TENSE

Tense is a grammatical category by means of which some natural languages express the temporal location of the event described by the sentence in which the grammatical tense occurs. (This definition assumes a distinction between grammatical and lexical categories. For the technically inclined, lexical categories are part of the lexicon of a language and are open classes [classes that allow new vocabulary through compounding, derivation, coining, and borrowing]. They become inflected, and do not contract, affix, or cliticize. Examples of lexical categories are nouns, adjectives, adverbs, verbs. Grammatical categories are part of the grammatical system of a language and are closed classes [classes that do not allow additions]. They may contract, affix, or cliticize. Examples include inflectional and derivational morphemes and function words, such as prepositions, determiners, conjunctions, and pronouns.) An instance of a tensed language is English. In the English unembedded sentence "Bill called," the grammatical tense "-ed" conveys the information that Bill's call happened before the time of speech. Similarly, in the English sentence "Bill will call," the grammatical tense "will" contributes the information that Bill's call occurs after the time of speech. When a language does not have grammatical tenses, as in the case of Chinese, the temporal information may be conveyed by lexical categories, such as adverbs.

The mapping between the grammatical tenses of a natural language and the expression of temporal location is very complex, and one of the goals of linguistic semantics is to investigate the relation between grammatical tenses and the expression of time. To achieve this goal, scholars in both linguistics and philosophy have proposed different theories of tense.

One type of theory, beginning with the work of the logician Arthur Prior, analyzes tenses as temporal operators. Prior (1957, 1967) treated the past and future tenses as sentential operators meaning "it was the case that" and "it will be the case that," respectively. The sentence "Bill called" is translated into P (^p ) and is true in a world w at a time t if and only if "Bill calls" is true in w at a time t ′ t (^p is the intension of p, and "〈" means "earlier than"). In his intensional system, Montague (1974) adopted Prior's tense logic by introducing tense operators for the past and future tenses, with the time parameter of the intensional expression embedded in the tense operator.

A different approach to the analysis of tense is that proposed by Reichenbach (1947). According to Reichenbach, tense is not a temporal operator but a complex structure built from a small set of primitives: the event time (E ), the speech time (S ), the reference time (R ), and two relations that can hold between these times, simultaneity (symbolized with a comma) and anteriority (symbolized with an underscore). One of these relations holds between S and R, and one relation holds between R and E. The relation between S and E is not represented but is inferred from the first two. With this small set of primitives, Reichenbach was able to define the set of possible English tenses. For example, the simple past, future, and present tenses have the structures [E, R _S ], [S _R, E ], [S, R, E ], respectively. The contribution of R becomes crucial in the analysis of complex tenses, such as the future and past perfect (which Reichenbach called "anterior future" and "anterior past"), where R overlaps neither E nor S. For example, the past perfect in "At 3:00 p.m., John had (already) called" has the structure [E _R _S ], where the calling time E precedes the reference time R (3:00 p.m.), which in turn is before S.

The case of the future perfect is a little more complex. Take the sentence "By 3:00 p.m., John will have called." Our intuition is that, while 3:00 p.m. must follow the speech time, the time of John's calling must be before 3:00 p.m. but does not have to follow the speech time. The availability of the reference time R allows Reichenbach to account for this intuition easily: R must be future relative to S, and E must be past relative to R, but the relation between S and E is left unspecified, leaving open the following three possibilities: [S _E _R ] or [S, E _R ] or [E _S _R ].

A third family of theories views tenses as temporal predicates expressing relations between times (or events). Zagona (1995), Stowell (1996), and Higginbotham (2002) are the main proponents of this view. According to these authors, tenses express temporal relations, such as anteriority, posteriority, and simultaneity, between two events (or times). However, unlike Reichenbach's theory, events (or times) are not introduced by the tenses but by verbs and adjectives instead. This view is also different from the operator analysis of tense since tenses are not operators shifting evaluation parameters

The operator theory of tense has been very influential and has inspired semantic analyses where tense is an existential quantifier binding the time argument in the predicate. Versions of the quantificational theory of tense have been proposed by David Dowty (1979), Arnim von Stechow (1995), Toshi Ogihara (1996), and Dorit Abusch (1997), among others (see Kuhn and Portner 2002 for an overview on tense logics for natural languages). Barbara Partee (1973, 1984) has observed that existential-quantifier theories are problematic when we consider some occurrences of tense in natural language. Her famous example is

