mimesis, dihegesis, style: socrates' paraphrase of homer iliad 1.11-42 (plat.rep.393d8-394a6)

Mimēsis, Dihēgēsis, Style: Socrates’ paraphrase1 of Homer Iliad 1.11-42 (Plat.rep.393d8-394a6)2 Donna Shalev I. Introduction The famous passage in the third book of the Republic, in which Homer’s partly mimetic scene, including the exchange of words between Khryses and Agamemnon, and Khryses’ prayer to Apollo, is paraphrased in a dihegetic form, is treated by Plato as a reworking of a source. The ‘source’ is Homer’s version of the mythical meeting between the Trojan priest Khryses and the Greek general Agamemnon, as it is fashioned in verses 11-42 in the first book of the Iliad, in hexameter form, where Khryses’ appeals for the release of his daughter from captivity are rudely refused by Agamemnon in a high-handed, threatening manner. Within the drama of the Republic, Socrates carries out this exercise in transposition,3 in order to illustrate to an aporetic Adeimantus how the recounting of these mythical events sounds without resorting to mimēsis — i.e. without Homer’s usurping the voices of Khryses or of Agamemnon. This exercise in transformation has both a literary and a theoretical context. Following a recurring scene conveyed through familiar verbal and structural patterns in Socratic dialogue as wrought by Plato into literary form, the interlocutor (in this passage 1 Nowhere does Plato refer to the paraphrase Socrates is performing of this Homeric passage as (a later term, in fact), but what takes place in this exercise has much common ground with the subsequent ancient definitions of paraphrase (using forms of the verb ), e.g. Philo Moses II.38: ; (‘Yet who does not know that every language, and Greek especially, abounds in terms, and that the same thought can be put in many shapes by changing single words and whole phrases and suiting the expression to the occasion?’ Tr. Colson in LCL). See VIII.2 below for comments on Plato’s use of forms of in the context of the passage in the Republic. 2 Research for this paper was supported in part by the Israel Science Foundation (Grant 837/01). 3 Labeled ‘transcription’, as Genette refers to it (1980: 145; = ‘réécriture’ 1972: 186), as an exercise of the writer Plato, not the interlocutor and oral educator/moralizer Socrates. See VIII.2 below. 2 Donna Shalev Adeimantus) is caught out not understanding; Socrates offers to show him ‘in another way’, by rephrasing in a different mode (see VIII.2 below). The theoretical context is the discussion, in the Republic, of the dosage of mimēsis involved in varieties of dihēgēsis, i.e. narration, and the ideal dihēgēsis. These topics are full worlds in themselves; the theoretical excursus of Socrates from the end of Pl.R.391 and up until the paraphrase beginning at 393d has been analyzed thoroughly by de Jong (1987: 2-5); Laird (1999) invests an entire chapter of his monograph to the theoretical passage introducing the paraphrase; Bers (1997: 11-3) also focuses his discussion on the introductory passage rather than the wording of the two versions.4 The definitive commentaries on Plato and Homer by scholars of strictly philological orientation comment in passing, but do not fully and systematically discuss, the differences in language and pragmatics. 5 Independent studies of the wording6 — mostly emphasizing similarity between the versions in the Iliad and the Republic — are not mentioned in the commentaries. Adam notes in passing occasional changes in wording in his primarily philosophically oriented commentary to the Republic. Adam’s traditional perspective is epitomized in the following observation on the passage: “the paraphrase is accurate and Plato leaves nothing essential out… .” In the Budé edition of Émile Chambry and Auguste Diès the same attitude is expressed in a note to the translation. Genette had at his disposal innovative theoretical and critical tools.7 His interpretation of this passage suggests however that he relied on, or was influenced by, a traditional approach such as that of the commentary of Adam, or, more plausibly, of Chambry and Diès, as seen e.g. in (1) below, a translation from his famous Figures III: (1) Genette, Narrative Discourse [orig. 1972] Tr. Lewin 1980: 165: “The most evident difference is obviously in length… Plato achieves this condensation by eliminating redundant information (‘So said he’, ‘obeyed’, ‘whom Leto bare’), and also by eliminating circumstantial and ‘picturesque indicators’ …especially ‘along the shore of the loud-sounding sea’ …a detail functionally useless in the 4 See III.1 below, the discussions on style and E.g., recently, Murray for Plato (1996), and Pulleyn (2000) for Homer. 6 See, e.g. Ludwich (1885), Des Places (1935), Labarbe (1949), and, indirectly, de Jong (1985). Interpretations in a similar vein, but not entering into specifics of wording are made, e.g. by Lynn-George (1988) and Griffin (2004), among others. 7 E.g. pragmatics, speech act theory, and conversation analysis, to name but a few. 5 Socrates’ paraphrase of Homer Iliad 1.11-42 3 story… A useless and contingent detail, it is the medium par excellence of the referential illusion, and therefore of the mimetic effect: it is a connotator of mimesis. So Plato, with a sure hand, suppressed it in his translation, as a feature incompatible with pure narrative” Genette’s interpretation centres on comparing length of text, laying little emphasis on the significance and type of reduced elements: in this respect it is reminiscent of Adam’s interpretation. More specifically, some of his examples of “redundant” information include frames (‘so said he’, ‘obeyed’) which are not distinguished by him from descriptive elements and formulae typical of epic (‘whom Leto bare’).8 The examples given together by Genette in (1) may be explained along lines of generic convention, framing in different modes of speech, or other important categories. 9 The wording in Socrates’ version has not been exhaustively compared with the wording in the Homeric model, although Des Places has summarized the main features. 10 Through a detailed comparison of the wording in the two passages, I try to show that the transition from original to reworking entails a change not only in how things are said, but also in essence: grosso modo the information passed on is the same, but the differences in genre, in style, in language and in the attitude of presenting the information amount to a rendering, and the resulting omissions, additions and restructuring all draw attention to the differences rather than the similarities.11 Recognizing a change in essence affirms the dramatic context in which this exercise in paraphrase is carried out: the teacher explaining to his disciple by changing the format. II. Synopsis In the synopsis given in (2) below Homer’s text is typed with one hexameter to each line (on the left), while Plato’s text is distributed so that the content more or less corresponds to its parallel in the relevant hexameter. Additional spacing is provided between different chunks in the Homeric text (for example dividing 8 By contrast the treatment of Lynn-George (1988: 52f), e.g., explains and emphasizes the importance of the ‘shore of the loud-sounding sea’ as a ‘liminal space in the epic theatre’ and the interplay between silence and noise, not only in our scene, but as a recurring feature in epos. 9 See also reservations by de Jong (1987: 5). 10 Inspiration for a synopsis and an exhaustive comparison comes from the desideratum of Des Places (1935: 133). 11 This is also the interpretation — based on other criteria, e.g. of Griffin (2004). Donna Shalev 4 between turns at talk in direct discourse ( ), and the ‘words between the 12 speeches’ ( ), e.g. the space between synoptic running lines 8 and 9). Within these divisions, further spacing, already typed into the Homeric text, reflects breaks in content/phrasing/chunking (e.g. the space between synoptic running lines 11 and 12, within the section of oratio recta set off by spacing). For convenience and economy we will refer in the remainder of this article to running line numbers in the synopsis (abbreviated as syn.) rather than verse number for the Homeric text and Stephanus page and section number for the Platonic text. I have also used some graphic markings in the synopsis to |highlight some of the formal and structural similarities and differences between wording in the two texts: parentheses on either side mark words or expressions omitted from (or added to) the alternate version; parallel expressions which involve different vocabulary or idiom are typed in italics; those involving the same words, often the nouns in different case forms and the verbs in different mood or aspect, are typed in bold. Finally, the passages in oratio recta in Homer, and their oratio obliqua parallels in Plato’s paraphrase, have vertical lines near the running line numbers (Khryses’ entreaty at 9-13; Agamemnon’s abusive threats and demands at 18-24; Khryses’ prayer to Apollo at 29-34). (2) S ynops is Hom.Il. 1.9-42 Plat.rep.393d8-394a6 1 (e1) 5 I9 “( I I10 I11 I I(A20) 12 To adopt the terminology of Plato in R 393b7 f. I9 I I10 I11 I I12 Socrates’ paraphrase of Homer Iliad 1.11-42 I13 5 ” I13 I18 “ I19 I20 (e5) I18 I19 I20 I I21 I(A30) I23 I | 24 I I21 I22 I23 I | 24 " 25 26 25 (394) (A35) 27 I29 “ I29 I30 I31 I30 I31 I(A40) I33 I I34 I I I I ” In the synopsis, the first thing to meet the eye is reduction on Socrates’ part, a surface observation which is compatible with analyses such as those of Genette. On the left, in Homer, all those expressions appearing in parentheses are expressions that are not represented in the Socratic version. On the right-hand side, in Plato, all the expressions within parentheses are the expressions which do not appear in the Homeric version. Sweeping claims have been made about the nature of the changes in the wording: that they are ‘not essential’ (Adam), or even ‘useless detail[s]’ (Genette). Less categorical critics may deliberate whether these Donna Shalev 6 changes in wording are mere ‘nuances’ or rather more significant. In the version put into the mouth of Socrates (the “reducer”), on the right, are also added expressions (those in parentheses): If one engages in simple word counts, perhaps “Socrates reduces”, but in fact there is a more complex change, because Socrates also adds elements, as one sees in particular in the consecutive lines of the synopsis (=syn.) 14, 22, 24, 25, 27, 29, 31bis, 33. Most of these added elements constitute expressions for framing speech, mainly with verba dicendi.13 On the whole, framing, and in particular reports of speech acts, are more frequent in the Socratic retelling than in the Homeric model (see VI and VII below). The comparison between the two texts was made through very sensitive reading by Des Places (1935). Much of what I will describe below was observed or implied by him, with a different emphasis: I share Des Places’ concentration