Thomas Habinek
Todd III
click here for the main site portal

Singing, Speaking, Making, Writing:
Classical Alternatives to Literature and Literary Studies

It scarcely needs arguing in the present context that the notion of literary discourse that underwrites the modern academic study of literature is the product of specific historical forces. Critics as diverse as E.D. Hirsch, Jr. and Raymond Williams have remarked upon the turn to the aesthetic that characterized late eighteenth and early nineteenth century approaches to writing, although they attribute the shift to forces of differing sorts.1 Ivan Illich reminds us that the habits of reading and writing that made possible the modern institutional framework of literature were to a large extent created by late medieval scholasticism.2 Following his account, we might redescribe the modern invention of a discipline of literature as the substitution of secular scripture for sacred, with the surrounding institutional apparatus adjusted accordingly. 3 Even to discuss the possibility of disciplining literature otherwise, as this collection does, is to put into question the subject, process, and context of literary studies. To paraphrase Tzvetan Todorov, "Literature in general does not exist, but variable conceptions of literature exist and will continue to exist, not only from one period or country to another but also from one text to another." 4

The multiple accounts of the interrelated origins of literature and literary studies, despite their important differences, are for my purposes variations on a theme. What they have in common, besides an interest in the inventedness of literary studies, is a relative lack of interest in antiquity. At best the ancients are regarded as prototypes of their neoclassical interpreters, with the result that ancient "literature" is taken to be more or less what the eighteenth century took it to be, that is, what various figures in the eighteenth century wanted literature to be in their own era. Ironically, classicists who have felt uncomfortable with this pigeonholing of antiquity have tended to respond by seeking out what is literary in the ancient material rather than by using their own understanding of antiquity to interrogate both the prevalent conceptualization of literature and the leading accounts of its development. In the brief compass allotted, I would like to commence just such a project by describing the diverse conceptualizations of verbal performance that characterize the ancient material as well as the struggle for social authority such competing conceptualizations entailed. Along the way, I hope to alert the reader to certain parallels and contrasts between the ancient struggle over the meaning of literature and more recent (and ongoing) contests and to suggest, if only indirectly, some possibilities for reconfiguration of classical and later "literary" studies.

A good place to begin is with one of the key texts in ancient criticism—indeed in all of Western literary theory—namely, Aristotle's presentation of katharsis as the final cause, or purpose of tragedy, in Chapter 6 of the Poetics. In an important recent discussion, Andrew Ford has sought to revive the interpretation of katharsis as a physical and emotional rather than an aesthetic or intellectual experience.

    I see no reason to resist the implication...that katharsis is a kind of pleasing and relaxing emotional experience which supervenes on certain stimuli; tragic katharsis would be the specific form of this response that attends the arousal of our emotions of pity and fear in an environment made safe by the fact that the tragic spectacle is only an imitation of suffering." 5

In the Poetics implicitly and in the Politics explicitly katharsis is differentiated from math¯esis, or learning. Katharsis is an experience of the lower orders of society, one that can safely be shared by the elites for whom and of whom Aristotle writes, only because they have the additional capacity to evaluate the imitation that prompted the katharsis. As Ford puts it,

    [Aristotle] maintains popular arts, with their popular satisfactions, but at the same time creates an intellectual and social space for elite art. As the solution to a political problem, katharsis presides not at the birth of tragedy out of the spirit of music, but at the birth of literary criticism out of the needs of cultivated leisure."6

Whereas Plato sought to regulate the popular and democratic art of tragedy directly, Aristotle engages in a more subtle form of regulation through evaluation and interpretation, specifically a type of evaluation and interpretation that focuses on emplotment and characterization rather than on the ritual, musical, and spectacular elements that are part and parcel of the ancient phenomenon known as tragedy. In turning tragedy—as an object of study—into something more closely akin to the modern concept of literature, Aristotle effectively purges it of the very katharsis that had made it such a potent and, in Plato's view, dangerous type of performance.

