Chapter 1 - General Conceptual Model

Summary

The use of the present tense to refer to past events may depend on two conceptual scenarios. First, the speaker may be mentally displaced to the past. Second, the speaker may pretend that the past events are currently accessible in the form of a representation. This 'representation' scenario is generally the most economic conceptual explanation for the use of the present tense to refer past events. Examples are discussed to illustrate the argument: passages from the novels of Alexandre Dumas and from Thomas Hughes' Tom Brown's schooldays; narratives accompanying security camera footage; a narrative by a character in an episode of Seinfeld; and passages from Thucydides. In all these cases the use of the present tense to refer to past events can be made sense of in terms of a conceptual representation scenario, where the difference lies in the exact nature of the representation. The more concrete the representation, the stronger the tendency for the speaker to use the present tense to designate the described events.

1.1 Introduction

In this chapter, I outline a theoretical framework for understanding tense- switching. I approach tense-switching here not just from the perspective of Classical Greek but as a cross-linguistic phenomenon, and I will discuss material from different languages and genres to illustrate the general viability of the model developed here. We will, however, move towards Classical Greek in the final sections of this chapter.

As explained in the Introduction, my account of tense-switching and, in particular, of the use of the present tense to designate past events is conceptual. This means that I try to explain how events that are temporally distal can be construed as if they occur at the time of the ground of the speaker/writer and addressees. This approach depends on the assumption that the present tense denotes present time reference. This assumption has been challenged by not a few scholars, who argue that the flexibility of use of the present tense can only be understood if we take it to be neutral with respect to time reference. I will argue in Section 1.2 that this is not a real solution to the problem and that a more intuitive approach is to adopt a broader understanding of the notion of present time reference (in terms of epistemic immediacy, Langacker [2011]; compare Allan [2011a] on Classical Greek).

With this point established, I turn to the question of how temporally distal events can be construed as temporally proximal (Section 1.3). I approach the issue from a wide perspective here, considering not just tense and other temporal expressions but also demonstratives. Taking my cue from Bühler (1990 [1934]: 154–7), I argue that the gap between the ground and distal event spaces can be bridged by two distinct conceptual scenarios. One involves mentally displacing our actual ground to a distal space, so that we imagine ourselves to be present in the distal scene. In the converse scenario, we assume that the distal entities are present in the actual ground in the form of a representation. Both scenarios can facilitate the use of proximal deictic expressions (including the present tense) to designate distal entities, but they differ in their associated linguistic characteristics. While the use of the present tense to refer to the past is generally accounted for in terms of the displacement scenario, I argue that the representation scenario constitutes a more viable explanation in most cases. Under this account, the present tense highlights the immediate accessibility of the designated past events through the medium of the representation (compare Vuillaume [1990]; Gosselin [2000]; Langacker [2011]; also von Fritz [1949]).

In Sections 1.4–1.7, I develop this argument by discussing different representation scenarios, each illustrated by a particular language or genre. In Section 1.4, we will be looking at material from nineteenth-century English and French novels. Here we encounter the idea that reading a novel is like watching a play. In this scenario, the present tense can be used to construe the past events as if they are occurring in the present story-play. We will also see, however, that there is a fluid boundary between the pretence of ‘watching a play’, on the one hand, and that of ‘displaced presence on the scene’, on the other. There is a point where the displacement scenario and the representation scenario converge, and a narrator may oscillate between the two.

Section 1.5 moves from the concept of ‘watching a play’ to an arrangement where the past events are shown on video. I begin by considering a narrative accompanying security camera footage as shown on a Dutch crime show. Here the present tense designates the past events as represented in the video. From this, we move to the use of the present in newspaper narratives reporting crimes. I argue that this use can be explained by assuming the fictive presence of a video showing security camera footage or a dramatic reconstruction of the described events (a covert scenario, Langacker [2008: 531–5]).

In Section 1.6, we will consider a conversational narrative told by a character in an episode of an American comedy show. Here I argue that the present tense designates the past events as represented in a simulation created by the narrator. I will focus on one particular aspect of narratorial simulation here, namely, depiction (which includes gesture and direct speech representation; see Clark [2016]), but I will extend the argument further in Chapter 2.

Finally, in Section 1.7, I discuss a representing entity that is purely conceptual. Here I argue that, as soon as we talk about past events, the discourse itself may be construed as a representation through which the designated past events are made immediately accessible. In this scenario, the present tense marks the ‘present occurrence’ of the designated event in the discourse, which is to say that it highlights this event as the current focus of joint attention. I illustrate this point with material from Classical Greek, and I will elaborate the argument in detail in Chapter 3.

I close the chapter (Section 1.8) with some general remarks on the nature of tense-switching patterns. First, I argue that the meaning of tense-switching is related to the nature of the underlying conceptual scenario, that is, what kind of representation is assumed to be present. Second, I point out that the ‘switched’ present tense in Classical Greek always seems equivalent to the past tense with respect to truth conditions and felicity conditions. This being the case, I identify my object of research in the subsequent chapters as the present for preterite. This allows me to avoid the term ‘historical present’, which is contentious in several respects.

1.2 The Semantics of the Present Tense

1.2.1 Temporal Neutrality

In many languages, the present tense can be used to refer to any temporal domain. It can refer, obviously, to the actual present; to the past, as we will see throughout this book; to the future, as in I leave for Paris next week; and, finally, to the generic or habitual domain: The Dean’s Conference meets on Thursdays (examples adopted from Fleischman [1990: 34]). This flexibility of use of the present tense has been thought by many to be such a great embarrassment to the idea that the present tense denotes present time reference that the linguist is forced to drop that assumption entirely.¹ As Fleischman (1990: 34) puts the problem, if we want to stick to a present time semantics for the present tense, ‘it is only through complex semantic circumlocutions that the range of other meanings can be explained, if at all’. The problem supposedly disappears if we choose ‘temporal neutrality’ as the basic meaning of the present tense. This would be reflected by the fact that the present is the unmarked tense category across languages.²

If the present has no inherent temporal meaning, the question is what motivates the variation between the present and the other tenses when they refer to the same temporal domain. There are, essentially, two possible answers to this question.

The first is based on an economy principle. If the present is unmarked with respect to time reference, then it can replace any of the other tenses when it is already clear from the context what temporal domain is referred to. Thus, Kiparsky (1968: 33–4) accounts for the ‘historical present’ in old Indo-European languages in terms of conjunction reduction, a mechanism ‘which optionally reduces repeated occurrences of the same tense to the present’. We can be swift in rejecting this thesis, as tense-switching patterns in Classical Greek do not bear out this principle. For one thing, conjunction reduction hardly explains the occurrence of single ‘historical present’ forms isolated among a large number of preterites (see example [4] in Section 1.3.2; compare McKay [1974]; Tristram [1983: 8]; Fanning [1990: 229–30]).

The second type of solution is to ascribe a specific value to the present tense but not in terms of time reference. Fleischman (1990: especially 52–63) makes a distinction between referential and non-referential or pragmatic meaning components of the tense categories. She argues that the present tense lacks a defined referential meaning component but does have certain pragmatic meaning components such as ‘subjectivity’ and ‘immediacy’. Such an approach potentially allows us to account for tense-switching in a meaningful way. However, the question is by what process the present tense acquired these pragmatic meaning components in the first place, if they are not derived from some original referential meaning.

The answer to this question is rather complex. Fleischman argues that there are two main modes of discourse: an ‘ordinary’ mode versus a narrative mode. (This echoes Benveniste’s [1959] distinction between discours and histoire and Weinrich’s [1971] besprechen versus erzählen.) In the ordinary mode, the speaker is mainly concerned with the present time. Here, reference to the past and future domains is covered by the (marked) past and future tenses. This leaves present time reference and generic reference for the (unmarked) present tense. This means that the association between the present tense and present time reference is accidental. In the ordinary mode, the present tense, being the temporally unmarked form, is simply left with the default temporal domain, which in this mode is the present. Due to this association with present time reference in the ordinary mode, the present tense acquires the pragmatic meaning components associated with present time reference – even if it never acquires the referential meaning component of present time reference itself.

