Methodology

Abstract

This chapter provides an account of the methodology employed, to a greater or lesser extent, in all the studies comprising this volume. First, it tackles two preliminary issues that have proved to be problematic in the literature: the extent to which corpus data may be said to reflect cognition, and the relationship between frequency and the two phenomena that, in the framework of the Gravitational Pull Hypothesis, create asymmetry in semantic networks-salience and connectivity. Then, the structure of the different corpora used in the individual studies is justified on the basis of previous work, most notably that of Hareide (2017a, 2017b). The main section in the chapter presents in detail the main steps in the methodology: choosing indicators, and the motivation for that choice; determining salience at the target end (magnetism), on the basis of corpus data, presence in normative texts and elicitation and translation tasks; determining salience at the source end (gravitational pull), drawing on corpus data and presence in normative texts; determining connectivity, on the basis of translation tests and previous studies, when available; hypothesis formulation; querying the translated component of the parallel corpus in order to compare frequencies to those in non-translations; statistical testing and hypothesis verification; querying the parallel corpus for connectivity patterns; and drawing conclusions. All the studies in this volume pivot around onomasiological, not semasiological, salience and focus on constructions, not lexical items.