(1) I didn't turn off the stove

uttered as the speaker is driving down the freeway. According to the existential analysis of tense, the sentence can be interpreted either as "There is no past time at which I turned off the stove" or as "There is a past time at which I did not turn off the stove," depending on the scope of negation with respect to the temporal quantifier. However, neither interpretation correctly captures the meaning of the sentence in the context we are considering. Clearly, the speaker did not mean to negate the existence of any time at which she turned off the stove, nor did she mean to assert the existence of some time at which she did not turn off the stove. She merely meant to assert that she did not turn off the stove at a contextually salient past time.

To solve this problem, Partee proposed a referential analysis of tense, in which tenses are linguistic devises by which we refer to times salient in the previous discourse. This analysis treats English tenses analogously to how Hans Kamp (1981) and Irene Heim (1988) treated pronouns and nominal anaphora. Variants of this idea have also been proposed by Enç (1986), Heim (1994), and Kratzer (1998). However, there are occurrences of tenses that are not about particular times, as in the sentence

(2) Einstein visited Princeton

where a quantificational analysis of tense seems more apt. Both quantificational and referential theories of tense need to account for the occurrences of tense in (1) and (2). One possibility is to analyze the past tense as a restricted quantifier, just like ordinary nominal quantifiers. In this analysis, (1) would assert that there is no time within a contextually salient past interval at which the individual turned off the stove. The indeterminate reading of (2) would arise when the restriction of the quantifier is Einstein's entire life span.

The discussion so far has been about the meanings of the English tenses, and we have been silently assuming that there is a one-to-one correspondence between grammatical tenses and these meanings. While this is generally true in simple clauses, there are exceptions. For example, (3) illustrates a use of the grammatical present tense with the so-called futurate meaning.

(3) The 4:00 o'clock train leaves in five minutes.

Example (4) from Enç 1996 illustrates a mismatch between the future tense morpheme will and the semantics of the future, since Pat's sleeping is understood to be overlapping the speech time. Similarly, example (5) illustrates a mismatch between the past tense morpheme on was and the semantics of the past, since the past tense is allowed to occur with the future adverb tomorrow.

(4) Pat will be sleeping now.

(5) Pat was leaving tomorrow.

The idea of a one-to-one correspondence between tense morphology and tense meanings turns out to be even more problematic when we consider subordinate clauses and the phenomenon of sequence of tense. Consider the following sentence, where the matrix verb and the embedded verb both occur in the past tense.

(6) Bill thought that Sue was pregnant.

There are two possible readings of (6). According to the first reading, the content of Bill's thought was that Sue was pregnant at some time before the time at which Bill was having the thought. This is the so-called shifted reading. According to the second reading, the content of Bill's thought is that Sue was pregnant at the time when Bill was having the thought. This is the so-called simultaneous reading. The possible simultaneous reading, where the embedded past morpheme is not interpreted as a past tense, seems problematic for a theory in which the morpheme "-ed" is always interpreted as a semantic past. The simultaneity relation, generally expressed in English by the absence of either past or future morphemes, is expressed in sequence of tense with the past morpheme. Furthermore, notice that when we actually embed a grammatical present tense under a grammatical past tense, as in (7), we obtain not a simultaneous reading but yet a third reading, the so-called double-access reading. As pointed out by Enç (1987), in (7) Bill's thought is that Sue's pregnancy extends over a period of time including both the time at which Bill had the thought and the time at which (7) was uttered.

(7) Bill thought that Sue is pregnant.

Let us go back to (6). Operator theories of tense try to reconcile the occurrence of an embedded past-tense morpheme with the simultaneous reading by proposing accounts where, at the level of semantic interpretation, the embedded past tense is deleted (Ogihara 1989, 1995; von Stechow 1995) or is semantically bound by the matrix past tense (Abusch 1997) and its temporal features are deleted (von Stechow 2003).

