1. Text and Sign

Following the Ethnography of Speaking, whose descriptive focus is the socially conventional ground rules for a community’s speech practices, this first chapter of Pindar’s Verbal Art explores a question fundamental to the study of epinician style: in what context of situation did people use epinician language? [1] In other words, the point of departure for an ethnographic analysis of epinician performance is to discover the communicative context of which Pindar’s text is a record from the point of view of epinikion itself. My analysis in this chapter takes its cue from the strong precedent in the history of Pindar scholarship for treating the opening lines of Pythian 1 as a passage indicative of the context of epinician performance: [2]
Χρυσέα φόρμιγξ, Ἀπόλλωνος καὶ ἰοπλοκάμων
σύνδικον Μοισᾶν κτέανον· τᾶς ἀκούει μὲν βάσις ἀγλαΐας ἀρχά,
πείθονται δ᾽ ἀοιδοὶ σάμασιν
ἁγησιχόρων ὁπόταν προοιμίων ἀμβολὰς τεύχῃς ἐλελιζομένα.
Pythian 1.1–4
Golden lyre, Apollo’s and the violet-haired Muses’
rightful possession: a footstep [basis], the beginning of celebration, hears it,
but singers [aoidoi] heed the signs [sâmata]
whenever with your vibrating strings you strike up the prooimia that guide the dancing.
This passage has provoked much discussion about the performance of epinician song, but there is a remarkably stable consensus in the interpretation of what the lines describe and virtually no variation in the method of interpretation applied by critics. Of the two branches of literal exegesis, linguistic and historical (Todorov 1981:xxix), commentators’ interpretations of Pythian 1.1–4 privilege historical description, an interpretive methodology that, when applied to the study of epinician performance, involves producing a reconstructed scenario for performance context; such a scenario, in turn, becomes the critical touchstone for interpreting a particular passage or song. On the basis of this passage, ancient commentators reconstruct choreographic and musical arrangements, blocking out a performance scenario whose sequence of events includes a musical ante-prelude (prooimia), followed by a prelude of dance (basis ‘footstep’), succeeded by the song (thus, aoidoi ‘singers’). Modern commentaries largely take their exegetical cue from the scholia. [3] I want to bring into focus my approach to epinician performance, generally, and to Pindar’s oral poetics, specifically, by thinking about the sâmata referred to at Pythian 1.3 in light of these methodological considerations. The scholiasts’ literal exegesis of the word leads to the reasonable view that these signs are some sort of cue, but this view gives way to interpreting sâmata in terms of the reconstructed performance scenario: the ‘signs’ are the rhythm guiding music and dance or some other form of previously agreed upon guiding element, including, according to some, written instructions in the form of a performance script. [4] In this chapter I will apply descriptive and interpretive methods of linguistic anthropology to discover the features of epinician language and performance that these sâmata signal. From this methodological vantage point, we can interpret Pindar’s sâmata as sociolinguistic and traditional conventions that cue the singers’ performance of epinician song and, crucially, the audience’s participation in that performance.

Epinician Text: Status and Voicing

To address the question of whether Pindar’s extant texts are records of spoken or written communication I begin by drawing from the work of linguistic anthropologist William Hanks (1989), who identifies two key issues in textual analysis: the status of the text and voicing in the text. The status of the text has to do with “how different approaches [to textual analysis] objectify different aspects of the total textual formation” (1989:102). The prevailing position is that written composition forms the epinician text. It is possible to discover whether or not this is actually the case from the point of view of Pindar’s language by asking: what communicative functions does epinician language claim as its own? [5] Hanks refers to this kind of evidence as “functional regimentation” (1989:106–107). There is a dialogical dynamic involved here. On the one hand, a text delineates its functional domain by giving empirical clues about the context(s) and function(s) for which its author(s) composed it. On the other hand, our understanding of a text’s functional domain regiments our description and interpretation of it, so that, in a dialogical way, textual analysis is responsive to the text’s representation of its language. Voicing in the text “subsumes the relation between the textual artifact and the framework of production and participation from which it arises” (Hanks 1989:102). In Pindar’s case voicing has everything to do with performance; epinikion is a performed verbal art form (the “framework of production”) and, as such, entails at least three categories of participation: composer, chorus, and audience. I follow five steps to assemble empirical evidence for the status of and voicing in Pindar’s texts:
  1. Identify Pindar’s most common forms of metalanguage;
  2. Determine whether each instance of metalanguage refers to representing or represented communication;
  3. Identify indices for the status of the text;
  4. Identify indices for voicing in the text;
  5. Identify methodological controls applied in steps 3 and 4.

Step 1: Metalanguage

“One particularly important instance of functional regimentation,” as Hanks writes, “is metalanguage, which consists of textual elements that refer to, describe, or otherwise characterize text itself, including the very one of which they are a part” (1989:107, emphasis in original). By analyzing the most common forms of metalanguage, we can discover whether epinician language characterizes itself as spoken or written communication. The forms of Pindar’s metalanguage that I analyze include: epos ‘word’, eipein ‘to say’, en(n)epein ‘to tell’, erein ‘to say (future tense)’, humnos ‘song of praise’, humnein ‘to sing in praise of’, humnetos ‘praised in song’, logos ‘word’, logios ‘chronicler’, and legein ‘to say, to tell’. [6]

Step 2: Representing and Represented Communication

The presence of metalanguage used in representing communication that indicates the same status of and voicing in Pindar’s texts as represented communication, and vice versa, will substantiate my analysis. Drawing from the method of frame analysis, I define representing communication and represented communication in terms of how they relate to each other. [7] For present purposes, the frame is the individual epinician song, a communicative event that embraces other communicative acts, the frameworks, that constitute the frame. The frame is representing communication—for example, in mythological narrative, the addresser’s communicative stance characterized by the third-person voice. The framework is represented communication—for example, in a mythological narrative, indirect or direct discourse.
Pindar’s texts evidence two types of representing communication. The first type is a singular or plural first-person utterance that describes an act or event of communication in terms of one of the forms of metalanguage that I analyze here. Consider the following passage:
          Τυνδαρίδαις τε φιλοξείνοις ἁδεῖν καλλιπλοκάμῳ θ᾽ Ἑλένᾳ
          κλεινὰν Ἀκράγαντα γεραίρων εὔχομαι.
          Θήρωνος Ὀλυμπιονίκαν ὕμνον ὀρθώσαις ἀκαμαντοπόδων
          ἵππων ἄωτον. Μοῖσα δ᾽οὕτω ποι παρέστα μοι νεοσίγαλον εὑρόντι τρόπον
5        Δωρίῳ φωνὰν ἐναρμόξαι πεδίλῳ                                                            
          ἀγλαόκωμον· ἐπεὶ χαίταισι μὲν ζευχθέντες ἔπι στέφανοι
          πράσσοντί με τοῦτο θεόδματον χρέος,
          φόρμιγγά τε ποικιλόγαρυν καὶ βοὰν αὐλῶν ἐπέων τε θέσιν
          Αἰνησιδάμου παιδὶ συμμεῖξαι πρεπόντως, ἅ τε Πίσα με γεγωνεῖν· τᾶς ἄπο
10      θεόμοροι νίσοντ᾽ ἐπ᾽ ἀνθρώπους ἀοιδαί.
Olympian 3.1–10
To please the descendents of Tyndareus, who are kind to friends, and to please Helen, whose hair is lovely,
I pray, as I honor famous Akragas,
now that I have erected an Olympian victor’s humnos for Theron—
for horses with indefatigable feet, the best thing. The Muse so stood beside me when I found a bright new way
to harmonize the celebratory sound with the Dorian sandal,
since wreaths bound to hair
give me this divinely established duty: to fittingly blend together the multisonic lyre and the voice of pipes and a foundation of lyrics [epea]
for the son of Ainesidamos. And Pisa urges me to speak out;
from there the god-allotted songs reach humans.
In this passage there are two instances of metalanguage: humnos (3) and epea (8). I characterize this passage as representing communication because first-person singular forms (lines 2, 4, 7, and 9) signal that the words in the passage are the addresser’s; the passage is an instance of representing communication that frames instances of represented communication in Olympian 3. [8]
The second type of representing communication in Pindar’s texts is the gnomic statement, which often expresses norms about appropriate or effective communication from the point of view of epinikion. Whereas I designate words that refer to Pindar’s language as forms of metalanguage, statements that describe Pindar’s language I designate metacommunication (cf. Babcock 1977). Analyzing gnomic statements as forms of metacommunication in conjunction with specific instances of metalanguage makes it possible to discover what these gnomic statements tell us about the status of and voicing in Pindar’s texts. The following passage is an example of the gnomic statement as representing communication:
ἄλλοις δέ τις ἐτέλεσσεν ἄλλος ἀνήρ
εὐαχέα βασιλεῦσιν ὕμνον ἄποιν᾽ ἀρετᾶς.
Pythian 2.13–14
Some men requite some
kings with a sonorous humnos as a reward for virtue.
This gnomic passage characterizes representing communication as orally generated and aurally received by virtue of the adjective euakhês ‘sonorous’, which modifies humnos. Since Pindar’s gnomic statements regularly express norms about language use, the gnomic style of Pythian 2.13–14 makes such a characterization all the more characteristic.
Frequently recurring types of represented communication in Pindar’s epinikia include indirect discourse, direct discourse, and what I call, for the sake of convenience, an “other” type of represented discourse. To explain this designation, I offer an example of reported communication that is neither direct nor indirect discourse:
Λάμπων δὲ μελέταν
ἔργοις ὀπάζων Ἡσιόδου μάλα τιμᾷ τοῦτ᾽ ἔπος,
υἱοῖσί τε φράζων παραινεῖ.
Isthmian 6.66–68
Lampon, [9] by exerting effort
in his deeds, especially honors this saying [epos] of Hesiod’s;
he relates it to his sons and advises them of it.
In this passage epos (line 67) names a form of represented communication that is not direct or indirect discourse, but an “other” type of represented discourse—in this case, Hesiod’s saying, which is not “quoted” but referred to in the passage. [10]
To summarize Step 2, forms of Pindar’s metalanguage occur in the follow-ing categories and subcategories:
I. Representing Communication
          A. First-Person Statements
          B. Gnomic Statements
II. Represented Communication
          A. Direct Discourse
          B. Indirect Discourse
          C. “Other” Reported Discourse

