Chapter 5. The Poetics of Deixis

Deixis is what speakers do to locate themselves in space and time, with respect to things, events, and each other. When speaking, it is impossible not to be deictic, not to “be in” the context of one’s discourse. Not being deictic is not communicating, not being in a situation, not being. This is what happens in some narratives, whose narrator disappears behind the events of the story and which seem to be deploying themselves without the intervention of any speaker. Such narrative, however, is strictly a written achievement, made possible by the fictional space that writing creates. [1] In oral narrative, it is just as impossible for a narrator to disappear as it is for any speaker, and to discover the signs of that presence is, I believe, an important aspect of the study of oral traditions that have come down to us in the form of text.
This Chapter is concerned with deixis in Homer and Hesiod. In particular, I will study the deictic demonstrative οὗτος, whose use, I attempt to show, can tell us something about the way in which an audience experiences Homer’s heroic and Hesiod’s theogonic narrative as a distant reality, a reality that is nevertheless shared by poet and audience as an immediate presence in the context of performance.

Near and Far

Narrative, in any form or variety, is an expression of the universal faculty of the human mind to focus on what is not present, or not perceptible in an individual’s physical here and now. Storytelling is a matter of remembering, of imagining, of recreating things past or absent. Some form of distance, temporal and/or spatial, in fact, is a necessary condition for narrative: what is close in time and space, actually happening here and now, lacks the essential properties of narrativity. And yet no story can exist in isolation from a here and now. For one thing, there is always the medium of language, which represents in the present what happened in the past. In Homer, the verse and the oral formulas studied by Milman Parry are an important aspect of this immediacy, this physical presence, of language. [2] Moreover, the Homeric performer is not only a versifier, but also a storyteller addressing an audience. The evocation of the past is necessarily a matter of the present, an act of narrative that needs two parties to succeed.
Any narrative will thus display a mixture of past and present, of far and near. This distinction is akin to Émile Benveniste’s distinction between “narrative” and “discourse.” For Benveniste, “discourse” (discours) is characterized by such “subjective” grammatical forms as “I” and “you,” and certain markers of the deixis of space and time. By contrast, “narrative” (histoire), as an “objective” discourse mode, is characterized by the absence of those deictic features. [3] For Benveniste, the distinction is not so much a matter of whole narratives as of sentences, so that a narrative (in the wider sense) can switch to and fro between the two discourse modes. In Homer, as we shall see, the “subjective,” discursive element, the language of the immediate present, is an integral part of the narrative texture. [4] And as in other languages, there are criteria by which we can identify the moments in Homeric narrative that are communication in the present, rather than reference to the past.
Deixis in a speaker’s immediate present is typically a matter of pointing at the reality of the speech situation, visible to the speaker and his or her addressee. In Homer, we see such immediate deixis frequently in the discourse of characters, as they react to the reality surrounding them. Narrators, on the other hand, usually have to “recall” the persons and things produced by their own narrative. This is typically done with anaphoric pronouns that refer back to their “antecedents.” [5] Those “antecedents,” the characters and other objects of the narrative, are merely “language”; yet as language they are meant to be the repetition, the re-creation, of their “prototypes” in the past. To designate the difference between deixis in the present and repetition of the past, I will simply speak of an opposition between deixis and anaphora.
Languages may or may not provide separate deictic elements for these two jobs. Homeric Greek does, in the anaphoric pronoun ὁ as against the deictic pronoun οὗτος. Such a situation is not only a grammatical expression of discours and histoire, of near and far in narrative; it also provides narrators with a means to use the deictic markers of the near, when they wish to pretend that the remote reality of their story is actually present in the performance, before the listeners’ eyes. This possibility, an important grammatical aspect of what ancient literary criticism called Homeric “vividness” (enargeia), is the main focus of this chapter.

Young and Old

The narrative dimension of “near” and “far” interacts in curious ways with a well-known, and controversial, feature of Homeric poetry: the diachronic dimension of Greek epic diction. Scholars have long noted that old forms coexist alongside recent ones in Homer in ways that go beyond anything similar in ordinary language. The German term Kunstsprache has become the standard term for this peculiarity. Covering both “artistic” and “artificial,” the term by convention characterizes discourse of the Iliad and Odyssey as a special, poetic language that was sung, not spoken, and that was removed from ordinary speech. [6]
The various attempts of Homerists to explain this coexistence of old and new can serve as an instructive way to review Homeric scholarship through the ages. In the nineteenth and early twentieth centuries, scholars were mostly concerned with deducing the relative age of contexts from the relative age of the linguistic forms found in those contexts. Their aim was to establish a chronology of compositional layers on the basis of linguistic evidence provided by our text. Thus, certain scenes, books, and discourse types, have been thought, on the basis of various linguistic and stylistic criteria, to be more recent than others, and so to belong to a different stratum of Homeric composition. This approach to Homer in terms of historical linguistics bears directly on our present problem, since among the types of passages that have been selected for “lateness” are “comments,” remarks or evaluations by the poet, [7] that is, precisely the passages that I have just characterized as discourse in an immediate present, as opposed to narrative. The question thus arises whether the linguistic findings can be reconciled with narratological considerations and which one of the two approaches has to take priority in the explanation.
The “discovery” of oral poetry by Milman Parry in the 1920s and 1930s put the linguistic observations of Analysts in an entirely different light. In a major paradigm shift in Homeric studies, Parry introduced the principle of “economy.” Old and new forms, Parry argued, are metrically different; more precisely, old forms are retained because of their metrical form, as a response to the demands of oral composition. [8] They are retained, not for their own sake, but only so long as no metrically equivalent form has become available yet. [9] Thus oral-formulaic theory yielded a picture of constant diachronic pressure on the Kunstsprache, the gradual replacement of old with new, regulated by the constraints and demands of meter.
Today’s conception of oral poetry in performance differs increasingly from that of Parry. We tend to emphasize communication rather than composition, and we conceive of the epic performance as the enactment, between a performer and his audience in their here and now, of a heroic past that is distant, yet alive, and accessible through the recognized medium of the poetic tradition. In this conception, the performer’s interaction with the audience in his evocation of the past is no less important than his mastery of a formulaic, metrical idiom. It is, I believe, a major task in Homer studies today to recover, if possible, some of that interaction.
The new, performative view of Homeric poetry has important consequences for the conception of Kunstsprache and its diachronic layering. To my mind, a crucial aspect that has been overlooked or underestimated in oral-formulaic theory, is archaism, the positive importance of old forms for Homeric poetry and narrative. Sometimes it is simply good to use an old form for the reenactment of those remote, prestigious events, quite apart from the “usefulness” of this old form from the standpoint of metrics and oral composition. The presence of old forms, in other words, can reflect a narrative intention, and the process of linguistic innovation is not as blind and relentless as has sometimes been thought. [10]
The retention of old forms, be it for their metrical utility or simply for their oldness, may have side effects that bear on the semantic depth of the Homeric Kunstsprache. When what has become obsolete in the ordinary language is retained, and continues to be used alongside the newer form, the result may be a semantic choice, and hence a richness, that is unavailable in the ordinary language. The case of ὁ and οὗτος, and the grammar of narrative deixis in Homer and Hesiod, is an example of this phenomenon.