on the differences between the two texts, and his consistent comparison of whole phrases and their composition rather than single words, in vacuo, even when comparing elements of vocabulary or other word-choice. However, whereas Des Places concentrates on differences in idiom and vocabulary in his analysis, and ultimately his interpretation of a transition in style (poetry to prose and archaic to contemporary) in these terms (1935: 130), I hope to bring out the different mechanisms for connecting phrases and sentences (III.2), the differences in structuring the text (VI, VII), and the variants of direct and indirect speech which seem to underlie many of the “omissions” and differences in Socrates’ retelling. III. Style and cohesion III.1 labels and parameters of style Style is a very broad term and includes choice of expression on the levels of vocabulary, euphony, register, generic norm, form of delivery, the interplay of current language with real and virtually earlier forms, geographical variation, emotional tone, aesthetic and rhetorical aims, and formal criteria in syntax, structuring of a text, phraseology, and sequence and combination of phrases and sentences. A brief perusal in the indices of Lausberg under the terms style, stilus, oratio, elocutio, or sermo indicates the breadth of scope of the concept and its 13 ( With the exceptions of synoptic lines 27 ( ). ) and 32 Socrates’ paraphrase of Homer Iliad 1.11-42 7 rich internal variety − especially of the Latin term stilus. Worth noting by contrast is the relatively narrow purview of the Greek terms and , centering on sentence architecture and combination.14 The style of the passage from Homer (A11-42) may be labeled: “Homeric”, but it is also “epic”, “archaic”, “oral”, “poetic”, “pathos-oriented”, “mimetic”, “hexametric”, “stichic”,15 “periodic”16 etc. Griffin (1986) speaks in terms of a Homeric style (or styles) when contrasting the nature of expression in the speeches vis-à-vis the narration between the speeches (implicitly following the Socratic distinction between and at 393b7). The style of the passage from Plato (393d8ff) may be labeled “Platonic”, “prose”, “succinct”, “narrative”, “dihegetic”, etc. In fact, as Thesleff has shown (1967), Plato has not one, but many styles. Thesleff conceives of style (as set out in 1967: 26 ff) as a “sum of those features in a text that somehow strike the reader as being remarkable”.17 In (3) is Thesleff’s analysis of R 393d ff in his conspectus of the stylistic affiliation and detailed tally of formal characteristics and shades of style: (3) Thesleff (1967: conspectus): “393d-394a: CE An example: a transcription of Il. 1.1242 in prose without oratio recta. Socrates (2).18 The style is quite specific, deliberately compressed and matter-of-fact, with participial clauses and some oratio obliqua. Many Homeric words have been exchanged in order to obtain a prose vocabulary, but the equivalents are still manifestly literary (e.g. 393e, note and ).” also reflexive 14 Cope (repr. 1970) surveys the concept and terminology in his sections on period and style, with elaborate discussion of ancient sources. 15 In the sense that the unit of the verse forms a backbone for units of thought and structure. The ‘stichic’ style of Homer elsewhere is manifest when the verse serves as a foil for breeches of this unity, e.g. in cases of enjambement or (uniquely rare) transition into direct speech in mid-hexameter. 16 In the sense that Homeric composition features Ringkomposition, and other significant structurally motivated repetitions, as well as passages with framing conventions and anatomical features signaling their beginning and end. This is most salient in A11-42 in the conventions for framing oratio recta. 17 See ibid. for important comments and distinctions on the notion, phenomena and function of style. 18 Style (2) in Thesleff’s classification is a ‘semi-literary conversational style’, used by the speaker Socrates here. CE means continuous exposition (E) embedded within an environment of reported dialogue (C). 8 Donna Shalev I agree with Thesleff in his approach, and on many points (the “compressed” and “matter-of-fact” nature of the style, the presence of “participial clauses” and “prose dictionary”). Perhaps I would put more emphasis on polysyndeton, hypotaxis, and relative connection (elements of cohesion); on extra framing of indirect discourse; and on the mimetic elements and manifestations approaching those of free indirect discourse (f.i.d.) which are not entirely eradicated from Socrates’ dihegetic conversion (VII). I would also mention the interesting grammatical features such as nominalization (IV.2 below); ellipse of object (IV.1). Finally, by “specific”, I understand Thesleff implies that our passage (Plato’s transcription of the opening of Homer’s Iliad) is unparalleled or unclassifiable. 19 Even if this transposition is not classified as a myth, or some other of the typical narrative passages inlaid by Plato into his Socratic dialogues, with shared formal characteristics and structural roles, as classified by Thesleff, it is narrative by definition, as opposed to non-narrative, and does exhibit some traits affiliated by Thesleff to his ‘mythic narrative style (6)’,20 in Platonic passages such as the Gyges story in the Republic, and the myths in the Protagoras and the Phaedrus. As we shall see in coming sections, our focus on the elements of structure and cohesion in our analysis of style, following the Greek preoccupation with this aspect ( ), bring out the differences in the two versions, and the change in essence wrought by the transformation. The two scholars who hone in on the wording of Socrates’ exercise in transposition of the opening of the Iliad, Des Places, and to some extent also Labarbe, both analyze much of the compared and contrasted elements in terms of style. Some of their observations come under the broader province of phrasing and cohesion. Des Places (1935:133) contrasts in Plato with in Homer. He calls the prose version more ‘synthetic’,21 intending perhaps that it is more compact or economical. Des Places makes other comments in terms which may be subsumed under the rubric of cohesion. 22 See Thesleff’s remark on the exceptional nature of this passage (1967: 171). Most saliently, connection by (see Thesleff 1967: 74). 21 ‘Remarquer … le tour plus synthétique de la prose, qui se contente d’un seul participe’. 22 On p. 134, on Plato’s for Homeric , Des Places explains the use of the participle as contributing to the ‘unity of the phrase’. He observes (ibid.) 19 20 Socrates’ paraphrase of Homer Iliad 1.11-42 9 Thus, though Des Places makes these comments within the context of his main difference, the transition in Plato to a briefer,23 less vivid,24 contemporary,25 prose26 style, he has not entirely ignored the dimension of cohesion and the different devices adopted in the two versions. Labarbe (1949) hardly visits the province of style in his analysis of the two texts. Rather, important motivations for Labarbe’s discussion of citations of Homer are the textual issues of interpolation, athetēsis, variant readings, and the Homeric version at Plato’s disposal.27 In contrast to Des Places’ phrasal perspective, Labarbe’s discussion centres on a tally of words and expressions, and whether or not they are represented in the paraphrase. All the same, Labarbe contributes to our discussion on paraphrase (348 n.3), and, of special interest to us, Labarbe also speaks (1949: 349) of the importance of this transposition for getting a feel of how a classical Athenian read Homer.28 Encapsulated in Labarbe’s words are his focus on the chronological (“période classique”, “surannés”), the registral (“idiome de chaque jour”), and the generic (“poésie”) in his assessment of the interplay between the Homeric text and that of Socrates’ paraphrase. The quotation marks around “sentait” in Labarbe’s text reflect perhaps his conception of a more pervasive but less that Plato produces hypotaxis by introducing infinitive where Homer has parataxis at v. 29 (in his terms, subordination and juxtaposition, respectively). 23 in Plato (for the more ‘juicy’ expression in Homer); Plato’s (for the reminiscent, but more wordy expression in Homer). 24 The ‘essential’ in Plato (for the more ‘savage’ Homeric expression)  more (less ‘pictorial’ (imagé) than Homer). current in Attic prose style. Plato’s 25 in Plato (less touching than the Homeric alternative, but more typical (instead of the aorist of Plato’s period). The future optative in Plato, subjunctive form in Homer which was not current in 4th century usage). 26 in Plato (for the periphrastic expression in Homer, more ‘majestic’, and fitting (rather than the more poetic alternative in Homer). the epic genre); 27 E.g. with respect to A31, where Labarbe argues (1949: 356) against the current use of Plato’s transposition to prove that his version of the Homeric text did not include this verse.Labarbe reminds us of Socrates’ consciously approximative nature of transposition, citing his introductory words (393d). Cf. VIII.2 below. 28 ‘[La paraph. ci-dessus] constitue le plus intéressant témoignage que l’antiquité nous ait transmis sur la manière dont un Athénien de la période class. «sentait» Homère et repensait, à l’occasion, dans son idiome de chaque jour cette poésie aux vêtements surannés.’ 10 Donna Shalev clearly defined dimension intuited in the paraphrase.29 The “feeling of the text”, as it were, is conveyed not only by mere and their equivalents (“idiome”, vocabulary), but by , the style, in the broad sense, including texture. III.2 Sentence and phrasal connection, textual cohesion, and style To take up where Des Places begins, and at the same time to address the “feel” or “sense” of the transition (to borrow Labarbe’s expression), I will describe the details of the transition between the text of Homer and Socrates’ version in terms of structural and cohesive styles, or modes of expression − to put it broadly, different .30 A synoptic tabulation of the two sequences of phrasal and sentence connection (4) below, derived from the synopsis of the two texts (2), reveals the different methods of tying together phrases in the two texts: 31 (4) Synoptic line Homer Plato nos. 4-7 art.+ nom. sj. ptcp. + ind. ind. ind. 9-13: speech oratio recta oratio obliqua 10-11 nom.sj. opt of wish) inf.+ art.+nom. sj.(acc.) in inf.) ptcp. +art.nom.d.o., inf. nom.d.o.,32 inf. 12-13 inf.33 inf. , ptcp.+d.o. inf. , ptcp., art.+d.o. ptcp. 29 Similar notions with an emphasis on the influence of the time gap are expressed by Ludwich (1885: 485) about post-classical and Byzantine prose paraphrases of Homer: ‘…sie lehren uns einmal, wie die Griechen selbst in den einzelnen Zeiträumen die Sprache ihrer älteren Dichter verstanden.’ 