply because Austin and those who came after used that term. For the sake of the argument, I call constative here any aspect of language which “constates” or describes its subject (for example, “she prays”), which might be true or false, to use Austin’s criterion. [22] Here I have already betrayed the difficulty in naming I noted above in order to capture the non-performative; when I use the term “constative” I mean the unnamable non-performative, but in fact I conjure up a speech act of description. I mean the force of κάλοι δέ σ᾽ ἆγον / ὤκεες στροῦθοι which creates what we think of as its poetic effect, the force which alongside the performative imbues the poem with its “descriptive meaning.” All of these terms are necessarily subjective; the problem of subjectivity is part and parcel of the problem of naming the non-performative. The non-performative force of the words “And the fine sparrows brought you…” exists only within the subject, the speaker and the hearer, whereas the performative force, at least in the context of the words themselves, has external reality in the tradition of like performance. [23] The ritual, performative aspect of the words refers to the externally perceived tradition of calling on a divinity to descend in a chariot drawn by her favorite animal, a tradition which manifests itself throughout cult worship; [24] the non-performative aspect has no such tradition—the hearer must rely on her own knowledge of the “constative” meanings of the words. I therefore leave the term “constative.” In the following discussion I use the word “descriptive” instead, in an attempt to avoid the difficulty of Austin’s initial formulation.
Clearly the non-performative, descriptive force of the language informs not only the “sense” of words like λίσσομαι but also their entire meanings in the same way the performative does. The two sides of the force at work in the language are mutually {20|21} dependent: without the motivation of the ritual enacted by the speech act, the descriptive force of λίσσομαι would have no motivation; without the descriptive force of language, the speech act could not take place. The element of the poem which describe represent a force just as telling as the performative elements. The example I used above, κάλοι δέ σ᾽ ἆγον / ὤκεες στροῦθοι, possesses a power of expression which we cannot define to any meaningful extent. If we call it “literary” or “poetic” or “lyrical” we invoke further performative traditions, accepted and recognized externally. Also, literary, poetic, lyrical, even strictly descriptive language, often occurs within speech acts. Sappho’s poem itself is an example, for the problems of interpretation which motivate my study arise in the joining of the performative prayer and the lyrical elements like the evocative epithets and the chariot drawn by sparrows. Poetry itself could be seen as an example of a performative tradition to which non-performative poetic effect gives shape. Barbara Johnson in her discussion of performative language and poetry demonstrates also the converse possibility: of poetry encompassing a speech act as a performance referring to the poem’s own lack of occasion, rather than presence of occasion. [25] Johnson’s study raises similar questions for the epiphany of Aphrodite in Sappho’s prayer. Sappho places the description of Aphrodite’s descent to earth and her words to the poet (which I explicate in detail in Chapter IV) in context outside of time with πότα κἀτέρωτα. This constative timelessness subverts any attempt to make of the poem a strictly occasional extension of ritual, the effort for which I criticize Gentili in Chapter I. In an attempt not to fall victim to such a constricted purely performative view of the poem we must eschew such terms in favor of a term which does not connote the loaded (not to mention performative) traditions of the “literary” and the “poetic.”
