Communicative pressures shape language during communication (not learning): Evidence from case-marking in artificial languages

Natural languages seem to be designed for efficient communication. A classic example is Differential Case Marking, when nouns are marked for their grammatical role only if this information cannot be derived from world knowledge (e.g. only atypical objects need to be linguistically marked as objects). Fedzechkina et al. (2012) present experimental evidence from an artificial language learning paradigm suggesting that biases in learning favour Differential Case Marking: learners exposed to a language with optional case-marking restructure the input, using case-markers more in situations where marking would reduce the uncertainty or ambiguity experienced by a listener, despite the fact that they never use the artificial language in a communicative task where a listener's uncertainty is a relevant consideration. This is surprising given previous studies suggesting that biases in learning favour simplicity and are agnostic with respect to communicative function. We report an experiment investigating whether biases for communicatively-efficient Differential Case Marking exist in learning. Contrary to Fedzechkina et al. (2012), we find no evidence for such a bias in learning: participants do not reliably produce Differential Object Marking in non-communicative recall tests, and their use of case is impervious to factors influencing message uncertainty or ambiguity, observations which are inconsistent with their hypothesis. However, we find good evidence that participants' behaviour in actual communicative interaction is driven by efficient communication considerations: in interaction participants exhibited the expected Differential Object Marking pattern. This suggests that languages adapt to communicative efficiency constraints as a result of being used in communication, rather than due to biases in human learning favouring communicatively-efficient languages.