Jordanian and Egyptian negative markers

This section explores the explanatory power of the multi-locus analysis of negation in accounting for the distributional contrasts between two negation strategies in Jordanian Arabic (JA) and Egyptian Arabic (EA). These two strategies are the use of the discontinuous marker ma…š and the non-discontinuous marker miš. I claim the distribution of these strategies follows from analyzing bipartite negation as a low negation occupying a position below TP. Following Benmamoun (2000), I assume that the discontinuous pattern is a result of predicate merger with negation in which the predicate undergoes head movement to the negative head for feature checking. I argue that this movement is tied to the person feature of the predicate, giving rise to discontinuous negation. I also depart from Benmamoun's (2000) idea that with non-discontinuous negation the subject NP fulfills the checking requirement of the negative head. Instead, I argue that a covert copular head fulfills the checking requirement as a last-resort mechanism when the predicate lacks the person feature.

I also argue that EA does not have this checking requirement and that, aside from with perfective verbs, predicate merger with negation is an optional post-syntactic operation. Assuming that perfective verbs undergo V-to-T movement (Benmamoun 2000; Soltan 2007), the obligatory merger between perfective verbs and negation is due to minimality constraints. Specifically, the verb must move into the negation head on its way up to the tense head to avoid minimality violations. All other contexts in EA, unlike in JA, tolerate non-discontinuous negation (cf. with imperfective verbs, with fronted prepositional predicates).