Chapter 8

There is a late twelfth-century lai by Marie de France, entitled Laüstic, about a nightingale that was killed by a jealous knight who had been told by his wife, when asked why she would leave the bed so often at night and stand by the window, that ‘there is no joy in all the world like hearing the nightingale’ (verses 84-85: il nen ad joië en cest mund | ki n’ot le laüstic chanter). [1] Till then, there had been a series of nightly contacts, purely by voice, between the wife and her secret lover, whose window was nearby:

To warn her lover that the secret has been discovered—an inference that follows, without explanation, from the death of the nightingale—the lady sends the body of the dead songbird to him. She starts by saying: {207|208} When the lover receives the body of the dead songbird, he enshrines it in a reliquary, which he carries around on his person for the rest of his life. [2]

This story of the nightingale, according to the interpretation of one critic, draws a parallel between the song of the bird and the medium of the lai, but the parallelism is enigmatic:

This formulation is offered in the larger context of a disagreement with Paul Zumthor’s hermeneutics of oral poetry, who claims that medieval literature is driven by the primacy of what he calls the living voice. [4] As an alternative to Zumthor’s model, it is suggested that “the Old French text is a tomb of the voice which it betrays.” [5] I agree, at least to the extent that the voice of the nightingale has indeed been ‘betrayed’ or revealed by the text:

And yet, it is not just the text that betrays the voice of the nightingale. Even the voice of poetic performance betrays the songbird’s voice. In the logic of the song, the nightingale sings a secret language, to be understood by lovers only, and the songbird is therefore betrayed if his love-song is made public, that is, if his secret language is sung to the public, the poet’s public.

To appreciate more fully the poetic implications of this theme of betrayal, it is instructive to begin with the poem Philomena praevia temporis amoeni by John Pecham (died 1292). This poet reshapes along religious lines the related theme of the nightingale’s love-song, and the legend it implies can be summarized as follows: {209|210} The songbird’s cry of love and death signals the theme of betrayal. In Song 18 of the troubadour Guillaume le Vinier, for example, the nightingale utters this cry oci! oci! ‘kill! kill!’ (verse 4) precisely because he is denouncing the trahitour ‘traitors’ (verse 7), that is, those who betray true lovers—and thereby cause the nightingale’s death. [7] It is a central convention of the troubadour traditions to represent the nightingale as a loyal messenger sent to the beloved by the lover, by the maker of a love-song: thus in Peire d’Alvernha, Song 1.1–4, the songbird is both the discreet communicator and the faithful guardian of the lovers’ secrets. [8] In the logic of this poetic tradition, the language of the nightingale is like the language of secret lovers: it cannot be understood by the uninitiated. This same language, it follows, is the language of the poet. In an early thirteenth-century poem by Peire Cardenal, Song 56 (verses 33–40), the homology is made explicit:

To betray the nightingale, then, is to betray the lovers’ secrets, which they communicate to each other through their love-song. [10] To betray the songbird’s secret love-song is to make it public. The irony here is that the death of the nightingale becomes essential for the continuity of the bird’s song as art: to betray and thus make public this love-song, with all its beautiful heartache, is to keep it alive—in the form of the song that the troubadour sings to his public. [11] If the poet compares himself directly to the nightingale, as we have just seen in the stanza from Peire Cardenal, it is more appropriate for him to stage himself at that very moment as being alone, without a public, singing to himself a song misunderstood by the uninitiated.

What, then, is the secret theme of the nightingale’s song in the lai of Marie de France? In the logic of the lai—or of the lady’s lament in the lai—the songbird’s betrayal and death signal both the lovers’ feeling of powerlessness to consummate their love and the expression of this feeling in song. It may be that this feeling of unfulfilled love is the secret theme of the nightingale’s song. This theme can be linked with a medieval belief that the nightingale ceased to sing and lost his singing voice once his songs finally led to the long-awaited moment of success in copulating. [12] There is a related theme in the troubadour tradition, as in Song 35 of Gaucelm Faidit, where the poet himself is pictured as dying from an inability to express his love fully in song. [13] {211|212}

My aim, however, is not to reflect any further on the feelings conveyed by the nightingale’s song but to explore the meaning of the song as a symbol of continuity in spite of death, even because of death. Within the story of Marie de France, to repeat the formulation that we saw earlier, the nightingale is “the sign of a ruse or lie told to calm the jealous husband’s suspicions, an invention synonymous with the lai itself.” Beyond the story, however, the song of the nightingale is the very opposite of an ad hoc invention: it is a sign of continuity, of a sad but compellingly beautiful song that cannot end with the death of the songbird. Marie de France draws attention to her use of a Breton word, laüstic, for ‘nightingale’: ceo est ‘russignol’ en franceis / e ‘nihtegale’ en dreit engleis ‘that is, russignol in French and nihtegale in proper English’ (Laüstic verses 5–6). Even in this word of Celtic origin, laüstic, we find an indirect historical indication of continuity in an oral and non-Latin tradition. [14] The image of the nightingale as oral poet persists to this day in the poetic traditions of the Celtic world: for example, the cognate of Breton laüstic in Welsh, eos or eosig, means both ‘nightingale’ and ‘bard’.

