V. The principle of unmarked temporal interpretation 427 V. The principle of unmarked temporal interpretation 8.40 Introduction When there is a shift of domain within the same absolute zone (as in She felt bored. It was raining and everybody else was sleeping), the tense forms them- selves do not express the temporal relation between the domains Ϫ see 8.21.1. It follows that such a shift of domain is only pragmatically acceptable if the temporal order of the situations is either irrelevant or recoverable in some other way, e. g. from the use of time adverbials, from the order in which the situations are reported, from the linguistic context, from pragmatic knowledge (i. e. the extralinguistic context and our general knowledge of the world) or from the bounded or nonbounded aspectual character of the various clauses. It is the role of (non)boundedness that is discussed in the following subsections. 8.41 The ‘Principle of Unmarked Temporal Interpretation’ 8.41.1 The role of (non)boundedness is summarized in the following ‘Prin- ciple of Unmarked Temporal Interpretation ’: In a sequence of unembedded clauses where each clause establishes a domain of its own within the same absolute zone and where there is no adverbial, contextual or pragmatic indication of temporal order, the unmarked temporal W-interpretation is as follows: (a) when the situations are represented as bounded, they are interpreted as following each other in the order in which they are reported; (b) when the situations are nonbounded (i. e. not represented as bounded), they are interpreted as W-simultaneous with each other; (c) when one situation is represented as bounded and the other as nonbounded, the bounded situation is interpreted as temporally included in the nonbounded one. Before we have a closer look at these claims, it is worth stressing the following point in order to rule out any misunderstanding. The Principle of Unmarked Temporal Interpretation (abbreviation: PUTI) holds only for sequences of clauses that satisfy the following two requirements: (a) each clause in the se- quence establishes a domain of its own within one and the same absolute time- zone; (b) the temporal order of the situations is neither indicated by an adver- bial nor recoverable from contextual information or from pragmatic knowl- edge. In order to ensure that we are dealing with domain-establishing clauses we will illustrate each claim not only with examples in the past tense but also with examples that refer to the post-present. (As we have seen, the difference 428 8. Temporal domains and relative tenses: theoretical foundations between absolute and relative tense forms is not always clear when we refer to the past, since a preterite form can be either a relative or an absolute tense form. The difference is, however, clear when the reference is to the post-pres- ent, because the tense form used to express T-simultaneity in a post-present domain (viz. the present tense) is different from the tense form used to establish the domain (viz. the future tense) Ϫ see 10.1Ϫ2. This means that we can assume as working hypothesis that the past tense clauses create a new domain if the corresponding clauses referring to the post-present use the future tense.) 8.41.2 The Principle of Unmarked Temporal Interpretation involves the fol- lowing claims: (a) The unmarked interpretation of a sequence of bounded clauses (which each establish a domain of their own in the same absolute zone) is in terms of what we may call ‘ iconic sequencing’: the situations are interpreted as following each other in the order in which they are reported. For example: John went to the door and knocked three times. The door opened slowly. John will go to the door and (will) knock three times. The door will open slowly. In sentences like these, the situations tend to be interpreted as following each other directly, unless such an interpretation is pragmatically impossible or un- likely. 26 This is the case in examples like He moved to London and found a job there as a teacher. He courted a girl called Mary and married her. Perhaps he will move to London and (will) find a job there as a teacher. And perhaps he will court a girl and (will) marry her. In sentences like these, the situations need not follow each other directly. There may be long intervals of time between them. The tendency to interpret a sequence of bounded clauses in terms of succes- sion is just a tendency, not a strict rule. When the clauses have different subjects and there is no contextual or pragmatic reason for adopting the succession interpretation, any of the three temporal readings is in principle available. A sentence like Bill studied chemistry and John studied mathematics tells us noth- ing about the temporal order in which these two situations actualized. We may wonder, then, why it is that a sequence of bounded clauses is some- times interpreted in terms of succession, as in the first two examples below, and sometimes allows any temporal order interpretation, as in the last two ex- amples: 26. In other words, the fact that the immediate succession reading is favoured is due to a conversational implicature, which can be cancelled. V. The principle of unmarked temporal interpretation 429 The man went to the door, opened it and shouted something. The man will go to the door, (he will) open it and (he will) shout something. [“Did any of them marry?”] Ϫ “Yes, Bill married Tina, and John married some Italian girl.” [“Will any of them marry?”] Ϫ “Yes, Bill will marry Tina, and John will