Bonifazi, Anna. 2012. Homer's Versicolored Fabric: The Evocative Power of Ancient Greek Epic Word-making. Hellenic Studies Series 50. Washington, DC: Center for Hellenic Studies. http://nrs.harvard.edu/urn-3:hul.ebook:CHS_Bonifazi.Homers_Versicolored_Fabric.2012.
Introduction. The Evocative Power of Word-Making
[In this on-line version, the page-numbers of the printed version are indicated within braces (“{” and “}”). For example, “{69|70}” indicates where p. 69 of the printed version ends and p. 70 begins. These indications will be useful to readers who need to look up references made elsewhere to the printed version of this book.]
versicolored, adj. archaic
1. changing from one color to another in different lights.
Oxford American Dictionary and Thesaurus
versicolorate = versicoloured
… when a surface changes its colour as the light varies.
Oxford English Dictionary
uersicolor, ~ōris, a.
Having colours that change (in time or, more usually, according to the angle from which they are viewed).
Oxford Latin Dictionary
The Latin verb vertere means “to turn,” “to change,” or “to address,” while the adjective versicolor, -ōris means “iridescent.” Iridescence “refers to the property of some surfaces to change colour with viewing angle.” [1] The archaic English adjective “versicolored” maintains the same semantic meaning. The vertere-component is crucial: an object turns out to be multicolored not simply by virtue of the different colors of the fabric, but by virtue of what happens once the object is turned, or once the viewer changes position, or both. [2]
I have chosen the visual image of a versicolored fabric for the title of this monograph because of this very vertere-component. The same object (or part of the object) can be seen by one person as red, by another as blue, and by yet another as yellow, without contradiction and without incurring a charge of falsification. One viewer can say “I see it red,” another can say “I see it blue,” and both can be right. The crucial distinction is in which direction the object has {1|2} been turned, and/or from which viewpoint it is observed. As the conclusion of my monograph will point out, various readers looking at the very same text, can see macroscopically different phenomena and can, as a consequence, be led to macroscopically different readings. It depends on how the text is used.
Such is the core idea that I am proposing with regard to the communicative and literary significance of some linguistic features of early Greek poetry—primarily, the Homeric poems. The grammatical features chosen as the starting point of my analysis are third-person pronouns and particles/adverbs. Their recurrence is accounted for in relationship not only to the clauses containing them but also to the narrative units including those clauses. My approach draws ideas and concepts from current studies in pragmatics, discourse analysis, and cognitive psychology. [3] The underlying view harmonizes with some innovative and relatively recent contributions that adopt linguistic approaches to ancient Greek literature. [4] The present investigation focuses upon one basic question: what advantages are producers and receivers of the text [5] thought to share through the use of various third-person pronouns and of various particles? Metrical constraints certainly play a role; however, to the extent that meter “is not an external constraint, independent of discourse,” as Bakker argues, [6] the linguistic competence of the producers must have permitted convenient choices at multiple levels—that is, the lexical, the syntactic, the semantic, and the pragmatic, with all of these levels being combined in the metrical outcome of epic telling. By “convenient,” I mean apt, effective, and economical. The present work specifically delves into one of these levels, namely, the pragmatic. The leading purpose is, therefore, to inquire into the pragmatic convenience of using some third-person pronouns and some particles. In a later subsection, I will underscore the relevance of pragmatic approaches to literary texts. Since some notions and concepts that the analyses rest on—such as, for example, {2|3} the term “pragmatics” itself—might not be familiar to classicists, I will first introduce them.