While Ford is attentive to the political implications of his account of Aristotelian katharsis, the final portion of his discussion characterizes Aristotle's "restrictiveness" as a "severe but enabling reduction," one through which "the theory of literature becomes possible, and"—in the final words of his essay—"drama enters the field of literature."7 Ford seeks to conciliate those scholars who prefer a more aesthetic or intellectual version of katharsis by showing that an interpretation of katharsis as an emotional experience does not negate the aesthetic and intellectual component of Aristotle's treatise; but in so doing he also reinforces the unfortunate tendency of classicists to accommodate their study (usually somewhat belatedly) to contemporary fashion rather than to use it, as I prefer, to challenge prevalent assumptions. Surely what Ford has brought to light—without articulating it as such—is the interestedness of Aristotle's formulation of katharsis and through it drama and through drama poetics. As with his discussion of virtue (aret¯e) within the Nicomachean Ethics, wherein Aristotle seeks to adjust an essentially aristocratic and theatricalized concept of virtue to the advantage of the business elites of the expanding Macedonian hegemony, so in the Poetics Aristotle seeks to "rescue" traditional forms of verbal production by privileging those aspects of them least likely to discomfit his fraction of society. 8 In this respect, his resolution of the tension between elite and popular cultural forms is repeated centuries later by Horace, who in his Letter to Augustus (Epistles 2.1) argues for the privileging of written literature that meets the standards of knowledgeable critics over revivals of theatrical productions that are diverse in both their modes of presentation and their social appeal. 9

It is important to understand that when Horace and to an even greater extent Aristotle restrict poetry chiefly to plot and character they are resisting much more than the performance qualities of the single genre of tragedy. As an emerging consensus among classical scholars makes clear, every genre of what we have come to regard as classical literature has its origin in a specific performance context. This recognition, fostered by the research of scholars such as Nagy, Calame, and Gentili, has in turn made possible a redefinition of ancient verbal artistry that has the potential to undo the preference for literariness articulated by Aristotle and his various successors and interpreters.10 Working from the language used to describe performance and to a lesser extent from the verbal artifacts of other cultures, Nagy in particular has described a system of verbal production in which the key distinction is between speaking and singing (or what Nagy also calls unmarked speech and SONG). SONG, defined as speech in the special context of myth and ritual, can be accompanied by movement and instrumentation and is frequently, although not necessarily, constrained by rhythm and/or melody. Just as singing constitutes a marked version of speaking, so too within the sphere of SONG, poetry, marked for the absence of melody, can be differentiated from song-types more generally. And within poetry a differentiation can be made between unmarked poetry (subdivided into various poetry-types) and prose, which is marked for its lack of meter. Despite the differences in mode of performance implied by the terms SONG, song-types, poetry, and prose, as well as the multiplicity of genres generated by the range of performance contexts in the culture of the ancient city, the distinction between SONG (encompassing the categories differentiated within it) and unmarked speech long remains a socially productive one. Moreover, the emphasis on performance that pervades the entire system outlined by Nagy—from SONG to prose, from lyric to drama to history—generates for the written text a pair of functions very different from what is found in either Christian scholasticism or its secular counterpart, specifically preservation of a canon of texts for reperformance and circulation of the performance script to new locales. As Nagy rightly notes, writing is convenient, but not essential, for the fulfillment of such functions. It is thus ancillary to, rather than constitutive of, the ancient tradition of "literature." 11

Because scholars such as Nagy have focused their attention on the earliest periods of Greek cultural history, it is easy for their discussion of the differences between SONG and poetry, poetry and prose, performance and composition, poetry and written literature, and so on, to be put to the service of teleological narratives. As I have already suggested, Ford's analysis of Aristotelian katharsis seems to fall into this pattern, placing Aristotle as an important milestone on the road from SONG to literature. Calame also creates an opening for such a move when he remarks in his recent study The Craft of Poetic Speech in Ancient Greece:

    It is true, however, that Greek poetry of the earliest period is, in essence, a poetry of occasion; each composition is ordered for a particular festival and composed for a specific public. And the poetic utterance bears many marks of this process.... With this in mind, the enunciative perspective is useful for comprehending a literature that, at least until Herodotus and although poetry gradually gave way to prose, was profoundly dependent on the practices surrounding its production.12

To some extent, the phrases I have emphasized can be taken to reflect an admirable scholarly caution, restricting the applicability of certain claims to the era Calame and others have investigated most thoroughly. But such a move plays into a too seductive tendency to exoticize the earliest periods of Western cultural history thereby containing their potential significance for an understanding of classical antiquity—as well as of later periods of Western culture. It is precisely this tendency that I wish to resist here, for it is my belief that the system of SONG outlined by Nagy (i.e., the setting apart of speech in the context of myth and ritual) pervades a far broader geographical and chronological range of antiquity than the careful periodization of Calame, Ford, and others might imply. If nothing else, I would like to suggest that to speak of literature at any point in antiquity is to answer a question that has scarcely even been asked.

