II.5 Particles and Anaphoric Reference: A discourse perspective on particles with third-person pronouns

§1. In ancient Greek, pronouns and particles have a special relationship: the two are often found together and intrinsically connected. They are not only frequently adjacent, but they work together to guide the discourse, and may even form a single unit. II.2 demonstrates how the Homeric and Pindaric performers produce their discourse piecemeal, each piece adding a bit of information to the preceding. As a speaker focuses on the ideas in her mind she verbalizes her words according to the flow of her thoughts, and in the form of discourse acts. Chafe argues that only one new idea can be in focus in the mind at one time. [1] This is reflected in the form and content that discourse acts take: established knowledge tends to appear towards the beginning of the act and new information tends to follow. [2] Because anaphoric pronouns recall referents that are already in the hearer’s mind, so they must appear within the body of established, given knowledge (more on “given” below) in order to be effective.
§2. Anaphoric reference directs attention to a referent about which something new will be added, and pronouns are the prototypical markers of anaphoric reference. Combinations or clusters of pronouns and particles are not used randomly; a deeper understanding of the pronoun illuminates the workings of the particle and vice versa. Only by comparing a larger number of instances of different collocations can it become clear that there are significant and consistent differences among them. The ideas established in the preceding chapters are relevant to the role that particles play in guiding reference. In particular the distinction between framed and unframed discourse (II.4) provides an important analytical tool for the study of anaphoric pronouns in Homer and Pindar.
§3. In this chapter I examine how the use of different particles interacts with different nuances of anaphoric reference. In order to understand these nuances it is first necessary to arrive at a good understanding of how anaphora works, both inside and outside the text. In the first section (5.1) I outline recent approaches that interpret the process of anaphoric reference as an interaction between a speaker and hearer rather than as a set of immanent relations between textual constituents. In the next section (5.2) I address the peculiar place of the nominative pronoun in a pro-drop language like Greek. Moreover, I discuss the ambiguous function of ὁ and ὅς as both demonstrative and relative pronouns. After addressing these issues, I present a representative case study (5.3) of combinations of the most frequent third-person pronoun in the nominative (ὅς and ὁ) and different particles in Homeric narrative: ὁ δέ, ὅ γε, ὁ δ᾽ἄρα, ὅ/ὅς ῥα, and ὁ/ὃς δή. [3] The analysis aims to show that the combinations have consistently and significantly different functions, depending on the particle used. [4] I do not engage in particular with pronouns (and particles) at the beginning of embedded narratives, but this theme has been studied extensively for both Homer and Pindar. [5] The final section (5.4) traces anaphoric reference through an entire Pindaric ode: Isthmian 2. In this close reading, I consider not only pronouns and particles, but also nouns and verb forms, to sketch a picture of the audience’s on-line processing of anaphoric reference.

5.1 A discourse approach to anaphoric reference

§4. Classically, anaphora has been regarded as a relation that functions within a text. [6] An anaphoric relation is typically expressed as one between an anaphoric pronoun and a textual antecedent. However, recent research that focuses on naturally occurring discourse demonstrates that the textual approach often does not explain what actually happens. Consider the following example:
Wash and core six apples. Put them into a fireproof dish. [7]
This simple example demonstrates the problem of a purely textual approach. If a reader regards “them” as referring to the textual antecedent “six apples,” the utterance would be infelicitous. After all, the referent of “them” must be described as “the six washed and cored apples.” In practice, of course, the reader or hearer has no trouble making the inferences necessary to understand the two sentences in the cookbook. This is a relatively simple example of anaphoric reference, but even here an analysis in purely textual terms does not explain the cognitive process sufficiently. [8] It is not enough to say that “six apples” is the textual antecedent of “them.”
§5. To solve problems such as this one, Cornish redefines the term “antecedent” as follows: “I take [antecedent] to be a description of the referent (...) in terms of its salient attributes.” [9] The generally understood meaning of antecedent, namely the textual antecedent, Cornish calls the “antecedent-trigger”:
The antecedent trigger “introduces an entity into the discourse via its predicational and utterance context, and an anaphor of a particular type and form accesses that mentally represented discourse entity at a later point in the discourse, adding to this representation further properties resulting from the processing of the anaphoric clause as a whole.”
Cornish 1999:4
Cornish’s description points to the importance of the cognitive processes involved in the production and processing of discourse. The textual antecedent problem is one of the factors that have led to a cognitive approach to reference.
§6. Another relevant factor is the relationship between different anaphoric expression and the kinds of referents that they can retrieve. In early pragmatic accounts, unaccented anaphoric pronouns were regarded as expressing “given” information, while the following predication contains the “new” information of the sentence. If a full noun phrase or name is used instead of a pronoun, however, this generally introduces new information into the discourse. The form of the referential expression was thus linked to its status of given versus new. [10] In English, the possible forms of referring expressions range from full noun phrases, via prosodically emphasized pronouns (SHE), to unaccented pronouns (she), and null anaphor (the absence of a verbally expressed subject or object). Along a scale between “given” and “new,” established information should receive light linguistic and prosodic marking, whereas new information is made explicit and receives prosodic emphasis. [11] Consider the following examples: [12]
“John called Mary a Republican, and then SAM [new] walked in.”
“Mary paid John and he [given] bought himself a new coat.”
It is important to be aware that the labels of “given” and “new” in this framework do not relate to the status of the referent in the speaker or hearer’s knowledge. In fact, (t3) reads most naturally when we assume that both speaker and hearer know who “Sam” refers to. The newness or givenness is rather a status relative to the ongoing discourse, which has led people to prefer the terms discourse-old and discourse-new. However, even this terminology is insufficient for explaining the forms of referential expressions used in naturally occurring discourse. Most obviously problematic are referents that are introduced into the discourse, then not mentioned for a certain span of text, and then retrieved. Although they are discourse-old (i.e. “given”), they cannot generally be retrieved by an unaccented pronoun.
§7. As an alternative to the discourse-old and discourse-new distinction, scholars have come to describe anaphora in terms of the “accessibility” [13] of referents or, alternatively, in terms of a givenness hierarchy. [14] The aim of these approaches is to account for the form of the referential expression in every conceivable reference relation in discourse. [15] What the approaches have in common is that they assume a match (more or less strict) between the accessibility/ givenness of a referent, and the different available referring expressions. [16] Since it is not my goal to be able to predict the referential expression for each instance of anaphora or deixis in Homer and Pindar, [17] I focus on the thing that these approaches have in common: the supposition that there is a relation between the accessibility of a referent and the referring expression a speaker may use. The next question, then, is what this accessibility consists of. [18]
§8. In the production of a discourse, the aim of the speaker is to guide the hearer’s creation of a mental representation of the discourse that is as close as possible to the speaker’s. This mental representation, the “discourse model,” is the framework within which reference functions. [19] With each new discourse act the discourse model is updated, while at the same time the production of each discourse act presupposes a certain state of the discourse model. There needs to be a minimal correspondence between the speaker’s and the hearer’s discourse model for us to be able to speak of successful communication. This minimal value is hard to establish, but one crucial factor is the mutual tracking of referents. That is, for us to speak of a successfully communicated narrative, the speaker and the hearer need to agree about who did what to whom. It is the speaker’s task, therefore, to assess the salience, accessibility, or givenness of a certain referent in the hearer’s current discourse model, if she wishes to successfully refer to the relevant character:
“[I]t is incumbent upon the speaker to use the discourse procedure which is in accordance with both his/her referential intention and with his/her assessment of the current state of the interlocutor’s discourse model.”
Cornish 1999:20 [my emphasis]
This description of the process of reference has been widely taken up, as witnessed by the following relatively recent quotes:
“[T]he grammar of reference and topicality in human language is keyed delicately to anticipate the epistemic mental states of the interlocutor.”
Givón 2005:133 [my emphasis]
Gundel, Hedberg, and Zacharsky focus on the presuppositions inherent in referential expressions:
“The major premise of the Givenness Hierarchy theory (Gundel et al., 1993) is that different determiners and pronouns encode, as part of their conventional meaning, information assumed by the speaker about the cognitive status of the intended referent in the mind of the addressee.”
Gundel, Hedberg, and Zacharsky 2012:251 [my emphasis]
What all authors agree on [20] is that anaphora is not to be understood as a relation between a referential expression and its antecedent in the co-text, but as an instruction to the hearer to focus on a certain referent in his discourse model. It is not primarily a process of verbal memory, but of cognitive focus within the mental representation of discourse. [21] As a result, the accessibility of a referent is its status in the current representation of the discourse: regardless of whether it has been mentioned before and how long ago, the referent’s current status as more or less in focus determines the anaphoric expression used.
§9. Over recent decades, Chafe has applied this cognitive perspective to narrative. Chafe follows the basic idea that accessing referents that are already in focus requires less cognitive effort than accessing referents that are out of focus at the moment of utterance. This difference in effort is what explains the use of different referring expressions, with generally less linguistic marking for referents that are in focus, and more for referents that are out of focus. [22] A good illustration of the cognitive approach to anaphoric reference in narrative is the case of (apparent) underdetermination:
“Jane hit Abby. She fell.”
Since we have two female referents in this narrative the “she” in the second sentence is underdetermined at first sight. However, we have no problem interpreting the line since we visualize the scene and imagine the victim falling, rather than the aggressor. From Homer, we might compare the moment of Patroclus’ death by the hands of Hector:
820    ἀγχίμολόν ῥα οἱ ἦλθε κατὰ στίχας, | οὖτα δὲ δουρὶ |
          νείατον ἐς κενεῶνα, | διαπρὸ δὲ χαλκὸν ἔλασσε· |
          δούπησεν δὲ πεσών, | (...)
Iliad 16.820-822
          close to him he [sc. Hector] came through the ranks, and thrust with his spear
          at the lower belly, and right through he drove the bronze.
          He [sc. Patroclus] clattered when he had fallen, (...)
In English we find the unaccented personal pronoun “he,” where in Greek we find the equivalent least marked reference, null anaphor, within the third-person singular verb (δούπησεν). Despite the change of grammatical subject (from Hector to Patroclus), the lack of an expressed new subject does not lead to confusion, for three reasons. First, the semantics of πεσών makes Patroclus the logical subject, since he has just been wounded. Second, δούπησεν δὲ πεσών is a formula that always refers to the stricken hero, so the audience’s knowledge of epic will prevent any potential ambiguity of reference. [23] Third, the form of the formula in itself suggests that not just “the referent Patroclus” is in focus, but the entire image of his fall. The finite verb does not strictly refer to the noise Patroclus makes, but to that of his armor and weapons. [24] The Homeric performer can present this complete, synaesthetic image, since from the moment that he is stabbed and wounded (line 821), Patroclus is in focus in the mind’s eye. It is the idea of “focus” that explains the two examples above: in both cases the mind’s eye inevitably moves to the recipient of the attack, putting them in mental focus, and thus making them accessible. [25] In earlier work, Chafe speaks of the “subject of consciousness” in terms of attention, “the mechanism by which the spotlight of consciousness is directed at one or another area of the material accessible to the mind.” [26] A higher accessibility leads to a lower “activation cost,” in Chafe’s terms, which in turn translates to a less specific referring expression. [27] The focus of the mind’s eye is an important factor in the process of anaphoric reference. It is especially relevant in passive constructions, where there is a clear distinction between subject/object and patient/agent. In Homer and Pindar, moreover, the grammatical subject can be a part of the body or a hero’s weapon or armor, as it might be here, while the subject of consciousness always remains the character (see also t13 below).
§10. In the field of ancient Greek literature, Bakker and Bonifazi have applied the cognitive approach to examine anaphoric reference in several works. Bakker has applied Chafe’s ideas on activation cost to expressions of anaphoric reference in Homeric epic. [28] Expanding on Bakker’s work, Bonifazi further has addressed the need for a cognitive perspective on anaphoric reference in ancient Greek literature more generally. She has argued that accessibility or activation cost does not sufficiently account for all forms of anaphoric reference, especially in literature. [29] Her study focuses on αὐτός and (ἐ)κεῖνος, but her brief analysis of anaphoric reference in the first ten lines of the Odyssey points the way for further research. [30] The present chapter builds on the work of Bakker and Bonifazi, applying the cognitive perspective particularly to third-person demonstrative pronouns in the nominative (ὁ and ὅς) followed by particles.

5.2 ὁ and ὅς

§11. To establish similarities and differences between the possible clusters of pronoun and particle I will take retrieval of a singular masculine referent in the nominative as the basic material. In the ongoing narrative, the Homeric performer constantly manages the relevant referents in multiple ways. Frequently, as in (t9) above, a verb form suffices to successfully select the correct referent in the discourse model. In this chapter I have chosen to focus on lexical items that serve specifically to retrieve referents: pronouns.
§12. Such lexical items still cover a wide range of words, including αὐτός, ἐγώ, σύ, μίν, ἑ, τοι, οὗτος, κεῖνος, [31] ὅδε, ὅς, and ὁ. The lexical item most frequently used to retrieve referents in a narrative is the anaphoric third-person pronoun referring to a character. Within this category, ὅς [32] and ὁ [33] give the lightest marking short of a verb form only (null anaphor). Because these forms are by far the most frequent, and are commonly accompanied by particles, they form the core material for my comparative study. Finally, I focus on the nominative rather than the oblique cases because the nominative form is often syntactically superfluous. The nominative forms are particularly interesting, since in a pro-drop language like ancient Greek the grammatical subject is encoded in the verb form. That is to say, ὁ is not equivalent to τόν in referential terms, since the former may in the right context be elided, while the latter is more often indispensable. [34] Although I do not focus on the oblique forms of the anaphoric pronoun, I discuss a small number of instances, mostly in the footnotes.
§13. Among the masculine singular nominative forms, the relationship between ὅς and ὁ is complex in archaic Greek. [35] In classical Attic prose, ὅς is the masculine singular of the relative pronoun, while ὁ is the masculine singular definite article or a weak demonstrative pronoun. [36] In Homeric epic, and through its strong influence in Pindar and other lyric as well, the distinction is not so clear. In this early Greek, the two words can have the respective values they have in classical Greek, but beyond that both can function as demonstrative or relative pronouns. This dual functioning presents significant problems, since it is not always clear from the Greek if a clause should be taken as a relative clause introduced by the pronoun or as a new main clause with a demonstrative pronoun as the grammatical subject. Consider the following example:
          (...) τοῖσι δ’ ἀνέστη |
          Κάλχας Θεστορίδης | οἰωνοπόλων ὄχ’ ἄριστος, |
70       ὃς ᾔδη τά τ’ ἐόντα τά τ’ ἐσσόμενα πρό τ’ ἐόντα |
Iliad 1.68-70
          (...) And amongst them stood up
          Kalchas son of Thestor, among augurs by far the best,
70      who knew what was, what would be, and what had been.
An alternative reading of the final line is: “...by far the best. He knew..,” if we interpret the pronoun as demonstrative rather than relative. [37] In (t10) it is perhaps unnecessary to choose a demonstrative reading, but consider this parallel:
          ἡ δ’ αἶψ’ ἐξ ἀγορῆς ἐκάλει κλυτὸν Ἀντιφατῆα, |
115    ὃν πόσιν, | ὃς δὴ τοῖσιν ἐμήσατο λυγρὸν ὄλεθρον. |
          αὐτίχ’ ἕνα μάρψας ἑτάρων | ὁπλίσσατο δεῖπνον |
Odyssey 10.114-116
          At once she called from the place of assembly the glorious Antiphates,
115    her husband, and he devised for them woeful destruction.
          Instantly he seized one of my comrades and made ready his meal, [38]
Here, Murray prefers to read the pronoun as demonstrative, and translates accordingly. In the language there is no formal distinction between the two possible readings of the pronoun, which suggests that Murray was guided by the context.
§14. In (t11) one reason for translating ὅς as “he” rather than “who” is the fact that it constitutes a narrative transition. The act following the mention of the husband (ὃν πόσιν) is not so much a description of this new referent – as might be expected in a relative clause – but rather an act introducing a new event, with the freshly introduced husband as the grammatical subject. [39] The fact that in (t11) an aorist (ἐμήσατο) follows the pronoun, whereas in (t10) it is a pluperfect (ᾔδη), standing in for the imperfect of οἶδα, contributes to this reading. [40] On a macrolevel the δή act in (t11) marks the beginning of a new scene within the narrative, which leads to Odysseus’ departure from the island of the Laestrygones. In discourse terms, ὅς in (t10) introduces an act that is unframed (the performer informs us about the character Kalchas), while ὃς δή in (t11) introduces a framed act, a continuation of the narrative. [41] This distinction may (unconsciously) play an important part in decisions of editors, regarding punctuation, and of translators.
§15. In Homer and Pindar both ὅς and ὁ are used to retrieve a masculine character. [42] To understand the use of ὁ/ὃς + particle, we must understand the nature of referent tracking in ancient Greek. To an English reader, ὁ may look like “he,” but the two pronouns are not equivalent. Because ancient Greek is a pro-drop language, neutral continuity of grammatical subject is signaled by null-anaphor constructions, with a predicate consisting of a finite verb only. [43] In a similar construction, English supplies the unaccented pronoun. ὁ in Homer, then, is not equivalent to unaccented “he” but to accented “HE.” [44]
          πρόσθε δ’ Ἀλέξανδρος προΐει δολιχόσκιον ἔγχος, |
          καὶ βάλεν Ἀτρεΐδαο κατ’ ἀσπίδα πάντοσε ἴσην, |
          οὐδ᾽ ἔρρηξεν χαλκός, | ἀνεγνάμφθη δέ οἱ αἰχμὴ |
          ἀσπὶδ᾽ ἐνὶ κρατερῇ. | ὁ δὲ δεύτερον ὄρνυτο χαλκῷ |
350    Ἀτρεΐδης Μενέλαος | (...)
Iliad 3.346-350
          First Alexander sent off his far-shadowing spear,
          and he hit on Atreides’ shield, perfectly balanced,
          but the bronze did not break through, and its point was turned
          on the mighty shield. And HE in turn rushed with his bronze,
350    Atreus’ son Menelaus (...)
In this passage the narrator describes the fight between Paris (Alexander) and Menelaus. For this narrative stretch, those two characters are the only relevant referents, and the focus of the mind’s eye shifts steadily from one to the other. In lines 346 and 347, agency remains with Paris and as a result the new act in 347 is introduced without a pronoun (∅ βάλεν). [45] A few lines later we have followed the thrown spear to its mark, Menelaus, who has now become focused in the mind’s eye. As Tomlin explains, the nominative case is a grammatical reflex of attention management: in English the nominative typically refers to the referent that has become the focus of attention in the directly preceding discourse. [46] However, in instances like (t12) the new grammatical subject (Menelaus) is in focus in the mind’s eye, but has not yet been primed in the language when he is activated with ὁ δέ in 349. This is the reason that we find the pronoun ὁ, and not just the verb. [47] Contrary to our intuition, then, ὁ is a marked rather than an unmarked anaphoric expression and should be taken as such. [48]
§16. In other words, it seems that Menelaus was not completely accessible in the middle of line 349, and therefore was activated with a pronoun. However, activation cost is only one of many factors affecting the use of ὁ or ὅς in Homer and Pindar. [49] The corpus of Homeric epic and Pindaric song differs to a significant extent from the discourse that contemporary linguists generally engage with. Although Homeric and Pindaric discourse must still, by and large, follow the unwritten rules of communication, they possess an additional layer of artfulness and traditionality that will have an effect on the linguistic form beyond cognitive requirements. Chafe’s account of referent retrieval, which proposes that accessibility is only one of many factors determining the form of a referring expression, offers additional avenues of analysis. [50]
§17. To handle a complex narrative, the speaker’s mind attempts to maximize clarity while minimizing explicit reference; [51] and the hearer’s mind works on the assumption that this is indeed what the speaker does. [52] Therefore, when a referential expression appears to give more information than required (i.e. when it exceeds the referent’s required activation cost) an explanation is called for. [53] In the vast majority of cases, such instances of apparent overdetermination can be explained by considering other factors. Beyond activation cost I consider the relevance of discourse transitions, frame switches, contrastiveness, and “zooming” of the mind’s eye. [54]

5.3 ὁ/ὅς + particle in Homer

§18. Τhe following discussion of ὁ/ὅς + particle combinations focuses on the pragmatic functions of the acts they introduce. It is from the pragmatic perspective that the particles’ force can be understood, and a comparative study shows that the speaker’s choice for one particle over another is rarely arbitrary. I first explore the combination ὁ δέ (and ὃς δέ) since it is one of the most common combinations, and has come to be associated with the very specific grammatical function of marking subject change. Our analysis aims to separate the different contributions of the two elements (pronoun and particle) in order to come to a better understanding of the whole in its many contexts. Subsequently I turn to the other extremely frequent collocation ὅ γε, which has also received some attention in the literature, in this case as a marker of subject continuity. As with ὁ δέ, the reality in Homer is more complex, but unlike ὁ δέ the combination ὅ γε appears to be working as a cluster. [55] After these frequent and known combinations, I turn to those pronoun and particle combinations that generally go undiscussed, but are in fact crucial in guiding the narrative: ὁ δ᾽ἄρα, ὅ(ς) ῥα, and ὁ(ς) δή.

