Navigation – Plan du site

AccueilNuméros47L’interaction entre modalité et t...

L’interaction entre modalité et temps verbaux en anglais

The interaction between modality and tense in English
Renaat Declerck
p. 53-76

Résumés

Cet article se veut une réflexion sur des emplois particuliers de temps verbaux qui engendrent une interprétation modale. L’auteur explique deux mécanismes qui sont à la base de l’interprétation modale et de ce fait indique les raisons pour lesquelles ces implications modales sont uniquement associées à certaines formes verbales. Il défend l’hypothèse que le premier mécanisme, un choix marqué de ’temporal focus’, est aussi utilisé différemment dans le système verbal en anglais, ce qui implique que la forme verbale en question est tout d’abord un indicatif. Mais le deuxième mécanisme, appelé « modal distancing », ne fait pas partie intégrante du système verbal en anglais. Ce constat implique qu’en anglais, il est nécessaire de distinguer un ’mode indicatif modal’ à coté du mode indicatif. L’analyse proposée permet à l’auteur de souligner l’importance d’un type de modalité épistémique qui est souvent négligé, à savoir l’idée que la situation n’est pas encore réalisée au moment du repère.

Haut de page

Texte intégral

1Declerck (2006) argues at great length that the core meanings and the basic uses of the English tenses concern the expression of temporal relations. However, it is well-known that certain tenses have uses that are at the same time temporal and modal (for example, the future tense) and that certain tenses have uses that would seem to be purely modal rather than temporal (for example, the forms had come and wouldn’t have found in If you had come tomorrow instead of today, you wouldn’t have found me at home). In this article we will discuss some of these uses with a view to finding the answers to the following questions:

  • (a) What is the origin of the modal interpretation of particular tense forms?

  • (b) Why do only some tenses have modal uses?

  • (c) Should we distinguish a ’modal indicative mood’ next to the ’indicative mood’, the ’imperative mood’ and the ’subjunctive mood’?

1. The I thought that . . . -construction

2This construction is analysed in detail in Declerck and Reed (2005). The main findings are the following. There are two modal uses of I thought that . . . , namely its use to express ’suspended factuality’ (implicating present counterfactuality) of the complement clause, as in (1a), and its use to express discourse tentativeness, as in (1b).

  • (1)

  • a. [What a surprise to meet you!] I thought you were in Australia!

  • b. [If you don’t mind my asking,] I thought you might do me a favour.

3In both cases thought is a normal preterite form: it locates the situation of thinking in the past and implies that the referent of the subject of thought did entertain the thought in question at some past time. The modal interpretations of the object clauses of thought arise by way of implicature from the fact that the thinking is not located at the time of speech in spite of the fact that this is the most relevant time for the speaker (unless the sentence forms part of a discourse in the past tense). In other words, the use of the past tense is a conscious violation of Grice’s (1975) ’Maxim of Relation (Relevance)’. Because he does not use I think, the speaker suggests that he is not committed to the current applicability of the proposition of the complement clause.

4In the case of suspended factuality, as in (1a), the motivation for the expression of noncommitment is that the speaker wants to suggest that he is not sure whether the content of the complement clause would be true if applied to the present world. In the case of discourse tentativeness, as in (1b), the noncommitment is suggested because the speaker wishes to make the content of the complement clause appear unimportant (to himself) in the present.

5These findings are relevant to the research questions (a)-(c) formulated above. The answer to question (a) [What is the origin of the modal interpretations?] is to be found in the conscious violation of the Gricean Maxim of Relevance by the use of the preterite where the present tense would be the unmarked choice. The answer to question (b) [Why do only some tenses allow modal interpretations?] is as follows. Although I have thought would also locate my thinking before S (speech time), it could not yield the two modal interpretations of I thought because in English there is an opposition between ’past time-sphere tenses’ and ’present time-sphere tenses’ (Declerck 2006: 152). Unlike the past tense (preterite), the present perfect locates the time of the situation referred to in the present time-sphere, more specifically, in the ’pre-present zone’ within the latter, and in doing so expresses a link with the present. The preterite, by contrast, implies a contrast between the past and the present. This is the reason why the preterite has to be used in sentences like He is no longer the brilliant scientist that he {was/*has been}. Needless to say, the modal interpretations of thought crucially depend on the presence of a sense of contrast between past and present.

6As to question (c) [Do we need to posit a separate ’modal indicative mood’?], the answer is that the data in connection with I thought that . . . fail to suggest that this is necessary. The two modal interpretations follow from the normal temporal use of the preterite, viz. locating a situation in the past, and the violation of a Gricean Maxim. The important thing is that thought is a quite normal indicative tense form. The fact that it can produce modal interpretations does not support the view that there is a separate ’modal indicative mood’ in English.

2. ’Formal anteriorization’ or ’distancing’

7By ’formal anteriorization’ or ’distancing’ (or ’modal backshifting’) we mean the phenomenon that in certain contexts verb forms are replaced by forms that seem to express a higher degree of ’pastness’ in order to produce a more ’distant’ (remote from reality) interpretation. Compare the following:

  • (2)

  • a. I will be happy if Sue comes.

  • b. I would be happy if Sue came

  • c. I would have been happy if Sue had come.