Background notions and concepts
A short definition of pragmatics is “the study of the use of context to make inferences about meaning.” [7] This branch of linguistics gathered momentum thanks to bold insights in semiotics and in philosophy in the first half of the twentieth century. [8] Pragmatics has become a widely acknowledged field of study, particularly after the appearance of seminal works by linguists such as John Austin, Paul Grice, and John Searle. [9] The idea that semantics cannot account for the workings of human linguistic communication developed, in part, in response to discussions within formal logic concerning the so called truth-conditions of statements. For example, the connective “and” in logic conveys symmetry (p and q = q and p): “the Brazilian flag is yellow and green” is as true as “the Brazilian flag is green and yellow.” However, linguists had, for many years, discussed innumerable cases of asymmetric “and,” such as “Smile and the world will smile with you,” where “The world will smile with you and smile” does not satisfy any truth-condition and does not make sense at all. [10] In a radically pragmatic view, linguistic messages convey meanings only within communicative settings that make the messages have some sense for somebody in some context. Situating a statement gives depth to that statement, as it is something being produced (which raises questions of the source, of the underlying intentions, and of the intended context of the message) and being received (which raises questions of implied meanings, of the balance between old and new information, and of the effects that are to be accomplished).
Communication—in any combination of linguistic, paralinguistic, and extralinguistic communication [11] —is assumed to be a “joint action,” to borrow H. H. Clark’s terminology. [12] Each time the present work refers to the communicative usage(s) of a certain word, the underlying idea is that the producer of a {3|4} linguistic expression communicates something that does not rest solely upon its logical, syntactic, or semantic value. For example, a certain amount of knowledge shared by the participants to a communicative activity allows a speaker to use words xyz in such a way as to permit the interlocutor (or the listener) to infer meaning B, even though the literal meaning of xyz is meaning A. [13] The joint action consists in the negotiation of meaning that is constantly launched and adjusted in order to have a successful interaction. Let us take an example from Homer, [14] which illustrates that communication may be successful if it is situated and if the participants share a sufficient amount of knowledge and are able to interpret the implied meaning. Towards the end of Iliad 1, Hephaestus tries to calm his mother Hera, who has just had a wrathful exchange with Zeus (573–594); he points out Zeus’ inevitable, dreadful revenges and says:
εἴ περ γάρ κ’ ἐθέλῃσιν Ὀλύμπιος ἀστεροπητὴς
ἐξ ἑδέων στυφελίξαι· ὃ γὰρ πολὺ φέρτατός ἐστιν.
ἐξ ἑδέων στυφελίξαι· ὃ γὰρ πολὺ φέρτατός ἐστιν.
Iliad 1.580–581; Hephaestus to the other gods [15]
For if the Olympian, god of the lightning, should have the intention
to hurl [us] out of our seats, … He is far too powerful, indeed.
to hurl [us] out of our seats, … He is far too powerful, indeed.
The lack of any apodosis—which probably corresponded to a gesture of substitution by the performer [16] —along with the lack of any accusative (for the ones whom Zeus might hurl) attest that communication (both linguistic and extralinguistic) was meant to be successful in that form; there was no need for more words—or, more words would have been less effective. This holds not only for the internal setting of the assembly of the gods, but also for the external setting of the performer impersonating Hephaestus in front of an audience.
“Use” and “force” are central terms in pragmatics, as they spotlight a simple but essential fact: namely, that any segments of written, as well as of {4|5} spoken, communication are produced and performed as acts. [17] The fundamental assumption of any pragmatic perspective is, indeed, that whenever we intentionally communicate by words, we do something. The use of language as a joint action is something underlying even written “non-personal settings” (such as news in a newspaper) and “mediated settings” (such as The Brothers Grimm writing down a folktale), as H. H. Clark explains. [18] The force of what is said concerns the point of the communicative act, as well as the effects in the reception of the utterance. [19]
Further crucial notions are those of discourse and discourse acts. “Discourse” may be taken in a broad sense as “actual instances of communication in the medium of language”; [20] as such, it is studied not as an abstract system but as a process. [21] We may have either spoken or written discourse; discourse embraces the complex of communicative intentions and purposes by the discourse producer and of meanings inferred by the discourse receiver. “Discourse act” is borrowed from Hannay and Kroon, who define such acts as “distinct steps which the language producer executes as a result of a strategic planning in order to realize her communicative intentions.” [22] Discourse acts are strategic elements {5|6} that may include a predicate or not (for example, a vocative may correspond to a discourse act). Following Hannay and Kroon 2005, discourse acts can be said to tend to correspond to punctuation units in written code and to intonation units in spoken code, rather than to syntactic units. [23] An increase of discourse acts shapes parts or segments of discourse (or discourse units), which in turn shape the (general) discourse structure.
