2.1 Defining hortative questions

For a question to qualify as hortative, it must contain the question word τί in the sense ‘why’ (not ‘what’), and the question must be modified by a negation. However, not all questions with the formula τί οὐ are hortative. A review of a corpus consisting of Aeschylus, Aristophanes, Euripides, Sophocles, Plato and Xenophon (searches in the Thesaurus Linguae Graecae on the exact phrases τί οὖν οὐ, τί οὐ, τί οὐκ, τί οὐχί, and proximity searches on τί οὖν … οὐ, τί … οὐ, τί … οὐκ and τί … οὐχί; cases are listed in the Appendix at the end of this article) reveals a distinction between three types of questions.

The first type concerns the idiomatic phrase τί δ’ οὐ μέλλω (also in the third person, as well as in the plural; one instance in the imperfect), literally ‘why/how am I not going to’, that is, ‘of course I do’. These cases will not concern us here.

The second type consists of what we may call ‘indignant’ questions. Here, the speaker is indignant, frustrated or surprised by the fact that something did not occur or that something is not going on. In these cases, the aorist is to be interpreted as a past tense, and the variation between the present and the aorist is temporal. Consider the following example of the aorist (Soph. OT 1391–3):

Oedipus bewails the fact that Μt Cithaeron allowed him to live as a child. A hortative interpretation (‘why don't you kill me’) is out of the question here, as the opportunity for carrying out the designated event has expired.Footnote ⁷

In the case of the present, the distinction between indignant questions and hortative questions is less clear. The difference between ‘I am frustrated that X is not happening’ and ‘Why don't you do X’ is much more subtle than that between ‘I am frustrated that X did not happen’ and ‘Why don't you do X’ (because, typically, there is an expiration date on the opportunity for carrying out X).

An ‘indignant’ interpretation is preferred when it is reasonable for the speaker to assume that the designated event should already be carried out at the present moment. This is especially clear when there has been a previous request or (implicit) injunction to carry out the described action. Consider the following example (Pl. Grg. 468c):

Socrates asks Polus a question. This requires an answer. Polus remains silent. This prompts Socrates to ask why Polus does not answer. Socrates is justified in assuming that Polus should at this moment be answering the question, and asks him what the reason is for his silence. This makes the import of the question more confrontational than in the case of a hortative question: the tone is ‘Why aren't you answering’, that is, ‘What is your deal that you are not answering’, rather than ‘Why don't you answer’, that is, ‘I suggest you answer’.

That the variation between the aorist and the present in indignant questions is temporal in nature is corroborated by the fact that the imperfect can also be used in such contexts. Thus there is aspectual variation in the past tense, which means that the choice of the present is unlikely to be aspectually motivated. An example is the following (Ar. Ran. 33):

Xanthias laments the fact that he did not take part in the battle of Arginusae, because then he would have more rights. In each case where the imperfect is used in indignant questions, the opportunity for carrying out the designated event has expired, as here (the sea-fight has been fought). The imperfect does not have a hortative use.Footnote ⁸

Hortative questions constitute the third type. A hortative question is actually a suggestion to carry out the described action, without the implication that it should already have occurred or should already be in the process of occurring. In the case of the aorist, we can define hortative questions as those questions with the phrase τί οὐ where the opportunity for carrying out the designated action is still actual. Consider the following example (Xen. Mem. 3.11.15):

Theodote asks Socrates to become her fellow-hunter. Socrates is in a position to fulfil her request. There is no reason for the speaker to feel that the designated state of affairs should have already been brought about before she asked the question.

In the case of the present tense, τί οὐ-questions are hortative when it is unreasonable for the speaker to suppose that the designated event should already be going on at the present moment. Consider Ar. Av. 828 τί δ’ οὐκ Ἀθηναίαν ἐῶμεν πολιάδα; ‘Why don't we let Athena Polias [wear the peplos]?’ Pisthetaerus and Euelpides, together with the chorus, are discussing some basic questions concerning their new city. One question is who should wear the sacred peplos, and Euelpides suggests that it should be Athena. As the issue is just being raised, and the city has not yet been founded, there is no reason for us to interpret Euelpides as saying that Athena should at this moment already be wearing the peplos.