5.3.1 ὁ δέ

§19. The cluster ὁ δέ is probably the most well-known combination of pronoun and particle, but despite its frequency the collocation is not well understood. In order to gain a better understanding of the combination, it is paramount to separate the functions of its two constituent parts, contrary to common practice. The pronoun ὁ works like an accented pronoun in English, roughly equivalent to emphatic “HE,” and as such it is used regularly in instances featuring discontinuity of grammatical subject or another kind of referent switch. In II.2 I outlined the main function of δέ as the “quintessential boundary marker,” following Bakker’s work. [56] This function suggests that the particle has nothing specifically to do with continuity or discontinuity in the tracking of referents in Homeric discourse. In fact, a combination of a finite verb + δέ is a common way of maintaining subject continuity. [57] If one wishes to link ὁ δέ to changes of subject one should be aware that the reason for this correlation is ὁ, not δέ. [58]
§20. There is more to be said about the combination and its relation to referential continuity or discontinuity. Janko writes that ὁ δέ “normally marks a change of grammatical subject.” [59] His claim is often true, especially after Homer, but it does not address the question of why ὁ δέ serves this purpose so well. In the combination ὁ δέ in Homer, the lightly emphatic pronoun invites the audience to find a reason for the emphasis – often a change of grammatical subject – whereas δέ simply marks the progress of the discourse.
§21. Since δέ marks the progress of discourse, and because most of Homeric discourse is framed narrative, δέ in Homer typically marks a continuation of framed discourse. In combination with the pronoun ὁ/ὅς, then, it is no surprise that ὁ δέ in Homer often introduces an act with a new grammatical subject. However, we must remember that ὁ as an anaphoric pronoun is only lightly emphatic; it is not as strong an anaphoric expression as a strong demonstrative, for example. Τherefore, even when ὁ δέ marks change of grammatical subject, the referent of ὁ must be highly accessible. Consider the following instance at the end of book twenty of the Iliad. After a long list of Achilles’ exploits on the battlefield the narrator caps the episode with two similes:
490    ὡς δ’ ἀναμαιμάει βαθέ’ ἄγκεα θεσπιδαὲς πῦρ |
          οὔρεος ἀζαλέοιο, | βαθεῖα δὲ καίεται ὕλη, |
          πάντῃ τε κλονέων | ἄνεμος φλόγα εἰλυφάζει, |
          ὣς ὅ γε [60] πάντῃ θῦνε | σὺν ἔγχεϊ | δαίμονι ἶσος |
          κτεινομένους ἐφέπων· | ῥέε δ’ αἵματι γαῖα μέλαινα. |
495    ὡς δ’ ὅτε τις ζεύξῃ βόας ἄρσενας εὐρυμετώπους |
          τριβέμεναι κρῖ λευκὸν | ἐϋκτιμένῃ ἐν ἀλωῇ, |
          ῥίμφά τε λέπτ’ ἐγένοντο | βοῶν ὑπὸ πόσσ’ ἐριμύκων, |
          ὣς ὑπ’ Ἀχιλλῆος μεγαθύμου | μώνυχες ἵπποι |
          στεῖβον ὁμοῦ νέκυάς τε καὶ ἀσπίδας· | αἵματι δ’ ἄξων |
500    νέρθεν ἅπας πεπάλακτο | καὶ ἄντυγες αἳ περὶ δίφρον, |
          ἃς ἄρ’ ἀφ’ ἱππείων ὁπλέων ῥαθάμιγγες ἔβαλλον |
          αἵ τ’ ἀπ’ ἐπισσώτρων· | ὁ δὲ ἵετο κῦδος ἀρέσθαι |
          Πηλεΐδης, | λύθρῳ δὲ παλάσσετο χεῖρας ἀάπτους. |
Iliad 20.490-503
490    As a portentous fire rages up deep glens
          of a dry mountain, and the deep forest burns,
          and driving it everywhere, the wind whirls the flame about.
          Thus HE rushed everywhere with his spear, like to a god,
          driving on his victims. And the black earth ran with blood.
495    As when someone yokes broad-fronted bulls,
          to crush white barley on the well-built threshing floor,
          and soon they are threshed out under the loud-bellowing bulls’ feet.
          Thus under brave-hearted Achilles the single-hoofed horses
          trampled corpses and shields alike; and with blood the axle
500    was all spattered below and the rims, those around the chariot,
          for them the drops from the horses’ hooves struck,
          and those from the wheels. And HE went to win glory,
          Peleus’ son, and with gore were spattered his invincible hands.
The audience cannot but picture the scene of Achilles tearing through the enemy ranks like a forest fire spurred on by the wind, trampling their bodies like grain on a threshing floor. In the final lines the narrator sketches an image of Achilles triumphant on a chariot spattered with blood, riding over the bodies of his adversaries. Although he has not been the grammatical subject since for the last eight lines, and has not been named in four lines, ὁ δέ suffices to restore Achilles as grammatical subject in 502. The images evoked are strong, but unlike some other similes they apply readily to the current situation on the battlefield. In 498 the horses are the subject (ἵπποι) and the axle (ἄξων) in 499, but Achilles is constantly at the forefront of our mind, literally in the center of the image, in focus. As in the case of Jason below (t32), ὁ in 502 does not just retrieve “Achilles” but it retrieves the raging and bloody Achilles that has just been created in the mental representation of the discourse.
§22. However, because ὁ/ὅς can retrieve all accessible masculine singular referents in the frame, ὁ δέ in framed discourse cannot be mapped directly onto the function “grammatical subject change.” There are several instances where a pronoun is used despite clear continuity of grammatical subject. Such cases of apparent overdetermination deserve closer inspection, as they may occur for several cognitive or stylistic reasons. In book nine of the Odyssey, Odysseus relates how he acquired the wine with which he would later intoxicate the Cyclops, saying:
          (...) ὅν μοι δῶκε Μάρων, | Εὐάνθεος υἱός, |
          ἱρεὺς Ἀπόλλωνος, | ὃς Ἴσμαρον ἀμφιβεβήκει, |
          οὕνεκά μιν σὺν παιδὶ περισχόμεθ’ ἠδὲ γυναικὶ |
200    ἁζόμενοι· | ᾤκει γὰρ ἐν ἄλσεϊ δενδρήεντι |
          Φοίβου Ἀπόλλωνος. | ὁ δέ μοι πόρεν ἀγλαὰ δῶρα· |
Odyssey 9.197-201
          (...) [the wine], which Maro gave me, son of Euanthes,
          the priest of Apollo, who had encompassed the Ismarus,
          because we had protected him with his son and his wife
200    out of reverence. For he [sc. Maro] lived in a wooded grove of
          Phoebus Apollo. And HE [sc. Maro] gave me splendid gifts:
After the first plural form περισχόμεθα, the singular verb form ᾤκει suffices to avoid ambiguity. It might seem all the more surprising, then, that the pronoun is used in the following act, even though there is total continuity of grammatical subject.
§23. An explanation for this overdetermination is readily available after the discussion of framed and unframed discourse in II.4. Both the imperfect tense of ᾤκει [61] and the particle γάρ suggest that the move in lines 200-201 is different from its surroundings. It is in fact a little piece of unframed discourse, where the performer turns to the audience and offers some information needed in order to understand the ongoing action in the narrative. In cognitive terms, the act starting with ᾤκει γάρ does not create the image of Maro in any kind of activity, but rather of his house in a sacred grove. The retrieval of Maro after that is therefore more fraught. [62] As we return to the contextual frame of the action, the pronoun (ὁ) turns attention from Maro “living in a grove” to Maro as he is now, having just been saved by Odysseus and his men.
§24. Similarly, in the following narrative about Poseidon intervening in the battle between Aeneas and Achilles, the anaphoric pronoun is used at one point despite continuity of grammatical subject:
          αὐτὰρ ἐπεὶ τό γ’ ἄκουσε | Ποσειδάων ἐνοσίχθων, |
          βῆ ῥ’ ἴμεν | ἄν τε μάχην καὶ ἀνὰ κλόνον ἐγχειάων, |
320     ἷξε δ’ ὅθ’ Αἰνείας ἠδ’ ὃ κλυτὸς ἦεν Ἀχιλλεύς. |
          αὐτίκα | τῷ μὲν ἔπειτα κατ’ ὀφθαλμῶν χέεν ἀχλὺν |
          Πηλεΐδῃ Ἀχιλῆϊ· | ὁ δὲ | μελίην εὔχαλκον
          ἀσπίδος ἐξέρυσεν | μεγαλήτορος Αἰνείαο· |
Iliad 20.318-323
          Now, when Poseidon the earth-shaker heard this,
          he set to go up through the battle and the hurtling of spears,
320     and he reached where Aeneas and glorious Achilles were.
          At once, he then shed a mist over the eyes of the one,
          Peleus’ son Achilles. And HE, the ashen spear well-shod in bronze
          he drew from the shield of great-hearted Aeneas,
In this passage ὁ δέ is problematic, since it would most naturally establish Achilles as the grammatical subject of the new act, instead of continuing to refer to Poseidon as it actually does. This apparent mismatch may have been Aristarchus’ reason for athetizing lines 322-324, since taking out these lines creates an attractive symmetry between τῷ μέν (sc. Achilles, 321) and Αἰνείαν δέ (325). [63] I follow the reading of the manuscripts and editions, however, which requires an explanation. In II.2.5 we devote some attention to small fronted acts that serve to guide the mind’s eye of the audience. These priming acts take the form of a referential expression + a particle (often δέ), and are followed by some performative discontinuity. Even though nothing linguistically suggests that there is a boundary after δέ (the accusative that follows seems in no way to be independent), the manuscripts suggest that some kind of discontinuity was assumed after ὁ δέ. [64] Here, the reason for such a discourse act is not obvious, but the motivation may be visual. After his arrival Poseidon sheds mist over Achilles’ eyes, which leads the mind’s eye to focus on Achilles. The expectation of the audience might have been that focus and agency stayed with Achilles, but in fact the scene moves back to Poseidon with ὁ δέ (322). The god then first enacts a ritual return of the spear to the hero, after which he performs a truly awesome deed: he throws Aeneas over the entire army to the other side of the battlefield. The scene is climactic and highly vivid, [65] and the use of the pronoun rather than a null anaphor may serve to prepare for the upcoming image that has Poseidon as its center.
§25. A study of ὁ δέ in Pindar reveals the same range of possibilities as in Homer, the difference being that ὁ δέ almost always marks change of grammatical subject. [66] Since in Pindaric song δέ marks boundaries between larger units of discourse (see II.3 §65), this need not come as a surprise. Besides this difference, ὁ δέ in Pindar has two additional functions. First, Pindaric ὁ sometimes comes close to its classical function as an article [67] – a use rare in the Homeric epics [68] – while retaining its Homeric function as a relative or demonstrative pronoun. [69] In two instances, ὁ δέ precedes a proper name, and there the boundary between demonstrative pronoun and article is more fuzzy than elsewhere in Pindar. [70] As in Homer, “there is an uncertainty about the relative vs. the demonstrative use of the pronouns, and (...) a complex interlacing of (...) functions among relative pronouns, definite articles, and articles used as pronouns.” [71] Second, the pronoun can be used as a forward-looking demonstrative in constructions like this gnomic expression:
(...) ὁ δ’ ὄλβιος, | ὃν φᾶμαι κατέχωντ’ ἀγαθαί |
Olympian 7.10
(...) He is fortunate, who is held in good repute [72]
Here and in the parallels, the sense of the act introduced by ὁ δέ is gnomic. [73] In all cases I read the first part of the thought as a copulative construction with ἐστί left out, which means that ὁ must be read as a demonstrative pronoun rather than as an article. Because in this context ὁ δέ introduces a gnṓmē – unframed, generalizing discourse – there is always a change of grammatical subject. However, in gnômai the referent of ὁ can be ambiguous; either it can be the indefinite “he”, or it can be the main referent of the preceding discourse, often the victor.
§26. Examples (t14) and (t15) demonstrate that ὁ δέ may accompany continuity of grammatical subject, provided that there is some other reason for emphasizing the referent. The combination ὁ δέ in Homer and Pindar reflects the whole range of its constituent elements: ὁ/ὅς as a marked anaphoric expression, sometimes only lightly marked, sometimes almost deictic, and δέ as the marker of discourse progression, marking stronger boundaries in Pindar than in Homer.

