5.1 A present of present things

Now is a subjective experience or an intersubjectively and culturally agreed-upon conceptual category – ‘the present’. The past and the future are subjectively understood as relative to now. As already discussed in Chapter 4, in the DST framework, S experiences ‘the present’ either as a point or as a peripersonal extension and normally experiences it as ‘real’, more real than the past and the future, which may be felt as less real than the past. A more abstract way of putting this is to say that now is experienced as maximally certain on a scale of epistemic modality. The relativity of the conceptual constructs of ‘the past’ and ‘the future’ to the now, where S is situated, is what makes reference to these ‘times’ deictic. This section is concerned with the effects produced by the use of the simple present tense and present progressive (ing form) in combination with various meanings expressed by particular verbs. As Chapter 4 has already explored two major types of present-time conceptualisation (instancing and presencing), this section is confined to summarising some extensions of these two types found in certain particular uses of the simple present and the progressive ing forms.

The English simple present does not always refer to the present time of the speaker, the speaker's now, though it generally includes it. In fact using the simple present to refer to a present instant is a rather specialised usage. It is generally recognised as conceptually connected with the following: states that include but are not restricted to now, generic concepts, habituals and narrative instancing (e.g. the ‘instant’ narration of audio commentators).

Frege thought of the simple present tense in terms of timelessness and time-setting:

By ‘giving a date’ Frege can be understood as meaning that an utterance of a sentence implicitly locates the time at which the uttered content of the sentence is asserted to be true. There are other, more language-oriented ways we can understand this in the DST framework. The present tense can be used to refer to a past event indicated in the discourse, perhaps with a date expressed in an adverbial adjunct and we below shall look at such cases in terms of the transformation of the base reference frame R.

Frege's idea of timelessness is important here because it characterises the abstract generalising use of the simple present, and is related to the logical notion of universally quantified propositions, which are timelessly true (or false). The important point is that treating the present tense as expressing temporal presentness (as opposed to temporal distance) is integral with epistemic certainty. And epistemic certainty inheres both in perceptual (therefore temporal) immediacy and in the conceptual transcendence of temporal boundedness. This characterisation emerges naturally from DST's geometric approach; it also seems phenomenologically natural. This way of understanding the meaning of what it is that the simple present tense corresponds to is similar to revised ideas about the English present tense found in cognitive grammar (Langacker Reference Langacker1991, Reference Langacker and Davis1995, Reference Langacker2001, Reference Langacker, Patard and Brisard2011, Brisard Reference Brisard and Brisard2002: 262ff).

We have already seen in Chapter 4 that although simple present (instancing operation) is the default for verbs denoting states and progressive (presencing operation) is the default for verbs denoting dynamic processes when S refers to them in S's present now, these default operations can be reversed. Applying simple present to dynamic process verb meanings produces the conceptual effect of atemporal generality, which can be understood in the DST geometric framework as a point on the t-axis.¹

The following sentences all use the simple present. Only (1b) involves a stative verb that, for the reasons described in Chapter 4, occurs by default with the simple present. In example (1d) the verb stand is semantically open to alternate construals by either the instancing or the presencing (is standing) operator. The point of grouping these examples is to consider further the conceptual similarities emerging from the application of instancing in all cases.

The simple present in example (1a) is not understood as a temporal instant or a temporal zone around the self. It seems rather to be understood as Frege's timelessness present. In DST we have proposed capturing this geometrically by ‘compressing’ the vector for revolves to t = 0 and simultaneously locating it at m = 0 – thus defining a point without duration that is also maximally certain for S. This is an abstract certainty that is different from the experiential certainty of being ‘close’ to some represented dynamic process.

This kind of characterisation of the conceptual implications of the simple present might seem inappropriate to (1b). The dissimilarities disappear if one thinks of both sentences as predicating permanent properties. In understanding (1b) it seems likely that there is a tendency to assume that Henry does not stop knowing the answer, and the time at which he started knowing it is conceptually not represented at all. One of Henry's properties is that he knows the answer. The DSM for (1a) and (1b) are the same on this account (see Figure 5.1).