In support of my claim that the distinction between singing and speaking remains as important as if not more important than any emerging notion of literature I would like to turn to evidence of two different types. The first consists, paradoxically, of the vigor with which a concept of poi¯esis, or making (in contrast to the more inclusive terminology of SONG, such as mythos and khoreia) is advanced in ancient treatises on literature and poetics and with which the boundary between poi¯esis and other related cultural practices is policed. The second consists of the vocabulary of verbal performance characteristic of Latin (in contrast to the almost exclusive emphasis on Greek that marks the work described heretofore). The continuing invalidation of musical performance that characterizes elite commentary from Aristotle onward together with the persistence of a lexicon that valorizes a different set of distinctions suggests, if nothing else, that the emergence and maintenance of a category of the literary was a locus of ongoing struggle in antiquity, as in later periods.

Aristotle, as we have already seen, by and large excludes the musical, religious, and spectacular elements of tragedy from his account of the means through which katharsis is generated, all the while acknowledging their presence in the production of tragedy. Horace, while arguing for the privileging of contemporary poets in the cultural environment of the Augustan revolution, nonetheless decries developments in musical performance as the result of an unseemly luxury (Ars Poetica 203, 211, 214ff.). Yet recognition of the musical nature of tragedy is carried in the very language with which Horace describes the performances he seeks to shape: "if you lack a (paid) applauder waiting for the curtain who will stay seated until the singer declares, 'You all applaud,' you'd better match the manners of your characters to their ages" (Si plosoris eges aulaea manentis et usque/ sessuri donec cantor 'vos plaudite' dicat,/ aetatis cuiusque notandi sunt tibi mores...; Ars 154-156). The reference to a professional applauder places the discussion in a contemporary Roman environment, and the use of the term cantor to describe the one who makes the final announcement betrays the extent to which drama is still song.13 Later in the Ars, Thespis, the legendary inventor of tragedy, is presented by Horace as having "carried about in wagons his poems, of the sort they used to sing and act after smearing their faces with wine-dregs" (dicitur et plaustris vexisse poemata Thespis/ quae canerent agerentque peruncti faecibus ora; Ars 255-256). Following his Aristotelian predecessors Horace emphasizes an understanding of poetry as "making," yet the language he uses carries with it a recognition of poetry's embeddedness in the category SONG.

Oratory, perhaps because its performance context so closely resembles the occasions that call for public song, is carefully policed throughout its history to prevent its reverting to the musical incantation from which it—and the terminology for its study—evolved. Thus Aristotle at Rhetoric 3.9.6 warns against producing prose periods of such length that they come to resemble dithyrambic preludes (anabolai). In his Orator Cicero recommends use of the full range of vocal modulation in order to take advantage of the characteristics of SONG to be found in speech-making (in dicendo quidam cantus obscurior; Orator 57), and he devotes about a quarter of the treatise to a discussion of rhythm. Yet he is careful to differentiate the musicality of Roman oratory from the so-called epilogus or epilogue of Phrygian and Carian orators, which he likens to the non-iambic songs, or cantica, of Roman comedy. Indeed, the whole of the Orator is prompted by attacks on Cicero for the resemblance of his oratory to that of Asianic orators with their strong emphasis on elements of SONG usually repressed in oratory. While Cicero does not expressly link rhythmical oratory to SONG, it is interesting to note that he regards the former as the more natural form of expression, with Atticist lack of rhythm a self-conscious affectation (Orator 170). Both Cicero and his successor Quintilian refer to the charges exchanged by the fourth century Greek orators Aeschines and Demosthenes concerning vocal inflection, but only to dismiss them as personal attack and not as a basis for disregarding such practice altogether.14