8Sentence (2b) is the result of applying distancing to both verb forms of (2a). The result is that the ’open’ condition (i.e. Sue may or may not come, but the speaker treats her coming as a real possibility) expressed by if Sue comes in (2a) is turned into a condition which is semantically ’tentative’ (Declerck & Reed 2001: 93), i.e. if Sue came expresses that it is not impossible but rather unlikely that Sue will come. Sentence (2c) is the result of applying distancing to both verb forms of (2b). The result is that the tentative condition expressed by if Sue came is turned into a counterfactual one: Sue did not come (or, if tomorrow is added, will not come).

  • 1 A temporal domain is a set of times which is such that each time is related to one of the other tim (...)

9It should be noted that distancing is not the same thing as ’backshifting’ (Jespersen 1932: 151) in indirect speech. Backshifing is the adaptation of the tense form of an independent clause when that clause is syntactically subordinated to a clause in the past tense: the original absolute tense (as in I have seen her) is replaced by a relative tense (as in He claimed that he had seen her). Unlike an absolute past tense form, which simply locates a situation in the past, a relative past tense form, which expresses simultaneity in a past ’temporal domain’,1is never backshifted — see Declerck (2006: 384). Consider:

  • (3) a. [He said:] “I will be happy when she is back.” (Will be is an absolute tense form, while is is a relative tense form expressing simultaneity.)
    b. He said that he would be happy when she was back. (Both forms are ’backshifted’ forms and are therefore relative tense forms: would be expresses posteriority, while was expresses simultaneity with the time of the situation referred to by would be.)

  • (4) a. [He said:] “I was happy until you came.” (Was is an absolute tense form, while came is a relative one — see Declerck 2006: 155).
    b. He said that he had been happy until I came. (Only the absolute form is backshifted.)

  • (5) a. [He said:] “I would be happy if she came.” (Both forms are distanced forms.)
    b. He said that he would be happy if she came. (Neither form is backshifted because both are distanced forms.)

  • (6) a. [He said:] “I would have been happy if she had come.” (Both forms are distanced forms.)
    b. He said that he would have been happy if she had come. (Neither form is backshifted because both are distanced forms.)

10Clearly, the gradual increase in semantic ’remoteness’ ( distance from reality) which we observe in conditionals like (2b) and (2c) is expressed by verb forms which are partly different from those resulting from applying backshifting to the same sentences in indirect speech: would be happy and came can become would have been happy and had come as a result of distancing — in (2c) — but not as a result of backshifting — see (5)-(6): (6b) is the indirect speech version, not of (5a) but of (6a). Moreover, we notice in (5b) and (6b) that distanced forms are not backshifted in indirect speech. Furthermore, while both absolute and relative past time-sphere tense forms can be distanced, as in (2b) and (2c), relative past time-sphere tense forms cannot be backshifted: this is why would be and came remain unchanged when (5a) is reported as (5b).

11Besides conditionals like (2b)-(2c), there are other cases in which a tense form can or must be distanced. Here we can mention the well-known uses of ’modal pasts’, which express either counterfactuality or discourse tentativeness, and of ’modal past perfects’, which express a combination of anteriority and counterfactuality. A first use is illustrated by the following, where the present tense could be used in the head clause but is replaced by the past tense for reasons of discourse tentativeness (i.e. to render the request less direct).

  • (7)

  • a. I {hoped/thought} you could help me clean up this mess.

  • b. I {wanted/hoped}to have a word with you.

  • c. I wondered if you could spare me a minute.

12Similarly, the preterite form of verbs like expect, intend, hope, mean, suppose, want, etc. can be distanced, so as to become a past perfect, in order to add the idea of counterfactuality (more specifically, the implicature of nonactualization) of the complement clause situation:

  • (8)

  • a. I {had hoped/had thought/had meant/had wanted} to see her {yesterday/today/tomorrow}.(Implicates: ’but I {did not/do not/will not} see her’.)

  • b. He {had supposed/had thought/had expected}that his mother would pay his debts.
    (Implicates that she did not.)

13After the verb wish (in any tense) and if only, a modal past expresses simultaneity plus counterfactuality, whereas a modal past perfect expresses a combination of anteriority and counterfactuality:

  • (9)

  • a. {If only/I wish/I wished} I had her telephone number.

  • b. {If only/I wish/I wished} she had given me her telephone number.

14Similarly, if the head clause uses a present time-sphere tense, a clause introduced by as if or as though uses a modal past to express simultaneity plus counterfactuality, and a modal past perfect to express anteriority plus counterfactuality:

  • (10)

  • a. He {has treated/treats/will treat} me as if I was his slave.

  • b. People now avoid me as though I had murdered my parents.

15In all these examples, the modal interpretations of the verb forms are due to the mechanism of modal distancing, and never to the context or to an implicature (though the modal interpretation itself may sometimes be an implicature, as in (8a-b).)

16The above observations about distancing (formal anteriorization) are relevant to the research questions (a)-(c) formulated at the beginning of this article. We have noticed that distancing always results in the use of a past time-sphere tense (i.e. preterite, past perfect, conditional tense or conditional perfect). A present perfect can never be used as a distanced form. We have also seen that the modal interpretations of the examples discussed are always due to the mechanism of modal distancing and never to the context or to implicatures. Moreover, it has been pointed out that distancing is a different phenomenon from backshifting. Backshifting is an adaptation of a tense when the temporal relation to be expressed changes, and is therefore a natural mechanism in the English tense system. By contrast, distancing has nothing to do with a change of temporal relation but has no other purpose than to bring forth some modal interpretation. Verb forms like would come can be distanced (would have come) but cannot be backshifted. All this suggests that the use of formally anteriorized tenses has nothing to do with the working of the English tense system and therefore does not belong to the realm of the indicative mood. Clearly, there is some positive evidence here supporting the existence of a ’modal indicative mood’. The fact that some modal preterite forms can be replaced by past subjunctives corroborates this conclusion:

  • (11)

  • a. I wish I {was/were} dead.

