The concept of productive ambiguity with some examples from Iliad 2

O conceito de ambiguidade produtiva com alguns exemplos da Ilíada 2

Alejandro Abritta
Universidad de Buenos Aires, Argentina

The concept of productive ambiguity with some examples from Iliad 2

Classica - Revista Brasileira de Estudos Clássicos, vol. 35, núm. 2, pp. 1-21, 2022

Sociedade Brasileira de Estudos Clássicos

Recepción: 07 Febrero 2022

Aprobación: 07 Abril 2022

Abstract: The goal of this paper is to introduce and elucidate the concept of “productive ambiguity” for the study of Homeric poetry and other literary texts. After the introduction, I present a theoretical elucidation of the concept, starting from a general notion of ambiguity and identifying three features of productive ambiguity: its irresolvable character (no alternative can be ruled out on textual or linguistic grounds), its persistence (both alternatives are appropriate to the context and contribute to the interpretation of the text), and its productivity (the ambiguity itself contributes to the interpretation of the text). In the third section, I analyze four passages from Iliad 2: 2.73, 285, 340-9, and 807, studying in each the source of the ambiguity and demonstrating that it fulfills the three features to be considered productive.

Keywords: Homer, Book 2, ambiguity.

Resumo: O objetivo deste artigo é apresentar e elucidar o conceito de “ambiguidade produtiva” para o estudo da poesia homérica e outros textos literários. Após a introdução, apresento uma elucidação teórica do conceito, partindo de uma noção geral de ambiguidade e identificando três características da ambiguidade produtiva: seu caráter insolúvel (nenhuma alternativa pode ser descartada por motivos textuais ou linguísticos), sua persistência (ambas as alternativas são adequadas ao contexto e contribuem para a interpretação do texto) e sua produtividade (a própria ambiguidade contribui para a interpretação do texto). Na terceira seção, analiso quatro passagens da Ilíada 2: 2.73, 285, 340-9 e 807, estudando em cada uma a origem da ambiguidade e demonstrando que ela cumpre as três características para ser considerada produtiva.

Palavras-chave: Homero, Livro 2, ambiguidade.

At the end of the movie Inception (Christopher Nolan, 2010), the script describes this final shot: “…on the table, the spinning top is STILL SPINNING. And we – FADE OUT.”1 As we see the movie, the spinning of the spinning top becomes a symbol of being trapped in a dream, so the open ending leaves us wondering if everything we have seen in the final scenes (or even the whole movie) has actually happened. While many viewers would (and have) defend(ed) one or the other interpretation of the scene, the movie gives us no evidence to definitively answer the question. The ambiguity of the events is, therefore, key to our understanding and enjoyment of the movie.

This use of ambiguity is common in ancient literature as well, but most critics seem to apply an intuitive approach to the matter. Il. 2.222-3, in the middle of the episode of Thersites, can serve as an example of this. There, the poet sings τῶι δ’ ἄρ’ Ἀχαιοί / ἐκπάγλως κοτέοντο νεμέσσηθέν τ’ ἐνὶ θυμῶι (“and at him, naturally, the Achaeans / were greatly angered and resentful in spirit”).2 Given the context, one may assume that the reference is to the ugliest of the Achaeans, and this is certainly a possibility. However, the previous sentence complicates matters: τότ’ αὖτ’ Αγαμέμνονι δίωι / ὀχέα κεκληγὼς λέγ’ ὀνείδεα (“But now it was against godlike Agamemnon / he noisily gave his litany of shrill abuse”). There are two obvious approaches to this issue: either to consider the ambiguity a literary resource or to consider it an interpretative problem. Since Leaf (1900, adloc), who claimed “τῶι is clearly Agamemnon,” scholars have assumed the second (the majority believing, however, that Thersites is a more natural referent).3

The goal of this paper is to establish a methodological approach to ambiguity to help in the identification of cases in which they function as literary resources. In those cases, ambiguities are like the one at the end of Inception, not only irresolvable but such irresolvability actually provides profitable grounds for interpreting the poem further. In the following section, I will introduce and elucidate the concept of “productive ambiguity” for this type of ambiguities and delimitate its extent with several counterexamples. Section 3 will then show its advantages in the analysis of four passages in Book 2 in which it applies, and, finally, section 4 will summarize the argument.

Defining productive ambiguity4

Ambiguity can be defined as the property of a statement to be understood in two or more possible ways.5 In any verbal utterance in any language, there is a potential for ambiguity since there are many things that can go wrong in the transmission of information from the sender to the receiver(s). In literature, where, for the most part, there is no possible feedback between the actors involved in the communicative act,6 this potential becomes an interpretative problem. Simply stated, the receiver cannot ask for clarification and any miscommunication must be resolved by studying the transmitted text further.

In the case of ancient literature, and literature from foreign, inaccessible cultures in general, we can distinguish two types of ambiguity, which I will call “intrinsic” and “exogenous.”7 In the second case, ambiguity results from our ignorance of some aspect of the original context or language. For example, since we cannot ask an Ancient Greek speaker what νῶροψ means in Il. 2.578; 11.16 and the other places where it appears, we cannot know whether we should translate it as “flashing,” “bright,” “resounding,” “blinding,” or something else,8 even if we can assume that both the poets and the audiences that used and heard the word knew (at least to some extent) what it meant.

