the medium of social negotiation, “an important part of a corpus of attitudes, behaviors, skills and science that kept civil society alive.”30
For lyric poets and audiences, the relationship between physical voice and literary or textual voice is codified in the rhetorical device of ethopoeia. The earliest known description of the figure appears in Aphthonius’s fourth-century Progymnasmata, a late antique rhetorical handbook: “Ethopoeia is imitation of the character of a proposed speaker. There are three different forms of it: apparition-making (eidolopoeia), personification (prosopopoeia) and characterization (ethopoeia). Ethopoeia has a known person as a speaker and only invents the characterization, which is why it is called ‘character-making’; for example, what words Heracles would say when Eurystheus gave his commands. Here Heracles is known, but we invent the character in which he speaks.”31 In Aphthonius’s definition, ethopoeia is the umbrella figure for all invented speech, with distinctions for the ontological status of the speaker. The ethopoetic speaker is “known” (a historical or literary figure) and alive, the eidolopoetic speaker is known and dead, and the prosopopoetic speaker is invented. There are further three modes of ethopoetic speech: affective, circumstantial, or a mixture of the two.32
The Progymnasmata gained new popularity in the early modern period following its 1572 translation by Reinhard Lorich, and no evidence exists of its use during the medieval period.33 Yet grammatical and rhetorical treatises popular in medieval schoolrooms discuss the composition of ethopoetic speeches, often under one of the following figures: adlocutio, conformatio, sermocinatio, fictio personae, prosopopoeia.34 In a rare medieval use of Aphthonius’s term, Isidore of Seville offers the following definition:
We call that ‘ethopoeia’ whereby we represent the character of a person in such a way as to express traits related to age, occupation, fortune, happiness, gender, grief, boldness. Thus when the character of a pirate is taken up, the speech will be bold, abrupt, rash; when the speech of a woman is imitated, the oration ought to fit her sex. A distinct way of speaking ought to be used for young and old, soldier and general, parasite and rustic and philosopher. One caught up in joy speaks one way, one wounded, another. In this genre of speech these things should be most fully thought out: who speaks and with whom, about whom, where, and when, what one has done or will do, or what one can suffer if one neglects these decrees.35
Isidore’s definition expresses the idea that written representations of speech can vary according to both affective and social factors: status, occupation, gender, mood, and situation of address. The Rhetorica ad Herennium describes, under the figure of conformatio, “making a mute thing or one lacking form articulate, and attributing to it a definite form and a language or a certain behavior appropriate to its character.”36 Priscian’s definition of adlocutio (“impersonation”) is ethopoetic:
Impersonation is the imitation of speech accommodated to imaginary situations and persons.… Speeches of impersonation can be addressed either to particular persons or to indefinite ones.… There are simple forms of impersonation, as when one creates a speech as though he were speaking to himself; and there are double impersonations, as though he were speaking to others.… Always, however, be careful to preserve the character of the persons and times being imagined: some words are appropriate to the young, some to the old, some to the joyful, some to the sad. Moreover, some impersonations have to do with manners, some with passions, and some with a mixture of the two.37
Geoffrey of Vinsauf provides an example of the speech of a pope as a kind of “refining by dialogue” (expolitio per sermocinationem): “Oh how marvelous the virtue of God! How mighty his power! How great I now am! How insignificant I once was! From a small stock I have grown in a trice to a mighty cedar.”38 Elsewhere in the Poetria Nova, Geoffrey parodies the figure. The discarded tablecloth grieves, “I was once the pride of the table, while my youth was in its first flower and my face knew no blemish. But since I am old, and my visage is marred, I do not wish to appear.”39 With its focus on the particularity of experience, ethopoeia had a natural affinity for the rhetorical study of personal “attributes,” elaborated in Matthew of Vendôme’s Ars Versificatoria (1175) and the anonymous Tria Sunt (1256–1400).40
Schoolroom exercises in ethopoeia and related figures were practiced from late antiquity (Augustine won a contest for his speech voicing Juno’s rage at her powerlessness to keep Aeneas out of Italy) to the early modern period.41 In antiquity and the early Middle Ages, the figure was used to teach letter writing, and a few examples, in which ethopoeia appears among a broader range of rhetorical figures, survive from the later Middle Ages.42 This practice encouraged students, some of whom became poets, to think about evoking a character with a voice that conveys his or her particular experience of emotions, social codes, and mores. It is perhaps difficult for us, who largely expect literary voices to do these very things, to appreciate how specific a use of voice this is and, further, how it establishes a literary convention that synthesizes a character’s inner and outer life by means of his or her voice. Some of the most poignant passages in Chaucer’s long poems are ethopoetic interludes, from Criseyde’s lament for her lost reputation (“Thorughout the world my belle shal be ronge”) to Dido’s lament in the House of Fame.43 (This figure was considered particularly suitable for the representation of female grief.) Further, to a medieval rhetorician, the soliloquy and the dialogue existed on an ethopoetic continuum, the dialogue being a juxtaposition of alternating ethopoetic utterances.44
Ethopoeia, then, is a figure of represented speech that is at once subjective and objective, affective and circumstantial. It is inherently tactical because it improvises rhetorically on a set of known circumstances from myth or history. Moreover, the ethopoetic voice of medieval lyric is distinct from the voice of the “speaking subject” that characterizes post-Romantic lyric theory. As I discussed in the introduction, Hegel defined lyric as the genre that takes as its content “not the object but the subject, the inner world, the mind that considers and feels, that instead of proceeding to action, remains alone with itself as inwardness, and that therefore can take as its sole form and final aim the self-expression of the subjective life.”45 This model of voice was most cogently critiqued by structuralist and poststructuralist thought. In Voice and Phenomenon, Derrida critiques the Western metaphysics that identifies an “inner voice” of thought with self-presence. For Derrida, the figure of the inner voice creates the fiction of self-presence, but the very fact of figuring it as a voice puts thought into the realm of signification. That is, an “inner voice” of thought is subject to the différance and the fundamental instability of representation that Derrida would more famously attribute to writing.46 Yet as David Lawton has recently discussed, literary voice, especially in medieval literature, is qualitatively different. It creates what he calls “public interiorities,” which project subjective expression into the public, or social, realm. These voices are characterized by “unstable reproducibility”; they can be endlessly iterated but without any presumption of fidelity to an original.47
Scholars of medieval literature have long recognized the slippery relationship between lyric voice and speaking subject. Who is it who says “I”? Anglo-American critics have frequently answered this question with recourse to Leo Spitzer’s classic essay on the “I” as an “everyman”: “in the Middle Ages, the ‘poetic I’ had more freedom and more breadth than it has today: at that time the concept of intellectual property did not exist because literature dealt not with the individual but with mankind: the ‘ut in pluribus’ was an accepted standard.”48 Spitzer’s model accounts for what he sees as the free substitution of one “I” for another, especially in prefaces and poetry. This suggestion was later developed by Rosemary Woolf, in her description of the “genuinely anonymous” religious meditative lyrics whose plain style willfully resists the development of an individual poetic voice in favor of universality, and by Judson Boyce Allen, who describes certain medieval English lyrics as “sublimat[ing]” the individual ego in the lyric ego.49 Most recently, A. C. Spearing has argued that in medieval literature, the poetic “I” inheres not in the speaker but in the text. In lyrics in particular, “the ‘I’ is little more than an empty