Ellipsis is one of the devices contributing to the cohesion of both written and spoken discourse. It is the omission of some elements from a clause whose meaning is understood even though they fail to be phonetically realized. This study aims to analyse the use of ellipsis as a cohesive device in Brendan Behan’s play The Hostage and in James Joyce’s composite novel Dubliners, trying to identify the similarities and differences with respect to this phenomenon in both genres. As a means to understand the results of this research, first it is essential to understand what ellipsis is, how it works and how it functions as a cohesive device. The quantitative analysis carried out in this study presents the number of elliptical structures occurring in The Hostage and Dubliners. These elliptical constructions are classified into three different varieties: nominal, predicate and clausal ellipsis. In the case of predicate and clausal ellipsis, each of their occurrences are further subcategorized. Predicate ellipsis is subcategorized into Verb Phrase Ellipsis (VPE), whereas clausal ellipsis is divided into occurrences of sluicing and fragment answers. This classification will be conducive to understanding similarities and differences between the different categories and subcategories in both literary works. A play and a narrative have been selected since it is expected there to be differences. The expectation is that ellipsis will be more profusely used in the play as the dialogues are closer to spoken language. On the other hand, it is likewise expected that in the composite novel fewer occurrences of ellipsis will be identified, though ellipsis will be more recurrent in the dialogues of the narrative work than in the narrated parts. The results obtained in this research have confirmed the cohesive function of ellipsis in written and spoken discourse, contributing to the uniformity of the text by avoiding repetitions or reducing clauses. Moreover, the results obtained show that the ellipsis phenomenon is more frequent in dialogical interaction than in narrative writing in these literary works.

34 p. -- Bibliogr.: p. 33-34