Formal and functional equivalence across typological diversity: insights from formulaic sequences in Korean, German and English

The way a language forms words (i.e. its morphological type) has been shown to affect the concept of formulaic sequences in a number of ways. This is especially the case if formulaic sequences are understood as fully specified sequences of word forms, e.g. as lexical bundles. Adopting a more flexible understanding of formulaicity as sequences of linguistic elements that manifest common ways of phrasing things in a speech community, this paper investigates how formulaicity is manifested formally and functionally in languages that are morphologically very different: what linguistic forms are used in the construction of formulaicity? Are the functions performed by formulaic sequences similar across those languages? To find out, a data-driven method of formulaic sequence identification was used to extract and compare relevant phenomena in a corpus of 30 million words of Wikipedia texts in Korean, German and English. Results indicate that while there are similarities, it is important to be sensitive to differing ways in which languages construct formulaicity. For example, while formulaicity is often expressed in multi-word constructs in more isolating languages, a polysynthetic language, as on might expect, features more morphological formulaicity at the sub-word level. The functions of formulaic sequences, however, turned out to be surprisingly similar across the three typologically different languages investigated, suggesting that formulaic sequences are a stable cross-linguistic concept