3.1.1 Mimesis, Diegesis and Experientiality

Scenic narrative and summary narrative are flexible categories. To begin with, there is no specific distance between discourse time and story time that defines the boundary between the two. Moreover, the category of summary narrative is extremely heterogeneous, because the distance between discourse time and story time may become infinitely large. For example, a paragraph of discourse may cover hours, days, weeks, months, years, millennia and so on. In reality, then, the distance between discourse time and story time is a continuous variable, and the distinction between two categories is mainly useful in so far as the category of scenic narrative is relatively circumscribed.

The point is that summary narrative may exhibit different degrees of experientiality. This depends mainly on the distance between discourse time and story time and the concreteness of the designated events. However, these factors can also be manipulated through linguistic construal: temporally extended and internally complex events can be described in a way that makes them more easily processible in terms of direct experience.Footnote ⁵ Consider the following passage:

While I would not consider this a typical example of summary narrative (the unity of time, place and action is strong), it would be problematic to classify it as scenic. We can be sure that the meeting between the two armies, the cavalry charge, the flight of the Pylians to the hill and the final defeat of the Pylians on the hill, took a substantial amount of time to occur. However, the events are construed in such a way as to minimise the relevance of the discrepancy between discourse time and story time. We get the impression that there is a quick succession of concrete and dynamic events, which is relatable in terms of ordinary experience.

For example, the part of (2) that tells of the encounter and the cavalry charge may be compared to the following passage:

As in (2), there is a chance meeting (compare the present περιτυγχάνομεν [‘happen upon’] in [3] to περιτυγχάνουσι [‘happen upon’] in [2]) and an attack (προσπίπτει [‘falls upon’] in [3], ἐμβάλλουσι [‘charge’] in [2]). The only difference is that the events are ‘downscaled’ in (3): they take up less time, there are fewer participants and they occupy a smaller space. The effect of the presentation of the events in (2) is to convey the sense of immediacy associated with scenic narrative, which is what we observe in (3). Similarly, the present ἐκκόπτουσι (‘knock away’) in (2) construes an internally complex event in terms of simple manual action.Footnote ⁶ All in all, the use of the present for preterite in (2) cannot be unambiguously assigned to either the mimetic or the diegetic category: it has aspects of both.

Further along the scale of diegesis are passages that seem more like summaries than actual narratives. Consider the following example:

Here the richness of reality is reduced to a ‘mere backbone’ of a story (Fludernik [Reference Fludernik1996: 28]). As dry and uninvolved as this passage may seem, however, aspects of experientiality may be identified: human agency, intentional action and action-reaction patterns.Footnote ⁷ Fludernik (Reference Fludernik1996: 29) argues that ‘[h]uman experience typically embraces goal-oriented behaviour and activity, with its reaction to obstacles along the way’ and that a ‘three-part schema of “situation-event(incidence)-reaction to event” … constitutes the core of all human action experience’. The narrative in (4) appeals to such experiential schemas: Tissaphernes’ accusation initiates an action-reaction sequence where Artaxerxes arrests Cyrus, which prompts the mother to rescue her son. A similar sequence in a scenic context can be found in Euripides, Heracles 994–7, which I discussed in Chapter 2, Section 2.6.2: Heracles goes off to kill his third son, but the mother snatches him away to another room and closes the door. Through such an association, the narrative in (4) arguably still retains something of the character of embodied experience, although in highly attenuated form.

To conclude, the mimetic present in scenic narrative and the diegetic present in summary narrative should be considered prototype categories. The relation between these two categories may be understood in terms of a balance between the parameters of experientiality and communicative dynamism. In Chapter 2, I argued that the mimetic present should be explained mainly in terms of experientiality, with communicative dynamism acting as a moderating influence on its use. In summary narrative, on the other hand, the ‘dynamics of experientiality’ are, to a greater or lesser degree, ‘cancelled’ (Fludernik [Reference Fludernik1996: 28]), and the function of the present for preterite is to be understood mainly in terms of the impact of the designated event on the story structure. At the same time, experientiality may remain a factor that finds expression in such variables as animacy of the subject, voice, number, et cetera (Allan [Reference Allan and Lallot2011a: 40–2], with reference to Wårvik [Reference Wårvik and Virtanen2004]). In this chapter, however, I will not focus on those aspects.

3.1.2 Selection of Material

In Sections 3.3–3.5 of this chapter, I will present quantitative analyses to support my claim that the diegetic present is structurally associated with certain discourse functions. In Chapter 2, I tried to isolate potentially mimetic uses of the present for preterite by selecting a corpus of eyewitness narratives in drama (Chapter 2, Section 2.1.1.1). The diegetic use of the present for preterite is associated with summary narrative, and this is what we typically find in historiography.

I have selected as my corpus Thucydides’ Histories,Footnote ⁸ Xenophon’s Hellenic affairs (which continues Thucydides’ work) and Xenophon’s Expedition of Cyrus (perhaps better known by its Greek title Anabasis). Two main omissions deserve mention. First, I have excluded Herodotus’ Histories. The problem with this work is that it exhibits a great variety of narrative styles, which makes it harder to zoom in on a particular usage of the present for preterite here. Also, I have excluded Xenophon’s Education of Cyrus because it is often anecdotal and educational rather than historiographical.Footnote ⁹

I have included all present for preterite forms in the selected texts, except those that fit the criteria for exclusion as specified in Section 2.1.1.2 of Chapter 2 and in the Appendix.Footnote ¹⁰ These amount to a total of 1,190 (Thucydides’ Histories: 546; Xenophon’s Hellenic affairs: 265; Expedition of Cyrus: 379). For the contrastive corpus, I have selected the first 50 aorist forms in each book of these works that did not fit the criteria for exclusion. These amount to 400 for Thucydides and 350 for each of the works of Xenophon, to a total of 1,100 aorist forms.

I have discussed the issues associated with performing quantitative analysis on Classical Greek texts in Section 2.1.1.3 of Chapter 2. It should at all times be remembered that the p-values reported here have not been corrected in any way and that they only represent the intrinsic probability of the occurrence of a reported distribution. As I have explained, the point of this is to give an indication of the markedness of a distribution as a function of the effect size and the quantity of the data.

Another issue is the heterogeneity of the data: each of the three historiographical works constitutes its own group. While Thucydides’ Histories and Xenophon’s Hellenic affairs seem similar in character with respect to the mimesis-diegesis opposition, Xenophon’s Expedition of Cyrus is generally more vivid. I therefore present the results for the individual works, except when there are very few cases in one of the categories of the predictor variable (as in Section 3.4.3).

3.2.1 Demonstratives and the Discourse in Classical Greek

Unlike English and other Western-European languages, the Classical Greek system of demonstratives has a tripartite division along the proximal-distal axis (e.g., Ruijgh [Reference Ruijgh, de Jong and Rijksbaron2006]).Footnote ¹³ These categories are typically characterised in terms of I-deixis, you-deixis and him-deixis. (This goes back to Brugmann [Reference Brugmann1904].) Demonstratives of the first category designate immediate proximity to the speaker and are translated with ‘this’. The morphological marker in Classical Greek is the suffix -δε, as in τό-δε (‘this [thing])’, ἐνθέν-δε (‘from here’). For the second category, I use the term relative proximity, as it focuses on the accessibility of the designated entity to the addressee. In English, these demonstratives can be translated with ‘this’ or ‘that’ depending on the context (for clarity’s sake I have consistently translated with ‘that’). These forms can be recognised by the morpheme του-, ταυ-, or τευ-, as in τοῦ-το (‘that [thing]’) ἐν-τεῦ-θεν (‘from there’). Finally, demonstratives derived from the basic adverb ἐκεῖ (‘there’), e.g., ἐκεῖ-νο (‘that [thing]’), ἐκεῖ-θεν (‘from there’), designate distance from the speaker and addressee and are translated with ‘that’.Footnote ¹⁴

Our main concern here is how this system relates to discourse deixis, which is concerned not with the actual spatio-temporal proximity of entities but with the accessibility of referents in the discourse. The distinction that is important to my argument is that between immediate and relative proximity. When referring to entities that are accessible from the previous discourse, the unmarked option is a demonstrative designating relative proximity.Footnote ¹⁵ Consider the following example:

The use of the pronoun ἐντεῦθεν (‘from there’) signals that the designated time frame needs to be retrieved from the preceding discourse. It thus signals a measure of distance to this time frame. Contrast the reference to the beginning of the same war in Thucydides’ Histories:

As I argued in Chapter 1, Section 1.7, the use of the demonstrative ἐνθένδε (‘from here’), designating immediate proximity, reflects the strong involvement of the narrator with the present development in the discourse.Footnote ¹⁶ The use of this adverb to mark discourse progression is exceptional in Classical Greek prose.Footnote ¹⁷ I suggest that its use here can be explained in terms of the exceptional discourse status of the designated event, that is, the outbreak of the Peloponnesian War, which is the subject of the entire treatise. The present tense form ἄρχεται (‘begins’) further highlights the present accessibility of the war in the medium of the discourse.

As I argued in Chapter 1, this type of accessibility of distal entities through an intervening medium should be distinguished from the type of accessibility that is achieved through mental displacement. In this respect, I disagree with Ruijgh’s (Reference Ruijgh, de Jong and Rijksbaron2006) account, which is otherwise illuminating, when he discusses a passage in Sophocles’ King Oedipus. Jocasta has just mentioned that King Laius was killed at a fork in the road (716). This leads Oedipus to the realisation that he himself was the killer (726–48). He then tells how it happened. In this narrative, we find the following passage:

Ruijgh (Reference Ruijgh, de Jong and Rijksbaron2006: 152) cites with approval earlier commentators who argued that the use of the demonstrative τούσδε (‘these’) suggests that the place is before Oedipus’ mind’s eye. As Ruijgh has it, ‘Oedipus relives with anxiety the scene he describes’. This is supposedly corroborated by the ‘vivid’ use of the present for preterite ἱκνοῦμαι (‘arrive’). I agree that the use of the expression τούσδε τοὺς χώρους (‘this place’) reflects the close involvement of Oedipus with the designated entity: the location of the murder has been a discourse topic for some time, and it is crucial to Oedipus’ realisation that he is the killer. But this does not imply mental displacement to the designated location. Similarly, the function of the present for preterite ἱκνοῦμαι (‘arrive’) can be explained as discourse-structural, introducing a new phase in the narrative (see Sections 3.2.2 and 3.4.1).Footnote ¹⁸ In the absence of positive linguistic cues for a displacement scenario, such an interpretation is more economical.

3.2.2 Narrative Structure: Beginnings and Culminations

In the previous section, I discussed Thucydides’ use of the adverb ἐνθένδε (‘from here’) to introduce new developments in the narrative. Here I will discuss a similar use of a demonstrative designating immediate proximity in the context of a local episode. This will illustrate the importance of discourse structure to the dynamics of tense-switching. At some point in the second book, Thucydides narrates a military expedition undertaken by the Ampraciots against Amphilochia:

This is followed by a digression in which Thucydides explains whence the enmity between these peoples originated (2.68.2–8). After this, the narrative is resumed in the following manner:

In the passage where the expedition is first mentioned, the reference is unambiguously to the past event space, as signalled by the adverbial phrase κατὰ δὲ τοὺς αὐτοὺς χρόνους (‘around the same time’) and the past tense form ἐστράτευσαν (‘made an expedition’). After the digression, however, things become more interesting. We still have an adverbial phrase designating the past event space: ὕστερον δὲ ἐν τῷ πολέμῳ (‘later in the war’). At the same time, the demonstrative τήνδε (‘this’) highlights the immediate accessibility of the expedition in the discourse: it is this expedition we are now talking about. The present for preterite ποιοῦνται (‘make’) designates the present ‘occurrence’ of this expedition in the discourse. The variation between the present here and the aorist at 2.68.1 reflects the difference in discourse status of the designated expedition in the two instances. At 2.68.1, Thucydides mentions that there was an expedition but immediately breaks off the narrative to provide background information to the conflict. Only at 2.68.9 does the narrative of the expedition begin. I submit that this sharp discourse-structural relief motivates the use of the demonstrative τήνδε (‘this’) and the present tense form ποιοῦνται (‘make’) here.Footnote ¹⁹

My claim is that the present for preterite by itself can carry the implication that the past event is presently accessible in the discourse. Consider the following example:

As in 2.68.9, the present tense form στρατεύει (‘makes an expedition’) signals the start of a military expedition. By highlighting the ‘present occurrence’ of the expedition in the story, Thucydides signals to his addressees that they are to update their mental model of the discourse at this start of new developments.