The letters embroidered on the silk that enshrouds the nightingale, preserved in the reliquary kept by the lover, are the transcript, as it were, of the song he once sang. To that extent, letters are indeed the tomb of performance. [15] In Greek traditions as well, “the poet seems to be saying that [his] poetry is his sêma ‘tomb’.” [16] This formulation has been applied in interpreting some cryptic verses of Theognis, where the poet’s words mirror the language of inscriptions actually found on tombs: [17] It appears that the poet here is picturing himself as already dead, speaking from a tomb. [18] The verb oikeîn in parallel contexts refers to the establishing of a corpse in a sacred precinct for the purposes of hero cult. [19] After the cryptic words of Theognis 1209-1210, and some further cryptic words that go beyond the scope of this inquiry (1211–1213), the poet reiterates that he is an exile (1213–1214), and then he announces that his abode is next to the Plain of Lethe (1215–1216)—clearly, the realm of the dead (cf. Aristophanes Frogs 186). [20]

And yet, the image of the tomb in this and other archaic Greek passages conveys a message of life after death, achieved through the dead poet’s words. [21] Moreover, sêma means not only ‘tomb’ but also the ‘sign, signal, symbol’ that is the poem. [22] So long as the sign of the dead poet is there, the song may continue to live. A sign authorizes, making the poet an author. [23] The same may be said of the medieval traditions that we have considered: so long as the sign of the dead nightingale is there, the song of the poet may continue.

Still, it would not be enough to think that the death of the nightingale—the death of the poet—ensures the continuity of the song. The one who continues the song must somehow find a point of engagement with the dead poet, through the dead poet’s words. In his 1919 essay “Tradition and the Individual Talent,” T. S. Eliot says that, for the modern poet, the “most individual parts of his work may be those in which the dead poets, his ancestors, assert their immortality most vigorously.” [24] For a culture like that of the ancient Greeks, where performance is still needed to bring the composition to life, Eliot’s words can be reapplied if we take for a moment the performer’s point of view: all I need {213|214} for the moment is to reword his work as his performance. Let me recompose, then: for a performer, the most individual parts of his performance may be those in which the dead poets, his ancestors, assert their immortality most vigorously. In other words, the reperformed composer becomes the recomposed performer.

Here we return to the subject of mimesis in Greek poetic traditions. This book has consistently stressed the fundamental role of mimesis in the performance of song and poetry in the theater, in choral events, in professional citharodic or aulodic events, in rhapsodic events, and even in the symposium. [25] So long as the authority of mimesis continues, we must reckon with its power to reshape the identity of those who take part in the process of performing a song or poem. Just as every performance becomes a potential re-creation in mimesis, that is, a virtual recomposition, so also the very identity of the performer stands to be re-created, recomposed. When the performer re-enacts an identity formerly enacted by previous performers, he or she is re-creating his or her own identity for the moment. That is to say, a performer’s identity is recomposed in performance.

From the standpoint of the Alexandrian period, an era that inaugurates terminal rigidity in the performance traditions of both tragedy and Homer, it is reasonable to expect the perception of a veritable chasm between the actor of, say, a Sophoclean drama and Sophocles himself, or between the immediate rhapsode and Homer, the ultimate Singer. We have seen a striking example of such a chasm at an even earlier date, already in the fourth century: it is the metaphor of the magnet in Plato’s Ion, with the rhapsode Ion pictured as the last and weakest link in a magnetic chain of rhapsodes leading all the way back to the genius of Homer (533d–536d). Still, I maintain my earlier objection to this idea that a rhapsode is a mere replica: such a mentality is contradicted by the more archaic mentality of mimesis, which shapes the idea of a recomposed performer, in that performers may still appropriate to themselves the persona of the composer.