30 In the Platonic corpus, the term is most often used to denote ‘form of expression’, whether as a style or diction. It is in fact used with most frequency and density in this passage (R.392ff) and further in book III, in the context of the discussion of the modes of speaking used by poets, whether the emphasis is on delivery (397b7), characterization (398b2) or production and form of presentation (397b1). Plato uses the term in the sense of ‘style’ familiar to us in Soph.225d8 and Apol.17d3 and 18a2. See also LSJ I.2 s.v., with reference to the passage from the Apology as well as to Lg.795e, and passages in Aristotle’s Rhetoric. 31 It must be kept in mind that the devices compared in the table are only those involved in expressions having parallels in either text. Other devices appear in the two texts, in phrases and expressions without counterpart, devices which ‘get lost’ in the comparison, e.g. direct object in Homer (syn.15), in Homer (syn.17), or the in , left out of account in synoptic line 25. Homer’s 32 The in 10 (A18) goes primarily with the addressees, in counterpoint to the girl Chryseis, with in 14 (A20). One cannot rule out some effect by on the first inf. , vis-à-vis the second. Socrates’ paraphrase of Homer Iliad 1.11-42 14-17:event 14 14-15 16 17 18-24:speech 18-19 20 21-23 24 25-28:event 25-26 27-2839 11 gen.abs.34 fin.v.d. + inf. inf. art. sj. fin. fin.35 fin. (‘logical’ sj. in dat.) sj. (prop.n.) fin. ptcp. [ ] fin. oratio recta oratio obliqua/f.i.d. sj. pers.pron. fin. (sbjv. of threat) now inf. later inf. now later ptcp. Sj. fin. (sbjv. of threat) fin. (opt.obl.) pron. d.o. sj. pers.pron. fin. conj. inf. (aor.pass.), art+nom. sj. (acc.), (fut.); adv. …fin. (fut.) inf. (fut.) 36 , d.o. fin. (impv.), …conj. inf. … inf.,37 conj. fin. (opt. obl.) fin. (sbjv) fin. nom.sj.,…38 fin. art.+nom.sj. …fin. ptcp. (praes., v.mot.) + ptcp. (aor.) … fin. fin. fin. + subj. 29-34: speech 31-3240 oratio recta If ever … ind., if ever …ind. 33 34 Cataphoric pronoun+ind+d.o. ind. (opt.wish), subj., possessive + 33 oratio oblique/f.i.d./rsa ever vbal noun (d.pl.)+gen.obj. vbal noun (d.pl.) + gen.obj. (opt.obl.) relative connective + postpos., ind. inf., subj. (acc.), ‘homeric’ possessive + complement, possessive + dat .instr. if ) alongside the The transmission here offers also a finite form (the optative inf. . See Pulleyn’s commentary and West’s recent edition for variae lectiones and their implications. 34 a framing expression with no counterpart in Homer. See further VI, end. 35 A rare instance of change in word order of content vis-à-vis its Homeric model. 36 , a framing expression not used in the Homeric There follows here counterpart, and the particle may be connecting this action to the previous, with connecting the first inf. ( ) and the second ( ). 37 Whereas in Homer, the expression, using the same vocabulary and structure — but in oratio recta rather than oratio obliqua — includes the direct object ( ), in the mouth of Socrates’ ‘strictly’ dihegetic prose narrativese, the expression features ) . See below (IV.1). ellipse of the direct object: ( 38 Note also the verbal expression with (in italics) — a phrase appearing in Homer only — and coming between the two verb phrases connected by : namely . 39 ) and in Plato ( ), The participles in synoptic line 27, both in Homer ( ) in the device termed , are lexical resumptions of the verbs in 26 ( . catena; see (13) below, with n. 47. 40 here ‘condenses’ the action expressed in Homer by specific Socrates’ verbs. Socrates adds the same detail in nominalized form (See IV.2 with (16) below). Donna Shalev 12 complement, possessive + dat. instr. The perspective gained from the synoptic comparison of the amalgam of syntactic and cohesive means used in the respective texts, as tabulated in (4) above, is a multidimensional picture of the sequence, distribution, and clustering of devices on either side of the synopsis, and the difference in cumulative stylistic effect. In this sense, it is not an absolute tally of particles chosen in Homer vs. those in Plato, but their dosage, distribution and composition. This bird’s eye view may be less useful than an integrative discussion of the choice (by Socrates) of different procedures for connection, coordination, and cohesion of syntactic and textual phrases, including the use of and “style ” sometimes but not always vis-à-vis (with passages 5-9 below), preference of participles (10-12) (including those in the catena structure (13)), hypotaxis vis-à-vis parataxis (14), and relative connection (15). Socrates’ use of and shades of “style ” In the mouth of Socrates, alongside a predilection for connecting participles 41 (see passages 10-13 below) and promoting the progress in the narrative, typical of Attic as a whole, we find a recurring choice of the pattern ‘verb verb’ where the Homeric text has , or some other connecting particle or device, such as a participle, as illustrated in the examples in (5-9) below: (5) synoptic lines 24-5 (Il. A 32-33; R. 394a1f): Homer: Socrates: / … Note that passage (5) occurs not in a speech, but in a narrative section (termed by Plato ). Two factors complicate this example: the phrase in Homer has no parallel in Plato, but may influence the (see e.g. Bakker, 1997: 72); the choice of particle subsequently used with in Plato is part of .The Homeric text in the passage paraphrased by So- 41 in Homer in synoptic In our passage alone, Socrates uses : (i) corresponding to lines 12, 21, 25, and 27; and (ii) departing from usage in Homer in synoptic lines 14 (with gen.abs., vis-à-vis Hom. ), 16 (Hom. ), and 24 (Hom. ? to connect, if not exclusively a prompt with ). For discursal and other values of viz., in Homer, see Bakker (1997: 62-74). Socrates’ paraphrase of Homer Iliad 1.11-42 13 crates does indeed produce one example (6, below) of two predicates expressed by two finite verbs connected by ( ) where the Socratic counterpart features a participle + finite combination ( ): (6) (synoptic lines 4-7) Homer (A 12ff.) Socrates (393d8-e1) But there are more instances where the text put into Socrates’ mouth involves the use of in paraphrasing some other form of phrasal connection in Homer: in (7) below, in addition to the transition from direct to indirect speech, is in lieu of adopted by Socrates cohesion through the use of participle uttered by Khryses R. 393e2-3): (7) synoptic lines 12-13 (Il. A 26-27; R.393e5): Khryses: Socrates: In passage (8) Socrates uses / instead of Agamemnon’s : (8) synoptic lines 18-19 (Il. A 26-27; R. 393e5): Agamemnon: / Socrates: These differences in passage (8) occur in a chunk of the narrative which also involves transposition from direct discourse to indirect discourse, and a change in distribution of negatives (cf. discussion on syn. 21 in notes to table (4)), and of connects two infinitives (in oratio verb semantics. In passage (9) Socrates’ obliqua for commands), whereas Agamemnon has two imperatives introduced without a connecting expression:42 (9) synoptic line 24 (Il. A 32; R. 393e8): Agamemnon: Socrates: 42 , formally an imperative, functions as a prompt for the command, In fact the first, similarly to other fixed forms such as (Schwyzer Griechische Grammatik II.583f). Interestingly, this seems to be taken literally as a distinct command ‘go!’ in Socrates’ conversion into dihegetic mode. 14 Donna Shalev The choice of (rather than, say, as in (5) above, for promoting the nar43 rative, is ultimately a stylistic one. It is true that and have different meanings, and do not entirely cover the same ground − these differences have been presented with illustration from short Herodotean passages in Ruijgh (1971: §§129ff).44 It is the free choice of the speaker or author, ultimately resting on style − says Ruijgh − whether to present these facts as one unit ( ) or two in Socrates correspond to in ( ).45 But as we saw, not all the instances of Homer, other s in Socrates’ version correspond to Homer’s (8 above), or to no connective (9). Although we will see in the next sections that Socrates uses other means of cohesion besides , we do not encounter many uses of in Homer (but see passage (6)). All this evidence indicates that, among other devices, the dihegetic, prose, Socratic style in this passage relies in no small measure on for connection and cohesion. This does not of course amount to ”, as described by Trenkner (1960), but one may safely claim that the “style this passage features a shade of the style.46 Socrates’ use of participles The participle is very much in use in Socrates’ dihēgēsis (as noted already by Thesleff in his conspectus, passage (3) above). Other than the example in (6) above (pitted against in Homer), in (10a-b below) the participle is preferred, 43 Similar sentiments, pointing ultimately to stylistic choice, are said to lie behind the varying usages of in contrast to (and other paratactic and hypotactic devices) in the different early prose writers, in the study of S. Lilja (1968: 75, 99). 44 and At the risk of treading over subtleties, I summarize these differences between as Ruijgh presents them: is (a) grosso modo analogous to French et in its contrast to the use of asyndeton in French; (b) accompanies the addition of a new fact, whereas combines two facts into one unit in the narrative; (c) coordinates phrases in a text, coordinates words in these phrases. whereas 45 See also Sicking – van Ophuijsen (1993: 11ff): for arguments against style as a main influence in distribution of and ; for a clarification of Ruijgh; and for an application to some Lysianic texts of their own view of the differentiae in the distribution of and (connection and discontinuity, respectively). 46 (which he terms copulative ) with his Interestingly, Thesleff associates this ‘colloquial style’ (1) and especially his ‘mythic narrative style’ (6). Trenkner herself affiliates with “style ” some of the passages mentioned by Thesleff in his style (6). Socrates’ paraphrase of Homer Iliad 1.11-42 15 twice, over a repetition of two finite or infinitive elements, paratactically connected with . (10a) synoptic lines 10-11 (Il. A 18-19; R. 393e1-2): Socrates: Khryses: (10b) synoptic lines 12-13 (Il. A. 20-21; R. 393e2-3): Socrates: Khryses: In the synoptic material reproduced in (11), the participle is preferred (in genitive and of the demonstrative must not be igabsolute − but the added force of nored) over the clause in Homer: (11) synoptic lines 14-15 (Il. A 22-23; R. 393e3-4): Homer: Socrates: In (12) Socrates prefers the (finite +) participle over Homer’s (finite +) (12) . synoptic lines 16-17 (Il. A 24-25; R. 393e5) Homer: Socrates: In this passage the style change involves also transition from Homeric tmesis ) to prose univerbation ( ), voice change, and shift from circumlocutory expression with explicit negative ( ) to a ). more contemporary, monolectic semantically negative expression ( Homer, and following him, Socrates, both employ a cohesion device involving repetition, in participial form, of the verb (or its lexical synonym) from the previous sentence. In both texts this is not immediately obvious, because rather than being genetic cognates of the finite verbs (Hom. , Socr. ), the participles are derived from other verbs ( respectively), with a close meaning, as we can see in (13) below: (13) synoptic lines 26-27 (Il. A 34-35; R. 394a2-3): / Homer: / Socrates: ’… On the one hand, the participle, the particle , and the lexical repetition in Socrates’ dihegetic paraphrase serve as cohesive ties; on the other, there is a transition from one event to another in the narrative. The use of this device, the catena, becomes more regular and visible in the later Greek prose paraphrases of Donna Shalev 16 Homer which pick up on this device. 