In introducing his discussion of the proposition that style reveals character, Seneca the Younger states that it is worth inquiring why sometimes inflated rhetoric is popular while at other times oratorical performance is effeminate and made to resemble cantica, or songs (ut aliquando inflata explicatio vigeret, aliquando infracta et in morem cantici ducta; Epistulae Morales 114.1). The ensuing discussion makes it clear that it is the latter fault that particularly concerns Seneca, for he castigates the Roman politician and artistic patron Maecenas both for his dissolute life and for his licentious language. One reason Seneca makes such a fuss over Maecenas (who was dead before Seneca was born) is that his behavior was emulated by later political figures, perhaps including Seneca's estranged pupil Nero. But Maecenas is dangerous to Seneca's broader readership as well, since to read his works is to re-present him. To the modern reader Seneca's attack may seem to anticipate that of contemporary cultural critics who fear that exposure to inappropriate literature or art may lead to imitation on the part of the suggestible reader or viewer. In fact, his critique harks back to Plato, who objected to certain kinds of music and poetry because the experience of such was inevitably mimetic. 15 To "read" Maecenas is to perform Maecenas and thus to reproduce the undesirable characteristics of the man. Even as he seeks to fortify the boundary between oratory and the rest of SONG, Seneca exemplifies the extent to which oratory—and indeed the study of oratory—is itself a performative activity, i.e., a subdivision of SONG in the sense described by Nagy.

The word used in several of the preceding passages to refer to the unseemly song to be avoided by orators is canticum, which is also the word for the non-iambic portions of Roman drama. As Nagy has noted, in Greek drama, the difference between iambic and other metrical passages "imitates the real-life opposition of SONG vs. speech" even though both are subdivisions of poetry which is itself a subdivision of SONG.16 In Latin the word canticum is built upon the same verbal root as cano, the verb frequently translated as "to sing," and it is with consideration of this lexical complex that we can commence the second part of our argument concerning the persistence and/or re-emergence of patterns of verbal production that transcend a simple dichotomy between literary and non-literary. My aim in this section of the discussion is to suggest—and it can be no more than a suggestion—that the Latin language itself enacts, in a different historical, political, and social context, distinctions and relationships strikingly similar to those described by other scholars as characteristic of archaic Greek. The terms to be examined here cluster around three verbs: cano "sing," dico "declare," and loquor "speak."

Loquor is the Latin word for everyday, unmarked speech. This definition is confirmed both by usage and by the testimony of several ancient authors, who seek to differentiate it from dico, which they describe—somewhat misleadingly, as we shall see below—as the word for oratorical speech.17 As the verb used of everyday speech, loquor can be modified with an adverb referring to the language in which the speech is carried out, e.g., Latine loquitur, meaning "she speaks Latinly," vs. Punice loquitur, "she speaks Punicly." What might be described as metaphorical uses of loquor in fact retain the general sense of unmarked communication, which is then made more specific by identification of the source or semiotic system. Thus users of sign language, wind blowing through trees, and a musical instrument can all be said to engage in the activity referred to in loquor. 18 As Ernout notes, compounds of loquor and their derivatives, such as elocutio or praeloquor, have a more specialized meaning in the context of ancient rhetorical studies than does loquor, but this feature is due to the fact that compounds of dico (e.g., edico, praedico) were already in use to describe socially authoritative linguistic functions when Greek rhetorical terms underwent linguistic and cultural translation at Rome.19

As noted above, Cicero and Quintilian differentiate loquor from dico in order to enforce the privileged position of rhetoric and its authority within Roman society. In fact, a more basic distinction is to be noted between loquor and cano, i.e., speaking and singing. A narrator in Apuleius' Metamorphoses remarks of a character that "she orders the cithara [a musical instrument] to speak: it starts to be played" (iubet citharam loqui: psallitur; Apuleius, Met. 5.15) Notice the difference between this application of a word for speaking to a musical instrument and Varro's statement that when trumpets and horns give the signal (in battle), they are described as "singing" (ut tuba ac cornua signa cum dent, canere dicuntur; De Lingua Latina 5.99) or Pliny the Elder's observation that "flutes must be taught to sing" (canere tibiae ipsae edocendae; Naturalis Historia 16.171). Loqui describes the commencement of the cithara's customary and distinctive sound (i.e., it begins to speak) while cano describes the instrument's performance either in a context that renders its sound authoritative or in a manner that is to be differentiated from mere sound (presumably it does not take much instruction of either flute or flutist to get the instrument to make some version of its characteristic sound). The same distinction obtains when reference is made to human utterance: loquor describes the making of our customary and distinctive sound (i.e., speech); cano describes an utterance rendered authoritative in context and/or made special through application of art or skill.