The following pair of examples illustrates the discourse-as-representation construal at the culmination of narrative developments. First, consider the following passage:

The chorus of Egyptian suppliants introduce the story of how Zeus had intercourse with Io. The king of Athens affirms he knows the story. In his reaction to the chorus in line 296, he uses the past tense form ἦν (‘were’) to designate the past event space. The pronoun ταῦτα (‘those’), designating relative proximity, similarly suggests a measure of distance to the events in the story (‘those entanglements you just mentioned’). By contrast, the chorus, in the next line, highlight the immediate accessibility of the dispute between Zeus and Hera through the medium of the discourse, both by the pronoun τάδε (‘this’) and by the present tense τελευτᾷ (‘ends’).Footnote ²⁰ This difference in referential strategy relates to two aspects. First, the chorus are talking about an important change in the structure of the story, that is, a culminating point, while the king refers to a situation. Second, the story is of crucial rhetorical importance to the chorus, as their descent from Io (the ‘mortal’ from the story) legitimises their present claim for asylum.Footnote ²¹ The king, on the other hand, still unsure whether to grant the chorus their wish, is more reserved.

Now compare the use of the present for preterite to highlight the climax of a dispute in the following passage:

The expression λήγει ἔρις (‘the dispute ends’) is, for all present purposes, synonymous to τελευτᾷ νείκη (‘[this] dispute ends’) in example (13). In that case, we saw how the immediate accessibility of this event in the discourse was highlighted by the demonstrative τάδε (‘this’). In (14), the present for preterite λήγει (‘ends’) carries this implication by itself. Its function is to signal a significant update to our mental model of the discourse, that is, the culmination of the current discourse segment.

3.2.3 Introducing Referents

The introduction of new referents impacts our mental model of the discourse in analogous fashion to the introduction of new narrative developments. A salient example of the use of demonstratives at the introduction of new referents is the following:

Demosthenes tells Nicias of the people who will rise to political prominence in the future. The first new referent introduced into the discourse is an oakum-seller. In his reaction, Nicias uses demonstratives designating relative proximity to designate the new referent (131 οὑτοσί [‘that (man) there’]) and to mark discourse progression (τοὐντεῦθεν [‘from there’]). When the second participant is introduced, however, Nicias switches to demonstratives designating immediate proximity (133 τώδε [‘these two (men)’], τόνδε [‘this (man)’]). The switch seems to reflect Nicias’ increased engagement with the discourse at this point, as it moves closer to the part that is actually relevant to his present situation. Moreover, at 131, Nicias immediately moves on, while here he asks for further information about the new referent. In any case, Nicias’ interruptions show how keeping track of referents is a processing task that occurs ‘here and now’ during the unfolding of the discourse. My argument is that such changes in the status of referents render the discourse structure cognitively salient, which facilitates the use of the present for preterite.Footnote ²²

Two pairs of passages will serve to make the connection between the accessibility of referents in the discourse and the use of the present for preterite more evident. In the following example, Agamemnon, stranded at Aulis, tells the audience the events leading up to the expedition to Troy:

The demonstrative ὅδε (‘this man’) in line 71 signals that the introduction of the referent is something that should be immediately attended to: Paris is ‘here’ as he enters the discourse at this point.Footnote ²³ Now compare the following instance of the present for preterite introducing a new referent:

The present tense marks the arrival of Hegesandrus, both as a character on the scene in the story world, and as a referent in the discourse. Like Paris in example (16), Hegesandrus will be the cause of conflict in the narrative. The familiarity of the audience with this new discourse referent, which Aeschines emphasises, is another similarity.

For my second pair, I reproduce example (6) cited above:

Aegeus asks Medea to update the discourse by identifying a referent whose existence he infers from her preceding assertion. At this point, Aegeus explicitly evokes the λόγος (‘discourse’). A comparable use of the present for preterite is found, again, in Aeschines’ speech Against Timarchus (compare example [17]). Aeschines seeks to establish that Timarchus has prostituted himself to different men in Athens. At the beginning of the narrative, Aeschines announces he will focus on the men in whose house Timarchus has lived (section 40 of the speech). The second of these men is a certain Anticles, and he is introduced in the following manner:

As at Medea 701 (example [18]), the present for preterite serves to identify a referent whose existence is presupposed in the discourse (because Misgolas was only the first to take up Timarchus). It is instructive to contrast the use of the aorist of the same verb not much later in the same speech, when another person takes Timarchus into his house:

Here, the assertion containing the verb ἀνέλαβεν (‘took up’) is lower in information status than was the case with ἀναλαμβάνει (‘takes up’) in (19), because the new referent has already been introduced in a separate, non-narrative clause.Footnote ²⁴

3.2.4 Conclusion

In this section, I have presented a theoretical account of the diegetic present for preterite in terms of the discourse as a representation space: the designated events are construed as immediately accessible through the medium of the discourse. Through this construal, the present for preterite signals to the addressees that they are to update their mental model of the discourse in the light of salient developments. Such developments concern either the segmentation of the discourse in terms of narrative structure or the status of referents.

In the next sections, I will argue in more detail that the present for preterite is structurally associated with certain pragmatic functions relating to discourse organization. In Section 3.3, I will discuss the use of the present for preterite in combination with certain attention-management strategies. In Sections 3.4 and 3.5, I will focus on two specific discourse-structural functions of the present for preterite: marking changes in the narrative dynamic (compare Section 3.2.2) and marking changes in the status of referents (compare Section 3.2.3).

3.3.1 Announcements and Cataphoric Reference

One way to call attention to the next event in the narrative is to anticipate it with an abstract reference. Consider the following two examples:

In both cases, we are not told immediately what actually happened, but the narrator heightens our anticipation by announcing the next event and emphasizing its remarkable character ([21] τὴν μεγίστην ἐπιβουλήν [‘their greatest scheme’]; [22] πάντων δεινότατον [‘the most terrible thing of all’]). These are moments of high narratorial mediation, and the present tense forms ἐπιβουλεύουσι (‘contrive’) and διοικεῖται (‘manages’) serve to focus the joint attention of the narrator and addressees on the next event in the discourse.Footnote ²⁶

In similar introductions, we often find demonstratives designating immediate proximity used as cataphoric pronouns. Consider the following passage from Xenophon’s Expedition of Cyrus, which marks a transition from one book to another:

This and other summary-like passages in this work are considered spurious, but that does not matter much to my argument. The point here is the variation between the past tense forms ἐγένετο (‘happened’) and ἐπολεμήθη (‘war was waged’), on the one hand, and the present ποιοῦνται (‘make’), on the other. All these verb forms designate events that actually took place in the past. With respect to the discourse space, however, a distinction is made between what has already been narrated (and is thus past in the discourse – note ἐν τῷ πρόσθεν λόγῳ [‘in the previous discourse’]) and that which will be narrated presently. The present tense form ποιοῦνται (‘make’), together with the demonstrative ὧδε (‘in this [i.e., the following] way’), signals that the following developments are the new focus of attention.

In example (23), the context makes clear that the reference of the present tense is to the discourse space, but in other cases, this is implicit. In the following passage, we have just been told that Xenophon had a portentous dream. Then the narrator comments:

The narrator invites the audience to ‘consider’ (σκοπεῖν) what happens next. The phrase γίγνεται γὰρ τάδε (‘for this [i.e., the following] happens’) reinforces this strategy of directing the audience’s attention to the next event in the narrative.Footnote ²⁷

3.3.2 Interacting with the Audience

A second class of attention-management strategies involves explicitly addressing the audience. In oratory, we find the present for preterite a number of times in rhetorical questions of the type Then what does X do? An example:

The speaker characterises what happens in the next episode as ‘a most outrageous plot’ (πάντων δεινότατον πρᾶγμα) and notes that it is ‘worth paying attention to’ (ἄξιόν ἐστι προσέχειν τὸν νοῦν). In the narrative proper, he reinforces this strategy by focusing attention with a rhetorical question: καὶ τί ποιοῦσιν; (‘and what do they do?’) This question conveys the pretence that the construction of the narrative is a joint activity between the narrator and his audience. The use of the present tense underscores this joint engagement with the story.Footnote ²⁸

It is instructive here to compare an instance of the present for preterite where the narratee actually participates in the construction of the narrative (‘narrative dialogue’, see Schuren [Reference Schuren2014]):

The use of the present for preterite in the question τί οὖν δὴ γίνεται; (‘so what happens?’) reflects the close engagement of the narratee with the evolving discourse.

Another way to call for the audience’s attention is to use an imperative (compare example [5]). In the following example, we find the combined use of an imperative and a rhetorical question (see also Nijk [Reference Nijk2013a]):

First, Demosthenes uses imperfect forms to describe Philip’s considerations. Then the rhetorical question τί οὖν; (‘what then?’) anticipates his next move. The phrase θεάσασθ’ ὡς εὖ (‘watch how cleverly!’) serves to focus the audience’s attention on the significance of this event; the present tense ἐπιχειρεῖ (‘tries’) correspondingly highlights the impact of this event on the structure of the story.Footnote ²⁹ The use of the verb θεάομαι (‘watch’) is interesting here. The verb does not invite the audience to literally ‘visualise’ the designated event. Indeed, it would be hard to see how Philip’s scheming here might be visualised at all. Rather, the concept of ‘vision’ is used metaphorically to express the immediate accessibility of the designated event through the medium of the discourse.Footnote ³⁰ That is, the cleverness of Philip’s act is available for close consideration in the story that Demosthenes is now telling.Footnote ³¹

3.3.3 Sentence Complexity

A more implicit way to manage attention is to use subordination (‘hypotaxis‘) to achieve a ‘hierarchical ordering of event-presentation’ (Toolan [Reference Toolan1988: 126]). The effect of subordinate clauses in a periodic sentence is to foreground the main clause assertion, and this effect is stronger as the sentence becomes more complex. This is illustrated by the following example:

By delaying the main clause for so long, Thucydides heightens the audience’s expectations: what will the Potideans finally do? This puts the main clause assertion into strong focus, and the effect is further enhanced by means of the phrase τότε δὴ κατὰ τὸν καιρὸν τοῦτον (‘then at that moment’) (for the particle δή [‘then’, ‘so’], see Section 3.3.4).Footnote ³²

I hypothesise, then, that clause complexity increases the odds of the present for preterite being used in the main clause. In counting subordinate clauses, I have disregarded those that followed the main clause verb (in example [28], the participial clause ξυνομόσαντες κοινῇ [‘making common cause with them’]). This is because I am concerned with the anticipation of the main clause assertion. I have counted three types of subordinate clauses:

1. (a) Conjunct participial clauses. These are participial clauses where the subject is coreferential with the subject of the main clause. For example, in example (28): ἐλθόντες δὲ καὶ ἐς τὴν Λακεδαίμονα μετὰ Κορινθίων (‘and going to Lacedaemon as well together with the Corinthians’).

2. (b) Absolute participial clauses. These are participial clauses where the subject is not coreferential with the subject of the main clause. These are in the genitive case or, rarely (with impersonal constructions), in the accusative. For example, βουλομένων δὲ τῶν τριάκοντα ἀποτειχίζειν (‘the Thirty wishing to wall off the place’; Xenophon, Hellenic affairs 2.4.3).Footnote ³³

3. (c) Temporal subordinate clauses introduced by ἐπεί/ἐπειδή (‘when’), ὡς (‘as’), ἡνίκα (‘when’), ὅτε (‘when’).

