| old_uid | 3430 |
|---|
| title | Tense parameters |
|---|
| start_date | 2007/11/12 |
|---|
| schedule | 10h |
|---|
| online | no |
|---|
| summary | I begin by rehearsing the account of English Tense, from Higginbotham (2002) and (2006):
(0) The actual time of utterance is default in root clauses.
(I) Tenses are binary, expressing one of the three relations ≈,<,or > between (the actual times of) events or situations, where '≈' is the relation between e and e' that holds when the time _(e) of e surrounds the time _(e') of e'.
(II) Anaphoric +Past is ambiguous (in English) between (a) facilitating anaphora, but having a -Past interpretation (B-past), and (b) expressing <(A-Past). The antecedent of a B-Past must be +Past, and the clause itselfmust be Stative.
(III) -Past in situ cannot be anaphoric to +Past.
(IV) Tenses in the C position of a complement clause are always anaphoric; movement of one copy of INFL to C is obligatory in these cases.
The theory applies both to the simple inflectional and paraphrastic tenses and to extensions to the English Perfect and the Progressive. It follows from (IV) that Sequence of Tense is obligatory in complement clauses, but not in relative clauses. The phenomena of English “double access” likewise fall out, as a joint consequence of (III) and (IV). There are a number of languages in which the forced double access interpretation, as in the well-known example (1), does not occur:
(1) John said that Mary is pregnant.
That is to say, in these languages the analogue of (1) means (and must mean) merely that John said that there was such a thing as Mary's being pregnant at the time of his, John's, speaking. Moreover, in some of these languages (e.g., Korean, Romanian) the complement +Past can be only a true Past; i.e., an A-Past in the sense of (II) above. Sharvit (2003) gives examples from Modern Greek as showing that, however common, this pattern is not forced: expressed in terms of my outline above, it may be that (III) fails whereas (II) holds, as it does in English. She fails to note, however, the requirement of stativity for the A-Past as in (II), a point that will form part of my theme here.
Double access appears in Italian and French with Imperfect complements, not manifestly represented in English. I shall assume for purposes of this discussion that an analogue of (III) is at work in these languages.
Tense is an indexical, but not a demonstrative, feature: the interpretation of an utterance using a Tense depends, as a strict matter of linguistic form, upon a contextual parameter, namely the time of utterance. In contrast to other temporal indexicals (e.g., now, a year ago), however, embedded Tenses sometimes do not, and sometimes cannot, behave as hey would in isolation, in root clauses. The reason, as suggested by the above outline, is not that they are in any way exceptions to the thesis that indexicals cannot shift their interpretations when embedded, but rather than they participate in anaphoric interclausal relations. The same, I conjecture, should be said for other cases where context-dependent elements shift their interpretations when embedded. I will illustrate with reference to (what I am told in Byun (2007) about) Korean cikum, which appears to be interpreted as now in root clauses, but then in some complements.
Sharvit (2003) asks the important question whether the condition (II) (permitting at least some occurrences of the B-Past) and condition (III) (forcing double access) can vary independently. She proposes to rule out languages in which (III) holds as in English, but the complement Past is never a B-Past. I will argue that her proposal is defective, for several reasons. In any case, the very thing that she proposes not to permit actually occurs in English, with non-Stative Past in the complement of a Past, as in (2):
(2) John said that Mary left the room.
The complement Past in (2) is only a A-Past, so there is no ambiguity. The reason this is so cannot be that the English Present of an achievement predicate is only "habitual," since (as noted in Giorgi and Pianesi (1997)) the same happens in Italian, where the simple Present can in fact be a report of what is presently occurring. A good test case for English is (3):
(3) John said that Mary remained in her room.
The V remain is about as non-active an "Activity" V as one can get: one can certainly say (4) in response to the question, "Where is Mary?"
(4) Mary remains in her room.
(not surprisingly, given the historical, and the obvious semantic, link between remain, become, and be). However, (3) can only be interpreted as a report of an alleged past, Past-Tense, utterance by John. Inversely, (5) shows double access, as expected:
(5) John said that Mary remains in her room. |
|---|
| responsibles | Copley |
|---|
| |