In Nicaraguan Sign Language, the second cohort of child learners made the grammar more systematic than the adults who invented it

The most-cited finding from Ann Senghas's studies of Nicaraguan Sign Language (NSL) is a cohort effect: the grammar of the language became more systematic in the hands of the generation that inherited it than in the hands of the generation that created it. Summarizing Senghas, Susan Goldin-Meadow writes that "the innovations introduced into NSL by the second cohort came primarily from deaf individuals who entered the deaf community as children, rather than as adults."

The concrete example is a grammatical device for tracking who does what to whom. Per Goldin-Meadow's commentary, "Cohort 2 consistently uses a rotated perspective across sentences when marking who does what to whom. Cohort 1 does not use this perspective reliably." The founders — the first cohort, who assembled the shared system out of their individual home-signs — did not stabilize this device; the children who acquired that emerging system as their first language did. The regularization tracks age of entry, not simply calendar generation: the systematizers are those who came to the community as children.

The finding is read as evidence for a child-specific language-creating capacity: grammar is not merely handed down intact but is actively sharpened by young learners, who impose structure the input did not fully contain. This is the mechanism behind the broader observation that NSL is a language observed emerging from zero language exposure, and it is a real-world instance of the learner-driven regularization that the iterated-learning framework models as structure surviving a transmission bottleneck.