Failure is always an option: Children's production of truth-functional negation and its implications for learning

Truth-functional negation presents a host of obstacles to a new language learner, ranging from abstractness (e.g., how can one point to it) to salience (e.g., it refers to things that are not occurring). It is therefore unclear how young learners learn truth-functional negation given the limitations in their knowledge about the world and still developing cognitive abilities. We explore the possibility that there are particular "gem" moments for "not" learning: when parents comment upon failures (e.g., failing to catch). Given recent findings that children are sensitive to abstract features of events before (successful) completion, failures may serve as null-discourse contexts for negator use and hence play an early role in learning what "not" means. We experimentally investigated the contributions of non-linguistic (i.e., visual) and linguistic (i.e., syntactic and/or lexical) evidence in inferring the presence of a negator in a message. In Study 1, adults (N = 144) guessed how parents might comment on videos depicting Failures and Successes. Across three experiments, negators were guessed at higher rates for Failures than for Successes even when no linguistic cues supplemented these videos. Linguistic cues promoted negator guessing when they included lexical or syntactic information, though the effect of language was strongest when both lexical and syntactic information were provided. In a child-friendly version (Study 2), children (4-7 years, N = 20) spontaneously used negators to describe Failures (64.5 %) much more often than Successes (1.5 %). This effect was observed without any linguistic support (i.e., just visual observation) and occurred in 19 out of 20 (95 %) of children and 20 out of 20 (100 %) adult controls. These findings support the hypothesis that parental comments on failures might serve as potent "gem" moments for learning negation merely by observing the referent world, even for linguistically inexperienced learners who do not yet have access to linguistic evidence for negation.