The second type of ambiguity is the one mentioned above, in which there is something in the text that is difficult to understand even for contemporary listeners or readers. In Il. 16.355, for example, the poet sings that wolves διαρπάζουσιν [ὄϊας] ἀνάλκιδα θυμὸν ἐχούσας. Here, we cannot tell whether the phrase is an epithet of all sheep (i.e. “sheep, which have a feeble heart”) or a specific descriptor of those sheep that the wolves snatch (i.e. “those sheep that had a feeble heart”). In this case, however, it is not our ignorance of the Ancient Greek language or culture that causes the ambiguity, and we can assume with very little risk that even contemporary audiences might have doubted which interpretation was better.9

Of course, it is not always easy to differentiate intrinsic ambiguity from exogenous ambiguity. The meaning of μέροψ in the formula μέροπες ἄνθρωποι, for example, was probably lost very early, maybe even already to the Homeric poet, given its very restrictive use in the poems, the fact that later poets use it only as a synonym to “mortals,” and its impossible explanation in Alexandrian times as a compound of μείρομαι and ὄψ,10 but we cannot be sure if that was the case. The above-mentioned example of intrinsic ambiguity could also be seen as a limit case: perhaps Homeric audiences would not have doubted to understand that turn of phrase as an epithet of all sheep or as a descriptor of some sheep in particular.

Even admitting a degree of uncertainty in individual instances, the distinction between exogenous and intrinsic ambiguity is clear,11 and the majority of cases can be attributed to one or the other type by an attentive study of the texts.

Productive ambiguity can be defined as a subtype of intrinsic ambiguity in which a) the ambiguity is ultimately irresolvable; b) different coexisting meanings on their own contribute something to the interpretation of the text; c) the coexistence of meanings contributes something to the interpretation of the text.12 The first condition is easily understandable, though in many instances it can be hard to define since some scholars would always argue that they have found the definite argument to solve an ambiguous sentence or phrase. It should be noticed when it comes to ambiguity that insolvability is a matter of degree. One should not rush to leave very dubious cases outside the category of productive ambiguity simply because one has found reasons to support one or another interpretation. For the most part, such reasons do more to contribute to the productivity of the ambiguity than to solve it.

The second condition is also relatively clear since it is nothing more than the necessary condition for the persistence of ambiguity.13 If somebody were to suggest that φηρσὶν in Il. 1.168 is not alluding to the Centaurs, since the word simply means “feral beasts,” we would quickly dismiss the hypothesis by noting that an unclear allusion to some unknown beasts does nothing for the interpretation of the passage 1.262-72, while Centaurs explain the mention of the Lapiths and connect Nestor’s words to a well-known mythological episode.

It is, of course, the third of the conditions that distinguishes productive ambiguity from every other type of ambiguity: it is not only that possible meanings contribute to the interpretation of the text, but also the fact that there are two or more possible meanings. As subtle as this may sound, it is much clearer in practice than one may think. In Il. 22.111-20, for example, in the middle of a soliloquy, Hector contemplates pledging to return Helen and the stolen goods to Achilles, to give everything back to the sons of Atreus, to divide everything else in Troy with the Achaeans, and to make the Trojan council swear not to hide anything ἀλλ’ ἄνδιχα πάντα δάσασθαι, “but to divide it all, equally.” One may consider the possibility that ἄνδιχα here means “equally among the Achaean leaders,” instead of “equally between Trojans and Achaeans,” given that it is somewhat strange that Hector is willing to give only half the goods in Troy to save it, considering that the alternative is the complete destruction of the city.14 There is little support for the idea in other instances of the word, but we cannot fully dismiss the possibility. Both meanings contribute to the interpretation of the text in different ways. However, the coexistence of meanings does nothing for it. It is impossible for Hector not to know what he means, since it is a soliloquy, and the one or the other meaning must be his actual meaning. The ambiguity itself also does little for the passage more broadly, since the point is not how the goods would be distributed – that is an Achaean, not a Trojan problem –, but the fact that they would be distributed.

Il. 22.120, though a somewhat artificial one given the relative certainty regarding the meaning of ἄνδιχα (see n. 11), is a good example of intrinsic persistent non-productive ambiguity. In the rest of this paper, I will explore the much more interesting cases of productive ambiguity that can be found in Book 2.15

Some cases of productive ambiguity in Iliad 2

Il. 2.73, ‛h θeμις ἐστίν

When introducing the (in)famous test, Agamemnon includes in the middle of his speech a very peculiar phrase:

ἀλλ' ἄγετ’, αἴ κέν πως θωρήξομεν υἷας Ἀχαιῶν.
πρῶτα δ' ἐγὼν ἔπεσιν πειρήσομαι, ἣ θέμις ἐστίν,
καὶ φεύγειν σὺν νηυσὶ πολυκλήϊσι κελεύσω·
ὑμεῖς δ' ἄλλοθεν ἄλλος ἐρητύειν ἐπέεσσιν. (Il. 2.72-5)
Come, let us arm the sons of the Achaeans –
but first I will test them with a speech, which is appropriate,
and I will order them to flee with their many-benched ships;
you, on all sides, check them with your words. (Il. 2.72-5)

The problem of what ἣ θέμις ἐστίν means here is an old one, as shown by the fact that it is already discussed in a scholium. What is Agamemnon’s right to test the troops? Or should we take θέμις here as “custom” and understand that it is customary for the commander to test the troops? The meaning of the formula can be inferred by studying the rest of its instances, as has been done by Kirk (1985, p. 122-3), Du Sablon (2009, p. 137-9), and, in more detail, Sampson (2009, p. 29-35). These analyses conclude that the phrase designates proper behavior in an undefined way, a behavior that respects the laws and/ or habits of the gods, of men, of family, of nature, etc. This, of course, does not help much in understanding the meaning of the formula in this case, since the problem is precisely which law or custom is Agamemnon talking about.16

