Bryn Mawr Classical Review 2002.08.37
Lowell Edmunds, Intertextuality and the Reading of Roman Poetry. Baltimore: The John Hopkins University Press, 2001. Pp. xx, 201. ISBN 0-8018-6511-5. $42.00.
Reviewed by David J. Kuyat, Penn-Trafford High School
Word count: 1670 words
Lowell Edmunds has written a book that provides what is expected and appreciated in a theoretical study: the scholarship is extensive and well organized into arguments which are themselves descriptive, provocative, challenging, and supported by a close reading of a variety of selections from Catullus, Horace, Vergil, and Ovid.
Edmunds utilizes linguistic philosophies including speech act theory to develop his own literary theory of intertextuality, successfully incorporating Jauss' conception of reading outlined in Edmunds' earlier work, From a Sabine Jar: Reading Horace, Odes 1.9. Noting the great interest in intertextual studies in the last two decades, Edmunds seeks to address two fundamental problems in these studies: the intentionality of the poet (which he shows has still not been obliterated from criticism) and the status of the poetic text. Edmunds locates intertextuality in the reader who understands it as essentially an aesthetic device that is "pleasing or intriguing, often unordinary, uses of language that convey or portend some meaning valuable to the reader" (p. xiii). It implies a reader whose primary approach to Roman poetry, following Jauss' theory, is an aesthetic one. Unlike Conte and others, therefore, Edmunds considers that an allusion exists not in the text but in the reading of the text. In fact, the acid test for an allusion is that it is "really not there materially or linguistically" (pp. xvii-xviii), i.e., "there is nothing in addition to the alluding words that causes the allusion to be made... Reference, or allusion, has no linguistic or semiotic basis" (p. xvii).
In the first five chapters, Edmunds discusses various components of poetry -- text, poet, reader, persona, addressee. In each, Edmunds discredits textual elements that appear to offer an unchanging, verifiable phenomenon upon which to base and judge intertextual studies. In chapter one, "Text," Edmunds explores the meaning of the text itself, employing but refashioning Kristeva's theory of linguistics (immediately challenging those of us for whom Conte's theory of allusion as a trope is a manifesto) to assert that any "poetic text is in principle, not secondarily and occasionally, intertextual" (p. 13). But Edmunds does preserve Conte's notion of the text as a "work" against Kristeva's regard for the text as nothing more than the "product of its historical circumstances" (p. 14). It is in the reader, in hermeneutics, that the text is salvaged from indeterminacy, the reader, that is, who treats the poem as an aesthetic creation with its own (however the reader interprets it) individuality. Already in challenging the boundaries of the text itself and restoring them through a readership, Edmunds has laid the groundwork for the reader-based approach to intertextuality, given full treatment in chapter eight.
In chapters two ("Poet") and four ("Persona"), Edmunds moves to ascertain whether or not the poet can be the irreducible element upon which to base a theory of intertextuality. Citing a variety of speech act theorists, Edmunds concludes that "it is ... only for as long as poetry is oral and performed that poems can be considered speech acts. Writing is the turning point... Thereafter the poem replaces the performer" (p. 29). The intentionality implicit in a speech act cannot and should not restore authorial intent to any discussion of intertextuality. In discussing the poetic persona, Edmunds distinguishes between the poet as the speaker of the poem, the poet as the speaking poet's persona, and the implied poet. All of these are proven to be fictive, while the implied poet appears as close to the historical poet as one may come. Of particular interest in this chapter is Edmunds' reading of Ovid's retelling of the Aeneid in books thirteen and fourteen of the Metamorphoses, a reading utilizing Gennette's three levels of narration: poet to reader, character to character, and narration within a character's narrative (i.e. subnarrative). The most extreme cases of intertextuality seem to occur when one character is narrating to another what was narrated to him, i.e. subnarration. The point here is that the further the reader appreciates these levels, the less likely it is that any poetic persona, the ego-speaker or a character, can be used as the objective standard for interpretation.
In chapter three, "Reader," Edmunds maintains that in any reading of Roman poetry one cannot avoid being a reader-interpreter in spite of scholars who seek to posit a Model Reader, i.e. a philologist reader who is detached and possesses the text's true meaning. As Edmunds points out, even the Model Reader is a construct garnered from a particular interpreter's individual reading of the text, i.e. through hermeneutics. For Edmunds, as poetry becomes a written phenomenon to be read and not recited (a phenomenon for which Edmunds gives ample evidence in chapter seven, "Reading in Rome, First Century B.C.E."), the performer is no longer the poet but the reader who with each reading fine-tunes, as it were, his performance of the poems read, knowing that he may alter his own performance at a later date and be altered by the readings of performers to come.
