Typology of Multiword Expressions in Czech and Frequency of Their Main Features in a Genre-Balanced Corpus Cover Image

Typologie víceslovných jednotek v češtině a frekvenční zastoupení jejich hlavních vlastností v žánrově vyváženém korpusu
Typology of Multiword Expressions in Czech and Frequency of Their Main Features in a Genre-Balanced Corpus

Author(s): Vladimír Petkevič, Marie Kopřivová, Milena Hnátková, Tomáš Jelínek, Pavel Kopřiva, Alexandr Rosen, Hana Skoumalova, Pavel Vondřička
Subject(s): Language studies
Published by: Univerzita Karlova v Praze - Filozofická fakulta, Vydavatelství
Keywords: multiword (lexical) expressions in Czech; typology of multiword expressions; frequency of types of multiword expressions; idiomaticity; lexical database; genre-balanced corpus

Summary/Abstract: The paper consists of two main parts: (a) In the first part, a typology of multiword expressions (MWE) in Czech is described in a detailed way. This typology is part of the description of MWE database entries in the lexical database LEMUR containing more than 10,500 MWE entries as of June 2020. MWE properties reflected in this typology are accounted for by categories and their values. Each MWE is identified by a unique lemma; a group of related MWEs is assigned a “superlemma”. A MWE is described by the following properties: a MWE definition, characteristic examples, lemmas and morphological features of MWE components (words), as well as the following key categories: MWE style/register, type of usage, syntactic structure (including its representation by a dependency and a phrase-structure tree), aspects of flexibility (variants and fragments, internal modifiability of individual MWE components, possibilities of syntactic transformations of the main MWE components and morphological constraints) and types of idiomaticity on the lexical, morphological, syntactic, semantic and pragmatic level. (b) In the second part of the paper, the authors focus on the frequency of the main features of the adopted typology in the real language material represented by the genre-balanced SYN2015 corpus, containing 100 mil. word forms (excluding punctuation): a type of usage correlated with a syntactic type and frequency of various kinds of idiomaticity. Our paper seems to be the first attempt at approaching the MWE properties from the point of view of MWE frequencies as types rather than tokens (i.e. frequencies of occurrences of a given MWE).

  • Issue Year: 11/2020
  • Issue No: 2
  • Page Range: 37-62
  • Page Count: 26
  • Language: English