Creole languages, which arise when speakers of mutually unintelligible languages are thrown into sustained contact and a new, stable vernacular emerges within a generation or two, display grammatical similarities across creoles with entirely unrelated parent languages: comparable systems for marking tense, mood, and aspect through preverbal particles recur from the Caribbean to the Indian Ocean. One explanation locates this convergence in substrate transfer: the enslaved and indentured populations who created these languages disproportionately spoke West African languages sharing certain grammatical structures, so the recurring patterns reflect those shared substrate features surfacing through an imperfectly acquired European superstrate vocabulary.

A rival explanation, the bioprogram hypothesis, instead locates the convergence in the language-acquisition process itself: children born into a contact community with no stable target language to acquire natively, the argument runs, draw on an innate grammatical template shared by the human species, so that creoles the world over converge not because their creators shared a substrate language family but because children everywhere resort to the same default grammar when no fully formed native language is available to learn.

Adjudicating between the two accounts has proven difficult because the historical contact situations that produced creoles typically involved both a substantial degree of linguistic diversity among substrate speakers, which should weaken any single-substrate explanation, and a first generation of children acquiring language under exactly the input-poor conditions the bioprogram hypothesis specifies, so that most documented cases cannot cleanly isolate which mechanism, if either alone, was doing the causal work.