In chapter five, "Addressee," Edmunds uses a dialogue format to argue against Citroni's conception that the authenticity of a text's meaning lies in the discovery or recreation of the original addressee. Of the many addressees to whom the poetic persona speaks -- himself, a patron, a divinity, etc. -- which should be privileged? Furthermore, Edmunds proves that the addressee of a poem "becomes more obscure as time passes; the personal relationship between poet and addressee into which ... the reader is called to enter disappears; the personal meaning of the [poem] fails; and yet the [poem] remains meaningful and enticing" (pp. 89-90). As Edmunds constructs his argument within a dialogue structure, he poignantly calls into question the legitimacy of giving primacy of position to any of the so-called historical personages addressed within a poem: Edmunds is speaking with both Citroni, a man he has never met, and Heinze, a man deceased. If Edmunds did not tell us this, we might inaccurately take the personal discussion among these three men as reality and not as a fictional construct. How then can we take it as given that such a reality exists between poet and addressee in Roman poetry?
In chapter six, "Possible Worlds," Edmunds utilizes the possible worlds theories of the speech act theorist Levin and the logician Kripke to illustrate the incompleteness of the poetic text, a condition that can be made comprehensible only by the reader, i.e. neither the text itself, the author, his personae, nor his addressee offers objective, substantiated boundaries upon which to base interpretation. The Augustan poet, for example, writes us into a possible world as shepherds (pastoral), lovers (elegy), etc., while simultaneously offering this world as a counterfactual one from which the poetic persona or poetic characters feel alienated even as he or they are described as existing within it. For Kripke, possible worlds were like the 36 ways in which a throw of the dice might come out, only one of which is the actual world. Edmunds maintains that poets have a way of offering a plurality of possible worlds whose collision in a poem manifests itself in a certain incompleteness that can only be made complete by the reader. No ancient possible world can be invented either by any resemblance between a modern reader's "world of reference" and an ancient reader's or by reconstructing the ancient reader's perspective (p. 103). Even as the re-reading of texts makes the discovery of other worlds of reference possible, the reader is ultimately prevented from any complete comprehension of them all.
In the final chapter, "Intertextuality," Edmunds develops his own reader-based theory of intertextuality. To summarize, an allusion is a text's (T1) quotation (Q1) of the words (Q2) of another text (T2), one which gives a new context (C1) to the context of the quoted text (C2). C2 is evoked in one of three ways: 1) through an expanded context in the quoting text; 2) through a continuously assumed context of the quoted text like the context of Homer's epics in the Aeneid; or 3) through parody in which the text "repeats or closely follows" another text (140). Edmunds also explores the process of Bakhtin's theory of dialogization when a text alludes to the language of a system, "verbal categories, literary and nonliterary, larger than single texts" (p. 143). Systems briefly explored are words from Roman law, social institutions, and business, as well as myth and genre. In allusions there is an inherent uncertainty as to whether the poetic meaning of the non poetic/mythic/generic words supplants, succumbs to, or integrates the meanings of the system words. As each meaning asserts itself, the text creates a dialogue in which "neither side has the last word" (p. 145). Thus, intertextuality presupposes a degree of undecidability that cannot be resolved through inherent intertextual signs or markers, primarily because even if such signs exist, their existence depends on whether a reader recognizes them or not.
Edmunds' major accomplishment is the systematic way in which he presents and integrates or rejects various linguistic theories in setting forth his own theory of intertextuality, filling a lacuna in the field of intertextual studies without being verbose even when confronting abstruse concepts. Theoretical studies, even those that deal with theories of reading, do not always offer a close reading of the text. The readings offered in Edmunds book are notable for their accuracy, thoroughness, and creativity, and as such they stand alongside the likes of Conte, Thomas, Farrell, and Hinds. His theory does nothing to jeopardize the significance of their readings. What Edmunds proves is that all critical interpretation of intertextuality is reader-based, an admission that many of the best interpreters refuse to make. We require a locus or at least a boundary; we fear uncertainty as much as solipsism. Edmunds shows that many of these boundaries are ambiguous, and all are subject to the particular reading of the interpreter. But critical chaos is avoided because the "interpretive community ultimately decides on validity, and classics is that community for the study of intertextuality in Roman poetry" (p. 168).