This meaning of cano is perhaps clearest in its nominal derivative carmen (from can+men, like germen, seed, from the verbal root gen-, to come into being).20 The word carmen applies to a whole range of productions from what we call poems to incantations, recipes, and legal formulae. The first definition given in the Oxford Latin Dictionary in fact encompasses all others: "a solemn or ritual utterance, usually sung or chanted and in metrical form." 21 Historically, it is not even clear that a carmen needs to be comprehensible (in the way the locutio does) to serve its social function: scholars have raised legitimate doubts about the comprehensibility of the song of the Arval Brothers at later periods of Roman history, a ritual production that does not for that reason cease to be called a carmen or to function as one.

If cano and loquor refer to different kinds of utterances (i.e., marked and unmarked), cano and dico refer to the same kind of utterance viewed from different perspectives. Cano calls attention to the materiality of the utterance, its shaping into rhythmical form or its accompaniment by music, dance, or special procedure; while dico, also used of delivering an authoritative utterance, asserts its intended efficacy or its relationship to referent or interlocutor. Livy writes of "songs spoken at the general" ( imperatorem dicta; 39.7.3). Here as frequently elsewhere in Latin there is no tension between describing an utterance as a carmen and a dictum simultaneously. The point is that its shape is that of a carmen, its function that of a dictum. Such an understanding explains a passage like Aeneid 9.621, where, after reporting the abusive speech of Remulus against the Trojans in their fortress, Vergil states that the young Trojan hero Ascanius could not put up with Remulus "who was making such taunts with assertions and singing in dread manner"—talia iactantem dictis et dira canentem. Remulus' remarks about the effeminacy of the Trojans lay claim to being true (that is to say they are dicta) and they take the shape of disturbing incantation (i.e., he "sings" them). Dico's use to call attention to the validity or invalidity of the referent explains why it can be applied both to assertions, as in the contrast between facta (what really happened) and dicta (what somebody said happened), and to performative utterances, i.e., dicere sententiam, used for example of the Senate declaring its deliberate assessment of a matter. The relationship between speaker and addressee implied by dico is usually an asymmetric one, either because the subject of the verb has information or judgment to impart to the addressee (dic mihi, Marce Tulli — "Tell me, Marcus Tullius"), or because the subject is formally in an authoritative position vis-à-vis the addressee (as when the Senate pronounces its judgment, dicere sententiam) or because the subject seeks through language to change the pre-existing relationship, as may be the case with the phrase from Livy, carmina in imperatorem dicta, describing soldiers' jesting putdown of their general during a triumphal parade. In contrast loquor is more likely to imply parity or interchangeability between interlocutors: hence its frequent use with the word cum, yielding the common verb colloquor and noun colloquium referring to conversation. Condico, on other hand, far from referring to a mutual exchange, means "inform," "make subject of a legal action," "fix a contract."22

If dico and cano offer two perspectives on the same activity, then it should be possible to sing a dictum and to speak out a carmen. We have already given examples of the latter expression. As for cano, it can take as its direct object any of a number of lexemes, including dicta (authoritative utterances), laudes (praises), verba (mere words), as well as the rubrics for various more familiar literary genres.23 The reason why the interrelationship between cano and dico may seem surprising or unfamiliar to many readers of Latin today is that the masters of speech-making had a strategic interest in obscuring the relationship between the two while calling attention to the contrast between dico and loquor. To differentiate dico from loquor is to remind the reader/listener of the power that accrues to the one who can speak out, i.e., dicere, in a public setting, as opposed to merely talking, i.e., loquor. But to call attention to the relationship between dico and cano is to raise the possibility of challenge to the masters of specialized speech by the purveyors of SONG. Just as Aristotle wants to restrict the study of tragedy to the making of plot and fashioning of character, so Cicero, Quintilian, and others want to focus on the aspects of dicta that can be made to differentiate them from carmina, i.e., absence of other potential elements of SONG, such as musical accompaniment, ritual dress and action, pervasive meter, etc. In fact, it seems likely that what we regard as formal prose in Latin, especially oratory, developed by process of differentiation within carmina, much as Nagy regards Greek prose as developing by process of differentiation within poetry. The earliest remains of Roman oratory show close similarity in rhythm and cadence to the polymetric cantica of drama. 24 Later rhetoricians, as we have seen, worry over the boundary between oratory and cantica, as well they might, given the tendency even of mainstream stylists such as Cicero to deploy rhythm, melody, and incantation when it is to their contextual advantage. But we should no more take their nervous prescriptions as descriptions of actual practice or as guidelines for our own inquiry than we should limit our interest in tragedy to the elements whose importance Aristotle promotes. SONG starts out and remains the verbal medium of communal authority at Rome.