A final point concerns person marking on the verb. It is generally understood that hortative questions occur only with first- and second-person verbs. This seems logical, as a hortative question seems to require the subject to be addressed directly. However, there are also instances of τί οὐ-questions with a third-person verb that function as hortative questions. In hindsight, this is not surprising, for if a language can have a third-person imperative, then why not a third-person hortative question? For example, (Soph. Ant. 1308):

The import of Creon's question is ‘Will someone please strike me with a sword?’Footnote ⁹

2.2 Accounting for the data

Before we turn to actual examples, a point of methodology needs to be addressed. I argue that the variation between the aorist and the present in hortative questions is aspectually motivated, as in grammatical categories such as the past tense indicative, the infinitive, the imperative, and so on. Even in those cases, however, the explanation for aspectual variations is often a matter of debate, and this issue cannot be solved in this article.Footnote ¹⁰ I do not pretend to present a full-fledged theory of aspect in classical Greek; rather, I adopt a working hypothesis of the semantic value of the two aspectual categories. When this working hypothesis does not fit the data, that does not automatically invalidate the claim that tense usage in hortative questions is aspectually motivated, because we may have to do with a peculiarity of Greek aspectual usage. Therefore, we also have to take the general aspectual behaviour of specific verb types into account.

2.2.1 The aorist in hortative questions

In this section I discuss a number of aspectual principles that account for the use of the aorist in hortative questions. I then turn to a few difficult instances.

Generally, the aorist (perfective aspect) construes the designated event as bounded, that is, as a complete whole. Such a construal is favoured when the verb phrase is telic (when it has an inherent endpoint: for example ‘destroy’, ‘cut’, ‘build a house’), and especially when the designated event is of short duration:Footnote ¹¹ for example Ar. Lys. 906 τί οὖν οὐ κατεκλίνης, ὦ Μύρριον; ‘Why don't you lie down, my Myrinne?’ The verb κατακλίνομαι ‘lie down’ designates an action that can be completed in a moment.Footnote ¹² Also, Soph. Ant. 1308 (cited above, page 5), the verb παίω ‘strike’ is punctual; Creon wishes to be put out of his misery by a single fatal blow.Footnote ¹³

Certain preverbs can make an atelic verb telic, or emphasize the aspect of ‘completion’ in the case of an already telic verb. In twelve cases of the hortative aorist, such a preverb is present (διά, lit. ‘through’, five instances; ἀπό, lit. ‘away from’, six instances; ἐξ, lit. ‘out of’, one instance). In the case of the present, by contrast, there is only one such instance (featuring διά). The case of διηγέομαι ‘fully discuss’ and διέρχομαι/διιέναι ‘go through’ is most instructive. An example is Pl. Symp. 173b τί οὖν … οὐ διηγήσω μοι; ‘Why then don't you tell me [about the symposium]?’Footnote ¹⁴ That the use of the aorist here is aspectually motivated is supported by the fact that in the imperative Plato only uses the aorist (three cases of δίελθε, four cases of διήγησαι). Moreover, the aspectual behaviour of this verb in other grammatical categories in the context is also of interest. In 172b, a friend asks Apollodorus to tell him about the symposium: σὺ οὖν μοι διήγησαι ‘so you tell me about it’. In 173b, he uses the hortative aorist: τί οὖν … οὐ διηγήσω μοι ‘why then don't you tell me about it?’ In 173c, Apollodorus acquiesces: εἰ οὖν δεῖ καὶ ὑμῖν διηγήσασθαι, ταῦτα χρὴ ποιεῖν ‘well, if it is necessary that I should tell you about it as well, then that is what I have to do’. In 173e, the friend repeats his request: ἀλλὰ διήγησαι τίνες ἦσαν οἱ λόγοι ‘but tell me what speeches were made’. Finally, Apollodorus says μᾶλλον δ᾽ ἐξ ἀρχῆς ὑμῖν ὡς ἐκεῖνος διηγεῖτο καὶ ἐγὼ πειράσομαι διηγήσασθαι ‘rather, I will try to tell you about it from the beginning, as he told me’. In all these cases where the verb designates the potential event of Apollodorus telling the story to his friend, the perfective (aoristic) aspect is used.Footnote ¹⁵

Another way in which the boundaries of the designated event become particularly salient is when it is emphasized that its duration should be as short as possible. This occurs in Pl. Plt. 279b–c τί δῆτα οὐ, καθάπερ ἐν τοῖς ἔμπροσθε τέμνοντες μέρη μερῶν ἕκαστον διῃρούμεθα, καὶ νῦν περὶ ὑφαντικὴν ταὐτὸν τοῦτ’ ἐδράσαμεν, καὶ κατὰ δύναμιν ὅτι μάλιστα διὰ βραχέων ταχὺ πάντ’ ἐπελθόντες πάλιν ἤλθομεν ἐπὶ τὸ νῦν χρήσιμον; ‘Why then, just as we divided each subject before by cutting off parts from parts, don't we now do the same with respect to the art of weaving, and, as quickly as we possibly can, go through all the steps and arrive again at what is useful for the present discussion?’ The aorist is used because the Stranger has the end firmly in view and wishes to reach it as quickly as possible (κατὰ δύναμιν ὅτι μάλιστα διὰ βραχέων ταχύ ‘as quickly as we possibly can’).