Step 3: Status of the Epinician Text

I have recorded whether each instance of metalanguage occurs in conjunction with indications of oral production and aural reception of messages or of written communication. [11] Olympian 3.1–10, for example, contains metalanguage that indicates oral production and aural reception of messages: phôna ‘sound’ (5), poikilogarus phorminx ‘multisonic lyre’ (8); boa aulôn ‘voice of auloi’ (8); and aoidai ‘songs’ (10). On the basis of how Pindar himself describes his communicative medium, I interpret the occurrences of humnos (3) and epos (8) at Olympian 3.1–10, as instances of metalanguage that indicate the oral production and aural reception of messages, especially in light of the complete absence of any indication of written communication. I have followed the same criteria throughout my analysis of Pindar’s metalanguage.

Step 4: Voicing in the Epinician Text

Here I simply record whether each instance of metalanguage occurs in a context characterized by face-to-face interaction or by reader-text interaction. In both representing and represented communication, oral production and aural reception of communication necessarily entail face-to-face interaction and do not at all indicate a reader-text interaction. In terms of the status of and voicing in the epinician text, both song and speech contrast with writing in terms of communication and context of situation: in antiquity, song and speech entail oral production and aural reception of communication and face-to-face interaction. [12]

Step 5: Methodological Controls

First, with respect to voicing in the epinician text, I interpret forms of metalanguage in represented communication as an index of speech only when the text explicitly represents reported speech as occurring in the context of face-to-face interaction. I do not assume that represented communication entails (represented) oral production and aural reception of messages, but seek evidence for the text’s voicing in the text itself. Take the following example:
          τὸν δὲ θαρσήσαις ἀγανοῖσι λόγοις
          ὧδ᾽ ἀμείφθη· “Φαμὶ διδασκαλίαν Χίρωνος οἴσειν. ἀντρόθε γὰρ νέομαι
          πὰρ Χαρικλοῦς καὶ Φιλύρας, ἵνα Κενταύρου με κοῦραι θρέψαν ἁγναί.
          εἴκοσι δ᾽ ἐκτελέσαις ἐνιαυτοὺς οὔτε ἔργον
105    οὔτ᾽ ἔπος ἐκτράπελον κείνοισιν εἰπὼν ἱκόμαν
          οἴκαδ᾽, ἀρχαίαν κομίζων πατρὸς ἐμοῦ, βασιλευομέναν
          οὐ κατ᾽ αἶσαν, τάν ποτε Ζεὺς ὤπασεν λαγέτᾳ
          Αἰόλῳ καὶ παισὶ τιμάν.”
Pythian 4.101–108
Then taking courage, with mild words
Jason answered him [Pelias] in this way: “I say that I will produce the teaching of Khiron; for, I come from a cave,
from Khariklo and Philura, where the revered daughters of the Centaur raised me. After rounding out twenty years, and saying to those people neither a deed
nor word out of place, I have come
home, retrieving the ancient honor of my father, which is being ruled
not according to destiny—the honor that Zeus once granted to the leader of the people, Aiolos, and to his sons.”
This passage occurs in a context in which Jason responds to Pelias’ direct questions before a group of onlookers and, so, is an example of how represented communication bears indices of face-to-face interaction. Quotation marks do not suffice to indicate the status of and voicing in the text; it is the context of situation depicted in represented communication that matters.
The second analytical control that I apply is to record only explicit indices of spoken or written communication and to consider closely whether these indices are consistent in both categories of communication, representing and represented. Consider, for example, the following passages:
ἐγὼ δ᾽ Ἡρακλέος ἀντέχομαι προφρόνως
ἐν κορυφαῖς ἀρετᾶν μεγάλαις, ἀρχαῖον ὀτρύνων λόγον…
Nemean 1.33–34
I willingly embrace Herakles
among the greatest heights of virtues, beginning with the old story…

μέγα εἰπὼν σκοποῦ ἄντα τυχεῖν
ὥτ᾽ ἀπὸ τόξου ἱείς· εὔθυν᾽ ἐπὶ τοῦτον, ἄγε, Μοῖσα,
οὖρον ἐπέων
Nemean 6.26–29
I hope,
by saying something great, to hit the target straight on,
shooting, as it were, from my bow. Go, Muse, direct toward that [home]
a wind of words,
a glorious wind.
In contrast to Olympian 3.1–10, quoted above, these two passages do not have explicit indices of the mode of communication involved. I do not interpret the occurrence of logos at Nemean 1.34 as an index for the status of or voicing in Pindar’s texts because there is no evidence in the passage for the type of communication, oral or written, that this logos is. It can be argued on the basis of linguistic pragmatics that imperative forms at Nemean 6.28 suggest face-to-face interaction between a first-person addresser and second-person addressee. But again, the passage bears no explicit indication of the mode of communication. I apply these methodological controls throughout my analysis of Pindar’s metalanguage.


Whenever Pindar’s metalanguage in both representing and represented communication evidences indices of status and voicing of the epinician text, it regularly describes the mode of communication as speech and never describes it as a written mode of communication. For the status of the epinician text, the metalanguage in representing communication indicates that the performance of epinician song entails oral generation and aural reception of communication [13] and never indicates written communication. For voicing in the epinician text, the metalanguage in representing communication indicates face-to-face interaction [14] and never indicates text-reader interaction. So too in represented communication: Pindar’s metalanguage evidences spoken communication for the status of the text [15] and face-to-face interaction for voicing in the text, [16] but never evidences written communication or a text-reader interaction. We can conclude, then, that Pindar’s epinician texts have the status of being records of spoken communication and that the voicing in the epinician texts indicates that the intended context for reception is face-to-face interaction in the performance of epinician song. Again, functional regimentation is dialogical: just as the language recorded in the text is functionally regimented, that language also functionally regiments description and interpretation of the text. From this it follows that epinician language can appropriately be described in its own terms as an oral medium. I stress a final point: even if material texts and writing play a role in some phase in the process of composition and transmission of Pindar’s epinikia, the speech event of performance—not a writing event—is constitutive of epinician language. [17]