What does this mean for tense-switching? When the present tense is chosen in a narrative context, the argument goes, it is to invest the narrative with the qualities associated with a current report (ordinary discourse): distance and objectivity, which are associated with the past tense, are replaced with immediacy and subjectivity. To quote Fleischman (1990: 253), the present tense ‘seeks to obliterate retrospection and transform narration into a current report where seeing and speaking ostensibly take place simultaneously’.

It will be noted that the result of Fleischman’s theoretical exercise is not much different from standard views of the ‘historical present’ as conveying the pretence that the past events are happening ‘right now’. There does not seem to be any real difference between saying ‘The present tense is used to give a narrative the pragmatic qualities of a current report’ and ‘The present tense is used to describe the designated events as if they were present’.

Moreover, the original objection to the temporal interpretation of the present tense is hardly compelling. The ‘semantic circumlocutions’ required to make the temporal interpretation work in the case of deviant usages may be complex, but that is not a real problem as long as these circumlocutions are psychologically plausible and conform to native speaker intuitions. Cognitive linguists have noted that complex mental constructions are in fact omnipresent in natural language. As Sweetser and Fauconnier (1996: 21) put it, what seems ‘improbable mental acrobatics’ is actually ‘the stuff of our everyday thinking and talking’. What seems convoluted to me is to strip the present of its intuitive meaning (present time reference) and then return that meaning to it in a roundabout manner.³

I do believe that the notions of ‘immediacy’ and ‘distance’ may be useful as schematic characterisations of the present and past tenses. What I object to is the idea that the concept of immediacy can be strictly separated from that of ‘present time’. This is well recognised in the cognitive grammar view of tense, which I will use as the starting point for my own investigations.

1.2.2 Epistemic Immediacy

In cognitive grammar, the distinction between the past and present tenses is characterised on the most abstract level in terms of ‘epistemic distance’ versus ‘epistemic immediacy’ (Brisard [2002], [2013]; Langacker [2011]).⁴ The qualifier ‘epistemic’ is somewhat confusing here. In fact, these categories denote domains in a temporal-epistemic continuum. The temporal axis represents temporal domains (past, present, future) and the epistemic axis represents degrees of reality. The point where present time and epistemic reality converge is immediate reality. This domain is denoted by the present tense. The past tense signals distance from immediate reality, either in terms of time (past time) or in terms of epistemic status (non-reality, in counterfactual statements).

The key point is that there is an asymmetry between the past and present tenses. Epistemic distance covers two domains in the continuum: that of non-present reality and that of present non-reality. Epistemic immediacy, by contrast, is confined to the intersection between present time and reality. The upshot is that the characterisation of the present tense in terms of the abstract notion of epistemic immediacy does not allow us to circumvent the notion of time reference, for epistemic immediacy entails both temporal actuality and epistemic reality.

This does not mean, however, that the present tense must always denote an actual event occurring in the here and now – the category ‘immediate reality’ is broader than that of ‘immediate experience’. A clear case is that of generics and habituals. Consider the sentence The Deans’ Conference meets on Thursdays. The view that the present tense form meets is temporally neutral because it refers to an ‘omnitemporal’ event is vitiated by the fact that the tense form can be changed: met, will meet.⁵ Obviously, the present tense signals that the utterance concerns the present somehow. But it is also true that the utterance does not refer to an event that is actually taking place in the present. As Langacker (2009: 267) puts it, the utterance describes a virtual event, an event that is mentally ‘conjured up’ to express a generalisation. We take it to be part of our present reality that the deans have the habit of meeting on Thursdays.

In conclusion, the notion of epistemic immediacy in cognitive grammar is not a replacement for the notion of present time. Rather, the concept of what is ‘present’ is understood in broad terms so as to include both actual events happening right now and virtual events that may be mentally conjured up. Such an approach allows us to accommodate recalcitrant uses of the present tense without us having to revise our most intuitive assumptions about its semantics.

1.3 Bridging the Gap between the Near and the Far: Two Conceptual Scenarios⁶

1.3.1 Displacement versus Representation

If we assume that the present tense signals that the designated event coincides with the time of the ground, then the use of the present tense to designate events that are temporally distal must be facilitated by some special conceptual scenario.⁷ This issue pertains not just to the present tense but generally to proximal deictic expressions (e.g., here, now, this) used to refer to distal entities. An example is the use of the adverb now in combination with the past tense, where the adverb seems to designate a past time (‘past + now construction’, e.g., Nikiforidou [2010], [2012]):

And now Tom for the first time saw his future school-fellows in a body.

(Thomas Hughes, Tom Brown’s schooldays, chapter 5)

The issue extends to certain uses of distal deictic expressions as well. Demonstratives such as there and that may be used to point to entities that are present in the shared environment of the speaker and addressees. Such expressions can be used to suggest the presence of an entity that in fact is nowhere to be seen. For example:

The new ball you may see lie there quite by itself, in the middle, pointing towards the School or island goal; in another minute it will be well on its way there. Use that minute in remarking how the Schoolhouse side is drilled.

(Thomas Hughes, Tom Brown’s schooldays, chapter 5)

Of course, the reader is unable to actually see the ball that is going to be used in the rugby match in the story.

The main argument in this chapter is that there are two distinct conceptual scenarios that may facilitate such ‘paradoxical’ usages of deictic expressions. This was recognised already by Bühler (1990 [1934]: 154–7) in connection with spatial deixis, but the distinction is rarely made by scholars working on time and viewpoint today.⁸ Bühler distinguished a type of deixis where demonstratives are used to point to entities that are not physically present. He argued that this ‘deixis ad phantasma’ may be supported by two conceptual operations: either the conceptualisers are mentally displaced to the distal space or the present space (the ground) is transformed into a stage on which the distal entities are represented.

I will illustrate this distinction with a contrived example (inspired by Bühler), which will facilitate our understanding of the data discussed in the rest of this chapter. Suppose that someone wants to describe the battle between Alexander and Darius at Gaugamela (331 bc) as if it is somehow ‘happening right now’. This may take two forms.

1. (1) Displacement scenario: ‘We are standing on the plain of Gaugamela. It is the first of October, 331 bc. Over there on the left are the forces of Alexander. On the right, you can see King Darius.’

2. (2) Representation scenario: Imagine that the speaker is standing in front of a model of the plain of Gaugamela. The forces of Alexander and Darius are represented by miniature soldiers. The speaker says, ‘This is the plain of Gaugamela. Here are the forces of Alexander. On the right is King Darius.’

In the first scenario, we are virtual observers of actual events. The present tense forms designate the actual past events as seen from a fictionally displaced ground. The expressions over there, on the left and on the right are anchored in this alternative ground. In the second scenario, we are actual observers of virtual events (for the notion of ‘virtual events’, see Section 1.2.2). Grounded in the actual present, we are looking at a representation of past events. The present tense refers to the time in which we are looking at the representation. Similarly, the proximal demonstrative this designates the model representing Gaugamela, and here designates a part of this model. The expression This is the plain of Gaugamela is the result of a blend, a conceptual integration of the actual entity with the representing entity (on blending, see, e.g., Fauconnier and Turner [2002]).

In actual discourse, a speaker may use the present tense to refer to past events without explicitly specifying the underlying scenario. That is, the speaker may simply act as if we are standing on the plain of Gaugamela, or act as if there is a model (or another representation) in front of us. It is then up to the listeners to infer what kind of construal is implied (a covert scenario, Langacker [2008: 531–5]). When the underlying construal is implicit, certain utterances may be underspecified with respect to which scenario is intended:

1. (3) Underspecified. ‘It is hot on the plain of Gaugamela. Alexander is on the left; Darius is on the right. Both are eager to start the battle.’

In such cases, an interpretation in terms of a displacement scenario and one in terms of a representation scenario may be equally valid: the utterances can be made sense of both in terms of a displaced perspective in the narrative world and in terms of looking at a model.