Finally, the aorist can be used to emphasize that the designated action needs to be carried out before another action can be carried out. Consider the example of Pl. Chrm. 154e τί οὖν … οὐκ ἀπεδύσαμεν αὐτοῦ αὐτὸ τοῦτο καὶ ἐθεασάμεθα πρότερον τοῦ εἴδους; ‘Why then don't we strip [Charmides] of exactly this and admire it before we do his looks?’ Socrates suggests they take a look at Charmides’ soul before admiring his looks. Now, the verb ἀποδύω ‘strip’ is telic, but θεάομαι ‘watch’, ‘admire’ is not. However, the point is that the interlocutors should admire Charmides’ soul before admiring his looks. The boundaries of the designated event are thus salient: before the latter can be done, the former must be completed.Footnote ¹⁶

Most instances in the data are relatively straightforward, but in some cases the aspectual interpretation of the aorist is less obvious. First, Eur. Heracl. 805 ὦ στρατήγ’ ὃς Ἀργόθεν | ἥκεις, τί τήνδε γαῖαν οὐκ εἰάσαμεν <lacuna>; ‘General coming from Argos, why don't we let this land be <…> ?’ Hyllus proposes a monomachy instead of an all-out battle, which would ravage the country. The verb ἐάω ‘leave be’ is in itself atelic (compare the present in Ar. Av. 828: n. 17). In my view, however, the point here is that, if this proposal is adopted, the parties involved will let Athens and Argos off the hook.

Second, Pl. Soph. 251e τί οὖν οὐ καθ’ ἓν ἀποκρινόμενος ἐφ’ ἑκάστου τὰ συμβαίνοντα ἐσκέψω; ‘Why don't you answer one by one and consider the consequences in each case?’ The Stranger confronts Theaetetus with three possible positions concerning the ability of entities to combine with each other. He then invites him to consider each option and see what the consequences of each view are. The verb σκέπτομαι ‘consider’ is atelic (one can keep considering ad infinitum), but the aorist suggests that each line of enquiry should be pursued to the end.

Third, Pl. Leg. 753a τί δῆτα οὐ καὶ σύ τε καὶ ὁ Μέγιλλος, ὦ ξένε, ἐκοινωνησάτην ἡμῖν τῆς πολιτείας; ‘Why then don't you and Megillus, stranger, take part in our state?’ κοινωνέω ‘have part in’ is atelic, but the aorist is ingressive, that is, it signals a transition from not having part in something to having part in that something. Compare Pl. Cra. 383a βούλει οὖν καὶ Σωκράτει τῷδε ἀνακοινωσώμεθα τὸν λόγον; ‘Do you want to make Socrates here part of our discussion?’

2.2.2 The present in hortative questions

The present (imperfective aspect) generally construes the designated event as unbounded. Such a construal is favoured especially when the verb phrase is atelic (unless there is a special reason to emphasize the boundaries of the event; see above on the aorist). An example is Ar. Thesm. 1193 τί οὐ κατεύδει παρ’ ἐμέ; ‘Why doesn't she sleep with me?’ The verb καθεύδω ‘sleep’ designates an atelic action, and there is no reason why the Scythian would be interested in the temporal boundaries of his intercourse with the girl.Footnote ¹⁷ An unbounded construal is also preferred in the case of habitual occurrences: Pl. Euthyd. 272d καὶ σὺ τί οὐ συμφοιτᾷς; ‘Why don't you too accompany me [as a student to Euthydemus and Dionysiodorus]?’ Socrates is thinking of repeated visits, extending indefinitely forward in time; note also the present infinitive φοιτᾶν in Euthyd. 272c.

I will discuss in more detail one instance where the aspectual interpretation of the present has been criticized by Lloyd (n. 6): Xen. Cyr. 4.1.11 τί οὖν οὐ διώκομεν ὡς τάχιστα, καταδήλων γε οὕτω τῶν ἀγαθῶν ὄντων; ‘Why don't we chase them as soon as possible, as the advantages are so obvious?’ I previously argued ([n. 5], 103) that the chase ‘is naturally presented as unbounded; it is unclear how long it will take and whether the enemy will be caught’. Lloyd (n. 6), 418 objects: ‘This seems rather an unlikely nuance for the soldier to import into an urgent suggestion. Why would anyone making such a proposal want to imply that it is unlikely to have any useful effect?’ But there is a difference between declining to express certainty about the result of a certain course of action, and implying that that course of action is unlikely to have any useful effect. Similarly, if someone were to say Let's try to catch them! the implication of uncertainty conveyed by the word try would not amount to the implication that the effort would be unlikely to be successful. More generally, διώκω ‘chase’ is atelic (a chase can be extended indefinitely), and the imperfective aspect is preferred with this verb: a search in the Thesaurus Linguae Graecae on the indicative forms of this verb in the historiographers Thucydides, Herodotus and Xenophon yields sixty instances of the imperfect against thirteen instances of the aorist.