  • b. He treats me as if I {was/were} his slave.

  • c. If Bill {was/were}here, Samantha wouldn’t be here.

3. The notion ’not-yet-factual at t

3.1. Definition

  • 2  We use situation as a cover-term for anything that can be expressed by a verb phrase. According to (...)

17As is well-known, finite clauses express tensed propositions, and propositions create ’t-worlds’, that is, ’possible worlds’ which are anchored to a particular time t. Thus, I will be in Paris tomorrow creates a future t-world (t = tomorrow) of which the untensed proposition ’I be in Paris’ is true. It is also generally assumed that finite clauses express one of the epistemic values figuring on the scale ’true — probably true — possibly true — unlikely to be true — false’. The actual world as we know it, and which we can refer to as the ’S-objective world’ because it is temporally anchored to S (speech time), comprises every proposition that is true at S, i.e. any proposition that represents the actualization of a situation as ’factualat-S’.2A situation is factual-at-S if it either actualizes at S (e.g. I am writing an article, London lies on the Thames) or has actualized before S (e.g. Caesar was a Roman emperor). A propositions is true of some ’hypothetical’ t-world if it is probably true, possibly true or unlikely to be true of the S-objective world. Thus, Jim may come tonight represents John’s coming as factual in a hypothetical future world. A proposition is true of a ’counterfactual’ t-world if it is false of the S-objective world. Thus, [If] John had been here now . . . represents John’s being here as factual in a counterfactual S-world (and hence as counterfactual in the objective S-world).

18A clause representing a proposition as true in the objective S-world (e.g. London lies on the Thames) is not normally treated as modal. As far as propositions are concerned, ’hypothetical’ and ’false’ are the only accepted epistemic modal values. It follows that only a hypothetical or counterfactual situation can be the referent of a modal clause.

  • 3 When referring to the future tense in this article we will especially be concerned with its modal i (...)

19Although it is widely accepted, this view of epistemic modality is incomplete. Some hypothetical t-worlds can be characterized in terms of a modal concept which has been neglected in the literature on modality, viz. the idea that the relevant modal world is envisaged by the speaker but not yet factual at the time t to which it is anchored. This kind of t-world, which we will refer to as ’not-yet-factual at t’, is evoked by any expression that has posteriority as part of its meaning. The clearest cases are those in which the reference is to a future world. Thus, John will take the exam tomorrow evokes a modal world which is not-yet-factual at S but is envisaged or predicted at S to become factual at some future time t. It follows that any situation that is temporally located in that future t-world by the future tense is ’not-yet-factual at S’.3Another type of example is I saw Sam before she had seen me. Here the situation of the before-clause is represented as not-yet-factual at the time of the head clause situation. The sentence can therefore be paraphrased as ’I saw Sam at a time when she had not yet seen me’, or ’When I saw Sam, it was not yet a fact that she had seen me’ — see section 3.2below.

20Sentences representing the actualization of a situation as not-yet-factual in a t-world combine two ideas, viz. ’counterfactual at t’ and ’predicted or otherwise envisaged at t’. The latter idea implies ’envisaged as possibly actualizing at a time posterior to t’, i.e. ’envisaged as factual in a hypothetical world that might exist at a time posterior to t’. Envisage thus refers to something weaker than prediction. It just means that the speaker reckons with the possibility that a situation will actualize.

21We thus claim that the factuality value ’not-yet-factual at t’ is a combination of the values ’t-counterfactual’ and ’t-hypothetical’. Thus I will do it tomorrow implies that I am not doing it at S and have not done it before S ( counterfactuality at S) and at the same time predicts or envisages my doing it in a hypothetical future world anchored to tomorrow. Since the combination of the values ’t-counterfactual’ and ’t-hypothetical’ is different from either of the constituting values, we conclude that there are not three but actually four theoretically possible factuality values, viz. [` factual], [(purely) counterfactual], [` (purely) hypothetical] and [` not-yetfactual at t]. The latter three are possible realizations of ’nonfactual’ ( ’not factual’). Unfortunately, this means that the term ’nonfactual’ is potentially ambiguous. We have to distinguish carefully between the factuality value ’nonfactual’ (which contrasts with ’factual’ only) and the three subvalues which ’nonfactual’ allows, viz. ’counterfactual’, ’hypothetical’ and ’not-yet-factual at t’. In order to avoid confusion, it is better not to use the label ’nonfactual’ in the sense of ’not-yet-factual at t’, as I have done in earlier publications (e.g. Declerck 1991a: 105, 1991b: 153-154) in connection with not-yet-factual before-clauses (as in I saw him before he had seen me, which is interpreted as ’He had not yet seen me when I saw him’ — see section 3.2below).

22As a further illustration of the above claims, let us consider the following sentences:

  • (12)

  • a. Abraham Lincoln abolished slavery.

  • b. Jimmy Carter is the President of the U.S.A.

  • c. [I’d be surprised] if Fred was helping his sister right now.

  • d. Joan will bake a cake tomorrow.

  • e. [John promised] that he would fix the car.

  • 4 It should be noted, however, that not all sentences in the future tense yield a ’not-yet-factual at (...)