1.3.2 Representations and the Present Tense

Now that the basic concepts have been clarified, let us consider the state of the question with respect to deixis and viewpoint in narrative discourse. Most scholars have taken for granted that the use of deictic expressions to designate absent entities must be resolved in terms of a displacement scenario. Under this account, these deictic expressions are anchored in the ground of a displaced consciousness from which perspective the designated entities are part of the actual environment (‘allocentric perspective’, Klum [1961: 164]; ‘decoupling of the deictic centre’, Langacker [1991: 267]; ‘base shift’, Cutrer [1994]; Adema [2019]; ‘allocentric reference system’, Linhares-Dias [2006: 2–3]). This construal is illustrated by the following extract from a Dutch university newspaper:

University president Sibrand Poppema can’t get the smile off his face today. It is May 15, 2014, and his beloved RUG [University of Groningen] is four hundred years old. The entrance to the Academy building, the stately headquarters of the university in the city centre of Groningen, has been newly painted, and a large red banner adorns the balcony. A carpet metres long lies ready for the king, the guest of honor.⁹

(UK Magazine, Roven van de RUG. Carlien Bootsma and Peter Keizer, 11 May 2016; http://archief.ukrant.nl/magazine/roven-van-de-rug, accessed 31 January 2017)

The article begins in medias res. By saying It is May 15, 2014, the narrators explicitly assume a ground that is different from their actual here and now. The proximal temporal adverb today is anchored in this displaced ground. The use of the present tense throughout is a natural consequence of this pretence of cotemporal narration. The attention paid by the narrators to visual details (the newly painted entrance, the red banner, the long carpet) strengthens the effect, allowing us to ‘take in the scene’ as if we were actual spectators.

Under this approach, ‘clashes’ between deictic elements, such as the use of the past tense in combination with the adverb now in (1), can be explained in terms of a multiple viewpoint construction (Sweetser [2013]; van Krieken et al. [2016]). In this particular instance, the past tense form saw then signals that the designated event is distal with respect to the ground; the adverb now, on the other hand, can be assigned to a viewpoint within the narrative world – in this case, that of Tom, the main character. The effect is that we partly align our viewpoint with that of the protagonist, so that we are immersed more deeply into the story. The use of the present tense in combination with distal adverbs, as in, for example, Then a man walks in, would be the inverse analogue of the ‘past + now’ construction: here, tense is grounded in a displaced viewpoint, but the adverb is anchored in the actual ground.

The central problem with this general approach is that the present tense can be used to refer to the past even when the way the narrative is construed suggests a decidedly retrospective viewpoint. Let us consider an example from the Classical Greek historian Thucydides:

οἱ δὲ Ἐπιδάμνιοι οὐδὲν αὐτῶν ὑπήκουσαν, ἀλλὰ στρατεύουσιν ἐπ’ αὐτοὺς οἱ Κερκυραῖοι τεσσαράκοντα ναυσὶ μετὰ τῶν φυγάδων ὡς κατάξοντες, καὶ τοὺς Ἰλλυριοὺς προσλαβόντες. προσκαθεζόμενοι δὲ τὴν πόλιν προεῖπον Ἐπιδαμνίων τε τὸν βουλόμενον καὶ τοὺς ξένους ἀπαθεῖς ἀπιέναι· εἰ δὲ μή, ὡς πολεμίοις χρήσεσθαι.

The Epidamnians didn’t listen to them at all, so the Cercyreans make an expedition against them with forty ships, together with the exiles, so as to restore them, and taking the Illyrians in addition. Besieging the city, they proclaimed that whoever of the Epidamnians wished to do so, as well as the foreigners, might leave unharmed; if not, they would treat them as enemies.

(Thucydides, Histories 1.26.4–5)

The present tense form στρατεύουσιν (‘make an expedition’) is flanked by two preterites: (οὐδὲν) ὑπήκουσαν (‘didn’t listen at all’) and προεῖπον (‘proclaimed’). In fact, the present form here is quite lonely in the wider context: the previous one occurs at 24.6 (πέμπουσιν [‘send’]), the next one at 46.1 (ἀφικνοῦνται [‘arrive’]). Nothing at this point seems to indicate a displaced ground. In fact, such an interpretation is actively discouraged by the way the designated event is construed. Thucydides reduces an event that takes up a long time to occur to a conceptualisation that can be processed in a moment. In other words, there is a large distance between discourse time and story time.¹⁰ Such compression of time is only possible from a retrospective viewpoint. The resulting conceptualisation retains very little of the character of immediate experience.

Such issues are not limited to one particular genre or language. I will furnish many examples in the following sections, but here let me just present another example to further illustrate my point. The following is an excerpt from a Dutch newspaper narrative (from the appendix in van Krieken et al. [2016: 167–8]):

For about two minutes, Vleerlaag hears shots [fired] at irregular intervals.¹¹

(NRC, Bij de groenteafdeling ligt een man. Joke Mat and Bart Funnekotter, 11 April 2011; www.nrc.nl/nieuws/2011/04/11/bij-de-groenteafdeling-ligt-een-man-12009982-a728227, accessed 9 December 2018)

The narrative here is certainly more vivid than in (4): discourse time comes closer to story time, and perception is evoked (hears). But the reduction of minutes of story time (For about two minutes) to a moment of discourse time and the characterisation of separate intervals within that time frame betray a retrospective viewpoint. An observer on the scene would not be able to oversee the temporal structure of the events in this way.

It is useful here to consider what an actual report by a commentator on the scene looks like, so that we can contrast its features with what we see in (4) and (5). The following is a brief excerpt from a commentary on a football match in the Dutch league:

De Ligt, hesitates, the ball bounces up in a weird way, now he has lost Kramer, he commits a foul.¹²

(NOS Studio Sport Eredivisie, 18 May 2018, Sparta against Ajax, at 4:40 playing time. https://nos.nl/uitzending/31958-nos-studio-sport-eredivisie.html, accessed 9 June 2018)

The possibilities afforded to a commentator on the scene to represent the observed events in discourse are constrained by temporal and epistemic factors (Langacker [2011]). The temporal constraint entails that the commentator is bound to real time. This means that the time it takes the commentator to describe the observed events will be roughly the same as the time it takes for the events to actually occur. The epistemic constraint entails that the commentator does not know beforehand what is going to happen, so that it is not possible to anticipate any relation between the event occurring at a particular moment and the events that follow.

The result is that this is no longer a narrative but a report (see Smith [2003] for the distinction between these modes of discourse). Narrators are able to edit their representation of the designated events because of their retrospective viewpoint. Features of such editing are compression of time, abstraction and the construal of a grammatically complex discourse. The commentator on the scene, by contrast, reports the events as they occur. The result is a series of disjunct discourse units (note the absence of connectives in [6]), each representing a momentary observation in a new ‘now’ (note now he has lost Kramer).

The ‘historical present’ has been associated with an immediate mode of representation, which seeks to replicate the features of an eyewitness report (with specific reference to Thucydides: Allan [2007], [2011a], [2013]; Aerts [2014]; see also Kroon [2002] on Latin). While this makes sense for some usages, examples such as (4) and (5) point to the need for an alternative account (compare, e.g., von Fritz [1949]; Willi [2017: 137–41]; Huitink [2019: 188–95]).

The question we need to answer, then, is how past events can be construed as part of the present, but in an edited, mediated form. The answer lies in Bühler’s second interpretation of ‘deixis ad phantasma’, where distal entities are brought into the present space. I argue that the use of the present tense as illustrated by examples (4) and (5) is facilitated by a conceptual scenario where the past events are assumed to be presently accessible in the form of a representation (see also Vuillaume [1990]; Gosselin [2000]; Langacker [2011]). Because a representation is itself an edited, ‘digested’ version of the events it represents, the temporal and epistemic constraints do not apply to a narrator who describes ‘what is happening’ in a representation in the same way as they apply to an observer of actual events.

Representations come in many forms, differing in the degree of editing that is done on the actual events. This editing may be understood in terms of iconicity versus schematicity. For example, a representation of the battle of Gaugamela in the form of a model and toy soldiers is highly schematic. To note only two aspects, size is compressed (the plain is reduced to a model), and when the pieces are moved, time will be compressed as well (it only takes moments to move entire contingents on a model). On the other hand, a live re-enactment, acted out on a large plain by actual people dressed up as Greek and Persian soldiers, is a highly iconic representation. Here, the effect of the representation is almost to displace those who witness it to the past.