I give the results for the total number of combined subordinate clauses per work. This number ranged from 0 to 6. In the uppermost regions, I have found only the present for preterite: four instances with six subordinate clauses, seven instances with five subordinate clauses.Footnote ³⁴ This finding is highly suggestive of a positive influence of clause complexity on present for preterite usage. I have excluded these cases from further analysis, however, to avoid zero counts for the preterite in the data (because the resulting infinite odds value is problematic in logistic regression). Moreover, in the case of Xenophon’s Expedition of Cyrus, there were only six cases with more than two subordinate clauses (five with three, and one with four). This is problematic because it results in low expected counts for both categories. Therefore, I have only included cases in which up to two subordinate clauses are used.

The results for each work are listed in Table 3.1.

Table 3.1Tense and combined subordinate clauses

^a B = .365 (constant = .009), standard error = .089, N = 937. I have checked the assumption of the linearity of the logit by looking at the interaction effect between the variable and a log-transformed copy of itself (Hosmer and Lemeshow [1989]). The p-value for this effect is .468, which does not approach statistical significance; therefore, I conclude that the assumption has been met.

^b B = .371 (constant = -.583), standard error = .097, N = 613. The p-value for the interaction effect between the variable and its log-transformed copy (see note ^a) is .357.

^c B = .287 (constant = -.037), standard error = .126, N =722. The p-value for the interaction effect between the variable and its log-transformed copy (see note ^a) is .710.

It should be understood that the odds ratio here expresses the average increase in odds of the present being used rather than the aorist for each additional subordinate clause. Thucydides’ Histories and Xenophon’s Hellenic affairs exhibit a highly similar trend (odds ratios of 1.44 and 1.449, respectively, with low p-values), which supports the hypothesis. In the Expedition of Cyrus, the observed effect is smaller (odds ratio 1.333) and statistically less certain (p = 0.022). As I explained in Section 3.1.1, such a difference may be expected because of the different character of the works. The Expedition of Cyrus is generally more vivid in character, and proliferation of subordinate clauses, as in example (28), is hardly ever observed in this work. It thus makes sense that other factors will be more central to tense-switching here.

3.3.4 Particle δή (‘Then’, ‘So’) Marking Discourse Progression

Greek narrators sometimes use the particle δή (‘then’, ‘so’) in order to ‘move forward narration by marking major steps in an account’ (Bonifazi et al. [Reference Bonifazi, Drummen and de Kreij2016: IV.4, §89]).Footnote ³⁵ Typically, the particle puts in focus an anaphoric pronoun or adverb, as in τότε δή (‘at that moment then’), οὕτω δή (‘under those circumstances then’), et cetera. According to Bonifazi et al. (Reference Bonifazi, Drummen and de Kreij2016: IV.4, §89), the particle serves to mark the author’s voice as narrator. What this means exactly is hard to pin down. Bonifazi et al. (Reference Bonifazi, Drummen and de Kreij2016: IV.4, §90), with reference to Herodotus, speak of ‘moderate involvement, a sort of investment … in the process of unfolding his historiographical discourse’. In other words, the particle puts the narrator’s mediating presence in the foreground. If my claim that the present for preterite is used at points where the discourse structure becomes cognitively salient is correct, then we should expect the odds of the present being used to increase when this particle is used.

Let me give some examples. The first illustrates the use of the particle in the collocations mentioned above:

The phrase οὕτω δή (‘under those circumstances then’) signals that at the designated point in the story, the local build-up of tension (will the Thebans listen, or will Cleombrotus have to invade their country?) reaches a critical point.Footnote ³⁷

A second example illustrates the less common use of the particle to mark narrative progression by itself. Here, Alcibiades tells how he tried to win over Socrates as his lover. When his first attempts fail, he tries another route:

Alcibiades tells how he wanted to make another attempt at Socrates, which raises the question what he tried to do next. Again, the particle δή (‘so’) puts into focus a new development after a build-up of anticipation.

Finally, the particle δή (‘then’, ‘so’) can be recognised in the grammaticalised conjunction ἐπειδή (‘so when’), which is a combination of the conjunction ἐπεί (‘when’) and the particle. According to Bonifazi et al. (Reference Bonifazi, Drummen and de Kreij2016: IV.4, §110), ‘the -δή component within the whole word (which functions as a temporal conjunction) corresponds to the pragmatic δή used for narrative progression’. The following passage is suggestive of the difference in value between the marked and the unmarked conjunction. Xenophon rides out to meet the Thracian ruler Seuthes. When he arrives, the following happens:

After the subordinate clause introduced by the neutral conjunction ἐπεί (‘when’), the narrator tells of some formal interaction preceding the actual meeting between Xenophon and Seuthes. The marked conjunction ἐπειδή (‘so when’) then signals that the narrator comes to the point: Xenophon actually went in and met Seuthes, as we were anticipating.

Now let me present the relevant statistics. First, I have looked at the use of the particle at the second position in the sentence (example [30]) or in the main clause (example [29]; on the position of particles, see, e.g., Denniston [1954: lviii–lxi]). As I explain in the Appendix, Section A.1.3, I also include non-restrictive relative clauses and subordinate clauses marking an endpoint or result in my data sample. In such cases, the particle can modify the relative pronoun or the conjunction, as in πρίν (γε) δή (‘until then/finally’). I regard this use as analogous to that modifying anaphoric expressions in the main clause (τότε δή [‘at that moment then’]), and have therefore included these cases as well.Footnote ³⁸ The results are listed in Table 3.2.

Table 3.2Tense and connective δή

^a B = 1.498 (constant = .294), standard error = .767, N = 946 (20 cases of δή).

^b B = .707 (constant = -.301), standard error = .464, N = 615 (14 cases of δή).

^c B = 1.809 (constant = .037), standard error = .626, N =729 (22 cases of δή).

The odds ratios and p-values are rather divergent. This must be due at least in part to the low number of cases where we find the particle (see the numbers in the notes to Table 3.2). Overall, however, the effect is clearly positive, so that the results support the hypothesis.

Things are more difficult with respect to the difference between ἐπειδή (‘so when’) against simple ἐπεί (‘when’). Thucydides almost always uses the marked conjunction.Footnote ³⁹ In Xenophon’s works, the marked conjunction is rare, but it does seem to have a predilection for the present for preterite. Grouping Xenophon’s two works together, the results are listed in Table 3.3.

Table 3.3Tense and the particle δή in conjunctions

^a B = 1.261 (constant = -.345), standard error = .619.

The results suggest that, for Xenophon, the use of the marked conjunction with the particle δή (‘then’, ‘so’) has a positive influence on the odds of the present for preterite being used (odds ratio 3.529), but the effect is not very certain (p = 0.042).Footnote ⁴⁰

3.3.5 Conclusion

In this section, I have aimed to support the claim that the diegetic present for preterite functions to mark salient changes in the discourse structure by showing that it has an affinity for certain attention-management strategies. In Sections 3.3.1 and 3.3.2, I discussed examples of the present for preterite in combination with announcements (in particular with cataphoric reference) and in questions and exhortations to the audience. In Sections 3.3.3 and 3.3.4, I discussed evidence suggesting that clause complexity and the use of the particle δή (‘then’, ‘so’) to mark discourse progression positively correlate with present for preterite usage. In sum, I hope this has convincingly shown that the diegetic present is associated with a strongly mediating stance of the narrator.

3.4.1 Typology

In the abstract, there are three types of changes in the narrative dynamic:

1. (a) the start of new developments against the background of a stable situation;

2. (b) a change in the direction of the story’s progression;

3. (c) the culmination of narrative complications.

The boundaries between these categories are fluid. The distinction between (a) and (b) depends on the degree of turbulence already present in the story at a certain point, and this is of course a continuous variable. Similarly, a change in the story’s direction (b) may be such that it puts a complete end to the narrative complications leading up to that point (c). Nevertheless, it is useful to make these distinctions for heuristic purposes.

For the first category (a), I use the term initiative when the event is brought about by intentional human agency. In the following example, a character in the story actively seeks a confrontation with another:

In this example, the initiative marks the beginning of a new episode (‘incipit’, see Section 3.4.2), which is neatly introduced by the embedding particle γάρ (‘for’, after a narratorial abstract which I have not cited) and a temporal subordinate clause setting the scene.Footnote ⁴¹

Another example illustrates the start of unexpected developments against the background of routine procedures (here we move towards [b]). The speaker and a certain Olympiodorus were making necessary arrangements after the death of a certain Comon, a relative of the two men. Then Olympiodorus raises an issue:

According to Fludernik (Reference Fludernik1991: 374, referring to Pollak [Reference Pollak1960] and Quasthoff [Reference Quasthoff1980]), ‘[t]he major occurrences of the historical present can be analysed as marking the incident which cuts into a setting – in the well-known “schema of incidence”’. Example (33) fits this description. The setting is provided by the participial clause ὄντων δ’ ἡμῶν περὶ ταύτην τὴν πραγματείαν (‘as we were engaged with this business’), and the unexpectedness of the event cutting into this setting is marked with the adverb ἐξαίφνης (‘suddenly’; compare the examples in Fludernik [Reference Fludernik1991: 375]).

However, I prefer to reserve the term incident for changes in the narrative dynamic where intentional agency is absent or de-emphasised. Consider the following passage:

The speaker first paints a picture of his circumstances. He tells us that on different occasions he was sent out as trierarch. Then he focuses on a specific occurrence where he was appointed as trierarch. This marks the end of the introductory section (‘orientation’, see Section 3.4.2) and the start of narrative complications. The present tense συμβαίνει (‘falls to [my] lot’) highlights this change in the discourse structure; note also how the narrative progression is marked by the particle δή (‘so’, Section 3.3.4).

Now we move to the second category: changes in the direction of the story’s progression. Again, I make a distinction between two subtypes in terms of the presence or absence of intentional human agency. The intervention is a purposeful action performed by a main character in the narrative seeking to change an unwanted course of events. We find two instructive examples in a narrative passage in Plato’s Symposium. In 189c–93d, Aristophanes gives his speech in praise of Eros, in which he tells a myth to explain man’s relationship to the god. According to him, the original race of man consisted of spheric creatures, with four arms, two heads, et cetera. Being high-minded, these creatures tried to ascend to heaven to attack the gods. So Zeus had to come up with a solution:

This is a typical ‘difficult situation-solution’ schema (note in the Greek ἠπόρουν [‘were at a loss’] and μηχανήν [‘contrivance’]; compare Section 3.4.3). If nothing were to be done, conflict between man and the gods would arise; so Zeus acts to change this course of events. His speech marks a change in the narrative dynamic, and this is why the present tense is used to introduce it.

Zeus proceeds to carry out his idea: he cuts man in half, so that each individual now has only two arms, one face, etc.; in this way, they are less powerful. However, this results in a new problem: each half-person is torn with desire for his or her other half, and they seek each other out, embrace each other and yearn to be stitched back together. In this way, they die, because they refuse to do anything else. This course of events is also deemed undesirable by Zeus, who does not wish the race of man to perish. So he comes up with a new solution:

Zeus’ solution allows people to have intercourse again (after having been cut in half, their private parts were in their backs), so that they will not perish. Note again the word μηχανήν (‘contrivance’) as a semantic indicator of a change in the narrative dynamic.

In other cases, the change in the progression of the story is the result of unintentional action or brought about by some external agent (e.g., a god or a character who is outside the ‘main cast’). Here I will use the term incident as well (the difference between [a] and [b] in this respect is not essential from a descriptive standpoint). Two passages in the narrative of the chariot race in Sophocles’ Electra illustrate the progression-incident pattern very neatly:

Up to a certain point, the race is progressing safely and steadily, each charioteer remaining upright in their chariot (723 ἕστασαν [‘stood’]). Then one of the competitors loses control of his horses, and a crisis ensues: the horses carry the man away and crash into another chariot (present tense forms 725 φέρουσιν [‘carry away’] and 727 συμπαίουσι [‘hit against’]). The change from stable progression to incident is marked by the contrast between the adverbs πρίν (‘first’, 723) and ἔπειτα (‘next’, 724) and by the switch in tense. A similar pattern occurs later on: Orestes safely finished all laps, except the last one, when he inadvertently hits the pillar. Again, the contrast is marked by the opposition between τοὺς μὲν ἄλλους πάντας … δρόμους (‘all the other laps’, 741) and ἔπειτα (‘next’, 743), and the present tense highlights the moment of change (744–5 λανθάνει … παίσας [‘inadvertently hits’]).