Deixis and Anaphora in Homeric Poetry

The basic facts concerning ὁ and οὗτος are straightforward and can be quickly pointed out. [11] ὁ is freely used in Homer for the purpose of anaphora in the sense introduced above, as a demonstrative that refers back to something that has just been introduced or identified in the narrative. Consider the first example of ὁ in the Iliad:
τίς τάρ σφωε θεῶν ἔριδι ξυνέηκε μάχεσθαι;
Λητοῦς καὶ Διὸς υἱός· ὁ γὰρ βασιλῆϊ χολωθεὶς
νοῦσον ἀνὰ στρατὸν ὦρσε κακήν, ὀλέκοντο δὲ λαοί,
Iliad I 8–10
Who of the gods then sent them together to fight in strife?
The son of Zeus and Leto, for he, angered at the king,
he raised bad disease among the army, and the people perished
The pronoun ὁ continues to be used in classical Greek, but its anaphoric demonstrative force is weakened to that of the definite article (by a development not unlike that of Latin ille into French le); only in some fixed combinations (ὁ μέν, ὁ δέ) does the erstwhile demonstrative retain its original force. Its anaphoric functions have been mainly taken over by οὗτος, for example:
Ἕλλησι μὲν Τεισαμενὸς Ἀντιόχου ἦν ὁ θυόμενος· οὗτος γὰρ δὴ εἵπετο τῷ στρατεύματι τούτῳ μάντις·
Hdt. 9.33.1
The priest of the Greeks was Teisamenos the son of Antiokhos. This man came with that army as a seer.
The contrast between ὁ as definite article and οὗτος as demonstrative, however, is not simply “post-Homeric,” but is found in Homeric narrative as well. In the first moments of the Iliad, for example, we have τὸν Χρύσην (I 11), τὰ δ᾿ ἄποινα (I 20), γέρων (I 33), and γεραιὸς (I 35). We might want to see some demonstrative force in these “articles,” especially in the first case, but it is clear that they are closer to classical usage than is the ὁ γὰρ of Iliad I 9. [12]
Οὗτος freely occurs in Homer as well. Its distribution, however, is highly restricted: while the demonstrative is freely used in characters’ speech, it is rarely used in narrative. [13] The difference between narrative and speech, even though it has been undervalued by Oral-Formulaic Theory, appears to be undeniable in this regard. [14] It has also been amply documented by Analyst scholars, who have found again and again that “young” features are more common in the speeches. [15] We might apply their basic question to our case: Is οὗτος younger than ὁ, thus testifying to the “lateness” of the contexts in which it is typically used? We have no reason to suppose so; but οὗτος appears to have won out over ὁ to become the unmarked demonstrative in the language for both anaphoric and deictic purposes. Homeric Greek gives us a glimpse of the earlier situation, in which οὗτος is still exclusively deictic and is as such very different from ὁ. [16] As a deictic marker, οὗτος confirms that the Homeric characters in their immediate present are different speakers from the Homeric narrator, whose primary concern is the far and distant. Thus rather than pointing to the recent date of speeches, the distribution of οὗτος suggests that Homeric language is able to meet the deictic needs of all its speakers.

