In English, the present perfect, unlike future, past, and non-finite perfects, cannot be modified by so-called ¡Æpositional¡Ç adverbials (Comrie 1976, McCoard 1978, a.o.). This phenomenon is known as the present perfect puzzle (Klein 1992).
(1)
a. *Alicia has danced on Monday / yesterday / at 10 o¡Çclock.
b. Alicia will have danced on Monday / at 10 o¡Çclock.
c. Alicia {had/must have} danced on Monday / yesterday / at 10 o¡Çclock.The prohibition against positional adverbials in the present perfect is not found in German (as seen in (4)), Dutch, Icelandic, or Italian. Notably, a present perfect morpho-syntax in these languages does not have the meaning of PAST , since it is compatible with present adverbials (Giorgi and Pianesi 1998, Musan 2001), a fact also illustrated in In English, the present perfect, unlike future, past, and non-finite perfects, cannot be modified by so-called ¡Æ positional¡Ç adverbials (Comrie 1976, McCoard 1978, a.o.). This phenomenon is known as the present perfect puzzle (Klein 1992).
(2)
a. *Alicia has danced on Monday / yesterday / at 10 o¡Çclock.
b. Alicia will have danced on Monday / at 10 o¡Çclock.
c. Alicia {had/must have} danced on Monday / yesterday / at 10 o¡Çclock.The prohibition against positional adverbials in the present perfect is not found in German (as seen in (3)), Dutch, Icelandic, or Italian. Notably, a present perfect morpho-syntax in these languages does not have the meaning of PAST , since it is compatible with present adverbials (Giorgi and Pianesi 1998, Musan 2001), a fact also illustrated in (3).
(3)
Hans ist {gestern um zehn / jetzt} weggegangen. (German)
Hans is yesterday at 10 now left (Musan 2001)
¡ÆHans has left yesterday at 10 / now.¡ÇThe puzzle has proved rather difficult to solve (see Dowty 1979, Klein 1992, Giorgi and Pianesi 1998, Kiparsky 2002, Katz 2003, Portner 2003, a.o.). Lack of space prevents us from discussing the previous accounts in any detail. We can only note here that none are without problems, and hence we consider the puzzle still unresolved.).
The puzzle has proved rather difficult to solve (see Dowty 1979, Klein 1992, Giorgi and Pianesi 1998, Kiparsky 2002, Katz 2003, Portner 2003, a.o.). Lack of space prevents us from discussing the previous accounts in any detail. We can only note here that none are without problems, and hence we consider the puzzle still unresolved.
We derive the present perfect puzzle by a new theory of semantic competition. The Perfect means the same in languages such as English and German. It denotes a time span that has no part located after the speech time but may be before the speech time or include it. The Present is different in English and German. In English, it denotes the speech time, in German a time not before the speech time. The Perfect operator is located at the place of the participle at d-structure. At LF it moves to the perfect auxiliary (if there is one). If the auxiliary is finite and located in the tense node, the complex tense Present + Perfect competes with the more specific Past in English. The result is that the Present Perfect has an Extended Now reading in this configuration and semantically excludes temporal frame adverbs specifying a past time. No such competition arises in German or for the Pluperfect. In certain other configurations of English, e.g., non-finite Perfects, the competition doesn't arise either.
1
We use capitalized regular font (e.g., Tense) for the syntactic category/node, small caps (e.g., PAST, PERFECT) to indicate the semantic feature merged in the syntactic structure, and lowercase font (e.g., past, perfect) for the morpho-syntactic realization of the semantic feature (e.g. a ./d/-suffixed verb, an auxiliary + past participle).
* Many thanks to Philippe Schlenker for extremely helpful discussions and ideas. Thanks also to theaudiences at NELS 34 at Stony Brook University, the University of Stuttgart, the University of Tubingen, the University of Texas, Austin, and UCLA.2
We use capitalized regular font (e.g., Tense) for the syntactic category/node, small caps (e.g., PAST, PERFECT) to indicate the semantic feature merged in the syntactic structure, and lowercase font (e.g., past, perfect) for the morpho-syntactic realization of the semantic feature (e.g. a ./d/-suffixed verb, an auxiliary + past participle).(4)
Hans ist {gestern um zehn / jetzt} weggegangen. German Hans is yesterday at 10 now left (Musan 2001)
¡ÆHans has left yesterday at 10 / now.¡Ç*The abstract can be downloaded in PDF : PanchevaStechowAbs.pdf