Finally, there is the culmination of narrative complications (category [c]). What is peculiar about this type of change in the narrative dynamic with respect to the other two is that the element of surprise is less prominent: the culmination is essentially the fulfilment of the expectation that the narrative complications need to reach a certain critical moment in order to be resolved (compare, e.g., Allan’s [Reference Allan, Bakker and Wakker2009: 187] definition of the ‘peak’; see Section 3.4.2). This often happens according to typical schemas, such as a battle ending in a loss or victory, a trial ending in a verdict, etc. For example:

Philocrates was indicted for proposing an illegal decree. This is the start of new complications in the story (the indictment is marked with the present γράφονται [‘indict’] in [14]). The episode proper is introduced with the clause μετὰ ταῦτ’ εἰσῄει ἡ γραφὴ εἰς τὸ δικαστήριον (‘after that, the case came to trial’). This episode reaches its culmination with the outcome of the trial, introduced with the phrase τὸ τελευταῖον (‘in the end’). The present forms ἀποφεύγει (‘is acquitted’) and ‘[οὐ] μεταλαμβάνει (‘does [not] get)’ highlight this peak.

Similarly, in the final part of the messenger narrative in Euripides’ Children of Heracles, the messenger tells how Iolaus went in pursuit of his foe Eurystheus:

This naturally creates anticipation: will Iolaus fail or succeed? The present marks the culminating point, which constitutes the answer to this question (859–60): αἱρεῖ δ’ ὁ κλεινὸς Ἰόλεως Εὐρυσθέως | τέτρωρον ἅρμα πρὸς πέτραις Σκιρωνίσιν (‘famed Iolaus captures the four-horsed chariot of Eurystheus by the Scironian rocks’).

To conclude, I use the following descriptive terms for identifying changes in the narrative dynamic:

1. (a) Initiative. The start of narrative complications, intentionally initiated by a human agent.

2. (b) Intervention. A change in the direction of the narrative progression, intentionally initiated by a human agent.

3. (c) Incident. Either the start of new complications or a change in the narrative progression, with agency being absent or de-emphasised.

4. (d) Culmination. The climax of narrative complications.

3.4.2 Episodic Structure

Changes in the narrative dynamic of the types I have just described are normally discussed with reference to a specific framework: that of episodic structure (Fleischman [Reference Fleischman1990]; Fludernik [Reference Fludernik1991]; Allan [Reference Adams2007], [2009], [2011a]). In this section, I present a critical evaluation of the explanatory value of this model for the Classical Greek material.

The theory of the episodic structure of narrative was introduced by Labov and Waletzky (Reference Labov, Waletzky and Helm1967) and further elaborated by Labov (Reference Labov1972) and Fleischman (Reference Fleischman1990). The idea is that narratives, in particular conversational ones, consist of certain fixed structural elements. There is an abstract, explaining what the narrative will be about; followed by an orientation, specifying the time, place and participants; then follows the complication, which is the actual action; the complication culminates in the peak; after that follows the resolution, which is what finally happened; and then an evaluation, the narrator’s reflection on the events.Footnote ⁴²

In the existing narrative-structural models of the present for preterite, the function of the present is tied to specific points in the structure of the episode. In Allan’s model, applied to Euripidean messenger speeches and Thucydides’ historical narrative (Allan [Reference Adams2007], [2009], [2011a]), the present for preterite is used to mark the peak and (less frequently) the incipit, that is, the beginning of the complication.Footnote ⁴³ An effective illustration is the following narrative episode from Andocides’ speech On the mysteries:

In this episode, all structural slots I mentioned above are filled and the present is found exactly in the designated positions. In the abstract, we are told what the story is about: a μήνυσις (‘laying of information’). The orientation gives us the main protagonist – a certain Teucer – and tells us of his circumstances. The complication starts when Teucer makes a proposal to the Athenian Council: he will lay information against the profanators of the mysteries and the mutilators of the Hermae if given immunity. The present ἐπαγγέλλεται (‘informs’) here thus marks an incipit – in my terms, this is an ‘initiative’. At the peak (culmination), Teucer fulfils his promise and writes down the names: note how ἄδειαν εὑρόμενος (‘having been granted immunity’) harks back at ἄδειαν (‘immunity’) in the incipit. As a result of this information, the denounced persons fled (resolution). Finally, the orator asks for the relevant document to be read, a typical part of a rhetorical evaluation.

The great merit of this approach is that it ties the use of the present for preterite to a criterion that is, in principle, objective and independent. That is, the structure of narrative episodes can be determined by factors other than tense usage. In practice, however, things are rarely as clear-cut as in the passage just cited. To begin with, we sometimes find the present in the ‘wrong spot’, and in such cases it is hard to resist the temptation to conform the structural analysis to the tense usage.Footnote ⁴⁵ More problematic than the existence of incidental counterexamples is the fact that narrative passages do not always submit to an analysis in terms of episodic structure. Many narrative assertions containing the present for preterite in the corpus are part of macro-structural gestures that extend beyond episodic units (which are circumscribed in terms of time, place, participants and action).Footnote ⁴⁶ This tends to be more the case as we move further to the diegesis end of the mimesis-diegesis continuum.

To begin, let me discuss an example where the present for preterite marks the beginning of new developments. The context is as follows. A man named Pittalacus had a boy with whom he was in love, by the name Timarchus, stolen from him by a certain Hegesandrus. One night, Hegesandrus, together with Timarchus and others, raided Pittalacus’ house and gave the man a flogging. The next day, Pittalacus went to an altar in the marketplace for refuge. He was ultimately persuaded by Hegesandrus and Timarchus to leave the altar based on the understanding that he would receive some sort of compensation. Then the following happened:

After Pittalacus was persuaded to leave the altar, Hegesandrus and his henchmen paid no further attention to Pittalacus. This is the resolution of the previous episode. Now Pittalacus files a suit against the two men (present λαγχάνει [‘files’]). This constitutes a new initiative; the participial clause βαρέως δὲ φέρων τὴν ὕβριν αὐτῶν (‘taking their outrageous behaviour against him badly’) signals an imminent change in the narrative dynamic (see Section 3.4.3). However, it is difficult to speak of an incipit here in terms of episodic structure. The problem is that the narrative assertion falls squarely in between the previous episode, which was closed off neatly with a resolution, and the next one, which is introduced by a subordinate clause that serves as an orientation (ὅτε δ’ ἐδικάζετο [‘when the case came to trial’]) and an announcement that refocuses the audience’s attention (σκέψασθε μεγάλην ῥώμην Ἡγησάνδρου [‘consider a powerful feat of Hegesandrus’]). The complications in this episode begin when Hegesandrus tries to take Pittalacus away as a slave. The point is that the present for preterite here can be accounted for in terms of a change in the narrative dynamic but on a higher level than that of the episode.

Let me discuss an example from historiography that is particularly illustrative of the use of the present for preterite in summary narrative to mark changes in the narrative dynamic. It is a long passage and it is hard to find a suitable starting point without quoting much material irrelevant to my analysis. The point is that some party at Athens is scheming to bring Alcibiades back from exile. Phrynichus, an Athenian general, does not wish this to happen, and the following events unfold:

The framework of episodic structure offers little help in accounting for the present forms here. If the entire passage is considered to constitute a single episode (which makes some sense given the unity of participants and action), then there are too many present forms. The alternative is to distinguish between different episodes within this passage, with the episode boundaries being constituted by the topic shifts and the (mostly implicit) changes of scene. In this analysis, the first episode consists of Phrynichus sending a letter to Astyochus; the next episode narrates how Astyochus betrays Phrynichus to Alcibiades; and so on. In this analysis, each episode is stripped down to its very core, to one central event, which becomes the peak by default.Footnote ⁴⁷ The problem, in my view, is that this allows so much flexibility to the application of the model as to be detrimental to its explanatory power.

In my analysis, we can simply explain these clustered instances of the present in terms of the impact of each individual event on the narrative dynamic. Phrynichus fears that Alcibiades will return; he therefore acts to prevent this from happening by sending a letter to Astyochus (intervention: τρέπεται [‘turns’], πέμπει [‘sends’]). Astyochus betrays him, thwarting Phrynichus’ plan and prompting Alcibiades to act (intervention: λέγει [‘tells’], γίγνεται μηνυτής [‘turns informer’]). Thereupon, Alcibiades plots to have Phrynichus killed (initiative: πέμπει [‘sends’]). Phrynichus is distressed and tries to persuade Astyochus once more (intervention: ἀποστέλλει [‘sends’]). Again, Astyochus betrays him (intervention: μηνύει [‘reveals’]). Thereupon, Phrynichus acts to prevent Alcibiades from accomplishing his goal of having him put to death by informing the Athenian army at Samos that the enemy is preparing an attack on the city (intervention: ἐξάγγελος γίγνεται [‘becomes messenger’, i.e., ‘informs’]).Footnote ⁴⁸

Finally, the present for preterite can be used to mark changes in the discourse structure on a much higher level. In Thucydidean historiography, this often concerns changes in the military/political map in Greece. Such usages are often more easily understood in terms of changes in the status of referents, which I discuss in Section 3.5. Let me just give one example here. The following passage narrates actions taken by the Lacedaemonians after having successfully completed a military expedition against a place called Iasus:

The aorist ἀνεχώρησαν (‘retreated’) marks the resolution of the military expedition. Before the summer ended (imperfect ἐτελεύτα [‘ended’]), the Lacedaemonians make some preparations, sending a commander to Erythrae (ἀποστέλλουσι [‘send’]) and installing a certain Philip as governor of Miletus (καθιστᾶσιν [‘install’]). To my mind, the concept of the ‘episode’ has no explanatory value here with respect to the use of the present for preterite. Rather, the present signals changes in the macro-level discourse structure. Our mental model of the discourse involves keeping track of what force or which commander is where at what moment. Philip is now the commander at Miletus, and Pedaritus is now on his way to Erythrae.Footnote ⁴⁹

In conclusion, the model of episodic structure may be a useful heuristic tool for identifying particular locations of changes in the narrative dynamic (the incipit and the peak), but only in those cases where the narrative submits to such an analysis in the first place. In summary narrative, it is often good economy to explain the use of the present for preterite in terms of changes in the structure of the narrative without reference to episodic structure.

3.4.3 Narrative Schemas: Corpus Analysis

Having discussed a number of suggestive examples, I now aim to show in a more systematic fashion that the present for preterite is structurally associated with the pragmatic function of marking changes in the narrative dynamic.

The evidence presented in Sections 3.3.3 (sentence complexity) and 3.3.4 (discourse progression marked by δή [‘then’, ‘so’]) already points in this direction. Another factor is verbal semantics. It has been noted many times that the present for preterite has a predilection for certain verbs or event types. For example, in the volume on the ‘historical present’ in Thucydides (Lallot et al. [Reference Lallot and Lallot2011]), there are individual contributions on the verbs αἱρέω/λαμβάνω (‘capture’) (Allan [2011a]), τρέπω (‘turn to flight’) (Mortier-Waldschmidt [2011]), πείθω (‘persuade’) (Jacquinod [2011]) and πέμπω (‘send’) (Lambert [2011:211–21]). These verbs all have a high present-to-aorist ratio and typically mark changes in the structure of the narrative.Footnote ⁵⁰ Other salient examples are verbs such as μηχανάομαι (‘contrive’; 5 present forms, 0 aorists in my selected corpus; compare Section 3.3.1, Footnote note 27), ἐπιβουλεύω (‘form a plot’; 3 present forms, 0 aorists) and the auxiliary verb φθάνω (‘do something before someone else can prevent it’; 9 present forms, 1 aorist; see also Chapter 2, Section 2.6.2).