The boundaries of the designated event may be kept out of focus for contextual reasons. An interesting case is that of Xen. Mem. 3.1.10 τί οὖν οὐ σκοποῦμεν … πῶς ἂν αὐτῶν μὴ διαμαρτάνοιμεν; ‘Why don't we consider how we can avoid being mistaken about these?’ A young man has studied generalship with Dionysiodorus. Socrates points out, however, that he has not learnt the most important thing: to distinguish good from bad men. He suggests making an enquiry. Before they make much progress, however, the young man is unable to answer questions, and Socrates suggests that he go back to Dionysiodorus. It seems that Socrates did not envision a complete enquiry from the start, but wanted to progress as far as necessary for the young man to realize that he had not been taught well by Dionysiodorus.

A category of verbs where we often find the present in hortative questions is that of verbs of ‘saying’. The aspectual variation with such verbs is one of the most elusive problems in classical Greek linguistics. Discussing Xen. Cyr. 4.1.12 τί οὖν … οὐκ ἐλθὼν Κυαξάρῃ λέγεις ταῦτα; ‘Why don't you go and tell Cyaxares this?’, Lloyd argues: ‘There is nothing “unbounded” about the present λέγεις, and it does indeed seem identical from the point of view of aspect to the request of Cyrus to Cyaxares with the aorist quoted above [ἔλεξας in 2.1.4; see note 15].’ But the same would seem to be true, prima facie, of Thucydidean speech introductions with the phrase ἔλεγε τοιάδε ‘he spoke words such as the following’ (ten instances) as compared to those with the phrase ἔλεξε/εἶπε τοιάδε (five); nevertheless, the aspectual variation is there. With respect to the example under discussion: in the following discourse, when the soldiers tell Cyaxares the idea, Xenophon uses the imperfective aspect as well: καὶ ἔλεγον οἷα ἐπιτήδεια ἐδόκουν εἶναι ὑπὲρ ὧν ἐδέοντο ‘and they told him what they thought was expedient with respect to what they were asking for’.

The point is that the aspectual usage of verbs of ‘saying’ in general is difficult to understand; therefore, the fact that an aspectual interpretation of verbs of ‘saying’ in hortative questions does not always conform to our basic intuitions about aspect does not, in itself, refute the validity of such an interpretation. We have to consider the aspectual usage of verbs of ‘saying’ in general and compare that to the variation between the aorist and the present in hortative questions. The use of the imperfective aspect with verbs of ‘saying’ is typically explained in terms of anticipating a reaction, while the perfective aspect has a more definitive force.Footnote ¹⁸ In the case of Xen. Cyr. 4.1.12, the unbounded construal is motivated by the uncertainty of Cyaxares’ reaction: ‘Why don't you tell Cyaxares [and see what he says]?’

I will discuss three further examples that will illustrate how the imperfective aspect can convey different nuances with verbs of ‘saying’. First, Ar. Ach. 358/9–61 τί οὖν οὐ λέγεις, ἐπίξηνον ἐξενεγκὼν θύραζ’, ὅ τι ποτ’, ὦ σχέτλιε, τὸ μέγα τοῦτ’ ἔχεις; ‘Why then don't you bring an executioner's block outside and say what it is, wretch, this big thing that you have?’ The import of the imperfective aspect may be ‘let's see if you have something to say’. Dicaeopolis himself uses the present infinitive λέγειν to refer to his anticipated speech a few lines earlier (355 ἐμοῦ ’θέλοντος ὑπὲρ ἐπιξηνίου λέγειν ‘me being willing to speak with my head over a butcher's block’) and the chorus do the same a few lines later (365 ἐγχείρει λέγειν ‘attempt to speak’).

Second, Pl. Lys. 211d τί οὖν … οὐκ ἐρωτᾷς; ‘Why then don't you ask him?’ Socrates uses the present infinitive to designate the same question in the previous line: ὅδε γάρ τι ὧν λέγω οὐ μανθάνει, ἀλλά φησιν οἴεσθαι Μενέξενον εἰδέναι, καὶ κελεύει τοῦτον ἐρωτᾶν, ‘This man here does not understand something that I am saying, but he says he thinks Menexenus knows, and he urges me to ask him.’ The import of the present ἐρωτᾷς ‘asks’ may be ‘ask him and see if he has something interesting to say’.Footnote ¹⁹