Methodological bearings
About thirty years ago, Traugott and Pratt (1980:20) stated:
As is the case with its other areas of application, linguistics is not essential to the study of literature. Certainly one does not need to know linguistics in order to read and understand literary works; and critical analysis has long been carried out without formal linguistic apparatus. However, linguistics can contribute a great deal to our understanding of a text. It can help us become aware of why it is that we experience what we do when we read a literary work, and it can help us talk about it, by providing us with a vocabulary and a methodology through which we can show how our experience of a work is in part derived from its verbal structure.
My monograph is intended to explore some elements of the Homeric verbal surface whose literary relevance has not been much appreciated. The specific linguistic standpoint is pragmatic. To associate pragmatic meanings with literary meanings is to underscore, on the one hand, that literary texts share fundamental communicative properties with non-literary texts and, on the other, that the “joint actions” [24] discernible through certain linguistic choices contribute to an understanding of their power in terms of artistry and of pleasure (τέρψις). {6|7}
My approach is comparable to that of Bakker and of Kroon, insofar as both adopt discourse perspectives in their analyses. [25] My work shares four major methodological premises with the studies of those scholars. First, the analysis shifts from the “what” to the “how” in the production of meaning. [26] Second, the principle of “deroutinization” in the reading of passages is constantly applied: [27] in order to track the discourse function of a particular utterance (or a word in an utterance), canonical or traditional interpretations are never taken for granted. Third, any remarks on poetic or literary behaviors are based on actual occurrences and recurrences of words (and often of single words or single features). Finally, the linguistic competence of the producer(s) of verbal messages is manifest in the choices, which are made from a multiplicity of options. [28] Fundamental insights from Bakker and Kroon have illuminated the methodological path of this study from its first beginnings: I have gained especially from Bakker’s elucidation of the “syntax of movement,” of “pretended immediacy” (together with its visual aspects), and of the discourse relevance of different particles in Homeric poetry; in Kroon’s work (which focuses on Latin), I mention especially her identification of general pragmatic frameworks for the definition of the discourse functions of particles and adverbs. [29]
As I have already mentioned, my investigation is guided by a fundamental question: what advantages do producers and receivers of the text share through their use of the words at issue? In sum, what I see behind their utterances is what I call “pragmatic and cognitive convenience,” that is, apt, economical, and effective ways to situate certain expressions in one or more contexts (on the pragmatic side) and to trigger or activate certain links in the mind of the {7|8} participants to the communicative event (on the cognitive side). [30] A major claim is that the pragmatic and cognitive significances the analyses shed light on may, in principle, be enjoyed by listeners as well as by readers. In other words, arguing about the pragmatic and the cognitive uses of some words does not correspond to qualifying those words as oral features per se; it suggests, rather, that some inferential processes might be (deliberately) entailed both in listening and in reading, and that much better effects of vividness and pretended immediacy are presumably accomplished in the hic et nunc of listening.