47 Exceptionally, Aristeides’ paraphrase parts ways with the others, resorting to a temporal clause. Parataxis in the mouth of Agamemnon converted into hypotaxis by Socrates48 is illustrated in (14): (14) lines 21-22 of the synopsis (Il. A 21; R. 393e7-8): Agamemnon: Socrates: Socrates’ use of relative connection Connection of independent phrases or sentences through relative pronoun (‘relative connection’, ‘relativer Anknüpfe’ etc.) is the form of cohesion chosen , to introduce the final by Socrates, in the expression wish ( ), to convert Khryses’ apodotic phrase opening with demon), reconstituted in essence in (15): strative ( (15) synoptic line 33 (Il. A 41; R. 394a6): Khryses: / Socrates: ( /… / The phrases are not strict parallels (see also IV.2 and VI with (21)); the whole structure of the Homeric text here is organized as a conditional sentence, in fact part of a structure typical in prayer, and is in direct discourse, an ), whereas in Socrates’ version the notion is expressed as a imperative ( separate phrase, as a report of request ( ), after a report of the act of . The use of relative connection ties this sentence to the in Homer’s text (‘for previous one, offering an anaphoric effect just as does this’, i.e. for my previous favour). Relative connection as described by Kühner47 Psellos: ; Moschopoulos: ... ; Cod. Gazae: ... Cf. also the variant with substantivized infinitive (instead of participle) in ms. Ven. 454: 48 A contrasting passage is syn. lines 14-15 structured by Homer (A22-23) as a complex verbum dicendi + content infinitival complement: , but simplified by Socrates (R. 393e4) as two paratactically connected finite verbs (with word order transposed): Socrates’ paraphrase of Homer Iliad 1.11-42 17 Gerth (ii§562.2) is used in an array of text types.49 The use is attested in Homer (mainly Odyssey), so I would not say Plato chose it here as a device to distance Socrates’ language from Homer’s style, or from a distinctly mimetic style. IV Other differences Other differences, identifiable albeit on a grammatical or syntactic level, also reflect choice, possibly based on a conception of style, or in imitation of typical features of a recognizable mode of presentation: omission of direct object, nominalization, and change in voice and aspect of the verb. IV.1 Omission of direct object The direct object is implied in the context rather than spelled out at syn. 24 in the Socratic rendering (393e8) for Homer’s (A32). The other possible omission of direct object in our passage also occurs in the Socratic version. 50 From Homer on, in fact,51 it is typical of Greek continuous text (as well as tightly-knit exchange) involving the same participants, to leave out, if the context permits it, a direct object, even of a verb which requires one in the lexicon. 52 The effect of such omission (ellipse) is often one of 49 Among typical patterns of use are causal phrases after questions (K-G §562.2a), and addresses and questions (§562.2c). See also Sicking – van Ophuijsen (1992: 18 ff). 50 > (syn.lines 14-15). Socrates’ at 394a4 summarizes the request in Khryses’ prayer, and has no counterpart in Homer, which fully quotes the prayer in direct speech; thus, the lack of object for is an omission in the internal context of Plato’s usage, rather than a product of a conversion process. Finally, Homer’s direct object pronoun in A29 appears as a subject (in acc.) in Socrates’ version due to passivization; this is not an omission, strictly speaking. 51 This becomes less frequent in later Greek, under the influence of evolution of sentence patterns, still extant in Modern Greek, where the direct object is required for a grammatically acceptable utterance. The progress and regression in this evolution from object deletion to object retention in its earlier stages is even felt in manuscript and papyrus readings of dramatic texts − my own investigations involve the dialogue passages in Sophocles, Euripides, Aristophanes, and Menander. 52 (and its variant ), in all other Homeric instances, has an exThe verb plicit direct object − often the pronoun . In our passage from Plato, Valckenaer and in his wake Ast, in fact, include the accusative pronoun . Possibly they are influenced by the addition of object in Ficino’s translation of this passage into Latin: seque…non irritare. After Aristeides the Byzantine prose paraphrases (discussed in VIII.3) retaining Homer’s distribution of direct/indirect speech, include in their versions of this phrase: (Psellos, Cod.Gazae); (Moschopoulos); Donna Shalev 18 stronger cohesion of threads in the text. Again, like the choice of as a connecting particle, it may be said that this is an issue of text-type and of style, but is also influenced by aspect, valency of the lexeme, and other grammatical factors. IV.2 Abstraction and nominalization Abstractions (on the semantic level) and nominalization (on the syntactic level) are involved in (16) in the contrast between Khryses’ verbal expressions conveying services and gifts granted to the gods ( ) ): and Socrates’ nominal counterparts ( (16) synoptic lines 31-32 (Il. A 39-40; R. 394a4-5): Khryses: / Socrates: The full wording is provided in order to demonstrate the complex nature of the conversion, not merely from verb into verbal noun, but rather from specific verbs in Homer ) to a general verb in Socrates’ rewording ), and specification of the original actions (expressed as verbs by ( Homer’s Khryses) in nominalized form, structured as prepositional phrases with the plural of the action nouns and , the plural here indicating the multiple, repeated nature of the action.53 Note also the transformation of the direct objects in Khryses’s wording ( , ) into objective genitives ) in Socrates’ nominalized conversion. The effect of nominalization ( may be a more compact, condensed packaging (not to be confused with brevity), and a feature suggested already in the analysis of Des Places, under the term ‘synthétique’.54 In the Socratic version in our passage, the periphrasis involving (Cod. Venetus 454). These ‘differences’ may reflect language type or diachronic development, respectively. 53 Plural forms of the verbal noun can, context permitting, also enforce a concrete interpretation. For the two meanings of plural verbal nouns in Latin, iterative and concretized, see Rosén (1981: 29 f, and references given there). Denniston (1952: 38) speaking about plural nouns in Greek abstraction, does not mention a connotation of repeated action, nor do Krarup (1949) or Griffin (1986), who discuss abstract or verbal nouns in Homeric Greek. 54 With reference to a participle (in lieu of a finite form), another form divested of some verbal categories. Socrates’ paraphrase of Homer Iliad 1.11-42 19 verbal nouns adds details to a rather general verb of granting ( ) without strict parallel in the Homeric text. If one reads Thesleff’s term ‘abstraction’55 to include also the nominalized construction, one may explain its use by Socrates in this passage as a deliberate choice in an ongoing exercise of defusing as many elements as possible contributing to the affective nature of Khryses’ prayer in the ‘mimetic’ version of Homer. This presumably fits in well with the effect sought in the ‘intellectual style’ mentioned by Thesleff in the context of abstractions. IV.3 Aspect and voice shifts Some other changes wrought by Socrates are more strictly grammatical in nature, such as changes in aspect and voice. Whether deliberate or not, these differences contribute to the general effect of change in style. Socrates changes verbal aspect: Homer’s present-stemmed is rendered by Socrates’ aorist-stemmed , involving also change in word order is rendered by Socra(syn. line 12).56 Conversely, Homer’s aorist tes’ lexical counterpart, in the imperfect, (syn. lines 14-15)57 (involving also a structural transition from verb + infinitive complement to two verbs paratactically connected by . Socrates also changes voice, twice, where the same verb is used as in the Homeric model. In synoptic line 17 Homer’s active is rendered by Socrates in the middle − but this is largely due to lexicon or idiom. The voice of the verb is manipulated already within the Homeric narrative,58 but the active uttered by Agamemnon in Homer (A29) is rendered in Socrates’ indirect discourse (393e7) by a passive 55 in his list of trends of the intellectual style (5). > Cf. also this aspectual shift in the lexical counterparts: (syn. line 11) involving also ‘reduction’, and shift to passive; > (syn. line 13) involving also shift from participial to connection; > (syn. line 27). 57 Cf. > (syn. line 26), where the limitations of lexicon may be the reason for not retaining an aorist form; finally, if the use of plural verbal nouns is interpreted as an indication of repeated action, the nominal expressions in syn. line 31-32 replace aorist representing Homer’s verb forms in Homer; however, there is the aorist verbs. 58 Medium when referring to Khryses and his beloved daughter, active in the crude mouth of Agamemnon with reference to his chattel (see commentators). If is to be read at A20, we have another instance of middle converted into active. 56 20 Donna Shalev construction: > , in a transformation also involving shifts from parataxis to hypotaxis, from explicit negative to implicit negative through complex sentence, from highly explicit agent expression, with a shift from elective personal pronoun of active subject to lack of agent expression in passive construction. Is this passivization and suppression of agent another pragmatic, rhetorical, stylistic choice recruited by Socrates for defusing the affective tone that is part and parcel of direct discourse? V Ancient theory of style and the notion of sentence connection To sum up what we have seen in a comparison and contrast of grammar, syntax and phrasing in Homer and Plato, the amalgam of features and devices for tying together phrases and weaving them into a text is different in the two accounts of the exchange between Agamemnon and Khryses, and these distinctions contribute to a recognizable difference in style. Ancient theory and criticism recognized formal distinctions between different styles or modes of expression ( , oratio, just a few of the terms used) 59 −distinctions drawn in terms of the form and degree of phrasal and sentential connection (using terms such as or their derivatives), often in combination with other factors and differentiae from the , covers (among other structural or phonetic levels.60 A blanket term, things) the notion of composition, including sentence architecture and connection, and textual cohesion. The significant difference recurring in most of these scales or axes of style (under various terms) is that of the nature and degree of connection. Aristotle’s perspective, in (17) below, is an early, well-known one, distinguishing different styles of prose based on objective formal critera as well as aesthetic judgment and generic labeling: (17) Arist.Rhet. III.9.1-3 1409a22-b1: 59 These very terms are mentioned together in an overview and judgment of ancient theories of style in Russell (1981: 129f). See also Cope (repr. 1970). 