Third, Pl. Grg. 504c τί δὲ οὐκ αὐτὸς λέγεις, ὦ Σώκρατες; ‘Why don't you say it yourself, Socrates?’ Here an ‘immediative’ interpretation of the imperfective aspect seems appropriate.Footnote ²⁰ In the imperative, the immediative use of the imperfective aspect entails that the expression is merely a prompt to carry out the designated action, when the speaker already takes it for granted that this is a desirable course of action. For example, in Dem. 24.32, Demosthenes asks the clerk to take law and read it aloud, using the aorist imperative (ἀνάγνωθι … λαβών). After a relative subordinate clause in which he describes the law, he uses another imperative, but this time the present, ἀναγίγνωσκε, whose value can be rendered ‘now read’.Footnote ²¹ Generally, the immediative use of the imperfective can convey a certain impatience: ‘do X already’.Footnote ²² A similar interpretation is available in the case of Pl. Grg. 504c. Callicles is tired with Socrates’ question-and-answer game and tells him to answer his own question. The use of the imperfective aspect in the hortative question is analogous to the ‘immediative’ use in the imperative: Callicles knows that Socrates knows the answer and asks him to ‘say it himself already’.Footnote ²³

A similarly recalcitrant category of verbs with respect to aspectual usage is constituted by verbs of movement. There is a relevant example in Pl. Prt. 310e ἀλλὰ τί οὐ βαδίζομεν παρ’ αὐτόν, ἵνα ἔνδον καταλάβωμεν; ‘Well, why don't we go to him, so that we may find him inside?’ The phrase ‘go to someone’ is telic, but verbs such as βαδίζω, πορεύομαι and χωρέω are often used in the imperfective, and the aorist of the first is particularly rare (in the classical dramatists and historians, there is only the aorist optative at Ar. Ran. 135).Footnote ²⁴ In Prt. 311a, we find the present subjunctive ἴωμεν ‘let's go’ three times; and when the party finally decides to go in Prt. 314c, we find the imperfect ἐπορευόμεθα.

Finally, we are left with two more difficult instances. First, Ar. Eq. 1207–8 τί οὐ διακρίνεις, Δῆμ’, ὁπότερός ἐστι νῷν | ἀνὴρ ἀμείνων περὶ σὲ καὶ τὴν γαστέρα; ‘Why don't you judge, Demos, who of us two is a better man to you and to your stomach?’ Demos uses the present infinitive κρίνειν in his reply, asking how he will be able to judge in a manner that will seem wise to the audience (1210). This corresponds aspectually with the use of the present in the hortative question. The verb διακρίνω ‘judge’ is telic; the imperfective aspect could be said to focus on the course of action that Demos is to adopt (‘make an enquiry’) rather than on the result (‘make a decision who is better’); but this is admittedly somewhat ad hoc.

Second, Ar. Av. 149–50 τί οὐ τὸν Ἠλεῖον Λέπρεον οἰκίζετον | ἐλθόνθ’; ‘Why don't you go to Lepreum in Elis and settle there?’ This is the most difficult case to explain aspectually. The aorist would seem to be more straightforward, and we find it further on when Pisthetaerus suggests to the Hoopoe that the birds should found a city (172 οἰκίσατε, 173 οἰκίσαιμεν). It may be that the concept of ‘settling’ is conflated with the concept of ‘living’ (οἰκεῖν), which is an atelic situation (note the present forms οἰκοῖτε in line 127 and κατοικεῖν in line 153). Alternatively, the present may convey the nuance ‘try to settle there and see how you like it’.

2.3 Hortative questions, aspect and politeness

The preceding discussion has shown that an aspectual account of the variation between the present and the aorist in hortative questions is perfectly appropriate. That does not mean that it is obvious in every single instance why the perfective or the imperfective aspect is used, but this is due to the difficulty of pinning down classical Greek aspectual usage in general. It may be objected that my approach is then unfalsifiable. While this may be true in terms of the impossibility of a single instance falsifying the entire theory, readers may judge for themselves whether the weight of the evidence as a whole speaks for or against an aspectual account of the variation between the aorist and the present in hortative questions.

Moreover, an interpretation of this variation purely in terms of politeness is open to similar objections. Specifically, a difficulty with Lloyd's account in terms of politeness is the use of the aorist with first-person verbs. Lloyd argues that hortative questions are potentially face-threatening to the addressee, and the aorist serves to mitigate this. However, when the addressee is the speaker himself (Soph. OT 1003), there is no face threat. In the case of the plural, Lloyd argues that the use of the first person (rather than of the second person) serves to mitigate the face threat, so here the present can be used more freely. But aorists are used in such cases as well, and here the criterion of politeness becomes rather too subtle to be meaningful.