Dialogic Deixis in Characters’ Speech

The pronoun ὁ, then, is anaphoric, [17] and οὗτος, in Homer, exclusively deictic. The pronoun ὁ is used, as its derivatives will be, to refer to any person or thing that the speaker cannot actually point at; from the point of view of the Homeric narrator, ὁ is thus the demonstrative of the far, of narrative. As part of the story world that is re-created in the narrative, it serves as a link to those real things and events of the past that the present narrative wants to replicate. In the following extract, for example, the demonstratives point to the reality that is (re)produced by the narrative:
ὣς οἱ μὲν τὰ πένοντο κατὰ στρατόν· οὐδ᾿ Ἀγαμέμνων
λῆγ᾿ ἔριδος τὴν πρῶτον ἐπηπείλησ᾿ Ἀχιλῆϊ,
ἀλλ᾿ ὅ γε Ταλθύβιόν τε καὶ Εὐρυβάτην προσέειπε,
τώ οἱ ἔσαν κήρυκε καὶ ὀτρηρὼ θεράποντε·
Iliad I 318–321
In this way these occupied themselves with these things around the army. But Agamemnon
did not put an end to the strife with which he had threatened Achilles in the beginning;
no, since he addressed Talthybios and Eurybates;
these two were his heralds and dedicated servants.
In the first line, ὣς, οἱ, and τὰ would have been οὕτως, οὗτοι, and ταῦτα in classical Greek. In the second line, it is customary to treat τὴν as a relative pronoun, but in the present context it is more pertinent to observe that the pronoun is fully anaphoric, building on evidence given earlier in the narrative. [18]
Οὗτος is different. Deictic rather than anaphoric, the pronoun does not repeat an earlier reality, serving as link to it. It points to what is really there, here and now. As such it is part of the system for “place deixis” in Ancient Greek. Most languages are able to convey the relative proximity of things to the speaker and his or her addressee. English has the contrast between “this” and “that”; other languages are more complex in this regard. [19] Greek has a set of three deictics: ὅδε for speaker-oriented deixis (sometimes ὅδε even refers to the speaker himself, e.g. Iliad XIX 140 and below); οὗτος for hearer-oriented deixis; and κεῖνος/ἐκεῖνος for the designation of what is more remote than the interlocutors in the current speech event. [20] Leaving aside ἐκεῖνος, I present examples of the first two deictics from the conversation between Helen and Priam on the walls of Troy. Priam asks Helen to identify certain conspicuous figures among the mass of the Greek army before their eyes:
ὥς μοι καὶ τόνδ᾿ ἄνδρα πελώριον ἐξονομήνῃς
ὅς τις ὅδ᾿ ἐστὶν Ἀχαιὸς ἀνὴρ ἠΰς τε μέγας τε.
Iliad III 166–167
… and so that you name me this man, the huge one,
who this is, this Achaian man noble and tall.
Priam has Agamemnon in mind, and he refers to him with the pronoun ὅδε: the object of pointing cannot yet be assumed to be a perception shared between him and Helen, and has to be presented as deixis from the point of view of the speaker himself. Helen answers Priam’s question with οὗτος, thereby indicating that she knows at whom Priam is pointing:
οὗτός γ᾿ Ἀτρεΐδης εὐρὺ κρείων Ἀγαμέμνων
Iliad III 178
That man there is Atreus’ son, wide-ruling Agamemnon
Unlike the Homeric narrator, Helen is dealing with a reality that is not produced by her own discourse; it exists before her, not in her speech but preceding it. With οὗτος she is actually pointing at the object of her reference, in the direct sense of “deixis.” Moreover, this pointing serves a distinct social function: Helen’s answer acknowledges Priam’s earlier perception. Helen’s and Priam’s joint seeing is in fact the very point of the use of οὗτος. We may say, then, that οὗτος is not only deictic, but also “dialogic.” The value of οὗτος as deictic of the second person is clearly brought out by the following example:
τίς δ᾿ οὗτος κατὰ νῆας ἀνὰ στρατὸν ἔρχεαι οἶος
Iliad X 82
You there, who are you walking through the ships and the army on your own?
Here we are close to the explicit linkage of οὗτος with the personal pronoun of the second person that we find in Attic drama. [21]
Thus far the difference between ὁ and οὗτος has been linked to the distinction between narrative and speech. But in the end this distinction is in and of itself not the final criterion. Before I discuss the use of οὗτος in narrative, I shall present some examples of ὁ in speech. In the following passage, for example, the two pronouns occur side by side in one context:
βάλλ᾿ οὕτως, αἴ κέν τι φόως Δαναοῖσι γένηαι
πατρί τε σῷ Τελαμῶνι, ὅ σ᾿ ἔτρεφε τυτθὸν ἐόντα,
(. . .)
τὸν καὶ τηλόθ᾿ ἐόντα ἐϋκλείης ἐπίβησον.
Iliad VIII 282–285
Go on shooting like this, so that you may become a light for the Danaans,
and to Telamon your father, who/he nourished you when you were little (. . .),
set him on the path to glory, though he is far away.
This is Agamemnon addressing Teucer; with οὕτως he refers to Teucer’s actual shooting, which is taking place right before his eyes. The reference to Teucer’s old father Telamon, on the other hand, is done with τὸν, not τοῦτον: the old man is not present; he cannot be pointed at and has to be established in Agamemnon’s speech, so as to be an antecedent for subsequent anaphoric reference. We observe, then, that the pronoun of anaphoric reference is not at all barred from the discourse of characters; it can, in fact, be quite effective in what they want to say. Consider, for example, how Agamemnon refuses to give back Chryseis to her father:
τὴν δ᾿ ἐγὼ οὐ δώσω
Iliad I 29
and her I will not give back
The choice of demonstrative, anaphoric τὴν instead of deictic ταύτην, signals that as far as Agamemnon is concerned Chryseis is not a reality shared between himself and her father. She is not present, not allowed to come out of the discourse that contains her reference, not allowed to become on object of pointing. Equally instructive is the way in which Eumaios the swineherd speaks about his absent master: after establishing him in the discourse with κεῖνος, the pronoun for remote deixis (xiv 70, 90), he refers to him with ὁ, the pronoun of anaphoric reference (xiv 133–137).

