Maria Persson, Lund University

VEGA consists of a study in the syntax of the finite and infinite verb forms of Gulf Arabic dialects. The aim has been to survey functions and areas of use for each of the main verb forms and to distinguish what other tools the dialects use to express tense, mood and aspect (TMA). Among other things, this meant exploring the use of auxiliary verbs and the correlation between main clause and subordinate clause predicates in various types of clause junctures. For the purpose of the project, a large database of authentic speech was collected from both the capital cities in the Gulf littoral and historically more isolated areas of Oman and the UAE. This dialectal area's encoding of tense/mood/aspect was of particular interest from a typological point of view since a process of increased grammaticalization had been observed in previous studies of individual dialects in the area. Hence, the survey of TMA markers in Gulf Arabic contributed to our knowledge of similar grammaticalization processes in languages in general. Gulf Arabic dialects were defined as the dialects spoken in Bahrain, Iraq (littoral), Kuwait, Oman, Qatar, Eastern Saudi Arabia and the UAE. Data was compiled from all these areas except Iraq where the political conditions did not permit research trips into the area.

Originally, the aim was also to compare the dialectal varieties included in the database with one another. Specifically, the dialects of the major urban centres were to be compared with rural varieties in the inland of Oman. However, an examination of samples collected from various parts of the area did not display such noticeable syntactic differences as to motivate a more systematic comparison between parts of the collected data within the scope and time frame of the project. Instead, a comparison regarding parts of the syntax was made between the total VEGA database and another database collected from a partly nomadizing population in Eastern Oman. Moreover, the VEGA database material was compiled and classified with such care as to easily allow for comparisons between various dialectal areas represented in the database within future projects.
Scholars differ in view as to whether the Arabic verb in one or all of its varieties signify tense or aspect or a mix of the two (Holes 2004; Ingham 1994; Comrie 1976). One important result of VEGA is to establish that the suffix form and the prefix form of the urban Gulf Arabic verb have aspectual values and that the active participle is tenseless and aspectless. The prefix form is imperfective, whereas the suffix form is perfective. The active participle describes a state. Furthermore, these rather clear cut differences in aspectual values, especially between the prefix form and the suffix form, are employed in various ways to signal hierarchy on the level of clause structure as well as discourse structure without a need to take resort to conjunctions or discourse markers. One of the goals of VEGA was to compare the dialects of urban populations with the dialects of more rural, nomadizing Gulf Arab communities. This was mainly accomplished in the study of the active participle where the VEGA database consisting of urban Gulf dialects was compared with data collected by Dr. Domenyk Eades among nomadizing people of the Sharqiyya region of Oman (Eades and Persson forthcoming). Persson and Eades have both found the comparison of their respective databases most promising and, hence, aim to continue their collaboration with research on other areas of verb syntax. This, then, is one area where the project has generated data and raised questions to be dealt with in further research.

A second important finding, and a corollary of the previous one, is the discovery of extensive use of asyndetic clause combining in Gulf Arabic. The article describing these structures (Persson forthcoming) must be regarded as the most important of the publications that have resulted from the project. This article covers the main goal of the project: to delineate the general use in Gulf Arabic of different verb forms and combinations of verb forms to express various facets of tense, mood and aspect and to discover areas of on-going grammaticalization processes. Asyndetic verb combinations are prime environments for grammaticalization processes whereby modal and aspectual modifiers develop through gradual grammaticalization of the first verb of two or more in a series. Several Gulf Arabic auxiliaries and verbs that may be under way of becoming auxiliaries were described in the article. Nevertheless, the degree of grammaticalization of these modifiers in the present Gulf Arabic corpus appeared to be less than what we find in descriptions of, for example, Egyptian and Levantine dialects. In addition to modal and aspectual modifiers, asyndetically juxtaposed verbs were found to function as circumstantial and/or adverbial modifiers. Furthermore, the study disclosed that they are used to serve as a verb phrase parallel to the well-known phenomenon of parallelism in the use of nominals in Arabic. The research on asyndetic clause combining in Gulf Arabic, and the discovery of a number of grammaticalization processes, raised several questions for further research. Firstly, it was found that the same structure is used for a variety of semantic functions and the intention of the language user in chaining verbs must often be inferred from context. Secondly, the processes of semantic bleaching of verbs, and the resulting spectrum of grammaticalization of particular verbs into verbal modifiers and particles, were found to have gone further in some other dialects, such as Egyptian and Levantine dialects, than in Gulf Arabic. However, due to lack of extensive databases from different dialects and, hence, lack of diachronic and/or synchronic comparative studies, the origin of preverbal particles is often unknown. From typological studies in general linguistics we know that some semantic categories of verbs and some syntactic structures are more prone to grammaticalize than others. By studying dialects where many of these verbs are still used in their full form, or where the process has only just started, we may learn more about the origin of modal, adverbial and aspectual particles in other dialects. This points to a need to detect similarities and differences between major Arabic dialect areas in the functions assumed by strings of juxtaposed verbs and processes of grammaticalization that appear in these. For this, databases similar to the VEGA recordings need to be collected from a number of hitherto uncharted dialects. Simultaneously, older text and available, fragmentary, data from previous studies need to be analyzed and presented in a coherent format that allows for comparisons.
The third most important result from the project was the detection of a distinctive mood marker: the b-prefix presented in the second most important project publication (Persson 2008). Quite different from the commonly known b-prefix in Egyptian and Levantine dialects, the Gulf Arabic b-prefix has been assumed to signal future tense and/or intent (Johnstone 1967; Brockett 1985; Al-Ma?t?q 1986; Brustad 2000; Holes 2004). However, a thorough analysis of the occurrences of this b-prefix in the present project clearly showed that it cannot be said to be a marker of future tense per se. Rather, the b-prefix is used for assumptions about a world that has not been experienced (futures, conditionals) or generalisations about past tendencies (habitual pasts). In other words, the basic function of this b-prefix in urban Gulf Arabic is as a marker of notionally irrealis categories. This, in turn, may - but need not - coincide with future temporal and/or intentional modal values. However, neither future tense nor intentionality are signalled by the b-prefix itself. The results of the study do not imply that Gulf Arabic has a binary system for marking the realis/irrealis distinction. Nevertheless, it was established that (urban) Gulf Arabic has a marker that can be used to mark some events as "purely within the realm of thought, knowable only through imagination" (Mithun 1999, 173). The survey also documented that another particle is rather widely used for future tense, but also at times for irrealis contexts, namely the particle râH/Ha. This particle has received little attention following Johnstone's descriptions from 1967. On the other hand, and in accordance with Holes (2004, 274 n.29), the verb(s) bagha/yabi, that have been mentioned as indicators of future time in the nearby Najdi dialect (Ingham 1994, 120f.) were, in the present database, found to be used only for the volitive, i.e. to express a wish but not at all to signal future time-reference.

In addition to published articles, results from VEGA were presented at several linguistic and Semitic research seminars in Lund, Uppsala and Gothenburg. Nine conference papers were read at international conferences on general Linguistics, Arabic dialectology and Semitic languages both in Europe and in the Arabian Gulf. A guest lecture presenting early results was held at the invitation from the United Arab Emirates University, Al-Ain, UAE already in 2006. A popular presentation of some of the results was given at the invitation of a local organization for retired people in Skövde, Sweden towards the end of the project time. Three other popular presentations related to the field work within the project were given in 2006 and 2008.