The “Oral Subterfuge” Hypothesis

Given that Pindar’s art is choral song, which implies per se that the epinician text has the status of being a record of speech (singing) and the voicing indicative of a record of face-to-face interaction, to identify Pindar as an oral poet is perhaps only a statement of the obvious. But existing studies of epinician performance do not adequately account for the fact that epinician performance is a speech event, and, as I am about to show, they in fact attribute to Pindar’s language characteristics that contradict well established views about the relationship between language and culture. To explore this problem, I introduce here the concept of markedness. Roman Jakobson’s definition is the locus classicus for this concept (1957:47):
[O]ne of two mutually opposite grammatical categories is “marked” while the other is “unmarked.” The general meaning of the marked category states the presence of a certain (whether positive or negative) property A; the general meaning of the corresponding unmarked category states nothing about the presence of A, and is used chiefly, but not exclusively, to indicate the absence of A. [18]
As linguist Linda Waugh has written, “it would seem that in the context of the history of humanity as a whole, ‘spoken’ language is the unmarked term and ‘written’ language the marked term. ‘Written’ language is more specialized in many ways than ‘spoken’ language” (1982:308). We might expect the language of Pindar’s performed works to reflect this cross-culturally observed pattern in human language use, but there is an entrenched consensus that his artistic medium is written. [19] To illustrate, and by way of further strengthening my thesis that Pindar is an oral poet, I consider in terms of markedness a serious attempt to reconcile a putatively literary Pindar with the performed quality of his epinikia, the “oral subterfuge” hypothesis.
Christopher Carey has described Pindar’s “oral subterfuge” as an “illusion created by the poet and shared by the audience” whereby Pindar “deliberately creates and sustains the impression of informal, extempore composition” (1981:5). Carey elsewhere describes the “oral subterfuge” as a “fiction” or “pretense” whereby Pindar presents already completed composition as if it were being composed in performance (1989:551–553). Here I cite Carey more fully: “This fiction is common in Pindar, who frequently speaks as though he were meditating on the contents or shape of his song prior to or during composition, whereas of course the ode is complete by the time of the performance” (1989:552). Especially since Carey describes Pindar’s songs as “written in literary ‘Doric’” (1989:562), his claim that the composition of the song would have been completed prior to performance seems to depend upon the assumption of a written mode of composition that took place prior to and apart from the context for which the composition was intended, namely performance. [20] In his study of Pindaric mimesis Andrew Miller (1993b:22) adopts Carey’s thesis
to draw attention to the central importance of Pindar’s “oral subterfuge” as a compositional technique and at the same time to demonstrate the practical utility of maintaining throughout the interpretive process a clear-cut distinction between the fictional (or at least quasi-fictional) speaker whose spontaneous utterance the poem purports to be and the hard-working professional poet who actually crafted it with care and skill. [21]
Carey and Miller segment performance into two stages: written composition, which they valorize as the truly creative act, and production of the song, which approaches interpretation as a matter, in part, of understanding how a written composition acts like a spoken one; the reality is written composition, the fiction, speech.
In terms of markedness, the “oral subterfuge” hypothesis contradicts the cross-culturally observed relationship between speech and writing in human history, suggesting that in the case of Pindar’s texts writing is unmarked and speech is marked. Existing scholarship on the question of how Pindar composed his songs generally reflects Gregory Nagy’s observations about the impulse “to define oral in terms of written” in the study of oral poetry: “if something is oral, we tend to assume a conflict with the notion of written.” [22] Just so, current understandings of epinician composition and performance upend the testimony of linguistic and anthropological evidence, so that writing is the given (i.e. unmarked) and orality is predicated upon that given (i.e. marked).
Note again that by stressing the constitutive place of speech for epinician composition and performance, I do not rule out the possibility that material texts were involved in the transmission of Pindar’s songs. It is possible to admit the hypothesis that a text form of libretto served as a mechanism for transmitting an epinician song from Pindar, say, to a production group, or for transmitting an epinician composition to contexts for reperformance, but only if it is understood that such a text entails not written art, but spoken art composed according to conventions for speaking and intended for actualization in a speech event, or, as Anna Bonifazi describes it, activation of the hic et nunc performance experience (Bonifazi 2000:82–83). From this point of view, the idea that the epinician libretto is brought to life in performance is not a fiction, but an immanent fact of the social reality upon which epinician performance bears.
Indeed, speech is so fundamentally constitutive of epinician style that, given a speech/writing opposition, we witness the plus-interpretation of speech in passages where Pindar characterizes his composition in terms of language associated with a material form of communication. To explain the designation “plus-interpretation,” recall that, from the point of view of linguistics, given an opposition between speech and writing, speech is unmarked and writing is marked. Waugh demonstrates that it is possible to interpret the opposition between speech and writing in three ways: zero-interpretation, minus-interpretation, and plus-interpretation (1982:302–306). The zero-interpretation is “that interpretation where the presence or absence of the unit of information is for the most part irrelevant”; given the speech/writing opposition, where speech is unmarked and writing is marked, the zero-interpretation of the marked member is “speech” (without regard for the marked term, “writing”). The minus-interpretation “signals the absence of the unit of information associated with the marked term”; in terms of the speech/writing opposition, the minus-interpretation is “speech, not writing” (the absence of writing is specifically signaled). Given a speech/writing opposition, indices for the status of and voicing in the epinician text occur in passages that evidence just such a minus-interpretation of that opposition; metalanguage that occurs without such indices affords a zero-interpretation of the speech/writing opposition. The plus-interpretation signals the presence of the marked term; in terms of the speech/writing opposition, the plus-interpretation is “writing.” In my analysis of Pindar’s metalanguage, the plus-interpretation of the marked term, writing, in the speech/writing opposition never occurs. Waugh points out, though, that the plus-interpretation of the unmarked term of an opposition can occur; again, in terms of the speech/writing opposition, the plus-interpretation of the unmarked term is “speech (as opposed to writing).” Bearing in mind Waugh’s caution that the plus-interpretation of the unmarked term “has to be forced by very specific contexts,” I will next demonstrate that passages in which Pindar metaphorically describes composition in terms of some material medium of communication provide just such specific contexts.
A crucial example of such a metaphorical representation of Pindar’s language is the only passage in which he refers to a skutala ‘message-stick’:
          ματρομάτωρ ἐμὰ Στυμφαλίς, εὐανθὴς Μετώπα,
85      πλάξιππον ἃ Θήβαν ἔτικτεν, τᾶς ἐρατεινὸν ὕδωρ 
          πίομαι, ἀνδράσιν αἰχματαῖσι πλέκων
          ποικίλον ὕμνον. ὄτρυνον νῦν ἑταίρους,
          Αἰνέα, πρῶτον μὲν Ἥραν Παρθενίαν κελαδῆσαι,
          γνῶναί τ᾽ ἔπειτ᾽, ἀρχαῖον ὄνειδος ἀλαθέσιν
90      λόγοις εἰ φεύγομεν, Βοιωτίαν ὗν. ἐσσὶ γὰρ ἄγγελος ὀρθός,
          ἠυκόμων σκυτάλα Μοισᾶν, γλυκὺς κρατὴρ ἀγαφθέγκτων ἀοιδᾶν·
          εἶπον δὲ μεμνᾶσθαι Συρακοσσᾶν τε καὶ Ὀρτυγίας·
          τὰν Ἱέρων καθαρῷ σκάπτῳ διέπων,
          ἄρτια μηδόμενος, φοινικόπεζαν
95      ἀμφέπει Δάματρα λευκίππου τε θυγατρὸς ἑορτάν
          καὶ Ζηνὸς Αἰτναίου κράτος. ἁδύλογοι δέ νιν
          λύραι μολπαί τε γινώσκοντι.
Olympian 6.84–97
My mother’s mother was Stymphalian, blooming Metopa,
who bore horse-driving Thebes, whose lovely water
I drink as I weave for warrior men
an elaborate humnos. Now urge our companions,
Aeneas, first to sing of Hera the Maiden
and then to know whether we escape the old reproach, “Boiotian pig,” with true
words. You are an upright messenger,
a message-stick [skutala] of the Muses, whose hair is lovely; you are a sweet drinking bowl of loudly uttered songs.
Tell them to remember Syracuse and Ortygia,
which Hieron oversees with an untainted scepter,
planning justly;
he is devoted to Demeter with her red foot, and to the festival for her daughter, whose horses are white,
and to the power of Aetnan Zeus. Sweetly worded
lyres and songs know him.
Pindar addresses an Aeneas (line 88), whom the scholia identify as the chorus leader, and metaphorically calls him “a message-stick of the Muses” (line 91). [23] Ancient commentators describe the skutalê as a Spartan practice: the message-sender inscribed a message on a leather strip that was wrapped around a stick in such a way that the edges of the strip touched and formed a writing surface; when re-wound around a stick with the appropriate diameter, the notation could be decoded by the recipient of the message. [24] The message-stick is a device that entails a material mode of communication, potentially introducing at lines 84–87 the plus-interpretation of the marked member, writing, of the speech/writing opposition. But, as I now show, the passage actually obviates such a reading by evidencing the plus-interpretation of the unmarked member (speech) of the speech/writing opposition.
The chiastic structure of the passage articulates this interpretation, if we take into account the following: first, the last sentence of the passage (line 96–97) is a gnomic statement, making it an instance of representing communication that parallels the opening lines of the passage (line 84–87), an instance of first-person representing speech; second, the composer characterizes the addressee as an angelos (line 90) and in the next lines introduces elements of the conventional epinician angelia—mention of the victor’s name and reference to his home polis. [25] With these considerations in mind, we can see that the passage has a chiastic structure that highlights song—witness humnos (line 87), “loudly uttered songs” (line 91), “sweetly worded lyres,” and “songs” (line 96–97)—and, consequently, the oral production and aural reception of messages, in the outer terms of the chiasmus:
A: Representing communication, first-person speaker (lines 84–87)
B: Aeneas as angelos (lines 87–92)
b: Angelia for Hieron (lines 93–96)
a: Representing communication, gnomic statement (lines 96–97)
Indices of spoken communication leak into the interior terms of this chiasmus as the result of a crucial parallel between the representing communication of the first-person addresser and the represented communication of the addressee: the first-person speaker characterizes himself as “weaving an elaborate humnos” (lines 86–87) and characterizes Aeneas as “a message-stick of the Muses” (line 91), pointing to the connection between the sung humnos and the model performance of the Muses, also song. Given that we must understand the first-person plural form pheugomen ‘we escape’ (90) as inclusive either of addresser and addressee or of addresser, the chorus, and addressee, the word’s occurrence in the middle of the passage near the crux of the chiasmus is yet another stitch binding together representing and represented communication, articulating still more emphatically the predominance of song and, so, speech in the passage. For further evidence that Olympian 6.84–97 is a record of spoken communication, first-person (lines 84, 86, and 90) and second-person forms (lines 87, 90, and 92) are indices of face-to-face interaction. Here, then, we have a compelling example to demonstrate that, in a context in which Pindar uses the metaphor of a material mode of communication, he applies the plus-interpretation of the unmarked term in the speech/writing opposition—namely, speech—to signal unambiguously the constitutive role of speech for this composition. [26]
In Pindar’s corpus there is another remarkable occurrence of the plus-interpretation of the unmarked term (speech) in the speech/writing opposition. It is well known that in the broader context of ancient Greece, ana-gignôskein ‘to recognize’ can refer to reading and graphein ‘to make a mark’ can refer to writing. Such forms of metalanguage occur extremely infrequently in Pindar, [27] especially in comparison to words that refer to song or music, such as aoida ‘song’, [28] aoidos ‘singer’, [29] melos ‘song’, [30] melpesthai ‘to sing’, [31] molpa ‘song’, [32] humnos, [33] and humnein. [34] The only Pindaric passage that contains occurrences of both anagignôskein and graphein bears indices of spoken communication:
Τὸν Ὀλυμπιονίκαν ἀνάγνωτέ μοι
Ἀρχεστράτου παῖδα, πόθι φρενός
ἐμᾶς γέγραπται· γλυκὺ γὰρ αὐτῷ μέλος ὀφείλων ἐπιλέλαθ᾽.
Olympian 10.1–3
Read out [anagnôte] to me the Olympian victor,
the son of Arkhestratos, [35] where
he has been inscribed in my heart; for I owe him a sweet song [melos] and have forgotten it.
The voicing in this passage indicates face-to-face interaction (not a text-reader interaction) by virtue of the second person plural imperative anagnôte ‘read out’ (line 1) and first-person singular forms (lines 1 and 3). The word melos ‘song’ (line 3) indicates the status of the text as a record of spoken, not written, communication. We can note as well that the passage is integrated as an instance of representing communication: the addresser represents the communication happening here as spoken (as always, taking song to mean also speech). As with Olympian 6.84–97, in the context of language that we might associate with a material mode of communication in Olympian 10.1–3, the plus-interpretation of the unmarked member (speech) of the speech/writing opposition surfaces, resolving any ambiguity about whether Pindar is representing his composition as spoken or as written. [36] Whereas the “oral subterfuge” hypothesis presents a theory of epinician composition predicated upon the putative given of written composition, Pindar’s theory of epinician composition is predicated upon the constitutive role that speech plays in his art, as status of and voicing in his texts and the test of markedness show. This brings us back to Pythian 1.1–4: in light of the fact that, when subject to an ethnographic analysis of metalanguage and metacommunication, Pindar’s texts represent themselves to us as records of speech, not writing, of what are the epinician sâmata signs?