The Poet, the Muse, and the Public

The distinction between anaphora and deixis and the specific sense of οὗτος can now help us appreciate the use of this demonstrative by the Homeric narrator. Οὗτος may be less frequent in narrative than in speeches, but it does occur in narrative. We might speak of a conflation of far and near: the remote reality of the epic tale “intrudes” into the immediate reality of the performance. Instead of being “contained” within the narrative, the past becomes now the real thing, a reality before everyone’s eyes at which the poet can point. Our first example, which was already cited at the end of the previous Chapter, occurs in Book Two of the Iliad: [22]
οὗτοι ἄρ᾿ ἡγεμόνες Δαναῶν καὶ κοίρανοι ἦσαν·
Iliad II 760
These then were the leaders and the princes among the Danaans
With these words the poet refers back to a list, and such a narrative situation would seem at first sight to be one that requires simple anaphoric reference, with ὁ. But this is a special case: this is the line that recapitulates the Catalogue of Ships in Book Two of the Iliad. It is hardly necessary to emphasize the central cultural significance of the Catalogue for the poet and his Panhellenic public, as an expression of the continuing presence of the glorious past in each new performance. And I suggest that the use of οὗτοι, in combination with the “evidential” particle ἄρα (on which more in the next chapter), [23] reflects the special nature of the moment. The narrator addresses his audience directly, as if the object of reference is cut loose from the narrative, as a reality before everyone’s eyes, as if he is saying: “There you have them; those were the leaders of the Danaans.” [24]
The closing formula has to be taken in close connection with the famous invocation of the Muses at the very beginning of the Catalogue, one of the rare moments at which the narrator speaks directly of himself and his public in terms of “me” and “us”:
ἔσπετε νῦν μοι Μοῦσαι Ὀλύμπια δώματ᾿ ἔχουσαι·
ὑμεῖς γὰρ θεαί ἐστε πάρεστέ τε ἴστέ τε πάντα,
ἡμεῖς δὲ κλέος οἶον ἀκούομεν οὐδέ τι ἴδμεν·
οἵ τινες ἡγεμόνες Δαναῶν καὶ κοίρανοι ἦσαν·
Iliad II 484–487
Sing now to me, Muses, who dwell in Olympian houses—
for you are goddesses and you are present, and have seen everything;
but we are hearing only the rumor of it and know nothing—
who the leaders and lords were of the Danaans.
I would propose that this sense of “us, here and now,” the poet and the audience experiencing a reality that is due to the Muse, reverberates in the οὗτοι of II 760. [25]
The next example, too, is concerned with the poet’s human limitations. At the onset of the Battle of the Ships, he is faced with the arduous task of putting the chaotic complexity of the Trojan War into words. This is how he expresses his mortal vision:
ἄλλοι δ᾿ ἀμφ᾿ ἄλλῃσι μάχην ἐμάχοντο πύλῃσιν·
ἀργαλέον δέ με ταῦτα θεὸν ὣς πάντ᾿ ἀγορεῦσαι·
πάντῃ γὰρ περὶ τεῖχος ὀρώρει θεσπιδαὲς πῦρ
Iliad XII 175–177
Various groups of warriors fought their fight at the various gates.
Difficult it is for me to put, like a god, all these things into words,
for all around the wall the god-kindled fire had risen.
These words have met with grave suspicion on the part of ancient critics, but I see no reason to doubt their authenticity. [26] Just as earlier, in the case of the Catalogue, the situation calls for a statement of the poet’s own humanity and its shortcomings, and this would seem to explain the use of ταῦτα. The demonstrative is here not used to point at something that is present; rather, the sheer difficulty of all that was happening there creates a special human bond between the poet and his audience, a dialogic moment in which the use of ταῦτα seems fully justified. In this way the analysis of the demonstrative lends support to the judgment of the scholiast, who, defending these lines against the criticism of Aristarchus, spoke of their “Homeric enargeia.” [27]
The passage just discussed is a good point at which to turn briefly to Hesiod. We saw that the Homeric narrator is aware of his mortal shortcomings when confronted with the formidable task of recreating the past, and this human condition with its cognitive limitations calls for the aid of the Muses. The persona of Hesiod, on the other hand, appears to be much more confident. He can take the Muses’ assistance for granted, since he has been personally initiated. His discourse, in fact, is much more personal than Homer’s. The passage in which he tells of his poetic vocation contains the most explicit self-presentation of any narrator in all of Greek literature:
αἵ νύ ποθ᾿ ῾Ησίοδον καλὴν ἐδίδαξαν ἀοιδήν,
ἄρνας ποιμαίνονθ᾿ Ἑλικῶνος ὕπο ζαθέοιο.
τόνδε δέ με πρώτιστα θεαὶ πρὸς μῦθον ἔειπον,
Μοῦσαι Ὀλυμπιάδες, κοῦραι Διὸς αἰγιόχοιο·
Theogony 22–25
Now these have once taught Hesiod beautiful song,
when he was herding his sheep below most holy Helicon.
Me here the goddesses addressed at first a speech,
Olympian Muses, daughters of Zeus who holds the Aegis.
Hesiod strikingly refers to himself with ὅδε (τόνδε με) which, as we saw, is the pronoun of proximal, speaker-oriented deixis: the pronoun here designates the speaker himself as he is physically present before his audience. Hesiod goes on to say that the Muses ordered him to sing of the race of the immortal gods, thus implying their continued presence in the performance of the poet’s song.
The resulting narrative is very different from Homer’s, offering a different ratio of histoire and discours. Even though the Theogony and the Works and Days are presented in the same metrical and formulaic idiom as the Iliad and the Odyssey, the stance of their speaker is quite different. The Hesiodic performer is concerned less with reenacting the past in all its complexity, and with impersonating what gods and heroes said in that remote time, than with explaining and justifying the present. [28] What Hesiod talks about, the reign of Zeus, the honors of the gods, or the agricultural calendar, is an important part of everyday reality. Accordingly, Hesiod’s way of speaking offers more opportunities than Homer’s to use the demonstrative of dialogic deixis, even though in Hesiodic poetry there is much less direct speech than in Homer. For example, the poet closes the preview of the contents of the Theogony in the following way: [29]
ὥς τ᾿ ἄφενος δάσσαντο καὶ ὡς τιμὰς διέλοντο,
ἠδὲ καὶ ὡς τὰ πρῶτα πολύπτυχον ἔσχον Ὄλυμπον.
ταῦτά μοι ἔσπετε Μοῦσαι Ὀλύμπια δώματ᾿ ἔχουσαι
Theogony 112–114
How they divided their wealth and how they distributed their honors,
and how in the beginning they conquered many-valleyed Olympos,
sing me of those things, Muses who hold your Olympian dwellings.
The Muses know what Hesiod is talking about, so that the pronoun of dialogic deixis is called for; ταῦτα locates the story of Zeus’ ascent, the justification of the known world, as a matter of the immediate present, of the communication between the Muses and the inspired poet. Later on, too, we note that οὗτος, the pronoun of the near, is used when the specific domain and honors (μοῖρα ‘share’, τιμή ‘honor’) of a given god are specified, reflecting their continuing importance, and the god’s continuing presence. See, for example, the statement about Aphrodite: [30]
ταύτην δ᾿ ἐξ ἀρχῆς τιμὴν ἔχει ἠδὲ λέλογχε
μοῖραν ἐν ἀνθρώποισι καὶ ἀθανάτοισι θεοῖσι
Theogony 203–204
And this is the honor she has and which is her
share among men and immortal gods.
Note also the present tense (ἔχει ‘she has’), a phenomenon that does not occur in Homer, where a god’s actions in the past are at stake, not his or her position in the present. [31]
It appears, then, that οὗτος is used when attention is drawn to the poet’s own speaking in the present. This would seem to apply to the next example as well, with which we return to Homer. Οὗτος is used here to characterize the words of the misguided Trojan warrior Asios:
ὣς ἔφατ᾿, οὐδὲ Διὸς πεῖθε φρένα ταῦτ᾿ ἀγορεύων·
Ἕκτορι γάρ οἱ θυμὸς ἐβούλετο κῦδος ὀρέξαι.
Iliad XII 173–174
So he spoke, but he did not succeed in persuading Zeus’s heart by uttering these words,
for to Hektor his heart wanted rather to extend glory.
These words refer to a preposterous speech in which Asios has blamed Zeus (calling him a liar, φιλοψευδὴς) for not allowing the Trojans to eliminate Polypoites and Leonteus, who successfully resisted the Trojan onrush. We have already been explicitly told that Asios is doomed:
νήπιος, οὐδ᾿ ἄρ᾿ ἔμελλε κακὰς ὑπὸ κῆρας ἀλύξας
ἵπποισιν καὶ ὄχεσφιν ἀγαλλόμενος παρὰ νηῶν
ἂψ ἀπονοστήσειν προτὶ Ἴλιον ἠνεμόεσσαν·
Iliad XII 113–115
nēpios, he was not to escape the evil spirits of destruction,
indulging as he was in his horses and his chariot,
and to return from the ships back again to windy Ilion.
This is one of the nēpios passages, in which the poet comments from his knowing position in the future, post factum, on an epic character’s folly or ignorance. Such passages, which will concern us in more detail in the next chapter, are replete with signs of speaking in an immediate present, one of which is the evidential particle ἄρα. [32] In other words, this is discours in Benveniste’s sense, direct interaction between the poet and his audience. I suggest that the use of dialogic ταῦτα, too, has to be understood in light of the mutual agreement on this quintessential nēpios.