Within the referential theories of tense, Enç (1987) proposed that the simultaneous reading of (6) is obtained when the embedded past tense is coindexed with the matrix past tense, and thus bound by it. Therefore, in her account, the embedded past tense refers to the past time referred to by the matrix past tense. Abusch (1988) points out that already in Kamp and Rohrer (1984) we can find some evidence against the claim that the morphological past tense in an embedded clause is interpreted as a semantic past tense. Abusch provides (8) as an example illustrating the fact that, the most embedded past tense, that associated with "were," cannot refer to any past time since, in the intended reading, the event of having their last meal together is understood as overlapping John's saying event.

(8) John decided a week ago that in ten days at breakfast he would say to his mother that they were having their last meal together.

Among referential theories of tense, a slightly different approach has been taken by Kratzer (1998). Kratzer's proposal, while inheriting several elements from Abusch's (1997) analysis of sequence of tense, is based on Irene Heim's observation that in some occurrences, pronouns have features that are not interpreted. For example, the second occurrence of "I" in Heim's example (9) is interpreted not as an indexical but as a bound variable in the so-called strict reading.

(9) Only I got a question that I understood.

According to Kratzer, the simultaneous reading of (6) arises when the embedded past tense is interpreted as a bound variable, just as the second occurrence of the first-person pronoun in (9) is interpreted as a bound variable, rather than as an indexical. The features on both the embedded "I" in (9) and the embedded past tense in (6) are not "interpretable" (in the sense of Chomsky 1995), that is, they do not contribute to the LF (logical form) representations of these sentences. They are zero pronouns, or zero tenses, whose morphological and phonological features probably derive from agreement with their antecedents and do not carry any semantic information. Kratzer's parallel between zero pronouns and sequence-of-tense tenses expands Partee's original insight about an analogy between pronouns and tenses. The parallel between pronouns and tenses is also at the center of recent work by Schlenker (2003) and von Stechow (2003).

The discussion of sequence-of-tense phenomena above has been concerned with sequences of tenses where th e matrix tense is a past. Hornstein (1990) claims that the availability of the simultaneous reading in sequence of tense is not restricted to the past tense but applies to all tenses. Enç (1996) challenges this claim on the basis of examples like (10), where, according to her judgment, only the shifted interpretation is possible:

(10) Mary will say that she will be tired.

Furthermore, Enç points out that the double-access reading is not forced by embedding the present tense under the future—a fact that thus sets the future tense apart from the past tense. In (11) the only reading is that Mary is upset at the time of John's assertion.

(11) John will say that Mary is upset.

On the basis of these asymmetries between the future and the past and on the basis of the observation that future-oriented modals behave like "will" with respect to sequence of tense, as in (12), Enç suggests that the future morpheme "will" is not a tense but a modal.

(12) John must claim that he is sick.

This last point raises the question of the relation between tense and two other grammatical categories: aspect and mood. Tense, aspect, and mood are intimately related, since they all contribute some information about the event that a given sentence is about: Tense, as mentioned, conveys information about the time of the event; aspect conveys information about the beginning, duration, completion, or repetition of the event; finally, mood conveys information about whether the sentence is about a possible or actual event. It is common to assume, however, that these categories are distinct, even though their boundaries are not always clear. (An example is the debate over the semantics of the present perfect in English and other languages. For a general overview of the topic, see Alexiadou, Rathert, and von Stechow 2003 and the references cited there.) Further comparative studies across Indo-European and non-Indo-European languages will, it is hoped, shed light on these intricate issues.

See also Artificial and Natural Languages; Chomsky, Noam; Intensional Transitive Verbs; Language; Montague, Richard; Prior, Arthur Norman; Quantifiers in Natural Language; Reichenbach, Hans; Semantics.

Bibliography

Abusch, Dorit. "Sequence of Tense and Temporal De Re." Linguistics and Philosophy 20 (1997): 1–50.

Abusch, Dorit. "Sequence of Tense, Intensionality, and Scope." In Proceedings of the Seventh West Coast Conference on Formal Linguistics, edited by Hagit Borer, 1–14. Stanford, CA: CSLI, 1988.

Alexiadou, Artemis, Monika Rathert, and Arnim von Stechow, eds. Perfect Explorations. Berlin: Mouton de Gruyter, 2003.

Chomsky, Noam. The Minimalist Program. Cambridge, MA: MIT Press, 1995.