Figure 5.1 Timeless simple present for (1a) and (1b)

This kind of timeless epistemic certainty concerning essential properties is not limited to culturally accepted scientific certainties but is extended to personal experiences, opinions, philosophical claims, etc.: broccoli is disgusting, Syd is a liar, politics is the pursuit of war by other means, and the like. The DST format relativises all such cases to S's cognitive–deictic space; they are not modelled as objective (for the modeller) truths. The key conceptual feature is timelessness, which distinguishes this kind of temporal–epistemic certainty from that expressed in e.g. the decorator is painting the door, Henry is sailing to Calais, where the progressive tense form expresses epistemic certainty that is time-bound and based on perception or perceptual evidence. Such statements as those in (1) are timeless, that is, temporally universal facts that hold true at the present moment, even if S does not perceive them. Only the simple present expresses this: Syd was a liar is not taken as a fact true as of speaking, and is thus not temporally universal.

As proposed in Chapter 4, DST models the timeless validity meaning of the present tense as in Figure 5.1.

In sentences (1b) and (1c) the simple present tense form produces different conceptualisations. In the case of (1b) the schema for the stative Aktionsart combines with the instancing operator (expressed by the simple present form) to produce a timeless vector relating Henry and the answer. Sentence (1c) combines the instancing operator (again via the simple present tense form) with a process schema, producing, again, a timeless vector. Since processes cannot be conceptualised as timeless or as instants, they are conceptually processed as timeless properties.

The concept ‘habitual’, which comes to mind in understanding (1c), can be seen to be closely related to the concept ‘timelessly valid’ which comes to mind in processing examples like (1a). Both are so to speak removed from particular time reference. Even in (1c), Henry's walking is conceptually rendered timeless by the instancing operator (simple present). The difference between a human individual's walking and the Earth's orbiting is settled by way of additional conceptual processing drawing on the background knowledge triggered by the semantic frame of the two respective verbs and of the nouns. Now this explanation requires further detail that cannot be undertaken here. But it should be noted that not only do the Aktionsart schemas and semantic frames for revolve, walk, Earth, and Sun come into play but also, in the case of (1c), frames concerning social knowledge and linked to expressions such as walk to work. This particular frame will contain knowledge about prototypical work routines in a particular human society and it is some such frame that makes (1c) ‘habitual’ in the sense that it evokes a series of repeated actions based on a daily cycle. It is possible that this is the case for all English expressions using the simple present that are understood as ‘habituals’. This is perhaps the reason why we do not think of (1d) as being a habitual, although there are plainly some similarities with the (1c), as well as more abstractly with (1a.): statues are not framed as having habits. Such additional semantic material is not part of DST, which is set up to model only the abstract deictic framework (here especially the temporal and epistemic dimensions of it) that is common to the meanings of the sentences in (1).

Both sentences (1a) and (1c) in this sense express timeless properties. Habituals therefore can be regarded not as a separate class of event types, but as a result of the cognitive instancing operator that ‘squeezes’ temporally extended process schemas into state-like, temporally unextended vectors at t = 0. The position at t = 0 is a truth condition of the sentence for S at t = 0 in a particular reference frame modelled by a particular DSM – the zero t in such diagrams is S's current reference frame relative to the universal space-time reference frame.

The most general way of stating the proposal here is that a proposition predicating a habitual action or state of some argument represents a conceptualisation according to which that predication is a temporally unbounded property of that argument.

In order to account for the use of the simple present to denote the habitual meaning and the general validity meaning, Langacker (Reference Langacker1991, Reference Langacker2001) and Brisard (Reference Brisard and Brisard2002) are obliged to postulate an extra mechanism. Langacker (Reference Langacker1991: 264–5) adopts a distinction between ‘structural’ and ‘phenomenal’ knowledge taken from Goldsmith and Woisetschlaeger (Reference Goldsmith and Woisetschlaeger1982) and uses the idea of ‘structural’ world knowledge to account for both habitual and generic sentences. He then proposes a schema to describe an ad hoc ‘rule’ (really an iconic diagram) by which the simple present turns a perfective (processual) verb into a ‘derived’ configuration in which what is expressed by the particular processual verb is now incorporated in ‘structural’ world knowledge, in effect rendering it imperfective. I want to suggest that, while compatible with Langacker's insights, the DST approach follows automatically and more simply from its geometrical principles. It does imply, however, that the habitual meaning emerges from pragmatic processing: (1b) and (1c) have the same schematic conceptual form, but (1c) involves a conceptual inconsistency between the process Aktionsart schema and the instancing operator (simple present tense) that is resolved by introducing the concept of habit.

The cases of simple present I have considered so far are somewhat different from cases where an event that is to some sufficient degree temporally concurrent with the speaker's speech time – as in sports commentary:

Such uses are not restricted to this particular genre but would also be natural in the following:

Cases (2) and (3) are not essentially different from conversational narrative uses of the simple present:

The example in (4) is, however, relativised to S's past by the tense of the first sentence. We shall look at a way of formalising this phenomenon in the next section.