However, as there are many different verb types and it is difficult to come up with objective criteria for determining their inherent ‘change-marking’ quality, I will not attempt a systematic review of tense marking and verb type. Instead, I will focus on how changes in the narrative dynamic are cued in the context surrounding the main clause assertion. I have identified seven narrative schemas involving a change in the narrative dynamic. Six of these are characterised by certain keywords, and one by a negation in a subordinate clause. I describe these schemas as follows:Footnote ⁵¹

Keywords: ἀπορέω (‘be at a loss’), πιέζομαι (‘be pressed’).

Keywords: ἔδεισα, φοβέομαι.

Keywords: ὀργίζομαι (‘be angry’), χαλεπῶς/βαρέως φέρω (‘find difficult to bear’), ἀγανακτέω (‘be grieved’), ἀλγέω (‘be pained’), λυπέω (‘be pained’).

Keywords: βούλομαι (‘want’).

Keywords: ὁράω (‘see’), γιγνώσκω (‘realise’), αἰσθάνομαι (‘perceive’), μανθάνω (‘learn’), πυνθάνομαι (‘hear/learn’).

Keywords: νομίζω (‘think’), ἡγέομαι (‘think’), δοκέω (‘think’, with the cogniser in the nominative case, or ‘seem’, with the cogniser in the dative case).

This schema is not characterised by a specific keyword but by a negation in a temporal subordinate clause.

The criteria for inclusion in the data and coding were the following:

1. (a) As sentence complexity in itself increases the odds of the present for preterite being used (Section 3.3.3), I included only complex sentences in my data sample. For schemas 1 to 6, I included cases with a minimum of one temporal subordinate clause or conjunct participial clause (either preposed or postposed). For schema 7, I included only cases with a preposed temporal subordinate clause (because all relevant instances containing a negation were preposed). Absolute participial clauses were ignored.

2. (b) For schemas 1 through 6, instances were coded as belonging to a certain schema when the main verb of one of the subordinate clauses belonged to one of the defined keywords (composites were included).Footnote ⁵³ Moreover, the subject of the subordinate clause had to be coreferential with that of the main clause (except in the case of δοκέω [‘seem’], schema 6).

3. (c) Instances were coded as belonging to schema 7 when the temporal subordinate clause contained a negation, without further conditions.

Two further thoughts. First, I have tried to keep my selection of keywords as simple as possible. There are many other cases I might have included in a more generous selection, but this would have made the selection more arbitrary. I have aimed at simplicity and consistency, not at maximizing positive results: the reader may judge whether I have left out keywords that should have obviously been included, or vice versa. Second, by looking only at subordinate clauses, I have excluded cases where the schema extends beyond a single sentence. To give an example:

In the first main clause, the character arrives at the idea that he should try to escape his dire situation (ἐδόκει [‘it seemed’] is a keyword for schema 6). The second main clause narrates the action taken by the character. The reason for not including such examples is that it is hard to decide where to draw the line: how great should we allow the distance between the two elements of the schema to be? Sticking to subordinate clauses makes the selection of the data more objective.

The results are listed in Table 3.4. Because the total number of instances of the schema in question (listed under ‘cases’) is generally low, I have grouped the three historiographical works together. The loss of individual differences is less important, in my view, than the overall picture. I have ranked the schemas according to odds ratio, from high to low. Schema 3 (‘angry’) was excluded because the data yielded only a handful of cases.

Table 3.4Tense and narrative schemas

^a B = 1.518 (constant = .273), standard error = .766. N = 1155.

^b B = .909 (constant = .261), standard error = .386. N = 1155.

^c B = .768 (constant = .274), standard error = .479. N = 1155.

^d B = .391 (constant = .245), standard error = .195. N = 1155.

^e B = .312 (constant = .199), standard error = .389. N = 254.

^f B = .260 (constant = .279), standard error = .342. N = 1155.

The generally positive effect associated with these schemas supports the claim that one of the main functions of the present for preterite is to mark changes in the narrative dynamic. Note that the increase in odds reported here is above the increase in odds due to the mere presence of at least one conjunct participial clause or temporal subordinate clause (odds ratio 1.511).Footnote ⁵⁴ The observed hierarchy makes sense in light of the hypothesis presented here. It stands to reason that the ‘difficult situation’ schema should be the surest sign of an imminent change in the narrative dynamic. Next, a character’s ‘fear’ is the strongest semantic cue. Of the other four, ‘wish’ (which has a reasonably strong odds ratio at 2.155 but a rather high p-value at 0.109) is the strongest indicator of a protagonist’s intention to effect a change in the narrative dynamic. ‘See’ and ‘think’ (a statistically negligible effect at p = 0.447) are more neutral terms.

As for the ‘go differently’ schema, many cases where the aorist is used are easily explained away in terms of narrative dynamics. The aorist is normally used when an attempt is abandoned altogether. For example:

The aorist is almost always used in such contexts, where the resolution of narrative tension is marked by characters ‘going away’ (ἀπέρχομαι), ‘retreating’ (ἀναχωρέω), ‘sailing away’ (ἀποπλέω) and the like. This is very different from when an initial setback inspires a new attempt to accomplish the original goal. At least 7 out of the 12 cases of the aorist counted for this schema fall into this category.Footnote ⁵⁵ Ignoring such cases would certainly make the effect (which is now statistically negligible at p = 0.423) stronger. Similar instances are found with the other schemas as well.

3.4.4 Conclusion

In this section, I have argued that one of the main discourse-structural functions of the present for preterite is to mark changes in the narrative dynamic. I identified four types: initiatives, interventions, incidents and culminations (Section 3.4.1). Next, I argued that the explanatory value of these concepts for tense-switching supersedes that of the model of episodic structure (Section 3.4.2). Finally, I presented evidence from the corpus to the effect that subordinate clauses that cue an imminent change in the narrative dynamic increase the odds of the present for preterite being used (Section 3.4.3).

3.5.1 Typology

I begin with changes in activation status. The activation of a referent entails either the introduction of an entirely new referent or the re-introduction of an already established referent in a particular episode or scene (‘cast change markers’; see van Dijk [1981: 181]). The following example illustrates the former principle:

Intaphrenes, Megabyzus and Hydarnes are mentioned here for the first time. Darius has been mentioned before by Herodotus but not as a character in the main narrative. These men, together with the already mentioned Otanes, Gobryas and Aspathines, will play the leading role in the following narrative, as they rise against the reign of Smerdis the mage.

An example of a local re-introduction of a character who has already been established in the macro-narrative is the following:

The present in ἐκκλησία γίγνεται (‘an assembly is held’) marks a new development in the macro-narrative structure, prompting the beginning of a new episode (the beginning of the actual episode is marked in the next sentence with the phrase ἐν δὲ ταύτῃ τῇ ἐκκλησίᾳ [‘in that assembly’]). The present λαγχάνει (‘is chosen by lot’) serves to highlight the introduction of Demosthenes, who is a main character in the macro-narrative, as an actor in this new episode.

Finally, the following instance illustrates the re-introduction of a referent on the scene on two levels of narrative organisation. The speech is Lysias’ Against Simon, and the context is as follows: After a confrontation between the speaker and the defendant over a boy, Theodotus, the speaker left Athens. After a while, he comes back:

The two instances of ἀφικνοῦμαι (‘arrive’) have been found problematic by those who explain the use of the present for preterite purely in terms of rhetorical impact (Sicking and Stork [Reference Sicking, Stork and Bakker1997]). In my account, these present forms mark the re-introduction of the speaker on the scene on different levels of the narrative. First, the speaker arrives back at Athens: this marks the start of a new phase in the narrative on a macro-structural level. The second instance marks the arrival of the speaker on the scene in the context of episodic developments. Note how the phrase ἐν δὲ τούτῳ τῷ καιρῷ (‘in that moment’) indicates that the event designated by the main clause verb constitutes an incision in the structure of the story.Footnote ⁵⁸

De-activation of referents occurs when referents are taken ‘off the stage’ (exit), usually by death. This is illustrated by the following example:

With his disappearance in the story, Therimenes disappears as a discourse referent that the addressees need to take into account in the processing of the narrative.

Changes in the role of characters often coincide with changes in their activation status in the discourse. One instance, however, where the referent is already highly active when his role changes, is the following:

Aeschines here acquires the role of ambassador. This is a key moment in Demosthenes’ narrative, as it concerns Aeschines’ alleged misbehaviour while acting in this function.

Finally, typical changes in the position of referents involve events such as marriage, adoption, financial transactions, transfer of property and the like.Footnote ⁵⁹ I discuss two cases that also illustrate the cumulative effect of the use of present forms in close succession to mark different changes in the status of referents. First, consider the following passage:

The present ἀφικνεῖται (‘arrives’) highlights the activation of a new referent: Timosthenes of Aegilia. Then, the narrative moves back to a previous occasion. The first instance of the present δίδωσιν (‘gives’) introduces a new referent (the bowls), re-activates a familiar referent (Phormio) in this new part of the story and marks a possession change (the bowls move from Timosthenes to Phormio). The second instance again marks a possession change and re-activates a referent who is familiar from the preceding discourse (Aeschrion).

My second example concerns the narrative in Isaeus’ speech On the estate of Hagnias (11). The facts as presented by the speaker are as follows: Hagnias did not leave his estate to his next-of-kin, but adopted a niece as his daughter. Should he die, his estate would go to whomever she married. However, he had written in his will that, in case his adopted daughter should die, the property was to go to Glaucon, his half-brother. Beside Glaucon, Hagnias’ next-of-kin were Eubulides, Stratocles, Stratius and the speaker. This was the situation, but then things changed:

The first two present forms (τελευτᾷ [‘dies’]) mark the exit of referents in the narrative who were important in that they were potential claimants to the disputed inheritance. The form λαμβάνει (‘takes possession of’) marks the transfer of the inheritance to the possession of Glaucon. Isaeus continues to use the present for preterite in the rest of this short narrative to mark changes of possession and exits of referents.Footnote ⁶⁰ This serves to facilitate our processing of the discourse, as we keep track of who is in possession of the estate, and which potential claimants are still alive, at each point in the narrative.

3.5.2 Mental Maps

The present for preterite is often used to mark stages in a journey. This use may be understood in terms of marking a change in the position of referents as well. However, I argue that the conceptual scenario in which the position of referents is traced in the discourse space may be supported by another conceptual scenario in which the position of referents is traced on a fictive map. This mental map is a ‘visual aid’ which helps to concretise the more abstract concept of referents changing position in the discourse space.

The most well-known cases of what we may call the ‘itinerary present’ are found in the first book of Xenophon’s Expedition of Cyrus. Here the stages of Cyrus’ expedition into upper Persia are consistently marked with present forms, in formulas of the following type: From there he marches X stages, Y parasangs, to location Z. The most remarkable cluster of such present forms is found in the following passage:Footnote ⁶¹

Rijksbaron (2002: 24) devotes a separate heading to this use of the present for preterite. He argues that ‘decisiveness’, which he considers the typical function of the present for preterite elsewhere (22–4), is ‘a less prominent feature’ here. Rijksbaron assigns a narrative-structural function to this use of the present: ‘[T]hey “punctuate”, as it were, the narrative, dividing it into narrative units.’ I think discourse segmentation is certainly an important consideration here. The narrative ‘moves along’ with the main character: as Cyrus marches to a next stage, we are led from one spatio-temporally defined narrative unit to the next (compare Willi [Reference Willi and Slater2017: 240]; Huitink [Reference Huitink, Willi and Derron2019: 196–8]).