In accord with the pragmatic perspective of my work, utterances are always situated, which means that it is always possible to presuppose one or more individuals performing them, and that they make some sense in one or more contexts to one or more receivers, no matter how many performers, how many contexts and how many receivers the work itself produces. For this reason, as others already did, [31] I will always make explicit who is speaking to whom in my analyses of passages. As for the main narrator of the Homeric poems, instead of assuming distinctions at the narratological level, I will simply refer to a unifying notion—that is, the primary speaking ‘I’, which most of the time in the Homeric epic is put in an “elliptic shade,” in Nagy’s terms. [32] Independently of the identity of any performer or transcriber (of which there were many), the fundamental pragmatic property of each epic word under discussion is that there is an ‘I’ who was and is supposed to be the source of the Homeric utterances constituting the texts that the tradition has transmitted to us, even though the verbal traces of this “I” are rare. As for whom the primary speaking ‘I’ addresses, I will use an analogously unifying notion—that is, “audience,” which broadly summarizes the most typical and chronologically continuous target of the poems. [33]
My analyses are synchronic; they presuppose that “old and young in Homeric diction exist side by side, and two uses of a given expression between which a diachronic relation can be established are often, synchronically, simply different senses or uses of the expression.” [34] My discussions focus upon the {8|9} distribution of the words or phrases under examination. Their functions are considered with respect to local as well as to global segments of the surrounding text. By “local,” I mean relating to immediately preceding and following utterances; by “global,” I mean relating to higher-level links between different parts of discourse. [35]
Finally, there is an interpretative constant throughout my work: while my proposed readings suggest one major interpretation, I also indicate additional or ambiguous values. An openness to more than one interpretation is important not only because no reading is absolute, but also because poetry always leaves space for polyvalence, and Homeric epic is no exception. There might well be a deliberate component in the multifunctional use of words—and beyond what I have analyzed here. This multifunctionality of words might fascinate our modern reception, just as it might also have fascinated the “there and then” producers and receivers. Ancient τέρψις might well have included such an experience.
Overview of chapters
This monograph is structured in two parts. The first, including chapters 1, 2, and 3, forms a whole, as it articulates the trends and the significances of a notable macro-feature displayed in the Odyssey, namely the pronominalization of Odysseus as κεῖνος and as αὐτός. I claim that the frequency and the distribution of these two pronouns correspond surprisingly well to fundamental aspects of the hero’s dichotomies: presence vs. absence, appearance vs. disguise, respectability vs. social vulnerability, external vs. internal identity. The interlacing of occurrences and of uses also provides a contextual frame that permits excursuses about the use of each pronoun outside of the Odyssey and/or outside of the Homeric poems (cf. chapter 1 for κεῖνος and chapter 3 for αὐτός). The outcome of the comparison runs in both directions: for instance, insights on Odysseus as κεῖνος both inform and are informed by insights on Oedipus as κεῖνος.
Chapter 1 devotes particular attention to Odysseus as κεῖνος in the first half of the poem—where it seems to be a recurrent feature—along with the different aspects conveyed by such a choice. Conversely, chapter 2 takes into consideration the exceptional and repeated co-occurrence of κεῖνος and αὐτός with Odysseus as the referent in book fourteen. This remarkable fact caused me—when I was already at an early stage of research—to completely re-think the subtlety of the Homeric verbal messages at a more general level. The overall point of this chapter is to suggest innovative elements of this subtlety and how they might {9|10} work. Chapter 3 illustrates the intricate and playful Homeric bravura of alternating Odysseus κεῖνος and Odysseus αὐτός in the second half of the poem, in accord with the multifarious suppositions and assumptions by the characters—including Odysseus himself—about who Odysseus is.
Chapters 4 and 5, which represent the second part of the monograph, have in common the treatment of particles and adverbs deriving from αὐ- and from αὐτός. The analysis of the occurrences stresses aspects of continuity with respect to the Homeric uses of αὐτός and also provides readers with complementary significances that complete the range of their functions. However, the main novelty of this part consists in the introduction of the notion of discourse markers, which can be partially applied to particles and to other adverbs and conjunctions. Several differences of meanings are explained in terms of a general distinction—namely, between lexical items contributing to the propositional content of what is said and lexical items contributing to the discourse function of what is said. Even though these lexical items may have an ambiguous function (sometimes they work at both levels), the distinction is helpful, especially as it spotlights the link between the use of words and discourse features. In this sense, it may allow readers to better understand what a particle is for in a certain passage; it may allow scholars to identify sets of discourse functions that are regularly made explicit by some words or phrases; finally, it may allow translators to realize how the overall meaning of a passage can change by virtue of the presence of discourse features and to find suitable criteria for translation. A final treat shared by these two chapters is the inclusion of instances from lyric poetry. My purpose is to offer some contribution to the scholarly bridging of genres that share fundamental mechanisms of communication.