23In (12a), the actualization is factual in the objective S-world. In (12b), the actualization is (purely) counterfactual in the objective S-world. The actualization referred to in the if -clause of (12c) is (purely) hypothetical in the speaker’s S-world because the speaker does not really know whether Fred is helping his sister or not. In (12d), the actualization referred to is not-yet-factual in the speaker’s S-world.4And the actualization of the situation referred to in the that-clause of (12e) is not-yet-factual in John’s past t-world, i.e. in the objective world holding at the time when John made his promise.

24It should be clear now that each of the terms ’factual’, ’(purely) counterfactual’, ’(purely) hypothetical’ and ’not-yet-factual at t’ can be used in two ways. Firstly, they can refer to one of the factuality values which the actualization of a situation has in a given t-world. Secondly, the four terms can also be applied to particular kinds of t-worlds. For example, the conditional world evoked by I would feel safer if he was with me is a subjective counterfactual S-world because the situations that are factual in this S-world are counterfactual in the objective S-world.

25This means that the following conclusions have been reached:

  • (a) There are four types of t-worlds. We can call them ’factual’, ’purely counterfactual’, ’purely hypothetical’ and ’not-yetfactual at t’, respectively.

  • (b) The same four labels can be used for the four (mutually exclusive) factuality values that (the actualizations of) situations can have in given t-worlds.

  • (c) ’Not-yet-factual at t’ combines the ideas ’counterfactual at t’ and ’hypothetical at t’. However, it is incompatible with the idea ’purely counterfactual at t’ as well as with the idea ’purely hypothetical at t’.

26In sum, the factuality values that may be applicable to a given nonfactual t-world can be represented as in figure 1.

Figure 1. —Representation of the factuality values in nonfactual t-worlds

3.2 Cases of situations represented as not-yet-factual at t

27As noted above, a situation that is referred to in the future tense is typically not-yet-factual in the objective S-world. Similarly, a situation that is temporally located posterior to a past time, as in [John promised that] he would fix the car, is not-yet-factual in the objective world holding at that past time.

28In certain contexts, 'not-yet-factuality at t' can be expressed by modal indicative tense forms, more specifically by the past tense and the conditional tense. In those cases the 'pastness' of the tense cannot be explained in terms of location in time or purely temporal relations. For instance, the preterite may refer to a future situation rather than to a past one. The 'pastness' of the tense must therefore be due to modal distancing. The following illustrate this:

  • (13)

  • a. It is (high) time we changed the code. (The situation of our changing the code is not-yet-factual in the objective S-world: though it has not actualized yet, its actualization is envisaged for the future.)

  • b. I'd rather you left next week. (similar)

  • c. I wish it would stop raining. (The stopping of the rain is not-yet-factual in the speaker's S-world. Its actualization in a future world is envisaged as a possibility, however slight.)

29In (13a-b), 'not-yet-factuality as S' is expressed by the preterite. In (13c), it has to be expressed by the conditional tense because (as we have seen in section 2) after I wish the preterite is reserved for the expression of S-counterfactuality. Thus, in I wish I had a son, the situation of my having a son is represented as counterfactual in the objective S-world.

30Needless to say, ’formal anteriorization’ (modal distancing) is essential in examples like It’s high time we fixed the car, since the ’not-yet-factual at S’ meaning cannot be expressed by *It’s high time we fix the car. In I wish John would kill his wife the expression of posteriority creates the not-yet-factual meaning, while the anteriorization of the marker of posteriority (will) produces a tentative interpretation, i.e. the possibility that John will kill his wife is seen as rather unlikely. The above sentence can be compared with I hope John twill/*wouldu kill his wife, where the second aspect of meaning cannot be expressed.

31The tenses used to express not-yet-factuality in t-worlds that hold at a time other than S (i.e. in the past or future) are the same as those used in connection with S-worlds:

  • (14)

  • a. It was (high) time we changed the code.

  • b. I wished it would stop raining.

  • c. [When we return from our holiday] it will be (high) time we changed the code.

  • d. Before the end of the day you will wish that it would stop raining.

32This illustrates that, unlike backshifting, modal distancing happens independently of the nature of the tense used in the head clause.

33Another typical context in which a not-yet-factual meaning can be produced by modal distancing consists of sentences involving a before-clauses in the past perfect. These can represent the actualization of the situation referred to as not-yet-factual in the t-world that holds at the time of (the actualization of) the head clause situation. For example:

  • (15)

  • a. I saw Sam before she had seen me.

  • b. [Another fifteen to twenty thousand snakes perished in the winter of 1998-9,] when a sudden frost caught them before they had descended below the frost line. (www)

34The actualization of the before-clause situation is represented here as not-yet-factual at the time of the actualization of the head clause situation. Thus, in (15a) it can be paraphrased in either of the following ways:

  • (16)

  • a. I saw Sam at a time when she had not yet seen me.

  • b. Sam had not yet seen me when I saw her.

35These paraphrases make it clear that the interpretation of the before-clause is neither factual nor purely counterfactual in the past t-world in which my seeing Sam is factual. A t-factual interpretation of the before-clause would be triggered by the past tense:

  • (17) I saw Sam before she saw me. (= ’We saw each other, but I saw her first.’ Note that there is reference to two different past t-worlds here: that in which my seeing Sam is t-factual and that in which Sam seeing me is t-factual. The latter t-world follows the former in time. This posteriority relation is exclusively expressed by before.)

36A purely counterfactual interpretation of the before-clause situation would be triggered by the conditional perfect:

  • (18)

  • a. I saw Sam just before she would have seen me [if I had not quickly turned into a side street].