In short, different types of representations imply different degrees of conceptual distance to the actual past events. Another way to put this would be to say that the fourth wall marking off the representation as a separate world may be more or less ‘solid’, depending on the degree of iconicity or schematicity in the representation. In Sections 1.4–1.7, I will discuss different representation scenarios that I believe underlie the use of the present to refer to the past in different genres. Each genre is represented by a particular language (or, in Section 1.4, two related languages). We will see how the present tense can be used to designate past events as represented in a virtual play (Section 1.4), a video (Section 1.5), a simulation or re-enactment (Section 1.6) and the discourse of the speaker itself (Section 1.7). The point of this overview is to illustrate the general viability of the account presented here and also to show that different scenarios carry distinct implications for the meaning of tense-switching (Section 1.8).

1.4 On Both Sides of the Fourth Wall: Nineteenth-Century Fiction

In this section, we will explore the boundary between the displacement scenario and the representation scenario. At what point does the pretence of presence at the distal scene (displacement scenario) transform into the pretence of a present re-staging of the distal events (representation scenario)? To answer this question, I will discuss material from nineteenth-century French and English literature. We will see how the narrators here explore the boundaries between these two construals both implicitly and explicitly.

As Vuillaume (1990) shows, novelists such as Alexandre Dumas often reflect on the viewpoint relationship between themselves and the addressees, on the one hand, and the described events, on the other. In the following example, we see how Dumas¹³ explicitly sets up a displacement scenario:

Let us transport in a full jump, without preface, without preamble, those of our readers who do not fear to step backwards with us, three centuries into the past, into the presence of the men with whom we are going to acquaint them, and into the middle of the events we are going to make them witness. It is [lit. ‘we are’] the fifth of May in the year 1555. Henry II reigns over France.¹⁴

(Alexandre Dumas, The page of the duke of Savoy, 1.1)

The readers are to imagine themselves being transported into the past, exchanging their current ground for a displaced one within the narrative world, at 5 May 1555. This arrangement also allows the narrator to construe himself and the readers as moving observers:

So, we are moving through the part of old France that people used to call ‘l’Artois’, but that nowadays we call the department of ‘Pas-de-Calais’.¹⁵

(Alexandre Dumas, The page of the duke of Savoy, 1.1)

This reinforces the idea of a displaced ground. As the displaced ground is a fictional construct to begin with, it is free to move around. In the representation scenario, by contrast, the past events are transported on a present stage, so that the readers stay where they are and have the events come to them.¹⁶

In other places, Dumas construes the reading experience in terms of a representation scenario. For example, he calls his novel Mohicans of Paris a ‘mirror to reflect faded events’ (un miroir à refléter les événements évanouis). Similarly, his novel Joseph Balsamo is a ‘moving tableau’ (tableau mouvant; both examples are discussed in Vuillaume [1990: 70]). The central metaphor I wish to focus on here is that of the narrator as a director, as in the following example:

So, when the new actor we introduce on stage ventured into the middle of the cheeky population of the hall, there was a huge outbreak of laughter.¹⁷

(Alexandre Dumas, Mohicans of Paris, chapter 111, cited in Vuillaume [1990: 68])

The narrator presents himself as a director who reproduces the past events on a present stage. I call this fictive reproduction the ‘story-play’.¹⁸ Interestingly, despite this pretence of present re-enactment being active, the narrator may still choose to use the past tense (and other distal deictic expressions) to designate the past event space, as here. We will see examples of proximal deictic expressions referring to the story-play in another novel shortly.

Finally, Dumas also lays bare the ambiguity between these two conceptual scenarios – being transported into the story world or watching the story-play from the actual ground – in a most salient manner. The following passage constitutes the set-up to a new chapter of his Mohicans of Paris:

As we suppose that the reader, from the moment when he attaches himself to us, is not hostile to these adventures, we are going to beg him to follow us to the spot to where we transport our camera obscura, to let pass before him a crowd of characters no less mysterious than the Chinese shadow shows of Mr. Seraphin. The theatre, as we have said, is located at Rue des Postes … the decor represents a small one-storey house.¹⁹

(Alexandre Dumas, Mohicans of Paris, chapter 116, cited in Vuillaume [1990: 69])

As in (7), Dumas asks his readers to assume a displaced ground in the story world (note follow, transport). What is different here is that we are limited to the enclosed space of a camera obscura. Even though we are transported into the story world, we are strictly separated from it by, literally, four walls. The same ambiguity is present in the next sentence, where the construal subtly changes. Now we are sitting in a ‘theatre’ looking at a ‘decor’. But like the camera obscura, the theatre is transported to the time of the actual events. This is rather paradoxical, as the whole point of a theatre is to transport distal events onto a present stage. In short, Dumas displaces us into the past, but at the same time, the experience of witnessing the past events is construed as the experience of looking at a representation (photographs, a play). This is the exact point where the displacement scenario and the representation scenario converge.

When the narrator is not explicit about the conceptual scenario underlying the presentation of the narrated events, the scenario may either be underspecified (i.e., both interpretations may be equally valid) or certain linguistic features may tip the scales towards one interpretation or the other (compare Vuillaume [1990: 42], who speaks of a ‘semantic equilibrium’ in this connection). I point to two aspects here, illustrating my arguments with examples from Thomas Hughes’ novel Tom Brown’s schooldays.

The first point concerns proximal temporal adverbials. These differ in the degree to which they are committed to actual proximity to the designated events. The adverb now is wholly ambiguous in this respect. Consider the following example:

The second act of Tom’s life may now be said to have begun.

The war of independence had been over for some time: none of the women now, not even his mother’s maid, dared offer to help him in dressing or washing.

(Thomas Hughes, Tom Brown’s schooldays, chapter 3)

First, note that the experience of reading is construed as watching a play (the second act). Now here designates the time of this story-play: at this point during our reading, we may say the second act has begun. In the next sentence, the narrator uses the past tense form had been to designate the actual event space. Then, we find the use of the past tense continued (dared) but combined with the adverb now. At this point, we can see how this use of the adverb is ambiguous. As I explained above, the ‘past + now’ construction is typically explained in terms of a displacement scenario, where now designates the present time of an observer within the narrative world (see, e.g., Nikiforidou [2010], [2012]). But it is also possible to have now refer to the time of the reading, so that it designates the present moment in the evolution of the story-play (compare Fauconnier [1984: 180]; Vuillaume [2008]). In that case, we may paraphrase the utterance as follows: ‘In the present moment the play has arrived at the point that represents the actual past time in which no one dared to help Tom.’²⁰ Both interpretations seem equally valid here. The activation of the representation scenario in the immediately preceding discourse (the second act, evoking a play) may favour the second interpretation, but this is rather incidental.

On the other hand, adverbs such as today or tomorrow cannot naturally be used to denote the time of the reading, because this process is normally not measured in days (see Fauconnier [1984: 180–1]; Vuillaume [1990: 97]). Such adverbs are thus more strongly indicative of a displaced viewpoint. We saw this in example (3), which I discussed in Section 1.3.2 (University president Sibrand Poppema can’t get the smile off his face today). An example from Tom Brown’s schooldays is the following:

The rest of the sixth go forwards into the close, to see that no one escapes by any of the side gates. To-day, however, being the School-house match, none of the School-house præpostors stay by the door to watch for truants of their side.

(Thomas Hughes, Tom Brown’s schooldays, chapter 5)

Because today cannot naturally mean ‘today as we are reading’, it seems to impose a displacement scenario. On the other hand, it might be argued (with Fauconnier [1984: 180–1]) that such usages of today may be paraphrased: ‘on the day we are talking about’. Such an extension of meaning could well be the result of a subjectification process, where the ‘textual’ implications of the word develop into a conventionalised meaning (see, e.g., Diewald [2011]). However that may be, the point remains that this is more of a stretch for today than for now.

On the other side of the spectrum are demonstrative expressions, which tend to be more compatible with a representation scenario than with a displacement scenario. Consider the following example:

At this moment Griffith, the itinerant vender of oranges from Hill Morton, enters the close with his heavy baskets.

(Thomas Hughes, Tom Brown’s schooldays, chapter 5)

A rugby match is in progress. Half-time has come, the sides change goals, and there is a break. Then Griffith enters the scene. The phrase at this moment can only be naturally used to refer to the actual present when the designated event is ongoing at reference time (At this moment, he is doing the dishes). A perfective construal, as implied by the simple present form enters, is incompatible with this phrase. For example, a basketball commentator will not say At this moment, he takes a shot. In my view, the demonstrative this requires a discourse-deictic interpretation – that is, it designates the time frame currently under discussion. Unlike now, it indicates a time that is not actually the present and therefore implies a representation scenario.