Now, as it is obvious, the ambiguity here is both irresolvable and persistent. We cannot define the exact intention of Agamemnon’s words, and several choices produce interpretative advantages. Taking the words with the meaning “this is what is usual,” Scodel (1999, p. 49-50), and Brügger; Stoevesandt; Visser et al. (2010, ad loc.),17 claim that the king is normalizing what the internal and external audiences of the speech may have considered abnormal, therefore avoiding and justifying the absence of an extensive debate regarding the Diapeira. Sampson (2009, p. 35-43) and Du Sablon (2009, p. 149-55), on the other hand, understand θέμις in the passage as an allusion to the wider issue of the order of the army and of Achaean society, therefore implying that there is much more than a mere rhetorical strategy.

I would suggest that these interpretative problems are inextricably implicated in the words of Agamemnon since the coexistence of these meanings contributes to the interpretation as much as each possible individual meaning. The apparently extemporaneous justification of the test leaves us thinking about its justification. On the one hand, it shows us that Agamemnon is sufficiently worried about the legitimacy of what he intends to do to add some rhetorical flourish to the proposal. On the other hand, this flourish draws attention to the issue of that legitimacy; as shown by Sampson, the very act of justifying the test makes us think about the θέμις of the situation and underlines the inability of Agamemnon as a leader. But these are not so much two coexistent interpretations as they are two sides of the same coin: it is the ambiguity that makes the passage function because a solution would destroy its power. Is Agamemnon a good leader that does what leaders should do? Or is he an incapable commander who almost destroys his chances of winning with a dumb move to show his power?18 The fact that the poem offers no answer to these questions is key to understanding the Diapeira, and its introduction anticipates this by including the ambiguous statement that testing the troops is θέμις.

Il. 2.285, πᾶσιν ἐλέγχιστον θέμεναι

At the beginning of his famous speech at the end of the Diapeira, Odysseus speaks directly to Agamemnon in an apparently sympathetic way:

Ἀτρείδη, νῦν δή σε, ἄναξ, ἐθέλουσιν Ἀχαιοί
πᾶσιν ἐλέγχιστον θέμεναι μερόπεσσι βροτοῖσιν,
οὐδέ τοι ἐκτελέουσιν ὑπόσχεσιν ἥν περ ὑπέσταν
ἐνθάδ’ ἔτι στείχοντες ἀπ’ Ἄργεος ἱπποβότοιο,
Ἴλιον ἐκπέρσαντ’ εὐτείχεον ἀπονέεσθαι. (Il. 2.284-8)
Son of Atreus, now, my lord, the Achaeans have a mind
to make you most contemptible in the eyes of all mortal men,
nor do they fulfill for you the promise that they undertook
at that time that they were leaving for here from the horse-grazed pastures of Argos–
that they would return home after sacking well-walled Ilion. (Il. 2.284-8)

The sequence of thought in the first three lines has caused interpretative problems since antiquity. Already a scholiast (Σb 284-5) provides a long explanation asserting that Odysseus is attempting to sympathize with Agamemnon, and most modern scholars assume with Kirk (1985, ad loc.), that “they want to make you a reproach among men, and do not fulfill their promise…” means “…by not fulfilling….”19 However, that is not what Odysseus is claiming. There is no γάρ in the second clause, and it is not ἐκτελέουσιν that governs θέμεναι, but ἐθέλουσιν, which means that nothing is said here regarding the Achaeans’ wanting to fulfill the promise.20 They simply do not fulfill it, and no explicit relation is made between that and their wanting to make Agamemnon the most contemptible man is made. While we can reconstruct the reasoning, there is a certain ambiguity of logical connection. Is one action the consequence of another, or are they two different attitudes of the Achaeans?

The usual interpretation has clear consequences, as shown since the scholium: Odysseus begins by indirectly praising Agamemnon and criticizing the troops, which is, after all, his goal. However, the ambiguity is persistent: if one takes both actions as independent from one another, then “to make you the most contemptible” can be understood not as proleptic, but as analeptic, that is, not as a consequence of their leaving Troy, but as the cause. The Achaeans want to make Agamemnon a reproach because of what he has done, because they despise him for quarreling with Achilles, and because they are tired of the never-ending war. Therefore, they will leave Troy because they will not keep on fighting for a leader they do not respect.

The productivity of the ambiguity comes from the realization that Odysseus is not talking to Agamemnon, but to the troops through the king.21 The goal is to convince them that, even if he understands them, he cannot abide by their behavior. Their wailing and longing to get back home are embarrassing, even if they are tired, even if their king is hateful. Now, it is obvious that Odysseus cannot tell this to Agamemnon, but by leaving his syntax a bit confusing at the beginning of his speech, he allows some room to recognize the legitimacy of the troops’ complaint. The ambiguity of the addressee and the contradictory positions of the possible targets of his message suggest not only that both interpretations are admissible, but, more importantly, that their coexistence is a key rhetorical strategy of the hero.

