Issues in the Syntax of Polarity-Sensitive Items

This chapter provides an overview of PSIs (polarity-sensitive items) in Arabic and briefly outlines the critical issues in the syntax of PSIs in Arabic and their contributions to Arabic syntax and linguistic theory. It also outlines the theoretical underpinnings of research on Arabic negation, relying on the author’s most recent publications. The author synthesizes the major and crucial findings from cross-linguistic studies on this topic and studies of PSIs in Arabic. This chapter also articulates the critical issues, methodology, and scope of content. A quick overview of polarity sensitivity reveals much of the microvariation in Arabic. Geographically, this variation comparatively grows at the phonetic, morphological, and lexical levels, moving from the west with the Moroccan dialect, into Egyptian and Jordanian along the Mediterranean, into the Arabian Peninsula with Qatari Arabic, and into the formal written Arabic.

PSIs with Head-Like Properties

This chapter discusses two main issues that arise from PSIs (polarity-sensitive items) with head-like properties. These PSIs seem to be outside the (immediate) domain of their licensor. The first issue is how these PSIs are licensed in syntax and how a unified analysis can handle their distribution. The author argues that these PSIs are adverbial phrases that do not project a clausal projection and that negation licenses these PSIs either in Spec-NegP or under c-command. This unified analysis does not appeal to the problematic head–complement relation as a putative licensing configuration. Another issue that arises from these NPIs (negative polarity items) with head-like properties is their ability to host clitics with accusative and genitive case marking. This issue raises interesting questions pertaining to case theory and dependent case licensing. The author argues that negation licenses the puzzling accusative case of the pronominal complement, a conclusion with far-reaching implications to dependent case licensing in natural language.

Licensing Negative Concord Items

This chapter discusses the syntactic configurations and processes that are involved in licensing the various types of NCIs (negative concord items) in Arabic. The author argues for a revised syntactic agreement approach. In this way, the chapter explains the behavior of Arabic NCIs with extra mechanisms (positing an abstract negative operator) to explain non-strict NCIs by conceptualizing a revised inventory of abstract negation features that the various types of Arabic NCIs carry, thus partially building on and departing from analyses for Germanic languages. The coordinators laa-wala ‘neither-nor’ are conspicuous analogues of the NCI wala, which can combine with NPs to create person-NCIs such as wala-ħada ‘no body,’ in their ability to co-occur with the negative adverb maʕumrhiš ‘never’ in the CP (complementizer phrase) layers without triggering a double negation reading. A covert negative operator licenses the coordinators laa-wala as a last resort in the absence of overt licensors.

Summary and Conclusions

This chapter summarizes the central points from each previous chapter and elaborates on the individual analyses of these PSIs (polarity-sensitive items) and their contributions to the critical issues in syntax and linguistic theory. This book studied the microvariation in the syntax of PSIs (negative and positive polarity, negative concord, and negative indefinites) in Standard Arabic and the major regional dialects represented by Jordanian, Egyptian, Moroccan, and Qatari. Overall, the syntactic licensing of PSIs in Arabic bears on key theoretical issues in the cross-linguistic studies of polarity sensitivity. Such issues include the syntactic licensing configurations for these items, the feature of structure/specifications of these items, and the availability of syntactic agreement in the context of negation. Lastly, the chapter presents residual issues for future research.

Licensing Negative Polarity Items

This chapter concerns the role that syntax plays in licensing NPIs (negative polarity items). Scholars have argued for a semantic approach to characterizing the unifying properties of the wide range of licensors, such as negation, disjunction, interrogation, and subjunctives, licensors that all share the semantic notion of nonveridicality. The author extends this approach to Arabic NPIs in this chapter but also argues that syntax is heavily involved in licensing these items. In particular, the role of the categorial status, syntactic configurations, and syntactic processes such as movement in licensing NPIs are discussed. The author argues for a unified analysis with a minimal set of basic syntactic operations such as Merge and Move, and licensing configurations of c-command and Spec-Head (specifier-head). The analysis captures the true nature of perceived mutual exclusivity between NPIs and the enclitic negative marker as an epiphenomenon of the availability of multiple loci for negation.

Classification of PSIs and Their Lexical Categories

In chapter 2, the author lays out a classification of polarity-sensitive items (PSIs) and their lexical categories. PSIs include negative polarity items (NPIs), free-choice items (FCIs), positive polarity items (PPIs), and negative concord items (NCIs). General indefinites display different distributions than do NPIs. Indefinite nouns like ħada and iši function as NPIs, and they are distinct from indefinite nouns (general indefinites) that occur in the context of negation. This chapter discusses the distinctive features of NPIs and PPIs, such as scope widening. Two different types of PSIs interact with negation in interesting ways: NPIs and NCIs. One key difference between the two is that NPIs cannot function as fragment answers without negation and can occur in nonnegative contexts, such as interrogative and conditional contexts. NCIs display the opposite behavior. This chapter describes the distribution of the disjunctive particles walla/willa/ʔam ‘or’ and the negative counterpart wala ‘nor’ in polarity contexts and their status as structures for coordinate complexes in Arabic.