  • b. I paid the electricity bill two days before the company would have cut offmy supply. (The conditional perfect is possible because of the implicit condition ’if I had not paid the bill’.)

37The interpretation which had seen does receive in I saw Sam before she had seen me is a ’not-yet-factual at t’ reading: in the world anchored to the time of my seeing Sam, the situation of her seeing me was counterfactual but its actualization was envisaged. ’Envisage’ here refers to a weaker notion than prediction. It just means that I reckoned with the possibility of her seeing me. Nothing is said about whether she also saw me or not — see footnote 5.

38Another case in which a not-yet-factual reading is possible concerns adverbial until-clauses and before-clauses depending on a finite or nonfinite clause which itself depends on a head clause whose verb is an intensional verb like intend, decide, promise, etc., which implies that the actualization of the situation of the complement clause is posterior to the time of the intensional attitude (i.e. the intention, decision, promise, etc.). In such a context, the subjective world in which the actualization of the situation of the temporal clause introduced by until or before is factual is posterior to the t-world in which the making of the decision, promise, etc. is factual. Compare:

  • (19)

  • a. Meg nagged Tim until he gave in. (Both situations are interpreted as factual in the objective S-world.)

  • b. Meg decided to nag Tim until he gave in. (The until-clause situation was not-yet-factual at the time of the beginning of the nagging and the nagging itself was not-yet-factual at the time when Meg made her decision. It follows that the until-clause situation was also notyet-factual at the time of Meg’s decision. Note that we do not know from this sentence whether or not Tim eventually gave in.)

  • c. Meg nagged Tim for three weeks before he gave in. (Both situations are represented as factual in the objective S-world.)

  • d. Meg decided {to wait/that she would wait} for some time before she replied to Tim’s letter. (The before-clause situation was not-yet-factual at the time of the beginning of the waiting, which was itself not-yet-factual at the time when Meg made her decision. It follows that the before-clause situation was also not-yet-factual at the time of Meg’s decision. Note that we do not know from this sentence whether or not Meg eventually replied to Tim’s letter.)

39Other temporal clauses depending on an infinitive clause functioning as object of decided, promised, intended, etc. are interpreted as not-yet-factual, not in the t-world in which the infinitival clause situation actualizes, but in the t-world in which the decision, promise, etc. actualizes:

  • (20) I intended to call up Bill {when he was at home/after Meg had left/ while he was at his office}. (The time clause situation is factual in the hypothetical t-world in which my calling up Bill is factual, but both these situations — my calling up Bill and the time clause situation — are not-yet-factual at the time of my intention, i.e. they are not-yet-factual in the objective S-world in which the existence of the relevant intention is a past fact.)

40In this example, the ’not-yet-factual at t’ interpretation is not induced by the temporal conjunction (as is the case in I saw Sam before she had seen me — see section 3.3) but by the relation of posteriority that is implicit in the meaning of the verb intend.

3.3. The origins of ’not-yet-factuality at t’ and the means of expression

41As noted in 3.1, ’not-yet-factuality at t’ always requires reference to a posterior time. This is self-evident in connection with examples like John will kill you, I thought that John would kill you and I wish John would kill you. It is also clear in It’s high time we fixed the car, where fixing the car can only be envisaged for the future, and in I decided to tell him the truth {when/after/while} he came in, in which not only the telling but also the actualization of the time clause situation can only be interpreted as following the decision.

42Posteriority also plays a crucial role in before-clauses. To see this clearly, it is necessary to say something more about the temporal structure that is expressed by a sentence involving a not-yet-factual before-clause. Let us consider the following example:

  • (21) I saw Tim before he had seen me.

43The before-clause here receives a not-yet-factual interpretation: the sentence can be paraphrased ’I saw Tim at a time when he had not yet seen me’ or ’Tim had not yet seen me when I saw him’. As explained in Declerck (1991b: 153-154, 2006: 709-710), the temporal structure of the sentence is as follows:

  • (a) Saw locates the situation (of my seeing Tim) in the past, thus representing it as a past fact.

    • 5 The conjunction before has developed from a phrase of the form ’before the time that’ or ’before th (...)

    (b) Before means ’before the time that’.5This means that before locates an implicit t1posterior to the time of saw.

  • (c) Since t1is posterior to the time (t) of saw, the t1-world anchored to t1is not-yet-factual att(t = the time of saw).

  • (d) Had seen represents Tim’s seeing me as anterior to the implicit t1.

  • 6  However, it will be pointed out at the end of section 3.4that at least for some speakers the tense (...)

44At first sight this seems a purely temporal structure, which only involves times and temporal relations. The times are the times of the actualizations of the situations referred to by saw and had seen, plus the implicit time t1inherent in the meaning of before. The temporal relations are the relation expressed by the preterite (= ’past relative to S’) and the relation expressed by the past perfect (= ’anterior to a t in a past temporal domain’), plus the posteriority relation (= ’t1posterior to t’) expressed by before (but not by a tense form). However, there are two factors that induce the reading that the situation of Tim’s seeing me was not yet a fact at the time when I saw him. One is the fact that the situation of Tim’s seeing me is temporally related to an implicit posterior time t1. The other is the fact that the tense in the before-clause is ’modally distanced’ (’formally anteriorized’): we use had seen instead of the past tense form saw. This tense choice is important, because I saw Tim before he saw me is naturally interpreted as meaning that Tim did see me after I had seen him.6The same is true of before-clauses referring to the future: to convey the not-yet-factual reading we will use the present perfect (I will see Tim before he has seen me) although there is no purely temporal reason to express anteriority. The reason for using has seen is that the unmarked interpretation of the corresponding present tense form sees in I will see Tim before he sees me is different, viz. ’Tim will see me, but only after I have seen him first.’ We therefore need to apply modal distancing to get a not-yet-factual reading.