Il. 2.340-9

The center of Nestor’s intervention at the end of the Diapeira has three ambiguous statements in a few lines:

ἐν πυρὶ δὴ βουλαί τε γενοίατο μήδεά τ’ ἀνδρῶν
σπονδαί τ’ ἄκρητοι καὶ δεξιαί, ἧις ἐπέπιθμεν.
αὔτως γὰρ ἐπέεσσ’ ἐριδαίνομεν, οὐδέ τι μῆχος
εὑρέμεναι δυνάμεσθα, πολὺν χρόνον ἐνθάδ’ ἐόντες.
Ἀτρείδη, σὺ δ’ ἔθ’ ὡς πρὶν ἔχων ἀστεμφέα βουλήν
ἄρχευ’ Ἀργείοισι κατὰ κρατερὰς ὑσμίνας,
τοῦσδε δ’ ἔα φθινύθειν ἕνα καὶ δύο, τοί κεν Ἀχαιῶν
νόσφιν βουλεύωσ’ – ἄνυσις, δ’ οὐκ ἔσσεται αὐτῶν –
πρὶν Ἄργος δ’ ἰέναι πρὶν καὶ Διὸς αἰγιόχοιο
γνώμεναι εἴτε ψεῦδος ὑπόσχεσις εἴτε καὶ οὐκί. (Il. 2.340-9)
Let the counsels and plans of men be burned in fire,
and solemn libations of treaty and the pledges in which we trusted!
For we contest with words in this manner, nothing expedient
are we able to find, for all the long time we have been here.
Son of Atreus, hold firm yet, as before, to your unshaken plan,
lead the Argives through the mighty combat,
and let those perish, the one or two who
make their plans apart from the Achaeans – they will accomplish nothing –
to return early to Argos, before knowing
whether the promise of Zeus who wields the aegis was false, or not. (Il. 2.340-9)

The ambiguity of all three statements can be reduced to a single question: what exactly is Nestor talking about? When he says “we contest with words,” is he talking about the recent dispute with Thersites, or is he talking about the events of Book 1? What before does “as before” mean, the Diapeira or the time before this misguided idea? And, of course, who are the one or two who make their plans apart from the Achaeans?

The three questions, as can be noted, pertain to a more profound ambiguity regarding the character of Agamemnon. In the first case, if we take the reference to be the assembly of Book 1, then this “quarreling with words” would be a direct criticism of the king, who fought with Achilles instead of fighting with the Trojans. However, if it is a reference to the Thersites episode, the “quarreling” would be a criticism of this low character and maybe of the troops in general, given that they needed speeches to go back to the war after the test.

In the second case,22 the meaning “as before the Diapeira” is, of course, a harsh criticism of Agamemnon’s failure and poorly conceived idea of testing the troops. However, the meaning “as you have been doing until now, including the Diapeira,” is a simple reminder of the status of the king and even praise of his commanding skills.

The final case is the most interesting one, since both possible referents, Thersites and Achilles, imply some approval of Agamemnon’s actions, but to vastly different degrees. If the one who will not succeed is Achilles, Nestor is essentially stating that he is now on Agamemnon’s side, which has far-reaching implications for the position of the army in the dispute. If it is Thersites, his criticism is little more than a corollary of Odysseus’ words before his and acts as an introduction to the new prophecy. It ties up the episode but has no long-term impact on the plot of the poem.

Scholars have, of course, produced several arguments to defend each position.23 However, the ambiguity of Nestor’s words is coherent with the context: by not being clear regarding the degree of his criticism to Agamemnon, as did Odysseus, he can appeal to the king and the troops at the same time. The listeners (both internal and external) can understand his speech in the way most suited to them, which is, of course, good rhetoric. The poet even illustrates this point in Agamemnon’s reply (370-93): the king catches some of Nestor’s expressions and in general draws some inspiration from his words,24 but at every turn, he restricts the interpretation of Nestor’s speech. He mentions the assembly of Book 1 (375-78), but only to blame the gods for his actions. This might seem to be poorly conceived rhetoric because he concedes that he was seemingly responsible for the quarrel; however, by choosing that interpretation he can make Nestor’s words in lines 346-7 refer to Achilles (379-80), putting the elder leader on his side of the dispute, a victory much more significant at this point than avoiding any mention of the quarrel with Achilles (for which he has avoided responsibility in any case). This “solution” to the ambiguity, however, actually enhances its productivity since the audience can perceive the gap between the two speeches, and in that gap, both characters are defined.

2.807, οὔ τι θεᾶς ’eπος ἠγνοίησεν

The final case of productive ambiguity that I will discuss is also the introduction of Hector in the poem.25 The focus of the narrative has turned for the first time towards the Trojans (786): the goddess Iris flows to them (786-90), adopts the appearance of Polites, one of the sons of Priam (791-5), announces the coming of the Greeks to Priam (796-801), and instructs Hector to marshal the troops and the allies (802-6). When her speech ends, for the first time in the story Hector becomes an active character (he has been mentioned by other characters in 1.242, 2.416, and 2.802):

Ὣς ἔφαθ’, Ἕκτωρ δ’ οὔ τι θεᾶς ἔπος ἠγνοίησεν,
αἶψα δ' ἔλυσ’ ἀγορήν· ἐπὶ τεύχεα δ’ ἐσσεύοντο. (Il. 2.807-8)
So she spoke, and Hector did not fail to recognize the word of the goddess,
and at once he broke the assembly; and the men rushed to their arms. (Il. 2.807-8)

