Annotating thematic features in English and Spanish

Abstract

In this paper we present the preliminary results of an empirical study designed to test contrastive features of the category of Theme in English and Spanish through corpus analysis and manual annotation. Using as our theoretical basis the more general features of the model of thematisation proposed in Lavid, Arús and Zamorano (2010), the study describes the different steps of the methodology used, starting with the selection of the corpus used as a ‘training suite’, followed by the design of the annotation scheme, and ending with a discussion of the results of two annotation experiments carried out so far to test the reproducibility of the annotation scheme. It is expected that the work reported in this paper has a theoretical impact on the area of contrastive corpus studies and serves as the basis for the (semi)-automatic annotation of thematic features in larger bilingual corpora.