On the other hand, I do not think this interpretation is entirely satisfactory. When the present is used in close succession, as in the example just cited, the idea that each present form marks the beginning of a new ‘narrative unit’ loses much of its explanatory value. The first two present forms do not introduce a new episode; nothing further happens after Cyrus arrives at his destination. In my view, this explanation should be supported by one that is more specifically geared towards the kind of conceptualisation that is evoked by these verbs – verbs such as ἐξελαύνω (‘march out’) in example (53), as well as (very frequently) ἀφικνέομαι (‘arrive’) and πέμπω (‘send’) and others such as ὁρμίζομαι (‘anchor’).Footnote ⁶²

I argue that the present for preterite of such verbs may activate a conceptual scenario in which the designated change in location is traced on a virtual map. The existence of maps in Classical Athens is well-attested, although the texts we have suggest that maps were something of a novelty.Footnote ⁶³ The best-known passage is probably the story in Herodotus’ Histories (5.49), where Aristagoras, the ruler of Miletus, comes to Sparta in order to convince King Cleomenes to undertake a military expedition into Persia. Aristagoras carries with him ‘a bronze tablet in which a map of the entire earth and the entire sea and all the rivers was engraved’ (5.49.1 χάλκεον πίνακα ἐν τῷ γῆς ἁπάσης περίοδος ἐνετέτμητο καὶ θάλασσά τε πᾶσα καὶ ποταμοὶ πάντες).Footnote ⁶⁴ On this map, Aristagoras shows Cleomenes the peoples that he will conquer on the way to Susa, pointing out their riches to entice the king to take on the enterprise:

Note how Aristagoras uses demonstratives designating immediate proximity to refer to the representations of the Ionian, Lydian and Phrygian peoples on the map (Ἰώνων τῶνδε [‘neighbouring the Ionians here’], οἵδε Λυδοί [‘the Lydians over here’], οἵδε Φρύγες [‘the Phrygians over here’]). In reality, these peoples live far away, but their representations on the map are within arm’s reach of Aristagoras.Footnote ⁶⁵

I argue that the use of the ‘itinerary’ present conveys the pretence that the narrated journey can be traced on a map that is currently accessible to the speech partners. Of course, there is no actual map and no actual tracing: we are dealing with a covert scenario, a conceptual substrate we supply in order to make sense of the utterance (Langacker [Reference Langacker2008: 531–5]). In Xenophon’s Expedition of Cyrus (example [53]), we mentally trace Cyrus’ journey inland on a map, moving our focus from one location to another.Footnote ⁶⁶ The way Xenophon describes the stages is iconic of this act of tracing. Xenophon starts with the reference location (ἐντεῦθεν [‘from there’]), then gives the verb (ἐξελαύνει [‘marches out’]), the distance travelled (σταθμοὺς δύο παρασάγγας δέκα [‘two stages, ten parasangs’]) and the destination (ἐπὶ τὸν Ψάρον ποταμόν [‘to the river Psarus’]).Footnote ⁶⁷

This explanation is not mutually exclusive with an explanation in terms of discourse structure. Rather, it reinforces it. Let me give another example to argue this point:

The present forms ἀφικνοῦνται (‘arrive’) and ὁρμίζονται (‘cast anchor’) signal a salient change in the position of referents in the discourse: the Athenian ships ‘are now’ at Corcyra, and the Corinthians ‘now anchor’ at Chimerium.Footnote ⁶⁸ The conceptual construct of a map serves as an aid to visualise this change in the discourse structure. In other words, a link is established between the referents in the discourse space and entities as represented in a mental map. When referents ‘move’ in the discourse, this moving is visualised by a repositioning of the entities on the map. This interpretative move is facilitated by the geographical description of the harbour inserted into the narrative by Thucydides.

3.5.3 Referent Designation: Corpus Analysis

In order to support the claim that the present for preterite is structurally associated with the function of highlighting changes in the status of referents, I will discuss evidence from the corpus pertaining to tense usage and the way referents are coded linguistically. The general hypothesis explored here is that the odds of the present for preterite being used increase as the subject of the main clause becomes less predictable. Concretely, this means that the odds of the present for preterite will increase, first, when the subject is given more coding material (according to Givón’s [Reference Givón1991] quantity principle), and second, when the referent is marked as indefinite.

Let me give an example to illustrate these points:

The subject (Aristides) is designated by proper noun rather than by a pronoun or by zero anaphora (i.e., no designation). This signals that the subject is not retrievable from the immediately preceding discourse context. Moreover, the subject is designated not only by proper noun but by patronym as well (‘Aristides the son of Archippus’), which suggests that the man is entirely new to the discourse.Footnote ⁶⁹ This becomes further apparent from the apposition with an indefinite description: εἷς … στρατηγός (‘one of the admirals’). Apparently, we are not supposed to be familiar with the referent or his role in the discourse. The object of the sentence is also designated by proper noun (‘Artaphernes’), and here too an indefinite description signals the referent’s all-new status in the discourse (ἄνδρα Πέρσην [‘a Persian’]). The introduction of two all-new referents constitutes a significant incision in the discourse structure, and we find the present for preterite in the main clause (ξυλλαμβάνει [‘arrests’]). In the next two sections, I will try to establish to what extent the association of the present for preterite with the unpredictability of referents is a structural property of the Classical Greek language.

3.5.3.1 Coding Material

When a referent is new to the discourse or newly activated, more linguistic material will be used to designate this referent. Givón (Reference Givón1991) proposes the following hierarchy: zero anaphora > unstressed pronoun > independent pronoun > full noun phrase. My hypothesis is that the odds of the present for preterite being used increase when we move higher in this hierarchy.Footnote ⁷⁰ I will confine myself to the analysis of subject designations. A subsidiary hypothesis I will discuss below is concerned with variation within the category of proper nouns: here I look at the difference between simple names and full names (see Mulkern [Reference Mulkern, Fretheim and Gundel1996]).

My focus here is solely on how referents are designated, with other aspects (e.g., animacy, number) being kept constant as much as possible. In order to address this, I have limited my data sample to cases where the main verb is singular and the subject is a person who can be identified by name (because the name is given somewhere in the text). Within this set, I have made a distinction between three kinds of reference:

1. (a) Noun phrase. This is almost always a proper noun (e.g., Δημοσθένης [‘Demosthenes’]). Cases where we find a type noun to refer to a character who is known by name (e.g., ὁ ἀνήρ [‘the man’]) are exceedingly rare. I have therefore excluded these from the data.Footnote ⁷¹

2. (b) Pronoun: ὁ (‘he’), αὐτός (‘he himself’), οὗτος (‘that [man])’, ἐκεῖνος (‘that [man])’. This is relatively uncommon, and I have therefore not found it useful to make more fine-grained distinctions within this group.

3. (c) Zero anaphora.

I have coded subject designations found in the main clause (Demosthenes did X), in a subordinate clause (As Demosthenes saw this, [he] did X), or as an extra-clausal theme-constituent (Demosthenes, seeing this, did X; see Matić [Reference Matić2003]; Allan [Reference Allan2014]). In the latter two cases, I have coded only the immediately following main clause verb for the relevant designation.Footnote ⁷²

The results are listed in Table 3.5. In Thucydides’ Histories, there is a substantial increase in odds of the present being used when the referent is designated by proper noun in comparison with zero anaphora (odds ratio 2.939, p = 0.002). The observed effect for pronouns is almost as strong (2.787), but this is statistically much less certain (p = 0.077). In Xenophon’s Hellenic affairs, we similarly see a positive effect of subject designation with a proper noun over the baseline condition, but it is less strong and less certain (odds ratio 1.718, p = 0.045). The results for the pronoun category are statistically negligible at p = 0.525. Overall, the results from these two works support the hypothesis.

Table 3.5Tense and subject designation

^a B = 1.078 (constant = .68), standard error = .343.

^b B = 1.025 (constant = .68), standard error = .579.

^c B = .541 (constant = -.932), standard error = .269.

^d B = -.284 (constant = -.932), standard error = .447.

^e B = -.31 (constant = .364), standard error = .218.

^f B = -.609 (constant = .364), standard error = .353.

This is not the case for Xenophon’s Expedition of Cyrus: here the results are negative both for the pronoun category (odds ratio 0.544) and for the proper noun category (odds ratio 0.734). However, the effects are uncertain (p-values of 0.084 and 0.156, respectively). The 23 cases of the ‘itinerary’ present ἐξελαύνει (‘marches out’) in book 1 (see Section 3.5.2) probably have an inordinate influence: there are only three cases with the proper noun against 20 with zero anaphora. More generally, the difference in results may reflect the difference in narrative character between the works (Section 3.1.2): the narrative in the Expedition of Cyrus is more vivid than in the other two works, so that the discourse-structural quality of the present for preterite may be less pronounced here (compare the results for sentence complexity, Section 3.3.3).

Now let us zoom in on the ‘proper noun’ category. As Mulkern (Reference Mulkern, Fretheim and Gundel1996) notes, the use of a simple name (‘John’) implies greater accessibility of the referent than the use of a full name (‘John Smith’). In Classical Greek, there are no surnames. I regard as the equivalent of a full name a noun phrase including the name of the referent and either an adjective specifying the referent’s citizenship (Θουκυδίδης Ἀθηναῖος [‘Thucydides, an Athenian’ or ‘Thucydides of Athens’]) or an attributive genitive specifying the identity of the referent’s father (Θουκυδίδης ὁ Ὀλόρου [‘Thucydides the son of Olorus’]).Footnote ⁷³ As I explained at example (56), such a phrase typically implies that the designated referent is new to the discourse (or at least has been out of sight for a long time).Footnote ⁷⁴

For simplicity’s sake, I present the results for all three works grouped together, as these are representative of the individual results as well. As can be seen from Table 3.6, I found no effect whatsoever.

Table 3.6Tense and simple name versus full name

Several considerations play a role here. First, in Thucydides’ Histories, the category ‘proper noun’ yields 87 present forms against 15 aorists. These are very high odds to begin with, so it is perhaps not surprising that no further effect can be observed (in the ‘full name’ category, we find 11 present forms against 2 aorists, exactly the expected counts). Second, in the Expedition of Cyrus, Xenophon is more generous with ‘full names’ than expected on the premise that this use is mainly reserved for introducing new referents. For example, Hagasias is mentioned 15 times in the entire work, and 7 times he is called (ὁ) Στυμφάλιος (‘a [the] Stymphalian’). Perhaps the reason for this is that the origin of the commanders is considered particularly relevant in a narrative about a mercenary army. This work accounts for 31 of the total 51 cases in the ‘full name’ category reported above.Footnote ⁷⁵ In light of such considerations, I believe the hypothesis is worthy of further investigation in other (later) historiographical works, even if it is not supported by the results presented here.

3.5.3.2 Definiteness

If the present for preterite is associated with the function of introducing new referents into the discourse, then we should expect the odds of the present for preterite to increase when the subject is indefinite. Consider the following example:

(57)

μετὰ δὲ ταῦτα, ὦ ἄνδρες, χρόνου μεταξὺ διαγενομένου καὶ ἐμοῦ πολὺ ἀπολελειμμένου τῶν ἐμαυτοῦ κακῶν, προσέρχεταί μοί τις πρεσβῦτις ἄνθρωπος ὑπὸ γυναικὸς ὑποπεμφθεῖσα ἣν ἐκεῖνος ἐμοίχευεν, ὡς ἐγὼ ὕστερον ἤκουον.

After that, men of the jury, after some time had elapsed, with me being completely clueless about my misfortunes, some old woman approaches me. She {was sent} by a woman with whom that man [Eratosthenes] was having an affair, as I later heard.

(Lysias, On the murder of Eratosthenes [1] 15)

The introduction of this character is important to the development of the narrative, because she reveals to the speaker that Eratosthenes has seduced his wife.

In coding for definiteness, I have adhered to the following principles:

1. (a) As it seems that the type of coding material influences tense usage (see Section 3.5.3.1), I have limited my investigation to a select group, including only subject phrases where the head is either a noun (but not a proper noun), an adjective, or a participle.Footnote ⁷⁶ Cases where the subject or one of the subjects was designated with a proper noun (whether a person, city, or – exceptionally – a ship) were altogether excluded. Subject phrases with a pronoun as head were ignored,Footnote ⁷⁷ but any other subjects in the same sentence that fit the criterion were coded.