Dowty, David. Word Meaning and Montague Grammar: The Semantics of Verbs and Times in Generative Grammar and Montague PTQ. Dordrecht, Netherlands: D. Reidel, 1979.

Enç, Mürvet. "Anchoring Conditions for Tense." Linguistic Inquiry 18 (1987): 633–657.

Enç, Mürvet. "Tense and Modality." In The Handbook of Contemporary Semantic Theory, edited by Shalom Lappin, 345–358. Oxford, U.K.: Blackwell, 1996.

Enç, Mürvet. "Towards a Referential Analysis of Temporal Expressions." Linguistics and Philosophy 9 (1986): 405–426.

Heim, Irene. "Comments on Abusch's Theory of Tense." In Ellipsis, Tense, and Questions, edited by Hans Kamp, 143–170. Amsterdam: University of Amsterdam, 1994.

Heim, Irene. "The Semantics of Definite and Indefinite Noun Phrases." PhD diss. University of Massachusetts, Amherst, 1982. Published by Garland, New York, 1988.

Higginbotham, James. "Why is Sequence of Tense Obligatory." In Logical Form and Language, edited by G. Preyer and G. Peter. New York: Oxford University Press, 2002.

Hornstein, Norbert. As Time Goes By: Tense and Universal Grammar. Cambridge, MA: MIT Press, 1990.

Kamp, Hans. "A Theory of Truth and Semantic Representation." In Formal Methods in the Study of Language, Part I, edited by J. Groenendijk, T. Janssen, and M. Stokhof, 277–322. Amsterdam: Mathematisch Centrum, 1981.

Kamp, Hans, and Christian Rohrer. "Indirect Discourse." Manuscript. University of Texas, Austin, and University of Stuttgart, 1984.

Kratzer, Angelika. "More Structural Analogies between Pronouns and Tenses." In Proceeding of Semantics and Linguistic Theory, VIII, edited by Devon Strolovitch and Aaron Lawson. Ithaca, NY: CLC Publications, 1998.

Kuhn, S., and P. Portner. "Tense and Time." In The Handbook of Philosophical Logic. 2nd ed., edited by D. Gabbay and F. Guenthner, 6: 277–346. Dordrecht, Netherlands: D. Reidel, 2002.

Montague, Richard. Formal Philosophy: Selected Papers of Richard Montague, edited by Richmond H. Thomason. New Haven, CT: Yale University Press, 1974.

Ogihara, Toshiyuki. "Double-Access Sentences and References to States." Natural Language Semantics 3 (1995): 177–210.

Ogihara, Toshiyuki. "Temporal Reference in English and Japanese." PhD diss. University of Texas at Austin, 1989.

Ogihara, Toshiyuki. Tense, Attitudes, and Scope. Dordrecht, Netherlands: Kluwer Academic Publishers, 1996.

Partee, Barbara. "Nominal and Temporal Anaphora." Linguistics and Philosophy 7 (1984): 243–286.

Partee, Barbara. "Some Structural Analogies between Tenses and Pronouns in English." Journal of Philosophy 18 (1973): 601–609.

Prior, Arthur N. Past, Present, and Future. Oxford, U.K.: Oxford University Press, 1967.

Prior, Arthur N. Time and Modality. Oxford, U.K.: Oxford University Press, 1957.

Reichenbach, Hans. Elements of Symbolic Logic. New York: Macmillan, 1947.

Schlenker, Philippe. "A Plea for Monsters." Linguistics and Philosophy 26 (2003): 29–120.

Stowell, Tim. "The Phrase Structure of Tense." In Phrase Structure and the Lexicon, edited by Johan Rooryck and Laurie Zaring, 277–291. Dordrecht, Netherlands: Kluwer, 1996.

Von Stechow, Arnim. "Feature Deletion under Semantic Binding: Tense, Person, and Mood under Verbal Quantifiers." In Proceedings of the North East Linguistic Society 33, edited by Makoto Kadowaki and Shigeto Kawahara. Amherst: GLSA, University of Massachusetts, 2003.

Von Stechow, Arnim. "On the Proper Treatment of Tense." In Proceeding of Semantics and Linguistic Theory, V, edited by Mandy Simons and Teresa Galloway. Ithaca, NY: CLC Publications, 1995.