5.3.2 ὅ γε

§27. ὅ γε enjoys a special status among combinations of pronoun and particle: the LSJ, for example, specifically lists ὅ γε as a special construction under its discussion of ὁ. They describe the combination as follows: “Pron. ὁ, ἡ, τό made slightly (if at all) more emphatic by the addition of γε.” [74] This description of the function of γε exemplifies most scholarship on the particle, and it covers well the sense that is common to most instances of the particle. Denniston discusses γε in terms of “concentration,” which leads to the two further functions of marking limitation and intensification. [75] About γε after pronouns, Denniston says the following: “Naturally, in many cases γε is limitative: but in many others it is determinative: often it seems to be otiose, the pronoun apparently requiring no stress, or at most a secondary stress.” [76] By limitative, Denniston means that γε marks its host word (group) as the thing that the current claim holds true for at least; “determinative” fits his idea of “concentration,” and what other scholars (and the LSJ) describe as “emphatic”; “otiose” appears to mean redundant.
§28. As we shall see, the limitative function of γε (as described by Denniston) does not emerge from the cluster ὅ γε, and yet the particle is clearly not redundant. Rather, γε in ὅ γε mainly lends emphasis, but this idea needs refinement. What does it mean for a particle to make a pronoun (and specifically ὁ) more emphatic? In the following section I demonstrate three things: (1) distributionally, ὁ and ὅ γε are complementary: they occur in mutually exclusive positions in the act and the verse; (2) functionally, ὅ γε serves to retrieve a referent of which some aspect has to be supplied through inference; (3) when ὅ γε appears in contexts where there is continuity of grammatical subject, several factors contribute to the choice of ὅ γε over a null anaphor, including frame switches and transitions between narrator text and direct speech.
§29. First consider the position of ὅ γε, which provides the clearest indication that we are dealing with a cluster rather than a combination. In the discussion above, I have taken ὁ as the equivalent of the accented pronoun in English. In its relative and anaphoric functions, the pronoun tends to occur in act-initial position in Homer. It is thus a statistical anomaly that ὅ γε never occurs in act-initial position. Of course, γε is more mobile than δέ, which means that it can occur later in the act as well as in peninitial position. However, the positional tendency of the pronoun would suggest that at least in some cases we would find ὅ γε at the start of a new act. For metrical reasons ὅ γε cannot occur at verse beginning, but this does not hold for οἵ γε, which still shares the same positional limitation: it never occurs at act or at verse beginning. That is to say, the nominative pronoun occurs in act-initial position, and γε occurs in peninitial position in the act, but ὅ γε as a combination is never act initial. This makes the combination different from the sum of its parts in a significant way, and therefore we shall treat ὅ γε as a cluster. If the positional limitations of the cluster differ from that of its components, the same may be possible for the cluster’s function. Unlike ὁ δέ, the function of ὅ γε may be more limited than the range of functions ὁ and γε have on their own.
§30. In commentaries, ὅ γε is regularly described as resuming the subject of the current sentence; that is, it is regarded as marking grammatical subject continuity. [77] Considering once more the masculine nominative pronoun’s range of functions, however, we may predict that this generalization will not hold, and the numbers show that it does not. [78] Consider the following passage from Iliad book one, where Achilles considers whether he should attack Agamemnon or not:
          ὣς φάτο· | Πηλεΐωνι δ’ ἄχος γένετ’, | ἐν δέ οἱ ἦτορ |
          στήθεσσιν λασίοισι | διάνδιχα μερμήριξεν, |
190    ἢ ὅ γε | φάσγανον ὀξὺ ἐρυσσάμενος παρὰ μηροῦ |
          τοὺς μὲν ἀναστήσειεν, | ὃ δ’ Ἀτρεΐδην ἐναρίζοι, |
Iliad 1.188-191
          Thus he spoke. Peleus’ son was distressed, and his heart
          in his shaggy breast debated two ways:
190    whether HE, having taken his sharp sword from his thigh,
          should make the others leave, and HE should kill Atreus’ son...
The text illustrates neatly how problematic generalizations about ὁ δέ and ὅ γε in Homer are: ὅ γε in 190 marks change of grammatical subject from ἦτορ, “heart,” to Achilles, whereas ὁ δέ in 191 accompanies subject continuity. [79] The choice of the pronoun over null anaphor is expected in 190, since null anaphor would have led to the jarring image of the heart drawing a sword. In 191 the anaphoric pronoun serves to juxtapose the image of the assembly dispersing with the image of Achilles and Agamemnon staying and fighting. This explains the use of the pronoun in both instances, but the question remains what the function of γε is in ὅ γε.
§31. The reason for the addition of γε, although hard to determine, may have been prosodic: ὁ alone is more easily lost than ὁ followed by a clitic and at times a whole syllable. [80] However, this is not enough to explain why we find γε instead of another enclitic, such as ῥα. Whatever the range of functions of ὅ γε when it grammaticalized as a cluster, a particular function of γε must have led to the development of the cluster in the first place. Before moving on to more instances of ὅ γε in Homer, let us examine this question more closely.
§32. One way to analyze the function of γε is to describe it in terms of focus. The term focus has multiple uses, both technical and intuitive, and use of the term in classical scholarship is diffuse. Denniston spoke of γε as a primarily “limitative” particle, reducing an expression’s applicability to “at least” the word (group) marked by γε. Building on Denniston’s ideas with the addition of the terminology of functional grammar, Bakker speaks of γε in wishes as marking “exclusive focus,” to contrast its function with the inclusive focus marked by περ. [81] Both Sicking and Wakker build on Bakker’s work but describe γε more generally as a “focus particle.” [82]
§33. A number of scholars have attempted to adapt or refine the idea of γε as a focus particle. About γε in Aristophanes, Tsakmakis says: “γε is not a focalizer which can be indiscriminately attached to any element of the utterance (even if that is focalized), but it can only be attached to a word which coheres with the preceding utterance.” [83] About ὅ γε specifically, Bonifazi says: “γε gives prosodic and semantic prominence to ὁ.” This interpretation conflates two elements that are consecutive, I believe: γε does indeed give prosodic prominence to ὁ, and it is this prosodic prominence that leads to an interpretation of ὅ γε as in some way emphatic (“semantic prominence”). Then she argues that in ὅ γε, γε “emphasizes something relationally new about somebody referentially old.” [84] This claim requires some unpacking. If the “somebody referentially old” is the referent of the pronoun ὁ, then the “something relationally new” must be contained in the rest of the discourse act in which ὅ γε occurs. However, I believe that γε in ὅ γε has scope only over the pronoun, not over the entire act.
§34. In my reading, γε emphasizes the pronoun, which would then have to be both “relationally new” and “referentially old.” I agree with Tsakmakis and Bonifazi that in ὅ γε the pronoun refers to someone who is referentially old and thus “coheres with the preceding utterance.” In line with Bonifazi’s argument, moreover, we may consider that the pronoun can refer to something at once old and new. More specifically, I argue that ὅ γε retrieves an accessible referent in a form that is to be inferred. [85]
§35. Consider the following passage from the Odyssey. While Odysseus is sailing past the Sirens, they call to him in an attempt to make him stop:
οὐ γάρ πώ τις τῇδε παρήλασε | νηῒ μελαίνῃ, |
πρίν γ’ ἡμέων μελίγηρυν ἀπὸ στομάτων ὄπ’ ἀκοῦσαι, |
ἀλλ’ ὅ γε τερψάμενος νεῖται | καὶ πλείονα εἰδώς. |
Odyssey 12.186-188
For never yet has a man rowed by here in a black ship,
before hearing the honey-sweet voice from our lips;
no, HE enjoys it and travels on, knowing more in fact.
The passage seems straightforward, but on closer inspection it involves an interesting shift. Whereas in the majority of instances of ἀλλ᾽ ὅ γε there is a single referent who functions as the agent in both parts of the construction, here the tracking of referents is more complicated. [86] The initial assertion is that “no-one sails by before hearing the Sirens,” so what does ὅ γε refer to? To no-one? The participle τερψάμενος makes it clear that who is meant by ὅ γε is the sailor who did indeed stay and listen, and was pleased as a result. The automatic inference triggered by “no-one sails by” creates a pool of available referents who did sail by, one of whom can then be referred to with ὅ γε. [87] A similar example occurs in the following gnomic thought that Odysseus imparts to the suitors:
τῷ μή τίς ποτε πάμπαν ἀνὴρ ἀθεμίστιος εἴη,
ἀλλ’ ὅ γε σιγῇ δῶρα θεῶν ἔχοι, ὅττι διδοῖεν.
Odyssey 18.141-142
Therefore let no one ever be an altogether lawless man,
No may HE keep in silence the gifts of the gods, whatever they may have given.
Again, an unproblematic reading belies the underlying referential complexity. As in (t18) ὅ γε directs attention to the indefinite referent that we infer from the preceding act. [88] Alternatively, one might say that ὅ γε triggers the creation of a generic referent, about which the only inferrable information at the point of utterance is that he has not “been a lawless man.” It is this cognitive action, I propose, that justifies the use of ὅ γε over null anaphor or the pronoun only. [89]
§36. The process of inference also applies in the occurrences of ὅ γε in the following passage from the Iliad. The episode reveals more about anaphoric reference than just the use of ὅ γε, however, so allow me a brief excursus. Pandarus is distraught by Diomedes’ relentless attack on the Trojans, and says to Aeneas:
180    “Αἰνεία | Τρώων βουληφόρε χαλκοχιτώνων |
          Τυδεΐδῃ μιν ἔγωγε δαΐφρονι πάντα ἐΐσκω, |
          ἀσπίδι γιγνώσκων | αὐλώπιδί τε τρυφαλείῃ, |
          ἵππους τ’ εἰσορόων· | σάφα δ’ οὐκ οἶδ’ εἰ θεός ἐστιν. |
          εἰ δ’ ὅ γ’ ἀνὴρ ὅν φημι | δαΐφρων Τυδέος υἱὸς |
185    οὐχ ὅ γ’ ἄνευθε θεοῦ τάδε μαίνεται, |
Iliad 5.180-185
180    “Aeneas, counsellor of the bronze-clad Trojans,
          to Tydeus’ battle-minded son I liken him completely,
          recognizing his shield and his crested helmet,
          and looking at his horses. With certainty I do not know if he is a god.
          And if HE is the man I mean, the battle-minded son of Tydeus,
185    then HE does not rage like that without a god.
Before we look at the two instances of ὅ γε, consider the use of μιν in line 181. μιν is an unstressed (enclitic) third-person pronoun in the accusative that serves to retrieve accessible referents. Pandarus’ speech starts in line 180, and he has not mentioned the referent of μιν (“that man”) up to this point. However, in the preceding turn Aeneas has already pointed the man out (τῷδ᾽...ἀνδρί 174). The referent of the strong demonstrative ὅδε is the man they both see, and Pandarus retrieves him with the much less emphatic μιν, since by now he is well established in both their mental discourse models (as well as in that of performer and audience, of course).
§37. To better explain this passage, we must turn to the difference between anaphora and deixis. I follow Cornish’ definition of the terms, which can be put in simple terms: a speaker uses deixis when she wants to bring a referent into focus in the current discourse model, and anaphora when she retrieves one that is already accessible. [90] Just before (t20), Aeneas introduces “the man” into the conversation with the expression τῷδ᾽...ἀνδρί, “that man,” a combination of demonstrative pronoun and noun—this is deixis. [91] This same referent, who is still in the discourse context (i.e. visible) for the two interlocutors, is from that moment onward in focus in their discourse model, and can be referred to with the enclitic anaphoric pronoun μιν. Deixis is thus not inherently linked to a reference to something outside of the discourse: the crucial question is whether a referent is part of the current discourse model or not. In fact, both anaphora and deixis function outside the text, since they do not primarily interact with the preceding text but with the mental representation of the discourse.
§38. This brings us to ὅ γε in lines 184 and 185. The first ὅ γε refers again to “that man there,” who Pandarus is not quite sure is even human. [92] Pandarus speculates that it may be Diomedes, since the man seems to bear Diomedes’ shield. The following conditional clause shows that the referent is interactionally clear, but undetermined in textual terms: “If THAT ONE is the man I mean, the battle-minded son of Tydeus...”. Both interlocutors know who they are talking about, but his identity is unknown. Thus, ὅ γε refers to the entirety of Aeneas and Pandarus’ suppositions about the man, including the possibility that he is a god or that he is Diomedes. Finally comes the apodosis, with the second instance of ὅ γε. If ὅ γε in 184 refers to the man in the middle of the spectacle, ὅ γε in 185 no longer has the same referent, since an assumption underlies the utterance of line 185 that the man is indeed human and in fact Diomedes: “if THAT (ὅ γε) is Diomedes, then HE [sc. Diomedes] (ὅ γε) cannot be raging without divine help.” The second ὅ γε marks this discontinuity of reference in the mental representation of the discourse: it refers no longer to “that man” but to “that man, Diomedes.” The audience can follow this interaction only by taking into account the development of the referent in the mental representation of the discourse, and this inferential process is foregrounded through the use of ὅ γε.
§39. The implied elements pertaining to the referent of ὅ γε can have local relevance, as in (t20) above, but they can also reach beyond the current narrative scene. Telemachus describes the fight between the stranger (Odysseus) and the beggar Irus to his mother Penelope:
οὐ μέν τοι ξείνου γε καὶ Ἴρου μῶλος ἐτύχθη
μνηστήρων ἰότητι, βίῃ δ’ ὅ γε φέρτερος ἦεν.
Odyssey 18.233-234
Although, the fight of this stranger and Irus did not end
according to the will of the suitors: in might, HE was stronger.
The situation here is different from most preceding examples, since the two available referents have both been mentioned only in an oblique case. Neither a null anaphor nor a pronoun suffices to retrieve Irus or the stranger. In the earlier narrative, however, it was told that the “stranger” was indeed the stronger, and in fact the whole co-text suggests that the suitors backed Irus. As a result, the referential ambiguity is resolved by the time that φέρτερος was uttered. [93] There is yet another layer to this reference. In the mind of Telemachus, who is speaking, and in the minds of performer and audience, ὅ γε refers not to a stranger but to Odysseus. The reference means something different to the internal audience (Penelope and the suitors) than to the performer and his audience, and this layering has its effect on the content. Bonifazi has demonstrated convincingly that references to the disguised Odysseus are particularly sophisticated in this part of the Odyssey, and there is no doubt in my mind that this instance is another example of this complexity. [94] The performer uses ὅ γε because the referent – and the sense of Telemachus’ utterance – can only be grasped fully if the listener makes the necessary inferences. It may be surprising to Penelope and the suitors that the stranger beat Irus, but this is decidedly not the case for Telemachus or the audience: obviously Odysseus is stronger in might than the resident beggar of his own palace.
§40. The consideration of the larger discourse is likewise indispensable in the following case of apparently superfluous reference to Telemachus. After he has been washed by Nestor’s youngest daughter, he is retrieved with ὅ γε:
αὐτὰρ ἐπεὶ λοῦσέν [sc. Πολυκάστη] τε καὶ ἔχρισεν λίπ’ ἐλαίῳ, |
ἀμφὶ δέ μιν φᾶρος καλὸν βάλεν | ἠδὲ χιτῶνα, |
ἔκ ῥ’ ἀσαμίνθου βῆ | δέμας ἀθανάτοισιν ὁμοῖος· |
πὰρ δ’ ὅ γε Νέστορ’ ἰὼν | κατ’ ἄρ’ ἕζετο, | ποιμένα λαῶν. |
Odyssey 3.466-469
Now, when she had washed him and anointed him richly with oil,
around him she threw a beautiful cloak, and a tunic.
Out of the bath he came, his body the immortals alike,
and HE went next to Nestor, and sat down by the shepherd of men.
The addition of γε in this instance may not appear to be needed, since Telemachus is surely constantly in focus as he is being washed and clothed. His continued prominence in the mental representation of discourse is confirmed by the fact that he is retrieved in line 468 as the unexpressed subject of βῆ. Since Telemachus remains the subject, ὅ γε looks marked in 469. I propose that we read the overdetermination here as an instruction to visualize not just Telemachus, but Telemachus as he is after the ablutions described in the preceding lines. The instruction is perhaps anticipated by the description δέμας ἀθανάτοισιν ὁμοῖος (469). It is not just Telemachus who joins Nestor, but it is Telemachus looking like a god. The same expression is used of Odysseus when he arrives at the palace of the Phaeacians and after he has been washed by Eurynome in book twenty-three. [95] Even more salient, Telemachus himself utters similar words when he mistakes his father for a god during the recognition scene in book sixteen. [96] The scene in book three thus reveals its importance on a macro-discursive level. The first books of the Odyssey are about how Telemachus finds himself in the role of man of the house. Perhaps the moment when he joins Nestor is the moment when he finally becomes a worthy heir to his father, and thus like his father looks “like a god.” If so, it really is a new Telemachus that sits down next to Nestor. [97]
§41. Now, since ὁ is itself a lightly emphatic pronoun, ὅ γε can also function to create an explicit or implicit contrast. [98] In this case, ὁ and γε work together: the emphatic pronoun suggests a contrast, while the particle triggers the implied opposite of the stressed referent. Contrastiveness can be marked with regard to referents already mentioned, to upcoming referents, or to implied referents. [99] Consider the following example of the construction ἦ τοι ὅ γε, from Theoclymenus’ words to Penelope:
“ὦ γύναι αἰδοίη Λαερτιάδεω Ὀδυσῆος, |
ἦ τοι ὅ γ’ οὐ σάφα οἶδεν, | ἐμεῖο δὲ σύνθεο μῦθον· |
Odyssey 17.152-153
“Revered wife of Laertes’ son Odysseus,
Let me tell you, HE does not know it clearly, do listen to my word.
ὅ γε creates the contrast: “Don’t listen to him, listen to me.” The referent of ὁ is highly accessible, since it refers to the last speaker (Telemachus), to whom both current interlocutors have presumably been listening. [100] Thus he is part of the speech situation, visually accessible, and in focus in the discourse model. [101]
§42. Consider one final example of subject continuity where ὅ γε serves to set up a contrast with an upcoming referent: [102]
          αἶψα δὲ νῆας ἔπηξε, | πολὺν δ’ ὅ γε λαὸν ἀγείρας |
665    βῆ φεύγων ἐπὶ πόντον· | ἀπείλησαν γὰρ οἱ ἄλλοι |
          υἱέες υἱωνοί τε βίης Ἡρακληείης. |
Iliad 2.664-666
          Quickly he built ships, and HE, after gathering many men,
          went fleeing over the sea. For the others threatened him,
          the sons and grandsons of mighty Heracles.
In the little narrative about Tlepolemus we hear how he kills his uncle, and then starts building ships: he has to flee, because the other children of Heracles are on his tail. The contrast is made explicit by the adjective ἄλλοι, which justifies the use of ὅ γε in 664. As (t24) demonstrates, contrastiveness is a factor that functions separately from activation cost: at that point Tlepolemus is clearly accessible. [103]
§43. In the following complex instance ὁ is placed relative to a contrasted, hypothetical version of its referent. This hypothetical version of the referent is triggered by the very utterance of ὅ γε. In her conversation with Aphrodite about Paris, Helen accuses the goddess of weakness:
μηδ’ ἔτι σοῖσι πόδεσσιν ὑποστρέψειας Ὄλυμπον, |
ἀλλ’ αἰεὶ περὶ κεῖνον ὀΐζυε | καί ἑ φύλασσε, |
εἰς ὅ κέ σ’ ἢ ἄλοχον ποιήσεται| ἢ ὅ γε δούλην. |
Iliad 3.407-409
May you not turn your feet toward Olympus yet,
but always suffer for him and guard him,
until he makes you his concubine or HE makes you a slave.
The passage is far from straightforward, but in line 409 ποιήσεται retrieves Paris, who is clearly very accessible at that point (κεῖνον, ἑ 408). Since the construction changes from imperative to indicative in 409, the third person verb form suffices to disambiguate among the three available referents: I (Helen), you (Aphrodite), and he (Paris). Then follows ὅ γε in the second part of the disjunction. The occurrence of ὅ γε in either part of a disjunctive construction is well documented, but little discussed. [104] Contrary to expectation, ὅ γε in disjunctions is not used to juxtapose two referents, but rather two possible events involving the same referent. [105] Here the two items in disjunction, ἄλοχον and δούλην, are syntactically symmetrical and semantic opposites. However, the addition of ὅ γε to the second suggests that the expression ἢ ὅ γε δούλην must be read with emphasis placed on ὁ (“or HE makes you a slave”). The opposite of this statement is not “until he makes you a CONCUBINE” [106] (this is the opposite of “until he makes you a SLAVE”) but rather of “YOU make him a slave.” It is exactly this presupposition that is triggered through the use of ὅ γε. [107] The expected situation in Greek culture is that the goddess of love makes a human her slave, not the other way around. [108] The markedness of the (hypothetical) situation is brought out by the combination of the pronoun and the particle. [109]
§44. The previous examples all illustrate that some of the force of γε remains in the cluster ὅ γε. However, because ὅ γε functions as a cluster, the force of γε sometimes appears to be extremely weak or even lost. In those instances, ὅ γε looks like a formal and metrical variant of ὁ, a weak demonstrative with no further pragmatic enrichment. In the second book of the Iliad we are told the story of Agamemnon’s scepter, a piece of unframed discourse introduced just before the king starts to speak in the council.
ἔστη σκῆπτρον ἔχων | τὸ μὲν Ἥφαιστος κάμε τεύχων. |
αὐτὰρ ὃ αὖτε Θυέστ’ Ἀγαμέμνονι λεῖπε φορῆναι, |
πολλῇσιν νήσοισι καὶ Ἄργεϊ παντὶ ἀνάσσειν. |
τῷ ὅ γ’ ἐρεισάμενος | ἔπε’ Ἀργείοισι μετηύδα· |
Iliad 2.101, 107-109
He stood up holding the scepter, over which Hephaestus toiled to make it.
Now this Thyestes in turn left it to Agamemnon to carry,
to rule over many islands and all of Argos.
HE, leaning on this, spoke words to the Argives:
After the story of the scepter, the current frame is recalled with τῷ ὅ γε, the first pronoun retrieving the scepter, and the second guiding attention toward Agamemnon. [110] The latter had been out of focus for several verses, and even though his name is mentioned in the oblique in line 107, this is not actually the referent of ὅ γε. Rather, the act containing ὅ γε re-establishes the frame of Agamemnon in the council, and that is why the pronoun is used rather than a simple verb form. There is, however, no need for further emphasis: γε is there only to accompany the pronoun in its non-initial position. Compare the following example from the Odyssey:
          οῖσιν δ’ Εὐπείθης ἀνά θ’ ἵστατο καὶ μετέειπε· |
          παιδὸς γάρ οἱ ἄλαστον ἐνὶ φρεσὶ πένθος ἔκειτο, |
          Ἀντινόου, | τὸν πρῶτον ἐνήρατο δῖος Ὀδυσσεύς· |
425    τοῦ ὅ γε δάκρυ χέων ἀγορήσατο καὶ μετέειπεν· |
Odyssey 24.422-425
          Among them Eupeithes stood up and spoke –
          for over his child comfortless grief lay on his heart,
          over Antinous, whom godlike Odysseus first killed –
          HE, weeping over him, addressed the assembly and said:
The co-text is essentially the same as in (t27): a speaker is introduced, then follows a brief piece of unframed discourse (γάρ), [111] and finally a pronoun referring to the unframed discourse (τοῦ) and a second pronoun with γε, in the act that re-establishes the narrative frame.
§45. In the above two examples ὅ γε is uttered just before the start of direct speech. This may not be a coincidence. In about a quarter of the instances where ὅ γε accompanies subject continuity, it is in the line right before or after direct speech. [112] The explicit marking of the referent in these instances serves to avoid confusion about the source of the upcoming thoughts or words, or to re-establish the speaker as an agent after his speech has finished. The construction ἦ τοι ὅ γ’ generally occurs in the following formulaic speech-capping verse: [113]
ἦ τοι ὅ γ’ ὣς εἰπὼν κατ’ ἄρ’ ἕζετο, τοῖσι δ’ ἀνέστη
Hey [114] HE, having spoken thus, sat down. And among them stood up
Since this verse follows direct speech, subject continuity is implicit. It is not hard to see that in this construction ὅ γε serves to set up a light contrast between this speaker and the next.
§46. Immediately after direct speech is one context where ὅ γε consistently differs from ὁ δέ. While ὅ γε marks continuity of grammatical subject after direct speech, ὁ δέ marks subject change. [115] This difference can be explained from the constituent order of the two constructions. With ὅ γε, we find “X ὅ γε | participle | finite verb,” whereas with ὁ δέ we find “finite verb | ὁ δέ.” In the latter construction, null anaphor would have been the natural marker of subject continuity.
§47. An extension of the use of ὅ γε to put particular cognitive focus on a referent can be found in the following example. Like the priming acts discussed in II.2.5, an emphatic pronoun may serve to direct extra attention to a certain referent. A priming act serves to mentally (or visually) turn to or zoom in on a referent, [116] which may in turn create the expectation that this referent will project over a significant piece of upcoming discourse. We can find a cognate of this construction in a small number of instances of ὅ γε, [117] especially just before direct speech or indirect thought. [118] In other contexts, it can also project the prominence of a character, as in the introduction of Alcinous, in Nausicaä’s description of the palace of the Phaeacians to Odysseus: [119]
ἔνθα δὲ πατρὸς ἐμοῖο θρόνος ποτικέκλιται αὐτῇ, |
τῷ ὅ γε οἰνοποτάζει ἐφήμενος | ἀθάνατος ὥς. |
Odyssey 6.308-309
And there is my father’s throne, leant against that .
On that HE sits and drinks his wine, like unto an immortal.
The reference to Alcinous here serves at once as the climax of the imagined entrance into the palace, and as the beginning of the long episode in which Alcinous and Odysseus are the main characters. At this crucial moment Alcinous is granted agency with some emphasis, which prepares the audience for his importance in the upcoming narrative.
§48. As a final consideration, I discuss the construction ἀλλ᾽ ὅ γε, which almost exclusively accompanies continuity of grammatical subject. [120] In the vast majority of cases, the construction takes the following form: “he did not do X, ἀλλ᾽ ὅ γε did Y.” [121] That is to say, these are occurrences of ὅ γε within the classical “οὐ X ἀλλά Y” construction, where ἀλλά equals German sondern. [122] This juxtaposing construction can be used with scope over single noun phrases or over (generally longer) verb phrases. The combination ἀλλ᾽ ὅ γε in Homer always juxtaposes an entire verb phrase with a preceding negated one: after positing what the referent will not do, the action resumes with the statement of what he will do. Consider the following typical instance, where Sarpedon has just heard of the wounding of Glaucus:
Σαρπήδοντι δ’ ἄχος γένετο Γλαύκου ἀπιόντος |
αὐτίκ’ ἔπειτ’ ἐνόησεν· | ὅμως δ’ οὐ λήθετο χάρμης, |
ἀλλ’ ὅ γε | Θεστορίδην Ἀλκμάονα δουρὶ τυχήσας |
νύξ’, |
Iliad 12.392-395
To Sarpedon sorrow came at Glaucus’ leaving,
right when he had noticed; still he did not forget the fight.
No, HE, coming upon Thestor’s son Alcmaon with his spear,
struck him,
There is clear subject continuity from line 393 onward, yet after two verbs with null subjects (ἐνόησεν, λήθετο) we find ἀλλ᾽ ὅ γε. There is a significant difference between the sense of the verbs before and after the pronoun and particle combination, marking the contrast between inaction and action.
§49. ἀλλ᾽ ὅ γε consistently introduces a new action, often marking the beginning of a new scene. The verb in the negated act of the construction is generally a reflection of some inner state of the referent (he was not afraid), [123] a static verb (he did not stay), [124] or a forward-oriented statement (he did not fulfill). [125] What these verbs have in common is that they interrupt the action by explicitly considering and dismissing a counterfactual situation. [126] Pragmatic projection is inherent in the construction, since the assertion that someone did not do one thing suggests that he did do another. The onus of the construction, then, is on the second part, for which the first part serves as a launching pad. From this perspective, ἀλλ᾽ ὅ γε works to resume the action, and this new beginning justifies the use of the pronoun even if the referent is already in focus. The occurrence of ἀλλ᾽ ὅ γε even when there is apparent subject continuity can thus be explained because some kind of cognitive redirection is needed – suggested by ἀλλά – just before the occurrence of the cluster ὅ γε. The apparent overdetermination can be read as an attempt to bridge this gap and direct attention toward the right referent.
§50. I have devoted a great deal of attention to ὅ γε for the simple reason that it is a very hard combination to pin down. With due attention to the direct co-text, context, and the larger narrative, however, its functions can be discussed. The function of γε that manifests in ὅ γε is the particle’s ability to activate an implication about its host word (group). This means that in a number of instances ὅ γε serves to retrieve an inferrable referent, or an accessible referent in an inferrable new form. Since ὅ γε is a grammaticalized cluster, this pragmatic enrichment of γε is not evident in every instance. In a number of cases ὅ γε appears to be nothing more than a formal variant of the demonstrative pronoun in limited positional contexts. The use of ὅ γε in these instances can be explained from the function of the lightly emphatic pronoun, and serves to put the referent firmly into cognitive focus. In this function, ὅ γε occurs particularly often right after frame switches and before and after direct speech. The function of the cluster must be regarded as a continuum with at one end ὅ γε as a variant of ὁ, and at the other end ὅ γε carrying a complex of inferences to be added to the referent. This continuum can be explained from the cluster’s origin in ὁ and γε. Every instance of ὅ γε in Homer represents a point on the sliding scale. In any case, it does not do justice to the discourse function of the cluster to remark that ὅ γε is “semantically redundant.” [127]

5.3.3 ὁ δ᾽ἄρα and ὅ(ς) ῥα

§51. In contrast to the flexible cluster ὅ γε, ἄρα only follows anaphoric pronouns in very limited contexts. In the present section, I first discuss the cluster ὁ δ᾽ ἄρ(α) in Pindar and Homer, which typically works as a marker of frame recall. Then follows an analysis of ὅ(ς) ῥ(α), a combination that can introduce two distinct constructions: either an unframed act with an imperfect, or a framed act with the aorist.
§52. In II.4, I argue that the use of ἄρα in Homer and Pindar is a sign of the speaker’s management of discourse memory. [128] The particle introduces acts containing information that the speaker expects to be accessible in the hearer’s mind. When ἄρα comes after anaphoric and relative pronouns, it likewise accompanies accessible information in the discourse memory that is not currently being attended to. Consider first this instance of ὁ δ᾽ἄρα in Pindar:
          (...) ξένου Λάκωνος Ὀρέστα.

