Semantic prosody in parallel corpora: An English-Slovene investigation

Abstract The article reports on a corpus-driven analysis of how Sinclairian extended units of meaning (EUMs) are treated in English into Slovene translation. The main goal of the underlying research was to test to what extent the EUM is perceived as the proper (extended) unit of translation as proposed by Zethsen (2021). The point of departure was a list of high-frequency EUMs that have previously been identified in a monolingual English corpus. Subsequently, these EUMs were checked in the OpenSubtitles 2018 parallel English-Slovene corpus. Although the overall number of EUMs was lower than expected, the results indicate that the rate of successful translation of EUMs in real-life circumstances is in the 30–40% range, while in the remaining majority of translations semantic prosody is either omitted or it undergoes explicitation by means of denotation. This low rate should encourage authors in the field of corpus-based translation studies to spread and raise awareness of semantic prosody. Another implication of the study in the field of higher education of translators is the need to alert translation students to this elusive phenomenon.