Support from creole languages for functional adaptation in grammar: Dependent and independent possessive person-forms

It seems to be a robust empirical observation that independent possessive person-
forms (such as English mine, yours, hers) are always longer than (or as long as)
the corresponding adnominal possessive person-forms (such as English my, your,
her). Since adnominal forms are also much more frequent in discourse than in-
dependent forms, this universal coding asymmetry can be subsumed under the
grammatical form-frequency correspondence hypothesis (Haspelmath et al. 2014).
In other words, the fact that independent possessive forms are longer can be seen
as a functional response to the need to highlight rarer, less predicatable forms.
In this paper, I present evidence from creole languages and show that irrespec-
tively of their young age and extremely accelerated grammaticalization processes,
these high-contact languages confirm the coding asymmetry. Moreover, creole lan-
guages, just as non-creole languages, show a diverse array of diachronic pathways
all leading eventually to longer independent possessive person-forms. Such a case
of multi-convergence of structures through very different diachronic processes
strongly suggests that the current patterns cannot be explained exclusively on the
basis of the sources and the kinds of changes that commonly give rise to indepen-
dent (and adnominal) possessive forms, but that there is an overarching functional
efficiency principle underlying these coding asymmetries.