          τὸν δὴ [129] φονευομένου πατρὸς
          θάνεν μὲν αὐτὸς ἥρως Ἀτρεΐδας
          ἵκων χρόνῳ κλυταῖς ἐν Ἀμύκλαις,

          μάντιν τ’ ὄλεσσε κόραν, ἐπεὶ ἀμφ’ Ἑλένᾳ πυρωθέντας
          Τρώων ἔλυσε δόμους ἁβρότατος. ὁ δ’ ἄρα γέροντα ξένον
35      Στροφίον ἐξίκετο, νέα κεφαλά, Παρνασσοῦ πόδα ναίοντ’·
Pythian 11.16-17 and 31-35
          (...) of Orestes the Laconian guest.

          Him actually, at the slaughter of his father [sc. Agamemnon],
          18 – 21 how Orestes escaped from Klytaemnestra
          22 – 25 Klytaemnestra’s motives
          25 – 30 gnômai
          He himself died, the hero son of Atreus [sc. Agamemnon],
          arriving in time in renowned Amyklai,

          and he brought death on the seer girl, after over Helen he had despoiled
          the burnt down houses of the Trojans of their luxury. So HE [sc. Orestes], the young boy, [130]
35      went to his aged host, Strophius, living at the foot of Parnassus.
After a gnomic passage, a demonstrative pronoun and a full noun phrase serve to retrieve Agamemnon (αὐτὸς ἥρως Ἀτρεΐδας 31). [131] Although he had been mentioned in the oblique before (φονευομένου πατρός 17, σὺν Ἀγαμεμνονία 20), he had not yet been fully attended to, only mentioned as part of an event relevant to Orestes. Then Pindar narrates a selection of Agamemnon’s adventures in reverse: he elliptically retrieves the story of Agamemnon’s taking of Priam’s daughter Cassandra, and then bringing her to Mycenae where she too is killed by Clytaemnestra (in Pindar’s version, lines 19-21) with the four words μάντιν τ’ ὄλεσσε κόραν. The particle τε here, I submit, serves to mark the sharedness of this episode, and allows Pindar to waste no more words in telling it. [132] The mention of Agamemnon’s death activates the event of line 16 again in the hearer’s mind. This makes Orestes, who has not been named or even mentioned for seventeen lines, accessible enough to be referred to with ὁ. In the preceding lines Agamemnon has been the grammatical subject, so the pronoun suggests a change of subject; the only other available masculine singular referent is Orestes. [133] The possible ambiguity in the anaphoric expression is perhaps the reason for the apposition νέα κεφαλά in line 35, [134] since this can only refer to Orestes. Finally, this passage demonstrates that ἄρα functions on the level of the larger discourse or interaction rather than as a link between two contiguous clauses. In narrative terms, ἄρα recalls the frame of the main storyline: lines 32-34 function as a little flashback, told regressively, and δ᾽ ἄρα serves to re-activate the main narrative frame of Orestes’ story. At the same time ἄρα marks its host act, as well as the upcoming narrative, as a part of the tradition shared between Pindar and his audience.
§53. This reading of (t31) is supported by both other instances of ὁ δ’ ἄρα in Pindar (Olympian 10.43 and Pythian 4.78), [135] one of which deserves fuller discussion. Pindar’s fourth Pythian ode is famous for its intricate narrative, an impression reinforced by the passage containing ὁ δ᾽ ἄρα. In a song that is at least partly about his quest, Jason is mentioned once in the beginning of the song (in an oblique case: εἶπε [sc. Medea] δ’ οὕτως ἡμιθέοισιν Ἰάσονος αἰχματᾶο ναύταις, “Medea spoke thus to the demi-god sailors of Jason the spearman,” lines 11-12), but is never named again. First Pindar gives Medea’s long prophecy and then he turns, in line 70, to the start of the journey of the Argonauts. This leads him to the episode of Pelias and Jason. Pelias receives a prophecy that a stranger with one sandal will come and be a threat to him and his throne. Later, this man actually arrives:
          ἦλθε δέ οἱ κρυόεν πυκινῷ μάντευμα θυμῷ, |
          πὰρ μέσον ὀμφαλὸν εὐδένδροιο ῥηθὲν ματέρος |
75       τὸν μονοκρήπιδα πάντως ἐν φυλακᾷ σχεθέμεν μεγάλᾳ, |
          εὖτ’ ἂν | αἰπεινῶν ἀπὸ σταθμῶν | ἐς εὐδείελον
          χθόνα μόλῃ κλειτᾶς Ἰαολκοῦ, |

          ξεῖνος αἴτ’ ὦν ἀστός. | ὁ δ’ ἆρα [sc. Iason] χρόνῳ
          ἵκετ’ αἰχμαῖσιν διδύμαισιν | ἀνὴρ ἔκπαγλος· |
Pythian 4.73-79
          And there came to him [sc. Pelias] a prophecy chilling to his shrewd heart,
          spoken at the central navel of the well-wooded mother:
75      to always be fully on guard against the man with one sandal,
          when, from the high dwellings, he came to the sunny
          land of famous Iolkos;

          be he a stranger or a townsman. And HE of course in time
          did come, with two spears, a terrible man.
I read ἆρα rather than the ἦρα proposed by Schroeder and printed in most editions. [136] Braswell regards ἆρα as a prosodically enriched form of ἄρα, and I would add that it may be regarded as pragmatically enriched as well. Again, ὁ δ᾽ ἆρα marks the recall of the main narrative frame, after the explanation of the prophecy. The “one-sandaled man” of the prophecy is Jason, as Pindar’s audience would have known. What makes this instance different from (t31) is that here there is apparent continuity of grammatical subject, so the use of the nominative pronoun must be explained otherwise.
§54. The reason for this overdetermination is that the pronoun ὁ has an ambiguous referent. Since Pythian 4 activates the narrative of the Argonauts right from the song’s beginning, we can say that Jason has been covertly present throughout the narrative. When Pindar turns to the specific episode of Pelias’ prophecy, the knowing audience will activate the figure of Jason, making ὁ in line 78 a referential nexus. ὁ refers to the one-sandaled man (τὸν μονοκρήπιδα, 75) who comes to Iolkos, but at the same time it refers to Jason. The pronoun conflates the stranger of the prophecy and the Jason of myth. The function of ἆρα is similarly layered: the particle accesses both the local expectations (“of course the prophecy comes true”) [137] and the global discourse memory that consists in the tradition shared by composer and audience (“as we both know, Jason did indeed come to dethrone Pelias”). [138] As in (t31) and the one other parallel (Olympian 10.43-45, see note 135), an appositional cap follows, ἀνὴρ ἔκπαγλος, to dispel any possible ambiguity: ἔκπαγλος strongly suggests that the referent is now supposed to be a hero rather than an unknown man with one sandal. [139]
§55. In the three instances in Pindar, the combination δ᾽ ἄρα (or δ᾽ ἆρα) after pronouns serves to recall the main narrative frame. I read ἄρα as having scope over the entire act rather than just over the pronoun (I imagine prosodic emphasis as falling on the most important part of the clause, perhaps the verb phrase: “And so he DID come...”). Thus, ἄρα modifies the act by marking its contents as shared between performer and audience. In narrative terms, the acts introduced by ὁ δ᾽ ἄρα are always framed: they continue the action of the narrative.
§56. The use of ὁ δ᾽ ἄρα in Homer is slightly more varied, but it falls within the same range as that in Pindar. The following example is representative for the majority of instances. During the battle at the wall of the Greek camp, Ajax gets angry and decides to take drastic measures:
          Αἴας δὲ πρῶτος Τελαμώνιος ἄνδρα κατέκτα |
          Σαρπήδοντος ἑταῖρον | Ἐπικλῆα μεγάθυμον |
380    μαρμάρῳ ὀκριόεντι βαλών, | ὅ ῥα τείχεος ἐντὸς |
          κεῖτο μέγας παρ’ ἔπαλξιν ὑπέρτατος· | οὐδέ κέ μιν ῥέα |
          χείρεσσ’ ἀμφοτέρῃς | ἔχοι ἀνὴρ | οὐδὲ μάλ’ ἡβῶν, |
          οἷοι νῦν βροτοί εἰσ’· | ὁ δ’ ἄρ’ [140] ὑψόθεν ἔμβαλ’ ἀείρας, |
          θλάσσε δὲ τετράφαλον κυνέην, | σὺν δ’ ὀστέ’ ἄραξε |
385    πάντ’ ἄμυδις κεφαλῆς· | ὁ δ’ ἄρ’ | ἀρνευτῆρι ἐοικὼς |
          κάππεσ’ ἀφ’ ὑψηλοῦ πύργου, | λίπε δ’ ὀστέα θυμός. |
Iliad 12.378-386
          Ajax Telamon’s son first killed a man,
          brave-hearted Epicles, a comrade of Sarpedon,
380    throwing a jagged rock. This of course lay inside the wall
          large and topmost near the battlements. And not easily
          with both hands could a man hold it, even in his prime,
          as mortals are now. So HE threw it, lifting it high,
          he crushed the four-ridged helmet, and crushed together
385    all at once the bones of his head. HE, then, an acrobat alike,
          fell down off the high wall, and life left his bones.
I discuss the use of ὅ ῥα to introduce unframed discourse, as in line 380, in §48-§58 below. For the present, I focus on the instances of ὁ δ᾽ ἄρα in lines 383 and 385. In 383 ὁ δ᾽ ἄρα retrieves Ajax, last named in 378, who picks up a stone no normal man could have lifted. The referential expression comes after the description of the rock, and ὁ δ’ ἄρα serves to retrieve the narrative frame. [141] After the image of Ajax throwing the rock at someone, his victim is made an agent by ὁ in 385, while what happens to him (he tumbles off the wall head first) is marked by ἄρα as the expected outcome. In addition, the next words (ἀρνευτῆρι ἐοικώς) suggest that the hearer is invited to imagine the scene by superimposing upon it the (shared) image of a diver. The vivid nature of the acts introduced by ὁ δ᾽ ἄρα is paralleled elsewhere in Homer, and it may be linked to Bakker’s reading of ἄρα as an “evidential” particle. [142]
§57. It cannot be established for every instance why we find ὁ δ᾽ ἄρα rather than ὁ δέ in Homer, but some pragmatic difference must underlie their linguistic divergence. The much higher frequency of ἄρα in Homer suggests that the range of functions of δ᾽ ἄρα is larger than in Pindar. Whereas in Pindaric song ὁ δ᾽ ἄρα consistently serves to recall the main narrative frame, in Homer the common element is that ἄρα expresses the performer’s assumption that the content of the act resonates strongly with the current state of the discourse memory. [143] What happens after ὁ δ᾽ ἄρα is expected, either based on common sense or social convention (obviously the messenger obeys Agamemnon), [144] or based on traditional knowledge (of course a plan by Odysseus works). [145] The combination in Homer always retrieves an accessible referent, but in many cases it accompanies not only a change of grammatical subject, but also a visual transition. [146]
§58. The combination δ᾽ ἄρα thus introduces new framed events that feature a referent who is thought to be retrievable but not necessarily in focus at the moment of reference. When one removes δέ from the combination, what is lost is the sense of a discourse boundary. That is to say, whereas pronoun + δ᾽ ἄρα moves the narrative ahead, no such thing need be expected of pronoun + ἄρα. In ὅ(ς) ῥα the particle still marks its host act as shared between performer and audience, but the act itself need not necessarily be framed, nor need it move the narrative forward.
§59. The combination ὅ ῥα is very rare in Homer and not attested in Pindar. Of the six instances, two do not in fact contain the demonstrative or relative pronoun, but rather the equivalent of ὅ τι, which is suggested by the preceding verb of perception or knowing. [147] Of the remaining four instances, two are followed by an imperfect and two by an aorist. [148] The meaning of these numbers becomes clearer when we consider that ὅς ῥα occurs in two constructions: (1) ὅς ῥα with an imperfect introduces unframed discourse, [149] and (2) ὅς ῥα with an aorist introduces framed discourse. [150] The following passage illustrates both constructions with ὅ(ς) ῥα:
          ἔνθά οἱ υἱὸς ἐπᾶλτο Πυλαιμένεος βασιλῆος |
          Ἁρπαλίων, | ὅ ῥα πατρὶ φίλῳ ἕπετο πτολεμίξων |
645    ἐς Τροίην, | οὐδ’ αὖτις ἀφίκετο πατρίδα γαῖαν· |
          ὅς ῥα τότ’ Ἀτρεΐδαο μέσον σάκος οὔτασε δουρὶ
Iliad 13.643-646
          There jumped on him the son of king Pylaemenes,
          Harpalion, who of course followed his beloved father, so that he could fight,
645    to Troy, and not again did he reach his fatherland.
          So HE then in the middle of Atreides’ shield thrust with his spear
ὅ(ς) ῥα introducing unframed discourse serves to recall little bits of information about the referent which the speaker expects to already be part of (or inferrable from) the discourse memory. [151] In line 643 the next assailant of Menelaus enters the scene, Harpalion son of Pylaemenes. After the name is stated in line 644 we find ὅ ῥα and a verb in the imperfect: the pronoun and particle introduce a piece of unframed discourse that consists of a flashback and a flash-forward. In this instance, one can just as well translate “Harpalion. He followed”; either way the act introduced by ὅ ῥα is unframed. Both the flashback and the flash-forward (οὐδ᾽...γαῖαν 645) are unframed discourse in the sense that the performer reveals his omniscience and informs the audience outside the frame. The use of ἄρα in unframed discourse marks a piece of knowledge from the discourse memory that is retrieved to become part of the current discourse model: it may be regarded as activating a piece of information in the long-term memory to become part of the working memory.
§60. After the unframed discourse in (t34) ὅς ῥα τότε follows at verse beginning, followed by an aorist. Functionally, there is a clear overlap with δ᾽ ἄρα: ὅς ῥα τότε here recalls the frame and introduces a new action, expressed by an aorist. It is possible that the ὁ δ᾽ἄρα and ὅς ῥα (τότε) functioned as metrical alternatives to effect the same transition in a different place in the verse. [152] In line with this use, ὅ(ς) ῥα followed by an aorist starts an embedded narrative in two instances in the Odyssey and once in a fragment of Pindar, where we might have expected to find γάρ. [153]
§61. In a small number instances ὅ(ς) ῥα is followed by an aorist when the act appears unframed:
          αὐτὰρ ὑπέρθυμον Πολυφείδεα | μάντιν Ἀπόλλων
          θῆκε | βροτῶν ὄχ’ ἄριστον, | ἐπεὶ θάνεν Ἀμφιάρηος· |
          ὅς ῥ’ Ὑπερησίηνδ’ ἀπενάσσατο | πατρὶ χολωθείς, |
255    ἔνθ’ ὅ γε ναιετάων | μαντεύετο πᾶσι βροτοῖσι. |
Odyssey 15.252-255
          And proud Polypheides, a seer Apollo
          made him, by far best of mortals, since Amphiaraos died.
          He had moved to Hyperesia, angry with his father.
255    Living there, he prophesied for all men.
My translation and punctuation reflect a reading of line 254 as unframed, despite the presence of the aorist. Two things support this reading: (1) ἀποναίω is not attested in the imperfect, and (2) there is a parallel passage in the Iliad. Compare this passage from the Catalogue of Ships, from the entry about the people from Doulichion:
τῶν αὖθ’ ἡγεμόνευε Μέγης ἀτάλαντος Ἄρηϊ |
Φυλεΐδης, | ὃν τίκτε Διῒ φίλος ἱππότα Φυλεύς, |
ὅς ποτε Δουλίχιονδ’ ἀπενάσσατο πατρὶ χολωθείς· |
Iliad 2.627-629
Them in turn led Meges, equal to Ares,
Phyleides, him bore the horseman Phyleus, beloved to Zeus,
who one day had moved to Doulichion, angry with his father.
It is clear that the resemblance between these two passages reaches beyond the repetition of the verb. [154] For the current purpose, it suffices to note that the information about Phyleus moving from Elis to Doulichion is necessary, yet clearly out of the current frame, as the use of ποτε confirms. I believe this justifies a similar reading of the passage from the Odyssey (t35): ὅς ῥα in line 254 introduces a piece of unframed discourse containing shared knowledge. [155] The passage preceding (t35) is a genealogy leading up to Polypheides and eventually his son, Theoclymenus. At this point in the genealogy, the family is in Argos, but it appears that the performer knows that Polypheides is connected to Hyperesia rather than to Argos. That would explain why he adds the line, with ὅς ῥα, to avoid a discrepancy between his performance and shared tradition.
§62. The combination ὅ(ς) ῥα thus serves two main purposes: (1) to introduce unframed discourse containing shared knowledge about a certain referent (accessing global discourse memory), and (2) to return to the frame after intervening unframed discourse of some sort (accessing local discourse memory). Moreover, there is a strong tendency for the performer to use the imperfect tense in unframed acts, and the aorist in framed acts.

5.3.4 ὁ δή and ὃς δή

§63. The combination ὁ δή/ὃς δή is not quite as flexible as ὅ(ς) ῥα. In II.3.3 I set out an analysis of the pragmatic functions of δή in Homer and Pindar. The particle is only used with a limited range of co-texts outside of direct speech: with temporal expressions, in certain combinations, and after pronouns. When used after pronouns, δή typically occurs in a relative clause. [156] The main functions of δή in early epic and lyric may be summarized as follows: (1) as a mobile (rare), it occurs in any position and has an intensifying force over the following word group, or else in act-initial position has scope over the entire act; (2) in peninitial position in the act, it generally has an intensifying force over the act it is in, or over the preceding word group, but sometimes still with scope only over the following word group; (3) in narrator text, in peninitial position with a temporal marker of some sort, it marks larger narrative moves.
§64. One might expect that δή in relative clauses would not mark larger narrative steps. This intuition is corroborated by the fact that of the twenty-seven instances of ὁ δή and ὃς δή in Homer only four are accompanied by a temporal adverb, and none of these occurrences appears to mark larger discourse steps. [157] Rather, the particle consistently intensifies either the entire act or the following word (group). [158] Before moving on to a discussion of the range of functions of δή in these clauses, let me establish some common characteristics of the clauses introduced by pronoun + δή. In contrast to δέ and δ᾽ ἄρα, δή after pronouns rarely serves to introduce a new event. Rather, it provides information about the internal or external state of the referent in focus: like the majority of ὅ(ς) ῥα instances, ὁ δή and ὃς δή introduce unframed discourse. Unlike ὅ(ς) ῥα, ὁ δή and ὃς δή introduce information about a referent that the speaker regards as inaccessible to the audience. [159]
§65. Since both anaphoric and relative pronouns most commonly occur in initial position, there are no instances of act-initial δή with scope over the act (function (1) in the list above) in this construction. [160] There is a link between the position of δή in the act and the scope of the particle: when δή occurs in peninitial position, it typically has scope at the minimum over the entire act, whereas in any other position its scope is more restricted, over only the following word. I first consider the instances where δή appears to have act scope. Since ὁ and ὅς δή are always at act beginning, however, the fact that δή occurs in second position does not exclude the possibility of mobile δή. After a discussion of the instances of δή with act scope, I explore the possibility of small-scope δή in some borderline instances. Finally, I show that sometimes the intensifying function of δή reaches beyond the reported interaction between two characters (in direct speech), and touches instead upon the interaction between performer and audience.
§66. δή occurs with intensifying force over the act in instances when the speaker urgently wishes to underscore an assertion. In English, this is often best rendered by adding an intensifier to the verb phrase (as in t37). [161] This construction only occurs in direct speech in Homer. Consider the following exhortation by Ajax to the other Greeks:
ἦ οὐκ ὀτρύνοντος ἀκούετε λαὸν ἅπαντα
Ἕκτορος, ὃς δὴ νῆας ἐνιπρῆσαι μενεαίνει;
Iliad 15.506-507
Do you not hear Hector encouraging all his people?
He is fairly raging to raze the ships!
The Trojans are at the wall, and Ajax does all he can to get the Greeks to fight back and keep them from the ships. His rhetorical question has a note of incredulity, and in the next utterance he vehemently adds some information that his audience apparently does not yet know. Although I do not see δή as directly connected to what is evident and visible, I do believe that some of the utterance’s tone of insistence comes from the implication that the speaker directly perceives what he is claiming. [162]
§67. In ten or eleven instances of ὃς δή / ὁ δή [163] the particle modifies first and foremost the following adverb or adjective, and is thus most likely unrelated to the preceding pronoun. This happens with forms of πολύς, for example in this passage from the Odyssey: [164]
δὴ τότε Φοῖνιξ ἦλθεν ἀνὴρ ἀπατήλια εἰδώς,
τρώκτης, ὃς δὴ πολλὰ κάκ’ ἀνθρώποισιν ἐώργει·
Odyssey 14.288-289
And at that moment a Phoenician arrived, a man skilled in wiles,
a greedy man, who had done very many evils to men.
This instance, like the majority of ὅς/ὁ δή instances, occurs in direct speech. Here Odysseus introduces a new character, a Phoenician (288), about whom he introduces several descriptions that his audience cannot yet know. One aspect of the “newness” inherent in relative clauses introduced by pronoun + δή is that it generally contains an expression of stance, a personal judgment or feeling of the speaker. [165]
§68. Similar is the use with superlatives (it is more frequent in historiography, see IV.4.6.2), which occurs twice in the Iliad, both in Aeneas’ boast to Achilles:
          Δάρδανος αὖ τέκεθ’ υἱὸν Ἐριχθόνιον βασιλῆα,
220     ὃς δὴ ἀφνειότατος γένετο θνητῶν ἀνθρώπων·
          Ἶλός τ’ Ἀσσάρακός τε καὶ ἀντίθεος Γανυμήδης,
          ὃς δὴ κάλλιστος γένετο θνητῶν ἀνθρώπων·
Iliad 20.219-220, 233-234
          Dardanus then begot a son, king Erichtonius,
220     who was by far the richest of mortal men.
          Ilus, Assaracus, and god-like Ganymedes,
          who was by far the fairest of mortal men.
Aeneas’ intensified superlatives serve to strengthen his boast about his forefathers. The second of the two examples shows that the link between relative clauses introduced by δή and “new information” is relative. There can be no doubt that both the internal audience (Achilles) and the audience at the performance was expected to know that Ganymedes was the most beautiful man of all. However, the presence of δή here, rather than ἄρα or τε, presents the statement not so much as a foregone conclusion, but as an expression of a personal opinion, with a clear rhetorical goal.
§69. Finally, we find small-scope forward-oriented δή in four (or five) final instances in temporal phrases, where again δή serves to intensify the sense of the following word. Here Eumaeus speaks to Telemachus of his trip to the palace:
ὡμήρησε δέ μοι παρ’ ἑταίρων ἄγγελος ὠκύς,
κῆρυξ, ὃς δὴ πρῶτος ἔπος σῇ μητρὶ ἔειπεν.
Odyssey 16.468-469
There joined me a quick messenger from among your friends,
a herald, who as the very first gave word to your mother.
Here, as in all the parallels, [166] the intensified expression occurs in direct speech. In all instances, the scope of δή is ambiguous, but here at least there seems to be a reference to a specific earlier scene (335-341). Eumaeus and the herald sent by Telemachus arrive at the palace at the same time, and the herald gives his news first (337). It is for this reason, I believe, that δή intensifies πρῶτος here: to render precisely what happened. [167]
§70. Finally, there is an instance of ὅς δή that appears to strongly appeal to the visual imagination of the audience. In narrator text, the performer introduces Dolon, a character central to the upcoming narrative, and gives a vivid description:
          ἦν δέ τις ἐν Τρώεσσι | Δόλων Εὐμήδεος υἱὸς |
315    κήρυκος θείοιο πολύχρυσος πολύχαλκος, |
          ὃς δή τοι | εἶδος μὲν ἔην κακός, | ἀλλὰ ποδώκης· |
Iliad 10.314-316
          There was someone among the Trojans, Dolon son of Eumedes,
315    the godlike herald, rich in gold and rich in bronze.
          And this guylet me tell you – he was ugly of face, but quick of feet.
This is one of only two instances of δή τοι in narrator text, and one of the relatively few instances in Homer where τοι is clearly the particle rather than the dative second person pronoun. Just as in (t39), this piece of discourse introduces a character. [168] The difference is that the introduction here first mentions the father of the character in focus. The performer brings the attention back to Dolon with the priming act ὅς δή τοι. [169] The priming act also projects Dolon’s importance in the upcoming long episode (316-457). [170] The presence of the particle τοι especially suggests that there may be one more factor at work in this passage. It is as if the performer turns to the audience and speaks to them directly, inviting them to imagine this Dolon, and to share the performer’s opinion of him. Since it occurs in the priming act, I believe that δή may be regarded as having scope over the entire line, the entire description of this antagonist.
§71. Whether δή following an anaphoric pronoun has small scope or act scope, it generally introduces a subjective view of a character, either by another character or by the narrator. The difference between unframed discourse introduced by ὅς ῥα on the one hand and by ὁ δή and ὃς δή on the other can be put in the following terms: unframed discourse introduced by ὁ δή or ὃς δή contains information that the performer assumes to be as yet unavailable to the audience. As a result, the function of pronoun plus δή overlaps partly with that of γάρ introducing unframed discourse. In general, however, unframed discourse introduced by γάρ is more directly associated with the preceding discourse than that introduced by δή. The following passage from Pindar will serve as illustration:
ἀπὸ Ταϋγέτου πεδαυγάζων| ἴδεν Λυγκεὺς | δρυὸς ἐν στελέχει |
ἡμένους. | κείνου γὰρ ἐπιχθονίων πάντων γένετ’ ὀξύτατον
ὄμμα. | (...)
Pindar Nemean 10.61-63
Peering down from the Taugetus, Lynkeus saw them in the trunk of an oak,
hidden. For of all the men on earth his was the sharpest
sight. (...)
The use of γάρ after pronouns in both Homer and Pindar often introduces unframed discourse, as here. [171] Here γάρ introduces a fact about Lynkeus’ sight, which is triggered by association with the preceding narrative. The difference between unframed discourse introduced by δή and γάρ is that δή typically introduces a personal judgment. This does not mean that γάρ cannot introduce a personal judgment, as δὴ γάρ in Homer illustrates (see II.4 §19). Unframed discourse introduced by ἄρα, conversely, is expressly presented as shared between performer and audience or between speaker and internal audience. The difference between ὅς ῥα on the one hand and ὁ/ὃς δή and ὁ/ὃς γάρ on the other may be represented quite well with the paraphrase “who of course” for ὅς ῥα and “who actually” for ὁ/ὃς δή and ὁ/ὃς γάρ.