This is the only time in Homeric epic that the verb ἀγνοέω has a speech as its object, and the only time it appears as a closing formula. The use was problematic enough as to deserve comment from Aristarchus (Σa ad 807), who explained that οὐ κεῖται δὲ συνήθως ἡμῖν τὸ ἠγνοίησεν, ἀλλ’ ἀντὶ τοῦ οὐκ ἀπίθησεν (“Customarily, we do not find ἠγνοίησεν, but rather οὐκ ἀπίθησεν”).26 Modern interpreters at least since Ameis and Henze (1884, ad 807), have not followed this interpretation (assuming that Aristarchus is implying that ἠγνοίησεν is merely a synonym) but understood that the words imply something more, namely, that Hector recognized the goddess. Again, however, that is not what the poet says, and Aristarchus is certainly right that, if we take the expression at face value, what we would expect here is a synonym to the regular indication that the receptor of the orders did not disobey them,27 particularly because that is what happens next. By changing the formulaic expression, however, the poet leads our attention to the issue, which is why most modern interpreters assume that ἠγνοίησεν implies “he recognized”, in spite of the fact that the word is not use in the sense “not recognize” in the Iliad.28

The persistence of the ambiguity is clear. Whether Hector recognized Iris or not, we know something more about the Trojan prince, but one may argue that it is not productive, since either Hector recognized the goddess, or he did not. But to that, there is a simple answer: our doubt regarding Hector’ ability to recognize the gods’ work will be quintessential in the resolution of the plot since he will ignore the role of Apollo in the killing of Patroclus (cf. 16.830-42) and, more importantly, his death will be caused by the deception of a goddess (cf. 22.226-305), whom he recognizes too late (22.296-305).29 By carefully choosing his words the poet leaves that doubt open, almost as a Chekhov’s gun that he will shoot twenty books later.30

As every literary resource, ambiguity is an appropriation of a common phenomenon in language with literary purposes. However, given that literary language can also be ambiguous, not all ambiguities can be considered a resource, and not all that can be easily identified. Intuitive approaches to the issue are insufficient, as shown by the fact that none of the above-studied cases of productive ambiguity have been identified as such by scholars, who have extensively debated in some of them which is the correct interpretation. Perhaps because the “single word” approach has dominated the study of the phenomenon,31 much of this debate has been misguided and could have been easily avoided with a better understanding of how to identify ambiguities that are not a mere accident of the language.

In fact, productive ambiguity is a regular tool in the arsenal of narrators. To mention another modern example, when Obi-Wan Kenobi says that Darth Vader killed Luke Skywalker’s father in the original Star Wars (George Lucas, 1977), in retrospect we know that he was being deliberately ambiguous, and he claims so himself in Return of the Jedi (George Lucas and Lawrence Kasdan, 1983) when he explains that “Your father was seduced by the dark side of the Force. He ceased to be Anakin Skywalker and became Darth Vader. When that happened, the good man who was your father was destroyed. So what I have told you was true... from a certain point of view.”32 This explanation is a different kind of ambiguity (what we may call productive retrospective ambiguity) from the one we have been exploring, but it is certainly not one unknown to Homer.33 It is not based on the multiple meanings of a single word or sentence, but on the intrinsic openness of many a statement in any language.

What all examples we have seen have in common is not only that the audience cannot tell what does what they are seeing or listening to means but that the fact that they cannot is an essential part of the experience.34 This takes us back to the opening example of this paper: Il. 2.222-3 and the referent of τῶι. The answer to the question presented in the introduction should at this point, however, have become obvious: the one with whom the Achaeans are greatly angered and resentful in spirit is both Agamemnon and Thersites, and both of them are also not the one with whom they are angry. As with the rest of the examples in the Diapeira, the poet leaves open who is really to blame for the situation and the degree of criticism that Agamemnon deserves. The ambiguity anticipates both Odysseus’ and Nestor’s speeches and permeates our understanding of the whole episode. Therefore, even if, in the end, Agamemnon’s authority is restored, we cannot really tell how much it has suffered from the events in the first two Books of the poem, which, in turn, will greatly affect our perception of his character in the rest of the narrative.

By taking productive ambiguity into account in our analysis of the Iliad we can better understand how the poet leads the audience in the construction of meaning.35 Ambiguity permeates the poem, and, while in many cases the struggle to decipher the correct interpretation of a phrase or scene is worthy, in many others it is little more than a misguided attempt to conceal a fundamental poetic device that audiences would not have missed, as we do not when watching the movies mentioned.36


ABRITTA, Alejandro. κόσμον ἐμῶν ἐπέων ἀπατηλὸν [“El engañoso orden de mis palabras”]: Un análisis coral del poema de Parménides, con traducción y comentario textual. Buenos Aires: Facultad de Filosofía y Letras, Universidad de Buenos Aires, 2020. in press. Available:

ALEXANDER, Caroline. The Iliad. New York: Harper Collins, 2015.

AMEIS, Karl Friedrich; HENTZE, Karl. Homers Ilias. Vol. 1, part 1. Leipzig: Teubner, 1884.

BATTEZZATO, Luigi. Ambiguity. In: ROISMAN, Hanna M. (ed.) The encyclopedia of Greek tragedy. London: Wiley-Blackwell, 2013, p. 96-8.

BEEKES, Robert. Etymological dictionary of Greek. With the assistance of Lucien van Beek. Leiden: Brill, 2010.

BRÜGGER, Claude; STOEVESANDT, Magdalene; VISSER, Edzard, et al. Homers Ilias. Gesamtkommentar, Band II: Zweiter Gesang (B), Faszikel 2: Kommentar. Berlin: De Gruyter, 2010.

BURGESS, Jonathan S. The Death and Afterlife of Achilles. Baltimore: The Johns Hopkins University Press, 2009.