In a performance tradition that is markedly more fluid, as in the case of the troubadour songs represented by Jaufré Rudel, we have seen that any distinction between the so-called “original” {214|215} composer and the performer is in fact so blurred as to lead a modern editor to talk about multiple authorship. Let us recall the striking formulation of Rupert Pickens: “The conventions and traditions of the courtly lyric have conspired to efface the author and to create at least as many Jaufré Rudels as there are medieval anthologies.” [26]

A comparative perspective, taking into account the differences we have observed so far, leads to the following two axioms:

1. The greater or smaller the degree of mouvance, the smaller or greater respectively is the distinction between composer and performer. [27]
2. Where the distinction between composer and performer requires it, the performer’s identity becomes reshaped—recomposed—to fit the ideology of his or her distinctness from the supposedly prototypical composer, the author.

All this is not to renounce the historical reality or even the very concept of authors, of earlier composers of earlier compositions in oral tradition. What I resist is simply the insistence of some scholars on the notion that an original composer of an original composition in oral tradition can be recovered as a synchronic reality. The parallel of historical linguistics imposes itself. Within the conceptual framework of this discipline, we can claim to be reconstructing a given earlier phase of a language on the basis of cognate forms, but we cannot ever say that we have recovered an original phase. [28] The different details that we reconstruct cannot be reassembled into one synchronic reality, one glorious instance of real speech as really spoken in one time and one place. All we can do is predict the relationships that the reconstructed details {215|216} maintain with each other within a continuum that is diachrony. Our predictions must take the form of hypothetical cross-sections of synchrony that correspond to genuinely attested cross-sections, that is, to recordings of living speech.

So also in the study of oral tradition, we cannot expect, as Lord warns us, to recover an “original,” in this case an original composition. Nevertheless, following the model of historical linguistics, we may indeed hope to reconstruct earlier stages or cross-sections of traditions in composition. In my earlier work, I used the letters L M N and so on as symbols for various reconstructed stages of authorship in oral traditions, avoiding the sequence of letters A B C and so on with the implicit purpose of emphasizing that a model of reconstruction cannot start with the beginning, only with a beginning. [29] To start with L M N and so on is thus symbolically apt, in line with the archaic Roman custom, derived from earlier conventions in the writing traditions of Semitic languages, of dividing the alphabet into two halves for teaching purposes, with the recto, as it were, starting at A-B-C and the verso, at L-M-N. Thus by learning the essentials of language, one would learn concurrently one’s A-B-C-s and L-M-N-s. The idea of L-M-N-s as implicit essentials, alongside the A-B-C-s as explicit ones, helps explain the etymology of Latin elementum, alongside abecedarium, as Michael Coogan has argued persuasively. [30] To use Coogan’s metaphor, the sequence L M N in one particular Qumran student’s practice abecedarium represents “a fresh start.” [31]

This etymology of elementum can serve as a fitting symbol for the elements of authorship in oral tradition. As we attempt to trace a progression of originators within an oral poetic tradition, we will predictably fail if we start with an originator standing at a starting line, as it were, but we may indeed succeed in catching up, along the way, with successive relays of continuators, each of whom becomes an originator for the next continuator.

The continuators, of course, need a continuum—a continuous setting, to match any original setting. We may link such “original” {216|217} settings as the hetaireía ‘assembly of comrades’ addressed by Alcaeus at one time and one place with such historically continuing settings as the symposium, in all its varieties throughout many times and many places, where the spirit of hetaireía writ large provides a fitting context for re-enactments of Alcaeus’ words in song. [32] To this extent, I accept Wolfgang Rösler’s dictum that the identity of Alcaeus as a lyric poet was a function of his social group, his hetaireía: “ohne Hetairie kein Lyriker Alkaios”. [33] I even accept the notion of Alcaeus as an author. I must insist, however, that the hetaireía is diachronic—and so too, for that matter, is the persona of Alcaeus. That is to say, the persona of Alcaeus may be adaptable through time, fitting a wide variety of situations—both positive and negative—affecting the very idea of hetaireía. Just as the society reflected by Alcaeus—let us continue to call it his hetaireía—changes over time, so also the persona of Alcaeus may change along with it. If indeed Alcaeus was transmitted primarily through the symposium, then Alcaeus the author will change as the symposium changes through time.