Zagona, Karen. "Temporal Argument Structure: Configurational Elements of Construal." In Temporal Reference, Aspect, and Actionality. Vol. 1: Semantic and Syntactic Perspectives, edited by Pier Marco Bertinetto, Valentina Bianchi, James Higginbotham, and Mario Squartini. Turin, Italy: Rosenberg e Sellier, 1995.

Michela Ippolito (2005)

TENSE The grammatical category, expressed in forms of the VERB, that locates a situation in time. In English, tense must be expressed in all finite verb phrases. It is marked by the choice of the first or only verb in the verb phrase: play versus played; has played versus had played; will play versus would play; is playing versus was playing. Since contrasts in number and person, where they apply, are also marked on the first or only verb, these choices combine with tense: present I/They play versus present She plays; present I am/She is versus present We/They are; past I/She was versus past We/They were. By definition, non-finite verb phrases do not have tense marking. There is also no tense choice for the imperative (‘Play harder’) or the subjunctive (‘We insisted that he play harder’).

Tense in English

In terms of morphology, English has only two tenses, the present or non-past (take/takes) and the past (took). The paradigm is extended by the use of the auxiliaries be and have: be followed by the present participle forms the progressive or continuous (is taking); have followed by the past participle forms the perfect (has taken). Although these are traditionally known as tenses, recent terminology refers to them as aspects (such as progressive aspect) and (for the perfect) phase. All three features can be combined: had been taking is past, progressive, and perfect. The passive voice is formed within the same paradigm, by be followed by the past participle, but is not a tense. The sequence of the auxiliaries is fixed: have + be + present participle, be + past participle, with the full verb in final position and a MODAL VERB preceding all other auxiliaries: may have been taken.

The simple present

With dynamic verbs, this tense expresses habitual activity and ‘timeless truths’: He goes to London every day; Water boils at 100 Celsius. In commentaries, demonstrations, and performatives, it serves to report events simultaneous with the speech event: He passes the ball to Smith, and Smith scores; I take three eggs and beat them in this basin; I name this ship ‘Fearless’. With static verbs, it refers to a present or timeless state: It contains sugar; Air consists of oxygen and other gases. With private verbs (of sensation, mental processes, etc.) it expresses how things are: I smell something burning; I think he'll come. In statements about the future, it shows that events have been arranged: We fly to Paris tomorrow. In literature and conversation, as the historic present, it reports past events dramatically and dynamically: He comes up to me and says … With verbs of communication, it states or informs: The Bible says …; John tells me that he is going to Spain.

The simple past

Generally, this tense refers to events, habitual activities, and states in the past: I talked to my brother this morning; The Normans conquered England in 1066; He went to London every day; It contained sugar. In the ‘sequence of tenses’ rule in reported speech, it restates the present tense of the original utterance: ‘He likes chocolate’ as reported in She said he liked chocolate. However, the present tense may be retained if the state of affairs being reported is covered by the time of speaking: John said he likes chocolate. It is used to express unreality, especially in unreal conditional sentences (If John came, Mary would leave; compare If John comes, Mary will leave), with wishes and recommendations, etc. (I wish I knew; It's time we went), and for tentativeness or politeness (Did you want to talk to me?). This accounts for some of the uses of the modal forms might, could, and would, as in: Might they want to see her?; Would you like us to come?

The progressive

The present progressive is most commonly used to indicate an event in progress at the time of speaking: He's reading a book. With the past progressive, the time of the continuous event is often explicitly shown to overlap a point of time or another briefer event: I was reading at ten o'clock/when he arrived. In contrast, the simple past would suggest that the event was subsequent to the point of time or other event: When he arrived, I left. In standard English, static and private verbs are non-progressive, in that they do not usually occur in the progressive, the simple present being used instead (not *I am loving you, not *I'm thinking he will come, although such usages occur in varieties of IndE and PakE). There are a number of verbs with inherent duration which may be used in the non-progressive form, even if the duration is clearly indicated: I worked all morning; She slept for eight hours. The progressive may indicate: (1) Incompletion: I was painting the house this morning versus I painted the house this morning. (2) Simple futurity, especially with verbs of motion: I'm flying to Paris tomorrow. (3) Limited duration of habitual activities or with non-progressive verbs: We're eating more meat now; We're living in London these days (compare We live in London). (4) Sporadic repetition: My car's always breaking down versus My car always breaks down when I forget to service it.