CHRISTENSEN, Joel P. Reconsidering ‘good’ speakers: speech-Act theory, Agamemnon and the Diapeira of Iliad, II. Gaia, v. 18, p. 67-82, 2015.

COOK, Erwin F. Agamemnon’s test of the army in Iliad Book 2 and the function of Homeric Akhos. American Journal of Philology, v. 124, p. 165-98, 2003.

DE JONG, Irene J. F. A Narratological commentary on the Odyssey. Cambridge: Cambridge University Press, 2004.

DU SABLON, Vincent. Le système conceptuel de l’ordre du monde dans la pensée grecque à l’époque archaïque. Τιμή, μοῖρα, κόσμος, θέμις et δίκε chez Homère et Hésiode. Dissertation (Docteur en Histoire) – PhD Program in Université Catholique de Louvain, Louvain-la-neuve, 2009.

EMPSON, William. Seven types of ambiguity. London: Chatto and Windus, 1949.

ERBSE, Hartmut. Scholia graeca in Homeri Iliadem (Scholia Vetera). v. I: Praefationem et scholia ad libros A-D continens. Berlin: De Gruyter, 1969.

FARRELL, Joseph. Intention and intertext. Phoenix, v. 59, p. 98-111, 2005.

GADAMER, Hans-Georg. Truth and method. Trans. rev. by Joel Weinsheimer and Donald G. Marshall. London: Continuum, 2004.

GARCÍA BLANCO, José; MACÍA APARICIO, Luis M. Homero. Ilíada. v. 1, reprinted. Madrid: CSIC, 2014.

HEBEL, Volker. Untersuchungen zur Form und Funktion der Wiedererzählungen in Ilias und Odyssee. Dissertation (PhD) – PhD Program in Ruprecht-Karls-Universität, Heidelberg, 1970.

HEIDEN, Bruce. Common people and leaders in Iliad Book 2: the invocation of the Muses and the catalogue of ships. Transactions of the American Philological Association, v. 138, p. 127-54, 2008.

JENSEN, Minna Skafte. Writing Homer. A study based on results from modern fieldwork. Copenhagen: Det Kongelige Danske Videnskabernes Selskab, 2011.

KELLY, Adrian. Homeric battle narrative and the Ancient Near East. In: CAIRNS, Douglas; SCODEL, Ruth (ed.). Defining Greek narrative. Edinburgh: Edinburgh University Press, 2014, p. 29-54.

KIRK, Geoffrey Stephen. The Iliad. A commentary. v. I. Cambridge: Cambridge University Press, 1985.

KLEIN, Jared S. Homeric Greek αὖ: A synchronic, diachronic, and comparative study. Historische Sprachforschung, v. 101, p. 249-88, 1988.

KUKKONEN, Karin. Fantastic cognition. In: BURKE, Michael; TROSCIANKO, Emily T. (ed.). Cognitive literary science. Oxford: Oxford University Press, 2017, p. 151-68.

LEAF, Walter. The Iliad, edited, with apparatus criticus, prolegomena, notes, and appendices. London: Macmillan and Co., 1900.

PORTER, Andrew. Agamemnon, the pathetic despot: Reading characterization in Homer. Washington DC: Center for Hellenic Studies, 2019.

RENEHAN, Robert F. Conscious ambiguities in Pindar and Bacchylides. Greek, Roman, and Byzantine Studies, v. 10, p. 217-28, 1969.

SAMPSON, Christopher Michael. Themis in Sophocles. PhD thesis – PhD Program in the University of Michigan, Michigan, 2009.

SCODEL, Ruth. Credible impossibilities. Conventions and strategies of verisimilitude in Homer and Greek tragedy. Stuttgart: Teubner, 1999.

SPINA, Luigi. L’oratore scriteriato. Per una storia letteraria e política di Tersite. Napoli: Loffredo, 2001.

STANFORD, William Bedell. Ambiguity in Greek literature. Studies in theory and practice. Oxford: Blackwell, 1939.

THALMANN, William G. Thersites: Comedy, scapegoats, and heroic ideology in the Iliad. Transactions of the American Philological Association, v. 118, p. 1-28, 1988.

TURKELTAUB, Daniel. Perceiving Iliadic gods. Harvard Studies in Classical Philology, 103, p. 51-81, 2007.

VON DER MÜHLL, Peter. Die Diapeira im B der Ilias. Museum Helveticum, v. 3, p. 197-209, 1946.

WEST, Martin Litchfield. Homeri Ilias. 2 v. Munich: Teubner, 2006.

WEST, Martin Litchfield. The Making of the Iliad: Disquisition and analytical commentary. Oxford: Oxford University Press, 2011.


1 I quote from (accessed: 02/12/2020).
2 I quote from West (2006). The translation is that from Alexander (2015), with some modifications.
3 See e.g. Spina (2001, p. 29); Brügger, Stoevesandt, Visser et al. (2010, p. 74). Thalmann (1988, p. 18, n. 44) supports Leaf’s positions with new arguments.
4 Before coping with the issue of defining and providing a method to identify “productive ambiguity,” I should clearly state that the concept of ambiguity and the use in literary studies for which I will advocate should not be confused with the concept and use of ambiguity in approaches such as Gadamer’s (e.g. Gadamer, 2004, p. 499), from which I will try to stay far away. Perhaps the most significant difference is that the “ambiguous” in my approach is nothing essential to a work of art, but merely an incidental device that a poet can use to produce meaning, as happens with alliteration, etymological plays, and others.