Langacker's (Reference Langacker1991, Reference Langacker2001) account of such uses of the simple present is appealingly straightforward in proposing that the use of the simple present in these situations is motivated by the approximate simultaneity of speaking time with the duration of the event being observed by the speaker. The DST account is the following. In an example like (2) we have process verbs that are not expressed in the progressive tense form (presencing operator), that is, they are not presented as portions of actions observed, as it were, close up. Given their event-type schema the expected default would be the progressive (Mary is bowling, Alice is racing). The use of the instancing operator (simple present) compresses the durational quality into a vector at unextended t = 0, as described in Chapter 4. This may be because bowling, racing to a position in cricket, making an incision, etc., though durational, can possibly be conceived as point-like relative to other types of actions. Be that as it may, application of the instancing operator (simple present) has to be understood here not in the sense of producing a single timeless vector that is then understood conceptually as an instance of a temporally unbounded set (as for examples (1a) to (1d)), but rather in the sense of producing an experiential instant. Such a temporal instant is still positioned on the t-axis at 0, but understood as the limit point of experienced peripersonal time.

In this way it is possible to construct a cognitively plausible account of uses of the simple present with process type verb schemas, one that is to some degree at least consistent with the geometrical account. But the main question is how to account for uses of the simple present that do not concern the present, that is, S's present, but to past or future times relative to S in S's reference frame R.

5.2 A present of past things

What is often called the historical present occurs widely in conversational narrative and literary narrative. In Langacker's formulation the historical present is a ‘radical mental transfer pertaining to the deictic centre’ in which ‘the speaker decouples the deictic centre from the here-and-now of the actual speech event and shifts it to another location’ (Langacker Reference Langacker1991: 267). We can also use such a shift to account for certain uses of the present tense to refer to future times relative to the speaker (the train arrives in five minutes). In writing of ‘displacement’ or ‘transposition’ of the deictic centre, Bühler (Reference Bühler and Goodwin1990[1934]) points us toward a geometric approach to describing uses of the present tense to refer to past events. In the DST framework, translation of reference frames on the t-axis seems like a natural solution and will be explored in this section. In both these cases, the relevant transformation is a translation of axes not of course a reflection. A second set of axes is copied which places a new 0′ (the deictic centre) at some point t_i > t₀. The way this works is outlined in Figure 5.2 below.

Figure 5.2 Transforming of reference frame for historical present: (5b) In June 1520 Henry sails to Calais

Sentence (5a) below is feasible when there is a conversational context located in a past jointly attended to by S and an interlocutor. Sentence (5b) shows that the date-setting can be fairly precise and (5c) is linked to a photograph as well as time located relative to a reader (who may be reading it at a pragmatically appropriate (not too soon after) the date mentioned in the adjunct.

Notice that such sentences can be temporally ambiguous. For example if (5b) is uttered by a twenty-first-century historian, it is present tense referring to the past relative to the historian and hearers. Equally, however, it could be uttered by, say, a privy councillor in the year 1518 planning a meeting with the French king to take place in 1520. The different possibilities are simply modelled by means of embedded reference frames.

Let us consider the case of the twenty-first-century historian's utterance. The speaker induces us to set up a second set of coordinates R′, a temporal reality space in which we take the position of S′. Figure 5.2 translates a copy of R to a point on t representing the publicly agreed-upon date June 1520. In this reference frame the event of Henry sailing is expressed in the simple present tense form, following an application of the instancing operator to the process event type schema, as outlined in Chapter 4. It is positioned in the same way and with the same conceptual effects, at t = 0 in the deictic space, i.e. the reference frame with the usual DST axes, as described in the preceding Section 5.2.1. The ‘instancised’ process verb sail is at t = 0 in R′, which is a copy of the base frame R, so carries the subjective deictic origo, now marked S′. In other words, S is ‘seeing’ an event taking place concurrently with his or her now, but this now is displaced into the past relative to the base frame and to S.

The discourse entities are labelled on the d-axis in the base space R: their coordinates run through into the virtual space R′, so we have trans-frame identity. The fact that the ‘ends’ of the time axes t and t′ do not coincide is of no consequence since the t-axes are potentially infinite. The essential point is that the vector for sail is simple present (the process sail has had the instancing operator applied to it) with respect to R′ and simultaneously past relative to S in R. This is exactly what is required in order to formally model a present tense form that refers to a past event.