The perfect

The non-progressive perfect refers to an event in the past with current relevance: I've broken the window indicates that I broke the window and that the window is probably still broken; I've seen John might suggest that I have told him what I intended to, or that he is now nearby. It is also used with just for events in the immediate past: I've just seen him. The progressive perfect relates to activity beginning in the past and continuing up to the present, or, for past-tense forms, to a point of time in the past: I've been reading for two hours; I'd been reading for two hours when he arrived. It may also indicate continuous activity in the past with current relevance: Someone's been moving my books—they are no longer where I left them. The present perfect is not normally used with past-time adverbials: not *I've broken the window yesterday. The simple past is often used in AmE where BrE uses the perfect: (1) BrE Have you washed your hands?, AmE Did you wash your hands? (2) BrE Have you done it yet?, AmE Did you do it yet? There is, however, wide variation in the use or non-use of the perfect in AmE.

The future

Traditionally, grammarians have taught that English has a future tense formed with shall and will, shall being used with first-person subjects (I shall be happy to see her) and will with the others (She will be happy to come). However, will is also commonly used with first person subjects (I will be happy to see her) to indicate futurity, though conversely shall is not used in the same way with the other persons (not *She shall be happy to come). The view that will and shall mark the future tense is widely held and often strongly asserted, but there are three arguments against it: (1) Morphologically, there are only two tenses, present and past; to talk about the future tense is to confuse time marking with grammatical tense. (2) Will and shall are formally modal verbs, and should be handled in the modal system, not the tense system. (3) Be going to is as good a candidate for the marker of the future tense as will and shall.

In the majority of instances, will and shall express a conditional future and are the forms used in the apodosis of future conditionals (the part without if): If you ask them, they will do it. Be going to indicates an envisaged progression towards a future event: It'll cost me a lot of money may imply ‘if I buy it’, whereas It's going to cost me a lot of money suggests that the decision to buy has been made. There are two arguments in favour of treating will and shall as markers of the future tense: (1) Future tenses in other languages also often express conditional futures. This is not unexpected, since the future is not factually known as the present and past are, and it is not surprising, therefore, if the future tense in English is marked by modal-type verbs. (2) Will and shall function in some ways more like tense markers than modal verbs, particularly in that they cannot be marked independently from the main verb for negation, as most modal verbs can. Thus, there is only one negative of You will see him tomorrow (You will not/won't see him tomorrow), as there is only one negative of You saw him yesterday (You did not/didn't see him yesterday), but there are two negatives of He may be in his office: (1) He can't be in his office (It is not possible that he is in his office). (2) He may not be in his office (It is possible that he is not in his office).

tense¹ / tens/ • adj. (esp. of a muscle or someone's body) stretched tight or rigid: she tried to relax her tense muscles. ∎  (of a person) unable to relax because of nervousness, anxiety, or stimulation: he was tense with excitement. ∎  (of a situation, event, etc.) causing or showing anxiety and nervousness: relations between the two neighboring states had been tense in recent years. ∎  Phonet. (of a speech sound, esp. a vowel) pronounced with the vocal muscles stretched tight.The opposite of lax.• v. [intr.] become tense, typically through anxiety or nervousness: her body tensed up. ∎  [tr.] make (a muscle or one's body) tight or rigid: carefully stretch and then tense your muscles.DERIVATIVES: tense·ly adv.tense·ness n.ten·si·ty / ˈtensitē/ n. ( dated ).tense² • n. Gram. a set of forms taken by a verb to indicate the time (and sometimes also the continuance or completeness) of the action in relation to the time of the utterance: the past tense.DERIVATIVES: tense·less adj.

tense2 drawn tight XVII; highly strung XIX. — L. tensus, pp. of tendere stretch, TEND2.
So tensile ductile XVII; pert. to tension XIX. — medL. tension XVI. — F. or L. Hence tensor (-OR1) (anat.) muscle that tightens some part XVIII; (math.) in quaternions XIX.

tense1 †time; (gram.) form of a verb indicating time. XIV. — OF. tens (mod. temps) :- L. tempus time.