Existing approaches to the representation of argument structure in grammar tend to focus either on semantics or on syntax. Our goal in this paper is to strike the right balance between the two levels by proposing an analysis that maintains the independence of the syntactic and semantic aspects of argument structure, and, at the same time, captures the interplay between the two levels. Our proposal is set in the context of the development of a large-scale grammar of Modern Hebrew within the framework of Head-driven Phrase Structure Grammar (HPSG). Consequently, an additional challenge it faces is to reconcile two conflicting desiderata: to be both linguistically coherent and realistic in terms of the grammar engineering effort. We present a novel representation of argument structure that is fully implemented in HPSG, and demonstrate its many benefits to the coherence of our Hebrew grammar. We also highlight the additional dimensions of linguistic generalization that our proposal provides, which we believe are also applicable to grammars of other languages.

This paper focuses on the interaction between raising, subject–verb inversion and agreement in Modern Hebrew. It identifies, alongside ‘standard’ (i.e., English-like) subject-to-subject raising, two additional patterns where the embedded subject appears post-verbally. In one, the raising predicate exhibits long-distance agreement with the embedded subject, while in the other, a colloquial variant, it is marked with impersonal (3sm) agreement. The choice between the three raising constructions in the language is shown to be solely dependent on properties of the embedded clause. The data are discussed and analyzed against a background of typological and theoretical work on raising. The analysis, cast in the framework of Head-driven Phrase Structure Grammar (HPSG), builds on research on raising, selectional locality, agreement, subjecthood and information structure, as well as verb-initial constructions in Modern Hebrew.

Within the Head-driven Phrase Structure Grammar (HPSG) community, one part of the Base Hypothesis concerning free relatives proposed by Bresnan & Grimshaw (1978) has gained wide support, namely that free relatives are headed by the wh-phrase. The second part of the hypothesis is that the wh-phrase is base-generated, and this has not gained support. In this paper, we will consider a subset of free relative constructions, i.e. non-specific free relatives, and provide support for this second part, restated in HPSG terms as a claim that there is no filler–gap relation between a free relative pronoun filler and a gap in the sister clause of the free relative pronoun.

This paper gives an account of the event and argument structure of past participles, and the linking between argument structure and valence structure. It further accounts for how participles form perfect and passive constructions with auxiliaries. We assume that the same participle form is used in both types of construction. Our claim is that the valence structure of a past participle is predictable from its semantic type, and that the valence structure predicts which auxiliary a past participle combines with in perfect constructions and whether the past participle may occur in passive constructions. Our approach sets itself apart from similar approaches, cf. e.g. Heinz & Matiasek (1994), Kathol (1994), Pollard (1994) and Müller (2003), with its strong emphasis on semantics.