60 For example the pairs periodic/open-ended; linear/disordered; metrical/ ‘resolved’; compact/loose. Socrates’ paraphrase of Homer Iliad 1.11-42 21 [ “The style must be either continuous and united by connecting particles, like the dithyrambic preludes, or periodic, like the antistrophes of the ancient poets. The continuous style is the ancient one…By a continuous style I mean that which has no end in itself and only stops when the sense is complete. It is unpleasant, because it is endless, for all wish to have the end in sight. [metaphor of double-course race] … Such is the continuous style. The other style consists of periods, and by period I mean a sentence that has a beginning and end in itself and a magnitude that can be easily grasped. What is written in this style is pleasant and easy to learn” (Tr. Fleese). The two styles, termed by Aristotle and in this specific context are distinguished first of all by the bare use of ) as a means of composition; the periodic style connectives ( (here termed ), also obviously uses connectives, but is contrasted from the non-periodic style mainly by its end-bearing structure ( …). In Aristotle’s definition, labeling the non-periodic style as the more primitive ( ), and making negative aesthetic judgments ( ), as well as associating the non-periodic style with generic orientations (dithyramb),61 all deflect attention away from an objective observation on the variety of forms of connection in different modes of presentation. Some modern authors discuss the phenomenon in a tone influenced by Aristotle’s bias for the periodic style of (and referring to it as the default), but all pay closer attention to the different connectives involved, and to the amalgam of features typical of this style, as well as to variety within a genre or an author.62 An ancient distinction of styles based on choice of connective viz., asyndeton may be seen in (18) below, from ps.-Demetrius’ discussion of the suitability of certain styles (defined by manner of connection) and delivery: 61 Note that this same genre, dithyramb, is associated by Socrates with as he defines it in his theoretical discussion running up to the paraphrase under discussion in this paper. 62 includes Eduard Norden (Antike Kunstrprosa, 37 Modern treatment of ff; Agnostos Theos, pp. 368-70), and Sophie Trenkner (1960 passim), as well as brief, but important comments in Thesleff (1967), Lilja (1968) and Guido (1983). Donna Shalev 22 (18) Dem. On Style 191-194: “ ” ” “ . “Above all, the diction must be clear. Now clarity involves a number of factors. … Secondly [it involves] the use of connectives. Sentences which are unconnected and disjointed throughout are always unclear…(194) To show that asyndeton suits an actor’s delivery, let this be an example: “I conceived, I gave birth, I nurse, my dear.” In this disjointed form the words will force anyone to be dramatic, however reluctantly - and the cause is the asyndeton. If you link it together to say, “I conceived and I gave birth and I nurse,” you will by using connectives substantially lower the emotional level, and anything unemotional is always undramatic.” (tr. Innes). Passage (18) may be read as contrasting asyndeton of members of a sentence with connection by , rather than contrasting sentence connection by asyndeton (or by other means of cohesion) with sentence connection by . All viz., of asyndeton. Note also the the same, style is differentiated by choice of association of use of connectives (possibly implying the specific choice of repeated ) with lowered emotional level. Trenkner (1960) found that in authors such as Xenophon and Aristophanes, among others, simple episodes and anecdotes have a high rate of mimetic elements, among them cohesion by means . Since this “parataxe ”, as she termed it, is the most salient feature of of an amalgam of features and patterns of distribution, the style in general was termed “style ”. Trenkner also affiliates this style with technical writing and non-composed recording of notes and epitomes. I would not go so far as to argue that the dihegetic transformation performed by Socrates is fully blown “style ”. Trenkner herself (1960: 16) warns that “style ” is not the only style used in Attic narrative − which also donned “style v”, the asyndetic style, , or a combination of features typical of a number of these styles. VI Relations between frame and content in oratio obliqua One of the salient differences between the (so-called ‘mimetic’) Homeric text and Plato’s (‘dihegetic’) version is the dynamics and economy between frame and content of speech events (Khryses’ request (passage 19); Agamemnon’s abusive response (passage 20); and Khryses’ prayer to Apollo (passage 21)). In line Socrates’ paraphrase of Homer Iliad 1.11-42 23 7 of the synopsis, Homer’s which introduces the direct quotation of Khryses’ supplication in synoptic lines 9-13, is rendered by Socrates (with AcI). At the end of Socrates’ version of the content of this supplication ) (synoptic line 14), he adds a framing expression ( not found in the Homeric model. The frames (19) are printed in bold for easier contrast: (19) synoptic lines 7ff (Il. A15ff) Homer and Khryses (R. 393e1-3) Socrates (9-13 direct discourse in one turn at talk ind.discourse in AcI constr.) In line 17 of the synopsis, opening passage (20) below, the Homeric expression introduces the quotation of Agamemnon’s ) is threat, insult and command in lines 18-24. Agamemnon’s turn at talk ( dissolved in Socrates’ dihēgēsis into three separate reports, one for each of the three speech acts enumerated: (1) the threat is framed by (with infinitive, and subordinates in optativus obliquus); (2) the insult is framed by (with a complex sentence with infinitive); and (3) the command is framed by (with infinitive and the subordinate clause in optativus obliquus). The end of Agamemnon’s turn at talk is framed by both Homer ( ) and Plato ): ( (20) syn. Lines 17-24 (Il. A25-32) Homer and Agamemnon 17 |18 “ |19 |20 |21 |(A30) |23 |24 25 ” (R. 393e5-394a1) Socrates 17 |18 |19 20 |21 |22 |23 |24 25 The different framing techniques in the two versions reflect the differences in nature both between epic convention and prose exposition, and between oratio Donna Shalev 24 recta and obliqua. This passage with the extra framing details is the prime illustration of Socratic additions. In line 27 of the synopsis (see (21) below) the verb which Homer uses , to introduce the words of Khryses’ prayer, appears in Plato’s version as accompanied not by direct quotation but by reports on the act of addressing ( ),63 on the act of reminding of past deeds ( 64 ), and on the act of requesting ( ).65 It is difficult to make at synoptic precise assignments, in particular regarding Socrates’ line 33: this partly represents the specific request, given in Homer by the imperative (also represented by in syn. line 31), and partly intro, framed in Hoduces the following prayer in oratio obliqua beginning mer by the closing formula in A43, which covers both the immediate wish ( ), and the prayer as a whole. The section of Khryses’ prayer reminding of , is represented in Socrates’ paraphrase not only past favours (A39f), the by a report of the act, but slips in elements of indirect ( with opt. 66 obl.), and free indirect discourse ( ) as may be seen in (21) below: (21) synoptic lines 27-35 (Il. A35-43; R.394a2-7) 27 28 |29“( |30 |31 |32 63 ) 27 28 29 |30 |31 |32 , and the attributive Representing, as it were, the direct vocatives relative clause. 64 Representing, as it were, the conditional clauses in A31-2. Note that Socrates also renders parts of this condition. The term for this act, , has come to signify a ritualized section of a prayer requesting a new favour from the gods based on past gifts and services bestowed upon them by the supplicant. This section, often coming in the form of ‘if ever’ conditional sentences followed by ‘grant me’ imperatives, studied by Norden (1913, and commentary on Aeneis) is particularly included, as a surrogate, in prayers not accompanied by an actual sacrifice; see Pulleyn (2000, ad 37-42). 65 Representing, as it were, Khryses’ direct expression , as well as the whole act of praying. 66 On the effect perceived by converting Khryses’ verb forms into verbal nouns, see above IV.2 with (16). Socrates’ paraphrase of Homer Iliad 1.11-42 |33 34 35 25 ” |33 |34 35 Ostensibly, the transition carried out here is completely clear-cut, but as we have seen, in actual practice the boundaries between direct and indirect are not so strict and impermeable. In synoptic lines 17ff, the infinitives in Socrates’ reworking may be interpreted as infinitives of command, 67 and the given in Socrates’ oratio obliqua belongs to the deictic realm characteristic of direct speech.68 The abrupt transition to a finite clause (with opt. obl.) opening with in synoptic line 18 has a mimetic edge to it. The expressions in in 33 are also affective elements normally associated with live line 31 and rather than reported speech. Oratio recta, oratio obliqua, and report of speech (act) Although in the theoretical context leading up to our passage, Socrates tries to create an extreme opposition between more mimetic modes of performance and strictly dihegetic performance,69 even he must succumb to the compromises of the intermediate stages in which mimetic elements trickle into the dihegetic modes. This concession (which may be viewed as failure) is one of the paradoxes of Socrates’ exercise in paraphrase, and may even be a design of Plato here. Both for linguists and for narratologists it is important to recognize the differences and the similarities between direct and indirect speech; frequently it has been associated with message and context. For the present purposes, the question of a text and its reworking also touches on this issue: when is a rewording a “faithful” translation, and when is it a process which generates change.70 It seems that even the perception of literal expression and its “fidelity” to the model are variable between one culture and another. Thus it is no wonder that 67 The status of infinitives after verba imperandi/voluntatis is equivocal, and grammarians do not always consider it indirect discourse (see e.g. Smyth §2622e); cf. Bers (1997 passim), and the literary implications of orders in indirect discourse in de Jong (1987, 116f) 68 See McHale (1978), Rimmon-Kenan (1989: 112), Baynham (1993). 