5.4 Participant tracking in a Pindaric ode: Isthmian 2

§72. Whereas until now I have limited my analysis to ὅς and ὁ, the masculine singular demonstrative or relative pronoun in the nominative, for my study of Isthmian 2 I take an more inclusive approach. Pronouns, nouns, and verbs all factor in the performer’s efforts to guide the audience’s attention. Within this overall process of participant tracking – that is, the cognitive process of monitoring which character is in focus and which others should be attended to – the specific role of particles bears close study. In mainstream commentaries, most entries concern nouns, adjectives, adverbs, and verbs. Moreover, the entries show a distinct focus on semantics over discourse progression or participant tracking. [172] This illustrates the point I make at the beginning of this chapter: participant tracking does not generally cause problems for comprehension. Nonetheless, the complex process of referring to people both inside and outside the direct performance context deserves close attention. In what follows I present a line-by-line commentary to Pindar’s Isthmian 2 with a focus on referent tracking.
§73. The second Isthmian is a special victory ode. Unlike most other odes written for an athletic victory, Isthmian 2 is addressed for the most part to the victor Xenocrates’ son Thrasyboulus. The reason for this appears to be the fact that by the time the song was actually composed, Xenocrates had already died. The exact context of the song has been debated, especially the questions of whether Pindar makes reference to the occasion of Pythian 6, an earlier victory ode for Xenocrates (probably at least twenty years before), and whether at that time Pindar and Thrasyboulus may have been romantically involved. [173] Finally, opinions differ on the question of date, with proposals ranging from 474 BCE to 471 BCE. If the song was composed and performed after 472, this means that Xenocrates’ brother Theron would have also passed away, and the reign of the Emmenids at Akragas would have ended. [174]
§74. Let me anticipate three overarching topics before moving on to the running commentary below: referents of first and second person forms, development of referents within the ode, and the juxtaposition of individual and collective. From the start of the performance, the first person refers most naturally to the performer—be it a singer or a chorus. [175] Behind that performer lies always the voice of Pindar, so first-person references will often have been ambiguous, possibly on purpose. In this song, Thrasyboulus is established as the main second-person referent immediately with a vocative in line 1. He is picked up again in the first epode, in a verb form, and the part of the discourse addressed directly to Thrasyboulus is rounded off in the second epode, with another vocative. Next, Pindar starts the third triad with a gnṓmē, which introduces an extensive eulogy on Xenocrates, addressed perhaps to the wider audience. If Thrasyboulus is here still to be understood as the second person, it no longer emerges in the language. The final epode reveals this implicit change, with a third-person imperative that has to refer to Thrasyboulus. This opens up the second-person slot for another referent, Nicasippus, whose identity remains unclear. It is typically reserved for the laudandus to be juxtaposed with Pindar in the final lines of the song, so whoever Nicasippus was, the very fact that he is referred to in the second person in the last epode strongly suggests that he was a person of some importance to Pindar, the laudandus, and/or the audience.
§75. With Thrasyboulus established as second person, Xenocrates – the actual victor – is automatically relegated to the third person. Since all referents besides the performer, Thrasyboulus, and Nicasippus occur in the third person, Pindar has to disambiguate every time he talks of Xenocrates. To that end the victor is introduced by name at the end of the first epode, which runs over thematically into the second strophe. Pindar describes his athletic victories, and Xenocrates remains in focus until the shift to his charioteer just before the end of the strophe. After rounding off the second triad, Pindar has to activate Xenocrates again by naming him, before launching into a praise of his virtuous life. The eulogy reaches a climax in the third antistrophe, where Xenocrates is really the only referent. Then at the beginning of the final epode the marker νυν brings the audience back to the here-and-now, and makes Xenocrates an unlikely referent for the third-person imperative. He moves to the background, referred to only as “the father of”, and Thrasyboulus takes his place as the main third-person referent.
§76. Throughout the song we find juxtapositions of individuals and groups, both mortal and immortal, but in the end the central figures are individuals. Thus in the first strophe and antistrophe the Muses are first activated as a group, but later represented by the Muse Terpsichora. Even more clearly, in the second strophe two groups are adduced, but only to better demonstrate the κλέος of Xenocrates: he was a light for the Akragantines, and he pleased the Erechtheids. [176] After the praise of Nicomachus the charioteer, there is an interesting exception: Pindar links Ainesidamus’ descendants, that is Xenocrates and his brothers, as a group to an Olympic victory. It seems likely that he is being purposefully vague since not Xenocrates or Thrasyboulus had been involved in this victory, but another of their clan. Finally, we find the townsmen and the gods mentioned as groups who were treated well by Xenocrates; again the collective serves as a foil for the praise of the individual.
§77. To sum up, the special nature of Isthmian 2 is reflected in the linguistic forms of anaphoric reference. Whereas the laudandus is typically in the second person, in this case it is an absent father in the third person who receives the most resounding words of praise. The son Thrasyboulus is the addressee in the first and second triads, but has to make way for the mysterious Nicasippus in the final epode. The first-person referent is embedded in the the here-and-now of the performance, and his linguistic presence is linked in particular to places where the song references the event. Thus besides in direct addresses to the second person we find the “I” in transitions and metanarrative statements.
§78. The text below visually highlights the elements of the discourse that guide the audience’s referent tracking. Linguistic markers of frame switches are instrumental in this process, since a change of frame changes which referents are accessible. For this reason, I have marked not only act boundaries (|) but also frame shifts, marked with a double vertical bar (||). [177] All particles are underlined (also when combined with a negative). In green I have marked all relative and demonstrative pronouns, including articles, as well as all pronouns, and nouns (including names) referring to characters outside or inside the discourse. The verb forms (finite verbs and participles) are also in green, when they are instrumental to participant tracking. These form the majority of the items discussed in the entries, but here and there I go slightly off topic in order to address an issue in the text or in current scholarship. After each metrical unit of strophe, antistrophe, or epode follows the translation, the list of referents, and then the commentary.
          οἱ μὲν πάλαι, | Θρασύβουλε, | φῶτες, | οἳ χρυσαμπύκων
          ἐς δίφρον Μοισᾶν ἔβαινον | κλυτᾷ φόρμιγγι συναντόμενοι, |
          ῥίμφα παιδείους ἐτόξευον μελιγάρυας ὕμνους, |
          ὅστις | ἐὼν καλὸς | εἶχεν Ἀφροδίτας
5        εὐθρόνου μνάστειραν ἁδίσταν ὀπώραν. ||
First strophe, 1-5
          The ancients, Thrasyboulus, men who mounted
          the chariot of the Muses with gold headbands, using the renowned lyre,
          lightly shot honey-sounding hymns for boys.
          Whoever, being beautiful, had
5        the sweetest bloom reminiscent of fair-throned Aphrodite.
Participants: the ancients, Thrasyboulus, the Muses, ὅστις, Aphrodite | 1-2 οἱ...ἔβαινον: The first line has οἱ and οἵ, the first demonstrative and the second relative. The placement of the vocative (Thrasyboulus, see §§74-75) between οἱ and φῶτες argues against reading οἱ as a definite article with φῶτες. This is supported by the fact that οἱ πάλαι on its own can mean “the ancients.” [178] φῶτες οἵ...ἔβαινον should then be read in apposition to οἱ πάλαι. | 2-3 συναντόμενοι...ἐτόξευον: Subject continuity, frame continuity: null anaphor. | 4-5 ὅστις...ὀπώραν: New grammatical subject introduced at line beginning with the relative ὅστις. Change of subject is accommodated by a possible pause before and/or after the independent participial phrase ἐὼν καλός. The relation between lines 3 and 4 is close at the semantic level, but syntactically loose. The asyndetic beginning of line 4 suggests that 4-5 stand somewhat apart, [179] which complements the lack of a clear syntactic link. The semantic coherence of 1-3 and 4-5, however, is promoted by strophic contiguity, and there is no need to create a more integrated syntactical construction in the translation. [180]
          ἁ Μοῖσα γὰρ | οὐ φιλοκερδής πω τότ’ ἦν | οὐδ’ ἐργάτις· |
          οὐδ’ ἐπέρναντο γλυκεῖαι | μελιφθόγγου ποτὶ Τερψιχόρας |
          ἀργυρωθεῖσαι πρόσωπα μαλθακόφωνοι ἀοιδαί. |
          νῦν δ’ ἐφίητι <τὸ> τὠργείου φυλάξαι |
10      ῥῆμ’ ἀλαθείας < ⏑ – > ἄγχιστα βαῖνον, |
First antistrophe, 6-10
          For the Muse, she was then not yet profit-loving, nor for hire.
          Not yet were they sold by honey-voiced Terpsichora,
          sweet songs with silvered faces and lovely voices.
          And now she [sc. the Muse] commands us to heed that of the Argive [sc. Aristodemus]
10      an adage that comes closest to <…> reality.  [181]
Participants: the Muse, Terpsichora, the Argive (Aristodemus) | 6 ἁ Μοῖσα γάρ: New grammatical subject introduced through article and noun, separated from the following in a priming act. γάρ accompanies the switch of grammatical subject, and serves to expand on the Muses mentioned before (line 2). The topic of the “mercenary Muse” persists throughout the first antistrophe. [182] Although lines 6-8 appear to continue the frame of πάλαι based on the continuation of past tense, the string of negations foregrounds the current situation over the old. The discontinuity of orientation, of reference (the men of old are no longer in the frame), and the occurrence of γάρ all suggest that a frame switch co-occurs with the performative discontinuity of strophe end after line 5. | 7-8 γλυκεῖαι...ἀοιδαί: The grammatical subject of this passage is ἀοιδαί, but the logical subject is Terpsichora. | 9 νῦν δ᾽ ἐφίητι: Combined with the present tense, νῦν δέ marks a clear transition to the present. Commentators wish to see νῦν δέ as anwering the οἱ μέν πάλαι in line 1, but for the reasons given ad 6 I believe that the switch in orientation occurs between lines 5 and 6. On a more local level, νῦν interacts with τότε (6): the antistrophe concerns the Muse (announced in a priming act in 6), both in the past (τότε) and in the present (νῦν). Therefore I take the Muse as the subject of ἐφίητι. The anaphoric and cognitive continuity across line 8 and 9 suggests that there is no frame switch here. | 9 <τὸ> τὠργείου... βαῖνον: τό (conjectured) may be a demonstrative pronoun or an article. As in line 1, I read it as a demonstrative, with the following noun phrase ῥῆμα ... βαῖνον, as an appositive. [183]
          “χρήματα χρήματ’ ἀνήρ” | ὃς φᾶ κτεάνων θ’ ἅμα λειφθεὶς καὶ φίλων. ||
          ἐσσὶ γὰρ ὦν σοφός· | οὐκ ἄγνωτ’ ἀείδω|
          Ἰσθμίαν ἵπποισι νίκαν, ||
          τὰν Ξενοκράτει Ποσειδάων ὀπάσαις, |
15      Δωρίων αὐτῷ στεφάνωμα κόμᾳ
          πέμπεν ἀναδεῖσθαι σελίνων, |
First epode, 11-16
          ‘Means, means make the man’ says he who has been left by both possessions and friends.
          You see, you are well-versed in song. Nothing unknown to you do I sing,
          the Isthmian chariot victory.
          Having granted that to Xenokrates, Poseidon
15      sent a wreath of Dorian wild celery to him
          to put over his hair,
Participants: a man, friends, Thrasyboulus, the “I” persona, Xenocrates, Poseidon | 11 ὃς φᾶ...φίλων: ὅς can be a relative or demonstrative pronoun. Although the relative pronoun can be postponed in Pindar (see II.2 §42n113), here I do not follow Verdenius in reading it as postponed until after the quasi-direct speech here. [184] The performative discontinuity between antistrophe and epode is too strong for such a complex relative construction to be felicitous. ὅς is better read as demonstrative. Moreover, I do not think we should take ὅς as referring to Aristodemus here (the Argive, τὠργείου), but as indefinite. Of course there is room for ambiguity, so depending on the audience ὅς may be interpreted as either general or specific. | 12 ἐσσὶ...σοφός: The transition from generic discourse to the hic et nunc is immediately accomplished with ἐσσί. The interaction between performer and the specific addressee Thrasyboulus initiated with the vocative in line 1 is picked up again here. The role of γάρ and ὦν in the transition has vexed commentators. I do not have the answer, but there are several things we can establish with a reasonable degree of certainty: (1) γὰρ ὦν in Pindar is not like the cluster γὰρ ὦν in Herodotus, where it tends to follow resumptive pronouns; [185] (2) γάρ introduces some salient information to the discourse model—even if the distinction framed/unframed may not be the most helpful in this non-narrative context; (3) ὦν in Pindar only occurs in combinations with other particles, which complicates the understanding of its function. [186] I have read γὰρ ὦν as establishing common ground: any relation with the preceding must be inferred in English, as it is in the Greek. | 12 οὐκ...ἀείδω: ἐσσί, a few words before, establishes the hic et nunc as the relevant frame, and by expressing the “you” it makes the “I” readily accessible: an act-final first-person verb with null anaphor thus creates no difficulty. If ἄγνωτ᾽ is indeed to be read as ἄγνωτα, then it must be indefinite. Thus, it is an expansion of the thought expressed in the first part of line 12. | 14-17 τὰν...φάος: τάν is anaphoric, directly retrieving the νίκα of line 13. The change in grammatical subject and the ring construction in lines 14 to 17 (participial clause – main clause – participial clause) suggest that 14 should be taken with the following rather than the preceding verse; thus I have marked a frame switch just before 14. | 15 αὐτῷ: The anaphoric pronoun looks superfluous, since Xenocrates has already been established as the indirect object of the construction. Verdenius proposes that it is added “for the sake of clarity: it prevents us from connecting κόμᾳ with πέμπεν,” but this explanation is insufficient. Rather, use of the pronoun may be connected to the fact that αὐτός here refers to the recently deceased Xenocrates. Whereas the participle concerning the victory is in the aorist (ὀπάσαις), πέμπεν is in the imperfect, suggesting a change in perspective. This change may be from the victory at the Isthmian games to the more current event of his death. The σέλινον was associated with death and sickness: “persons dangerously ill were said δεῖσθαι τοῦ σελίνου.” [187] Pindar appears to combine this expression with that of crowning victors with σέλινον, to create a complex reference to the victory and to Xenocrates’ death. It is in this context that we may reconsider αὐτός, which has a special relationship with the body of a hero as a deictic center. [188]
          εὐάρματον ἄνδρα γεραίρων, Ἀκραγαντίνων φάος. ||
          ἐν Κρίσᾳ δ’ εὐρυσθενὴς εἶδ’ Ἀπόλλων νιν | πόρε τ’ ἀγλαΐαν ||
          καὶ τόθι | κλειναῖς Ἐρεχθειδᾶν χαρίτεσσιν ἀραρώς |
20      ταῖς λιπαραῖς ἐν Ἀθάναις, | οὐκ ἐμέμφθη
          ῥυσίδιφρον χεῖρα πλαξίπποιο φωτός, |
Second strophe, 17-21
          honoring the man who knows the chariot well, a light for the Akragantines.
          In Krisa mighty Apollo saw him, and gave him splendor.
          And there, pleasing to the renowned favors of the Erechtheids,
20      in shining Athens, [189] he did not complain about
          the chariot-saving hand of the horse-striking man,
Participants: Xenocrates, the Akragantines, Apollo, the Erechtheids, the charioteer | 18 ἐν Κρίσᾳ...νιν: There is a change of grammatical subject, introduced by a name (Ἀπόλλων); the tense returns to aorist (εῖδε, as ὀπάσαις), and the object (Xenocrates) remains the same and can be picked up with the enclitic νιν. | 18 πόρε τε: Subject continuity, frame continuity: null anaphor. τε adds an entire clause, generally called “sentential τε” or “consecutive τε.” | 19 καὶ τόθι: There is discussion about whether this should be connected to the preceding (in which case τε is inserted after κλειναῖς): “And he [sc. Apollo] granted him glory, there too. And to the renowned...” or to the following: “And he [sc. Apollo] granted him glory. And there pleasing to the renowned favors of the Erechtheids.” [190] Prior to answering the question of how καὶ τόθι should be constructed, we need to consider the referential transition. In the middle of a symmetrical construction about gods granting Xenocrates victories, Xenocrates is here made an agent and grammatical subject himself. This kind of transition is unlikely after τόθι, since τε with a participial clause and no other marker would strongly suggest subject continuity (there is subject continuity in lines 18, 25, and 38; in 23 the relative in the oblique projects subject change: ὅν τε). Therefore I propose keeping the reading of the manuscripts (without τε), as do Fennell, Dornseiff, and Thummer, and I take καὶ τόθι as a transitional priming act, [191] projecting “there” forward. Thus there are three contextual frames: one in the Isthmus (introduced by [Ἰσθμίαν νίκαν] τάν 14), one in Krisa (near Delphi, introduced by ἐν Κρίσᾳ δέ 17), and one in Athens (introduced by καὶ τόθι). That is to say, καί works de dicto, not de re. The reason for the variation in form may be that καὶ τόθι introduces a longer section of discourse, from line 19 to line 27. The length in turn may be because the Athenian victory was the most recent. [192] | 20 οὐκ ἐμέμφθη: The change in contextual frame effected by καὶ τόθι means by definition that there is a new pool of potential referents (there is no covert continuity across frame boundaries). In the preceding discourse, across the different frames, the only common element is Xenocrates. Therefore, he is the most accessible, and the only referent that can be retrieved with null anaphor (∅ ἐμέμφθη) at this point. | 21 φωτός: A new referent is introduced into the frame with a full noun and adjective. The full line devoted to introducing him – without naming him – at the end of the strophe projects his relevance into the antistrophe. φωτός thus works cataphorically: a name is expected and anticipated.
          τὰν Νικόμαχος κατὰ καιρὸν νεῖμ’ ἁπάσαις ἁνίαις· |
          ὅν τε καὶ κάρυκες ὡρᾶν ἀνέγνον, | σπονδοφόροι Κρονίδα
          Ζηνὸς Ἀλεῖοι, | παθόντες πού τι φιλόξενον ἔργον· |
25      ἁδυπνόῳ τέ νιν ἀσπάζοντο φωνᾷ |
          χρυσέας ἐν γούνασιν πίτνοντα Νίκας |
Second antistrophe, 22-26
          that Nicomachus plied rightly to all the reins.
          And him the heralds of the season also recognized, the truce-bearers of Cronus’ son
          Zeus from Elis; they had probably experienced some deed of hospitality.
25      And they welcomed him with sweetly breathing voice,
          as he fell in the lap of golden Victory,
Participants: Nicomachus, the heralds from Elis, Zeus, Victory | 22 τάν Νικόμαχος: The relative only has one locally available referent: χεῖρ, and the name is automatically connected to the charioteer introduced in 21. His name at the very beginning of the antistrophe suggests that he will be its main topic (cf. ἁ Μοῖσα in line 6). | 23 ὅν τε: The pronoun retrieves the most accessible masculine referent: Nicomachus. Although τε occurs right after a relative pronoun, it appears to be at least partly copulative, in a bisyndetic construction with the τε in line 25. [193] | 23 καὶ κάρυκες: καί should not be connected with τε, but has small scope over κάρυκες. [194] This can mean either: “the heralds, too” (as Privitera 1982:37), or καί can have a pinning-down function, [195] perhaps to be rendered through prosodic emphasis: “the HERalds.” Since the introduction of the κάρυκες ὡρᾶν marks the transition to the Olympic games, “also” is not unfelicitous. | 24 παθόντες πού: The plural participle picks up κάρυκες, and I agree with Bury 1892:45 that the aorist participle places παθόντες before ἀνέγνον in time. [196] που might serve to interact with the audience at the performance, if we follow Koier’s argument that που with verbs of knowing marks something as shared between speaker and hearer. [197] Thus it serves to reinforce a sharedness in the praise of Nicomachus’ hospitality. | 25 νιν: The enclitic pronoun marks continuity of object. | 26 πίτνοντα: Co-referential with νιν. Despite the semantic difficulties in this line (see Verdenius 1982:22), it is clear that the referent concerned is Nicomachus. | 26 Νίκα: This ode is only marginally concerned with an athletic victory, but Pindar still introduces personified Victory here. It is remarkable that the word Νίκα is so far separated from the names of Xenocrates and Thrasyboulus, and instead directly associated with the charioteer Nicomachus.
          γαῖαν ἀνὰ σφετέραν, || τὰν δὴ καλέοισιν Ὀλυμπίου Διός
          ἄλσος· | ἵν’ ἀθανάτοις Αἰνησιδάμου
          παῖδες ἐν τιμαῖς ἔμιχθεν. ||
30       καὶ γὰρ οὐκ ἀγνῶτες ὑμῖν ἐντὶ δόμοι |
          οὔτε κώμων, | ὦ Θρασύβουλ’, | ἐρατῶν, |
          οὔτε μελικόμπων ἀοιδᾶν. ||
Second epode, 27-32
          in their land, which men actually call a sanctuary of Olympian
          Zeus. There the descendants of Ainesidamus
          joined in immortal honors.
30      And thus, your house is no stranger
          to either beloved revels, Thrasyboulus,
          or to sweet-sounding songs.
Participants: the heralds, , Zeus, Ainesidamus’ descendants, Thrasyboulus | 27 γαὶαν... σφετέραν: Nicomachus won “in their land,” σφετέραν still refers to the κάρυκες. | 27 τὰν δή: The pronoun followed by δή consistently marks a new beginning of some sort in Pindar. [198] Here it introduces unframed discourse, which allows for a new, generic referent “men.” [199] The act introduced by τὰν δή resolves any possible ambiguity in the preceding passage, where Olympia has only been referred to obliquely. | 28-29 Αἰνησιδάμου παῖδες: The new grammatical subject is introduced by a full noun phrase. The re-introduction of the descendants of Ainesidamus (his sons Xenocrates and Theron, but including grandson Thrasyboulus too, I believe; see note ad ὑμῖν 30) also accommodates the upcoming return to the hic et nunc. Only Theron won at the Olympic games, so I would take it as a generic statement that the family gained honor at the games. [200] | 30 καὶ γάρ...δόμοι: γάρ introduces the return to the present, and is best left untranslated. I disagree with Verdenius and Thummer that καὶ γάρ should be taken as affirmative and that the transition should be read as an asyndeton. [201] γάρ serves precisely to mark the associative link between the preceding move about the past and the current move about the present, and as such has scope over the entire move; καί, conversely, has scope over the act only. | 30 ὑμῖν: The combination of ὑμῖν with the vocative Θρασύβουλε in the next line suggests that Pindar here takes all descendants of Ainesidamus together: Thrasyboulus, his father Xenocrates (deceased), and his uncle Theron (perhaps still alive). δόμοι “house,” the grammatical subject in this move, reflects this inclusiveness.
          οὐ γὰρ πάγος | οὐδὲ προσάντης ἁ κέλευθος γίνεται, |
          εἴ τις εὐδόξων ἐς ἀνδρῶν ἄγοι τιμὰς Ἑλικωνιάδων. ||
35      μακρὰ δισκήσαις ἀκοντίσσαιμι τοσοῦθ’, | ὅσον ὀργάν
          Ξεινοκράτης ὑπὲρ ἀνθρώπων γλυκεῖαν
          ἔσχεν. || αἰδοῖος μὲν ἦν ἀστοῖς ὁμιλεῖν, |
Third strophe, 33-37
          For there is no obstacle, nor is the road steep,
          if one brings honors of the Heliconians to of famous men
35      Having thrown the discus far may I cast the javelin just as far, as
          Xenocrates has a temperament sweeter than that of all other
          men. He was respectful in dealing with his townsmen,
Participants: one (= the “I” persona), famous men, the Heliconian Muses, Xenocrates, men, townsmen | 33 οὐ γὰρ πάγος: γάρ is directly followed by a new grammatical subject, which cannot naturally be connected to the preceding move. In combination with the strong performative discontinuity between epode and strophe, this would probably have created the expectation of a frame switch. Only in the course of the following line does it become clear that γάρ introduces a gnomic reflection on the preceding. | 34 τις: The indefinite pronoun is only nominally indefinite: it refers generically to poets who sing the praises of good men, but at the same time specifically to Pindar himself singing the praises of Thrasyboulus’ clan. [202] | 34 τιμὰς Ἑλικωνιάδων: The “honors of the Heliconian muses” here stand for Pindar’s song. | 35 μακρά... ἀκοντίσσαιμι: The tightly knit combination of participle and finite verb directly projects an athletic image onto the poetic activity. The first-person verb places the act firmly in the hic et nunc, however metapoetic the comment may be. | 35-37 ὅσον...ἔσχεν: The athletic image leads into further praise of Xenocrates, who is introduced by a full name, and in an integrated construction. | 37 αἰδοῖος μέν ἦν: After the introduction in lines 35-36, a frame switch occurs in 37 with μέν and a switch from aorist to imperfect. μέν is used here, as often in Pindar, to project a topic (and a grammatical subject) across the performative boundary between strophe and antistrophe. [203]
          ἱπποτροφίας τε νομίζων ἐν Πανελλάνων νόμῳ· |
          καὶ θεῶν δαῖτας προσέπτυκτο πάσας· | οὐδέ ποτε ξενίαν
40      οὖρος ἐμπνεύσαις ὑπέστειλ’ ἱστίον ἀμφὶ τράπεζαν· |
          ἀλλ’ ἐπέρα | ποτὶ μὲν Φᾶσιν θερείαις, |
          ἐν δὲ χειμῶνι πλέων Νείλου πρὸς ἀκτάν. ||
Third antistrophe, 38-42
          practicing horsebreeding according to the Panhellenic ways,
          and he welcomed all the feasts of the gods. And never did an
40      adverse wind furl his sail at his hospitable table;
          no, he travelled on, toward Phasis in summer,
          and in winter sailing to the shore of the Nile.
Participants: Xenocrates, the gods | 38 τε νομίζων: Without explicit marking to the contrary, τε suggests continuity of grammatical subject. | 39 καί: As in 19 (∅, δέ, καί), καί introduces a third consecutive item: μέν, τε, καί. [204] | 39 οὐδέ ποτε: The linguistically more extensive transition introduces the fourth topic of hospitality, which will remain in focus until line 42. | 41-42 ἀλλά, μέν, δέ: The discourse progresses in clear steps. Because of constant subject continuity there are no (pro)nouns referring to the grammatical subject.
          μή νυν, || ὅτι φθονεραὶ θνατῶν φρένας ἀμφικρέμανται ἐλπίδες, ||
          μήτ’ ἀρετάν ποτε σιγάτω πατρῴαν, |
45       μηδὲ τούσδ’ ὕμνους· || ἐπεί τοι
          οὐκ ἐλινύσοντας αὐτοὺς ἐργασάμαν. ||
          ταῦτα, | Νικάσιππ’, | ἀπόνειμον, | ὅταν
          ξεῖνον ἐμὸν ἠθαῖον ἔλθῃς.
Third epode, 43-48
          Now may he not, since envious hopes surround the minds of men,
          may he never be silent about his paternal virtue,
45      nor about these hymns. Since, you know,
          I did not craft these to remain idle.
          Give him this, Nicasippus, as his due, when
          you come to my trusted guest-friend.
Participants: men, Thrasyboulus, Xenocrates, Nicasippus, the “I” persona | 43 μή νυν: Right after the discontinuity between antistrophe and epode we find νυν: even in its enclitic form it will have activated the hic et nunc after μή at the beginning of act, verse, and epode. | 43 ὅτι...ἐλπίδες: The unframed nature of this insertion gives it the sense of a parenthetical right after the beginning of the epode. The gnomic thought serves as a backdrop for the wish started by μή νυν. | 44 μήτε: The negative resumes the interrupted construction in 43, and τε is probably to be understood as anticipatory in some sense to the negatives at the start of the following two lines. | 44 σιγάτω: One of the hardest forms in the song as regards referent tracking. Xenocrates has been the grammatical subject throughout the last passage, but that is no longer the case after the strong discontinuity marked by μή νυν. As soon as the frame of the here and now is activated, the most accessible referents are the performer (first person) and the audience or Thrasyboulus (second person). There is no logical referent for the third person imperative σιγάτω, however, since the wish is unlikely to refer to the dead Xenocrates. On the basis of sense, all commentators read σιγάτω as referring to Thrasyboulus, albeit in the third person. Bury rightly notes that πατρῴαν immediately helps to disambiguate. [205] Verdenius suggests that the performer addresses Nicasippus from μή νυν onward (see the vocative in line 47). [206] This would certainly explain the shift to third person, and it may be supported by one more argument. Verdenius believes that ταῦτα refers to the present song, [207] but I believe Privitera rightly reads it as referring specifically to the two pieces of advice introduced by ταῦτα: “riferisci a Trasibulo questa mia raccomandazione di celebrare suo padre.” [208] If ταῦτα indeed refers to these specific elements, then it is likely that the performer already turned (mentally or physically) to Nicasippus with μή νυν in line 43, and σιγάτω refers to Thrasyboulus. | 45 μηδέ: Perhaps the second limb of μήτε...μηδέ is indeed more poignant than the first. [209] This asymmetry appears to be corroborated by the following line, which expands on the hymns. | 45 ἐπεί τοι: As in line 43 (ὅτι), a causal conjunction introduces a motivation, in this case for the preceding wish. τοι must be the particle, since if it were the second person, it would only make sense if it referred to Thrasyboulus, which it cannot do here. | 46 αὐτούς: Since there is clear object continuity, Bury is probably right in regarding αὐτούς as implying a contrast with “other hymns.” [210] | 47 Νικάσιππε: Whoever this man was, [211] I believe his presence at the performance must be presupposed. If he was indeed a khorēgós, the last epode is apparently some kind of personal message. However, in performance the addressee of the chorus might then have been unclear to the audience. In any case, he must have been present at the occasion, and Nicasippus must have been an accessible referent for (the majority of) the audience. For σιγάτω to be easily understandable, a physical shift of gaze by the performer would have been helpful. I find it doubtful if Nicasippus was a mere professional of any kind: the position so close to the end of the song is extremely marked. | 48 ξεῖνον ἐμόν: The guest-friend must be Thrasyboulus, but the first-person reference is inherently ambiguous, referring at once to the composer Pindar and the current performer. The ambiguity may not have been felt by the audience at all, but a full understanding of the final expression depends again on the question of who Nicasippus was.
§79. Many of the passages discussed in the running commentary above may appear at first sight not to be problematic at all. However, it is my point to reveal how complex the process of referent tracking is, which normally works automatically in our minds as we read or listen. The focus in this section has not been to solve problems, but to attempt to explain why we have relatively few problems in following Pindar’s discourse. Along the way, I have discussed more general problems with the second Isthmian from the perspective of anaphoric reference. The analysis only proposes an alternative road to solutions, arriving sometimes at conclusions similar to, sometimes different from, the commentaries. Finally, considering an entire discourse illustrates best how referent tracking and particle use mutually influence each other. Understanding better what particles do helps us to gain a fuller understanding of the process of reference tracking, and vice versa.