Epinician Sâmata ‘Signs’

In his essay “Sêma and Noêsis: Some Illustrations” Nagy (1983) demonstrates that in Archaic Greek poetic tradition a sêma ‘sign’ serves as a key to recognition and reception of a message, both for figures within a narrative and between the composer and audience. By looking closely at the “working relationship” between sêma and noos ‘mind’, the instrument of noêsis ‘recognition’, we can draw from Nagy’s collection of comparative evidence within Archaic Greek poetic tradition to show that Pindar’s use of the word sâmata at Pythian 1.3 refers to the epinician way of speaking. [37] I will ultimately argue that Pythian 1.1–4 records an act of composition, encoding Pindaric sâmata, that is expressed in terms of recognition, the decoding of sâmata, which turn out to be conventional rules for epinician speech and performance. The following analy-sis of patterns in Pindar’s use of the words noein ‘to recognize’, noos ‘mind’, and sâma ‘sign’ supports this claim.
At Olympian 13.48 noêsai expresses Pindar’s recognition of social conventions for appropriate speech. [38] Pindar claims that he is unable to enumerate the athletic achievements of Xenophon’s (the laudandus’s) family, likening them to the number of pebbles in the sea. In the gnomic passage where the verb noêsai occurs, Pindar gives his rationale for passing over a detailed account of these achievements, then goes on to give a catalogue of the laudandus’s mythological ancestors, framing this as praise of Xenophon’s polis, Corinth:
40      ἐν δ᾽ ἀμφιάλοισι Ποτειδᾶνος τεθμοῖσιν 
          τοιοδώρῳ σὺν πατρὶ μακρότεραι
          Τερψίᾳ θ᾽ ἕψοντ᾽ Ἐριτίμῳ τ᾽ ἀοιδαί·
          ὅσσα τ᾽ ἐν Δελφοῖσιν ἀριστεύσατε
          ἠδὲ χόρτοις ἐν λέοντος, δηρίομαι πολέσιν
45      περὶ πλήθει καλῶν· ὡς μὰν σαφές 
          οὐκ ἂν εἰδείην λέγειν ποντιᾶν ψάφων ἀριθμόν.
          ἕπεται δ᾽ ἐν ἑκάστῳ
          μέτρον· νοῆσαι δὲ καιρὸς ἄριστος.
          ἐγὼ δὲ ἴδιος ἐν κοινῷ σταλείς
50      μῆτίν τε γαρύων παλαιγόνων  
          πόλεμόν τ᾽ ἐν ἡρωίαις ἀρεταῖσιν
          οὐ ψεύσομ᾽ ἀμφὶ Κορίνθῳ.
Olympian 13.40–52
In Poseidon’s festival between the seas,
longer songs will follow his [39] father, Ptoiodoros,
and Terpsias [40] and Eritimos. [41]
As many achievements as you enjoyed at Delphi
and in the lion’s fields, [42] I dispute with many
over the abundance of excellence.
I would not know how to count out accurately the number of sea pebbles.
In every matter
moderation matters. The best discretion is to recognize [noêsai] this.
As a private individual sent before the public
singing out about their ancestors’ ingenuity
and warfare among their heroic successes,
I will not speak falsely about Corinth.
Pindar’s (rhetorical) inability to list all of the achievements of Xenophon’s family results from his recognition of propriety as expressed in the gnomic statement, “In every matter moderation matters” (line 47). In the context of Pindar’s praise poetry kairos ‘timing’ can entail, as at Olympian 13.48, what Bundy (1962:73–76) has called the siga ‘silence’ motif, which is a conventional acknowledgement that excessive praise runs the risk of violating decorum or inciting jealousy. If, as Nagy demonstrates, the poetics of noêsis entails recognition, then at Olympian 13.48 Pindar represents the act of composition as recognition of social conventions concerning appropriate speech, as witnessed by his use of the verb noêsai. [43]
This relationship between communication and social rules for appropriateness is crucial from the point of view of the Ethnography of Speaking. In order to discover how and why members of a community select, organize, evaluate, and participate in the speech acts and speech events that constitute a community’s speech economy, it is necessary to understand the social rules for appropriateness that condition that selection, organization, evaluation, and participation. Bauman (1977:11) has described the relationship between performance and the social rules for appropriateness as follows: “Fundamentally, performance as a mode of spoken verbal communication consists in the assumption of responsibility to an audience for a display of communicative competence. This competence rests on the knowledge and ability to speak in socially appropriate ways.” [44] Pindar’s uses of the verb noein entail the display of such knowledge and ability. If, as Nagy writes, “a true recognition of the sign, a true nóēsis of the sêma, can be achieved only by recognizing the internally coherent system of signals” (1983:38–39), my analysis of the occurrences of noein in Pindar’s epinikia recommends that we identify that system as a social semiotics, not semiotics in the strict sense of the relationship between a sign and its referent(s), but a pragmatic (social) semiotics of the relationships among signs and their users. As an “internally coherent system of signals” the epinician way of speaking comprises a host of appropriateness rules that regiment the composition of each victory song and the audience’s participation in its performance. Pindar’s social semiotics entails recognition of the ready-made social conventions for praise poetry, as well as the display of competence in creative adaptation of those conventions. If this observation bears out, we should expect that upon further analysis Pindar’s sign language will continue to evidence recognition of, as Bundy (1962:11) describes it, the “propriety that determines the relationship between song and merit.”
Occurrences of the word noos in Pindar’s songs express just such recognition of this propriety. Consider the following example:
          ἔπεχε νῦν σκοπῷ τόξον, ἄγε θυμέ· τίνα βάλλομεν
90      ἐκ μαλθακᾶς αὖτε φρενὸς εὐκλέας ὀιστοὺς ἱέντες; ἐπί τοι 
          Ἀκράγαντι τανύσαις
          αὐδάσομαι ἐνόρκιον λόγον ἀλαθεῖ νόῳ,
          τεκεῖν μή τιν᾽ ἑκατόν γε ἐτέων πόλιν φίλοις ἄνδρα μᾶλλον
          εὐεργέταν πραπίσιν ἀφθονέστερόν τε χέρα
95      Θήρωνος. 
Olympian 2.89–95
Aim now the bow at its target—let’s go, heart. Whom do we strike
when we launch glorious arrows from our subtle mind?
Taking aim at Akragas
I will voice a claim pledged with a true mind [noos],
that for one hundred years no city has produced a man who is for his friends a greater
benefactor thanks to his thoughtfulness or more generous with his strength
than Theron.
Pindar vouches for his logos with an alathês noos, where this ‘true mind’ serves as the faculty of recognition of the praiseworthy qualities (lines 93–95) of Theron, the song’s laudandus. The representation of the composition in terms of an interior dialogue conditions Pindar’s use of noos (line 92). In this connection, witness the self-address in the form of a prayer directed to the speaker’s thumos (line 89) and the description of metaphorical arrows of song being launched from the speaker’s malthaka phrên ‘subtle mind’ (line 90). In the absence of evidence indicating otherwise, it is methodologically appropriate to interpret inductively such an interior dialogue as representing composition as a process of recognition, a process that is in progress. From this point of view Pindar’s exercise of noos to enact recognition can be interpreted as a live display of the composer-performer’s competence in an original performance, as composition-in-performance. In Olympian 1 the same dynamic occurs in a prayer addressed to composer’s êtor ‘heart’: “But take the Dorian lyre from its peg, if at all the grace of Pisa and Pherenikos put your mind [noos] under the influence of the sweetest thoughts” (ἀλλὰ Δωρίαν ἀπὸ φόρμιγγα πασσάλου / λάμβαν᾽, εἴ τί τοι Πίσας τε καὶ Φερενίκου χάρις / νόον ὑπὸ γλυκυτάταις ἔθηκε φροντίσιν, lines 17–19). The mention of Pisa is a topographical reference to the site of the Olympic Games, where the laudandus, Hieron, enjoyed his victory; Pherenikos is the name of his horse; and these features of Hieron’s athletic contest motivate recognition by the composer’s êtor. This occurrence of noos is accordingly a display of “propriety that determines the relationship between song and merit.” Like Olympian 2.89–95, Olympian 1.17–19 depicts the composition process by representing the composer as engaged in an internal dialogue, as evidenced by the word noos and the phrase ὑπὸ γλυκυτάταις…φροντίσιν ‘under the influence of the sweetest thoughts’ (line 19), about how to compose his song. [45]
In Pythian 1 Pindar describes Apollo’s reception of a prayer addressed to him in the language of recognition: “I hope that you are willing to put these things in your noos and to make the land a place for noble people” (line 40). The context of the passage shows that the referent for “these things” is a passage celebrating Hieron’s recently founded city of Aetna:
          ναυσιφορήτοις δ᾽ ἀνδράσι πρώτα χάρις
          ἐς πλόον ἀρχομένοις πομπαῖον ἐλθεῖν οὖρον· ἐοικότα γάρ
35      καὶ τελευτᾷ φερτέρου νόστου τυχεῖν. ὁ δὲ λόγος 
          ταύταις ἐπὶ συντυχίαις δόξαν φέρει
          λοιπὸν ἔσσεσθαι στεφάνοισί ν<ιν> ἵπποις τε κλυτάν
          καὶ σὺν εὐφώνοις θαλίαις ὀνυμαστάν.
          Λύκιε καὶ Δάλοι᾽ ἀνάσσων Φοῖβε Παρνασσοῦ τε κράναν Κασταλίαν φιλέων,
40      ἐθελήσαις ταῦτα νόῳ τιθέμεν εὔανδρόν τε χώραν. 
          ἐκ θεῶν γὰρ μαχαναὶ πᾶσαι βροτέαις ἀρεταῖς,
          καὶ σοφοὶ καὶ χερσὶ βιαταὶ περίγλωσσοί τ᾽ ἔφυν. ἄνδρα δ᾽ ἐγὼ κεῖνον
          αἰνῆσαι μενοινῶν ἔλπομαι
          μὴ χαλκοπάραον ἄκονθ᾽ ὡσείτ᾽ ἀγῶνος βαλεῖν ἔξω παλάμᾳ δονέων,
45      μακρὰ δὲ ῥίψαις ἀμεύσασθ᾽ ἀντίους. 
Pythian 1.33–45
The first gift for sea-traveling men
when they set out on a voyage is for a favorable wind to come; then it is likely
that at the completion of their journey they have a secure return. This statement,
given the present success, suggests
that in the future this city will be renowned for crowns and horses
and famed for its lovely sounding celebrations.
Lycian, you who rule Delos; Phoibos, you who cherish Parnassus’ Kastalian spring,
I hope that you are willing to put these things in your mind [noos] and to make the land a place for noble people;
for all devices for mortal virtues originate with the gods,
and the wise, the mighty, and those gifted at speech are born that way. And that man [Hieron], I
want to praise him and hope that,
as if in a contest, brandishing in my hand the bronze-cheeked spear, I do not throw it out of bounds,
but that by casting it far I surpass my opponents.
In this passage the application of noos and the criterion of alatheia again serve as a display of the composer’s recognition of the “propriety that deter-mines the relationship between song and merit.” In the prayer addressed to Apollo, the zeugma [46] involving the infinitive tithemen ‘to put’ (line 40) effectively links the two requests uttered by the praise poet: first, to recognize the validity of the praise communicated in lines 35–37 and, second, to actualize the doxa ‘glory’ (line 36) of Hieron’s city by making the land prosperous.