45In sum, the use of a past perfect or present perfect in a before-clause that is to receive a not-yet-factual-at-t interpretation is not the result of the application of a principle that is inherent in the English tense system (like backshifting) but is the result of the application of the principle of ’modal distancing’ (’formal anteriorization’) which has no other purpose than to produce a modal interpretation. In general, distancing strengthens the sense of hypotheticality (compare If he comes . . . with If he came . . . ) or expresses counterfactuality (as in If he had come . . . ). However, in before-clauses it evokes a ’not-yet-factual at t’ reading. As noted before, distancing can produce shifts (e.g. from would be to would have been) which cannot be effected by backshifting. This means that, unlike backshifting, modal distancing is not a mechanism that is inherent in the English tense system (see also section 2above). As noted before, backshifting typically occurs in indirect speech, for example, when John {died/has died} is reported as I told Bill that John had died. This kind of backshifting is no more than the phenomenon that anteriority to a past orientation time has to be expressed by the past perfect whereas anteriority to S has to be expressed by the past tense or the present perfect. By contrast, the modal distancing which we observe in the before-clauses of I saw him before he had seen me and of I will see him before he has seen me is the phenomenon that a tense form expressing anteriority (to the implicit t inherent in the meaning of before) is used instead of a tense form expressing simultaneity. The purpose is purely modal, in this case to convey a ’not-yet-factual at t’ interpretation of the before clause situation. The fact that an anteriorized form is used in not-yet-factual before-clauses is therefore not surprising. We noted in section 2that using an anteriority form where there is no purely temporal reason for using it is a widespread device to trigger particular epistemic modal meanings. For example, instead of saying I intended to visit him, which merely expresses that at some past time I had the intention of visiting him, we can use I had intended to visit him to express that, in spite of my past intention, I did not and will not visit him. Distancing here triggers a counterfactual interpretation of the situation referred to in the infinitive clause. As we have seen, the same mechanism can also induce tentative modality, as in I wanted to have a word with you and I would be happy if she arrived tomorrow, which are tentative versions of I want to have a word with you and I will be happy if she arrives tomorrow.

4. ’Strong’ versus ’weak’ not-yet-factual interpretations

46We can speak of ’strong not-yet-factuality’ if, as in the above examples, the expression of posteriority is accompanied by formal anteriorization of the tense form. We can speak of ’weak not-yet-factuality’ if there is posteriority but no anteriorization (distancing), as in the following examples:

  • (22)

  • a. I will visit them tomorrow. (The actualization of the situation is located in a t-world that is posterior to the speaker’s S-world. This implies that my visit is not-yet-factual at S.)

  • b. John promised that he would visit me the next day. (The actualization of the situation is located in a t-world that is posterior to the past t-world in which the making of the promise is factual. This implies that John’s visit was not-yet-factual at the time of his promise.)

  • c. I decided to wait until the shop opened. (similar)

47These examples illustrate the obvious fact that any situation whose actualization is conceived of as posterior to a time t is not yet factual at t. This is ’weak not-yet-factuality’: the idea of ’not-yet-factual at t’ inevitably forms part of the interpretation of the clause, but it is not the predominant aspect of meaning. Strong not-yet-factuality means that the extra mechanism of modal distancing puts emphasis on the idea of not-yet-factuality. In other words, whereas there is weak not-yet-factuality at t whenever a situation is posterior to the t in question, there is strong not-yet-factuality only if the ’not-yet-factual at t’ meaning is linguistically emphasized by the use of an anteriorized tense form. In some cases this use is obligatory. The following examples further illustrate the distinction:

  • (23)

  • a. “[We think] it’s time something’s done,” says Donald Skadden. (Wall Street Journal) (The sentence in the present tense only expresses weak not-yet-factuality. It just means ’It is time to do something’ — a meaning which can also be expressed by It’s time something should be done. The idea that nothing has been done yet is not foregrounded.)

  • b. It’s time something was done [about it]. (In this case strong not-yet-factuality is expressed. The idea ’It is time to do something about it’ forms part of the interpretation, but the foregrounded aspect of meaning is the idea that nothing has been done about it yet. The combination of these two ideas naturally leads to the interpretation that something should already have been done about it, i.e. that taking measures has been postponed for too long.)

48For at least some speakers, before-clauses need not use an anteriorized form if a t-factual reading is ruled out by the context in which the before-clause is used. For example, compare the following:

  • (24)

  • a. She read the letter before I read it.

  • b. She read the letter before I had read it.

  • c. She tore up the letter before I {had read/read}it.

49Because of the posteriority relation expressed by before, the situation of my reading the letter is automatically weakly not-yet-factual in the t-world in which the actualization of the head clause situation is factual. The function of using the anteriorized (distanced) form had read in (24b) is therefore to bring not-yet-factuality into focus, because the use of read in (24a) suggests that I read the letter too. In the latter case the speaker focuses on the t-factuality of the before-clause situation in the objective S-world rather than on the not-yet-factuality of the situation in the anterior t-world established by the head clause in the past tense. In (24c), the head clause pragmatically excludes the possibility of actualization of the before-clause situation, so that the idea of not-yet-factuality comes to the fore regardless of whether there is anteriorization or not. This explains why at least some speakers can replace had read with read without creating confusion.