5.5 Conclusions

§80. In the discussion of particles that come after ὅς or ὁ in Homer, we must consider at least the following factors in order to perceive the relevant patterns. First, what is the exact referent of the anaphoric pronoun? That is, not its textual antecedent, but the mental representation of the referent at the moment that the pronoun is uttered. Second, does the anaphoric pronoun continue framed discourse, mark a transition from framed to unframed discourse, or a transition from unframed to framed? Third, who is the referent for the speaker or audience: is the referent present in the physical discourse context? Is there a particular emotional connection between speaker and referent? Does the referent have a particular relevance to the larger narrative or tradition? Do speaker and audience have the same referent in mind?
§81. As regards the difference between ὁ and ὅς, it is clear that they are used interchangeably in Homer. For both third-person pronouns the following pattern occurs: when ὁ/ὅς introduces framed discourse, all masculine singular referents within the frame are in principle accessible. When ὁ/ὅς introduces unframed discourse, conversely, the performer picks out one character who must be highly accessible: that is, the referent must be overt in the directly preceding discourse, typically as subject or as object. Therefore ὁ/ὅς introducing framed discourse more often marks change of grammatical (and logical) subject than ὁ/ὅς introducing unframed discourse.
§82. The different particles that follow the pronoun introduce very particular kinds of acts. ὁ δέ marks a continuation of framed discourse or a resumption of framed discourse after unframed discourse. The cluster ὅ γε can carry the force of γε to a greater or lesser extent, thus offering a functional continuum. At one end of the continuum ὅ γε serves to activate a referent who is completely or largely to be inferred, that is, who has not been expressed linguistically in the preceding discourse. At the other end, ὅ γε is indistinguishable from the anaphoric pronoun ὁ itself, serving particularly to retrieve or pin down the character currently in focus near discursive transitions. As for Homer, neither the numbers nor current understanding of the anaphoric pronoun justify the widespread belief that ὁ δέ marks change of grammatical subject, whereas ὅ γε serves to mark grammatical subject continuity.
§83. Both ἄρα and δή after anaphoric pronouns have commonly been described as lending emphasis to the pronoun. [212] In translations, in fact, their presence is often not reflected at all, particularly in Homer. Thus, these particles go undiscussed, and patterns of use have remained unstudied. In the sections above I have provided one possible way of describing the differences between the use of ὅ/ὅς ῥα and ὁ/ὃς δή. When it introduces unframed discourse, ὅ/ὅς ῥα accompanies a verb in the imperfect tense that imparts a piece of shared or expected knowledge about the referent. In framed discourse, conversely, a verb in the aorist typically follows ὅ/ὅς ῥα, describing an action that is either already known or logically expected. ὁ/ὃς δή in Homer always introduces unframed discourse, barring one instance, mostly in direct speech, and offers new information about a referent. Often the newness of the act lies in the fact that it is a personal reflection of some sort. It is this last aspect that sets ὁ/ὃς δή slightly apart from ὁ γάρ. This last combination also typically serves to introduce new information, but γάρ, unlike δή, betrays no particular personal involvement, and can therefore be used freely by the narrator as well as by internal speakers.
§84. Both the corpus study of particle use after anaphoric pronouns in Homer and the “anaphora commentary” to Pindar’s Second Isthmian are meant as sorties into a huge field that has yet to be explored further. Building on the work of Bakker and Bonifazi I have attempted to demonstrate the importance of taking a discourse approach to anaphoric reference. Not only does this offer a deeper understanding of the process of anaphoric reference itself, but it provides a solid basis for better explaining certain aspects of difficult Homeric particles like ἄρα and δή. The complexities of anaphoric pronouns are yet another element of discourse to take into account when searching for patterns of particle use in Homer and Pindar.