Pindar does not limit his language for representing the act of composition as a recognition of epinician social semiotics to words morphologically related to noêsis. [47] One salient example of another way in which Pindar expresses this recognition, as suggested by Olympian 1.17–19 discussed above, is to use the word êtor ‘heart’. Of the nine occurrences of êtor ‘heart’ in the Pindaric corpus, four involve workaday applications of the word. [48] The rest of Pindar’s uses of êtor operate like noein and noos: to represent the act of composition as an enactment of existing social rules for appropriate speech. In another passage of Olympian 1, for example, Pindar addresses his êtor in a form of representing communication that stages composition as an internal dialogue: “But if you wish to sing of victory prizes, my heart [êtor]…” (εἰ δ᾽ ἄεθλα γαρύεν / ἔλδεαι, φίλον ἦτορ…, lines 3–4). In addition to contextualizing the use of noos at Olympian 1.19, discussed above, this passage explicitly reflects propriety concerning the relationship between song and athletic achievement. [49]
To illustrate another pattern in his sign language, Pindar regularly uses the verb peithein ‘to persuade’ or ‘to obey’ to describe responses to specific speech acts. This pattern suggests that peithein has a special application that entails recognition on the part of the addressee that an appropriate response is recommended according to the social conventions of the context in which a speech act is uttered. To consider occurrences of peithein in narrated events (i.e. represented communication), Pindar relates the story of how Herakles brought the olive tree to Olympia, where he founded a sanctuary for Zeus, “after he persuaded [peisais] with his speech the Hypoborean people, attendants of Apollo” (δᾶμον Ὑπερβορέων πείσαις Ἀπόλλωνος θεράποντα λόγῳ, Olympian 3.16). In another passage Pindar describes the confidence accorded to the hero Aiakos: “without a summons the leading heroes who dwelt nearby willingly wished to obey [peithesthai] his commands” (ἀβοατὶ γὰρ ἡρώων ἄωτοι περιναιεταόντων / ἤθελον κείνου γε πείθεσθ᾽ ἀναξίαις ἑκόντες, Nemean 8.9–10). These examples of narrated events represent both the persuasive power of the speakers and the recognition on the part of the addressees of the propriety—i.e. effective performance—of the speech acts that the speakers utter. [50]
In Pythian 1 there is an especially remarkable occurrence of peithein in a passage of representing communication in which the first-person voice of the composer-performer addresses a prayer to the Muse using an imperative form of the verb: “Muse, obey [pitheo] me and for Deinomenes too sing a song of praise as a reward for the four-horse chariot” (Μοῖσα, καὶ πὰρ Δεινομένει κελαδῆσαι / πίθεό μοι ποινὰν τεθρίππων, lines 58–59). Here the middle voice of the verb peithein with a dative object has the meaning ‘obey’. Rather than requesting the Muse’s guidance in the composition of song, as we might expect, the composer entreats the Muse to heed him. Based upon the uses of peithein that involve the addressee’s recognition of a successfully performed speech act, the composer’s request implies that his addressee recognizes his competence as a speaker. One gauge of this competence, as we have seen, is that the passage is an exact expression of the “propriety that determines the relationship between song and merit,” to refer again to Bundy’s formulation, because it links song and merit, explicitly characterizing the song of praise as “a reward for the four-horse chariot” (Pythian 1.59). [51]
At this point I return to a last example of Pindar’s use of êtor to represent the process of composition as an act of recognition. In this passage êtor occurs as the object of the verb peithein: [52]
          ὕδατι γὰρ ἐπὶ χαλκοπύλῳ
          ψόφον ἀϊὼν Κασταλίας
          ὀρφανὸν ἀνδρῶν χορεύσιος ἦλθον
10      ἔταις ἀμαχανίαν ἀ[λ]έξων
          τεοῖσιν ἐμαῖς τε τιμ[α]ῖς·
          ἤτορι δὲ φίλῳ παῖς ἅτε ματέρι κεδνᾷ
          πειθόμενος κατέβαν στεφάνων
          καὶ θαλιᾶν τροφὸν ἄλσος Ἀ-
15      πόλλωνος.
Paean 6.7–15
At the water from the bronze gates
I heard the sound of Kastalia
bereft of men’s dancing and came
to ward off helplessness
from your kinsmen and from my honors.
Obeying [peithomenos] my heart like a child obeys his cherished mother
I came
to the nurse of crowns and festivities,
the sanctuary of Apollo’s.
If, like noos, êtor functions as the faculty of recognition and if peithein is a word that Pindar frequently uses to characterize the effect of a speech act upon an addressee, then at Paean 6.12–13 we have an especially outstanding case of composition represented as recognition. In some ways similar to Pindar’s entreaty to the Muse at Pythian 1.58–59, the phrase êtori de philô…peithomenos ‘obeying my heart’ (Paean 6.12–13) describes the persuasive power of a speaker; in this passage the speaker represents his composition as an interior dialogue, as an act of recognition-as-reception that guides composition and performance.
Now that we have seen some ways in which Pindar uses the language of recognition to describe composition, we can turn to a consideration of Pindar’s applications for the word sâma, which occurs only five times in the epinikia. [53] Corresponding to Nagy’s observation that “sêma bears not only the general meaning of ‘sign’ but also the specific meaning of ‘tomb’” (1983:45), Pindar uses sâma in just those two ways: first, to name a tomb; second, to describe the recognition of ‘signs’. In Olympian 10 Pindar selects the topic of his mythological narrative: “The proclamations of Zeus urged me to sing about the outstanding contest that Heracles founded with six altars beside the ancient funeral mound [sâma] of Pelops” (ἀγῶνα δ᾽ ἐξαίρετον ἀεῖσαι θέμιτες ὦρσαν Διός, ὃν ἀρχαίῳ σάματι πὰρ Πέλοπος / ✝βωμῷ ἑξάριθμον ἐκτίσσατο, lines 24–25). A mythological narrative in Pythian 9 describes the burial of Iolaos: “After he [Iolaos] laid waste to Eurustheus’ head with the edge of a sword, they buried him [i.e. Iolaos] below, under the earth, in the funeral mound [sâma] of the charioteer Amphitryon” (τὸν, Εὐρυσθῆος ἐπεὶ κεφαλάν / ἔπραθε φασγάνου ἀκμᾷ, κρύψαν ἔνερθ᾽ ὑπὸ γᾶν / διφρηλάτα Ἀμφιτρύωνος σάματι, lines 80–82). A gnomic statement in Nemean 7 stresses the importance of moderation: “The wise learn of the wind to come on the third day and are not harmed by the influence of success. Rich and poor to death’s funeral mound [sâma] go” (σοφοὶ δὲ μέλλοντα τριταῖον ἄνεμον / ἔμαθον, οὐδ᾽ ὑπὸ κέρδει βλάβεν· / ἀφνεὸς {τε} πενιχρός τε θανάτου παρά / σᾶμα νέονται, lines 17–20). [54]
In the two occurrences of sâma that remain, the word refers to an act of communication in which the focus is upon the reception end of the communication process—to the process of decoding sâmata. The context for the first instance of the word that I will consider is a mythological narrative in Pythian 4. The crew has just embarked upon the Argo and the seer Mopsos is prophesying with birds and lots. Jason prays to Zeus, who answers favorably with lightning flashes. [55] The Argonauts acknowledge the signs of Zeus’ thunderbolts, an omen that bears a message requiring recognition: “Heeding the signs [sâmata] of the god, the heroes took heart” (ἀμπνοὰν δ᾽ ἥρωες ἔστασαν θεοῦ σάμασιν / πιθόμενοι, Pythian 4.199–200). The reaction of the Argonauts indicates that they interpret correctly Zeus’ sâmata, which the narrative describes as “the propitious sound of a thunderbolt” (βροντᾶς αἴσιον φθέγμα, lines 197–198). The working relationship between sêma and noêsis that takes place in narrated acts of communication is a reflex of how it operates in actual performance, in the interaction between performer and audience (Nagy 1983:51). This crucially communicative process of encoding a message by way of a sêma happens, in a sense, both within poetry and without: between figures in mythological narrative within the boundaries of a given work; between composer and audience at the surface of those boundaries. This observation is another way of formulating the reflexive relationship between representing and represented communication. Just as the Argonauts must have the right interpretive framework to effect recognition of Zeus’ sâmata in Pythian 4, the audience to epinician performance must be familiar with the social conventions for performance that are constitutive of epinician poetics and that enable composition and reception.
At the beginning of this chapter, I reviewed the strong precedent in existing scholarship for approaching Pythian 1.1–4 as an outstanding record for the performance of epinikion. Where I part company with those interpretations of the passage is to apply Nagy’s discovery of the communicative and functional relationship between sêma ‘sign’ and noêsis ‘recognition’ in order to read the Pindaric sâmata at Pythian 1.3 as “signpost[s] for nóos” (Nagy 1983:50) that point the way toward the existence of a social semiotics, a set of social conventions for epinician speech and performance that are to be discovered ethnographically. In particular, I have shown that Pindar frequently uses noos to describe the poet’s competence at observing propriety concerning the relationship between praise poetry and the athletic achievement that merits it. We have further observed that Pindar regularly displays such competence in the performance of praise as an act of reception in the form of recognition. In this connection I call attention to a common feature of Pindar’s uses of sâma at Pythian 1.3 and Pythian 4.199. In both passages a form of the word sâma is the object of peithein in the sense of ‘to obey’, a word, as I have shown, that Pindar uses to describe recognition of an appropriately and effectively performed act of communication. Just as the heroes at Pythian 4.199–200 recognize Zeus’ sâmata, so the singers of Pythian 1.1–4 recognize certain sâmata. What is remarkable about the occurrence of sâmata at Pythian 1.3 is that the act of composition is described in such language of reception.
My analysis of Pindar’s sign language has focused upon how that language evidences the existence of a social semiotics, a system of conventions for speaking that Pindar and his audience share. [56] I have stressed one dimension of this social semiotics, the “propriety that determines the relationship between song and merit.” Now I will broaden the terms to suggest that the epinician way of speaking itself is constituted by social conventions for speech and performance that we can call a social semiotics, a code for the social practices entailed in the speech event of epinician performance and the speech acts that constitute performance events. The next three chapters of Pindar’s Verbal Art accordingly present a practical taxonomy for the social semiotics of epinikion, detailing what signs (i.e. social conventions) singers and audience heed in the composition and reception of epinician song.