The second point concerns compression and abstraction. A problem with the displacement scenario is that it is generally difficult for a narrator to consistently maintain the pretence of cotemporal narration. For example, if the narrator of Tom Brown’s schooldays were to describe the entire rugby match as if in real time, this description would take up too much space. Some parts need to be passed over more quickly than others. Here the narrator starts to ‘edit’ the description, and the pretence of presence at the scene transitions into a more schematic representation of the designated events.

Let us take a closer look at some passages in the rugby match narrative to see this point. In the part leading up the match, the narrator employs a displacement strategy:

The new ball you may see lie there quite by itself, in the middle, pointing towards the School or island goal; in another minute it will be well on its way there. Use that minute in remarking how the Schoolhouse side is drilled.

(Thomas Hughes, Tom Brown’s schooldays, chapter 5)

Two features suggest a displacement scenario here. First, the focus on direct visual access (The new ball you may see lie there). Second, the pretence that time elapses for us the same way as it does in the story world – we can ‘use the minute’ before the ball is kicked to focus our attention on another feature of the scene.²¹

Note also the use of the adverb now when things become more dynamic:

Old Brooke is talking to the captain of quarters; and now he moves away.

By using the adverb now to move forward narrative time, rather than then or next, the narrator pretends to be bound by the epistemic constraints of an actual observer on the scene (see example [6] in Section 1.3.2).

Now compare the following description of the match in progress:

The ball is returned, and they meet it and drive it back amongst the masses of the School already in motion. Then the two sides close, and you can see nothing for minutes but a swaying crowd of boys, at one point violently agitated.

Time is compressed in this passage: the verbs is, meet and drive designate a complex of events that takes more time to occur than the few moments it takes to narrate it. The effect becomes stronger in the next sentence, where minutes of actual time are compressed into one utterance (you can see nothing for minutes). The use of then here, instead of now, conforms to this change in narrative strategy: while the events are ‘present’ in the sense that they are evolving in the present story-play, we are looking at them from a more distant perspective than in (14) and (15). The narrator is able to zoom in and out as he pleases, and as he does so, our understanding of the implicit viewing arrangement changes.

To conclude, let me summarise the main points made in this section:

1. (a) The boundary between a displacement scenario and a representation scenario – in this case, between virtual presence at the scene and watching a virtual play – is fluid. Specific instances of ‘deixis ad phantasma’ may be underspecified with respect to which scenario is presupposed.

2. (b) Certain linguistic cues can tip the balance in one direction or the other. This concerns, in particular, the use of adverbs (today – now – in this moment) and the degree of temporal compression and abstraction. Changes in these parameters affect our implicit understanding of the underlying conceptual scenario.

1.5 Video Footage and Dutch News Narratives

In this section, we move from the idea of ‘watching a play’ to a scenario where the past events are made presently accessible in the form of video footage. My examples will be taken from Dutch news media. Here, the representation scenario moves further away from the displacement scenario, as the representation is confined to a physical object (the television). The fourth wall solidifies, and the viewing arrangement remains static throughout.

The following passage is a (translated) transcription of a narrative accompanying security camera footage, as shown on a Dutch television programme:

(a) Cameras on the terrain register the first movements of that morning at about six-thirty, right in front of the recycling company in De Mortel. It is someone who first potters about a little, crosses the road and then disappears from view.

(b) Was this the perpetrator?

(c) And then it becomes quiet again for a while. Until a little before six; then the first employee arrives at work. …

(d) Pay attention to the man who now walks the street, suddenly turns around and who runs towards the company.²²

(Bureau Brabant, Bewakingsbeelden overval recyclingsbedrijf Eindhoven. 31 October 2014; www.youtube.com/watch?v=WNy3ZDV167c, accessed 26 March 2018)

In this short fragment, we see how the narrator uses three distinct deictic strategies. Each of these seems to have a particular communicative function. First, the narrator may exclusively highlight the present representation space, as seen in part (d): Pay attention to the man who now walks the street etc. The speaker is no longer narrating in the strict sense but telling the audience to pay attention to the present representation. The description in the subordinate clause (who now walks the street etc.) is a report, that is, a real time description of what is happening right now (see Section 1.3.2). The speaker’s aim here is to direct the audience’s full attention to what can be observed in the representation. The designated person is also visually highlighted in the video at this point.

Second, in (b) the narrator uses the past tense: Was this the perpetrator? This can be interpreted in two ways. Under one interpretation, the past tense highlights the actual past event space. In this way, the narrator focuses on what actually happened, that is, the question whether the person presently shown in the video actually was the person who would later commit the robbery. Alternatively, the past tense may be taken to refer to the past within the context of the representation. After all, video footage has a temporal dimension as well, and the appearance of the designated person in the video can be construed as anterior to the present moment of viewing.²³

Present tense narrative, as observed in (a) and (c), allows the narrator to mediate between the present representation space and the actual past event space. On the one hand, distal temporal expressions (that morning, then, a little before six) are used to designate the original temporal structure of the past events. This is important, as the narrator’s aim is to give an accurate reconstruction of what actually happened so that the crime may be solved. At the same time, the use of the present tense serves to keep the narratees engaged with the video, so that they closely follow what happens.

Notice how different this narrative style is from an actual on-the-scene report. I reproduce example (6) below:

De Ligt, hesitates, the ball bounces up in a weird way, now he has lost Kramer, he commits a foul.

As I pointed out in Section 1.3.2, the cotemporal perspective imposes temporal and epistemic constraints on the commentator’s reporting strategies. The video commentator in (17), by contrast, is not constrained in this way. In the first place, the narrative was construed after the fact. Second, the representation is itself an edited version of the actual past events. This translates into narrative editing. For example, the narrator is able to skip time (And then it becomes quiet again for a while) because there is a cut in the video.

I propose to extend this argument to present tense crime reports in newspaper articles. The use of the present tense here has been explained in terms of a multiple viewpoint construal (van Krieken [2016]; van Krieken et al. [2016]). Without wishing to exclude this possibility, I propose an alternative explanation. Consider the following passage:

Saturday afternoon at about twelve o’ clock, Tristan van der Vlis parks his black Mercedes at the Carmenplein by the shopping centre ‘de Ridderhof’. He is carrying three firearms. He gets out and shoots someone. Then he goes up a stone staircase on the side and into the shopping centre through a door.²⁴

(NRC, Bij de groenteafdeling ligt een man. Joke Mat and Bart Funnekotter, 11 April 2011; www.nrc.nl/nieuws/2011/04/11/bij-de-groenteafdeling-ligt-een-man-12009982-a728227, accessed 9 December 2018)

In my view, the narrators are telling the story as if they have access to a video showing the described events, either as footage of actual events or as a dramatic reconstruction.²⁵ No video is mentioned, but we, the addressees, supply the conceptual scenario in order to make sense of the consistent use of the present tense here.

This interpretation is cued, first, by the subject matter, as we are familiar with the experience of seeing crime reports on television; and second, by the narrative style of passages such as (19), which reflects the character of video commentaries. First, there is the combination of present tense forms with distal adverbs (Saturday afternoon at about twelve o’clock, then). Second, the narrators are able to compress time (the time it takes for these events to actually occur is longer than the time it takes to process the narrative). Third, they have knowledge that was unavailable to people who were actually on the scene (the fact that Tristan was carrying three firearms when he got out of the car). Fourth, the way the discourse is construed shows the narrators are able to anticipate what will happen. A case in point is the verbal ellipsis in the final sentence: goes is supplied first by up a stone staircase and then by into the shopping centre. All these features can be made sense of in terms of a conceptual scenario where the narrators are telling the story as if they are reporting what is happening in a replay/reconstruction as seen in a video.

In conclusion, I have argued for the following points:

1. (a) The accessibility of video footage as a representation allows the narrator to construe the past events as present in edited form.

