Time in Language

Wolfgang Klein

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781136151729
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

This book looks at the various ways in which time is reflected in natural language. All natural languages have developed a rich repetoire of devices to express time, but linguists have tended to concentrate on tense and aspect, rather than discourse principles. Klein considers the four main ways in which language expresses time - the verbal categories of tense and aspect; inherent lexical features of the verb; and various types of temporal adverbs. Klein looks at the interaction of these four devices and suggests new or partly new treatments of these devices to express temporality.

1 Introduction

1.1 Time and Finiteness

Time and space are the basic categories of our experience and our cognition, and without efficient communication about them, no well-coordinated collective action, hence no human society, would be possible. Therefore, all natural languages we know of have developed a rich repertoire of means to express temporality and spatiality. It has often been noted, however, that there is a certain asymmetry in the way language treats them: whereas the speaker is free to talk about space or not, this is not so for time: each finite verb obligatorily includes temporal information – it expresses tense, aspect, or both. Certainly, not all utterances must necessarily have a verb, let alone a finite verb. There are even some languages without finite verbs (Chinese, for example), although this is the exception rather than the rule. But whenever there is a finite verb, and this seems to be the normal case, the speaker has no choice: the expression of time is necessarily a consequence of the way in which language is structured.

Why this asymmetry? Time and space seem to be equally fundamental to cognition and experience, and there is no apparent reason why time should be deeper rooted in the structural organisation of language than space. In fact, this asymmetry has hardly ever been the subject of explicit discussion; but it has led to a two-fold bias in the linguistic research on the expression of time and space. First, there is a continuous and rich research tradition on temporality from Aristotle to our days, whereas spatiality, although not completely ignored, has never been a focal area of linguistic investigation until very recently. Second, research on temporality exhibits a strong bias towards its grammatical marking by the finite verb, notably on form and meaning of the two verb categories TENSE and ASPECT. There is also work on the inherent temporal characteristics of different verb types (state verbs, process verbs, event verbs, etc.), sometimes subsumed under the label ‘aspect’. But there is much less research on temporal adverbials or particles, although they are not only ubiquitous – not all languages have tense or aspect, but all languages have a wealth of temporal adverbials – they are also much more refined and richer in their expressive power. Their analysis is often considered to be a part of lexical semantics, whereas tense and aspect are deeply rooted in the structural organisation of language, and hence are more prone to excite the linguist’s attention.

Two and a half millennia of research have not led to a consistent, clear and generally accepted theory of tense and aspect (Binnick 1991); but on a somewhat global level, most linguists share a certain picture of what these categories are and how they function. The International Encyclopedia of Linguistics (1992, entry ‘Tense, aspect, and mood’) aptly summarises this ‘canonical view’ as follows:

TENSE refers to the grammatical expression of the time of the situation described in the proposition, relative to some other time. This other time may be the moment of speech: e.g., the PAST and FUTURE designate time before and after the moment of speech, respectively…. TENSE is expressed by inflections, by particles, or by auxiliaries in connection with the verb…. ASPECT is not relational like tense; rather, it designates the internal temporal organization of the situation described by the verb. The most common possibilities are PERFECTIVE, which indicates that the situation is to be viewed as a bounded whole, and IMPERFECTIVE, which in one way or another looks inside the temporal boundaries of the situation…. These aspects are usually expressed by inflections, auxiliaries, or particles.

In this informal characterisation of the ‘canonical view’, the notion of FINITENESS, although implicitly present, is not directly mentioned. I believe we can gain a better understanding of how time is reflected in language if this notion is given more profile. In a nutshell, the idea is as follows. Consider a simple utterance such as (1):

(1)	The light was on.

In (1), we can distinguish between a ‘finite’ and a ‘non-finite’ component, abbreviated FIN and INF, respectively. INF is a (selective) description of a possible situation; in this particular utterance, INF simply consists of the subject the light and the (non-finite) predicate be on. This selective description can, but need not, be made more specific by adding some temporal (or other) adverbial, such as for two hours, two times, or yesterday. It is less clear what the content of FIN is. In particular, it varies with sentence type. In normal declaratives, such as (1), the content of FIN is the ASSERTION made by the utterance, TENSE, in this example the simple past, does not serve to localise the situation in relation to the moment of speech, as is assumed in the ‘canonical view’. It rather imposes a temporal constraint on the assertion: it narrows down the assertion to some particular time. There are various ways to relate this FIN-time to INF-time, that is, to relate the time for which the claim is made to the time of the situation which is selectively described by INF; it is this relationship between FIN-time and INF-time which is traditionally described as ASPECT. Both tense and aspect can be defined in terms of temporal relations, such as before, after, simultaneous; they only differ in what is related to what. Under this perspective, the relation between finiteness, on the one hand, and tense and aspect, on the other, is not an accidental one: it is finiteness which intervenes and mediates between the moment of speech and the situation depicted. In the next section, I will try to motivate this idea.

1.2 Topic Time, Time of Utterance, and Time of Situation

If, on some occasion, an utterance such as (1) is made, then a distinction is to be made between the time at which the light was on, on the one hand, and the time for which such a claim is made, on the other. I shall call the latter the ‘topic time’ (TT) of that utterance, and the former the ‘time of situation’ (TSit). The situation is here the being-on of the light. This situation is selectively described by the non-finite component of (1). Hence, TSit is simply the time which corresponds to INF. The situation at issue could also be the eating of an apple by George, the being red of a pen, or the revolving of the sun around the earth. TT is the time which corresponds to the ‘finite component’ of the utterance. Both TT and TSit are different from the time at which the utterance is made, the time of utterance TU, and so is their relation to TU. In (1), TT clearly precedes TU. But this does not preclude that the light was still on at the time of utterance, that is, that TU is included in TSit, rather than preceding it. Hence, it is not TSit which is related to TU but TT.

