Abstract Research has shown that when ‘xxx English’ is used in reference to an English variety, the attribute ‘xxx’ is more likely to be an adjective than a noun. This is true of traditional ENL and ESL varieties, whose speakers can legitimately claim that ‘English is our language’. EFL varieties tend to follow the same trend, with ‘China English’ being a glaring exception to date. Corpus analysis shows that the pre-head attribute of a noun phrase may be filled by a noun (e.g., Belgium, Canada) or an adjective derived from it (e.g., Belgian, Canadian). A pre-head nominal attribute expresses the meaning ‘a type of’ (e.g., ‘communication skill’, ‘generation gap’), while a pre-head adjective signals either a quality (e.g., ‘smart city’ – with quality of being smart) or membership of a class (e.g., ‘smart card’, distinct from other cards). A pre-head nominal attribute and its adjectival counterpart signaling ‘class membership’ thus share a classificatory function. This helps explain stylistic variation such as free alternation between ‘Scottish flag’ and ‘Scotland flag’ in the same text; nor is there anything unusual using names of countries in sports reporting (e.g., ‘Brazil national football team’). But such a stylistic shift is semantically nuanced and not always licensed (consider, e.g., ‘Ukrainian security’ and ‘Ukraine war’, where the pre-head attributes ‘Ukrainian’ and ‘Ukraine’ are not interchangeable). The adjectival form of a toponym is usually longer in speech and writing than the toponym itself. Apparently, phonology matters when naming an English variety, witness the rarity of ‘Guamanian English’ and ‘Peruvian English’.
David C. S. Li (2026) studied this question.