Topic: HUMAN SEMANTICS
Roumyana Pancheva has a paper on the present perfect tense puzzle, a linguistic phenomenon:
Another Perfect Puzzle
The interaction of the perfect with temporal adverbials is the domain of
the well-known present perfect puzzle (Klein 1992) - the fact that certain
adverbials are prohibited with the present perfect in English (though not
some other languages) while acceptable with non-present perfects. As is
generally agreed, the prohibition is against past specific adverbials (cf.
Heny 1982, Klein 1992, Giorgi and Pianesi 1998, a.o.).
This paper adds yet another puzzle to the area of perfect-adverbial
interactions. It establishes a new generalization regarding the modification
of perfects by both past and non-past specific temporal adverbials. The
puzzling facts are illustrated in (1).(1) a. ?? We saw John last night. He had arrived yesterday...
b. We saw John this morning. He had arrived yesterday...
c. We saw John last night. He had arrived the same day...
Adverbials like yesterday are allowed in past perfects, and they may specify
the time of the event, as in (1b).However, their presence is restricted,
depending on what the reference time in the past perfect is. The reference
time is the interval which tenses relate to the speech time, and which the
event time is situated relative to. In the case of the past perfects in (1), the
reference time is a past interval anaphoric to the reference time of the
preceding past sentences: last night in (1a) vs. this morning in (1b). The
choice of a reference time contained in the interval denoted by the adverbial
modifying the perfect results in degraded acceptability, as in (1a).
When the reference time in the past perfect is not contained in the denotation
of the adverbial modifying the perfect, the result is an acceptable sentence, as
in (1b).
Read it
Posted by Tony Marmo
at 07:24 BST
Updated: Monday, 9 August 2004 08:11 BST