4.1. Strong ’not-yet-factuality at t’ and the research questions

  • 7 As we have seen, there is only one exception to this rule, viz. the use of the present perfect (rat (...)

50The first two research questions were: ’Why do only some tenses have modal uses?’ and ’What is the origin of a modal interpretation?’. The answers are simple. The origin of a strong ’not-yet-factual at t’ interpretation is the application of the rule of formal anteriorization (modal distancing). The possibilities allowed by this rule are well-defined. As a rule, the distanced form has to be a preterite (If he came . . . ), a past perfect (If she had come . . . ), a conditional tense form (I would be happy if . . . ) or a conditional perfect (I would have been happy if . . . ). In other words, a distanced form is a past time-sphere form, not a present time-sphere tense form like a present tense, a future tense, a present perfect,7or a future perfect. So, it naturally follows from the origin of the modal ’not-yet-factual at t’ interpretation that only some tense forms can express this reading.

51The third research question was whether we should distinguish a ’modal indicative mood’ next to the ’indicative mood’, the ’imperative mood’ and the ’subjunctive mood’. The fact that unlike backshifting, distancing is not a mechanism inherent in the English tense system suggests that the answer to this question is positive.

5. The status of the ’future tense’: tense or mood?

  • 8 Declerck (2006: 105-106) argues that the ’predictability’ sense of That will be the milkman (which (...)

52The issue is also pertinent to the moot point whether will in Bill will arrive tomorrow is a future tense auxiliary or a modal auxiliary. Quirk et al. (1985: 213) claim that “there is no future tense in English” and that will is a “modal verb”. The modality in question is said to be prediction. Huddleston & Pullum (2002: 208) speak of “the lack of a future tense in English” and claim that futurity is a use of will that falls “within the general idea of epistemic modality” (p. 188) but they do not specify what kind of epistemic modality is expressed in Bill will arrive tomorrow. This point of view is shared by many other linguists. However, our position, defended in Declerck (1991a: 8-13, 2006: 102-106), is that will arrive is a future tense form because it locates Bill’s arrival in the future. We subscribe to Lyons’ (1977: 68) and Comrie’s (1985: 9) view that tense is a grammaticalized means of locating the actualization of a situation in time. Linguistically speaking, will arrive does exactly this (viz. locate a situation in time) and nothing else. It is therefore a future tense form.8Of course, it is true that the predicted situation is not-yet-factual at S, but this is simply because any situation that is predicted to actualize later than S is not yet factual at S. This is a question of how the objective world is rather than of how it is linguistically represented. As is true of any tense, the function of the future tense is to represent a situation as actualizing at a particular time. If the time in question is represented as posterior to S, the situation is automatically not-yet-factual at S (in the weak sense of the term). There is nothing in a future tense form to hint at the possibility of strong not-yet-factuality: there is no anteriorization or any other formal means of focusing attention on ’not-yet-factuality at S’.

53Moreover, in connection with the future tense, the idea ’not-yet-factual at S’ appears to have the strength of an implicature only. There are sentences in the future tense in which the idea is blocked (= prevented from arising) or cancelled by an element in the sentence or in its context or by our pragmatic knowledge of the world. Consider:

  • (25)

  • a. I will still be here tomorrow. (Still blocks the implicature that I am not yet here at S.)

  • b. [For the moment you can’t beat me. You’ll be able to beat me when I’m fifty] because you will be ten years younger than me. (The addressee is already ten years younger than the speaker, but it is only when the speaker becomes fifty that this difference of age will play a decisive role. The future tense is used here to put the ’temporal focus’ (Declerck 2006: 573-576) on that future time.)

54Similarly, the sentence John will be in London tomorrow does not rule out the possibility that John is already in London at S, so that the sentence can be used in a context inducing this interpretation. However, it actually only predicates something of tomorrow and locates John’s being in London at that time. It does not say anything about the present. In the same way as John was in London yesterday implicates that John is no longer in London at S, John will be in London tomorrow implicates that John is not yet in London at S. These implicatures are due to the Gricean Maxim of Relation (relevance): other things being equal, the present is more relevant to the speaker than the past or the future. So, if a situation is actualizing at S, it should in principle be referred to in the present tense. If another tense is used, the hearer has a right to conclude that the situation is not actualizing at S. However, this conclusion is a cancellable implicature, as is clear from John was in London yesterday — in fact, he still is.

Conclusion

55We have drawn attention to a number of cases in which a tense form has a modal implication or implicature. Some of these (for example, the modal interpretations of the object clauses of I thought that . . . ) follow naturally from the use of a mechanism that is inherent in the English tense system (e.g. using the past tense to put the ’temporal focus’ on the past rather than on the present). However, there are also cases in which a modal interpretation, especially that of strong not-yet-factuality at t, is the result of formal anteriorization (modal distancing) of a verb form — a mechanism which has no other use and therefore does not form part of the English tense system. We can hardly escape the conclusion that these anteriorized verb forms do not form part of the indicative mood in English, hence that we have to assume the existence of a separate ’modal indicative mood’ in English.

Haut de page

Bibliographie

Comrie, B.A. 1985, Tense, Cambridge, Cambridge University Press.

Declerck, R. 1991a, Tense in English: its structure and use in discourse, London, Routledge.