Nevertheless, my aim is not to invalidate Lloyd's claim that the aorist in hortative questions is, typically speaking, more polite in tone than the present. Lloyd's discussion of the evidence from a politeness perspective does indeed suggest that this may be a factor. The question is from what semantic properties of the tense forms this pragmatic difference derives. Lloyd suggests that the politeness value of the aorist indicative lies in its being a past tense. He points to the existence of the ‘attitudinal past’: the use of the past tense in questions such as I wanted to ask you something, instead of the more direct I want to ask you something. The problem with adducing the attitudinal past as a parallel is that it is confined to specific contexts and specific verb types, and in those cases it is easy to find an intuitively plausible explanation for why the past tense came to carry the implication of politeness. Such a grammaticalization path from past time value to politeness is hard to postulate in the case of the hortative aorist.

Let us take the example I wanted to ask you something. Why does the use of the past tense convey politeness? We can imagine a scenario where a graduate student is at a conference and spots a professor from another university he has been wanting to speak to. The student walks up to the professor, but feels awkward putting his question to her directly. To mitigate the directness, he does not communicate his current desire, but refers back to an unspecified moment in the past where he felt he wanted to ask the professor a question. We may take this moment to be the moment the student saw the professor and decided to go up to her. The import of the past tense wanted, then, would be something like I saw you and at that moment I was aware of a desire to come and ask you something. Through such scenarios, we may imagine how the past tense with verbs of states of mind or intention came to convey politeness.

Such an explanation is unavailable in the case of the hortative aorist. As I have argued in section 2.1, when the aorist is used as a genuine past tense in questions with τί οὐ, it serves to express indignation at the fact that something did not occur in the past. There is no implication of politeness in such questions. If the hortative use of the aorist derived from this more ‘literal’ use, we would expect the hortative aorist to convey impatience, suggesting that the designated event should have occurred already. This is why the traditional interpretation of the hortative aorist—which Lloyd rightly rejects—is precisely that it conveys impatience.Footnote ²⁵

Another objection is that this account leaves unexplained why there is no hortative imperfect tense (section 2.1). It is unclear why this should be so if politeness in hortative questions is conveyed by the past tense.

When we interpret the hortative aorist not as a past tense but as the perfective counterpart of the imperfective present, the aspectual distinction in hortative questions becomes analogous to the case of the imperative and other moods.Footnote ²⁶ If the aorist is more polite in tone, while the present is more impatient, this can be explained on the grounds that the distinction between the indignant and hortative interpretations is much more difficult to make in the case of the present than in the case of the aorist. With the aorist, there are basically two options: either it refers to something that did not occur in the past and a hortative interpretation is blocked (because the window of opportunity for carrying out the designated action has expired) or it is hortative. However, questions in the present such as τί οὐκ ἀποκρίνῃ; are ambiguous between ‘What is your deal that you are not answering’ and ‘I suggest you answer’. Therefore, the present may be felt to be more impatient in tone than the aorist. Another factor is the immediative interpretation that is open with the present, which also may convey a note of impatience (section 2.2.2).

3.1 Introduction

In this section I revisit the tragic aorist. The arguments will be in the same vein as in the previous sections, so I will be more brief here. To begin with, I fully agree with Lloyd's analysis of the tragic aorist in terms of politeness. At the same time, I believe that, semantically, the difference between the tragic aorist and its present competitor is aspectual. Lloyd has made no attempt to falsify this explanation by considering the broader evidence for the aspectual behaviour of the verbs in question, so I will present a case study of a key verb in this connection: ὄμνυμι ‘swear’. First, however, some preliminary remarks need to be made about the nature of the utterances in question.

3.2 Two groups

Lloyd argues that tragic aorists are always performative. Performative verbs are verbs in utterances that do not describe something but perform an act. For example, to say I promise is to perform the act of making a promise; to say I swear is to perform the act of swearing.

Lloyd's cases of the tragic aorist can be classed under two headings (these do not entirely correspond to his ‘notional actions’ vs ‘polite aorists’). The first concerns cases where the verb is not in itself performative. These are cases such as ἥσθην ‘I enjoy’, ἀπέπτυσα ‘I spit out’, ὤμωξα ‘I lament’. The verbs ἥδομαι ‘enjoy’, ἀποπτύω ‘spit out’ and οἰμώζω ‘lament’ are not performatives in themselves; for example, in Eur. IA 1613 ὡς ἥδομαί τοι ταῦτ’ ἀκούσασ’ ἀγγέλου ‘How pleased I am to hear that from the messenger’ the verb ἥδομαι describes the attitude of the speaker (the chorus); in Aesch. Cho. 1014 νῦν αὐτὸν αἰνῶ, νῦν ἀποιμώζω παρών ‘Now, being present, I praise him, now I bewail him’ the verb ἀποιμώζω describes Orestes’ present feelings regarding his father.