Another important theoretical problem I should clarify before the main argument is that of the “author’s intention.” This is a very complicated issue (see e.g. Farrell, 2005), mainly in the field of intertextual studies. My position is fairly straightforward: both the author and the receptor of a literary work are not real persons but theoretical constructs, and therefore it makes no sense to debate their actual capacities and intentions (though it does and it is essential to debate their possible capacities and intentions). If, for the sake of simplicity, we exclude the issue of intertextuality (as we can for the purposes of this paper), “author’s intention” is merely the terminology used to indicate that the words in any utterance are not random, and that we can extract meaning from them. Therefore, if a statement is ambiguous, we can assume that it was deliberately made ambiguous by the author (that is, the theoretical construct, not the actual person, about whose intentions we know nothing – at least in the case of ancient authors). Note that this is valid both for written and for oral literature since for our purposes there is no difference between them regarding the “intentions”. This also means that throughout this paper, I will use much “intentionalist” terminology since there is no reason not to do so if one keeps in mind that “the poet” is not the real poet (but, and this is important, it is also not not the real poet), but the theoretical construct we use to exclude randomicity in a text (which also means, therefore, that “the poet”, “the poem”, “the text”, etc. are for the most part functional synonyms).

5 See Battezzato (2013), who clearly refers to productive ambiguity throughout the entry, as most have done. This seems to be because most scholars assume that important ambiguities for the interpretation of a text are obvious for any reader, and therefore an intuitive approach to them is sufficient. This is, naturally, incorrect, both from a methodological standpoint (even if they were obvious, we can certainly profit from a delimitation) and from a practical one (none of the ambiguities that will be explored in this paper have been, to my knowledge, ever recognized as productive). Therefore, I build up from linguistic ambiguity rather than down from obvious cases of literary ambiguous statements.
6 “For the most part” is necessary, since in oral literature it is not uncommon for the audience to interact with the singer in different ways (see Jensen, 2011, p. 80-3). While, of course, the Homeric texts are oral literature, the fact that we can relate to them only as written literature (at least for the purposes of this paper) allows us to dismiss this possibility. I would claim, however (see the conclusions), that even contemporary audiences would have enjoyed the ambiguity.
7 Note that these “types” of ambiguity should not be confused with the types of ambiguity studied by Stanford (1939), following partly the classical work of Empson (1949) (1º ed. from, 1930), which classify the ways in which a statement can be ambiguous. Many if not all of the cases studied by Stanford can be considered cases of productive ambiguity (see also Renehan, 1969), but the fact that he claims that the Iliad has very little ambiguity proves the limits of his approach. Besides the obvious differences between Stanford’s classification and the one presented here, there is one that is theoretically very significant: while many ambiguities can and do escape from Stanford’s net, any ambiguous statement must necessarily fall into one of the categories presented below. Needless to say, this does not mean the authors were not aware of the issue (see in general Empson’s chapter VIII), but the fact that they do not systematize it is telling in itself and explains the need for a methodological approach to the problem.
8 See on the problem Beekes (2010), s.v.
9 A reviewer insightfully notes that this ambiguity might be solved with better knowledge of Homeric/ Iliadic tradition, if sheep were traditionally considered feeble. However, we have a good number of other mentions of sheep and goats in the poems, in none of which they are classified as ἀνάλκιδα θυμὸν ἐχούσας, nor do they receive epithets linked to their being weak or cowards. Therefore, I believe we have enough evidence to support the idea that this is not an ambiguity produced by our ignorance.
10 See on this Kirk (1985, p. 79-80); García Blanco and Macía Aparicio (2014, p. 16-7), both with further references.
11 It should be noted, however, that the limit between exogenous and intrinsic ambiguity depends on how much one is willing to include “within” the target culture. Is it something that Archaic audiences understood but Hellenistic audiences did not intrinsic or exogenous? It depends entirely on the goals of the researcher using the concepts, and I believe such fluidity is more advantageous than inconvenient. Although it does not affect the results, for the rest of this paper I will assume that the limit of intrinsic ambiguity is the audience contemporary to the rhapsodes.
12 “Contribute something” might be seen as a somewhat vague formulation, but that is the point: the delimitation of both “contribute” and “something” must be made within the context of the study of specific texts, authors, and cultures.
13 As a shorthand, I will classify ambiguities that fulfill this condition as “persistent ambiguities”.
14 I am certainly not claiming such a thing. The example merely illustrates a case of non-productive ambiguity. The parallels in Il. 18.511 and Hes., Op. 13 almost guarantee that ἄνδιχα means “in two.”
15 I have chosen Book 2 somewhat arbitrarily, but any other of the Iliad or the Odyssey would have certainly provided enough cases. Productive ambiguity, if not ubiquitous, is quite a usual device of the Homeric poet. Book 2, however, offers not only an interesting but also a diverse number of examples.
16 Since Kelly (2014) has successfully refuted the claims of those that saw links between the Diapeira and Near Eastern traditions, I will not deal with the issue.
17 See also Leaf (1900, ad loc.); West (2011, ad 73-5).
18 The question has inspired much debate amongst scholars, which, I would say, further proves the productivity of the ambiguity. See Porter (2019), sec. 4.2.2.