In this example the instancing operator has applied to the process verb sail (as in examples (1) above), resulting in a particular construal of the event. We can think of the configuration similar to that of the action-by-action commentator or observers (cf. (2), (3) and (4)), with the process event compressed to a conceptual instant of time in which the event is completed. A connected motivation may be that process events in the past are viewed as complete when presented in the usual way by means of the simple past: Henry sailed. In fact, the temporal translation of R is only part of the construal potential. Once R has been shifted the event of sailing can be represented either by the using the cognitive operator provided by simple present or by using the cognitive operator provided by the present progressive as in (6):

In such a representation R′ is positioned on the t-axis as in Figure 5.2, but is also translated on the d-axis, so that it is closer to the coordinate for the sail vector – in accordance with our characterisation of the presencing operator in Chapter 4. The resulting DSM would in fact look like Figure 4.11 in Chapter 4, repeated above as Figure 5.3, with labelling appropriate to (6).

Figure 5.3 Present progressive in the past: (6) In 1520 Henry is sailing to Calais

5.3 A present of future things

The relationship between the English present-tense forms and reference to future time is particularly complex. It is often pointed out that reference to the future has a strong modal element and this is assumed to be connected to the fact that the future is inherently unknown, though it is mentally representable because of the cognitive phenomena of intending, planning and expecting.²

Jaszczolt (Reference Jaszczolt2009: 38–45, 50–5, 140–1) argues strongly for a modal view of the expression of deictic future times and proposes a scale of future time markers corresponding to a scale expressing ‘a certain of detachment from the certainty of now’. Following Jaszczolt's scheme, but excluding be to, be about to and be on the point of, we might arrange the following sentences in the following order:

According to Jaszczolt these are graded from highest certainty (least detachment) to least certainty (highest detachment). While this scale seems intuitively satisfactory, it is not completely clear how these particular tense forms encode graded modal meanings. Jaszczolt outlines in some detail how Default Semantics, including its revised version, explains meaning as an emergent phenomenon arising from societal, cultural and other pragmatic information (Jaszczolt Reference Jaszczolt2009: 132). But the tense form must also have core meaning and what is not clear is whether and how these core meanings per se encode modal gradations. I shall leave aside the question of whether the semantic structure of the different tense forms have inherent modal meaning that enables us to place them on a modal scale, and instead focus on the temporal axis as set up in DST. The question we are concerned here is how DST can model the use of tense forms that appear to refer to the speaker's present in order to refer to the speaker's future. We shall consider the following examples:

These examples suggest that present tense forms cannot refer to a subjectively future time without further cognitive input. This means some explicit lexical indicator such as this Thursday, next year, in a few days' time is required. Note that a vague indication such as in the future is not enough:

It seems that some more precise time-indicating expression is required. This is not necessarily a lexical time expression in the clause, however. It may be a mutually shared piece of knowledge in the conversational context.³ These considerations will lead us to a detailed model of present-for-future expressions that is somewhat more detailed than the conventional accounts and which is primarily temporal rather than modal. I turn now to simple present and progressive present forms used to conceptualise times that are in the future relative to the speaker and hearer.

5.3.1 Simple present and the future

In order to account for the ‘scheduling’ use of the simple present in (1), Langacker (Reference Langacker2001) makes use of a notion of ‘virtual entities’, understood by him in terms of a ‘non-canonical viewing arrangement’. Brisard deals with the problem primarily by emphasising a modal meaning of the simple present that he calls ‘immediate givenness’ (Brisard Reference Brisard and Brisard2002: 263–8). Brisard retains Langacker's idea that the simple present has a double meaning (immediate phenomenal experience and general structural knowledge of the world), but offers a different description, in which the processual verb is ‘projected’ (by use of the simple present) onto a ‘virtual plane’ equated with representation of ‘structural aspects of the world’ (Brisard Reference Brisard and Brisard2002: 274). The DST approach, however, provides a way of dealing with such a matter within the overall theory of frames of reference. Lonergan proposes three kinds of reference frames: the personal, the public and the special. The latter are mathematical and physical and need not concern us here. The first corresponds with personal cognition of three-dimensional physical space:

…everyone has his personal reference frame. It moves when he moves, turns when he turns, and keeps its ‘now’ synchronized with his psychological present …

(Lonergan Reference Lonergan1957: 144)