We can imagine that (1) is given in answer to the question of a judge at court, addressed to a witness:

(2)	What did you notice when you looked into the room?

In this case, the judge’s question fixes a definite TT, and the witness is advised to talk about this time, and this time only. But his testimony is still truthful if the light was on before and after TT. He is not asked for the time at which the light was on, but for what was the case at exactly that particular TT, and no claim is to be made about any other time. In other words: TT is THE TIME SPAN TO WHICH THE SPEAKER’S CLAIM ON THIS OCCASION is CONFINED. This time span may be long – and then, the claim is quite unrestricted – but it may also be relatively short, as in this particular example.

Suppose now that the witness continues (1) by (3):

(3)	There was a book on the table. It was in Russian.

Again, it seems clear that the time of the book’s being on the table – the TSit of the first part of (3) – is not confined to the TT of that utterance, that is, to the time about which the witness is making claims. The witness does not want to say that the book was on the table just at TT, nor would we normally understand his utterance in this way. On the other hand, we would not assume that the book was there forever, that is, at any possible topic time one might want to speak about. Somehow, the time of the book’s being on the table before TT and after TT will be restricted, although nothing is explicitly said about this. It simply results from the kind of situation expressed here: being on the table is a TEMPORARY property of this book. We can imagine another topic time, say TT′, at which the book is not yet on the table, as well as a third topic time TT″, at which it is no longer on the table.

This is apparently different in the second utterance of (3). We would assume that the book was not just in Russian when the witness looked into the room: it was also in Russian before TT, and after TT; so far, there is no difference. But we would not assume that there is any other time the witness might want to speak about at which the book was NOT in Russian. Being in Russian is generally considered to be a PERMANENT property of books. Nevertheless, it would be inappropriate for the witness to say It is in Russian in this case, rather than It was in Russian. He would even stress this was, if someone, say the judge, questioned the claim: Are you sure? – Oh yes, it was in Russian. The letters were Cyrillic. What is stressed in this case is neither the fact that the book’s being in Russian is before TU, in contrast to the possibility that it might be after TU or simultaneous to TU, nor the relation of TT to TU. The fact that TT in this case is before TU is beyond any discussion; what is stressed is rather the claim that this situation obtains at the given TT. In other words: was is not in contrast to is or will be, but in contrast to was not.

Three lessons can be drawn from what has been said so far:

1. There is a distinction between two types of temporal relationships – TT to TU on the one hand, and TT to TSit on the other.
2. Tense marking applies to the relation between TT and TU, rather than to the relation between TSit and TU (TSit is not before TU in the case of the book’s being in Russian). It does not matter for tense whether the event, state, or process is before, at, or after the time of utterance. This is in remarkable contrast to widely held assumptions about tense.
3. Descriptions of situations may differ in their behaviour towards possible TTs, as has been illustrated above by the contrast between the book’s being on the table and the book’s being in Russian. Some situation descriptions, like the former, show a ‘TT-contrast’, others do not, and this is one of the factors that determine our assumptions about the relation between TSit and TU.

This last point will now be considered in more detail. In the case of the book’s being on the table, or the light’s being on, there may be many possible TTs before and after the TT at issue, which are also included in TSit (for example, the TT which is just one second before the witness looked into the room). But there are also many possible TTs for which this does not hold, unless we assume that the book was, is, and will be forever there, or the light was, is, and will be forever on. There is a contrast between the TT of the utterance and at least some other possible TTs. This is not so for a situation like the book’s being in Russian. Whatever TT we try to associate this situation with, TT falls within TSit (barring the case of language-changing books). When it is stated that this situation holds at some given IT, then this is not said in contrast to some other possible TT at which it might not hold.

Consider now those situations which have such a ‘TT-contrast’. In the two examples with TT-contrast considered so far, this is an ‘outside contrast’: in addition to many other possible TTs which fall inside TSit, there are still others which fall outside TSit. There are also situations with an ‘inside contrast’ between possible TTs, for example the book’s being taken away from the table. Such a situation includes, first, a number of possible TTs at which the book is on the table, and second, a number of possible TTs at which it is not on the table. A speaker, when linking such a situation to some TT, may choose one of the former, thus depicting the situation in its first state, or one of the latter; the speaker may also want to make a claim about a TT long enough as to include part of the first and part of the second state. Hence, we can distinguish (at least) three types of situation descriptions, which I will call ‘0-state’, ‘1-state’, and ‘2-state’, respectively. The first involves no TT-contrast at all, like the book’s being in Russian; the second involves an outside contrast, like the book’s being on the table; and the third involves in addition an inside contrast, like the book’s being first on the table and then not on the table, hence the book’s being taken away from the table.

These are, in a nutshell, the concepts needed to describe the way in which time is encoded in natural language. Essentially, I will make the following claims:

...

(4)

Tense concerns the relation between TT and TU. Aspect concerns the relation between TT and TSit – the way, or sometimes ways, in which some situation is hooked up to some TT. For the analysis of tense and aspect, it suffices to distinguish three types of ‘situation descriptions’, called here 0-state, 1-state, and 2-state. Adverb modification may apply to both TT and TSit.