So what describes? What creates literary effect? Each of the different performatives I have applied to the non-performative in an effort to name it has in {21|22} common one characteristic: they all differ from the speech act of prayer I identified at the beginning of this chapter. The concept of difference underlies all language, of course. My simple example above, of the distinction between λίσσομαι and εὔχομαι, illustrates this principle. As Saussure explained in his Course in General Linguistics, meaning is differential. [26] Λίσσομαι means nothing unless we consider it in relation to εὔχομαι and the other prayer words in the language. When we think of the two words in relation to each other, it is easy to conceive of the force of the unnamable constative—the force is the difference between them, and the difference between each and the other words of the language. Consider “I entreat” and “I boast” (approximate translations). The part of the conceptual meaning of the two words which deals with the attitude of the speaker arises from difference with the other word. If we were to create a new prayer-word, its meaning would obviously depend on its difference from other words. The binary relation of λίσσομαι and εὔχομαι does not encompass all of their non-performative force, for they differ not only from each other but from all of the other words of the language. But it is easy to see that the descriptive force of the words comes from that difference, though in order to document it fully (and pointlessly) we would have to examine its difference from every other word with which it comes into contact. In reality we can never completely capture these differences, for another difference can always appear and change the constative element, like a new word for prayer. The differential constative recedes forever into an infinity of differences. [27] This concept of difference is practical, not mystical, for this retreat of difference allows words their meaning in the first place. Without the unlimited capacity for difference a language could not mean at all. {22|23}
My example of λίσσομαι and εὔχομαι has further implications for a reading of Sappho’s poem. Each of these words is an explicit performative utterance. Each represents an act of language as I outlined at the start of this chapter. I have just tried to show that the so-called constative represents in part the difference between them. If that definition stands, we can extend it to include other speech acts, such as, on a simple level, description. The constative then is the difference between “(I describe you as) rich-throned” and “(I describe you as) deathless” (ποικιλόθρον᾽ ἀθανάτα). More importantly, when Aphrodite speaks within the poem, her own speech act, her questions, generate a constative difference from Sappho’s initial λίσσομαι. Of course these differences do not form the entire constative meaning of the various speech acts, but they do represent an important part, to be taken into account along with the other differences which form the constative.
So we see a sort of structure of difference emerging in the poem. While λίσσομαι provides the impulse for the overarching performative of the poem, the speech act of ritual prayer, the complementary constative of the poem, its “poetic effect,” arises in the differences from that initial speech act which the other elements of the language, like description and statement, create. The description of the epiphany, for example, where the sparrows flutter their wings (στροῦθοι περὶ γᾶς μελαίνας / πύκνα δίννεντες πτέρα) differs in its descriptive speech act from the ritual performative of λίσσομαι in that it is light-hearted and witty—the sparrow, according to Page, is known for wantonness, [28] and the hearer could hardly expect it to pull a chariot; the image is novel [29] and so differs from the traditional, taking on a constative force through its difference. It is this web of difference which I now propose to explore. {23|24}
In this inquiry, it seems essential to find one term which differs from the ritual performative of prayer through which to explore the constative difference in the poem. Through the use of such a term I will be able to show in a manageable fashion the force of the difference which arises in Sappho’s poem. An attempt to explicate every differing speech act would end with an impressionistic haze of various differences, when the actual effect of the differences in the poem is of unity. Because the weight of the poem, as I showed above, rests on the one performative utterance λίσσομαι, the array of differences created by the other speech acts provides finally a “unity of difference” from the enacted prayer. In reading the poem critically, however, it is difficult to capture this unity without betraying the multiplicity which gives rise to it, for selecting one speech act to cover the whole variety of them present in the poem betrays the difference within the unity. My solution is to find a term which is not a speech act but captures the differing potential of the other speech acts in the poem: the “different.” I intend to indicate not a mystical force of language but the precise—though unnamable in one term—speech acts which differ from the ritual speech act motivated by λίσσομαι. The term “different” also functions etymologically as the foil for the absolute element of the performative ritual I detailed above, which identifies with the unchanging god. The word is derived from differo, to carry apart. The force I ascribe to the different “carries apart” the absolute in that it creates constative difference from it and within it, as I will detail much more fully in Chapter III.
I admit that no reader, critical or casual, can capture the force of every differing speech act, because of the multiplicity of difference I noted above. Simply contemplating the myriad of different speech acts in the epiphany should bear out this point. To attempt to find every nuance of the shifts in performative language from Sappho reporting Aphrodite’s words (ἤρε᾽ ὄττι δηὖτε πέπονθα κὤττι / δηὖτε κάλημμι) to Sappho quoting Aphrodite directly (τίς σ᾽, ὦ / Ψάπφ᾽, ἀδικήσι;) to Sappho quoting Aphrodite expressing a sentiment which clearly comes from Sappho’s own spirit (καὶ γὰρ αἰ φεύγει, {24|25} ταχέως διώξει…) would be useless. To express the tenor of the differences which these shifts create by indicating them in the single concept of the different yields valuable insights into the workings of the poem.