Declerck, R. 1991b, A comprehensive descriptive grammar of English, Tokyo, Kaitakusha.

Declerck, R. 2006, The grammar of the English verb phrase. Vol. 1: The grammar of the English tense system, Berlin/New York, Mouton de Gruyter.

Declerck, R., Reed, S.2001, Conditionals: a comprehensive empirical analysis. Berlin/New York, Mouton de Gruyter.

Declerck, R., Reed, S.2005, “What is modal about I thought that . . . ?”,English Language and Linguistics, 9, 311-332.

Grice, H. P. 1975, “Logic and conversation.” In: Cole, P., Morgan J. L., (eds), Syntax and Semantics, Volume 3: Speech Acts, New York, Academic Press, 41-58.

Huddleston, R., Pullum G.K.2002, The Cambridge grammar of the English language, Cambridge, Cambridge University Press.

Jespersen, O. 1932, A modern English grammar on historical principles. Part 4: Syntax (3rd volume), London, Allen and Unwin.

Lyons, J. 1977, Semantics, Cambridge, Cambridge University Press.

Mitchell, B. 1987, Old English syntax. Vol. II: Subordination, independent elements, and element order. Oxford, Clarendon Press.

Quirk, R., Greenbaum, S., Leech, G., Svartvik, J.1985, A comprehensive grammar of the English language, London, Longman.

Haut de page

Notes

1 A temporal domain is a set of times which is such that each time is related to one of the other times by a tense form. See Declerck (1991a: 20, 2006: 121) for a more detailed definition.

2  We use situation as a cover-term for anything that can be expressed by a verb phrase. According to Lyons (1997), a situation is either a state, an action, a process (= change, development) or an event (= a nonagentive dynamic situation, e.g. a fall). The term actualize is used as a cover-term for all the verbs that are typically associated with a kind of situation. The sentence The situation is actualizing can thus be said of a state that is holding, an action that is being performed, a change that is taking place or an event that is happening. In what follows, ‘the situation’ will often be shorthand for ‘the actualization of the situation’. Thus, the formulation ‘The preterite locates the situation in the past’ is short for ‘The preterite locates the actualization of the situation referred to in the past’.

3 When referring to the future tense in this article we will especially be concerned with its modal implication. This does not alter the fact that we treat the future tense as a real tense — see noteon page 62.

4 It should be noted, however, that not all sentences in the future tense yield a ’not-yet-factual at S’ interpretation. As we will see in section 4, this reading is an implicature which can be blocked by another constituent of the sentence (as in I will still be here tomorrow) or cancelled by contextual information or by pragmatic knowledge.

5 The conjunction before has developed from a phrase of the form ’before the time that’ or ’before then that’, variously realized in Old English as toforan þam timan þe, foran to þam timan þe and toforan þam þe — see Mitchell (1987: 379). The reduction to before that in Middle English, and later to before, has not changed this meaning. Before still means ’before the time that’, in which the time functions as an implicit orientation time as far as the use of the tense in the before-clause is concerned. That is, in I saw Tim before he had seen me, the form had seen represents Tim’s seeing me as anterior to an implicit t, thus locating it in the t-world which is not-yet-factual at t. Hence the not-yet-factual interpretation of the before-clause, which is vague as to the question whether or not Tim saw me after I saw him. This vagueness can be resolved by the context, as in I saw Tim before he had seen me, so I called out his name versus I saw Tim before he had seen me, so I had time to hide into a porch because I didn’t want him to see me.

6  However, it will be pointed out at the end of section 3.4that at least for some speakers the tense of the before-clause does not need to be ’anteriorized’ if a t-factual reading is already ruled out by the context in which the before-clause is used. An illustration of this is He burned the letter before I read it.

7 As we have seen, there is only one exception to this rule, viz. the use of the present perfect (rather than the present tense) in the not-yet-factual before-clauses of sentences like I will try to see him before he has seen me.

8 Declerck (2006: 105-106) argues that the ’predictability’ sense of That will be the milkman (which is similar to French Ce sera le laitier) follows from the fact that the speaker uses the future tense to put the ’temporal focus’ on some future time: “That’ll be the milkman implies something like ’as you will see when you open the door’.” This means that will be is a normal future tense form which is used in accordance with one of the principles inherent in the English tense system, viz. that the choice of tense can be determined by the choice of ’temporal focus’.

Haut de page

Table des illustrations

Légende Figure 1. —Representation of the factuality values in nonfactual t-worlds
URL http://journals.openedition.org/praxematique/docannexe/image/2795/img-1.png
Fichier image/png, 21k
Haut de page

Pour citer cet article

Référence papier

Declerck, R. (2006) L'interaction entre modalité et temps verbaux en anglais. Cahiers de praxématique 47. Montpellier : Pulm. 53-76

Référence électronique

Renaat Declerck, « L’interaction entre modalité et temps verbaux en anglais »Cahiers de praxématique [En ligne], 47 | 2006, mis en ligne le 01 janvier 2010, consulté le 26 février 2021. URL : http://journals.openedition.org/praxematique/2795 ; DOI : https://doi.org/10.4000/praxematique.2795

Haut de page

Auteur

Renaat Declerck

K.U. Leuven Campus Kortrijk E. Sabbelaan 53, B-Kortrijk 8500, Belgium
renaat.declerck@kuleuven-kortrijk.be

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page
  • Logo Praxiling
  • Logo Presses universitaires de la Méditerranée
  • Logo ERIH PLUS
  • OpenEdition Journals
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search