Unregistered words in everyday language and a proposal for their optimal lexicographic microstructure

Abstract

This article looks into lexicographic adaptation to media change. Instant messengers in Korea function as the most popular communication medium. According to the latest survey by Gallup Korea, instant messengers are used by 92% of the population overall. It means that the instant messenger corpus provides an ideal resource for accessing the language of the masses from a corpus linguistic point of view. In this contribution, we analyze an instant messenger corpus of 1.4 million words, and look into the prevalent unregistered words in the corpus to propose a microstructural model for them. Section 2 introduces the normalized parallel corpus of Messenger used in this study, and discusses the extraction methodology for unregistered words. We discuss the operational definition of unregistered words for dictionary inclusion and their extraction process. Section 3 examines the prevalence of unregistered words in the defined Messenger corpus and categorizes them based on the characteristics of messenger language. These characteristics encompass deviations from the pre-existing writing system, deviations from linguistic norms, deviations from socio-ethical criteria, incidental omissions, and non-verbal expressions. Section 4 proposes an optimal lexicographical structure incorporating unregistered words and their characteristics identified in the previous sections. Additionally, we discuss the extension and modification of microstructures in existing dictionaries, which could be made to effectively represent this new medium’s language.