19 See Cook (2003, p. 181); Brügger; Stoevesandt; Visser et al. (2010, p. 89).
20 There might be, as a reviewer points out, a phonetic play in ἐθέλουσιν- ἐκτελέουσιν, but that certainly does not disambiguate the sequence of thought.
21 See Hebel (1970, p. 43-4). The narrator’s introductory words (v. 281-2, ὡς ἅμα θ’ οἱ πρῶτοί τε καὶ ὕστατοι υἷες Ἀχαιῶν / μῦθον ἀκούσειαν, “so that the sons of Achaeans, both in front and behind, / might hear his word”) anticipate this.
22 On which see Brügger; Stoevesandt; Visser et al. (2010, p. 104), with references.
23 See Leaf (1900, ad 2.346); Von der Mühll (1946, p. 207-8); West (2011, ad 2.342-3); García Blanco and Macía Aparicio (2014, p. 58). While he does not deal with the ambiguities, Christensen (2015) studies the complex construction of Agamemnon’s character in the episode.
24 νόσφιν βουλέυωσ’ (347) can be linked to ἔς γε μίαν βουλεύσομεν (379); ἄνυσις (347) with ἀνάβλεσις (380); and in both speeches, we find a threat to those who refrain from the fight (v. 357-9 and 391-3).
25 I have excluded from consideration potentially the most interesting case in the book, that is, the final verse of the invocation to the Muses before the Catalogue of Ships (v. 493), since the interpretation of this passage requires a more detailed analysis. On the problem, see Heiden (2008, p. 128-34). The main source of ambiguity here is the meaning of αὖ in 493, which can be understood as adversative (as does, for example, Klein, 1988, p. 257) or as continuative (as Kirk, 1985, p. 167-8). In the second case, we would have a circular structure: request to the Muses to name the kings, the problem of the mass, the possibility of naming the mass, announcement of the list of kings and ships. If αὖ is taken as adversative, however, the point would be that, though the poet cannot name the masses, he would list the kings. The fascinating part of the ambiguity is that it emphasizes one of the main issues with the catalog, as observed by Heiden: is it mainly about the leaders, or is it mainly about the people and their places of origin?
26 The quote is from Erbse (1969), and the translation is mine.
27 E.g. in this very same Book, 2.166: Ὣς ἔφατ’, οὐδ’ ἀπίθησε θεὰ γλαυκῶπις Ἀθήνη.
28 Cf. 1.537 and 13.28; in both cases, the meaning is something like “to ignore/ disregard the intention or nature of a person,” but note that in both cases the object of the verb is a person well known by the subject(s) of the verb, while in 2.807 the object is the speech and Iris is certainly not well known by Hector (though Polites is!). “To not recognize” is the sense of the word in the Odyssey (cf. Od. 5.78, 20.15, 23.95, and 24.218).
29 On Hector’s encounters with the gods and epiphanies in the poem in general, see Turkeltaub (2007), who does not count this encounter with Iris as a recognition (see p. 61, n. 31), “because Hector in 2.807 (…) does not explicitly recognize Iris’ voice, her ops, but only her speech, her epos.”
30 A “seed,” in the terminology of De Jong (2004), esp. xvii-xviii, but I am not sure if this ambiguity would count as a “piece of information.” A reviewer points out that “that Hector recognizes the goddess here but not latter (…) doesn’t lessen the passage as seed,” but I disagree: how can Hector’s recognition of Iris be a seed for his not recognition of Athena? It might contribute something to the surprise (“he failed to recognize the goddess this time!”), but I find that that is a much weaker and uninteresting effect.
31 See, in fact, the definition of “ambiguity” in Renehan (1969, p. 217).
32 I quote from (accessed: 02/12/2020).
33 When Achilles mentions Apollo two times in four verses in Il. 16.94-7, the first time noting that he loves the Trojans and the second one praying for his and Patroclus survival, from the point of view of the end of the Book, after Patroclus has been killed by Apollo, those mentions acquire a very different meaning, and we cannot be entirely sure if Achilles did not have in mind the possibility of the god’s violent and direct intervention. Of course, this ambiguity is retrospective only if the audience did not know (at least for this particular case) that Apollo had a role in the killing of Patroclus, which is not a risky hypothesis, given the variations even regarding the death of Achilles (see Burgess, 2009, p. 38-9, though he assumes that Apollo+Paris is the traditional version). As a side note, one may observe that retrospective productive ambiguity is tragic irony without foreknowledge, which explains why it is far less common in ancient poetry than in modern cinema.
34 A very similar idea has been used by Kukkonen (2017) to define the fantastic as a genre, the difference being perhaps that, while productive ambiguity is a resource that can be used locally in any work of art, its equivalent in fantastic is a macro-technique that permeates every aspect of the narrative. I say “perhaps” because one could easily argue that the difference is merely one of degree. I cannot dive here into the Bayesian aspect of the device, but it is without a doubt a very profitable approach to productive ambiguity.
35 Given the scope of this paper, I limit my conclusions to the Iliad. I believe, however, that productive ambiguity is an essential technique in Ancient Greek literature and, as the example of Heraclitus’ surely shows, philosophy (see also Abritta, in press: 21-2, 38-9, 81-2, 91-3, who demonstrates that the same is true for Parmenides). As noted above, in fact, most if not all of Stanford’s many examples can be classified as productive ambiguities.
36 Does standing for productive ambiguity means that we cannot attempt to resolve ambiguities? Certainly not. But it does mean that solving apparently irresolvable ambiguities means that we must eliminate both its productivity and its persistence. And it is important always to take into account that these are very different things to do: persistence is not eliminated by showing that one option works better in a certain context, but by showing that the other option does not work in that context. If one cannot show that, namely, if one cannot eliminate persistence, it seems to me to be a better idea to seek for productivity than to insist on trying to find a solution.
HTML generado a partir de XML-JATS4R por