2. (b) In this scenario, distal temporal adverbial expressions designate the past event space, which is important because the audience will be interested in the details of what actually happened. At the same time, the present tense keeps the audience engaged with the replay/reconstruction that is evolving in the here and now.

3. (c) In crime reports in newspaper articles, the use of the present tense can be explained in terms of a fictive arrangement whereby the narrator tells the story as if providing a commentary to video footage.

1.6 Simulation in a Performed Story

A play and a video have in common that they represent events with a high degree of consistency. In this section, I will discuss a type of representation that is more mutable. As the degree of iconicity in the representation changes, so does the propensity for the narrator to use either the past or the present tense to designate the narrated events. Here we come closer to the kind of tense-switching patterns we see in the Classical Greek corpus (see Chapter 2).

I will be discussing a conversational performed story, which means a story where the act of narrating acquires a theatrical character (Wolfson [1978], [1982], [1989]; Fleischman [1986], [1990: 8–9]). This is achieved mainly through depictions (Clark [2016]) such as gestures, sound effects and direct speech representation. These effects allow the narrator to stage a simulation of the designated past events in the surrounding physical space. With this construal in place, the present tense may be used to signal the present occurrence of the designated past events in the simulation.²⁶

My case study is a narrative told by a character in the American comedy series Seinfeld. This is not a natural narrative in the strict sense, but that is not a real problem. One way to approach the issue is to assume that art must imitate life and that an actor who is allowed to improvise will intuitively make the speech of his character like actual speech. Another perspective would be to regard ‘staged conversational narrative’ as a genre that is distinct from natural conversational narrative; in that case, it is still perfectly legitimate to explore tense-switching here in its own right. As the material I discuss in Chapter 2 will be taken from theatrical texts, the present example of a ‘narrative on stage’ is actually opportune.

The passage I will discuss is taken from season 4, episode 15 (‘The visa’), and the situation is as follows: Jerry Seinfeld (the eponymous character) and Elaine (his friend) are hanging out in Jerry’s apartment, when Kramer (Jerry’s highly eccentric neighbour) comes in. He has just returned from a ‘fantasy baseball camp’, but he is a few days early; so Jerry and Elaine ask him what happened. Kramer’s answer is that ‘there was an incident’, which turns out to be the fact that he punched Mickey Mantle in the mouth. Kramer begins his story as follows:²⁷

[Kramer] Well, you know, we were playing a game and, uh, you know, I was pitching, and I was really, you know, throwing some smoke. And uh, Joe Pepitone, he was up, and man, that guy, you know, he was crowding the plate.

In terms of episodic structure, this is an orientation section (Labov and Waletzky [1967]; Labov [1972]; Fleischman [1990]; Allan [2009 etc.]). The starting situation is described: Kramer is pitching and Joe Pepitone is batter. At this point, Kramer is not yet acting out the described actions. The backgrounded status of this passage, together with the lack of acting on Kramer’s part, is reflected in the use of the past tense, conveying conceptual distance.

Next, Kramer starts to set up a stage:

[Jerry] Wow, Joe Pepitone!

[Kramer] Well, Joe Pepitone or not, I own the inside of that plate!

As Kramer says that plate, he points to his right, as if the plate may be seen there. Kramer is setting up a simulation space (Liddell and Metzger [1998]; Liddell [2003]), which means that he assigns a part of the actual physical surroundings to serve as the stage for a re-enactment of the narrated past events. By pointing at the virtual pitcher’s plate, Kramer implies that part of the room now represents the baseball field.

The simulation becomes more elaborate in the next section, where Kramer starts to use iconic gesture to depict the narrated events:²⁸

[Kramer] So you know I throw one, you know, inside, you know, a little chin music, put him right on his pants. Cause I gotta intimidate, you know, when I’m on the mound.

Here, Kramer enters full acting mode. As he says I throw, he faces the imaginary plate and makes an elaborate and violent throwing gesture. When he says a little chin music, he suddenly takes on the part of Joe Pepitone, bending backwards as if to avoid the ball. At put him right on his pants, Kramer raises his arm and throws it down.

I argue that the present tense form throw and the other present tense forms referring to the past we will see below refer to the events as they occur in the simulation staged by Kramer. As for put, the tense is ambiguous, depending on who or what we take to be the subject: Kramer (I put, which may be present or past tense) or the ball/the entire event (it put, past tense). The present gotta seems to have a generic character: a pitcher generally needs to intimidate when on the mound (the same logic applies to own in [21]).

The next phase in the story is introduced as follows:

[Kramer] Well the next pitch, he’(i)s right back in the same place! So … I had to plunk him.

[Jerry] You plunked him?

As Kramer says he’s right back in the same place, he points to the imaginary pitch. The present tense designates the situation as represented in the simulation space. Kramer then switches to the past tense: I had to plunk him. This is commentary rather than narrative proper. Moreover, at this point Kramer halts the simulation, putting his hands in his pockets and looking away uncomfortably. The past tense, therefore, refers to a situation that cannot be witnessed in the simulation. Finally, the past tense in Jerry’s question You plunked him designates the past event space, as Jerry focuses on whether Kramer actually plunked Pepitone.

Then follows the main crisis, or peak:

[Kramer] O yeah. Well he throws down his bat, he comes racing up to the mound. Next thing both benches are cleared, you know. A brouhaha breaks out between the guys in the camp, you know, and the old Yankee players, and as I’m trying to get Moose Skowron off one of my teammates, you know, somebody pulls me from behind, you know, and I turned around and I popped him. I look down and whoa man, it’s Mickey! I punched his lights out.

It would take too long to describe all of Kramer’s gestures here, but they are especially iconic at throws (throwing gesture towards the ground), pulls (grabs his own coat by the neck and violently pulls it back) and I look down and whoa man, it’s Mickey (looks down and turns his head away in horror). Overall, Kramer’s acting gives the strong impression of a re-enactment, and the present tense is used to designate the past events as they occur in this representation.

It is worth pausing at the past tense forms for a moment. The form punched is easily explained, as it breaks the narrative progression. Kramer had already punched Mickey, but it is only at this point that he realised it. What is striking is that the past tense is used at the critical moment of the story, when Kramer actually punches Mickey Mantle: turned, popped. This may have to do with Kramer’s embarrassment at his actions here: he wants to make the incident seem like an unfortunate accident and distance himself from the act to a certain degree.

With respect to his acting, a few things are noteworthy. First, when Kramer says I turned around, he does turn his body, but in an unobtrusive manner, and the turn is only about 90 degrees; the point of this seems to be rather to face his interlocutors again (he had his back to them during the previous gesture) than to simulate his past action. As for I popped him, the gesture he makes here is not a violent swing in the air, but he strikes his right fist against the inside of his left hand. I believe this shows a certain inhibition that is absent at throws and pulls, for example. But what is more, the hitting gesture precedes the corresponding utterance. As Kramer says I popped him, he pauses after I, then makes the gesture, and then says popped him. This is markedly different from the rest of the narrative, where gesture and utterance coincide. Perhaps the fact that the event in the simulation (the gesture) is anterior to the utterance of the verb phrase prompts the use of the preterite.

In the final part, Kramer takes on the role of another baseball player:

[Kramer] Then Hank Bauer, you know, he’(i)s screaming ‘Mickey, Mickey, what have you done with Mickey, you killed Mickey!’

[Elaine] So what’(di)d you do?

[Kramer] Well I got the hell out of there.

Recreating speech is a highly iconic form of depiction.²⁹ The simulation is made even more vivid by the fact that Kramer actually changes his voice and makes trembling motions with his hands and head. The use of the present tense is screaming reflects the intensity of the simulation at this point.

Then Elaine asks what finally happened, and Kramer responds: I got the hell out of there. At this point, Kramer sits down. The past tense reflects both the backgrounded discourse status of this narrative resolution (release of tension after the peak) and the end of the simulation.

To summarise the main points of this section:

1. (a) A conversational narrator may set up a simulation space as a stage on which the designated past events are re-enacted. Aspects of this simulation are pointing to imaginary entities, iconic gesture and direct speech representation.

2. (b) The present tense can be used to designate events as they occur in the simulation. The past tense designates the actual past event space and seems appropriate at switches from narrative proper to commentary, pauses in the simulation and in backgrounded sections in the story structure (orientation and resolution).

