The experimentally backed and hitherto overlooked empirical observation of the paper is a contrast among indefinite Positive Polarity Items regarding their possibility of being rescued under certain operators with different rescuing potential. If/surprise/only/don’t think can rescue some-indefinites, suspending their anti-licensing (i.e., their impossibility to occur in the scope of a clausemate negation): while some-pronouns (in English and French) and des-indefinites in French exhibit the expected rescuability, English some-NPs remain unexpectedly degraded. Our account relies on the hypothesis that ‘rescuing’ is due to sentential negation being interpreted as ‘external’ (vs. nullified as in most literature). The definition we propose for external negation is syntactic: rescuing operators allow sentential negation to raise to an illocutionary functional projection above Tense Phrase (TP). Thus at LF (Logical Form), the negation takes that higher projection (rather than TP) as complement and becomes harmless for some-indefinites. The semantic correlate of this syntactic proposal is the interpretation of external negation as a propositional operator. As it involves the illocutionary periphery, rescuing is pragmatic in nature. The different rescuing potential between some-pronouns and some-NPs arises from the interplay between their distinct LF-representations and a minimal-event pragmatic constraint on rescuing.

This paper investigates nga-marked numerals in Albanian. They qualify as distributive numerals, since the presence of nga on the numeral yields a distributive reading of the sentences they belong to. Beyond their differences, most of the previous accounts rely on the hypothesis that distributive numerals introduce some kind of semantic feature, e.g. a covariation feature; an evaluation plurality requirement, also called a post-suppositional plurality requirement; or a distributivity force. Our main claim goes against this trend of thinking. We propose that distributive numerals do not carry any semantic feature but only a formal syntactic feature that needs to enter a syntactic dependency relation with a distributivity feature. The analysis is implemented in terms of Zeijlstra’s (2004) upward agree.