2. (b) For the position of the subject phrase I have adhered to the same rule as specified in Section 3.5.3.1, coding for the subject phrase as found in the main clause, in a subordinate clause or as an extra-clausal theme constituent.Footnote ⁷⁸

3. (c) I have determined (in)definiteness mainly based on the presence or absence of the definite article. There are two important exceptions. First, the word βασιλεύς (‘king’) is used without the article to specifically designate the King of Persia. This was coded as definite. Second, in expressions designating a group of men belonging to a certain city or country, the article is optional (e.g., Ἀθηναῖοι [‘the Athenians’]). These cases were coded as definite by default.

4. (d) Cases with multiple subjects were coded as ‘indefinite’ if at least one of the subjects was indefinite.Footnote ⁷⁹

5. (e) I have distinguished between two groups: one with singular verbs and one with plural verbs. In Classical Greek, a singular verb can be used with a plural subject if its gender is neuter. I have excluded these cases.Footnote ⁸⁰

The results are presented in Table 3.7. Let me begin with the plural. The results do not support the hypothesis; in fact, a negative effect is observed (odds ratio 0.653), but we need not take this very seriously in light of the high p-value (0.232). The results for the individual works do not add much nuance to the picture.Footnote ⁸¹

Table 3.7Tense and definiteness

^a B = .567 (constant = -.115), standard error = .352.

^b B = -.426 (constant = .138), standard error = .356.

For the singular, a positive effect is observed (odds ratio 1.763), but again this is highly uncertain (p = 0.107). Breaking down the results per work, we see that the effect is due mainly to Thucydides’ Histories.Footnote ⁸² An individual analysis of this work yields an odds ratio of 2.942, with a p-value of 0.033.Footnote ⁸³

As to the interpretation of these results, I offer the following thoughts. In example (57), we saw how an indefinite description was used at the introduction of a new character who played an important role in the narrative. However, the cases included in the dataset are rarely this straightforward. In the case of the singular, indefinite subjects are rarely persons or concrete objects. In most cases, the noun denotes either an action (ναυμαχία [‘naval battle’]), a natural phenomenon (χειμών [‘storm’]), or sometimes an emotion (φόβος [‘fear’]). Out of the 29 cases in Thucydides, for example, only 4 subject phrases designate a person or a concrete object (the present is used in all four cases but that may of course be accidental).Footnote ⁸⁴ With the plural, there are only 36 cases in the indefinite category, and there is some ‘noise’ here as well – that is, few cases straightforwardly conform to the type where new entities are introduced that will play an important role in the narrative.Footnote ⁸⁵

Definiteness, then, does not seem to be a predictor of tense usage, but it is uncertain to what extent this affects the claim that the present for preterite is used to introduce new referents due to the heterogeneity of the ‘indefinite’ category.

3.5.4 Conclusion

In this section, I have argued that the second main discourse-structural function of the present for preterite is to mark changes in the status of referents. I divided this into changes in activation status and changes in role or position (Section 3.5.1). For verbs of ‘travelling’ and the like, I argued that the conceptual scenario of the discourse as representation is supplemented by a scenario in which we trace the designated journey on a mental map (Section 3.5.2).

In Section 3.5.3, I translated these theoretical considerations into the hypothesis that the odds of the present for preterite being used increase as the subject becomes less predictable. This was partly corroborated by the observation that designation of the subject by proper noun increases the odds of the present being used over zero designation (but only for Thucydides’ Histories and Xenophon’s Hellenic affairs; Section 3.5.3.1). A comparison between designation by simple name and by full name yielded no effect. Definiteness (Section 3.5.3.2) did not turn out to be a significant predictor either, except perhaps with respect to the singular in Thucydides’ Histories; but the heterogeneity of the data constituted a major problem. All in all, the degree to which the present for preterite can be structurally associated with the function of marking changes in the status of referents is a matter to be investigated further.

3.6.1 The Adoption of Boeotus in Demosthenes’ Against Boeotus 1 and 2

The Demosthenic corpus contains two speeches written for a certain Mantitheus against his half-brother Boeotus.Footnote ⁸⁷ Mantitheus’ father, Mantias, had extramarital relations with a woman named Plangon. She gave birth to two sons and claimed they were by Mantias. When the man wouldn’t recognise them as his own, one of the boys threatened to take legal action. Mantias, unwilling to go to court, saw himself forced to make an arrangement with Plangon. In return for a sum of money, to be given to a third party for her sake, Plangon agreed to the following: Mantias would challenge her to declare under oath that he was in fact the father of the two boys, and she would refuse, so that Mantias would not be forced to recognise them as his own. Plangon, however, violated the agreement, accepted the oath and swore that both her sons had been fathered by Mantias. Thereupon, Mantias was forced to acknowledge them as his own and entered them into the clan register, naming the one Boeotus, the other Pamphilus.

Subsequently, the boys had to be entered in the official deme register when they reached the age of eighteen. In the meantime, however, Mantias died, and when the time came, Boeotus had himself entered under the name Mantitheus. But this was the name of Mantias’ legitimate son, who proceeded to take legal action against Boeotus to force him to drop his claim to the name Mantitheus.

I will discuss the narrative section of the first speech in its entirety and the overlapping portion of the second speech. The narrative in the second speech is longer because it deals with more issues; I will come back to this below. I divide the narrative into two portions: first, the events leading up to, and including, the trial where Plangon deceived Mantias; second, the registration of the sons of Plangon into the clan register.

The relevant passages in the first part are presented as parallel texts below. Both passages lead up to the arbitration, where Mantias administered the oath and Plangon, against the agreement, accepted it. The central point here is the difference between the aorist κατωμόσατο (‘swore’) in 39.4 and the present ὄμνυσιν (‘swears’) in 40.11. As I will argue, this reflects a difference in the way the events are presented in the two narratives.

Both passages start with the situation where Boeotus, the son of Plangon, sued Mantias (39.2 ἐδικάζετο [‘went into court’], in a main clause, 40.9 ἐδικάζετο [‘brought suit’] in a subordinate clause), aiming to be acknowledged as his legitimate son. To prevent being forced to acknowledge the boy as his son, Mantias came to an agreement with Plangon. This agreement is discussed in a little more detail in the second speech, where it gets a main clause verb (40.10 ὡμολόγησεν [‘agreed]). In the first speech, the conclusion of the agreement is more implicit: we are told in a participial clause that Mantias was ‘deceived’ (39.3 ἐξαπατηθείς) by Plangon and that he administered the oath ‘on those conditions’ (ἐπὶ τούτοις). In both cases, the conclusion of the agreement only serves as the background for what happens during the arbitration.

So let us consider how this event is construed in the two narratives. In the second speech, only the acceptance of the oath is narrated, and this constitutes the peak (culmination) of this section. The speaker uses certain attention-management strategies here that draw attention to the impact of the designated event on the narrative (compare Section 3.3). First, the event is introduced with a ‘cut-to-the-chase’ formula (τί ἂν ὑμῖν μακρολογοίην; [‘why should I make my story a long one?’]), which signals the speaker moves on to the main point. Second, the outrageousness of Plangon’s act is emphasised by the participial clause παραβᾶσα πάντα τὰ ὡμολογημένα (‘violating the entire agreement’) and by the note that the oath was ‘the opposite of her previous one’ (ἐναντίον τῷ προτέρῳ).Footnote ⁸⁸ Also, the speaker appeals to the audience’s own knowledge of the matter, claiming that it was ‘notorious’ (περιβόητος) at the time.Footnote ⁸⁹ The present forms δέχεται (‘accepts’) and ὄμνυσιν (‘swears’), then, highlight Plangon’s act of deceit as the peak of this portion of the narrative.

In the first speech, both the administration of the oath by Mantias and the acceptance of the challenge by Plangon are narrated. Here the administration of the oath is construed as a change in the narrative dynamic of the intervention type. Mantias sees himself confronted with a problematic situation: Boeotus is going to court. The administration of the oath is Mantias’ attempt to prevent this from happening. Note how the typical ‘fear’ pattern is indicative of the intervention (39.3 φοβούμενος [‘feared’]; see Section 3.4.3). After Mantias falls into the trap, Plangon’s acceptance of the oath is here presented with a high degree of presupposition. What is a main clause verb in 40.11 (δέχεται [‘accepts’]) is here only a participle (δεξαμένη [‘accepting’]). And where in 40.11, the verb ὄμνυσιν (‘swears’) is part of the focus of the clause, in 39.4 the verb κατωμόσατο (‘swore’) is presupposed material: the focus of the assertion is on the heavy preverbal constituent οὐ μόνον τοῦτον, ἀλλὰ καὶ τὸν ἀδελφὸν τὸν ἕτερον πρὸς τούτῳ (‘not only the defendant, but his brother too, her other son’).

The second part of the story is presented below. In both cases, the speaker notes that Mantias was forced to register the boys as his sons (39.4 ἦν ἀνάγκη [‘it was necessary’]; 40.11 ἀναγκασθείς [‘was compelled’]). The registration of Boeotus and Pamphilus gets a present form in 39.4 (ἐγγράφει [‘registers’]), but in 40.11 we find the aorist ἐνέγραψε (‘registered’). In the first speech, the registration of the boys is presented as a peak. After the aorists εἰσήγαγεν (‘entered’) and ἐποιήσατο (‘adopted’), referring to protocollary actions, the speaker uses a ‘cut-to-the-chase’ formula (ἵνα τἀν μέσῳ συντέμω [‘to cut short the intervening matters’]) to highlight the registration of the boys. In the second speech, the registration is simply presented as a specification of what was narrated in the previous sentence (note how the verb ἐνέγραψε [‘registered’] constitutes presupposed information).

This difference in presentation ties in with the difference in the rhetorical importance of the designated event in the context of the two speeches.Footnote ⁹⁰ In the first speech, the main concern of the speaker is to show that Mantias did in fact register Boeotus under that name and that the man should not be allowed to call himself Mantitheus. When the speaker calls witnesses to establish the truth of his narrative, this is the point he focuses on (6 ὃν μὲν τοίνυν τρόπον ἡμᾶς ἐνέγραψεν ὁ πατήρ, ἀκηκόατε τῶν μαρτύρων [‘now, you have heard from the witnesses in what manner my father registered us’]).

In this connection, it is noteworthy that we find the present marking the same fact later on in a predominantly argumentative passage, which is rare:

The use of the present tense to mark this fact both here in the argument section and in the narrative seems to underscore its rhetorical importance to the speaker. Incidentally, I think this is also the reason why the next event in the narrative in this speech is designated only with an aorist: 39.5 ἐνέγραψεν (‘registered’). That Boeotus laid claim to the name Mantitheus is not disputed; the point is that that is not the name his father gave him.Footnote ⁹¹

In the second speech, by contrast, the speaker’s goal is to regain his mother’s dowry. It seems that the speaker had lost the previous case and was now forced to sue his step-brother under the name Mantitheus (40.18). When the speaker talks about his ‘original misfortune’ (ἀτύχημα ἐξ ἀρχῆς) in the abstract (40.2) – that is, the Plangon episode which corresponds to the entire narrative in the first speech – he limits himself to Plangon’s deceit and the fact that he was forced to acknowledge two brothers and share his inheritance with them. After this, the speaker goes on to the main issue, which is the dispute surrounding the dowry (40.14–18). All in all, the main point of the Plangon story in the second speech is that her deceitful act forced Mantias to acknowledge her sons as his own; the naming issue has faded to the background.

Let me conclude. In the two speeches Against Boeotus, we see two instances where the tense of verbs referring to the same event is reversed: κατωμόσατο (‘swore’) in 39.4 becomes ὄμνυσιν (‘swears’) in 40.11, and ἐγγράφει (‘registers’) in 39.4 becomes ἐνέγραψε (‘registered’) in 40.11. These differences can be explained in terms of narrative presentation and the rhetorical concerns of the speaker. In 39.4, the aorist κατωμόσατο (‘swore’) follows upon a change in the narrative dynamic (3 δίδωσι [‘administers’]), while in 40.11, the present ὄμνυσιν (‘swears’) constitutes the peak, introduced by a ‘cut-to-the-chase’ formula and with emphasis on the outrageousness of the act. In 39.4, ἐγγράφει (‘registers’) designates the crucial point to the speaker’s case and is presented as a peak with a ‘cut-to-the-chase’ formula. In 40.11, the same fact has little bearing on the speaker’s case, and it is presented simply as a matter of protocol.