1.7 The Discourse as Representation in Classical Greek

The final step in our exploration of representation scenarios is to move from the physical to the purely conceptual. In this section, I argue that, whenever we talk about past events, the discourse itself can be construed as a representation of the designated events. As the discourse progresses, we construe a mental model of it that keeps track of the events and entities that are evoked (for the mental discourse model see, e.g., Cornish, [2010], [2011]; Kroon [2017]). In this scenario, distal entities and events are proximal purely in the sense that they are the current object of joint attention. For example, when someone says Then this man walked in, the demonstrative this directs the addressees’ attention to the designated entity as a new referent in the discourse (compare, e.g., Dancygier and Vandelanotte [2016]). This ‘metaphorical’ use of deictic expressions, where spatio-temporal proximity is translated into referential proximity, is the result of a mapping of the distal event space onto the mental space representing the discourse structure (compare Fauconnier [1984: 176–82]).³⁰

As I will argue in Chapter 3, this ‘discourse-as-representation’ scenario is central to the Classical Greek use of the present tense to refer to the past. Here I will only briefly state the argument. To begin, let us consider the beginning of the second book of Thucydides’ Histories. Thucydides has just ended the first book on the note that there was still intercourse between the Athenians and the Peloponnesians without the intervention of heralds (1.146). At the start of the second book, Thucydides marks the beginning of open hostilities between the Athenian and Peloponnesian alliances:

Ἄρχεται δὲ ὁ πόλεμος ἐνθένδε ἤδη Ἀθηναίων καὶ Πελοποννησίων καὶ τῶν ἑκατέροις ξυμμάχων, ἐν ᾧ οὔτε ἐπεμείγνυντο ἔτι ἀκηρυκτεὶ παρ’ ἀλλήλους καταστάντες τε ξυνεχῶς ἐπολέμουν. γέγραπται δὲ ἑξῆς ὡς ἕκαστα ἐγίγνετο κατὰ θέρος καὶ χειμῶνα.

From here then begins the war between the Athenians and the Peloponnesians and the allies of both, in which they did not mingle with each other without the intervention of heralds and, once begun, fought continuously. [The events] have been written down in chronological order by summer and winter.

(Thucydides, Histories 2.1.1)

What interests us here is the use of the present tense ἄρχεται (‘begins’) in combination with the adverb ἐνθένδε (‘from here’). These proximal deictics refer to the time of the ground, which is constituted by the interaction between the narrator and narratees.³¹ ‘From here’ means ‘from this point onwards as we are progressing through the discourse’. The present tense form begins designates not the beginning of the actual war, but that of the war as represented in the discourse. What Thucydides means is essentially that here, the account of the war begins. However, by just saying ‘the war’ and not ‘the account of the war’, Thucydides maps the past event space onto the discourse space, resulting in the pretence that somehow the war itself is immediately accessible in the discourse (compare Loraux [1986]; Edmunds [1993]; Bakker [2006]; see already Classen and Steup [1919] ad loc.). This construal is maintained throughout the narrative, as Thucydides consistently uses the proximal deictic pronoun ὅδε (‘this’) to refer to the war. An example:

καὶ [τὸ] δεύτερον ἔτος ἐτελεύτα τῷ πολέμῳ τῷδε ὃν Θουκυδίδης ξυνέγραψεν.

And the second year ended to this war which Thucydides chronicled.

(Thucydides, Histories 2.70.4)

The war is ‘right here’ (note τῷδε [‘this’]) in the sense that it is the focus of attention throughout the entire discourse. The act of narrating somehow reconstitutes the designated events but in a highly schematic form; witnessing the actual war is a wholly different experience from reading the account of the war.

I argue that the present tense can be used to highlight the present occurrence of the designated past events in the medium of the discourse.³² I reproduce example (4) from Section 1.3:

οἱ δὲ Ἐπιδάμνιοι οὐδὲν αὐτῶν ὑπήκουσαν, ἀλλὰ στρατεύουσιν ἐπ’ αὐτοὺς οἱ Κερκυραῖοι τεσσαράκοντα ναυσὶ μετὰ τῶν φυγάδων ὡς κατάξοντες, καὶ τοὺς Ἰλλυριοὺς προσλαβόντες. προσκαθεζόμενοι δὲ τὴν πόλιν προεῖπον Ἐπιδαμνίων τε τὸν βουλόμενον καὶ τοὺς ξένους ἀπαθεῖς ἀπιέναι· εἰ δὲ μή, ὡς πολεμίοις χρήσεσθαι.

The Epidamnians didn’t listen to them at all, so the Cercyreans make an expedition against them with forty ships, together with the exiles, so as to restore them, and taking the Illyrians in addition. Besieging the city, they proclaimed that whoever of the Epidamnians wished to do so, as well as the foreigners, might leave unharmed; if not, they would treat them as enemies.

(Thucydides, Histories 1.26.4–5)

If the war is present in the discourse, then the events that constitute the war, such as military expeditions, may be construed in the same way. The present tense στρατεύουσιν (‘make an expedition’) highlights the present occurrence of the designated event in the discourse, thus focusing the joint attention of the narrator and addressees on this new development.³³

This naturally raises the question why the present is used here but not with the other two main clause verbs in this passage. In fact, as I pointed out in Section 1.3.2, this present form stands quite isolated among preterites in the wider context. My claim is that the present is used at points where the discourse structure becomes cognitively salient – in particular, at changes in the narrative dynamic (or turns, e.g., Fludernik [1991]) and at the introduction of new referents (e.g., Thoma [2011]). That authors such as Thucydides use the present to highlight the importance of the designated events to the story structure or to their own rhetorical concerns is a widely held view in Classical scholarship (Sicking and Stork [1997]; Rijksbaron [2002]; Allan [2009]; Lallot et al. [2011]; Nijk [2013a]). I will elaborate this argument in detail in Chapter 3. For now, I point out that my account in terms of a discourse-as-representation scenario explains how the pragmatic functions often associated with the present tense can be derived from its semantics. The present tense construes the designated past event as occurring at the time of the ground; this can be made sense of by presupposing a conceptual scenario in which the present tense highlights a virtual occurrence in a representation space; this representation space is identified as the discourse; and this implies that the author wishes to underscore the impact of the designated past event on the structure and evolution of that discourse.

Finally, the discourse-as-representation scenario is not confined to narrative discourse.³⁴ The following example is taken from the prologue to Sophocles’ play King Oedipus. Creon has just returned from his mission to ask the oracle at Delphi how Thebes might be rid of the plague that is destroying the city. Creon tells Oedipus that the oracle told him they need to drive out the murderer of Laius, the previous king. Then Oedipus asks a series of questions, trying to find out who was the perpetrator. We are concerned with the following lines:

(29)

πότερα δ’ ἐν οἴκοις, ἢ’ν ἀγροῖς ὁ Λάιος,

ἢ γῆς ἐπ’ ἄλλης τῷδε συμπίπτει φόνῳ;

Does Laius meet with this murder

in his home, or in the fields, or in another country?

(Sophocles, King Oedipus 112–13)

Laius’ death is construed as immediately proximal to the speaker by means of the demonstrative pronoun in τῷδε … φόνῳ (‘this murder’) and by the present tense συμπίπτει (‘meets with’). The motivation for this seems to be that the matter is the subject of close enquiry and essential to the speaker’s immediate concerns: the entire fate of the city hinges on Oedipus’ ability to solve the mystery of the murder of Laius (compare Jebb [1914] ad loc.; Nijk [2016a]).³⁵ This highlighting is especially poignant here, as the location of the murder will turn out to be tragically revealing: when Oedipus learns where it happened, he soon realises that he himself is the killer (729–45).

To summarise the main points in this section:

1. (a) Whenever we talk about past events, the discourse itself may be construed as a representation through which these past events become immediately accessible.

2. (b) In this scenario, the present tense highlights the designated event as the current object of joint attention. This may be because the designated event impacts the structure and evolution of the story, or because it is central to the present concerns of the speaker.