Indirect Speech and the Telos of the Odyssey

Asios’s words are referred to with ταῦτα, yet closed in the usual Homeric way with ὣς ἔφατ᾿, “so he spoke.” This is different from the examples from the Odyssey to which we now turn. In Book Eight, the three performances of Demodokos at the banquet in honor of Odysseus are all closed in the following way:
ταῦτ᾿ ἄρ᾿ ἀοιδὸς ἄειδε περικλυτός
Odyssey viii 83, 367, 521
this then the famous bard was singing
Let us first observe that the ταῦτα here is not a οὕτως; it applies, as direct object to a verb of saying, to the words themselves, and not to the manner in which they are presented (as does the adverbial anaphoric ὣς). Second, the ταῦτα is not a τά; the pronoun is deictic, not anaphoric. That is, Demodokos’s three performances are not the reenactment of a hero or god speaking in the past, referred to as ὣς ἔφατ᾿ (“this is the way he spoke [then and there]”). Such a speech may be the most dramatic of impersonations, but it remains encapsulated within the narrative framework; the ὥς, as demonstrative of the far, effects a repetition, a link between the hero’s speech of the past and the poet’s speech of the present, without creating the illusion that the poet’s performance is actually the hero’s speech. [33]
Such a total coincidence, I suggest, is exactly what ταῦτα effects. But then Demodokos’s performances are far from being ordinary speeches; in fact, they are not direct discourse at all. They may seem to end as speeches, but they do not begin in this way. Demodokos’s songs are explicitly introduced as well-known, recognizable songs. The first is called “the οἴμη (‘lay’) whose “fame” (κλέος), was reaching up to heaven, the quarrel of Odysseus and Peleus’ son Achilles” (viii 74–75); the second simply “on the love of Ares and Aphrodite of the gold wreath” (viii 267); and the third one, in Odysseus’s words, “the ἵππου κόσμον (‘construction of the horse’),” followed by a brief description in which Odysseus’s own role is emphasized (viii 492–495).
This means that Demodokos’s songs start, quite uncharacteristically for Homer—but see the next example—as indirect speech (ὡς ‘how’, viii 76, 268, 500). Immediately after, however, the indirect speech construction verges into a discourse mode that seems at first sight to be direct speech, with Demodokos as speaker. Yet on closer inspection, it appears that we have here in fact a curious blend of Demodokos, the poet of the past, and Homer, the poet of the present. The voice of Demodokos is allowed to intrude into the discourse of the present. The song of the blind mythical bard, performed on the rarest of occasions, with an eyewitness in the audience whose scrutiny it withstood, is in fact appropriated by the Odyssey of the present; it becomes the Odyssey. Its reality is now nothing less than the reality of the Homeric performance itself. The immediate deixis of ταῦτα, then, bridges the gap with the past; far from merely repeating the earlier speech, it pulls that speech into the present, while at the same time characterizing the song of the present as just as good as the song of the past.
Demodokos’s third song, on the wooden horse and the sack of Troy, remains more overtly direct speech than the other two songs: by creating grammatical distance between himself and the story, the narrator of the present is more fully in control of the song of the past. [34] Furthermore, the lay of the sack of Troy has been specifically commissioned by Odysseus, its principal hero, by one of the neatest narrative moves in the Odyssey: Odysseus’s own telling of his adventures, which is to follow shortly, is immediately preceded by the bard’s telling of what immediately precedes the events of Odysseus’s narrative. This creates an almost seamless transition between the stories of the bard and of the hero. Nor is this the end of the similarity between the two performances. Just as Demodokos’s song is appropriated by the Odyssey, so is Odysseus’s. Yet this happens much later, on the night of the reunion of Penelope and Odysseus, when Odysseus tells once more the story of his wanderings. Or does he? Consider the way in which the poet closes Odysseus’s story:
τοῦτ᾿ ἄρα δεύτατον εἶπεν ἔπος, ὅτε οἱ γλυκὺς ὕπνος
λυσιμελὴς ἐπόρουσε, λύων μελεδήματα θυμοῦ.
Odyssey xxiii 342–343
this then was the last word he spoke to her, when the sweet sleep
that loosens limbs sprang to him, releasing the anxieties from his spirit.
This is, of course, exactly the way in which Demodokos’s songs are recapitulated. Odysseus’s narrative, like the stories of the Phaeacian bard, begin as direct speech (ἄρξατο δ᾿ ὡς πρῶτον [“and he started how first…”] xxiii 310); and just as Demodokos’s third song, it remains indirectly reported: all along, the formula ἠδ᾿ ὡς (“and how”) is used, keeping Odysseus’s narrative firmly in control of the Homeric narrator. Odysseus’s narrative is not allowed, this time at least, to become a performance, not even one he shares with the narrator. At the night of the reunion of the hero with his wife, the principal moment of nostos and the ultimate telos of the Odyssey, the poem reasserts its own authority. Homer, the speaker of the Odyssey, the speaker of the present, finally puts his stamp on the bewildering chaos of voices and performances that has preceded. [35] Indeed, it is the Odyssey itself that is here summarizing itself, in the same way that it had summarized and appropriated Demodokos’s mythical song. The poem of the present is obviously an epos that makes up the present experience of the poet and his audience. Accordingly, I characterize the use of οὗτος here not as a reference to the past, but as a pointing, a being in the present.

