Networks of meanings: Complementing collostructional analysis by cluster and network analyses

GND
1190300125
Affiliation
University of Hannover, Hannover, Germany
Dekalo, Volodymyr;
GND
122070135
ORCID
0000-0002-2884-9677
Affiliation
University of Erfurt, Erfurt, Germany
Hampe, Beate

Abstract This paper contributes to the development of collostructional analysis by taking up one of the long-standing issues in collostructional analysis, viz. the assessment and interpretation of the collexeme lists created. It uses the DWDS corpus documenting 20th century written German to study two modalverb constructions which contain either vermögen or bekommen as the first verb plus an infinitive. When expressing ‘possibility’/‘capability’, these constructions compete with the ubiquitous and strongly grammaticalized könnenconstruction. To compare their constructional semantics, the paper combines collostructional analyses with a close investigation of the verbal meanings characterizing their V inf slots. To classify the collexemes occurring in these slots comprehensively, the semantic information provided about each V inf collexeme by “GermaNet” is utilized in three different ways: A manual, context-sensitive classification of the top collexemes into the 15 generic verb classes provided by GermaNet is followed by cluster analyses over the semantic-relatedness values given out by the SemRel tool of GermaNet for the pairwise comparison of all possible collexeme combinations. This is complemented by a network analysis identifying collexemes that are salient or “central”, i.e. located on the largest number of “shortest paths” connecting any two collexemes. The implications of the results for the functional differential between the two constructions as well as the methodology itself are discussed.

Cite

Citation style:
Could not load citation form.

Rights

License Holder: © 2018 by Walter de Gruyter Berlin/Boston

Use and reproduction:
This publication is with permission of the rights owner freely accessible due to an Alliance licence and a national licence (funded by the DFG, German Research Foundation) respectively.