Classical literature as literature surely exists to the extent that later centuries read and interpreted the written remnants of antiquity through the protocols of their own literary institutions. As such classical literature is a product not of antiquity but of the late eighteenth century and of each generation that accepts comparable approaches to the study of verbal artistry. But if classical literature is a product of the present, or at least the recent (to a classicist) past, then there seems no justification for the exclusively literary study of classical literature in the context of a Classics department, which by definition limits itself chronologically to antiquity. More appropriate to a Classics department would be the investigation of special utterances differentiated from everyday speech through a variety of features. Such study would necessarily include meter and figurative language, as well as music, ritual, performance occasion, and social context. It might include the investigation of genres and systems of genres, but it might also find the intrinsic terminology of performance—perhaps especially in non-Hellenic contexts—to be a more useful indicator of historically relevant typologies. To be sure, it would leave itself open to the possibility that a concept closely akin to the still dominant view of literature can be isolated in antiquity. But it would also leave itself open to the possibility that the development of such a concept was little more than the spoilsport strategy of a self-serving elite. Some years ago E.D. Hirsch proposed a "stipulative definition" of literature as including "any text worthy to be taught to students by teachers of literature, when these texts are not being taught to students in other departments of a school or university."25 Throughout much of antiquity, literature consisted of nothing but the texts being taught to students by teachers of literature, and there were, in effect, no other departments. To re-impose, or rather to continue to impose, such a restrictive definition on antiquity is to take sides in an ongoing ancient struggle over social authority and to risk stifling what little sound the dead still make.

Thomas Habinek



(1)  E.D. Hirsch, Jr., "What Isn't Literature?" in What is Literature?, ed. Hernadi (Bloomington: Indiana University Press, 1978) 24-34, discusses the influence of Kant as an impulse toward aesthetic evaluation and mentions the waning of the classics as generating the need for an inclusive term for fine writing. R. Williams, Keywords: A Vocabulary of Culture and Society (Oxford: Oxford University Press, 1983) focuses on broader economic and social trends generating the category of literature.

(2) I. Illich, In the Vineyard of the Text: A Commentary to Hugh's Didascalicon (Chicago: University of Chicago Press, 1993).

(3)  A similar argument can be found in W.B. Carnochan, The Battleground of the Curriculum: Liberal Education and the American Experience (Stanford: Stanford University Press, 1993).

(4)  T. Todorov, Genres in Discourse, tr. C. Porter (Cambridge: Cambridge University Press, 1990) 71. Todorov says "poetry" where I have placed "literature," although I do not think the sentiment contained in my paraphrase significantly misrepresents his remarks in the chapter entitled "The Notion of Literature."

(5)  A. Ford, "Katharsis: The Ancient Problem," in Performativity and Performance, ed. with an intro. by A. Parker and E. Sedgwick (New York: The English Institute, 1995) 109-132, quotation from 113.

(6)  Ford 124.

(7) All quotes from 127.

(8)  The relevant sections of Nicomachean Ethics are 1.5 and 5.3ff. I have discussed Aristotle's treatment of arete in an unpublished lecture entitled "The Origins of Virtue in the Western Tradition."

(9)  See in particular lines 197-200 where Horace contrasts spectacle with writing and laments the inability of writers to be heard; also 214-216 where he asks Augustus to heed those authors who seek readers rather than spectators. The whole discussion is strongly political in tone, with Horace mocking the tastes of the "little people" (plebecula 186) and disparaging the tendency of certain art forms to encourage peasants to intimidate their betters (149-50). For fuller discussion see T. Habinek, The Politics of Latin Litérature: Writing, Identity and Empire in Ancient Rome (Princeton: Princeton University Press, 1998) ch. 4: "Culture Wars in the First Century BCE."