It is the case that DST does not work with the physical space but with a personal conceptual space of a kind that specifically underlies language. To that extent DST seeks to advance beyond Lonergan's idea of reference frames. Nonetheless, Lonergan's general point and the link he makes with temporal as well as spatial deictic expressions is consistent with DST's deictic space. What is of special concern in the present context is his second class of reference frame, ‘public reference frames’. These are both spatial and temporal. In the following Lonergan is primarily describing temporal reference frames:

[people] are familiar with alternations of night and day, with the succession of weeks and months, with the uses of clocks and calendars. Now such relational schemes knit together extensions and durations. But they are not personal reference frames that shift about with an individual's movements. On the contrary, they are public, common to many individuals, and employed to represent the here and now of the personal reference frame into generally intelligible locations and dates…

(Lonergan Reference Lonergan1957: 144)

Now cognitive linguistics, coming from a different angle, has a complementary point to make concerning the nature of this publicly shared knowledge of times (Lakoff Reference Lakoff1987: 68–9), drawing on Fillmore's work on cognitive frames (e.g. Fillmore Reference Fillmore1985). Lakoff describes what English calls a ‘week’ as an ‘idealized cognitive model’ – we could equally call it an idealised public reference frame – that contrasts with for example Geertz's description of the Balinese calendar (Geertz Reference Geertz1973: 392–3 cited by Lakoff Reference Lakoff1987: 68). Names of days, such as ‘Tuesday’, can only be defined relative to a reference frame, and this is clearly a conceptual construct:

Tuesday can be defined only relative to an idealized model that includes the natural cycle defined by the movement of the sun, the standard means of characterizing the end of one day and the beginning of the next, and a larger seven-day calendric cycle – the week. In the idealized model, the week is a whole with seven parts organized in a linear sequence …

(Lakoff Reference Lakoff1987: 67)

Such frames (or ‘models’) do not exist in nature but are conceptualisations that are agreed upon by long processes of cultural coordination.

Some formal semantic accounts of the scheduling present take it that such cases arise because they have the ‘connotation’ that they are predetermined with respect to the present (Kaufmann et al. Reference Yu2005: 90). I suggest that this ‘connotation’ arises because of the insertion of a cognitive frame such as the week frame, where the days of the week and the cyclic recurrence of weeks are in a sense ‘predetermined’.

I am proposing that the notion of idealised frames can be usefully thought of in terms of reference frames: Fillmore's frames and Lakoff's models, at least those that concern time, can usefully be understood as reference frames that enter into relations with personal reference frames of the kind that DST models in its particular geometry. The crucial point is that now is a personal cognitive experience not a socially agreed-upon time in a public frame of reference: the personal and the public frames of reference have to be ‘coordinated’ – their separate coordinate systems have to be brought into some sort of alignment, the one relatively to the other. We shall make use of this idea in looking at the meanings of English present-tense morphology.

In DST we already have the apparatus to model time-related frames. The first step in considering how to use them to model the ‘scheduling’ use of the simple present tense form is to note that its use suggests a secondary set of axes R′ whose origin is located at some time t in the future relative to S. Further, we may say that the origin of the scheduling axes R′ is not at S′, but some other referent such as a timetable or some pragmatically given ‘public reference frame’ (PRF), which might include a shared mental representation of a written or verbal agreement, or shared cognitive frame representing a time cycle such as the calendar or the seasons, in the encyclopaedic memory of the interlocutors.

In using the term ‘schedule’ here, one should not be misled into importing contemporary associations, since modern scheduling is merely a kind of cooperative planning for the future that is characteristic of all human societies from early beginnings that may very well be closely tied to the emergence of human language. More needs to be said about how DST can geometrically model PRFs but this cannot be done here. We focus on the relations between frames of reference, depicted in Figure 5.4. In this and subsequent figures, the peripersonal temporal region is shown as a shaded plane.

Figure 5.4 Example (7a) Henry visits Calais this Thursday

PRF is a public frame of reference. Because it is public there is no S defining a now, so there is also no m-axis representing the modal component of S's conceptualisation, and no d-axis on which S ‘locates’ discourse referents at subjectively relative ‘distances’. It is simply an ordered set of seven arbitrarily named time periods, each corresponding to the diurnal cycle. These intervals lie on a finite directed time-line that defines a one-dimensional plane. Though inherently it is not deictically centred, it is directed in the direction of ontological time. In common with other individuals in the culture, S holds PRF in long-term memory but also aligns it, in virtue of socially shared knowledge about ‘which day it is’, with his subjective conceptualisation of peripersonal now.