69 ‘Single’ or ‘one-layered’ (rather than ‘pure’), de Jong (1987: 3), see Laird (1999: 54). 70 This is partly a question of principle, and partly based on subjective judgment. The essence of paraphrasis is touched on in later sections of this article. 26 Donna Shalev the differences between direct and indirect speech were treated differently, and the boundaries were drawn differently, by scholars of different languages, or even exclusively of Greek, with different mother tongues. One division of modes of speech performance espoused71 consists of direct speech (with its internal varieties), indirect speech (with internal varieties), and report of a speech act. According to this division, de Jong (1987: 115) counted, in the Iliad, 677 direct speeches, 88 in indirect discourse, and 33 occasions on which Homer merely mentions the speech event. Smyth adopts for Greek a traditional binary distinction between direct and indirect speech. Kühner-Gerth’s taxonomy of speech modes (ii. §592) relies in part on the level of syntactic (in)dependence involved.72 Linguistic research far removed from Ancient Greek (e.g. Li, 1986) has observed different degrees of syntactic integration in different patterns and constructions. Patterns range from those with no integration between the frame and the content (oratio recta), those featuring a degree of integration (oratio obliqua with clausal constructions), a tighter integration (oratio obliqua formed with infinitives or participles) and complete integration of frame and content, when the content is not represented at all but only reported − for example the verbs ‘he prayed’, ‘he agreed’, or expressions such as ‘he made a threat’.73 The material between the speeches, , blends into the dihegetic paraphrase with less trauma than do the speeches. The material in the speeches, , undergoes not only changes in the broad stylistic areas ranging from idiom to cohesion, but also undergoes a dilution of the mimetic effect, with accompanying changes in grammar and syntax. VII. The nature of the material omitted (and added) by Socrates 71 Among others by the linguist Coulmas (1986), by the classicists de Jong (1987), Richardson (1990), and Laird (1999), and compare also the later Genette (1988). 72 ‘Diesen zwei unabhängigen Redeformen entsprechen die folgenden zwei abhängigen Formen, in denen die berichteten Worte oder Gedanken einem im Hauptsatze stehenden Verbum der Wahrnehmung oder Mitteilung grammatisch untergeordnet wurden’, where syntactic and discursal interdependence are both implied. 73 One of the issues addressed by Fónagy (1986) is the status of individual expressions, whether they frame direct discourse, indirect discourse, or reports of speech (also known as “dihegetic summary”). Socrates’ paraphrase of Homer Iliad 1.11-42 27 I hope to show that material ‘conceded’ in the transition from epic mimetic style (both in the turns at talk, , and in the material between them, ) to strict dihēgēsis is not ‘superfluous’ material, but important on the pragmatic, rhetorical level of the verbal exchange 74. Some of these so-called superfluous elements are inexpressible in strictly indirect discourse, in the idealized, but not applied with absolute success, by Plato. Other such elements lose their function when taken out of the environment of epic codes and conventions.75 Only one omission may be considered censorship in the sense of bowdlerization (synoptic line 23), although censorship is given heavy emphasis in Gadamer (1980). The importance and place of the verbal action within the amalgam of action in the Iliad has been acknowledged by many scholars,76 and has a central role in the arguments of works such as those of Martin (1989). With respect to Iliad A12ff, Pulleyn (2000) rightly emphasizes in the introduction to his commentary that the essence of this scene is the pitting of Agamemnon’s rudeness against Khryses’ politeness. Socrates, for a combination of reasons, is less interested in emphasizing and detailing the attitudes of the main players than in the sequence of events, the “action” (as elaborated, e.g. in Blondell (2002)): in sum, his tally involves a couple of acts of motion, one act of emotion, and at least five verbal acts.77 As already stated, many of the events are speech events and action of the verbal kind. The emphasis, in Homer, is on interpersonal relations. Moreover, the speech consists not of statements but of more affective, pathos-laden utterances, such as begging, insults, and threats, or ritualistic ones such as supplications to powerful captors, and prayers to gods. The content of such speeches is difficult to convey authentically in indirect modes (as shown by discourse analysts, in the wake of linguists). Socrates is obliged to compromise 74 On the theory and use of in Greek and Roman rhetoric see G.Calboli (1993: 214-217) and in particular L.Calboli Montefusco (1988: 33-77). 75 Synoptic lines 4-7, 7-8, 10, 13, 18, 26, 28. Also the framing conventions typical of epic, or in contrast to their distribution in Socrates’ prose version, as discussed in §5, above. 76 E.g. Létoublon (1983); Richardson (1990: 70); Taplin (1992: 50f); Laird (1999: 65); Griffin (2004). 77 The priest came (4) and prayed (7). The Achaeans gave him respect (14). Agamemnon threatened (17) and made a declaration (22), and commanded Khryses (24) who was frightened (25) and departed (26) and prayed (27). 28 Donna Shalev his ideal of in transforming this passage of the Iliad, and therefore his report of the contents is formed in indirect discourse of varying degrees of freedom. The Greek in the Socratic version in synoptic lines 18, 20, 22, 31, and 33 reveals signs of a sort of indirect discourse which is open to mimetic elements. The omissions in synoptic lines 9, 18, 20, 22, 27, 29-31, and 33 also reflect Socrates’ tendency to avoid mimetic and pragmatic elements in his dihegetic report, when he ‘translates’ both the events and the speech events. Elements associated to direct speech and perceived as marking the mimetic mode, omitted by Socrates, are vocatives and address: Khryses’ formal address, (syn. line 9); Agamemnon’s (syn. line 18), and the vocatives and eponyms in Khryses’ prayer vocative to Apollo (syn. lines 29-31). Vocatives and other elements78 which are difficult (and in some languages impossible) to express in indirect speech are, by natural extension, elements which will be lost in a strict dihēgēsis. Other affective and emotional elements omitted by Socrates are the particle in Agamemnon’s aggressive threat (syn. line 20), the pathos-laden phrase ,79 and Khryses’ ethical dative at the climax of his prayer to Apollo (syn. line 33). All three omissions involve features typical of Greek and of a number of languages,80 certainly not universal. Elements of epic character, such as the end of the Homeric hexameter in synoptic line 10 are omitted from Socrates’ version. This is a legitimate example for a generic feature of epic, 81 which tends to add 78 See, e.g. Rimmon-Kenan (1989: 112), Baynham (1996: 64) and Coulmas (1986: 2f). Aside from addresses, other elements mentioned in McHale, Baynham and Coulmas, are different dialects, sentences with incomplete syntax or repetitions, exclamations, other affective utterances, deixis of time and place, politeness, ‘perhaps’ and other expressions of uncertainty, and other subtle attitudes. It is important, I think, to accompany such a collection of features with a reminder that they are not universal in essence or degree, but to some extent language-specific, as emerges, e.g., in the treatment of Fónagy (1986). 79 Forming the epically rhythmic bucolic dihaeresis. Agamemnon’s ruthlessness, and the counterpoint between his character and that of Khryses, are kept in focus. 80 The particle and ethical dative are not common in some of the modern languages familiar to discourse analysts. 81 Labarbe (1949: 358): ‘Platon a eu le souci d’exclure les expressions figées, propres au style aédique.’ According to Labarbe not only the formula ’ , but noun epithet groups belonging to this epic style are omitted. Socrates’ paraphrase of Homer Iliad 1.11-42 29 what is sometimes considered less informative material in order to fill in a hexameter.82 Some literary scholars would have an easier task if all the omissions in Socrates’ version were of this order. But even this epithet in fact contributes to the portrayal of Khryses’ posture: a reverent man of god, in particular here (in Homer’s portrayal) when he is in a deliberate counterpoint with the rude and insolent man of war Agamemnon, whose character is painted, among other things, by the lack of polite address (line 18). In line 26, , again the material ‘filling out the line’ so to speak, is the example Genette gives for insignificant material. In the Homeric tradition the noisy seashore in the face of a still person is a formula indicating the Leitmotiv of a sorrowful context, and the connection between the motif and this formula must have been recognized as such by the original audience.83 The direct speech maintained by Homer presents Khryses using polite expressions of request (in the form of optative of wish:84 (syn. line 10) (13), if this is the reading), in contrast with the mainly negative, harshtoned, assertive, threats and prohibitions in the mouth of Homer’s Agamemnon (syn. line 18) (20) (21), ( (24)). In the dihegetic style of Socrates’ paraphrase, this pragmatic-rhetorical contrast of Khryses and Agamemnon is lost. Literary nuances of character attitude forfeited by Socratic omission: In lines 6-7 of the synopsis, omitted Homeric material / ) is important for a more subtle understanding of the attitude of Khryses, who, in Homer’s version, is painted with suggestions of authority. The fillets and the sceptre symbolize Khryses’ status as a priest, and are meant to be suggestive to Agamemnon (and to the audience) of the respect due to Khryses. Apollo and his ballistic powers are on the side of Khryses, as is emphasized in the Homeric presentation of this scene, with its implicit threat. These subtleties are silenced in Socrates’ version. 82 (syn. 13), (syn. 18), (syn. 28), and of course (syn. 26) discussed below. 83 For further detail see passage (1) above, and discussion. See also Taplin (1992). 