Spoken Thoughts, Thought Speech

The remaining examples of οὗτος in Homeric narrative take us from represented speech to represented thought. In what ways does Homeric poetry represent the thoughts of its heroes or gods? The best-known strategy is the hero addressing his own thumos: the hero’s thoughts are performed on the Homeric stage, and there is no formal difference from overt speech presented to another person, for example:
ὀχθήσας δ᾿ ἄρα εἶπε πρὸς ὃν μεγαλήτορα θυμόν· (. . .)
ὣς εἰπὼν
Iliad XXI 552–571
And deeply annoyed he spoke to his own great-hearted spirit (. . .).
Having spoken thus….
Just as in the case of direct speech, the anaphoric adverbial demonstrative ὥς is used, marking the speech as a performance pointing to the past.
Yet there are other strategies. A character’s inner thoughts can also be represented by means of indirect discourse, governed by such verbs as ὁρμαίνειν or μερμηρίζειν (‘ponder’, ‘deliberate’). Such a representation often involves disjunctions of the form ἦ . . . ἠέ (‘whether … or’) and its end can be marked in two ways in the narrative. The first is the simple report on the character’s decision, for example: [36]
στῆ δ᾿ ἄντα σχομένη· ὁ δὲ μερμήριξεν Ὀδυσσεύς,
ἢ γούνων λίσσοιτο λαβὼν εὐώπιδα κούρην,
ἦ αὔτως ἐπέεσσιν ἀποσταδὰ μειλιχίοισι
λίσσοιτ᾿, εἰ δείξειε πόλιν καὶ εἵματα δοίη.
ὣς ἄρα οἱ φρονέοντι δοάσσατο κέρδιον εἶναι,
λίσσεσθαι ἐπέεσσιν ἀποσταδὰ μειλιχίοισι,
Odyssey vi 141–146
And she stood opposite him, firm; and he pondered, Odysseus,
whether clasping her knees to supplicate the fair-eyed girl,
or standing there from a distance with gentle words
to supplicate her, hoping she would show him the city, and give him clothing.
This way then as he was thinking it seemed most profitable to him,
to supplicate there from a distance with gentle words.
The narrator closes his rendering of Odysseus’s thoughts with a simple report on the outcome: “he considered A and B, and he decided B.” Again, the anaphoric ὥς is used. Notice, however, that ταῦτ᾿ ἄρα οἱ φρονέοντι would have been metrically possible, a phrase that actually occurs elsewhere (see the next example). In choosing the anaphoric adverbial demonstrative ὥς instead of the deictic accusative ταῦτα, the narrator contents himself with repeating the thought process as such, without allowing it to break through the narrative framework to become part of the immediate reality of the performance.
Consider now the second possibility. The character’s thought processes may also be interrupted by the sudden appearance of a god who determines his line of action. When this happens, the indirectly reported thought is referred to with the pronoun of dialogic deixis: ὥς recedes in favor of ταῦτα. Consider for example:
Ἕκτωρ δ᾿ ἐν Σκαιῇσι πύλῃς ἔχε μώνυχας ἵππους·
δίζε γὰρ ἠὲ μάχοιτο κατὰ κλόνον αὖτις ἐλάσσας,
ἦ λαοὺς ἐς τεῖχος ὁμοκλήσειεν ἀλῆναι.
ταῦτ᾿ ἄρα οἱ φρονέοντι παρίστατο Φοῖβος Ἀπόλλων
Iliad XVI 712–715
But Hektor, in the Skaian Gates he kept his single-foot horses,
and wondered whether he should drive back and fight in the carnage,
or call out to his people to assemble inside the wall.
As he was pondering these things Phoibos Apollo came and stood by him.
Let us first observe that Apollo’s true identity is a matter between the narrator and his audience: Hektor himself is not aware of it. Second, Hektor’s deliberation has potential consequences that transcend the framework of the narrative: had the second alternative, retreating to the city walls, prevailed in his mind, there would have been no encounter with Patroklos, and the course of events would have been very different from the received story. Apollo’s intrusion is thus crucial for the action of the Iliad. The deictic pronoun ταῦτ᾿(α) objectifies Hektor’s thought, and sets it up for the confrontation with Apollo, who as an external agent decides the outcome. The following example, from the Quarrel of Achilles and Agamemnon, is similar:
ὣς φάτο· Πηλεΐωνι δ᾿ ἄχος γένετ᾿, ἐν δέ οἱ ἦτορ
στήθεσσιν λασίοισι διάνδιχα μερμήριξεν,
ἢ ὅ γε φάσγανον ὀξὺ ἐρυσσάμενος παρὰ μηροῦ
τοὺς μὲν ἀναστήσειεν, ὁ δ᾿ Ἀτρεΐδην ἐναρίζοι,
ἦε χόλον παύσειεν ἐρητύσειέ τε θυμόν.
ἧος ὁ ταῦθ᾿ ὥρμαινε κατὰ φρένα καὶ κατὰ θυμόν,
ἕλκετο δ᾿ ἐκ κολεοῖο μέγα ξίφος, ἦλθε δ᾿ Ἀθήνη . . .
Iliad I 188–194
So he spoke. And grief came on Peleus’ son, and his heart
within his shaggy breast pondered in two separate ways,
whether he should draw his sharp sword from beside his thigh,
make all of them stand up, and kill the son of Atreus,
or else put and end to anger and restrain his spirit.
As he considered these things in his mind and in his spirit,
and started drawing from its scabbard his mighty sword, Athene descended
As in the previous example, Achilles’ deliberations, known and told by the narrator, [37] are not a process leading up to a decision, but specify what might have happened had Athene not intervened. The possibility of Achilles killing Agamemnon is not allowed to become narrative; it remains, in the form of an objectification by means of tauta, a matter of the present, of the communication between the narrator and the audience. [38]
I suggest that in these two cases ταῦτα, the demonstrative of the near, is used to underscore the vividness inherent in these situations: the character’s inner thought, potentially resulting in a line of action that would diverge from the Iliad, is objectified, and presented by means of οὗτος as something you can point at. This marked use of οὗτος is consistent with the poetics of the Iliad, where the verbalization of what could have happened, but did not, is a recurrent theme: the well-known “if not” situations speak for themselves in this regard. [39]
The remaining instances of οὗτος in Homeric narrative are all part of the same formulaic line that served, as we saw, as link between Achilles’ thoughts and Athena’s intervention: [40]
εἷος ὁ ταῦθ᾿ ὥρμαινε κατὰ φρένα καὶ κατὰ θυμόν.
When he considered these things in his mind and in his spirit
One of the instances of this formula (X 507) involves a divine epiphany, but the other cases are different. They do not involve a divine intervention that is responsible for the action of the poem, nor is the thought presented as indirect speech. All but one of the instances of this formula follow thought presented as direct speech, a hero addressing his own thumós, in the line we have already seen:
ὀχθήσας δ᾿ ἄρα εἶπε πρὸς ὃν μεγαλήτορα θυμόν·
and deeply annoyed he spoke to his own great-hearted spirit
Does this invalidate the analysis of ταῦτα just presented? Before we conclude that this is the case, we have to take into account that ταῦτα is prepackaged in a formulaic context. In other words, it is possible that the use of ταῦτα is conditioned by the formula. A quick search, in fact, reveals that the verb of the formula, ὥρμαινε, is not used according to its basic lexical value: like μερμηρίζειν, ὁρμαίνειν is used primarily for inner thought, not public speech. Both verbs are often used for (hidden) intentions, with direct objects such as φόνον (‘murder’) or δόλον (‘ruse’), or they govern a purpose clause. [41] The use of the formulaic line in question derives from the ambiguous status of speech addressed to one’s own thumós: such “performed thought” can either be categorized as overt, direct speech (hence the anaphoric reference with ὥς, see above), or as inner thought, to be referred to with the verb ὁρμαίνειν. This means that the formulaic line containing ταῦτα is used in a “grey” area where categorization can go either way, outside the proper locus that originally generated its existence. This does not necessarily mean that these instances are more recent or less “original.” Rather, we would have the more or less routine use of an expression in contexts that are similar to the one for which it was originally designed. This is a frequent phenomenon, not only in the deployment of the formulaic idiom of an oral epic tradition, but also in language in general. [42]
With this formulaic connection, οὗτος is a truly Homeric element. More important than formulas and their relative age, however, is the function of the demonstrative as part of the Homeric grammar of deixis, serving a fundamental goal of the epic tradition and its performers: the vivid representation of a heroic past that is alive thanks to the power of words that can reveal its presence.


