The representation of ‘immigrant’ in news discourse: a corpus-based study of semantic preference and semantic prosody

The representation of immigrants in the media has been a long-standing and contentious issue in public, political, and social discourse. In today’s increasingly complex geopolitical landscape, the portrayal of immigrants has become even more controversial. This study employs Sinclair’s model of Extended Units of Meaning to examine the representation of immigrants in written media discourse using the enTenTen21 corpora on Sketch Engine. Utilizing a corpus-based approach, the study focuses on collocates, semantic preference, and semantic prosody of the node word ‘immigrant’. Semantic preference is determined using the UCREL Semantic Analysis System (USAS), while semantic prosody is identified through a textual analysis of concordance lines of collocates of the node word. The findings reveal a dominant transnational narrative framed by a semantic preference for government, law, and order. Government-related collocates such as ‘illegal’ and ‘deportation’ carry a dominant negative semantic prosody that constructs immigrants as a threat, while ethnically-specific collocates such as ‘Mexican’ and ‘Asian’ exhibit a positive prosody, highlighting a humanitarian narrative towards vulnerable immigrant groups. Yet, competing narratives (threat and vulnerability) coexist around collocates like ‘refugee’ and ‘undocumented’. The coexistence of these patterns demonstrates that the meaning of immigrant is constructed through competing extended units of meaning within the global news media context.