Search

This paper provides evidence that medial adjunct PPs in English are possible. On the basis of corpus data, it is shown that sentence-medial adjunct PPs are not unacceptable and are attested. Our corpus data also reveal a sharp asymmetry between negative and non-negative adjunct PPs. The analysis of the corpus revealed the following pattern: Non-negative adjunct PPs such as at that time resist medial position and instead tend to be postverbal; negative adjunct PPs such as at no time appear medially rather than postverbally. In the second part of the paper, we broaden the empirical domain and include negative complement PPs in the discussion. It is shown that when it comes to the licensing of question tags, English negative complement PPs, which are postverbal, pattern differently from postverbal negative adjunct PPs. That is, sentences with a postverbal negative adjunct PP pattern with negative sentences in taking a positive question tag, while sentences containing a postverbal negative argument PP pattern with affirmative sentences in taking a negative tag. To account for the observed adjunct-argument asymmetry in the licensing of question tags, we propose that clauses are typed for polarity and we explore the hypothesis that a polarity head in the left periphery of the clause is crucially involved in the licensing of sentential negation. [ABSTRACT FROM AUTHOR]