3.6.2 The Beginning of the Peace Process with Philip in Aeschines’ On the false embassy and Against Ctesiphon

In 348, Athens was at war with Philip of Macedon. Encouraged by reports that Philip wanted peace, Philocrates wrote a decree that allowed Philip to send ambassadors to Athens. Philocrates was indicted on the charge of having written an illegal decree, but he was acquitted. After further positive reports from Macedon, Philocrates wrote another decree that proposed to elect ten ambassadors to Macedon to have peace talks with Philip. Among those chosen to serve were Demosthenes and Aeschines. The ambassadors went to Macedon and came back with optimistic reports.

These and subsequent events are discussed in the two pairs of opposing speeches by Aeschines and Demosthenes: Demosthenes’ On the false embassy (19) versus Aeschines’ On the false embassy (2) and Aeschines’ Against Ctesiphon (3) versus Demosthenes’ In defence of Ctesiphon or On the crown (18). I will discuss here the two passages that lend themselves best to a side-by-side comparison: Aeschines 2.12–19 and 3.62–3.Footnote ⁹²

The starting point for my comparison is the passing of Philocrates’ first decree. Aeschines’ narrative in speech 3 begins with this event, but in speech 2, he provides some background, which I will briefly discuss here.Footnote ⁹³ Some messengers from Euboia came to Athens to talk about peace. They mentioned that Philip was also interested in making peace with Athens. Some time later, Ctesiphon was sent as envoy to Macedon in order to recover some ransom money that was paid to Macedonian privateers who had captured a certain Phryno. When Ctesiphon came back, he made a report to the assembly about the business on which he was sent, but he also added that Philip had said that he wished to end the war with Athens. This is where my discussion begins. The parallel passages from Aeschines’ two speeches are printed below:

It should immediately strike the reader that the passage on the left is full of present for preterite forms, but the passage on the right is devoid of them. This can be explained in terms of narrative presentation. First, the decree of Philocrates is marked with the present δίδωσι (‘offers’) in 2.13 but with the aorist ἔγραψε (‘made a motion’) in 3.62. In speech 2, the event is presented as a peak (culmination). All events narrated so far lead up to this point. Multiple reports have come to Athens that Philip wants to make peace, so now Philocrates proposes a decree that allows Philip to send ambassadors. The phrase ἐνταῦθα ἤδη (‘then finally’) is a typical marker of changes in the narrative dynamic, suggesting that ‘now (in the story) finally’ things start to happen. The five absolute participial clauses leading up to the main clause serve to heighten the audience’s anticipation (compare Section 3.3.3). Also, Philocrates’ introduction as a character in the narrative is highlighted by the fact that he is mentioned by a fuller designation than necessary: Φιλοκράτης ὁ Ἁγνούσιος (‘Philocrates of Hagnus’; he was widely known and was already mentioned in the proem at 6 and 8; compare Section 3.5.3.1, Footnote note 74). In speech 3, by contrast, the event constitutes the very beginning of the story. The aorist is normally used in such contexts.Footnote ⁹⁴

Similar considerations apply to the indictment of Philocrates, marked with the present γράφονται (‘attack [by legal means]’) in 2.14 but with the aorist ἐγράφη (‘was attacked’) in 3.62. In principle, the event constitutes a change in the narrative dynamic in both cases. However, in speech 2 this is more strongly highlighted. Aeschines introduces the event with a narratorial comment: πρότερον μὲν γὰρ καὶ αὐτὸ τοῦτ’ ἐκωλύετο ὑπό τινων, οἷς ἦν τοῦτ’ ἐπιμελές, ὡς αὐτὸ τὸ πρᾶγμα ἔδειξεν (‘for up to this time even that [allowing Philip to send ambassadors] had been prevented by certain men who made it their business to do so, as the event itself proved’). The indictment is presented as an intervention by men who sought to prevent peace from being made. In speech 3, on the other hand, the agency behind the indictment is backgrounded, as Aeschines uses a passive construction to describe the event.

This backgrounding of agency is part of a larger strategy of narrative presentation. In the passage from the third speech, the events are treated in a strikingly summary fashion. Several linguistic features bear this out:

1. (a) Absence of connective particles where clauses move forward narrative time.

2. (b) Absence of subordinate clauses.

3. (c) Main clauses contain only syntactically essential material: verb, subject and, in a few cases, complement.

The aim of this narrative strategy seems to be to deal quickly with facts that the speaker presents as background information.Footnote ⁹⁵ In speech 2, Aeschines pretends to give a clear and honest account of the peace of Philocrates: 11 ὅθεν δ’ ἡγοῦμαι σαφεστάτους μοι τοὺς λόγους ἔσεσθαι καὶ γνωρίμους ὑμῖν καὶ δικαίους, ἐντεῦθεν ἄρξομαι, ἀπὸ τῶν περὶ τῆς εἰρήνης λόγων καὶ τῆς αἱρέσεως τῆς πρεσβείας (‘but I will begin with those events which I think will enable me to make my presentation most clear and intelligible to you, and fair; these events are the discussion that took place concerning the peace, and the choice of the ambassadors’; trans. Adams [1919]). The narrative in speech 3 has a stronger focus. Considerations of space play a role here, as many other things had happened since the ‘false embassy’ trial that Aeschines wishes to discuss as well. With regard to the Peace of Philocrates, Aeschines explicitly states that his main concern is to discuss ‘the decrees that Demosthenes wrote together with Philocrates’ (60 τὰ ψηφίσματα … ἃ μετὰ Φιλοκράτους ἔγραψε Δημοσθένης). Philocrates’ initial proposal, which was defended by Demosthenes but not written by him or passed under his auspices, is preliminary material; as we will see shortly, we start finding present forms when Demosthenes and Philocrates start working in concert.

To conclude this section, I argue that the narrative style of the entire passage in 3.62 explains the absence of the present for preterite here. Aeschines wants to do away with these facts as quickly as he can to move on to the part that is immediately relevant to his main concern. This extends to the form ἀπέφυγε (‘was acquitted’) as well, which I have not discussed individually. For the corresponding present forms ἀποφεύγει (‘is acquitted’) and [οὐ] μεταλαμβάνει (‘does [not] get’) in 2.14, which mark a peak (culmination), see Section 3.4.1.

The second pair of parallel passages concerns Philocrates’ second decree, which proposed to elect ten ambassadors to go to Philip to talk about peace. Again, Aeschines provides background material in speech 2 (15–17) which has no parallel in speech 3. Olynthus was taken, and many Athenian citizens were captured; the Athenians sent an envoy to Philip on behalf of those captured; when the envoy came back, he reported that Philip wished to become an ally of Athens. This is where the narratives start to run parallel again:

Now the situation is exactly opposite from above: we find present forms in speech 3 where we find preterites in speech 2. In the period of Themistocles’ archonship, Demosthenes ‘enters’ (εἰσέρχεται) the Council as a member (3.62). The present marks the introduction of a main character into a new part of the narrative (see Section 3.5.1 with example [47]). Moreover, Aeschines presents this event as the start of the machinations between Demosthenes and Philocrates, which is the main concern of his narrative: Demosthenes bought his position so that he could help Philocrates in any way possible. In speech 2, there is no dramatic introduction of Demosthenes. Aeschines simply mentions at a certain point that Demosthenes was Council member (2.17 ἦν [‘was’]).

Philocrates’ proposal is marked with the present νικᾷ (‘carries’) in 3.63 but with the aorist ἔγραψεν (‘wrote a decree’) in 2.18. As I have just explained, in speech 3, the passing of the decree is presented as part of the joint machinations of Demosthenes and Philocrates, which form the main rhetorical concern of this part of Aeschines’ narrative. Note how it is introduced with the phrase ὡς αὐτὸ ἔδειξε τὸ ἔργον (‘as the event itself made evident’), which is strongly reminiscent of ὡς αὐτὸ τὸ πρᾶγμα ἔδειξεν (‘as the event itself proved’) in 2.13. In both cases, the comment serves to direct attention to the next event in the narrative as something significant.

In 2.18, on the other hand, the decree is not presented as in any way remarkable. It seems that this was simply the natural thing to do, after the first decree had been approved by the Athenian assembly, and now news had come that Philip definitely wanted to make peace. The differences in the way this second decree and the first decree are presented in the same speech are telling. In 2.13, the present in δίδωσι ψήφισμα (‘offers a motion’) is accompanied by several cues indicating a change in the narrative dynamic: the accumulation of absolute participial clauses, the phrase ἐνταῦθα ἤδη (‘then finally’), and the introduction of Philocrates by a fuller designation than necessary (Φιλοκράτης ὁ Ἁγνούσιος [‘Philocrates of Hagnus’]). Such cues are absent here.

The nomination of Demosthenes and Aeschines is presented with an aorist in 2.18 (προεβλήθην [‘was nominated’]). This corresponds to the imperfect ἦν (‘was’) in 3.63, referring to Demosthenes’ role as ambassador. In 2.18, the fact that Demosthenes and Aeschines were nominated is presented as presupposed information, as witnessed by the ellipsis of the verb in the second part (‘I was nominated by Nausicles, but Demosthenes by Philocrates himself’).Footnote ⁹⁶

To conclude, we have seen how tense usage is exactly reversed in the two parallel passages I discussed here. In the speech On the false embassy, Philocrates’ initial proposal constitutes a peak in the narrative, and the indictment of Philocrates and the outcome of the trial are also significant changes in the narrative dynamic. In Against Ctesiphon, by contrast, the whole affair is backgrounded with respect to the following narrative, where Demosthenes and Philocrates ‘write decrees together’. Consequently, we find present for preterite forms in the former version but not in the latter. Conversely, Philocrates’ second decree is presented as unremarkable in On the false embassy, while it is presented as a deceitful act in Against Ctesiphon. Here the situation with respect to tense usage is reversed, with the present being used in the second case but the preterite in the first.

3.6.3 Conclusion

In this final section, I have presented two case studies that aimed to demonstrate the relevance of the analytical tools I have laid out in previous sections for explaining tense-switching in discourse. I would not claim that these case studies are entirely representative of the corpus as a whole (large as it is), but I do think these parallel discussions have revealed some important aspects of narrative presentation that influence tense-switching. The running thread throughout my discussions was that the present for preterite is used when the narrator highlights the impact of the designated event on the structure of the story. We saw how this was reinforced in a number of ways:

1. (a) Subordinate clauses evoking narrative schemas (‘fearing’, ‘seeing’), as explained in Section 3.4.3.

2. (b) Generally, proliferation of subordinate clauses (Section 3.3.3).

3. (c) Narratorial comments: ‘cut-to-the-chase’ formulae as indicative of peaks (Section 3.6.1), ‘as-the-event-proved’ formulae as indicative of significant acts (Section 3.6.2).

4. (d) Emphasis on the outrageousness of the designated events: for example, παραβᾶσα πάντα τὰ ὡμολογημένα (‘violating the entire agreement’) in Demosthenes 40.11 (Section 3.6.1).

5. (e) Use of the marker ἤδη (‘finally’ or ‘now’).

On the other hand, the preterite is associated with backgrounded material:

1. (a) Routine or predictable events. Low information status of the verb is indicative of the predictability of the designated event.

2. (b) Resolution material (events that follow upon a change in the narrative dynamic).

3. (c) Passages characterised by a marked brevity: simple sentences with only the necessary constituents, without connective particles (Section 3.6.2).

Finally, I have argued that what narrators choose to construe as changes in the narrative dynamic is ultimately determined by their main rhetorical concerns. While tense usage is certainly not the sole diagnostic for rhetorical importance, the present tense can be used – as part of a complex of strategies – to signal to the audience that the designated event is something they should pay particular attention to.