(10)  Works I have found most useful—part of a large and growing bibliography—include C. Calame, Les Choeurs des jeunes filles en Grèce archaique (Rome: Edizione dell'Ateneo & Bizzarri, 1977); C. Calame, The Craft of Poetic Speech in Ancient Greece, tr. J. Orion (Ithaca: Cornell University Press, 1995); F. Dupont, L'invention de la litterature: de l'ivresse grecque au livre latin (Paris: La Découverte, 1994); B. Gentili, Poetry and its Public in Ancient Greece, tr. T. Cole (Baltimore: The Johns Hopkins University Press, 1988); G. Nagy, "Early Greek Views of Poets and Poetry," Cambridge History of Literary Criticism, vol. I, ed. G. Kennedy (Cambridge: Cambridge University Press, 1989) 1-77; G. Nagy, Pindar's Homer: The Lyric Possession of an Epic Past (Baltimore: The Johns Hopkins University Press, 1990); and G. Nagy, Poetry as Performance: Homer and Beyond (Cambridge: Cambridge University Press, 1996). Also important to my thinking has been the tentative reconstruction of archaic Roman culture to be found in two papers by N. Zorzetti, viz. "The Carmina Convivalia," in Sympotica, ed. O. Murray (Oxford: Oxford University Press, 1990) 289-307 and "Poetry and the Ancient City: The Case of Rome," Classical Journal 86 (1991): 311-329.

(11)  On the social function of writing in the Greek world see D. Steiner, The Tyrant's Writ (Princeton: Princeton University Press, 1994); on writing and Latin literature, see T. Habinek, The Politics of Latin Literature (Princeton: Princeton University Press, 1998), esp. ch. 5, "Writing as Social Performance."

(12)  Calame 14. This teleological approach also characterizes Dupont, L'invention de la litterature. Despite her emphasis on the transformation of the literary system during the Roman period of antiquity, Dupont provides valuable evidence for the continuing significance of performance in the Roman world. See also her earlier study, L'acteur-roi (Paris: Société d'édition "Les Belles Lettres", 1985).

(13)  For a plosor, or plausor, as an applauder arranged for in advance, see Suetonius, Nero 25.1 and Quintilian, Institutio Oratoria 12.9.4.

(14)  The passages in question are Cicero Orator 57 and Quintilian Institutio Oratoria 11.3.168.

(15)  On mimesis as re-enactment or impersonation of either character or composer, see Nagy, "Early Greek Views of Poets and Poetry" 47 and 63. Seneca's letter suggests that anxiety over the pleasures and dangers of re-enactment persists, even when the re-enactment consists of what we would call "reading aloud" alone or among a circle of friends.

(16)  Nagy, Pindar's Homer 47.

(17)  On the meaning of loquor, see A. Ernout, Dictionnaire étymologique de la langue latine (Paris: C. Klincksieck, 1979) 366. Passages that focus on the differentiation between dico and loquor are Cicero Orator 32 and Quintilian Institutio Oratoria 12.6.5.

(18)  See Ovid, Tristia 2.453, Catullus 4.12, and Apuleius, Metamorphoses 5.15 respectively.

(19)  Ernout, Dictionnaire étymologique 366.

(20) Ernout, Dictionnaire étymologique 100-101, following L. Havet.

(21) Oxford Latin Dictionary, ed. P.G.W. Glare (Oxford: Oxford University Press, 1992), s.v.carmen.

(22) Oxford Latin Dictionary, s.v. condico.

(23) For examples of each, see Varro, De lingua latina 7.70, Cicero, Tusculan Disputations 4.3, Ovid, Fasti 3.388, and Apuleius, Florida 20, respectively. On singing dicta, see also the passage from Horace, Ars Poetica cited earlier in the text in which a cantor is described as declaring an utterance (cantor 'vos plaudite' dicat, Ars 155).

(24)  T. Habinek, The Colometry of Latin Prose (Berkeley: University of California Press, 1985) 187-200.

(25)  Hirsch, "What Isn't Literature?" 34.

© 1998 Stanford Humanities Review unless otherwise noted.