ON CANONICAL FORM(S) OF CO-OCCURRENCES IN DIFFERENT LSP AND/OR DOMAIN SPECIFIC CONTEXTS

Thanks to technological advances in recent decades, both the possibility of compiling highly specialized large corpora and the availability of high-performing corpus-linguistic tools allow us to carry out fine-grained comparative studies to detect regularities in language use in different domains and to heighten its predictability (see, for example, Abel/Zanin 2011; Cap/Heid 2011). Nonetheless, given the complexity of natural languages, the quantitative-qualitative analysis of huge amounts of empirical linguistic data remains a challenging issue (see, in particular, Gries 2014). As will be shown, with the help of so-called concgrams, which allow us to display semantically- or pragmaticallymotivated co-occurrences, it is feasible to explore more systematically sameness and difference, repetition and variation in the realization of salient patterns (see, among others, Cheng W. et al. 2006; Cheng W. et al. 2009; Höhmann 2013; forthcoming). By focusing on phraseological units in highly specialized subcorpora from different areas (history of art, environmental law and the promotion of tourist destinations), this paper examines and supports the hypothesis that not only the use of core vocabulary items but also the respective canonical forms of cooccurrences are affected by microscopic LSP and/or domain-specific regularities which can be detected with the help of fine-grained corpus linguistic analysis.