Lexicographer’s Lacunas or How to Deal with Missing Representative Dictionary Forms on the Example of Czech

Abstract

When compiling a list of headwords, every lexicographer comes across words with an unattested representative dictionary form in the data. This study focuses on how to distinguish between the cases when this form is missing due to a lack of data and when there are some systemic or linguistic reasons. We have formulated lexicographic recommendations for different types of such ‘lacunas’ based on our research carried out on Czech written corpora. As a prerequisite, we calculated a frequency threshold to find words that should have the representative form attested in the data. Based on a manual analysis of 2,700 nouns, adjectives and verbs that do not, we drew up a classification of lacunas. The reasons for a missing dictionary form are often associated with limited collocability and non-preference for the representative grammatical category. Findings on unattested word forms also have significant implications for language potentiality.

Introduction

“…While cognitively this approach makes a great deal of sense in that it shows how people negotiate the common absence of forms in their experience (Janda and Tyers, 2018), it can lead in practice to counterintuitive results in which an accidental corpus lacuna deriving from a pragmatic situation is interpreted as an inherent paradigmatic gap. The logical results are documented in Kováříková et al, 2020: the form zavraždila 'murder-PAST-FEM' is a lacuna, and yet fails our 'inherent defectivity' test, as native speakers have no difficulty in providing it or accepting it. Its near-absence in the corpus turns out to be a function of the fact that far more men are murderers than women.…”