Chapter 4 is particularly extended because it includes the theoretical arguments upon which the following analysis of αὖ, αὖτε, and αὐτάρ is based. Furthermore, the words at issue represent an important constituent of the Homeric telling, especially in regard to some crystallized formulaic expressions. Chapter 5 shows the to-and-fro propositional, as well as non-propositional, values of αὔτως, αὖτις, αὐτίκα, and αὐτοῦ, which overall turn out to be strongly linked to those of αὖ, αὖτε, αὐτάρ, and αὐτός. As for their propositional meanings, different pragmatic and cognitive implications associated with their utterance will be highlighted.
General aim
The general aim of this work is to contribute to an update of the grammatical accounts of some words in accord with notions and concepts from contemporary linguistics that are applicable to Homer. This does not mean underestimating {10|11} the insights that have long been provided by dictionaries, lexica, and grammars. Rather, my intent is to expand upon those insights by making use of new theoretical tools and terminology. My proposed readings are forward-oriented, scholarly speaking: they do not represent the only possibilities, of course; in more than one case, they are simply a first suggestion about how to look anew at familiar Homeric words. And this is where the idea of the evocative power of word-making combines grammar and literature: more uses and more meanings ultimately serve the richness—sometimes obvious, more often latent—of the “dialogic interrelations” of Homeric epic. [36] More generally, the “how” underlying the production of meaning explored in this work might be applied to other Homeric phenomena and to the same phenomena in other ancient Greek texts as well. Despite the explicitly poetic side of my findings, the heuristics concerning the words that I have examined might also be relevant to a fuller understanding of anaphoric choices and of discourse markers in prose. The overarching idea of successful communication is, after all, a strong condition for the work of any literary composition.
The results of my textual analyses will, I hope, be of interest in multiple academic spheres, such as didactics, literary criticism, and scientific dialogue. At the level of teaching, explanations of the meanings of particles gain in range and precision. At the level of literary criticism, the pragmatic differences characterizing the uses of αὐτός and of (ἐ)κεῖνος, for example, shed more light on the standpoint of either the author or the internal characters. More generally, even seemingly less meaningful words can turn out to contribute to local as well as to global literary intentions. At the level of scientific dialogue, finally, my greatest hope is that there can (still) be ways of making the study of literature and the study of linguistics merge, so that the former might be enriched by the latter, and vice versa. Such an interdisciplinary endeavor is ultimately meant to break methodological walls that impede learning from advances in both fields. {11|}
Footnotes
[ back ] 1. Meadows et al. 2009.
[ back ] 2. The notion of “color” itself may not correspond to any univocal perception. “Colours are not intrinsic to objects in the physical world or to the neural processes in the visual system; rather, they are properties of the world taken in relation to the perceiver. Thus on the question of whether colours are intrinsic properties or relational properties, I side with the received view that they are relational” (Thompson 1995:177).
[ back ] 3. In an article published more than 20 years ago and titled “Thoughts on the Pragmatics of Ancient Greek,” Collinge identifies several sectors of potential analysis, “evidence for which should not be lacking in any sophisticated natural language. We should therefore hope to see, despite the lack of native informants and reactors, what sort of pragmatic behavior betrays itself in Ancient Greek.” (Collinge 1988:3).
