Profile-based measures of lexical variation. Four case studies on variation in word choice between Belgian and Netherlandic Dutch.

In their pivotal work Convergentie en divergentie in de Nederlandse woordenschat: Een onderzoek naar kleding- en voetbaltermen Geeraerts, Grondelaers & Speelman (1999) introduce the notion of profile-based measures of lexical variation. The adopted profile-based perspective takes its starting point in the profile of a concept, which can be defined as all alternative lexicalisations of that concept in a given source, distinguished by their relative frequencies in that source. It is precisely this distributional nature of the profile of a concept that allows to quantify different measures of lexical variation. The first profile-based measure is the external uniformity measure and it calculates the degree of overlap between two sources. In other words, the external uniformity measure looks at how much two sources have in common regarding word choice. Next, the internal uniformity measure allows to quantify the homogeneity of concepts. Lexical homogeneity refers to the degree to which a concept is lexicalised uniformly in a given source. If a concept is lexicalised by only one term, the internal uniformity is maximal in that source. The internal uniformity decreases when there are more alternatives to express the concept. Both the number of alternative terms as well as their distribution influence the degree of homogeneity. The third profile-based uniformity measure is the feature proportion measure, which calculates the proportion of terms which demonstrate the feature in question (e.g. 'of foreign origin'). In Geeraerts, Grondelaers & Speelman (1999) these three measures are used to investigate how the Dutch lexicon in Belgium and that in the Netherlands relate to one another and how they develop. In particular, data from two lexical fields, that of CLOTHING and that of FOOTBALL, three points in time, the 1950s, 1970s, and 1990s, and two registers, standard language and more informal regionally coloured language, are analysed for this purpose. In this dissertation, we continue to pursue the original research goals set out by Geeraerts, Grondelaers & Speelman (1999). Moreover, by adding a new point in time for the CLOTHING terminology and by introducing three new lexical fields, each with its own characteristics, we raise new research questions, and develop the representativeness of the study. Besides the profile-based approach to lexical uniformity introduced above, we extend this profile-based approach to two other areas of lexical variation. First, we will investigate which concept and word features influence lexical success. These features include characteristics such as 'is it a concrete or an abstract concept' or 'is it a simplex or compound word'. In this respect, we assume lexically successful words, as calculated by a high relative frequency for a specific word, are characterised by a number of features. More advanced inferential statistics are applied to simultaneously measure which of these features influence lexical success the most. Second, we will study lexical homogeneity more closely. In particular, we will examine which concept features are typically associated with very homogenous concepts, as measured by the internal uniformity measure introduced above. This dissertation centres around four case studies which each introduce a different lexical field. The first case study replicates the CLOTHING terminology case study from Geeraerts, Grondelaers & Speelman (1999) and adds a new temporal measuring point to further track the evolution of word choice trends in Dutch. In the second case study, the same focus on lexical uniformity is applied to TRAFFIC terminology. Additionally, we also examine lexical success for this lexical field. For the third case study, we zoom in on the lexical field of IT and again study lexical uniformity and lexical success. The fourth case study is concerned with EMOTION terminology. Besides looking at lexical uniformity and lexical success, we also investigate lexical homogeneity for this particular lexical field. Overall, this dissertation contributes to the study of Dutch by showing how the once clear converging trend between the Belgian Dutch lexicon and the Netherlandic Dutch lexicon, on account of the former moving towards the latter, has come to a halt. In addition, in studying lexical variation, we demonstrate the influence of the lexical field and raise the importance of the identification and operationalisation of relevant concept and word features. References Geeraerts, Dirk, Stefan Grondelaers & Dirk Speelman. 1999. Convergentie en divergentie in de Nederlandse woordenschat: een onderzoek naar kleding- en voetbaltermen. Amsterdam: P.J. Meertens-Instituut.

Jaar van publicatie:2022

Toegankelijkheid:Closed