Consider Figure 5.4. First the base axes R are aligned with the PRF so that S's now is aligned with Tuesday in the week frame. Next we set up R′, a copy of S's base R, a virtual frame of reference in which S's now is aligned with Thursday in the PRF. It is in R′ that Henry's sailing to Calais is located, a reference frame located at a specific future that has both subjective and public reality for S, as the diagram shows. Now this future event is expressed by the simple present, and the implications of this tense as distinct from the progressive have to be explained.

The vector visits is located at 0 in PRF. Either this means the event of leaving on Thursday is a timeless property, or it means that it is viewed as a timeless instant – the relevance here is that the event is viewed instantaneously as a whole, in contrast with the presencing conceptualisation cued by the progressive form, which excludes beginning and end. This is an aspectual distinction but formulated in the terms just used we can see why it is appropriate to express the type of meaning under consideration.

5.3.2 Present progressive and the future

Neither simple present nor present progressive can refer to the future without extra lexical or pragmatic specification and presumably therefore additional cognitive structure. Compare the following:

Henry is visiting Calais (now)

Henry is visiting Calais this Thursday.

In (10) the default understanding is that Henry is visiting Calais now, at the time of speaking. In (8a) the speaker sets up a mental space R′. But this future is located in a PRF, as shown in Figure 5.5. The deictic space model for the conceptualisation induced by this use of the progressive combined with lexical (or pragmatic) time indicator proceeds as for the case of the simple present (Figure 5.4 above).

Figure 5.5 Frame shift for (8a) Henry is visiting Calais this Thursday

In this new space, represented by the bold dashed axes in Figure 5.5, S (and the hearer) have a present now located at a deictically future time, indicated by a temporal deictic this Thursday, defined relative both to the PRF and to base R. Future now and present this Thursday are conceptually co-located. Within the embedded frame, i.e. relative to it, we have the presencing operation due to the progressive form is visiting. The ‘up close right now’ element is represented by the conceptual presencing operation within that frame. Sentence (8a) seems to bring us ‘close’ to the action, ‘puts us in the picture’. It does not seem necessary to consider this effect ‘modal’. Moreover, it is often noted that the intuited sense of such examples is that the future event is somehow viewed as rooted in the speaker's present. According to the present model, this is accounted for within the shifted reality frame R′.

How should we characterise the difference between the simple present and the progressive present referring to a future time? The elements of a description are made explicit in the notions of coordinate geometry, vectors and shifts of reference frame. The temporal component is the same for both tense forms (see Figures 5.4 and 5.5). The difference is the structure of the instancing operator (simple present) and the presencing operator (progressive present). In the former the whole event is viewed as a completed instant, but since the simple present also serves to conceptualise timeless generic properties, it carries a potential implication of high subjective certitude. This is not the case for the progressive form (presencing operator), which has the slightly contradictory effect of close-up viewing and incompleteness (as the ‘ends’ of the action are not ‘in view’). This effect may or may not, it seems to me, result in a sense of greater or lesser certitude with respect to the simple present. It is thus not clear to me that these forms should be viewed as primordially ‘modal’, though this may arise by implication under the influence of various contextual factors.

5.3.3 Going to the future

We are concerned here with the ‘periphrastic future’ tense forms be going to/gonna, as in (5)

Henry is going to/gonna visit Calais this Thursday.

Such forms are discussed in terms of grammaticalisation by Hopper and Traugott (Reference Hopper and Traugott2003). In the case of gonna, which is only possible with a following verb (not a NP), the fact that going assimilates to indicates that we are dealing with a coalesced concept of directed motion, i.e. we are not dealing with an infinitive form to V, e.g. to visit. In terms of cognitive metaphor theory, the underlying metaphor as is well known is Purpose Is Directed Movement (cf. Lakoff and Johnson Reference Lakoff and Johnson1980). Hopper and Traugott's (Reference Hopper and Traugott2003) explanation of the grammaticalisation of going to as an auxiliary expressing future time rests on a pragmatic inference of futurity from purposive action as expressed in the lexical verb go. However, this fails to generalise over the symmetrical pair found in a language like French: venir de V (literally ‘come from V’, i.e. to have just done V) and aller V (‘go V’). Purpose cannot be inferred from ‘come from’. While Hopper and Traugott's account is not necessarily wrong, the French example suggests that this kind of temporal auxiliary results from a direct conceptual transfer from space to time, preserving deictic relations. The DST account makes this assumption and the coordinate and vector components of that theory are well placed to model what is at issue. Spatial translation to/from a location is directly mapped onto temporal translation to/from a time point.