[ back ] 1. See Chafe 1994:108-119 for the discussion of his “one new idea constraint.” He applies it to intonation units, but we use the term “discourse act,” see II.2.
[ back ] 2. For early literature on the idea of topic, see Chafe 1976 and Givón (ed.) 1983. H. Dik 1995 and 2007 applies a pragmatic approach to word order in Ancient Greek. Scheppers 2011 employs similar methodology to his idea of the “colon” in prose, a close cognate of what we call a discourse act.
[ back ] 3. So-called “epic” τε after pronouns has already been discussed for Homer in II.4.3.1 and II.4.4.4 and for Pindar in II.4.5. See also IV.2.3 for a more complete discussion of τε’s functions; for ὁ καί I refer the reader to the extensive discussion of the particle’s functions in IV.2.4. Finally, the function of γάρ after pronouns falls under the discussion of “γάρ introducing unframed discourse” in II.4.2.1.
[ back ] 4. The topic of referents in narrative is to a significant extent more relevant for Homer than for Pindar. First of all, the Homeric epics offer a reasonable number of instances of the phenomena, whereas Pindar’s Victory Odes have only a very limited number of the kinds of constructions under examination. This is the result of the fact that the Pindaric corpus is smaller, and less of it is narrative. Second, even in narratives, Pindar is much less concerned with scenes involving multiple characters. As a result, Pindar is less prominent in section 3, the comparative analysis, but the close reading of Isthmian 2 in 5.4 balances the asymmetry.
[ back ] 5. See e.g. Des Places 1947 and Bonifazi 2004c for Pindar, and Slater 1983 and Calame 1985 for Homer.
[ back ] 6. Cornish 1999:116-117.
[ back ] 7. This is an example from a cookbook, quoted by Halliday and Hasan 1976:2. One may compare the more famous, but constructed, example from Brown and Yule 1983:202, “Kill an active, plump chicken. Put it in the oven.” Consider another example, from Dinsmore 1987:15, “If J. Edgar Hoover had been born a Russian, he would have been a Communist.” Here the named character refers to a historical person, but the personal pronoun “he” refers to a hypothetical referent. Emmott 1997: 179-180 discusses this gap in coreferentiality.
[ back ] 8. See Berrendonner and Reichler-Béguelin 1995 for their arguments against what they call the “antecedentiste” (26-27) approach to reference.
[ back ] 9. See Cornish 1999:7 and 41-51; this description of the antecedent is influenced by Ariel 1996:17.
[ back ] 10. See especially Halliday 1967 and Halliday and Hasan 1976.
[ back ] 11. See e.g. Prince 1981 and Brown and Yule 1983:190-222.
[ back ] 12. Slightly adapted from Prince 1981:226-227.
[ back ] 13. Ariel 1988, 1990, and 1991.
[ back ] 14. Gundel, Hedberg, and Zacharsky 1993, expanded in multiple later publications.
[ back ] 15. The main difference between the two approaches is that whereas Ariel (accessibility) maps referential expressions directly on a status on the accessibility scale, Gundel’s approach (givenness hierarchy) allows for upward implication (see most recently Gundel, Hedberg, and Zacharsky 2012:252-254). That is to say that a referential expression that is marked low for givenness may be used for entities that are in fact higher on the givenness hierarchy in the minds of speaker and hearer. In simpler terms, it is “allowed” to overdetermine referents, but not to underdetermine them (since in the latter case the communication would probably be unsuccessful).
[ back ] 16. Consider the scheme proposed for English by Gundel et al. 2012:251: [ back ] in focus > activated > familiar > uniquely identifiable > referential > type identifiable [ back ] (it) > that/this/this N > that N > the N > indef. this N > a N [ back ] In the Accessibility Marking scale by Ariel 1991:441, the list is even more extensive; see Cornish 1999:6-8 for a discussion of both approaches.
[ back ] 17. Bakker 1997:111 attempts to create such a scheme for Homer.
[ back ] 18. For the reader’s convenience, I henceforth use the term accessibility to cover both the ideas of givenness and accessibility.
[ back ] 19. Cornish 1999:5-6, but the term goes back to Prince 1981:235. On page 5 Cornish defines discourse model: “This model is a coherent representation of the discourse being evoked via the co-text and its context in terms of the speaker’s or writer’s hypothesized intentions.”
[ back ] 20. See also Ariel 1991:444 “[T]he claim is that addressees are guided in antecedent retrievals by considering the degree of Accessibility signalled by the marker, rather than by noting the contextual source marked (general knowledge, physical salience, linguistic material), as had commonly been assumed by pragmaticists (Clark and Marshall 1981, Prince 1981, inter alia).”
[ back ] 21. Bonifazi 2012:19-38 proposes a similar approach to anaphoric markers in Homer, with a main focus on (ἐ)κεῖνος and αὐτός.
[ back ] 22. Chafe 1994:75 “For the most part, both new and accessible information are expressed with accented full noun phrases, whereas given information is expressed in a more attenuated way.”
[ back ] 23. It occurs 21 times: Iliad 4.504, 5.42, 5.540, 5.617, 11.449, 13.187, 13.373, 13.442, 15.421, 15.524, 15.578, 16.325, 16.401, 16.599, 16.822, 17.50, 17.311, 17.580, 20.388; Odyssey 22.94, 24.525; see Kirk 1985:392.
[ back ] 24. Although the LSJ s.v. δουπέω takes the main meaning of the verb to be “sound heavy or dead,” I follow Chantraine 19992:282 in reading it as “the clatter or noise (of battle)” (“fracas des lances” “bruit de la bataille”), since outside of this formula the verb is used chiefly to describe the sound of battle or of the sea.
[ back ] 25. See also Chafe 1994:175: “Whether or not a referent is assumed to be newly activated in the listener’s consciousness is a different question from whether or not it is assumed to be already part of the listener’s knowledge.” Chafe describes these two separate domains in terms of “active/inactive” versus “shared/unshared.”
[ back ] 26. Chafe 1974:122.
[ back ] 27. Chafe 1994:71-81.
[ back ] 28. Bakker 1997:108-111 and passim.
[ back ] 29. Bonifazi 2012:19-26.
[ back ] 30. Bonifazi 2012:28-38.
[ back ] 31. For κεῖνος and αὐτός forms referring to Odysseus in the Odyssey see Bonifazi 2010 and 2012:38-183.
[ back ] 32. ὅς: 375x in Iliad and 229x in Odyssey.
[ back ] 33. ὁ/ὅ: 751x in Iliad and 164x in Odyssey.
[ back ] 34. This distinction is not addressed in Des Places 1947:35-50, Hummel 1993:174-177, or Bonifazi 2004c.
[ back ] 35. For a concise exploration of the issue, see Bonifazi 2004c (with a focus on Pindar); see also Bakker 1999 and 2005:77-80 on ὁ and οὗτος in Homer.
[ back ] 36. ὁ represents the pronoun from the PIE root *to, i.e. ὁ, ἡ, τό, while the relative pronoun from the PIE root *yo gives the paradigm ὅς, ἥ, ὅ. The TLG edition of the Odyssey does not accentuate the demonstrative pronoun, while the edition of the Iliad does. For reasons of consistency, and in accordance with common practice, I have chosen to give ὁ throughout.
[ back ] 37. Probert [forthcoming] section 2.5 quotes this passage as an early example of a relative clause in ancient Greek, and says: “it is not difficult to think of the structure found in (2.1) [sc. “Calchas, who...”] as having come from a structure of type (2.2) [sc. “Calchas. He...”].”
[ back ] 38. Translation Murray. This is in fact a unique instance of ὃς δή in Homer; only here does it introduce framed discourse. See below §59-§65 for the more common pattern of use of this pronoun and particle combination.
[ back ] 39. Behind Murray’s translation there might also be a presupposition about the status of main clauses and subordinate clauses. The former are generally regarded as carrying the narrative forward, while the latter offer “background” information. This view is challenged by Cristofaro 2003, and specifically for ancient Greek by De La Villa 2000 and Bonifazi 2004c; see also Probert 2015. For further discussion of background and foreground see II.3 §§25-27, II.4.2, and IV.5 §68
[ back ] 40. See below, especially §49-§66, for the possible relevance of the use of the imperfect (sometimes present or pluperfect) in a context of narrative told in aorists.
[ back ] 41. See II.4 §§11-14 for the terms “framed” and “unframed” discourse.
[ back ] 42. I exclude the feminine singular pronoun in the nominative for practical reasons. First, the feminine pronoun (ἡ and ἥ) occurs much less frequently; second, the constructions in which it partakes do not differ from that of the masculine pronoun that I discuss below.
[ back ] 43. For a discussion of pronoun use in pro-drop languages, see Frascarelli 2007:694-696, with extensive references.
[ back ] 44. See Cornish 1999:63 on pronouns in English and French: they “signal referential and attentional continuity.” He adds, “this is the case where they are unaccented in English, and clitic in French. Where they are accented, their indexical properties change: in particular, they are capable of referring to entities which, though assumed to be recoverable by the addressee, are not the ones enjoying the highest degree of focus at the point of use.”
[ back ] 45. If there is no ambiguity of anaphoric reference, there is a tendency in English to keep reference to the subject “light,” called the “light subject constraint” by Chafe 1994:82-92; Greek appears to function similarly.
[ back ] 46. Tomlin 1997:181-186 discusses the nominative in terms of attention: the nominative marks the referent being attended to, and this “being attended to” is generally initiated in the preceding act. Consider especially the note on 182: “That is, the grammar of English does not look at the semantic role of an argument when determining subject selection; it only looks for the current output of attention system [sic] – the attentionally detected event parameter.”
[ back ] 47. Compare the following constructed example from Prince 1981:227 “John called Sam a Republican and then HE insulted HIM.” The accented pronouns (as opposed to “John called Sam a Republican and then he insulted him”) suggests that there is a shift of subject, leading to the assumption that “HE” refers to Sam, whereas “he” would most naturally have referred to John. See also the example in Prince 19n34. Fox 1987:172 shows that in written English narrative the accented pronoun would most probably take the form of a full noun phrase. For English, Givón 2005:136 shows that zero anaphora and unstressed pronouns signal maximal referential continuity whereas constructions containing a stressed pronoun or even stronger marking signal referential discontinuity.
[ back ] 48. I use marked/unmarked in the sense proposed by Givón 2005:139: “maximal-continuity anaphoric devices – zero-anaphor and unstressed/clitic pronoun – are the least marked devices, carrying the smallest phonological weight and lacking independent lexical status” [emphasis original].
[ back ] 49. See Bonifazi 2012:26.
[ back ] 50. First in Chafe 1976.
[ back ] 51. See Levinson 1987:68 “The less you say, the more you mean.”
[ back ] 52. See Cornish 1999:6 “[T]he speaker’s task in referring must be to choose a referring expression marking the level of cognitive accessibility of the intended referent which matches that which s/he assumes the entity in question currently enjoys in his or her addressee’s mental model of the discourse under construction,” and 20, “the type of signalling device (...), which is most likely to get the addressee to grasp the referent intended in the most economical manner possible.”
[ back ] 53. Gundel et al. 2012:251: “A speaker, in producing a particular determiner or pronoun, thus provides a processing signal to the addressee that helps restrict the set of possible referents.” On pages 252-253 they explain how the hierarchy works on the basis of how informative the linguistic referring expression is.
[ back ] 54. Chafe 1994:77-78 discusses the factor of contrastiveness (creating contrast between one referent and another, which may or may not be expressed) as a reason for using accented forms in spoken English when unaccented forms might have been expected; Emmott 1997:86 (with reference to Longacre 1974 and 19962) notes that there may be literary reasons for “lexical reiteration” instead of pronominalization.
[ back ] 55. I use “combination” as a neutral term for two or more particles or other words that co-occur, and “cluster” for recurrent combinations whose resulting function either extends beyond, or is significantly different from, the sum of its parts; see I.1.
[ back ] 56. See II.2 §§31-36.
[ back ] 57. See for example ἷξε δέ in (t15) below.
[ back ] 58. Chantraine 1953:159 remains vague when he says “[l]a particule et l’article servent souvent à indiquer un changement de sujet.” Since he discusses ὁ much more as a pronoun than as an article, it is striking that he calls it an “article.”
[ back ] 59. Janko ad Iliad 16.467, where ὁ δέ occurs despite continuity of grammatical subject. It is probably comments like this that lead to claims like that in Raible 2001:593 “Languages using this [zero anaphor] technique tend to develop a special morpheme signaling a different subject (...) in the subsequent clause. In classical Greek this is the function of the particle de.” While Janko’s generalization holds, Raible’s claim for δέ alone oversimplifies. He might rather have said that in combination with a pronoun in the nominative, δέ often signals a change of subject.
[ back ] 60. For this use of ὅ γε – to help retrieve the referent after an intervening discourse discontinuity of some sort (here a simile) – see §§27-50 below.
[ back ] 61. ᾤκει is a single imperfect among aorists (δώκε, 197, περισχόμεθα, 199, πόρεν, 201): this is typical for unframed discourse in Homer; see II.4.2, and §§51-62 below on ἄρα.
[ back ] 62. See II.4 §14 and Emmott 1997:239 and 248-252.
[ back ] 63. The reason given in the scholia is that Poseidon could not have drawn the spear from the shield since in lines 276-279 the spear is described as landing on the ground. Edwards 1991:327 plausibly defends the lines by taking it to mean that the shield had been pinned to the ground by the spear, “which is realistic enough.”
[ back ] 64. The codex Marciana 458 has a comma after ὁ δέ, while both Escorial Ω and Venetus B have a sign of a double grave over δέ: “ δὲ` ”, which I interpret to be some kind of instruction for prosodic discontinuity. This is why I use a comma in the English translation. It resembles Iliad 20.455-456, where there is a clear boundary after ὁ δέ: ὣς εἰπὼν Δρύοπ’ οὖτα κατ’ αὐχένα μέσσον ἄκοντι· / ἤριπε δὲ προπάροιθε ποδῶν· ὁ δὲ | τὸν μὲν ἔασε. In this passage ὁ δέ does mark a change of subject (ἤριπε has Druops as its subject), but as in 20.322 there is a transition to a new episode of the narrative. See IV.3.4-5 on more about ancient and medieval punctuation and its links to our act boundaries.
[ back ] 65. The use of αὐτίκα may also have contributed to the vividness of the scene, see Bonifazi 2012:273-281, with reference to the present passage in 280n40.
[ back ] 66. ὁ δέ accompanies continuity of grammatical subject only in Pythian 4.78, see (t32).
[ back ] 67. See Gildersleeve 1885:ci and Bonifazi 2004c:49-54. In the present and following notes I give a list of the uses of ὁ δέ in Pindar. I read the article ὁ in Olympian 1.1, 8.28; Pythian 1.35, 9.78, 11.30; Nemean 7.67; Isthmian 7.39.
[ back ] 68. See Chantraine 1953:160-162, 165-166; Chantraine notes that in the Homeric books generally regarded as more recent, the use of the “article” is closer to that in classical Greek (165), and adds “on a pu supposer qu’à l’époque d’Homère, la langue courante connaissait déjà l’article, mais que l’épopée conservait traditionellement l’emploi démonstratif de l’article.” Bakker 2005:76n12 adds that “in many cases the “article” is more marked [in Homer] than in Attic Greek.”
[ back ] 69. Ι read ὁ as a relative pronoun in Olympian 1.73, 10.43 (δ᾽ ἄρα); Pythian 1.8, 4.78 (δ᾽ ἆρα), 3.92, 6.33, 9.17, 11.34 (δ᾽ἄρα); Nemean 1.43, 1.61, 7.36, 10.13; Isthmian 6.41.
[ back ] 70. Pythian 2.73 ὁ δὲ Ῥαδάμανθυς (“(that) Rhadamanthys”) and 5.60 (“(that) Apollo”). Bonifazi 2004a links this to the idea of “recognitional deixis,” from Diessel 1999, to mark a referent that is new in the discourse, but already known to speaker and listener. Using an article with a name is the exception rather than the rule in Pindar, which suggests that at least some demonstrative force may be attributed to ὁ in these cases.
[ back ] 71. Bonifazi 2004c:50.
[ back ] 72. Literally: “whom good rumours hold.”
[ back ] 73. The parallels are: Olympian 10.66 (“He, who”); Pythian 8.48 (“He, who”; in this instance as for Pythian 4.78-79, (t33) below, Giannini in Gentili 1995:575 reads ὁ as an article), 8.88 (“He, who”); Nemean 5.34 (“He, (...) Zeus”), 9.24 (“He, (...) Zeus”). [ back ] The construction does occur in the Hymns: Homeric Hymn to the Muses and Apollo 4 ὁ δ᾽ ὄλβιος and Homeric Hymn to the Earth Mother 7 ὁ δ᾽ ὄλβιος; and there is a rare instance in Homer Iliad 1.139 ὁ δέ κεν κεχολώσεται ὅν κεν ἵκωμαι “he will be angry, to whom I will come.”
[ back ] 74. Liddell, Scott, Jones, and McKenzie 19409:1195b (s.v. ὅ γε).
[ back ] 75. Denniston 1950:114-115.
[ back ] 76. Denniston 1950:122. He provides an analogy from English: “The same tendency occasionally shows itself in English, as when we say ‘Not I’, meaning ‘I certainly did not.’” De Jong 2012:68 ad Iliad 22.33 (ὅ γε) comments that “the anaphoric pronoun is often redundant and γε unnecessary (..). The combination is found very often, however, and may have been of metrical use to the singer.”
[ back ] 77. See Leaf 1900:149 on Iliad 3.409, about ὅ γε in general: “ὅ γε (...) merely resumes the original subject” and Neitzel 1975:47 “ὅ γε steht bei Homer immer demonstrativ als masc. nom. Es nimmt das Subjekt des Satzes betont wieder auf.”
[ back ] 78. Out of a total of 127 instances in the Iliad, ὅ γε marks grammatical subject change in 58 instances. This number is 27 out of 62 for the Odyssey, so slightly less than half in both epics.
[ back ] 79. Leaf 1900:18 rightly notes that “ὁ δέ as often repeats the subject of the first clause.”
[ back ] 80. Consider also that Homer has the elided form ὅ γ᾽ 104 times and the full form ὅ γε 86 times.
[ back ] 81. When used in scalar wishes, περ marks something that is still attainable (“inclusive”) while γε marks something that is impossible to attain (“exclusive”), see Bakker 1988:97-98.
[ back ] 82. See Sicking 1986:125 and Wakker 1994:308.
[ back ] 83. Tsakmakis 2010:350; see also Slings 1997:126, who believes that the idea of γε as a focus marker “cannot do justice to its use in adding constituents to already complete sentences (quite apart from the fact that γε hardly ever accompanies the true Focus of a sentence).”
[ back ] 84. See Bonifazi 2012:31.
[ back ] 85. See III. for an exploration of this aspect of γε marking dialogic resonance across utterances in tragic and comic dialogue, and see III.4.5.2 for γε in answers to express a speaker’s stance.
[ back ] 86. See §46-§47 below for more on ἀλλ᾽ ὅ γε.
[ back ] 87. Compare Berrendonner 1990:28 “Au pronom ne s’attache donc pas nécessairement une assomption d’existence: il ne comporte en lui-même aucune présupposition concernant des objets de connaissance qui devraient déjà figurer dans M [=mémoire discursive].”
[ back ] 88. If one wishes to find a textual antecedent for ὅ γε, one can choose to read “Let it not be that someone is a wholly godless man, but let him own his gifts from the gods in silence.” In this reading ὅ γε could be taken as referring to τίς, and a similar reading could be wrangled out of (t19). However, ὅ γε here is better understood as triggering the creation of a generic referent than retrieving an earlier one.
[ back ] 89. See Iliad 21.113 for another instance of ὅ γε retrieving τις, and similarly ὅ γε retrieving ᾧ κεν in 24.530; compare Cornish 54-56 with examples 2.20a and 2.20b for similar constructions in English and French.
[ back ] 90. See Cornish 1999:112-148 for a discussion, with extensive literature.
[ back ] 91. See Cornish 1999:30, who discusses a speaker using an “accented demonstrative pronoun (THAT) fulfilling a deictic function, in order to render accessible and salient an item of information which (...) was in the background, not the foreground, of attention.” There are several similar cases: Iliad 13.53 (Poseidon speaking of Hector who is in sight), 13.70 (about Poseidon/Kalchas), and 19.344 (κεῖνος ὅ γε, Zeus to Athena about Achilles).
[ back ] 92. Especially in instances like this I cannot agree with Bakker 1999:6 that “ὁ is used (...) to refer to any person or thing that the speaker cannot actually point at.”
[ back ] 93. Compare (t8) and (t9) above.
[ back ] 94. Bonifazi 2012, especially 159-172. Compare also ὅ γε in Odyssey 19.575 (reference to Odysseus by Penelope, spoken to Odysseus-in-disguise).
[ back ] 95. Odyssey 8.14 and 23.163.
[ back ] 96. Odyssey 16.182-183: ἄλλα δὲ εἵματ’ ἔχεις καί τοι χρὼς οὐκέθ’ ὁμοῖος. / ἦ μάλα τις θεός ἐσσι.
[ back ] 97. Another possible reason for the strong referring expression is the fact that there is a scene boundary after line 468. Compare Odyssey 1.443 for ὅ γε marking a similarly strong visual focus on Telemachus, as he sits on his bed and ponders his future.
[ back ] 98. See also Monro 1882:258, who argues that γε after pronouns serves “to bring out the contrast which (...) Pronouns [sic] more or less distinctly imply.” More generally, Hartung 1832:371, Kühner 1835:398, and Stephens 1837:92 remark that γε can mark a contrast with something left implicit; see III. for a closer analysis of γε marking a contrast with something implicit in tragic stichomythia.