The symposium can serve as an ideal example of a setting for performance, since this institution happens to be more conservative than most in maintaining a continuum of traditional values in the history of Hellenism. [34] And the stronger the continuity, the stronger we may expect to be the sense of potential identification between composer and performer. For example, the sympotic persona of Alcaeus, conveyed in the varieties of ethos that are being acted out in the songs attributed to him, makes it all the more natural for any sympotic performer of Alcaeus to develop a relatively strong sense of identification with him in performance. The same can be said of monodic poetry composed in elegiac couplets, as in the case of Theognis: “the figure of Theognis speaks less as a generalized choral personality and more as a specialized sympotic personality” (cf. especially Theognis 239–243). [35] Even the verses of Archilochus, at least {217|218} those composed in elegiac couplets, were suited for performance at symposia. [36]

We may recall in this connection the opinion of Aristotle, for whom the paideía ‘education’ that a younger man acquires in the symposium, presumably by way of performing as well as hearing the kind of songs that were traditionally performed in that context, provides an immunization against the potentially harmful effects of attending theatrical performances of íamboi ‘iambs’ and comedy, where we would expect the mimesis to concentrate on negative varieties of ethos: τοὺς δὲ νεωτέρους οὔτ᾿ ἰάμβων οὔτε κωμῳδίας θεατὰς θετέον, πρὶν ἢ τὴν ἡλικίαν λάβωσιν ἐν ᾗ καὶ κατακλίσεως ὑπάρξει κοινωνεῖν ἤδη καὶ μέθης, καὶ τῆς ἀπὸ τῶν τοιούτων γιγνομένης βλάβης ἀπαθεῖς ἡ παιδεία ποιήσει πάντως ‘it should be ordained that younger men not be theater-goers [theataí] of íamboi or of comedy until they reach the age where they have the opportunity to participate in lying down together at table and getting intoxicated [that is, to participate in symposia], at which point their education [paideía] will make them altogether immune to the harmful effect of these things’ (Politics 1336b20–22). [37] By implication, the paideía of mimesis in the symposium—even the cumulative ethos of the symposium—provides a proper balance for Hellenic youth in their educational experience of learning the variations of ethos. The symposium, it seems, provides a “safe” occasion for morally vulnerable younger men to hear things that might be “unsafe” to hear in the theater. Presumably even the roguish blame poetry of Archilochus becomes a “safe” topic in the symposium.

In this connection, we may note Aristotle’s tendency to use specific passages from the poetry of Archilochus when he tries to reconstruct the prototypes of comedy (for example, Poetics 1449a9ff, apparently with reference to Archilochus F 120 W); [38] we may note too his theory that íamboi are a prototype of comedy and that they stem from blame poetry (Poetics 1448b32–1449a6). [39] And yet, it seems that Aristotle does not rule out the symposium as {218|219} a context for performing this kind of poetry, despite its frequent representation of rogues who exemplify a negative morality.

An example of such roguishness is Archilochus F 4.7–8 W, where the speaker says that ‘we’ cannot endure guard-duty without drinking wine. The question imposes itself: is the speaker referring to a real situation? One critic offers this answer: “I think it far more probable that Archilochus is evoking a situation with which his audience was all too familiar but which they could thank the gods was not their actual situation while they sang.” [40] I agree, but my point of emphasis is different: the negative morality that is being represented in this composition is being framed by the positive morality of the symposium as the setting of the representation. If I understand Aristotle correctly, the same representation in a setting that is different from the symposium, such as the theater, would make it easier for impressionable youth to become vulnerable to the negative morality that is being dramatized. In other words, Aristotle seems to be saying that the symposium provides a proper frame for moral discrimination, whereas the theater is more hazardous.

In the mimesis of a rogue’s persona, as in the case of Archilochus, we may expect an intensification of distinctions between composer and performer in the symposium. A similar point can be made about other personae as well, as in the case of Sappho. We know that Sappho’s songs, like those of Archilochus, were performed at symposia. In one source, Solon himself is pictured as becoming enraptured by a song of Sappho as sung by his own nephew at a symposium (Aelian via Stobaeus 3.29.58). [41] The point is, for a male singer to act out a woman’s persona implies a radical reshaping of personality in performance. One critic {219|220} remarks, in arguing for a contrast between the ease with which a symposiast may perform the words of Theognis addressing his boy-love Kyrnos and the difficulty with which the same symposiast may perform the dramatized words of Aphrodite addressing Sappho: “Contemporaries will have had little difficulty in singing a song addressed to [Kyrnos]; they might, however, have felt some oddity in singing Sappho [F] 1, with its give-away τίς σ᾿ ὦ Ψαπφ᾿, ἀδικήει; [‘who, Sappho, is doing you wrong?’]; or reciting such lines as Hipponax [F] 32.4 W δὸς χλαῖναν Ἱππώνακτι [‘give a cloak to Hipponax!’].” [42] So also with such songs as Alcaeus F 10 V and Anacreon PMG 385: in each case, the dramatized persona who is speaking is clearly female.

