Children are more adept at learning language than adults. This phenomenon is often ascribed to innate language-learning abilities available only during an early sensitive period. Such a period presumably evolved to be of sufficient duration in an environment providing a fully-developed, natural language.
In those rare situations in which the linguistic environment is incomplete, children's natural language-learning capacities may be more evident as they take on a greater load. Are they sufficient to create a new language within a child's learning-lifetime? A generation of deaf Nicaraguans who were not exposed to a fully-developed language provides a test case.
The cohort of deaf Nicaraguans entering school in the late 1970s spontaneously began developing a new sign language, the first known in Nicaragua. A second cohort of children, entering school in the decade that followed, learned this system from their older peers. Current discrepancies between the first cohort of children (now adults) and later arrivals (now adolescents) reveal language creation processes in action.
Changes in particular linguistic constructions reveal how learning shapes languages. For example, the use of the signing space to link elements together in a sentence has changed to such a degree that, in shielded communication tasks, second-cohort signers predictably "misinterpret" first-cohort signing, even though such signing represents the input they received as children. The younger signers have reanalyzed and reshaped their input into a more precise and specific system.
These changes give insight into the nature of language-learning abilities. Given sequential cohorts of children, unsystematic gestures become a systematic grammar, and a new language emerges.