Blurb:

The aim of this paper is to make a first attempt towards measuring semantic differences between translated and non-translated language. More particularly, we present a quantitative bottom-up corpus-based method for the identification of lexical items in a semantic field. The proposed method will enable us to measure and to visualize semantic similarity between the elements in that field (i.e. the field of inceptiveness in Dutch), using bidirectional parallel corpus data (Dutch-French). This method builds on the successful implementation of parallel corpora


within contrastive linguistics to discern semantic fields (Dyvik 1998; 2004;


Aijmer & Simon-Vandenbergen 2004, 2006; Simon-Vandenbergen 2013),


while simultaneously overcoming one of its drawbacks, viz. the accurate, statistics-based visualization of the observed fields.