[ back ] 1. I take the concept “context of situation” from Malinowski 1923:306. Because of its emphasis upon situated practice, “context of situation” became a theoretical and methodological rallying point for formulating an ethnographic approach to the study of texts, a concern that is central to Pindar’s Verbal Art.
[ back ] 2. Except where noted, I use Snell and Maehler 1997 for Pindar’s text. All translations are my own.
[ back ] 3. This performance scenario is the consensus view in the scholia, but ancient commentators offer different interpretations of what the key words—basis, aoidoi, and sâmata—mean (Σ Pythian 1.1-4, Drachman 1910:8–9). For Heyne basis is “the step of the chorus to the rhythm, with which begins the dancing, the ‘beginning of celebration,’ of festivity, of festal happiness. Then too the song of those very dancers hangs on the music of the lyre; [Pindar] refers to ‘singers’ instead of ‘song’” (1798:146–147). Dance and song obey the lyre, according to Boeckh, and basis refers to the entrance of dancers; aoidoi are simply the singers of the chorus (1821:226–227). Dissen 1830(II):163–164 endorses Boeckh’s interpretation of the passage. Gildersleeve reproduces the scholiasts’ understanding of the sequence of events—the lyre’s music, the dancers’ foot-step, followed by the chorus’ song: “In the first part of the poem the lyre is the organ of harmony, in the second the organ of praise. In the first part everything is plain. Apollo and the Muses are to the Greek the authors of all harmony, artistic, political, social, spiritual… Everything that owes allegiance to Zeus obeys his son Apollo, obeys the quivering of the lyre’s strings. So the foot-step of the dancer, the voice of the singer” (1890:240). According to Fennell, aoidoi refers to the chorus, but not to bards, presumably to distinguish between choral poetry and the solo Homeric aoidos ‘singer’ of tales (1893:145). Farnell gives an exceptionally imaginative reading of ambalai ‘strikings up’ (line 4) based, it seems, on vase-painting conventions: “The word was probably suggested by the singer flinging back his head to give full voice to his song” (1932:107). Burton writes: “The first illustration of the function of the lyre arises naturally from what the audience could see and hear: vv. 2–4 give in fact a vivid impression of the elements that constitute a performance of choral lyric. Βάσις [basis] is to be taken literally and signifies the actual steps of the dancers as they begin the spectacle” (1962:94). Kollmann reproduces the consensus view on Pythian 1.1–4: “Die Leir, die am Beginn des Prooimions angerufen wird (V. 1), beherrscht den Tanz und den Gesang (V. 2–4)” (1989:9). Gentili observes that choral performance is referred to in the passage and that it draws a parallel between divine celebration and the epinician celebration in honor of Hieron, the poem’s laudandus (1995:12). For other applications of literal exegesis to Pythian 1.1–4, specifically in connection with the question of epinician performance, see Heath 1988:185, Burnett 1989:286, Heath and Lefkowitz 1991:180, Carey 1991:199–200, and Morrison 2007:68–69 (note that Morrison 2007:99–100 explores how the opening of Pythian 1 anticipates context(s) of reperformance). Hubbard 1985:90–92 analyzes thematic oppositions of light/dark and harsh/delicate in the proem of Pythian 1. Herington 1985:181–183 assembles internal references to performance in Pindar’s and Bakkhulides’ songs.
[ back ] 4. For the scholiasts’ interpretations of sâmata (Pythian 1.3), see Drachman 1910:9. Kirkwood 1982:131 follows the scholiasts. Schroeder 1922:5 speculates that the sâmata are marked out by the composer. Burton 1962:94 endorses Schroeder: “As for σάμασιν [sâmasin = dative plural of sâmata] Schroeder is probably right in referring it to the various directions given by the accompanying lyre player to the chorus, such directions being marked, together with musical notation, on the performance-copy of the text; and the natural Greek word for them is σάματα [sâmata], signs which give a cue to the singers.” Kollmann 1989:54 cites Schroeder’s interpretation of sâmata. Note that the historical evidence actually contradicts the hypothesis that the sâmata are written notations on the manuscript: we know, for example, that accent marks were an innovation of Hellenistic scholars. Herington writes: “We must emphasize once more that on our present evidence the archaic and classical poetic texts, however and wherever preserved, are most unlikely to have been anything more than mere rows of capital letters without word division, without line division (except, probably, where stichich meters were concerned), without accents, with no more than a minimum of punctuation, and without musical or choreographic indications in any form” (1985:46). As a further illustration, consider Nagy 1990:29 on the role of writing in rhapsodic reperformance of Homeric song: “There is no compelling reason to believe that the medium of writing had anything to do with the traditions of the rhapsodes. In fact there is positive evidence that their mnemonic techniques were independent of writing. The textual tradition of Homeric poetry as we have it stems from Hellenistic Alexandria, where the practice of accentual notation was invented. This textual tradition bears witness to certain archaic patterns of Homeric accentuation that were no longer current in the everyday Greek language—patterns that can now be verified through the application of Indo-European linguistics. This comparative evidence leads to the conclusion that these patterns were preserved through norms of recitation inherited by rhapsodes; the factor of writing seems to be ruled out, since a textual tradition for the notation of accents was evidently lacking before Alexandrian times.”
[ back ] 5. For Goldhill, Pindar’s reflexive language works together with kleos-building and Pindar’s first-person to illuminate dimensions of the poet’s voice in Greek poetry (1991:129). García 2002 applies an ethnographic analysis of reflexive language in his study of the role of ritual speech in Homeric performance.
[ back ] 6. I do not include phâmi ‘I say’, phatis ‘speech’, and rêma ‘word’ or ‘saying’ in this analysis because they self-evidently name speech, not writing, practices.
[ back ] 7. On frame analysis see Goffman 1974 and Hanks 1993.
[ back ] 8. E.g. Herakles’ logos at line 16.
[ back ] 9. Father of Phylakidas, the laudandus of Isthmian 6.
[ back ] 10. The saying is from Hesiod Works and Days 412 (cf. Farnell 1932:362 and Kurke 1990:89).
[ back ] 11. Cf. Aloni’s view: “Melodia, ritmo, intonazione e dizione sono dunque altrettanti tratti e vincoli che sanciscono la separazione della lingua testuale da quella non testuale” (1998:23).
[ back ] 12. Of Pindar’s use of the verb garuein ‘to sing of’ at Olympian 1.3, Gerber notes that “the distinction between ‘speech’ and ‘song’ is much greater in English than in Greek” (1982:16). See Nagy 1990:17–51 for the most important discussion of the relationship between song and speech in Archaic Greek poetry, as well as Nagy 1994–1995:12.