[ back ] 1. On fictionality, see also Chapter Three above (p. #). On deixis in Greek poetry, see recently Felson, ed. 2004:253–266.
[ back ] 2. Compare my efforts (e.g. Bakker 1993a, 1997a, 1997c) to study Homeric discourse from the point of view of spoken language. See also Chapters Three and Four above.
[ back ] 3. Benveniste 1966:238–239 (histoire), 241–242 (discours). Benveniste speaks primarily of the distribution of tenses in French, whereby the passé simple is confined to histoire, the passé composé being the tense of discours. More on Benveniste’s conception in Chapter Seven below, in connection with verbal augment.
[ back ] 4. See also de Jong 1987:41–99, although her literary point of view is different from my performance-oriented approach.
[ back ] 5. For an account of anaphora in these terms, see Lyons 1977:660.
[ back ] 6. On the diachronic and dialectal heterogeneity of Homeric diction, see recently Forssman 1991; Meier-Brügger 1986; Ruijgh 1995; Shipp 1972; Hackstein 2002.
[ back ] 7. The best-known formulation of this observation is Shipp 1972, who includes (p. 3–4) nēpios comments, which will briefly occupy us later on in this Chapter and in more detail in Chapter Six below.
[ back ] 8. E.g. Parry 1971:332: “by the constraint of his technique of epic verse-making, the singer keeps the formula though its language has become archaic.”
[ back ] 9. See e.g. Horrocks 1981:152–53 on the usefulness of tmesis (shown to be already an archaism by the time of the formulaic diction; see also Janko 1992:11n15, 17). From a slightly different angle, linguistic modernization has been studied as a factor beneficial to the flexibility of epic formulaic diction (Hoekstra 1965), or as a window on the relative chronology of the various works of early Greek hexameter poetry (Janko 1982).
[ back ] 10. For a new account on Homeric Kunstsprache, see Hackstein 2002, who stresses the contemporary Ionic character of Homeric diction and allows for deliberate archaizing.
[ back ] 11. See Monro 1891:217; Kühner and Gerth 1898–1904: 1:575–581, 641–651; Chantraine 1963:158–169.
[ back ] 12. Still, in many cases the “article” is more marked than in Attic Greek. For example, Kirk 1985:145 is probably right in seeing contrastive force in II 278 ἀνὰ δ᾿ πτολίπορθος Ὀδυσσεύς  ἔστη σκῆπτρον ἔχων, though I would see the noun-epithet formula as apposition to ὁ, rather than as noun phrase modified by a definite article; see also VIII 532; X 363, 563; XX 320. On apposition of nouns or noun-epithet formulas to ὁ, see Bakker 1997a:92–93, 198–200.
[ back ] 13. In making this observation I am indebted to Ruijgh 1991.
[ back ] 14. Note the reactions from the literary camp, e.g. Griffin 1986, De Jong 1987:151–60; 231–33. For a renewed interest in speeches within the framework of orality, see Martin 1989.
[ back ] 15. See e.g. Drewitt 1908; 1912:117–118, using verbal augment (on which, see Chapter Seven below), the infinitive ending ειν, and elision in the caesura as criteria. Note that for the same reasons, scansion and prosody, speeches have more recently been found, ironically, to be more archaic than narrative (Kelly 1990:20–27).
[ back ] 16. Οὗτος is usually derived from the coalescence of the pronominal root *so/ho with a particle *u (related to Greek αὖ?) and the flectional forms of the pronominal root (-to); see Meillet-Vendryes 1948:494; Rix 1976:184; Klein 1996:35. This could mean that the composite form is less “basic” than the original, but it could also be argued that the formation is Proto-Indo-European, and is hence of no relevance for the study of relative chronology in Greek.
[ back ] 17. See already Monro 1891:224.
[ back ] 18. In this connection we can also observe that it is usually ὁ, and not οὗτος, that figures in correlative constructions (e.g. Il. II 36 τὰ φρονέοντ᾿ ἀνὰ θυμὸν ῥ᾿ οὐ τελέεσθαι ἔμελλον). The reason seems to be that in correlative constructions the demonstrative specifies what is dealt with elsewhere in the utterance, and so has no immediate deictic force. This principle is independent of the distinction between narrator and character. See e.g. I 125, 554; II 38; IV 361; XVIII 4; i 257; ii 116; v 188.
[ back ] 19. See Fillmore 1997:64–66.
[ back ] 20. See also Klein 1996:26–27.
[ back ] 21. E.g. Aristophanes, Acharnians 564; Sophocles, Oedipus Tyrannus 532; Euripides, Orestes 1567. For further details, see Dickey 1996:154–158.
[ back ] 22. This example is also discussed by Ruijgh (1991), who speaks of an “expressive” use of οὗτος.