As suggested by the grammatical construction in question, the model in Figure 5.6 shows two vectors, a component for ‘be going’ and one for ‘visit’, following the modelling already discussed for progressive forms (presencing operations). This corresponds to the reported sense that be going to futures are closely connected to the present, to some process already in train at S's now. The main verb (here the visit component vector) appears as a timeless event vector at some point t_i > t₀ that might be specified pragmatically or, lexically, by a PP (here ‘this Thursday’).

Figure 5.6 Example (11) Henry is going to/gonna visit Calais this Thursday

Presencing operations usually leave the culmination of activities ‘out of view’ but in the case of the periphrastic future construction that uses them the preposition to specifies the end point, namely the event expressed in the main verb (here visit). Nonetheless, the model also captures the fact that from S's present viewpoint, the culmination is ‘outside’ her peripersonal space. And as for all progressive forms, the presencing operator also leaves ‘out of view’ the beginning point of the activity, which equally seems to be appropriate for the conceptual effects of the periphrastic future that are often described.

These features seem to correspond well with the sense that periphrastic futures refer to future events that are expected with a high degree of certainty because the speaker has experiential evidence in the present that they will take place in the subjectively ‘near’ future. Another way of putting this (Jaszczolt Reference Jaszczolt2009: 65, reporting Eckardt Reference Eckardt2006), is to say, in Reichenbachian terms, that E (event time) is in the future and R (reference time) is in the present and that be going to sentences are about ‘the present time and what is imminent in it’, or perhaps, one might say, ‘immanent in it’.

Note that there can be a subtle conceptual difference in Henry is going to visit Calais next week. The speaker S may be imagining Henry's intention, or S may be communicating her/his own prediction. DST modelling handles this. In Figure 5.6 we have the case where S projects her own world R′ in which Henry visits Calais at a future time. To model the meaning in which S is communicating Henry's intention, the origin of R′ has Henry's coordinate on the d-axis.⁴

5.4 The putative future: a reference frame solution

Across languages it is not uncommon to find that a future-tense form occurs in contexts that are interpretable and that yield a cognitive effect generally described as modal, or epistemic, or as we shall prefer here ‘putative’. I am not concerned with the precise pragmatic factors that allow such understandings, merely with their conceptual structure when they do occur.

As Jaszczolt argues, the fact that the conventional future form will + V can express present probability may be evidence of the close cognitive connection between future time reference and the inherent cognitive (or even metaphysical) uncertainty of the future, but this is not a necessary reason to conflate modality and temporality when there is good reason, for the purpose of semantic modelling, to think that they are, or can be, separate in linguistically expressed cognition. The English putative future meaning of will + V is not just a conventionalised expression of high probability. The meaning of Mary is probably writing the report is not the same as Mary will be writing the report (now) and is unlikely to be used in the same pragmatic contexts, i.e. have the same pragmatic potential effects.

Furthermore, the meaning of will + V seems to have a different internal conceptual structure from must, may and might + V. French linguists have noted that the equivalent French form, the futur putatif, presupposes a future point of view at which the present event is predicted to be verified (Damourette and Pichon Reference Damourette and Pichon1911–36, Sthioul Reference Sthioul, Moeschler, Jayez, Kozlowska, Luscher, Sthioul and de Saussure1998, Saussure and Morency Reference Saussure and Morency2012, Saussure Reference Saussure2013). These authors speak of an ‘imaginary’ future ‘perspective’ or ‘viewpoint’. They also insist that this viewpoint is allocentric, in the sense that it is distinct from the current speaker in the actual world, a claim that DST frames in a slightly different way. The general approach of these authors, however, is adopted here.

The conceptualisations we have to consider arise in the context of certain kinds of prompts for deictic times, illustrated in the following:

A simple present version of (14)Henry will visit Calais now in the putative sense is not valid because of the regularities governing non-statives.

Sentence (12) is the ‘regular future’ and (13) is a ‘regular future’ use of will combined with the progressive form (presencing operator) that is represented by the insertion of a new viewpoint in a new R′ located at some time in the future relative to S – this is shown in Figure 5.7 below. What do we have for (14), where we have a future progressive form understood as referring to the present now? We can simply put the visit verb vector at high probability on the m-axis. But this is not explanatory. We want to see if the frame-shift model of DST can account for this use of future form with auxiliary will.

Figure 5.7 Example (7) Henry will be visiting Calais this Thursday [non-putative]

Figures 5.7 and 5.8 show the different conceptual structures, in DST terms, of the different potential meanings of Henry will be visiting Calais, Figure 5.7 modelling the non-putative ‘regular future’ meaning, Figure 5.8 modelling the putative use of regular future to refer to a present event.

