The paper studies the nature of the interplay between viewpoint aspect and tense in the temporal systems of Modern Hebrew and Biblical Hebrew, focusing on the suffixed verb form qaṭal and on the participial verb form qoṭel common to both. The examination of these temporal systems is couched in a neo-reichenbachian framework, taking the categories of viewpoint aspect and tense to consist of relations between temporal intervals (also known as the “two dimensional theory of tense”). It follows from this framework that viewpoint aspect and tense are encoded in any given temporal system, and languages differ in whether they contrastively mark tenses and/or aspects. When only one of the categories is morphologically marked, it is assumed that the other marks a default value. In the current context, it will be shown that viewpoint aspect properties of the verbal forms remain essentially unchanged over the two periods: qoṭel expressed, and still does, imperfective viewpoint aspect, with some characteristics of a progressive, whereas qaṭal expressed, and still does, a default viewpoint aspect that is interpreted as perfective according to the lexical aspectual properties of the underlying VP. In contrast, the properties of the tense categories in the systems have undergone a significant change: in Biblical Hebrew, the category of tense, in these particular forms, but possibly also more generally in the system, mark only a default value where the two temporal intervals, R and S, are not ordered in any specific manner, but rather present a temporal overlap; in Modern Hebrew, the category of tense contrasts temporal values, where R and S are clearly ordered with respect to each other: overlap, precedence. It will be suggested that this state of affairs is correlated with the (im)possibility of the forms to express narrative progression. Given that narrative progression involves update of R, this is not possible when the temporal relation between R and S is one of general overlap, blocking R progression. In Biblical Hebrew, narrative progression is achieved via the sequential w-forms, which have later disappeared from the system.
In this paper we will provide a description of what we term here the Coreferential Dative Construction. The languages under consideration are Syrian Arabic, which has never been studied before from this respect, and Modern Hebrew. We will show that this construction, related to other constructions containing non-selected datives, expresses the speaker's stance or emotional attitude towards the described eventuality by seeing it as having weak relevance. We will also show that the most important grammatical difference between the two languages is that in Syrian Arabic the presence of the Coreferential Dative obligatorily triggers a special type of modification in the VP: it must be modified by an attenuative vague measure. The comparative approach will help to shed new light on previous analyses of Modern Hebrew data.
In this paper we describe the pragmatic, lexical and syntactic properties of the Syrian Arabic Coreferential Dative Construction (CDC), featuring a dative element bearing agreement features which are identical to those of the subject in the clause, the Coreferential Dative (CD), and an obligatory expression of attenuative vague measure, described by us in Al-Zahre & Boneh (2010). We first show that the CD, which has no truth conditional meaning, contributes to the creation of a Conventional Implicature (Horn 2004, Potts 2005). Second, we propose a way to compositionally integrate the CD into the derivation of these constructions by arguing that the visible pronominal features are non-referential but rather the morphological reflex of checked uninterpretable phi-features on a defective applicative head. To couch the analysis in a wider context, we show how it can extend to other categories of non-core dative in Syrian Arabic.