We must distinguish, however, between a dramatized ‘I’ who simply plays out a conventional role in a conventional situation and a dramatized ‘I’ who claims to be the author, as when a rhapsode intones ‘tell me, Muses’ or ‘tell me, Muse’, thereby becoming Homer the author, Homer the culture hero of epic. Here we return to the second axiom: where the distinction between the composer and performer requires it, the performer’s identity becomes reshaped—recomposed—to fit the ideology of his or her distinctness from the supposedly prototypical composer, the author. When a rhapsode performs the lament of a woman, the lamenting ‘I’ is surely distinct from the narrating ‘I’ of Homer the author. [43] Granted, the equation of the rhapsode’s ‘I’ with Homer’s ‘I’ is itself an act of mimesis, but the further equation with a lamenting woman’s ‘I’ surely intensifies the mimesis. In the same way, a sympotic performer’s ‘I’ is surely less mimetic when it renders an ‘I’ overtly equated with Archilochus as author or with Sappho as author and more mimetic when it renders an ‘I’ that seems distinct from the authors.

Even so, the identity of the author is at risk. Let us reconsider the various songs in which an “author” is speaking through what is understood to be his or her own persona. The variety of situations conjured up even in such appropriated songs may lead to a commensurate variety of speaking personae. In other words, the demands of mimesis may lead toward an intensified multiplicity {220|221} in ethos even for the author, with the persona of an Archilochus or a Sappho becoming transformed into multiple personalities that fit multiple situations. Just as the performer may be recomposed in multiple ways, so too this multiplicity may be retrojected all the way to the supposedly prototypical composer, the author. A case in point is the persona of Sappho, which becomes refracted into multiple personalities that eventually become distinguished from the “real” poetess in various Life of Sappho traditions: one such “fake” Sappho is a lyre-player who reputedly jumped off the cliff of Leukas (Suda Σ 108, iv 323 ed. Adler; cf. Strabo 10.2.9 C452), while another is a courtesan (hetaíra: Aelian Varia Historia 12.19, Athenaeus 13.596e), even a prostitute (publica: Seneca in Epistles 88.37).

Despite the verifiable reality of recomposition-in-performance, of change in identity within the process of mimesis, the songmaking tradition may continue to insist on its unchangeability. The tradition may even claim that mimesis itself is the visible sign or seal of unchangeability for the song and, by extension, for its author. Such a traditional mentality is evident in two passages from Theognis of Megara.

In the first passage, the persona of Theognis claims that he is placing a sphragís ‘seal’ upon his words as he identifies himself by name: It has been argued about the “seal”:

With his “seal,” then, the figure of Theognis is authorizing himself, making himself the author. [45] There is an explicit self-description of this author as one who succeeds in sophía, the ‘skill’ of decoding or encoding poetry. [46] On the basis of this success, the author lays claim to a timeless authority, which resists the necessity of changing just to please the audience of the here and now, who are described as the astoí ‘townspeople’. [47] The author must risk alienation with the audience of the here and now in order to attain the supposedly universal acceptance of the ultimate audience, which is the cumulative response of Panhellenic fame. [48]

Such fame is achieved, as we see from the second Theognis passage, through the authority and authenticity of mimesis. Implicitly, only the pleasure of exact reperformance, which is {222|223} supposedly the ongoing achievement of mimesis, is truly lasting, while the pleasure elicited through changes in response to an immediate audience is ephemeral. In this second passage, the persona of Theognis declares that only the one who is sophós, that is, ‘skilled’ in the decoding and encoding of poetry, can execute a mimesis of Theognis: Here the notion of mimesis becomes an implicit promise that no change shall ever occur to accommodate the interests of any local audience in the here and now, that is, of the astoí ‘townspeople’. The authorized reperformance of a composition, if it is to be a true re-enactment or mimesis, can guarantee the authenticity of the “original” composition. The author is saying about himself: “But no one who is not skilled [sophós] can re-enact my identity.”

Here is an occasion to conjure up, yet again, the reworded words of Eliot: for a performer, the most individual parts of his performance may be those in which the dead poets, his ancestors, assert their immortality most vigorously. These words bring us back to the paradox of mimesis, which demands a never-changing identity for the author through an ever-changing identification with whatever the author chooses to represent: {223|224}

To this extent, there is indeed such a thing as an author in oral tradition—or at least, there are different kinds of author in different traditions. In pre-Islamic poetry, for example, as critics contemplate the myriad variants constituting the corpus of a single poet, they claim they can sense the author’s presence in the creation—and re-creation—of his poetry:

To recognize the song, then, is to recognize the singer. And yet, the singer cannot be independent of the song, as it continues to be performed and re-performed. We may heed the words of Thomas Hardy, echoing the poems of Wordsworth, Keats, and many others:The Selfsame Song