Figure 5.8 Example (14) Henry will be visiting Calais (now) [putative]

Unlike the future tense form will + V, the progressive form introduces the viewing frame (presencing operator), which involves an embedded set of coordinates, transposing a represented or ‘imaginary’ S to the future time point in the base coordinates indicated pragmatically by will at the deictic time t represented by this Thursday. (To simplify the account, the PRF has been omitted here.) The cognitive effect seems to be one in which the utterer of this sentence is ‘closer’ to the projected future event than in the sentence Henry will visit Calais next week. Next we have to consider whether this modelling approach can handle the conceptual effects triggered when the adverbial this Thursday is replaced by, for example, now, at this moment, etc. or by some equivalent but unexpressed pragmatic indicator, as in (14).

The frame-shifting principles of DST seem to be able to accommodate this putative future use of the future. The putative future is the reverse of (8). Whereas in the latter we have a present tense form referring to a future time, in (14) we have a future tense referring to present time, always relative of course to the speaker's now.

Broadly following the proposals of Sthioul (Reference Sthioul, Moeschler, Jayez, Kozlowska, Luscher, Sthioul and de Saussure1998) and Saussure and Morency (Reference Saussure and Morency2012), we have in Figure 5.8 a metarepresented point of view at some t_i > t₀. This additionally represented point of view does not have to be an unspecified allocentric other in the sense of a different individual S_i. It is a cognitive avatar of S, S′, and can still be regarded as ‘allocentric’ in that sense. This is also consistent with examples from contexts in which it is pragmatically manifest that the speaker herself will be in a position to verify, in the future, a current situation. In Figure 5.8 from the viewpoint of S′ the vector visit is in the past and it is true (located in the plane m = 0); from the viewpoint of S visit is in the present. The use of coordinate systems as reference frames makes it possible to model this property of the putative future construction.

Just as ‘regular future’ tense forms can, in the putative construction, refer to S's present time, so ‘regular future perfect’ tense forms can refer to events in S's past time. Such a reading is available for (15):

In the account proposed by Sthioul and by Saussure and Morency, sentence (15) may be said to take a view of a past event from a future vantage point at which it will be verified. The sentence is of course ambiguous, since it can also be read in a simply temporal sense as referring to a time t_i in S's future but before some future vantage point (i.e. t_j > t_i). One may indeed say that it is doubly ambiguous, since for some speakers this temporal reading of (15) may be either simply temporal or putative, again in the sense of Sthioul, Saussure and Morency. If this is so, the ambiguity indicates not that future time is inherently modal but that there are two possible conceptualisations of the future, one predictive and the other something short of predictive.

The geometric DST model for these sentences will have the same structure as in Figure 5.8 except that the vector representing visit will be located at some time prior to S's now, as in Figure 5.9, which illustrates a model for the non-progressive in order to simplify, and for a reading that is putative. (The embedded axis system has its origin at d = 0, in contrast with Figure 5.8, which illustrates a progressive tense form, i.e. presencing operator.)

Figure 5.9 Example (15) Henry will have visited Calais [putative reading]

In this model for the putative reading of (15) the embedded axes are still in the future relative to S, since, in the account of the putative meaning that we are adopting, verification is still in the future relative too S. What is different between (15) and (14) is the location of the event in the past relative to S, i.e. at some t_i < t₀. The diagram can show how the have been component of the construction, referring to the past relative to S, is conceptually represented simultaneously with the will component associated with future reference relative to S.⁵

In this chapter I have paid attention to only a small number of temporal expressions, in particular those that refer to the future, that is, the deictic future relative to some speaker S: simple present for future, progressive present for future, periphrastic present going to, with brief mention of the ‘regular’ willfuture. The overall aim has been to demonstrate how a geometrical approach formulated in DST can, perhaps surprisingly, elucidate temporal relations and in particular bring out the role of shifting reference frames. This approach separates the conceptual dimensions of modality and time, and one of the ancillary aims has been to characterise the distinctions between the different constructional means of referring to the future in terms of shifting point of view rather than modal scales. This is not to deny that modal effects may arise, but the present model does not treat them as inherently conflated with time as linguistically expressed. If and when they arise, they arise through contextual factors interacting with the conceptual structures built from reference frames. The solution proposed here to the problem of giving an account of present tense forms referring to past and future times consists of using translation transformations, which are part and parcel of coordinate geometry. The next chapter investigates a further type of reference frame transformation.