[ back ] 99. Chafe 1994:76-78, examples on 77: “ín” in example 7b (backward), and “dóctor” in example 8c (forward). Compare Grégoire 1930:163, who notes that when γε in Homer has the ictus, it usually follows a form of ὅ. In these cases, he believes that the pronoun-particle combination is in some kind of opposition with a preceding element.
[ back ] 100. The other instances of ἦ τοι ὅ γε where there is no subject continuity are Iliad 11.94 and 19.100.
[ back ] 101. Compare the discussion of ὅ γε in direct speech to refer to “that man there” in (t20).
[ back ] 102. It is not “semantisch redundant” as Latacz 2003:II.2.215 claims.
[ back ] 103. Chafe 1994:77, “Contrastiveness is independent of activation cost.”
[ back ] 104. See Monro 1882:258-259, Denniston 1950:119, Chantraine 1953:II.159.
[ back ] 105. See Gesamtkommentar III.2.143, “ὅ γε betont im zweiten Satzglied die Identität der (...) unterschiedlich handelnden Person,” with parallels.
[ back ] 106. ἄλοχος does not mean “wife” here, but “concubine of equal class” (Gesamtkommentar III.2.143, “‘Konkubine, Geliebte’ von ebenbürtigem Stand.”).
[ back ] 107. It does not “merely resume the original subject,” as Leaf 1900:I.149 believes.
[ back ] 108. Shipp 1961:14-15 and 19722:240 rather believes that δούλη here means “slave-concubine,” but this is rejected by Krieter-Spiro in the Gesamtkommentar, with references. Regardless of the exact meaning of ἄλοχος and δούλη here, the scalar sense is clear: “until he makes you a concubine” (unexpected), and “until HE makes you a slave” (even more unexpected).
[ back ] 109. See Bonifazi 2012:37 “The particle γε might contribute (...) by stressing the paradoxical novelty introduced by the ongoing discourse act.”
[ back ] 110. See II.4.3.2 and below §§51-62 for frame recall after unframed discourse. The same construction occurs after an intervening relative clause, as in Odyssey 15.252-255, or after a simile, see ὅ γε in (t13) above. Sometimes the gap between the final mention of the referent and the retrieval through ὅ γε is rather long, as in Iliad 17.108 and Odyssey 17.514. A change of subject while the referent is in focus occurs in Odyssey 18.398.
[ back ] 111. De Jong 2001:584 reads lines 423-424 as focalized through Eupeithes, upon which the narrator intrudes with his knowledge of whom Odysseus killed first: “The narrator intrudes upon his embedded focalization (...) by adding the detail ‘first’ (something which the father cannot know).” Unlike the instances of δὴ γάρ (see II.4 §19), however, there is nothing in the language here to suggest a blurring of perspectives. I read the lines as unframed discourse, directly from performer to audience, where he shares knowledge only he can have. The Homeric performer knows Eupeithes’ mind and he can tell the audience that his son Antinous was in fact the first to be killed.
[ back ] 112. ὅ γε right before direct speech: Iliad 1.93, 2.55, 2.109, 4.357, 8.138, 13.94, 13.480, 17.219, 19.100, 21.367, 23.5, 23.42; Odyssey 1.31, 2.24, 4.189, 13.254, 17.466, 18.110, 24.425, ; ὅ γε right after direct speech: Iliad 1.68, 1.101, 2.76, 2.207, 4.250, 7.354, 7.365, 9.620; Odyssey 2.224.
[ back ] 113. Iliad 1.68, 1.101, 2.76, 7.354, 7.365; Odyssey 2.224.
[ back ] 114. For this interjection-like reading of ἦ at strong discursive discontinuities, see II.3.2.3.
[ back ] 115. In both Iliad (8x) and Odyssey (8x) there is the speech-capping construction ὣς φάθ᾽, ὁ δέ, always marking subject change. A similar construction occurs in the Odyssey only, but accounts for 14 out 41 instances of ὁ δέ: ὣς ἐφάμην, ὁ δέ.
[ back ] 116. ὅ γε is especially visually relevant in Iliad 5.585 (the image of the warrior standing upright with his head in the sand) and Odyssey 17.302 (Odysseus’ dog Argos).
[ back ] 117. There are even a few instances where the two constructions intersect, as in Iliad 21.550, 21.581, and Odyssey 11.190, 20.140, 22.116.
[ back ] 118. The cognitive priming is primarily local and visual/experiential in the construction φῆ δ᾽ ὅ γε: Iliad 2.37; Odyssey 17.142, 24.470. For the loci of ὅ γε near the beginning of direct speech, see note 112 above.
[ back ] 119. I read a similar projecting function of ὅ γε in: Iliad 15.455, 24.189; Odyssey 22.480.
[ back ] 120. There are also 6 instances of the construction where there is no subject continuity: Iliad 2.3, 15.676, 17.705, 24.14, Odyssey 5.82, 14.526.
[ back ] 121. 20 times in the Iliad and 9 times in the Odyssey; the 4 exceptions are Iliad 1.281, 5.434 (but this instance is textually uncertain), and 8.311 (=13.518), where we find a different construction. I am not sure what Denniston 1950:12 means by: “Often the emphatic word or phrase in the ἀλλά-clause (which word or phrase follows immediately, or almost immediately, after the particle) is limitatively qualified by γε (...). Homer never has ἀλλά ... γε,” especially since he quotes Iliad 1.281 ἀλλ᾽ ὅ γε on page 11.
[ back ] 122. See V.ἀλλά passim.
[ back ] 123. The parallels are (line numbers refer to the instances of ἀλλ᾽ ὅ γε): Iliad 1.320 (λῆγ᾽ ἔριδος), 4.389 (τάρβει), 5.321 (ἐλήθετο), 12.305 (μέμονε), 12.394 (λήθετο), 13.523 (πέπυστο, but the text is uncertain for ἀλλ᾽ ὅ γε), 15.676 (Αἴαντι .. ἥνδανε θυμῷ), 17.705 (ἤθελε), 21.581 (ἔθελεν), 24.14 (μιν λήθεσκεν); Odyssey 5.82 (ἠγνοίησεν), 9.554 (ἐμπάζετο), 14.422 (λήθετο), 14.526 (συβώτῃ ἥνδανεν).
[ back ] 124. Iliad 2.3 (Δία δ’ οὐκ ἔχε .. ὕπνος), 6.504 (δήθυνεν), 15.586 (μεῖνε), 23.5 (εἴα ἀποσκίδνασθαι); Odyssey [5.33 the negation takes the form of an adverbial phrase “without interference by gods or men”], 11.190 (οἱ <ἐστι>), 12.188? (παρήλασε, see t18), 18.142 (εἴη).
[ back ] 125. Iliad 2.420 (ἐπεκραίαινε), 22.92 (ἔπειθον); Odyssey 9.288 (ἀμείβετο).
[ back ] 126. For the more well-known variant of a counterfactual construction (“καί νύ κεν...εἰ μή) see Louden 1993. His argument is that these constructions occur at pivotal moments in the narrative.
[ back ] 127. See Gesamtkommentar II.2.215, “semantisch redundant.”
[ back ] 128. See II.4 §§34-37 and §§46-49.
[ back ] 129. See below note 159 as well as (t48) with note 198 for δή after pronouns in Pindar.
[ back ] 130. The parenthetical apposition comes much later in the Greek, but sticking to this position would yield an ambiguous English translation.
[ back ] 131. See Bonifazi 2012:137-184 for more on the layered meaning of αὐτός, noting especially the strong link with death (141-143 for the body of Patroclus); I merely point out the proximity of θάνεν here.
[ back ] 132. See II.4 §§57-71 for more on the link between τε and tradition in Pindar.
[ back ] 133. I take issue with Des Places’ reading, 1947:45 “11, 34 : ὁ δ᾽ (= Oreste) après (θάνεν) μὲν αὐτὸς ἥρως (31), qui tient lieu d’ὁ μέν.” From the point of view of discourse and performance there is no reason to see a link between μέν and δέ in this passage.
[ back ] 134. Snell/Maehler give the nominative νέα κεφαλά, following Heyne’s reading, rather than the dative νέᾳ κεφαλᾷ of the manuscripts (followed by Gentili 1995); either reading suits my interpretation, although in the manuscript reading it is not an appositive.
[ back ] 135. Olympian 10.43-45 is a good parallel (ὁ δ’ ἄρ’ ἐν Πίσᾰͅ ἔλσαις ὅλον τε στρατόν / λᾴαν τε πᾶσαν Διὸς ἄλκιμος / υἱὸς) with the difference being that the referent in that song, Heracles, is not retrieved by association as Orestes is here. However, Heracles is available at the occurrence of ὁ δ᾽ἄρα by virtue of the death of the main referent in focus (θάνατον αἰπὺν οὐκ ἐξέφυγεν, 42). Just as in Pythian 11, the pronoun is followed by a clarifying appositive later in the sentence (Διὸς ἄλκιμος / υἱὸς).
[ back ] 136. The manuscripts have ἄρα, which in most current editions is emended to ἦρα since a heavy syllable is desired; see Braswell 1988:173-174 and De Kreij [forthcoming 2014b] for discussion; Gentili 1995:128 and Liberman 2004:100 also read ἆρα.
[ back ] 137. Even if some members of the audience may not have known the episode well enough, the fact that something is fated (θέσφατον ἦν, 71) means that it will happen.
[ back ] 138. See Bonifazi’s comment at 2004c:63 about pronouns in Pindar: “[t]hrough them both the performer and the audience enter in the dimension of epic memory.”
[ back ] 139. ἔκπαγλος is most likely (through dissimilation *ἔκπλαγλος > ἔκπαγλος) from ἐκπλήσσω “expel,” “hunt.” If this is correct, the original sense of “outcast” might interact here with the more generic “terrible.” Pindar refers at once to the “terrible” hero and the “exiled” hero.
[ back ] 140. Most editions print ὃ δ᾽ ἄρ᾽ both here and in 383, but since I regard ὁ as referring to Ajax, I give ὁ δ᾽ ἄρ᾽, in this instance supported by the best manuscript (Venetus A gives ὁ δ᾽ἄρ᾽ in both 383 and 385).
[ back ] 141. Similar instances of ὁ δ᾽ ἄρα clearly marking frame recall in Iliad 2.268, 10.354, 19.367, 21.174, 21.246; Odyssey 5.392, 5.453, 19.447, 19.464, 20.275, 23.90.
[ back ] 142. See Bakker 1993:15-25 and 1997:17-20. As for the combination ὁ δ᾽ ἄρα, compare Iliad 19.367, where the terrifying image of Achilles as he prepares to re-enter the battle is followed by ὁ δ᾽ ἄρα; similarly Iliad 12.462 (Hector leaps through the broken gates, followed by a vivid description), Odyssey 5.456 (note Murray’s translation “So he lay breathless...”). Finally, in 23.90, it appears the hearer is invited to share a character’s perception, when Penelope has just entered the hall and sits across from Odysseus, after which: ὁ δ᾽ ἄρα πρὸς κίονα μακρὴν / ἧστο κάτο ὁρόων, which I am tempted to read as “And there he sat...”
[ back ] 143. See the discussion of discourse memory in II.4.1
[ back ] 144. Iliad 3.120, compare Iliad 5.836, 7.188, 7.416, 8.319, 10.374, 12.385, 13.192, 15.543, 16.413, 16.579, 16.742, 21.118; Odyssey 5.456, 8.450, 12.411, 12.413, 14.485, 15.98, 15.121 (this instance is special in the sense that it retrieves specific information that was added to the discourse memory in lines 102-104), 18.396.
[ back ] 145. Iliad 10.350; the boundary between common sense and traditional knowledge was probably fuzzy in the Greek world, but I would offer the following loci as parallels: Iliad 6.154 (genealogy), 17.196 (Achilles’ armor), 20.239 (genealogy), 21.189 (genealogy), 23.642 (the sons of Actor); Odyssey 11.257 (genealogy).
[ back ] 146. The one exception where ὁ δ᾽ ἄρα accompanies grammatical subject continuity is Iliad 11.426.
[ back ] 147. Namely Iliad 16.120 and Odyssey 24.182.
[ back ] 148. Iliad 12.380 (imperfect, see t34 above), 13.644 (imperfect), 22.470 (aorist); Odyssey 22.327 (aorist).
[ back ] 149. The parallel instances of ὅ(ς) ῥα + imperfect introducing unframed discourse are: Iliad 2.77, 2.752, 5.70, 5.612, 5.708, 6.18, 6.131, 11.123, 13.665, 15.431, 15.461, 16.178, 16.464, 16.572, 17.611; Odyssey 1.154, 2.225, 9.187, 16.396, 22.331. Sometimes with the pluperfect: Iliad 5.77, 12.445, 13.364, 17.350; Odyssey 24.445 or present: Iliad 15.411, 22.23, 22.27, 23.517; Odyssey 15.319, 22.403. The present tense is generally used with ἄρα when it occurs in the first part of a simile, see e.g. II.4 (t16).
[ back ] 150. The parallel instances of ὅ(ς) ῥα + aorist to effect frame recall are: Iliad 1.405, 10.318, 11.231, 15.584 (after an apostrophe of a character), 15.644, 17.72 (after ἔνθα κε...εἰ μή flash-forward), 23.384 (after καί νύ κε...εἰ μή flash-forward); Odyssey 10.158 (beginning of little narrative), 14.380, 20.291, 21.184.
[ back ] 151. See for another interpretation of ὅς ῥα (τε) Ruijgh 1971:432-443; he follows Hartung and Denniston and believes that ἄρα in this construction serves to mark a certain measure of surprise or interest on the part of the speaker.
[ back ] 152. Odyssey 12.281 is different: a present in direct speech which actually refers to the present.
[ back ] 153. Odyssey 10.158, 14.380. The only extant instance of ἄρα after a pronoun in Pindar is found in fr. 125.1 τόν ῥα, where it also appears to introduce an embedded narrative about how Terpander invented the bárbiton (note the aorist εὗρεν). See II.3.2.2 on γάρ beginning embedded narratives.
[ back ] 154. These are the only two instances of πατρὶ χολωθείς in Homer. Moreover, there is a strong phonetic resonance between Φυλεΐδης and Πολυφείδης; see Minchin 2001:88-90 for this kind of “auditory memory” in Homer.
[ back ] 155. Other instances of ὅ(ς) ῥα apparently introducing unframed discourse with the aorist: Iliad 5.612 (in direct speech), 5.650 (in direct speech), 6.158, 16.328; Odyssey 3.161 (in a σχέτλιος comment).
[ back ] 156. However, whether to read a pronoun as relative or demonstrative is always a subjective decision to some extent.
[ back ] 157. ὁ δή occurs four times in the Iliad and five times in the Odyssey, ὃς δή eight and ten times, respectively. See my discussion of those four instances in §69 below; also compare the plural 3.134 οἳ δὴ νῦν.
[ back ] 158. About the latter combination, note Denniston 1950:123 “The particle which normally stresses a relative relation is δή.”
[ back ] 159. In Pindar this pattern does not hold, as τὸν δή in Pythian 11.17 starts a little embedded narrative.
[ back ] 160. In fact, even καί sometimes abandons its strict initial position (with scope over the act) for the sake of a pronoun, see IV.2 §70.
[ back ] 161. ὁ δή/ὃς δή (and because of its frequency I include neuter ὃ δή) with act scope is paralleled in Iliad 1.388 ὃ δὴ τετελεσμένος ἐστί (Achilleus about Agamemnon’s threat), 2.436 ἔργον ὃ δὴ θεὸς ἐγγυαλίζει, 17.202 ἆ δείλ’ οὐδέ τί τοι θάνατος καταθύμιός ἐστιν / ὃς δή τοι σχεδὸν εἶσι· (Zeus, as if to Hector); Odyssey 4.777 ἀλλ’ ἄγε σιγῇ τοῖον ἀναστάντες τελέωμεν / μῦθον, ὃ δὴ καὶ πᾶσιν ἐνὶ φρεσὶν ἤραρεν ἥμιν (Antinous about the suitors’ plan), 10.514 Κώκυτός θ’, ὃς δὴ Στυγὸς ὕδατός ἐστιν ἀπορρώξ (Circe to Odysseus), and 15.490 ἠπίου, ὃς δή τοι παρέχει βρῶσίν τε πόσιν τε. Compare Pindar Isthmian 2.27 τὰν δὴ καλέοισιν.
[ back ] 162. See II.3.3.2n192 for scholars linking δή to perception or evidentiality. A similar link between δή and perception is drawn in III.2 §§73-79 for tragedy and comedy and in IV.4.5.2 for historiography.
[ back ] 163. Ten or eleven depending on the reading of Odyssey 2.16.
[ back ] 164. The parallels are Iliad 15.291 (ὃς δὴ πολλῶν), 2.117 (ὃς δὴ πολλάων), 9.24 (ὃς δὴ πολλάων). See II.3 §61n201 for more parallels of δὴ + πολύς and similar adjectives.
[ back ] 165. δὴ γάρ also often introduces personal viewpoints of some kind, see II.3 §62 and II.4 §19. For the concept of stance and for its relevance to the use of δή in Herodotus and Thucydides, see IV.4.4-6.
[ back ] 166. Parallels: Iliad 21.315 δὴ νῦν; Odyssey 1.49 δὴ δηθά and 2.48 δὴ τάχα; perhaps also Odyssey 2.16 δὴ γήραϊ κυφός “bowed through great age.”
[ back ] 167. Compare the use of δή after τόσσον in Iliad 17.522, discussed in II.4 §§38-41.
[ back ] 168. Compare also ὅς δή in Odyssey 7.156; the line recurs in Odyssey 11.343 (about the same character), but is there omitted by many manuscripts.
[ back ] 169. The other instance is Odyssey 20.289, where ὅς δή τοι similarly forms a priming act, but where the visual component is not clearly present; Hainsworth 1993:186 notes the structural similarity of Iliad 10.314-318 and Odyssey 20.287-291.
[ back ] 170. Book ten of the Iliad is often called the Doloneia after this episode; see e.g. Danek 1988, Dué and Ebbott 2010, and Finkelberg 2011:I.216-217.
[ back ] 171. Compare in Homer for example Odyssey 17.256-257: αὐτίκα δ’ εἴσω ἴεν, μετὰ δὲ μνηστῆρσι καθῖζεν,/ ἀντίον Εὐρυμάχου· τὸν γὰρ φιλέεσκε μάλιστα; “At once he went inside, and sat among the suitors, / across from Eurymachus; for him he liked most”; note especially the iterative φιλέεσκε which demonstrates the unframed nature of the act. The parallels in Pindar are Olympian 3.36, 6.25, 7.23, 9.28, 13.6; Pythian 4.281; Nemean 6.17; Isthmian 1.17.
[ back ] 172. By way of example, in Verdenius’ 1982 commentary on Isthmian 2 there are 138 lemmata, of which only 18 do not concern verbs, nouns, adjectives, or adverbs; 10 of these 18 lemmata concern particles. Out of the total 138, only 11 lemmata discuss problems concerning the referent of a (pro)noun or verb.
[ back ] 173. Bury 1892:26-37 offers an excellent introduction on the song’s context. Bury believes that there may have been an amorous connection between Pindar and the young Thrasyboulus (33), but he is not followed by Privitera 1982:29 or Verdenius.
[ back ] 174. Bury 1892:31 prefers the later date, Privitera 1982:27-28 the earlier; Verdenius 1982:1 proposes 472 BCE, but does not discuss the implications.
[ back ] 175. See II.1 §5 for the issue of the “I” persona in Pindar, and its relevance to questions of performance.
[ back ] 176. Athanassaki 2012:199 takes this presentation as emphasizing the audience’s reaction to Xenocrates’ and Nicomachus’ athletic feats in the games.
[ back ] 177. Frame shift always coincide with move transitions, but there may be (and often are) multiple moves between frame shifts. In IV.3 79, we use || to mark the boundaries between moves. Since the commentary concerns anaphoric reference, frame transitions are most important to mark.
[ back ] 178. LSJ s.v. πάλαι I.2.
[ back ] 179. See II.3 §§ 49-50 for asyndeton at move beginning.
[ back ] 180. Bury, ad ὅστις: “The antecedent is παῖδες implied in παιδείους”; Race: “shot...at any boy who was beautiful”; Privitera 1982:35, “per chi era bello.”
[ back ] 181. I follow the reading of line 10 suggested by Verdenius 1982:9.
[ back ] 182. For a productive analysis of this passage, see Cairns 2011.
[ back ] 183. For parallels for demonstrative τό followed by a genitive, yielding the meaning “that of...” or “that pertaining to...,” see e.g. Euripides Ion 742 and Trojan Women 43.
[ back ] 184. Verdenius 1982:10.
[ back ] 185. Contra Bury 1892:42.
[ back ] 186. Unlike for other particles, Hummel 1993:410 does not offer any explanation for ὦν in Pindar.
[ back ] 187. LSJ s.v. σέλινον I.1.
[ back ] 188. See Bonifazi 2012:141-143; Bury 1892:43 also explains αὐτῷ with reference to Xenocrates’ recent death, but rather because the wreath was intended for Xenocrates himself, but once it arrived he was no longer alive to receive it.
[ back ] 189. Pindar appears to denote Athens alternately with and without the article. The extant instances do not show a clear pattern; I have here not translated the article to avoid awkward English.
[ back ] 190. Bury 1892:44 reads καὶ τόθι as forward looking, but retains τε and says that “τ᾽ connects τόθι and ταῖς λιπαραῖς ἐν Ἀθάναις ἀραρώς.” He translates: “And, both there and in rich Athens...” He offers no parallels of this use of καί...τε in Pindar, and he appears to translate καί twice: both as a sentence connective (“And...”) and as anticipating τε (“both...”); Hummel 1993:399 remarks that και...τε only ever connects 3 items: X καί Y, Z τε.
[ back ] 191. Privitera 1982:36 adopts καὶ τότε, but reads it like me as effecting a transition. His reading is attested only in a scholion, and I see no need for the emendation (pace Privitera 1982:161 who calls καὶ τόθι “grottesco”).
[ back ] 192. Bury 1892:30-31. Both Xenocrates’ Pythian and Isthmian victories are mentioned in the 476 Olympian odes for Theron (Olympian 2 and 3), while his Athenian victory is not.
[ back ] 193. See II.4 §§58-61 about the rarity of so-called epic τε in Pindar.
[ back ] 194. See II.4 §66 about τε καί in Pindar: it is never used as a combination with sentence scope (pace Thummer 1969:II.44 and Verdenius 1982:20).
[ back ] 195. See IV.2.4.2 for this function of καί.
[ back ] 196. Bury 1892:45.
[ back ] 197. See Koier 2013:258-259.
[ back ] 198. In Pindar, δή occurs twice after a pronoun (here and in Pythian 11.17, I do not count Olympian 9.9 τὸ δή ποτε, since I believe that δή should there be taken with ποτε). In both cases it marks the introduction of a frame switch (into an embedded narrative at Pythian 11.17 and into unframed discourse at Isthmian 2.27). In Pythian 11.17 Pindar introduces the element that Arsinoe saved Orestes, which may have been an innovation by him. This could explain the intensification conveyed by δή.
[ back ] 199. As Bury 1892:45, contra Verdenius 1982:22.
[ back ] 200. Thummer 1969:II.45-46, Privitera 1982:162, and Verdenius 1982:23 believe it only refers to Xenocrates and Theron, and is purposely vague in order to imply that Xenocrates won at Olympia too.
[ back ] 201. Thummer 1969:II.46 and Verdenius 1982:24.
[ back ] 202. Bury 1892:46 construes τις with Ἑλικωνιάδων.
[ back ] 203. See II.2 §55 and II.3 §73 with note 241 for more on this function of μέν.
[ back ] 204. Thummer 1969:II.49 adduces Olympian 4.14-16 for the exact same sequence of particles.
[ back ] 205. Bury 1892:48.
[ back ] 206. Verdenius 1982:34.
[ back ] 207. Verdenius 1982:35.
[ back ] 208. Privitera 1982:166.
[ back ] 209. Compare Thummer 1969:II.53, “das μή wird aufgegliedert in μήτε...μηδέ”; Verdenius 1982:34 further adduces Denniston 1950:193 on οὔτε...οὐδέ “giving the effect of climax in the second limb.”
[ back ] 210. Bury 1892:49; other commentaries do not discuss the pronoun.
[ back ] 211. Bury 1892:49 believes that Nicomachus was a chorēgós, and is followed by Thummer 1969:II.53, and Verdenius 1982:35; Privitera 1982:165 says that he was perhaps “just a messenger.”
[ back ] 212. See also Ruijgh 1971:435, “Après tout, ἄρα a donc a peu près la même valeur que δή, particule qui est moins fréquente chez Homère.”