[ back ] 13. Epos: Olympian 3.8, 9.47, 13.98; Pythian 2.66, 3.2, 3.113; Nemean 7.16, 7.104, 9.3, 9.7. Eipein: Isthmian 1.46, 4.41. En(n)epein: Nemean 7.69. Erein: Pythian 5.108. Humnos: Olympian 1.8, 2.1, 9.48, 11.4; Pythian 2.14, 3.64; Nemean 1.5, 3.65, 4.16, 7.81, 9.3; Isthmian 2.45, 7.20. Logos: Olympian 1.28, 2.92, 6.90, 7.21, 9.12, 9.35; Pythian 2.66, 6.16; Nemean 4.94, 6.30, 9.54; Isthmian 5.13. Legein: Olympian 13.12, 102; Pythian 5.108.
[ back ] 14. Epos: Olympian 13.25; Pythian 2.66; Nemean 9.3. Eipein: Olympian 6.92. En(n)epein: Nemean 7.69. Humnos: Pythian 1.60; Nemean 5.42, 7.81, 9.3; Isthmian 2.45, 4.3, 5.20, 7.20. Logos: Olympian 6.90, 9.12, 9.35; Pythian 2.66, 3.80, 5.48, 6.16; Nemean 7.51. Legein: Isthmian 5.39.
[ back ] 15. Epos: Olympian 1.86, 6.16; Pythian 4.9, 4.57, 4.138, 6.37; Nemean 2.2, 10.80; Isthmian 4.39, 6.42, 6.67. Eipein: Olympian 1.75, 6.16, 8.46, 13.71, 14.22; Pythian 3.40, 4.11, 4.86, 4.100, 4.156, 4.229, 8.43; Isthmian 6.51. En(n)epein: Olympian 1.86, 8.41; Nemean 1.69, 10.79; Isthmian 8.45. Erein: Pythian 9.51. Humnos: Isthmian 2.3, 8.60. Humnein: Nemean 5.25. Logos: Olympian 2.60, 11.5; Pythian 4.59, 9.43; Nemean 7.21, 9.6; Isthmian 5.27. Legein: Pythian 4.189.
[ back ] 16. Epos: Olympian 1.86; Pythian 4.9, 4.29, 4.57, 4.138, 4.299, 6.37; Nemean 10.80; Isthmian 6.42, 8.46. Eipein: Olympian 1.75, 4.23, 6.16, 8.46, 13.71, 14.22; Pythian 4.11, 4.86, 4.100, 4.105, 4.156, 4.229, 8.43, 9.66, 9.119; Isthmian 6.51, 6.55, 8.31. En(n)epein: Olympian 1.86, 8.41, 8.82; Pythian 4.242; Nemean 1.69, 10.79; Isthmian 8.45. Erein: Pythian 9.51. Humnos: Isthmian 2.3. Humnein: Nemean 5.25. Logos: Pythian 4.59, 4.101, 4.116, 4.128, 4.132, 4.240, 8.38, 9.43. Legein: Pythian 4.189.
[ back ] 17. This position is contra D’Alessio 2004, who argues that in performance Pindar’s songs index an original written composition. Herington 1985:45–47 describes the functions of written texts—preservation of texts in temples and family archives and the use of texts in schools—but cites two Aristophanic passages (Acharnians 383–479 and Thesmophoriazusai 25–265) as the only pre-fourth-century BCE representations of the composition process, neither of which, tellingly, figures writing. Mackie writes: “Although the victory ode would have been composed and rehearsed in advance of the performance, the performance itself was oral and public. In other words, these poems were composed to be performed orally” (2003:1). Discussing Aristophanes’ Clouds 1354–1357, where Strepsiades describes the inability of his son Pheidippides to sing a Simonidean song, Morrison 2007:21, following Nagy 1990:400–413, says that written texts would have been “the basis for the teaching and learning of the odes in schools.” Cf. Irigoin 1952:11–20 and Currie 2004:52.
[ back ] 18. See also Waugh 1982. For applications of the concept of markedness in the study of the speech of Archaic Greek poetry, see Martin 1989:29–30, Nagy 1990:5–8, and Aloni 1998:21–22.
[ back ] 19. Among many possibilities, see Irigoin 1952:5–9, Hubbard 1985:65–70, Segal 1986:153–164, Race 1990, and D’Alessio 2004.
[ back ] 20. Carey 1989 explicitly claims that Pindar composed his songs prior to performance, but is less explicit about his view of what mode of composition Pindar used.
[ back ] 21. Miller 1993a also advances the “oral subterfuge” hypothesis. See also Kurke on Pindar’s “scripted spontaneity” (1991:113 and 113n13), Carey 1995:99–103, Pfeijffer 1999a:34–37 and passim on “fictional mimesis of ex tempore speech,” Bonifazi 2000, and Burnett on the “familiar choral conceit by which well-rehearsed performers mime the spontaneity of a kômos as they begin to sing” (2005:109-110) and on “the fiction of spontaneity,” one of “the tricks of inclusion and interaction that Pindar took from the sacred choruses,” (241) exemplified in her discussion by Alkman 3 PMG.
[ back ] 22. Nagy 1990:8, emphasis in original.
[ back ] 23. Nicholson 2005:90–91 interprets this passage as a representation of the victory song as a gift, as opposed to a commodity.
[ back ] 24. See West 1988:42 for an evaluation of ancient sources on the meaning of skutalê.
[ back ] 25. Note the parallel between Pindar’s Aeneas as angelos and as skutala and Archilochus’ use of skutalê in conjunction with the vocative form Kêrukidê (recalling kêrux ‘herald’) at Fr. 185, as discussed by West 1988:46–47.
[ back ] 26. Hubbard 2004:91 interprets Olympian 6.87–91 as evidence that Pindar viewed writing as crucial to the preservation of his songs. Cf. Descat 1985:69–70.
[ back ] 27. Anaginôskein (Pindar’s variant for anagignôskein) occurs twice and graphein, three times.
[ back ] 28. Thirty-two occurrences in the epinikia (s.v. Slater 1969b).
[ back ] 29. Three occurrences in the epinikia (s.v. Slater 1969b).
[ back ] 30. Thirteen occurrences in the epinikia (s.v. Slater 1969b).
[ back ] 31. Three occurrences in the epinikia (s.v. Slater 1969b).
[ back ] 32. Three occurrences in the epinikia (s.v. Slater 1969b).
[ back ] 33. Forty-three occurrences in the epinikia (s.v. Slater 1969b).
[ back ] 34. Four occurrences in the epinikia (s.v Slater 1969b).
[ back ] 35. Father of Hagesidamos, the laudandus of Olympian 10.
[ back ] 36. Nagy interprets Olympian 10.1–3, as follows: “Thus the image of reading out loud can even serve as the metaphor for the public performance of a composition, and the image of writing, as the metaphor for the composition itself” (1990:171). Segal 1986:10 interprets the few references to writing in Pindar’s songs as an indication of the poet’s awareness of writing; if so, the plus-interpretation of the unmarked term of the speech-writing opposition foregrounds speech over and against writing. Another passage where the plus-interpretation of the unmarked member of the speech-writing opposition, speech, surfaces, is Nemean 8.46–51.
[ back ] 37. Nagy 1983:35. My discussion of epinician semiotics is heavily indebted to Foley 1999.
[ back ] 38. There are in all four occurrences of noein ‘to recognize’ in Pindar’s corpus (s.v. Slater 1969b). The occurrence of the verb at Nemean 10.86, where the verb’s sense is ‘to intend’ rather than ‘to recognize’, is necessarily an exception to the pattern that I describe here.
[ back ] 39. I.e. Xenophon’s.
[ back ] 40. According to the scholia, Ptoiodoros’ brother.
[ back ] 41. According to the scholia, Terpsias’ son or grandson.
[ back ] 42. I.e. at Nemea.