Chapter III, in which I explore the relation of the performative and the constative through the difference of ritual prayer and the different, the other speech acts such as the descriptive, the constative, and the persuasive, will clarify the roles of the two forces greatly. Towards this end I begin a reading of Sappho’s prayer informed by the concerns of the performative elements of the poem and their differences from one another. {25|26}


[ back ] 1. Austin, esp. 4–7.
[ back ] 2. Austin 6.
[ back ] 3. Austin 9–24.
[ back ] 4. See the definition of ritual I give from Burkert, Greek Religion below.
[ back ] 5. Burnett 247–249 has a comprehensive bibliography.
[ back ] 6. Again, Burnett 247 offers a useful survey.
[ back ] 7. Burkert, Greek Religion 8.
[ back ] 8. Walter Burkert, Structure and History in Greek Mythology and Ritual (Berkeley: U of California P, 1979) 35–39.
[ back ] 9. Burkert, Greek Religion 272.
[ back ] 10. Burkert, Greek Religion 55.
[ back ] 11. Burkert, Greek Religion 8.
[ back ] 12. Bruno Snell, Poetry and Society: the Role of Poetry in Ancient Greece (Bloomington: U of Indiana P, 1961) 44–47.
[ back ] 13. Segal offers an informed exploration of the oral, ritual nature of Sappho’s poetry.
[ back ] 14. This assumption underlies Burkert, Structure and History 37–38 and 49–51 and is a fundamental tenet of modern psychoanalysis as Lacan, “The Mirror Stage” introduces.
[ back ] 15. Burkert, Structure and History 48–52.
[ back ] 16. Burkert, Structure and History 57.
[ back ] 17. Burnett 246.
[ back ] 18. Marcel Mauss, Les Fonctions Sociales du Sacré, Oeuvres, 3 vols., ed. Victor Karady (Paris: Les Editions Minuit, 1968) 1:409.
[ back ] 19. Burnett 247.
[ back ] 20. Austin 3.
[ back ] 21. Roman Jakobson, “Signe Zéro” (1940), Selected Writings, 2nd, expanded ed., 7 vols. (The Hague: Mouton, 1971) II: 211–222.
[ back ] 22. Austin 9–24.
[ back ] 23. The play of internal and external, inside and outside, which subverts the distinction and allows the external performative to effect the internal constative has an important formulation in Jacques Derrida, “Structure, Sign, and Play in the Discourses of the Human Sciences,” Writing and Difference (1967), trans. Alan Bass (Chicago: U of Chicago P, 1978), rpt. in Adams and Searle, 83–94
[ back ] 24. Page 7.
[ back ] 25. Barbara Johnson, “Poetry and Performative Language: Mallarmé and Austin,” The Critical Difference: Essays in the Contemporary Rhetoric of Reading (Baltimore: John Hopkins UP, 1980) 52–66.
[ back ] 26. Ferdinand de Saussure, Course in General Linguistics (1913), trans. Wade Baskins, ed. Charles Bally and Albert Sechehaye in collaboration with Albert Riedlinger (New York: McGraw-Hill, 1959) 65–70.
[ back ] 27. One of the main contributions of the post-structuralist theorists to Saussurean linguistics has been the transformation of Saussure’s closed system of difference to an open, infinite one. See, notably, Derrida, Of Grammatology (1967), trans. Gayatri Chakravorty Spivak (Baltimore: Johns Hopkins UP, 1976) 30–65 and Lacan, “The Agency of the Letter in the Unconscious or Reason Since Freud” (1958), Ecrits, trans. Alan Sheridan (New York: Norton, 1977), rpt. in Adams and Searle, 738–756.
[ back ] 28. Page 7–8.
[ back ] 29. Most commentators are in agreement here: Leah Rissman, Love as War: Allusion in the Poetry of Sappho, Beiträge zur Klassischen Philologie H. 157 (Königstein/Ts.: Hain, 1983) 9 in particular has a useful analysis.