84 See Chantraine, Grammaire homérique §§315, 320. Cf. Donna Shalev 30 But this omission may be the inadvertent fallout of an orthodox, almost ideological adherence to a strict dihegetic mode. Failure to omit all ritual elements in Khryses’ prayer: As a whole it is difficult to transfer the act of prayer into indirect speech, and Plato must have been aware that here he cannot omit it. So he chooses merely to report the occurrence of the prayer: . The Homeric text as given in synoptic lines 29-34 is a typical expression for the contents of prayer in ancient texts: the conditional sentences preserve the liturgical ritual version. The Socratic version, if it were faithful to the demands of strict dihēgēsis which Socrates preached, would content itself with the expression . But stages of ritual cannot be brushed over, even in dihēgēsis, so Socrates succumbs to a compromise in his ideal dihēgēsis, and gives mention to the events: . These acts are fully verbalized in Homer, trickles through even into Socrates’ but some of the wording of the version. VIII. Varieties of Paraphrase VIII.1 Retelling A17-42 within the Iliad Socrates is not the first character to paraphrase parts of this opening scene of the Iliad. For different reasons, and with an agenda of his own, Achilles recounts to his mother, Thetis, his version of events (A365-393). Achilles’ narrative is introduced by Thetis’ request, posturing ignorance, and Achilles response that retelling the tale is needless, since his mother knows all (i.e. is omniscient, immortal).85 This context of knowledge and lack of knowledge foreshadows the nature of the setting of Socrates’ retelling, revolving around an impasse in understanding (see VIII.2 below). Achilles’ version begins, ab ovo, with a brief account of the capture of Khryseis (366-369), whereas the sequence narrated in A11-34 (and in Socrates’ paraphrase) is in medias res. In his commentary, Pulleyn (2000, ad locum) aligns the parallels in the two passages, and notes in 85 362-5: Thetis: “ ”/ / “ compares this with Odyssey 12.145f. ; / ; De Jong (2001: 486f) Socrates’ paraphrase of Homer Iliad 1.11-42 31 particular the ommisions (Khryses’ speech A22-25, and Agamemnon’s speech A26-32) and differences. Pulleyn concludes by pointing out that Achilles’ narrative may be characterized as not reporting direct speech. The opening scene, besides being paraphrased at A365ff, may be said to be a member of a group of mirror scenes, most famously in the final book of the Iliad (see Macleod (1982) for alignment of matching or similar wordings); these two, and other scenes of supplication and ransom have been compared for nuances of similarity and difference in dramatic setting, ritual detail, epic structure, motifs, and language form (see Taplin (1992) for summary and reference to previous works). The small twists of difference contribute to the overall appreciation within the Iliad, and these reworkings belong to another, more complex level, which cannot be included in our contrast of devices of cohesion, grammar, and dosage of mimetic elements in prose conversions. VIII.2 The performed by Socrates and the notion of paraphrase What, in fact, is the activity in which Plato has Socrates engage here? Is it telling, retelling, transforming, paraphrasing, explaining, teaching, or a combination of these, not to be captured by any one term? In referring to this passage, as well as the common reference to “paraphrase” (e.g. Adam, Labarbe), the terms used are “transformation” (de Jong, Laird), “retelling” (Laird), “reformulation” (Laird), “transposition” (Des Places, Labarbe), even “rewriting”/ “transcription” (de Jong, Genette) − referring, presumably, to the activity of Plato rather than Socrates.86 The comparison of wording and its analysis suggest that this activity of Socrates (or Plato) in our passage involves retelling, rewording, reforming, rephrasing, with differences on a variety of levels: among others verse and prose (in Homer and Plato respectively); a blend of mimēsis + dihēgēsis in Homer, and dihēgēsis (with residues of mimēsis) in Plato; a blend of indirect + direct discourse in Homer, and indirect discourse in Plato; of poetic and pedestrian language and style; of and ; of participles, clauses, particles, and the style. The term used by Socrates himself in this passage is . Before performing the exercise of rephrasing − the act of 86 Cf. also Des Places (1935: 133) explicitly referring to a transcription by Plato, and Thesleff (1967: 62 n.3). Donna Shalev 32 as it were − he uses the term twice, perhaps not with identical connotation: (22a) R 393d2-3: And lest you may say again that you don’t understand, I will explain to you how this would be done (tr. Shorey), and immediately before the paraphrase (22b) R 393d8-9: ’ − It would have been somewhat in this wise. I will state it without meter… (tr. Shorey). In translations of the passage, these expressions are at times glossed over.87 Definitions for offered by LSJ emphasize the “explaining”, and are put in opposition with “merely saying” (see I.1. and I.2 end, s.v.). The contexts of many of the passages referred to in these lemmata suggest not only explanation, but some form of elaboration in the telling. The passages at the end of LSJ I.2, e.g. those from Sophocles, illustrate the use of forms of in the context of misunderstanding and a request for clarification through retelling. Not many passages from Plato are listed. The dramatic backdrop and the roles of the interlocutors, Socrates and Adeimantus, in the run-up to the paraphrase of the early passage from the Iliad, revolve around a difficulty in understanding. At 392c6, Socrates announces his intention to move from to the dimension of . The transition and the distinction it requires, prove rather subtle and challenging for Adeimantos as he is characterized by Plato here ( ). At first (392d1) Socrates tries another way of informing Adei). Without announcing what this way is, he mantos ( performs a robust question-and-answer session by which he hopes to achieve clarification and understanding. It is not long (392d7) before Adeimantus again admits trouble in understanding ( ). Socrates condemns, in self-irony, his own poor teaching skills, and announces (392e1) another attempt at clarification ( ). This he tries to achieve, again, by the method of sustained question and answer. 87 e.g. Chambry, Jowett, Laird. Shorey’s choice of ‘state’ is not optimal. Socrates’ paraphrase of Homer Iliad 1.11-42 33 Adeimantos gives the impression of grasping,88 and it is Socrates who reverts to Adeimantos’ past difficulties in understanding. Socrates is returning to a recurring motif turned literary convention of the interlocutor declaring aporia. It is in this context that Socrates instigates (393d2) the performance of an instructive move which he thinks will once and for all clear up the issue at hand, and rid Adeimantos of the feeling of not understanding: (24)R.393d23: And lest you may say again that you don’t understand, I will explain to you how this would be done (tr. Shorey) Socrates’ clarifications through the question-and-answer routine were not successful, so he will explain and elaborate. He intends to achieve this by performing a . The illustration of approved poetry in which the poet does not take on the voice of a character, of narrating without mimetic elements, is carried out of it without meby retelling the beginning of the Iliad, by performing a ter.89 So in this passage, the retelling in a different form, in order to clarify and explain to someone who cannot understand by definition (or, in this instance, by repeated attempts at instruction through question and answer), is termed by forms of 90 are often used, it is true, to denote ‘to In Plato, forms of speak’.91 Yet in a fair number of passages elsewhere in Plato, 92 especially on a backdrop of the motif of not understanding, along with other expressions of clarifying,93 an expression involving a form of announces an attempt or intention by Socrates to clarify for his interlocutor: then follows either a question- Reactions of confident agreement and acquiescence to Socrates’ questions and testing statements: ; b6); ; ; (c4) 88 And without mimēsis, and all it entails for Socrates (direct speech, , poetic vocabulary and register, epic touches, pathos, etc.) 90 , which is not found in the Platonic corpus. But not the verbal noun 91 See Ast’s Index Platonicum for passages. 92 Euthyph. 10a5; Crat. 404e7; Gorg. 463e5; Thaet. 146d6; Soph. 262e14; Theag. 129a8. A full tally of passages falling under this contextual and dramatic pattern requires further research; I have not been able to find commentary referring to such a motif. 93 E.g. (Thaet. 164d3); (Symp. 206b10). (Parm. 136c7); or 89 34 Donna Shalev and-answer session,94 a retelling,95 or a blow-by-blow explanation,96 all presumably exegetical and/or didactical in intent and nature. The de facto use of different or parallel words, or rephrasing, is sometimes also implied in the announcement ‘I’ll rephrase it as it seems to me’,97 implying, I’ll give you my take on it, my version. Socrates may be said to be performing in R 392d an , along lines described in Marrou (1956: 165-8), uniquely by telling it in another way, in continuous prose exposition. Elsewhere, one may say Socrates is performing or extracting by catechism, shown by Marrou (1956: 168) to be a popular method of clarifying Homeric plots, albeit in the period subsequent to Socrates ) and of narrative ( and Plato. Characteristics of catechism, of the fable ( ) practiced by the authors of progymnasmatic literature, resonate already in Socrates’ paraphrase. In most instances in Socratic dialogue, this , and is framed by expressions involving activity falls under the title of (± ) or their near equivalents. The term is post-Platonic, and no direct link between the use of the simplex in Plato and this (or other compounds) has been established, but the similarity in ignored. Context, purpose, and results cannot be ignored. VIII.3 Aelius Aristeides and the later Greek prose paraphraseis of Homer Apart from the paraphrase put into Socrates’ mouth by Plato within the framework of a dialogue, in an ad hoc situation of staged lack of understanding, and the internal paraphrases within the Iliad, there exist a series of later paraphrases, of two types: rhetorical and grammatical. 98 The paraphrase of Aristeides, from the 2nd century CE, appears as chapter 14 (in book I), entitled in his work On Rhetoric. Ludwich (1885: 483 ff) treats this prose paraphrase separately from the later ones,99 which, in an appendix on Greek para94 Euthphr. 10a5; Gorg. 463e5; Tht. 146d6. R. 392d2; d8. 96 ? Gorg. 463e5. 97 Crat. 404e7; Gorg. 463e5; Thaet. 146d6; 164d3. 98 This distinction is discussed, along with the phenomenon, in Lehrs (1873: 51f). 99 Some only in ms. form. Ludwich’s conspectus of prose paraphrases of Homer includes on the one hand those of Plato and Aristeides; on the other those of Psellos, Moschopoulos, Theodoros Gaza, and an anonymous one in the 15th century ms. Venetus 454. 