[ back ] 4. Dutch scholars have taken the lead: see, among other recent publications, Bakker 1997a and 2005; Van Ophuijsen and Stork 1999; H. Dik 1995 and 2007; De Jong and Rijksbaron 2006; Rijksbaron 2007; Allan and Buijs 2007 (including new approaches to Latin texts as well). For an updated online bibliography concerning different linguistic approaches to ancient Greek, see the website <vkc.library.uu.nl/vkc/antiquity/knowledgeportal/Wiki/A%20Bibliography%20of%20Ancient%20Greek%20Linguistics.aspx> (“A Bibliography of Ancient Greek Linguistics,” maintained by M. Buijs).
[ back ] 5. I use “producers” and “receivers” as terms applicable to both performers/composers and listeners/readers; my reasoning will be explained in the next subsections.
[ back ] 6. Bakker 1997b:187.
[ back ] 7. Schiffrin 2006:316. Among the many handbooks of pragmatics that have been published in the last decades, I mention Levinson 1983 (the earliest one), Verschueren 1999, Grundy 2000, Mey 2001, Horn and Ward 2006, Huang 2007, and the online Handbook of Pragmatics (www. benjamins.com/online/hop/).
[ back ] 8. An early definition of pragmatics occurs in Morris’ Foundations of the Theory of Signs: “by ‘pragmatics’ is designated the science of the relation of signs to their interpreters” (1955:108).
[ back ] 9. J. L. Austin 1962; Grice 1975; Searle 1969.
[ back ] 10. See, for instance, Lakoff 1971; Schmerling 1975; Grice 1981; Carston 2002; Schiffrin 2006. The example is quoted from Schmerling (1975:211).
[ back ] 11. Bara 1999:25–55; for an application of this view of communication to Pindaric poetry, see Bonifazi 2004a.
[ back ] 12. H. H. Clark 1996:3.
[ back ] 13. A very famous example from the scholarly literature is the question “Can you pass me the salt, please?” (originally discussed in Searle 1975). Meaning A (that is, the literal meaning) calls for the answer “Of course I can,” but the point of the question is actually different. Meaning B is conveyed through xyz as xyz represents a conventional polite request. The answer to such a request is supposed to be a positive extralinguistic reaction (the gesture of passing the salt), possibly accompanied by a linguistic agreement such as “Sure,” or the like. The mechanisms underlying this simple case of interaction are explained in pragmatics by way of two crucial notions: “implicature” (from Lat. plicare meaning “fold”) and “indirect speech act.”
[ back ] 14. “Homer” throughout this monograph is intended simply as a metonymic reference to “Homeric poetry.”
[ back ] 15. All Greek texts are quoted from the online TLG. Unless otherwise indicated, translations are my own; rather than aiming at elegance, they try to render the force of the words under discussion.
[ back ] 16. Boegehold 1999:42.
[ back ] 17. The meaning of words basically rests on the use of words, as Wittgenstein argues in his Philosophische Untersuchungen: “Wie ein Wort funktioniert, kann man nicht erraten. Man muß seine Anwendung ansehen, und daraus lernen” (1953:109; italics in the text). “Force” is invoked by J. L. Austin (1962:95, 199, 121) to capture the “point” of a linguistic act as it is performed. The tripartite distinction concerning speech acts runs as follows: we may perform the act of using words with a certain “sense and reference” (which effects the “meaning” of those words); we may perform an act in using some words (which effects the “force” of those words, according to “the different types of function of language”); finally, we may perform an act by using some words (which effects “the achieving of certain effects” [Austin’s italics]). By force, Austin means the “illocutionary force” of what we say, which is what we achieve in saying. Illocutionary force is connected to the primary intentions of the speaker, on the one hand, and to the conventional aspect of language use, on the other hand.
[ back ] 18. H. H. Clark 1996:6–7.
[ back ] 19. Throughout this monograph, I will use “utterance” to stress the process and the act of uttering any sentence, either in spoken or in written form. Lyons (1977, II:26) chooses “an act of uttering” as the translation of French énonciation (as it arguably differs from énoncé), about which see Benveniste 1966–1974, II:79–88 and Ducrot and Todorov 1972.
[ back ] 20. Johnstone 2002:2.