However, in the aorist, Lloyd argues, these verbs can function as performatives. For example, the expression ἥσθην ‘I enjoy’ can be used as a restrained alternative to actually laughing, like saying Ha ha. Similarly, ὤμωξα ‘I lament’ is a restrained alternative to crying οἴμοι and ἀπέπτυσα ‘I spit out’ is an alternative to actually spitting. Thus these expressions have the value of actions.Footnote ²⁷

In my view, the performative function of the aorist is to be related to its perfective aspect.Footnote ²⁸ Because the perfective aspect limits the designated event to the boundaries of the utterance itself, the utterance and the event become one and the same. Bary (n. 4) provides evidence from other languages to establish the link between performativity and perfective aspect. According to Lloyd, on the other hand, the performative aorist is a kind of ‘attitudinal past’. But it is unclear why the past tense should be polite with performative verbs, and if tense is the key factor, it is unclear why there is no performative imperfect.

The second group concerns cases where the verb is inherently performative, such as ὄμνυμι ‘swear’ or ἐπαινέω ‘praise’: saying I swear constitutes an act of swearing, and saying I praise constitutes an act of praising. In these cases both the present and the aorist are performative. According to Lloyd, the aorist is more formal or reserved in tone than the present. Bary, on the other hand, argues that there is no difference: both the primary present indicative and the secondary aorist indicative are suboptimal forms for the non-existent primary aorist indicative.

Lloyd's explanation for the difference in tone between the aorist and the present is convincing. For example, we find the performative aorist in Eur. IT 1023, where Iphigeneia rejects Orestes’ proposal to kill Thoas: οὐκ ἂν δυναίμην⋅ τὸ δὲ πρόθυμον ᾔνεσα ‘I would not be able to; but I praise your enthusiasm’. Here the praise is merely a matter of courtesy intended to mitigate the rejection. We find the present, by contrast, in Eur. Phoen. 1683, where Oedipus seems to reject Antigone's offer to join him in exile: ὦ θύγατερ, αἰνῶ μέν σε τῆς προθυμίας ‘Daughter, I praise you for your enthusiasm’. Here the praise can plausibly be interpreted as more heartfelt: Oedipus is touched that his daughter wants to join him in exile. Oedipus does not explicitly reject Antigone's offer, and, indeed, ultimately accepts it.

This difference, however, is a matter of aspect, not of tense. The aorist is more reserved in tone because it binds the verbal content to the utterance itself: as soon as the utterance has been made, the matter is dropped.Footnote ²⁹ The imperfective aspect of the present, on the other hand, signals that the designated event transcends the moment of speech in its duration, suggesting a greater commitment on the part of the speaker to the designated verbal content. This is analogous to how, in English, the progressive of a performative verb seems to convey stronger insistence than the simple present: I am begging you vs I beg you.

If this account is correct, then it should be corroborated by the evidence of the aspectual behaviour of these verbs in categories other than the indicative. I argue that this is in fact the case, taking the imperative of ὄμνυμι ‘swear’ as the prime example.

3.3 The case of ὄμνυμι ‘swear’

According to Lloyd (n. 2), 30, ‘the most solemn and explicit oaths in Greek drama’ are in the present tense. For example, when, in Eur. Med. 752–3, Aegeus swears to Medea that he will protect her, he uses the present tense in an emphatic oath: ὄμνυμι Γαῖαν φῶς τε λαμπρὸν Ἡλίου | θεούς τε πάντας ἐμμενεῖν ἅ σου κλύω ‘I swear by the earth and by the shining light of the sun and by all the gods that I will abide by your words’. Oaths performed with the aorist are more perfunctory, as in Eur. Or. 1516 τὴν ἐμὴν ψυχὴν κατώμοσ’, ἣν ἂν εὐορκοῖμ’ ἐγώ ‘I swear by my life, an oath I would keep’.Footnote ³⁰

This difference is mirrored in the imperative; I have found five instances of (κατ)όμοσον and three instances of ὄμνυ in the dramatists. To begin with, Aegeus’ use of the performative present ὄμνυμι in Med. 752 corresponds to Medea's use of the present imperative ὄμνυ some lines earlier (746–7): ὄμνυ πέδον Γῆς, πατέρα θ’ Ἥλιον πατρὸς | τοὐμοῦ, θεῶν τε συντιθεὶς ἅπαν γένος ‘Swear by the plain of Earth, by Helios, the father of my father, and add the entire race of the gods’. Conversely, the Phrygian who makes his oath with the performative aorist in Eur. Or. 1516 was instructed to do so with the aorist imperative: ὄμοσον—εἰ δὲ μή, κτενῶ σε—μὴ λέγειν ἐμὴν χάριν ‘Swear—if you will not, I will kill you—that you do not say that to please me’. Finally, there is a similar correspondence in Soph. Trach. 1185–8:

Heracles uses the present imperative to instruct Hyllus to swear an oath, and Hyllus answers with a present tense performative.Footnote ³¹

In one case, there is a discrepancy between the performative verb and the corresponding imperative (Ar. Av. 444–7):

Pisthetaerus uses the aorist imperative κατόμοσον: his request is not particularly emphatic (he does not mention gods or conditions), in contrast with those of Medea and Heracles discussed above. The chorus, on the other hand, take the opportunity to make an emphatic oath with the present, which takes up two and a half lines and includes a self-curse.Footnote ³²

Three cases of the aorist imperative are found in Ar. Ran. 305–6:

Dionysus is not asking for a full, official oath: he merely wants Xanthias to utter an affirmative. In so far as he is not satisfied with Xanthias’ oaths, it is not because he desires a fuller one; he simply wants multiple affirmations.