[ back ] 43. The same concern for appropriate speech is involved in the occurrence of the verb noein at Nemean 5.16–18, where Pindar uses language of noêsis to express his selection of mythological topics according to the criterion of alatheia ‘truth’. He declines to tell the story of how the heroes Peleus and Telamon killed Phokos, their half brother: “I will stop; every truth [alatheia] is not more advantageous when it starkly shows its face; to be silent is often the wisest thing for a person to recognize [noêsai]” (στάσομαι· οὔ τοι ἅπασα κερδίων / φαίνοισα πρόσωπον ἀλάθει᾽ ἀτρεκές· / καὶ τὸ σιγᾶν πολλάκις ἐστὶ σοφώτατον ἀνθρώπῳ νοῆσαι). As at Olympian 13.47–48, we see at Nemean 5.16–18 that Pindar uses the siga motif to explain why he will not relate mythic material whose content is inappropriate according to the criterion of alatheia and expresses his observance of social conventions for appropriate speech in the language of noêsis. In Paean 2 (= Fragment 52b), lines 54–55, Pindar uses the language of recognition to describe the process of composition by addressing a potential response of hearers of his song: “once it recognizes enmity, envy for those who have died long ago passes away” ([ὁ δ]᾽ ἐχθρὰ νοήσαις / ἤδη φθόνος οἴχεται τῶν πάλαι προθανόντων). The potential for praise to provoke phthonos ‘envy’ is a recurrent concern in Pindar’s songs. Here he abstractly represents phthonos itself as recognizing (noêsais, line 54) social conventions for praise, describing in the next line one such convention in the form of a gnomic statement: “It is necessary for a person to convey the profound glory due to ancestors” (χρὴ δ᾽ ἄνδρα τοκεῦσι<ν> φέρειν βαθύδοξον αἶσαν, Paean 2.56). See Bulman 1992 for a detailed study of phthonos in Pindar’s songs, as well as Kirkwood 1984 and Mackie 2003:9–37.
[ back ] 44. Cf. Hymes’s critique (1974:94–95) of Noam Chomsky’s conception of communicative competence on the basis of generative grammar’s inability to account for appropriateness understood as “whether and to what extent something is in some context suitable, effective, or the like.”
[ back ] 45. Cf. Gerber 1982:45–46. Another occurrence of noos is analogous to Pindar’s use of noein at Nemean 5.16–18, cited in note 42 above, where he similarly expresses recognition of the criterion of alatheia: “If any mortal stays the path of truth with his mind [noos], then it is necessary for him to fare well because he gets this from the blessed ones” (εἰ δὲ νόῳ τις ἔχει θνατῶν ἀλαθείας ὁδόν, χρὴ πρὸς μακάρων / τυγχάνοντ᾽ εὖ πασχέμεν, Pythian 3.103–104). Other occurrences of noos that entail the use of the word as a faculty of recognition of a message include: Olympian 9.75, 10.87; Pythian 1.95, 2.89, 3.5, 3.29, 5.44, 5.110, 5.122, 6.46, 6.51, 8.67, 10.68; Nemean 3.42, 6.5, 7.88; Isthmian 1.40, 5.61; Paean 5.45; Fragment 43.2.
[ back ] 46. As I interpret the passage, taking euandros ‘noble’ as a predicate adjective modifying khôra ‘land’. For discussion of the zeugma in this passage see Gildersleeve 1890:245, Gerber 1982:136–137, and Gentili 1995:342.
[ back ] 47. Pelliccia 1995:292–306 analyzes Pindar’s use of the words êtor ‘heart’, kear ‘heart’, thumos ‘impulse’, phrên/phrenes ‘mind’, psukhê ‘soul’, and noos as physiological or psychological organs that motivate or enable speech.
[ back ] 48. In the direct discourse of Erginos at the end of Olympian 4, whose brief mythological narrative concerns the hero’s victory in a race in armor, the hero’s words are: “Such am I when it comes to speed; hands and heart [êtor] are alike” (οὗτος ἐγὼ ταχυτᾶτι· / χεῖρες δὲ καὶ ἦτορ ἴσον, lines 24–25). In Pythian 9 Apollo invites Kheiron to witness firsthand the qualities of Kyrene: “a young woman with heart [êtor] superior to toil” (μόχθου καθύπερθε νεᾶνις / ἦτορ ἔχοισα, lines 31–32). In Nemean 8 Pindar explains that even fortitude does not secure renown without song: “Any man who is inarticulate but brave in his heart [êtor] obscurity suppresses in ruinous quarrel” (ἦ τιν᾽ ἄγλωσσον μέν, ἦτορ δ᾽ ἄλκιμον, λάθα κατέχει / ἐν λυγρῷ νείκει, lines 24–25). In Isthmian 3, Pindar characterizes the song’s laudandus in the following way: “what is due to Melissos for his twin prizes is to direct his heart [êtor] to sweet delight” (ἔστι δὲ καὶ διδύμων ἀέθλων Μελίσσῳ / μοῖρα πρὸς εὐφροσύναν τρέψαι γλυκεῖαν / ἦτορ, lines 9–11).
[ back ] 49. In Olympian 2 Pindar lists figures dwelling on the Isle of the Blessed, among whom, thanks to his mother’s intercession, is Achilles: “And his mother brought Achilles, after she persuaded the heart [êtor] of Zeus with entreaties” (Ἀχιλλέα τ᾽ ἔνεικ᾽, ἐπεί Ζηνὸς ἦτορ / λιταῖς ἔπεισε, μάτηρ, lines 79–80). In this occurrence of êtor we witness the reflexive relationship between narrated events and the act of narration. In a passage of mythological narrative Pindar represents the persuasive power of Thetis’ entreaties in terms of her ability to move Zeus’ êtor (line 79). Here I would suggest that what moves Zeus’ êtor is as much the effective performance of the speech act of entreaty (litai ‘entreaties’, line 80) on the part of Thetis and Zeus’ accordingly positive evaluation of that performance as any emotional quality that may be implied in Thetis’ unreported words. Just as Zeus recognizes Thetis’ use of appropriate speech within the framework of narrated events, so in performance Pindar uses the term êtor to express his recognition of the social semiotics of epinikion. In Nemean 4 the composer represents himself as being moved to include another topic in his song: “I am drawn in my heart [âtor] by the influence of a love charm to touch upon the feast of the new moon” (ἴυγγι δ᾽ ἕλκομαι ἆτορ νεομηνίᾳ θιγέμεν, line 35). Note that here Snell and Maehler 1997 give the Doric form âtor for êtor. Pindar attributes the identification of the topic of “the feast of the new moon” to the faculty of recognition as expressed by âtor. Note too that êtor can describe an instance of misrecognition, as at Nemean 7.20–27, where Pindar illustrates how a blind êtor is susceptible to uses of êtor that do not satisfy the criterion of alatheia. I discuss below the occurrence of êtor at Paean 6.12.
[ back ] 50. See also Olympian 2.79–80 (quoted in previous footnote) where persuasion is a consequence of Thetis’ speech act identified as litai ‘entreaties’.
[ back ] 51. Of twenty-one occurrences of forms of peithein in Pindar’s corpus, I have so far analyzed four: Olympians 2.80, 3.16, Pythian 1.59, and Nemean 8.10. Below I consider another of Pindar’s uses of peithein at Paean 6.12 and discuss two other occurrences of the verb in connection with Pindar’s [ back ] use of the word sâma (Pythian 1.3 and Pythian 4.200). I do not claim, then, that Pindar’s use of forms of peithein are characteristically linked with the poetics of sêma and noêsis; I only want to demonstrate that this connection exists in the examples that I have given. Cf. Bakkhulides 5.21, where the functionally analogous construction pisunos + dative occurs.
[ back ] 52. My reading then is contra Kurke 2005:106n77, who, following Radt 1958:117–118, writes that “it is impossible to understand these lines to mean that the speaker obeys his own heart as a child its mother; therefore we must supply Pytho as implied object of πειθόμενος [peithomenos].”
[ back ] 53. Olympian 10.24; Pythian 1.3, 4.199, 9.82; Nemean 7.20.
[ back ] 54. On this passage see also Nagy 1983:49.
[ back ] 55. Nagy 1983:43 identifies lightning as “the most ubiquitous sêma of Zeus.”
[ back ] 56. Referring specifically to deixis in Pindar’s epinikia, Bonifazi captures well this process of producing situated meaning through the use of a shared code, what I am calling a social semiotics, writing that the victory song’s “deictic system constitutes a linguistic sub-code of various shared contexts and of associated shared meanings, including the symbolic one” (2004a:413–414).