Lexical patterns in role-playing games: a corpus-based analysis of vocabulary profiles and keyword distributions

The present study investigated the vocabulary profile of role-playing games (RPGs) through a corpus-based approach. To this end, three corpora comprising 685,356 tokens were compiled from the narrative content of three widely acclaimed RPGs: The Elder Scrolls V: Skyrim , The Witcher 3: Wild Hunt , and Baldur’s Gate 3 , including monologues, dialogues and dialogue options. Firstly, the corpora were subjected to a vocabulary profile analysis to find out about the distribution of the vocabulary items across the first 1,000 word families, the second 1,000 word families, the academic vocabulary list and non-list words. Secondly, keyword analysis was carried out for each corpus to find out to what extent the corpora were similar to and different from the general American English corpus. The findings indicated that RPGs predominantly use the first 1,000 word families, with lexical coverage ranging from approximately 76 % to 85 % , while also exposing players to relatively less frequent words from the second 1,000 word families and academic words, and a notable proportion of non-list items mainly consisting of proper nouns. Compared to general language corpora, RPGs show significant differences in the frequency of keywords, particularly with regard to pronouns, auxiliaries, conjunctions, and proper nouns, reflecting the distinctive lexical patterns of RPG monologues and dialogues. Based on the findings, the study offers academic and practical implications for language learning.