1.8 Tense-Switching Patterns and the Classical Greek Present

In the previous sections, I have distinguished four representation scenarios that may support the construal of past events as presently accessible. This allows us to understand how the use of the present tense to refer to the past is possible in the first place. The second main issue concerns the alternation between the tenses in these contexts. What motivates an author or speaker to designate either the representation space or the past event space? I have already touched upon this issue in the previous sections, but here I will draw the strands together. I will also prepare for the next chapters by pointing to the peculiar character of tense-switching in Classical Greek.

As a general principle, I argue that tense-switching patterns depend on the nature of the underlying representation. In literary fiction (Section 1.4), tense seems generally consistent throughout large stretches of discourse. For example, in chapter 5 of Tom Brown’s schooldays, the narrative tense is the preterite, until a switch occurs (Then the præpostor who stands by the master calls out the names), and the present tense is used until the end. The part narrated in the present tense describes the rugby match, and it is not hard to see why the narrator would wish to engage his narratees more strongly with this climactic part of the chapter than with the preceding developments. The consistency in tense usage may be due to the artificial nature of the narrative set-up in such novels. The kind of spontaneous tense-switching associated with natural speech may appear unsophisticated in the context of a work of literature.

In Dutch news media (Section 1.5), we saw that present tense usage was quite consistent in the material I discussed. I argued that the underlying representation in this case is a video showing security camera footage or a dramatic reconstruction. A video is a highly iconic representation that constantly imposes itself upon the senses. The virtual presence of the past events is thus cognitively salient throughout, and this results in consistent present tense narration.

In the simulation scenario (Section 1.6), on the other hand, we saw some instances of tense-switching, even on the narrative main line. The representation underlying the use of the present tense here is more mutable than a video. A simulation exhibits varying degrees of iconicity, depending on how the narrator acts out the designated events. Moreover, the representation is only accessible when the narrator keeps it running; it may be ‘paused’ at any time. I argue that this makes tense-switching more natural in this scenario than in the scenarios previously discussed.

Finally, in the discourse-as-representation scenario (Section 1.7), we find that the present tense may be used in relative isolation among past tense forms. Here, the representation is merely an abstract conceptualisation. The construal of the discourse as a representation is not consistently foregrounded in the way that physical representations are. As I argue, the present will be used to highlight this construal mainly at points where the discourse structure becomes cognitively salient.

The overall point is that the meaning of tense-switching differs per genre. The narrative present tense has been associated with particular narrative personae: when using the present tense, the narrator assumes the role of ‘performer’ (Fleischman [1990: 61–3]) or ‘eyewitness’ (Allan [2013] on Thucydides). In my account, different representation scenarios involve different types of narrative persona: there is the director (nineteenth-century fiction), the video commentator (Dutch news media), the actor/performer (performed stories) and the discourse manager (Thucydides).

In the next chapters, we will be focusing on the Classical Greek data. The key feature of tense-switching in this language is that the past-referring present seems completely interchangeable with the preterite (compare Robar [2016: 311, with references]). This may seem to be in contradiction with the assumption that the present tense has a specific value of its own that distinguishes it from the preterite. What I mean by calling the tenses interchangeable is that the present tense can be replaced with the past tense without impacting either the truth conditions or the felicitousness of the utterance.

This is due to two main reasons. First, in Classical Greek, the present tense never designates the past events as seen from a displaced ground. My justification for this claim is the fact that the present is never used to refer to the past in combination with the adverb ‘now’ (νῦν), let alone ‘today’.³⁶ In the displacement scenario, replacing a present tense form with a past tense form can yield a strange effect. Consider the following rewrite of example (3):

University president Sibrand Poppema can’t get the smile off his face today. It was May 15, 2014, and his beloved RUG [University of Groningen] is four hundred years old.

I will not say the switch from the present tense (can’t get) to the past tense (was) would be impossible, but it does seem rather marked – especially if we switch back to the present tense in the next sentence (is). This markedness lies in the necessity to accommodate our viewpoint arrangement back to the normal mode, right after the displacement scenario has been set up in the previous sentence. This is never an issue in Classical Greek.

Second, the present is never used in Classical Greek with the kind of consistency we see in certain novels or news media. When an entire narrative section is told in the present tense, an isolated past tense form may seem odd to the addressee. On the other hand, when the present tense is found in relative isolation among past tense forms, as we observed in Section 1.7, then changing that present tense to the preterite makes for an entirely natural result. Clustering of present tense forms does occur in Classical Greek, but rarely longer than four main clause verb forms in a row, and even in the most vivid passages, present forms will be interspersed with preterite forms. Therefore, it does not seem that a change of any individual present form to the preterite would be registered by the addressee as being infelicitous.

This being the case, I will from now on use the term ‘present for preterite’ to designate the use of the present to refer to the past in Classical Greek. This allows us to avoid the contentious term ‘historical present’, which is problematic for two main reasons: (a) its liberal use obfuscates the essential difference between the two conceptual scenarios distinguished here; (b) it is usually associated exclusively with narrative discourse, so that deviant instances of the present to refer to the past (as in (29), for example) are either neglected or explained away (compare Nijk [2016a]).

1.9 Conclusion

Let me recapitulate the main arguments made in this chapter and then look ahead to the following chapters.

1. (a) Semantics of the present tense. I have argued that we should retain the assumption that the present tense denotes present time reference, even in the face of usages that seem to contradict this rule. Such usages should be explained by assuming a special conceptual scenario whereby the designated distal events are construed as occurring at the time of the ground.

2. (b) Displacement versus representation. There are two ways in which the gap between the ground and a distal event space can be bridged conceptually. The first involves a displacement of the ground to the distal space (‘virtual observer of actual entities’). The second involves bringing the distal entities into the ground in the form of a representation (‘actual observer of virtual entities’).

   Both scenarios facilitate the use of the present tense to describe past events. There is, however, a fundamental (even if not entirely absolute) difference between the two: the representation scenario allows for the designated events to be edited (by compression and abstraction) in a way that the displacement scenario does not.

3. (c) Different representation scenarios. The ‘switched’ use of the present tense may be facilitated by different kinds of underlying representations, depending on the language and genre. As we have seen, the representation may be conceived in terms of a (virtual) play, a video, a simulation or performance and the discourse. These scenarios involve different tense-switching patterns and different ‘narrative personae’ associated with the present tense.

4. (d) Tense-switching in Classical Greek. The ‘switched’ present in Classical Greek can, as far as I can judge, always be replaced with the past tense without changing the truth conditions or felicitousness of the utterance. I therefore label this use the ‘present for preterite’.

In the next chapters, I discuss the uses of the present for preterite in Classical Greek in terms of three distinct representation scenarios. In each case, I give a theoretical description of the representation in question and its associated pragmatic implications and corroborate this description with analyses of corpus data. My methods of analysis here include discussions of salient individual examples, quantitative analyses and extended case studies.

The following is a brief overview of the argument of each chapter:

Chapter 2, ‘Scenic Narrative and the Mimetic Present’. Here I further explore the idea of a simulation as representation, which I introduced in Section 1.6. I argue that the use of the present for preterite in scenic narrative depends on the pretence that the designated past events are presently being re-enacted in a simulation. This pretence of re-enactment consists in an analogous relationship between the narrative experience and actual experience – in other words, in narrative mimesis. This involves mental simulation, physical depiction (through gesture, etc.) and iconic use of grammar.

Chapter 3, ‘Summary Narrative and the Diegetic Present’. Here I develop the argument outlined in Section 1.7, that is, that the discourse can be conceived as a representation through which past events are made presently accessible. I argue that this scenario becomes more salient as we move from scenic to summary narrative, which involves compression of time and a greater degree of abstraction in the way the events are presented. In the discourse-as-representation scenario the present for preterite draws attention to the impact of the designated event on the structure of the discourse. This occurs in particular at changes in the narrative dynamic or at the introduction of new referents.

Chapter 4, ‘Zero-Degree Narrativity and the Registering Present’. In this final chapter, I focus on the use of the present for preterite in non-narrative contexts. I argue that the present tense here serves to construe the designated event as being somehow ‘on record’. This ‘record’ may be a concrete entity, such as an iconographical representation or a document recording a transaction, but also a purely mental representation, such as a culturally established ‘canon’ of mythological or historical events. The pragmatic function of this ‘registering’ present is to elevate the status of the designated event (as being ‘canonical’ or ‘official’) and to underline the legitimacy of the speaker’s assertion.