marry some Italian girl.” There appear to be two factors that are conducive to a succession interpreta- tion. The first is pragmatic knowledge. We know that going to the door must take place prior to opening it. The second is the kind of text or discourse that the sequence of L-bounded clauses is part of. If this is a story that progresses in time, then each bounded clause that is added will normally be interpreted as carrying the action forward. That is, each new situation will be interpreted as following the situation of the previous clause. The first two of the above examples appear to be part of such a story with a linear temporal structure. The answers in the last two examples are not, and therefore need not be inter- preted in terms of succession. (b) The second claim made in the Principle of Unmarked Temporal Interpreta- tion presented in 8.41.1 is that in a sequence of (unembedded) L-non- bounded clauses the unmarked interpretation is for the situations to be interpreted as W-simultaneous. 27 John stood by the window. Mary was in the kitchen. The man seemed reliable. He worked in a bank. John carried the suitcase. Bill pushed the cart. He enjoyed and admired the scenery of the Lake District. John will stand by the window. Mary will be in the kitchen. In ten years’ time the man will seem reliable. He will work in a bank. John will carry the suitcase. Bill will push the cart. [When he is older] he will enjoy and (will) admire the scenery of the Lake District. In cases like these all the verb forms establish W-simultaneous domains. As we have seen, L-nonbounded situations are represented as homogeneous, which means that there is no particular reference to an initial or terminal point. For this reason, nonbounded situations may in principle go on indefinitely. This explains why the two nonbounded situations referred to in each of the above examples are not normally interpreted as preceding or following each other. There is simply no reason why one of them should be interpreted as beginning only after, or as terminating before, the other situation. This is true even if the 27. Remember that we are dealing with unembedded domain-establishing clauses that do not contain an adverbial indicating the temporal order of the situations. 430 8. Temporal domains and relative tenses: theoretical foundations clauses form part of a story. Nonbounded clauses (not containing a time adver- bial) do not normally carry the action forward. They have a ‘ backgrounding’ rather than a ‘ foregrounding’ function. (In a narrative text, the ‘foreground’ is the linguistic material which charts the progress of the narrative through time, while the ‘background’ consists of durative and descriptive material which serves to elaborate on the foreground.) There is one exception to this rule. Nonbounded clauses do carry the action forward if they are interpreted inceptively (inchoatively) and therefore suggest that the situation in question did not obtain earlier. This is the case, for exam- ple, in the following, where the inceptive reading is induced by the presence of suddenly: [We were sitting in the dining-room.] Suddenly there was a loud noise in the street. (c) The third claim made in the Principle of Unmarked Temporal Interpreta- tion set out in 8.41.1 is that, when one of the two situations is L-bounded whereas the other is not, the unmarked interpretation is for the bounded situation to be included in the nonbounded one. (This is a kind of W- simultaneity relation.) Mary was in the drawing-room. Suddenly Bill came in. John went out of the room. He was feeling very tired. Mary will be in the drawing-room. Suddenly Bill will come in. John will go out of the room. He will be feeling very tired. The reason why the nonbounded clauses are here interpreted as W-simulta- neous with the bounded ones is again that they represent backgrounded infor- mation. Here too, nonbounded clauses that are interpreted inceptively form an ex- ception to the rule. In each of the following examples the nonbounded clause pushes the action forward because its situation is not interpreted as holding at the time of the situation referred to in the preceding clause: [There was a loud explosion.] Suddenly all kinds of things were flying through the air. [After a while the car left the main street and turned into a side street.] Shops, restaurants and beautiful houses were gone at once. [We only saw small grimy houses] and then suddenly we were outside the town. [This time her lover threw her out of his house.] There she was, in the middle of the narrow street, with a strange man staring at her curiously. [Bill asked where the children were.] Mabel looked anxiously at the playground. [The next thing I knew,] she was taking off her clothes. [I ran out of the room in panic.] V. The principle of unmarked temporal interpretation 431 8.41.3 It should be stressed that the Principle of Unmarked Temporal Inter- pretation set out in 8.41.1 stipulates only what is the unmarked temporal W- interpretation of sequences of unembedded clauses. As noted before, the un- marked interpretation may easily be ruled out by pragmatic factors. This is clear from examples like the following: John was in the kitchen. Suddenly Bill came in. John was in the kitchen. Suddenly he went out into the yard. In both cases we have a combination of an nonbounded clause and a bounded one. But whereas the first example yields the unmarked inclusion interpreta- tion, the second can only be interpreted in terms of sequence. This is because of pragmatic considerations: the W-simultaneity interpretation is ruled out by the simple fact that John could not be in two different places at the same time. In the following sequences (in which both sentences are bounded) the second sentence is interpreted as referring to the same action as the first, so that there is an interpretation in terms of W-simultaneity rather than in terms of temporal progression: Yes, I fired the gun. I shot the old devil through the head. [It was legitimate self- defence. I had no choice.] We threw a party last night and most of our friends got drunk. 