[ back ] 23. See also Bakker 1993a:15–25; 1997b:17–20.
[ back ] 24. Compare the very similar use of οὗτος ἄρα much later in the narrative: οὗτοι ἄρ᾿ ἡγεμόνες Δαναῶν ἕλον ἄνδρα ἕκαστος (‘so these, lords of the Danaans, they killed each his own man’, XVI 351). The narrative section to which this statement refers has clear catalogic properties (on the special relation between battle narrative and catalogues, see Beye 1964). The special reality of the “list” just presented seems motivated by the special importance of this turning point in the battle, the prelude to Patroklos’s death.
[ back ] 25. Note that the Muse is addressed again in the very next line (II 761).
[ back ] 26. On ancient criticism of XII 175–178, see Hainsworth 1993:336–337. Aristarchus’s criticism of the premature mentioning of fire (which is not actually thrown into the ships, in spite of numerous attempts, until XVI 122) is unfounded: the passage is programmatic for the whole Battle of the Ships all the way to Patroklos’s appearance in the battle in Book XVI. Such orienting statements may have features that are strictly speaking not chronological (note that the pluperfect ὀρώρει is frequently used as a descriptive, non-sequential verb in previewing descriptions, e.g. IV 449, XII 289, XIII 540, XVII 384). On problems with our notion of narrative chronology for the Battle of the Ships in general, see Whitman and Scodel 1981; on orientation and previewing in general, see Bakker 1997a:86–122. At pp. 57–58 I discuss the passage in question as a confrontation of the linearity of speech with the spatial, simultaneous complexity of the battle.
[ back ] 27. Scholion ad XII 175–181: ἄλλως τε καὶ Ὁμηρικὴν ἐνάργειαν ἔχουσιν οἱ στίχοι (“and besides the lines have a Homeric vividness”).
[ back ] 28. See also Nagy 1992:125–126; Clay 2003:2, 57.
[ back ] 29. See also Hesiod, Theogony 75 ταῦτ᾿ ἄρα Μοῦσαι ἄειδον, where the demonstrative refers to the Muses’ song of Zeus’ victory over Kronos and the establishment of his reign. In Works and Days, too, οὗτος is used in cases of direct address: 27, 274 (Perses), 263 (the greedy kings).
[ back ] 30. See also Hesiod, Theogony 348, 520.
[ back ] 31. On present tense, see further Chapters 6–9 below.
[ back ] 32. On ἄρα as well as on the evidential verb μέλλειν, see Chapter Six below. See also Bakker 1997b:17–23; 30–36, 1998a:71–73. On nēpios passages, see also de Jong 1987:86–87. Note that these passages, too, have been considered “late” (Shipp 1972:3–4).
[ back ] 33. See also Chapter Six below, p. #, where I characterize this use of ὥς as “a sign pointing to the past.”
[ back ] 34. Note the repetition of the embedding construction: ἤειδεν δ᾿ ὡς . . . ‘and he sang how…’ (viii 514).
[ back ] 35. In defending the passage against the attacks of ancient and modern critics, Heubeck 1992:346 calls the passage a “retrospection,” and makes a comparison with I 365–392, where Achilles recounts the events of the Quarrel to his mother. The two passages, however, are not comparable. Achilles’ account, a first-person mirror story (de Jong 1985b), gives a character’s personal vision of previously recounted events, whereas the bedroom narrative recapitulates Odysseus’s “objective” story of his wanderings, guaranteeing its truth.
[ back ] 36. See also II 5; V 671–675; VIII 169; XIII 455–459; XIV 159–162; XVI 647–655; x 50–53, 151–155, 438–442.
[ back ] 37. Note the anaphoric pronouns τοὺς and ὁ in line 191: the pronouns suggest that Achilles’ thought is presented from the narrator’s standpoint.
[ back ] 38. A third case in which a god intervenes in a character’s (Diomedes) thought processes is X 507. The same ὥρμαινε formula is used, though the importance of the god’s intervention is less crucial.
[ back ] 39. See Bakker 1997a:178–180.
[ back ] 40. X 507; XI 411; XVII 106; XVIII 15; iv 120; v 365, 424.
[ back ] 41. φόνον: ii 325; iv 843; xix 252; δόλον: ii 93; xxiv 128; purpose clause: XXIV 681; ix 554; xv 169; xx 29. The two verbs are functional synonyms, as appears from the fact that the plural and inflected forms of the formula πολλὰ φρεσὶ μερμηρίζων || i 427; xx 10) feature the participial forms of the other verb: πολλὰ φρεσὶν ὁρμαίνοντα (X 4; cf. iii 151; iv 843), φρεσὶν ὁρμαίνοντι (XVI 435).
[ back ] 42. On the semantic “catachresis” of formulas, see Bakker 1988a:186–195; 1997a:190–195. This issue touches on the “improper” use of epithets, signaled by Parry 1971:151–152; see also Chapter One, n.29 above.