Departmental Bulletin Paper 否定辞繰り上げ述語と否定辞notの位置

森, 貞

It is well known that in English the entence like I {think/believe} he will come may be negated either as I don't {think/believe} he will come (main clause negation, henceforth MCN) or as I {think/believe} he won't come (subordinate clause negation, henceforth SCN) with no difference of cognitive meaning. Many linguists argue that MCN is strongly preferred to SCN. The computer-aided analysis of the CNN Transcripts Corpus, however, reveals that that is not the case when such verbs as suppose and suspect are used as the main clause predicate. It is claimed that the difference in the frequency of appearance mentioned above originates from the differences in the degrees of confi dence conveyed by think, believe, suppose, and suspect , all of which are NR predicates. It is also claimed that the functions of I {think/believe} [ ¬ p] are signifi cantly different from those of I {suppose/suspect} [ ¬p]; the former conveys the speaker's strong inclination toward [ ¬p] while the latter conveys the speaker's weak nclination toward [ ¬p], which is also conveyed by I guess [ ¬p].

Number of accesses :  

Other information