95 Socrates’ paraphrase of Homer Iliad 1.11-42 35 phrase literature, he describes as ‘word for word’, and classifies as ‘grammatical’. In fact the salient but little-noticed difference from our perspective is that Aristeides’ paraphrase follows Socrates’, remaining (as much as possible) in oratio obliqua. Aristeides’ paraphrase begins with a lemma, line 1 of the Iliad, an incipit of the original text. After the background, in [8] according to Ludwich’s lineation (following the Homeric content) equivalent to §183 in the Teubner edition, Aristeides frames the long sequence of AcI construction, hinging formally on the verbum dicendi , appearing in a question, possibly for a pedagogic effect: ; The other paraphraseis are closer to Homer, from our particular viewpoint, because they maintain the Homeric distribution between direct and indirect speech, rather than attempting a sustained transposition from Homer’s blend of direct speech and intermediate narrative material, into unadulterated indirect discourse, as Aristeides does, following Socrates. What the later (“grammatical”) prose paraphrases share with Aristeides is attributive discourse (framing) more typical of prose writing, and the later vocabulary; both classes of paraphrase differ from Socrates’ reformulation in that they maintain the content of much of the epic formulaic phraseology which Socrates omits. In fact Aristeides’ “fidelity” to Homer and to the later paraphrasts in some details of grammar and style100 only underscore his similarity to Socrates in maintaining oratio obliqua. Even Aristeides’ paraphrase fails to sustain the oratio obliqua until the very end. Just as Socrates has to make concessions, so does Aristeides − different ones. A more mimetic use of slips into a string of indirect discourse at [32], and he omits the phrase ‘don’t disturb me’ (see n.52) which even Plato retains in his transformation (albeit without the object). The oratio obliqua finally breaks down in Khryses’ prayer to Apollo, which includes the pronouns in a sort of oratio obliqua, a list of Apollo’s eponyms, and the conditionals of the are in indirect discourse with elements of direct discourse, and elements of dihegetic summary. 100 + verb (A24); retains original voice of the Unlike Socrates, Aristeides reverts to verb (e.g. ); does not subdivide Agamemnon’s speech, but frames it all with a verb of threat. Unique, even vis-à-vis Socrates, is Aristeides’ adherence to hypotaxis, when he uses a temporal clause instead of a resumptive participle at A35. Clearly a fuller comparative study of these paraphraseis is desired. 36 Donna Shalev IX. Concluding remarks The detailed comparison of the wording in the two texts, Homer Iliad A12-42 and Plato Republic 392dff, surprisingly seldom pursued by classicists, has produced some very interesting findings, the most important of which is the amount of change involved in the conversion. Whether paraphrase is seen as striving for fidelity or for change, some of its aims include rewording for clarification, or for change of format, and often for the purpose of teaching. The change wrought on the Platonic text in the non-speech sections ( ) involve style in the sense of idiom, vocabulary, currency, and composition (including cohesion); in the speeches ( ) along with changes in style as just detailed, the transition from direct to indirect discourse adds change in another dimension, the dosage of mimetic elements, as the author tries to recalibrate the mimetic effect. If Socrates’ professed ideological aim in this exercise is to convert a text with mimetic parts into “simple” dihēgēsis, unadulterated by elements of direct speech, he does not entirely succeed: since the three speeches in the Homeric text are all affective, it is not surprising that they cannot be rendered entirely in indirect discourse, with all the mimetic elements suppressed: the vocatives, the addresses, the emotional force, the ritual elements of prayer, all trickle into the indirect discourse, making it to some extent free (VII, and n.77). The moral need to defuse the pathos has been interpreted as central to Socrates’ ideology in Republic III, and the impetus behind his exercise in eradicating mimetic modes in lore (e.g. Blondell, 2002). And indeed some emotional elements of content are omitted from Socrates’ version (VII). Plato, the master of stylistic diversity and versatility, can promote this defusing of emotional effect by shifting, in as obvious a way as he can, from affective to intellectual language, form, and structure: deliberate steps are shifts to abstract expression, to passivization, to shades of styles, such as that involving repeated connection by , which already Demetrius associated with lowering of emotion: (passage (18) above). Scholars have commented on Socratic brevity in relation to the Homeric text, some (e.g. Genette) in terms such as “condensation”, “reduction”, which imply more than mere size. A different quality of construction, of economy, lies at the basis of such terms, as referred to by others (e.g. Thesleff), or the related term ‘synthetic’ as captured by Des Places. Within Socrates’ paraphrase of Homer Iliad 1.11-42 37 the more complex connotations of this notion I would include specific phenomena of language, structure and style: hypotaxis, ellipse, nominalization and other abstraction, monolectic verbs of negative semantics, and participial connection. Condensation is also achieved by integration of frame and content of indirect discourse, by preference for non-finite constructions and reports of speech − devices of reduction which are in fact Socratic additions (VI and VII). I would go one step further, and say that Socrates’ deliberately fashioned style in his paraphrase integrates formal and syntactic choices which contribute to a lowered emotional effect, and those which contribute to condensation. Finally, a study of Platonic text, even of a most technical aspect, cannot be complete without some consideration of the dramatic and literary setting. Socrates’ pedagogic paraphrase is staged in this passage of the Republic as an exhēgēsis in response to a real or imagined impasse in Adeimantos’ comprehension (VIII.2), a motif begging further investigation, and echoing the context of ignorance/ omniscience summoning another paraphrase, within the Iliad, by Achilles, of the same scene (VIII.1). References Adam, James (1965), The Republic of Plato. Cambridge: Cambridge University Press. Bakker, Edgert (1997), Poetry in Speech. Ithaca - London: Cornell University Press. Baynham, Mike (1996), “Direct Speech: What’s it doing in non-narrative discourse?”. Journal of Pragmatics 25.61-81. Bers, Victor (1997), Speech within Speech: Studies in Incorporated Oratio Recta in Attic Drama and Oratory. Boulder - New York - London: Rowman and Littlefied. Blondell, Ruby (2002), The Play of Character in Plato’s Dialogues. Cambridge: Cambridge University Press. Calboli, Gualtiero (1993), Cornifici, Retorica ad C.Herennium, Introduzione Testo Critico Commento, 2° ed., Pàtron, Bologna. Calboli Montefusco, Lucia (1988), Exordium Narratio Epilogus, Studi sulla teoria retorica greca e romana delle parti del discorso, CLUEB, Bologna. Cope, Edward Meredith (1970 reprint), An Introduction to Aristotle’s Rhetoric. Hildesheim: Olms. Coulmas, Florian (1986), Reported Speech: Some general issues, in: Coulmas, F. (ed.), Direct and Indirect Speech. Berlin: de Gruyter, 1-28. Denniston, John Dewar (1952), Greek Prose Style. Oxford: Clarendon. Des Places. Edouard (1935), “Citations et paraphrases de poètes chez Démosthène et Platon”. In: Mélanges … Octave Navarre. Toulouse: Privat, 129-137. Ferrari, Giovanni (1989), “Plato and Poetry”. In: Kennedy, George (ed.), Cambridge History of Literary Criticism, I. Cambridge: Cambridge University Press, 108-119. 38 Donna Shalev Fónagy, Ivan (1986), “Reported Speech in French and Hungarian”. In: Coulmas, Florian (ed.), Direct and Indirect Speech. Berlin: de Gruyter, 255-309. Gadamer, Hans (1980), “Plato and the Poets”. In: Dialogue and Dialectic, tr. C. Smith. New Haven: Yale University Press, 39-71. Genette, Gerrard (1980), Narrative Discourse, tr. J. Lewin. Ithaca: Cornell [=1972, Figures III (Paris)]. Genette, Gerrard (1988), Narrative Discourse Revisited, tr. J. Lewin. Ithaca: Cornell. Griffin, Jasper (1986), “Homeric Words and Speakers”. JHS 106.36-57. Griffin, Jasper (2004), “The Speeches”. In: Fowler, Robert (ed.), The Cambridge Companion to Homer. Cambridge: Cambridge University Press, chapter 10. Guido, Roseanna (1983), “ e (Ps. DEMETRE. De Eloc. IV.193)”. In: Studi in onore di Dinu Adamesteanu. Galatina: Congedo. de Jong, Irene (1985), “Iliad 1.366-92: A Mirror Story”. Arethusa 18. 5-22. de Jong, Irene (1987), Narrators and Focalizers: The Presentation of the Story in the Iliad. Amsterdam. Krarup, Per (1949), “Verwendung von Abstrakta in der direkten Rede bei Homer”. CM 10.1-17. Labarbe, Jules (1949), L’Homère de Platon. Liège. Laird, Andrew (1999), Powers of Expression, Expressions of Power. Speech Presentation and Latin Literature. Oxford: Oxford University Press. Lausberg, Heinrich (21973), Handbuch der literarischen Rhetorik. Eine Grundlegung der Literaturwissenschaft. München: Max Heuber. Lehrs, Karl (1873), Die Pindarscholien: eine kritische Untersuchung zur philologische Quellenkunde. Leipzig: S. Hirzel. Létoublon, Françoise (1983), “Défi et combat dans l’Iliade”. REG 96.27-48. Li, Charles (1986), “Direct Speech and Indirect Speech: A functional study”. In: Coulmas, Florian (ed.), Direct and Indirect Speech. Berlin: de Gruyter, 29-45. Lilja, Saara (1968), On the Style of the Earliest Greek Prose. [Comm. Hum. Litt. 41.3]. Helsinki: Societas Scientiarum Litterarum. Ludwich, Albert (1885), Aristarchs Homerische Textkritik nach den Fragmenta des Didymos, II Leipzig: Teubner. Lynn-George, Michael (1988), Epos: Word, Narrative, and the Iliad. Basingstoke: Macmillan. Macleod, Colin (1982), Homer: Iliad XXIV. Cambridge: Cambridge University Press. Marrou, Henri-Iréné (1956), A History of Education in Antiquity, tr. G. Lamb. London: Paulton. Martin, Richard (1989), Language of Heroes. Ithaca - London: Cornell University Press. McHale, Brian (1978), “Free Indirect Discourse: A Survey of Recent Accounts”. PTL 3.249-287. Murray, Penelope (1996), Plato on Poetry: Ion, Republic 376e-398b, Republic 595-608b. Cambridge: Cambridge University Press. Norden, Eduard (1913), Agnostos Theos. Untersuchungen zur Formengeschichte religiöser Rede. Leipzig. Pulleyn, Simon (2000), Homer: Iliad I. Oxford: Oxford University Press. Socrates’ paraphrase of Homer Iliad 1.11-42 39 Rimmon-Kenan, Shulamith (1989), Narrative Fiction: Contemporary Poetics. London: Penguin. Rosén, Hannah (1981), Studies in the Syntax of the Verbal Noun in Early Latin. München: Fink. Ruijgh, Cornelius (1971), Autour de ‘ épique’ . Amsterdam. Russell, Donald (1981), Criticism in Antiquity. Berkeley: University of California Press. Sicking, C.M.J. – van Ophuijsen, J.M. (1993), Two Studies in Attic Particle Usage: Lysias and Plato. Leiden: Brill. Thesleff, Holgar (1967), Studies in the Styles of Plato. Helsinki. Taplin, Oliver (1992), Homeric Soundings. The shaping of the Iliad. Oxford: Oxford University Press. Trenkner, Sophie (1960), Le style dans le récit attique oral. Assen: Van Gorcum.