Coffee and Tea? A diachronic corpus-based, collocation and colligation analysis of the words coffee and tea

This paper aims to analyze the absolute and relative frequencies of the words coffee and tea in the COHA corpus and to conduct collocational and colligational analysis using log-likelihood ratio (LLR) (Gries. 2009; Bybee et al. 2001), pointwise mutual information (PMI) (Evert et al. 2016; Church et al. 1990), as well as KWIC (Keyword in Context) methods. The frequency analysis uses a diachronic approach, and collocation and colligation analysis is conducted using a synchronic approach. A sample language dataset of 3.6 million words is collected from the COHA (Corpus of Historical American English) website and analyzed using the Python programming language. The frequency analysis suggests that coffee occurs 46.67% more often than tea and that coffee is gaining more popularity while tea sees the opposite. The collocational and colligational analysis reveals that coffee collocates more diversely than tea in all categories, including verbs, adjectives, and nouns. Moreover, coffee and tea is frequently used in the context of social interaction, as well as in the sense of social gathering (instead of a beverage). They often collocate and are used with other food-and-drink-related words. The dominant constructions identified are possessive noun phrases and direct objects. Lastly, coffee is specifically found to be used in the domain of commercialization.