Multiword units and synonymy: interface between collocations, colligations, and semantic prosody

Abstract

Appropriate use of near-synonyms in English often poses difficulty for learners. This study aims to compare and contrast the near-synonyms predict and foresee, focusing on genres, collocations, colligations, and semantic prosody which specifically characterize each synonymous verb. The study consulted the Corpus of Contemporary American English (COCA), i.e. the largest corpus representing American English (Davies, 2020), for the distribution of the target synonyms across genres and the top-20 noun collocates the MI-score of which is 3 or above. The total frequency of predict in the corpus is far higher than that of foresee. Of all the eight genres of COCA, predict is the most common in academic texts, whereas the highest number of occurrences of foresee is found in webpages. Interestingly, both target synonyms are not characteristic of colloquial English as their frequencies in speaking, fiction, and TV and movie subtitles are relatively low. The data from COCA also reveal that both synonymous verbs are common in written English, with predict being far higher in frequency. Although sharing certain object noun collocates, it is indicated through semantic prosody, i.e. occurrences of adjacent collocates and surrounding lexical items, that foresee is associated with negative connotations, while predict does not primarily express adversity. As near-synonyms of each other, the target words share colligational patterns. However, two corpus-informed syntactic structures, i.e. foresee + somebody + V.ing and predict + somebody + to-infinitive, clearly distinguish between both verb synonyms. It is concluded from the corpus data that predict and foresee are near-synonyms rather than absolute synonyms since they clearly differ in connotations and collocational/colligational patterns.