Across Languages and Genres: Creating a Universal Annotation Scheme for Textual Relations

The present paper describes an attempt to create an interoperable scheme using existing annotations of textual phenomena across languages and genres including non-canonical ones. Such a kind of analysis requires annotated multilingual resources which are costly. Therefore, we make use of annotations already available in the resources for English, German and Czech. As the annotations in these corpora are based on different conceptual and methodological backgrounds, we need an interoperable scheme that covers existing categories and at the same time allows a comparison of the resources. In this paper, we describe how this interoperable scheme was created and which problematic cases we had to consider. The resulting scheme is supposed to be applied in the future to explore contrasts between the three languages under analysis, for which we expect the greatest differences in the degree of variation between non-canonical and canonical language. 1 Aims and Motivation The aim of the present study is to create a scheme which will allow us to use existing annotations of textual phenomena, and which will be applicable to multiple languages and genres, including noncanonical ones. The annotations were created within two separate projects: German-English Contrasts in Cohesion (GECCo, Lapshinova and Kunz (2014)) whose focus was on English and German on the one hand, and the Prague Dependency Treebank (PDT 3.0, Bejček et al. (2013)) with the analysis of Czech, on the other hand. The resulting scheme will serve our overarching goal to unify the two approaches in a joint analysis of contrasts between English, German and Czech on the level of discourse. We assume that the greatest differences between these languages lie in the degree of variation between non-canonical and canonical language (here we especially mean spoken language). Previous findings on lexico-grammatical and also cohesive phenomena have evidenced that there is more variation between written and spoken dimensions in German than in English, even though they are closely related, cf. Mair (2006) or Kunz et al. (forthcoming). Studies with respect to spoken and written Czech (see, e.g., Cvrček et al. (2010)) suggest that the differences between written and spoken language are even more pronounced in Czech than in German, at least with respect to lexico-grammar, hence we expect that this also holds for the level of text/ discourse. We therefore suggest that if we draw a line of differences between spoken and written English, German and Czech, we would observe a continuum in the degree of variation between these languages, as seen in Figure 1. The graph also reflects the above assumption that the differences are less pronounced between English and German than if we compare English and German with Czech. The reasons for this lie in the linguistic heritage of these languages (English and German have a common WestGermanic origin while Czech is a Slavic language) and in sociolinguistic factors that influenced their evolution (for example, Czech purism at the beginning of the 20th century, described, e.g., in Havránek and Weingart (1932)). To our knowledge, there is no