[ back ] 21. “Discourse is a process of interpretation through which intentionality is recognized and a contextual connection is activated” (Widdowson 1995:164). For a discussion of different theoretical approaches to discourse, see Schiffrin 1994:20–43.
[ back ] 22. Hannay and Kroon 2005:121. This notion of “discourse act” draws from different traditions of scholarship, included the so-called “Geneva Discourse Model” (Roulet et al. 1985) and the “Functional Discourse Grammar” (Hengeveld 2004). I prefer this notion to that of speech acts for the following reasons: first, the identification of discourse acts does not depend on the presence of verbs; second, this notion of “discourse act” provides a scholarly clean-sweep of the debated and the misused applications of “speech acts”—including to literature—throughout the last decades.
[ back ] 23. In a future project, I will argue that ancient Greek particles mark discourse acts beyond the hyposyntactic articulation of texts.
[ back ] 24. In the previous section, I associated the notion of “joint activity” with that of “meaning negotiation.” The latter is understood by Widdowson as the matter of language use and is referred to as “the continuous process of plotting a position and steering an interpretative course by adjustment and prediction” (1990:105); see also Widdowson 2007:53–56. The readers’ co-production of meaning with the author of a written text is regarded as a “pragmatic act” by Mey in the following terms: “in reading, as in other pragmatic acting, it is the general contextual conditions and presuppositions that make any understanding (literary or otherwise) possible.” (Mey 2001:256).
[ back ] 25. See Bakker 1993a, b and c, 1997b and c, 1999, 2001, 2002, 2005, and Kroon 1995, 2004, 2005, 2007 and 2008.
[ back ] 26. Bakker 1997a:1
[ back ] 27. My use of “deroutinization” is a metaphorical transfer of Hopper and Traugott’s employment (1993:65), the latter being adopted by Bakker (1997b:186).
[ back ] 28. In Bakker’s view, the linguistic “outcomes” of verbal communication can be understood in terms of focuses of consciousness being activated in the mind one by one (1997b:44–47). Intentionality in Homeric linguistic choices is regarded as interconnected to memory constraints and to metrical patterns. Linguistic choices may be conscious or unconscious, more or less traditional, “routinizing” or “deroutinizing”; in any case, at the synchronic level, they reflect adherence to what Grice calls the “superordinate” cooperative principle of communication: “Make your contribution such as is required, at the stage at which it occurs, by the accepted purpose of the talk exchange in which you are engaged” (Grice 1975:47). The idea of pragmatic and cognitive convenience in linguistic choices will be discussed further below.
[ back ] 29. At a later stage of the writing of this manuscript, I had the chance to compare my results on some Homeric anaphoric choices with Kroon’s analysis of pronouns in Latin literature (Kroon 2009); the consonance is striking.
[ back ] 30. In this monograph, cognitive psychology is invoked together with discourse analysis and pragmatics when the analysis concerns mental activities involved in the use of language, such as the mental representation of a referent.
[ back ] 31. E.g. Bakker 1988.
[ back ] 32. The Homeric “I” “highlights the prototypical singer of tales, elliptically shading over an open-ended succession of rhapsodes in the lengthy evolutionary process of countless compositions-in-performance over time” (Nagy 2004a:173). The infrequent Homeric occurrences of an explicit “I” mark referring to the performer constitute moments in which “the singer steps out of the elliptic shade” (Nagy 2004a:175). By “‘I’ mark,” I mean any linguistic codification of the zero-point of utterance, that is, the “I, here and now” (first-person verbs, “here” adverbs, “now” adverbs, and first-person pronouns); Bühler calls it origo (1965:102–120).
[ back ] 33. For a grosso modo distinction between external and internal audience, see ch. 1, n2.
[ back ] 34. Bakker 1997b:189.
[ back ] 35. See Georgakopoulou and Goutsos 1998:888–889.
[ back ] 36. Bakhtin’s concept of “dialogic interrelations” will be addressed at the beginning of ch. 3.