We are left with one instance of the imperative: Eur. IT 743 ὄμνυ⋅ σὺ δ’ ἔξαρχ’ ὅρκον ὅστις εὐσεβής ‘Swear: you, begin an oath which is pious’. The phrase ἔξαρχ’ ὅρκον ‘begin an oath’ suggests a full-fledged ritualistic oath.Footnote ³³

In conclusion, the aspectual variation between the aorist and the present in the performative first-person indicative of the verb ὄμνυμι ‘swear’ is mirrored by the aspectual variation of the same verb in the imperative.

3.4 The genre issue

At the end of his article, Lloyd (n. 6), 424 puts forward a theoretical objection to an explanation of the tragic aorist in terms of aspect: that such an explanation ‘integrate[s] [the performative aorist] into a general account of the Greek aorist so successfully that one would expect it to be much more widespread than it actually is’. He points out that ‘the performative aorist is only used in a limited stylistic and chronological context’.

This criticism is valid in so far as Bary's account is concerned. She argues that there is no difference between the performative aorist and its present counterpart, and under that account we would indeed expect the distribution between performative aorists and presents to be about fifty–fifty, which is not what we find. Lloyd is right that there is a difference and that the performative aorist has a distinct semantic and pragmatic value. While my account is not vulnerable to this objection to the same extent that Bary's is, the potential criticism remains that it may seem to make a grammatical normality out of a rare usage. Let me offer three thoughts on this point.

First, I do not think the rarity of the usage of the performative aorist affects the discussion concerning its semantics. The politeness value of the performative aorist must derive from its grammatical properties one way or the other. Why would it be acceptable for the past tense to be used in a polite manner only in certain genres but not for the perfective aspect? We can maintain both that the performative aorist is bound to certain genres and that its use is aspectually motivated.

Second, I agree with Colvin (n. 3), 116–17 that ‘it should be clear from looking at the examples that the phenomenon is grammaticalized, and calls for a different order of explanation [different than one in terms of style]’.

Third, and most importantly, we should be careful in our evaluation of the negative evidence. Take, for example, the following observation by Lloyd (n. 6), 423–4: ‘To take just three verbs, there are no examples in Plato and Xenophon of the performative aorist of ὄμνυμι (‘I swear’), δέχομαι (‘I accept’) or ἐπαινῶ (‘I approve’), although there are numerous opportunities in their works for swearing, accepting and thanking.’ What needs to be shown for this argument to be effective, however, is the following:

1. a. That first-person forms of these verbs are used in the first place. The fact that the aorist is not used is meaningless if the present is not used either.

2. b. That, if we do find first-person present forms with these verbs, the consistent use of the present instead of the aorist cannot be explained in terms of the different values of the two categories. There may be swearing, accepting and approving, but not in the reserved tone that warrants the use of the performative aorist.

With regard to the second point, Colvin (n. 3), 118 argues that Platonic dialogue ‘lacks the interactive and reactive features of dramatic dialogue, and the linguistic markers of (realistic) turn-taking’. This would account at least partly for the absence of performative aorists in these texts. Lloyd (n. 6), 424 n. 31 objects: ‘[B]ut such passages as the beginning of Lysis, to take one example from many, are likely to be closer to natural dialogue than is the verse of fifth-century tragedy.’ The question is: do we find swearing, accepting or approving in those passages in Plato? In the beginning of the Lysis, we do not. And when we do, can the use of the present not be explained in terms of the unreserved tone it conveys?

A brief look at the evidence for ὄμνυμι in Plato and Xenophon will illustrate this point. There is only one instance of ὄμνυμι in Plato (Phdr. 236e); it does not stand at the beginning of a speaker's turn, and the fulness of the oath fits the profile of the present as discussed by Lloyd (see section 3.3). In Xenophon, we find eight instances, none at the beginning of a turn—except in An. 6.6.17, where Hagasias rises to speak and makes an oath without being prompted:

Neither this nor any other oath in Xenophon is of the perfunctory nature that explains the use of the aorist in some cases in drama. Overall, I believe that Lloyd's objection to an aspectual account of the performative aorist on the ground that the phenomenon is limited to certain genres is invalid.