8.41.4 Up to now we have identified three possible types of temporal W- relation in sequences consisting of two unembedded domain-establishing clauses (A and B): (a) situation B follows situation A; (b) the two situations coincide; (c) situation B includes situation A, or vice versa. However, there are still other (marked) possibilities, in which the Principle of Unmarked Temporal Interpretation set out in 8.41.1 is not observed. (Remember that we have stressed that the principle is merely a tendency, based on implicatures, not a strict rule.) (a) The situation of the second clause may sometimes be interpreted as W- anterior to that of the first clause, irrespective of the (non)bounded nature of the two clauses: My grandfather died at the age of 89. He was a very healthy man, who had a long walk every day. (The first clause is L-bounded, whereas the second and third are L-nonbounded. Instead of the unmarked reading, in which the first situation is included in the second and third, we get the interpretation in which the situation of the second and third clauses are W-anterior to the situation referred to in the first clause.) Yesterday afternoon my neighbour was arrested.Hestabbed his wife in the back. (The two L-bounded situations are not interpreted in terms of ‘iconic sequencing’. The second is interpreted as W-anterior to the first.) 432 8. Temporal domains and relative tenses: theoretical foundations Such a shift of domain (in the second clause) is only allowed if the correspond- ing unmarked interpretation is pragmatically impossible. For this reason it is relatively rare. To represent a situation as anterior to the situation of the previ- ous clause we normally use a relative tense (as in John could not take a taxi. He had gambled away all his money.) or add a time adverbial signalling anteriority. (b) In sequences consisting of more than two unembedded clauses, we may sometimes find a mixture of the three possibilities referred to in the Prin- ciple of Unmarked Temporal Interpretation set out in 8.41.1: We made a long journey through Italy. We visited Rome and Venice, and afterwards we went to Naples. We will make a long journey through Italy. We will visit Rome and Venice, and afterwards we will go to Naples. In each of these examples the two clauses in the second sentence establish domains of their own (with afterwards indicating the temporal order between them), but the situations they refer to are interpreted as W-simultaneous with that of the very first sentence. This is what typically happens when a clause referring to a situation with a complex structure is followed by several clauses describing subsituations. (c) In some sequences of more than two unembedded clauses the temporal order of the situations is open to interpretation (and is in fact irrelevant): [In that year several of John’s relatives had an accident.] Betty fell from the stairs. Bill drove into a lorry, and aunt Mildred got knocked over by a motorcycle. [Next year several of John’s relatives will have an accident.] Betty will fall from the stairs. Bill will drive into a lorry, and Aunt Mildred will get knocked over by a motorcycle. Here the various domains established by the clauses following the introductory sentence are presented as an unordered set. This is also the case in examples like the following: [“What did you do this morning?”] Ϫ “Betty did the washing-up, I fixed the car, and John planted the tulips.” [“What will you do this morning?”] Ϫ “Betty will do the washing-up, I will fix the car, and John will plant the tulips.” 8.41.5 In 8.41.2Ϫ4 we have identified all the kinds of W-relations that are logically possible between situations that are not bound to each other by the tense forms, i. e. which are allowed by absolute tense forms. In order to pick out the possibility that is intended by the speaker, the hearer can often be guided by (a) his pragmatic knowledge of the world, (b) the information pro- vided by temporal adverbs (if any), and (c) the nature of the text or discourse V. The principle of unmarked temporal interpretation 433 containing the sequence, which may or may not be a story with a linear tempo- ral structure. In the absence of these factors the interpretation of the temporal order of the situations will be in accordance with the ‘Principle of Unmarked Temporal Interpretation’ presented in 8.41.1. . the temporal order of the situations is either irrelevant or recoverable in some other way, e. g. from the use of time adverbials, from the order in which the situations are reported, from the linguistic. interpreted as following the situation of the previous clause. The first two of the above examples appear to be part of such a story with a linear temporal structure. The answers in the last two examples. Naples. In each of these examples the two clauses in the second sentence establish domains of their own (with afterwards indicating the temporal order between them), but the situations they refer