A Dual-Process Activation Model: Processing definiteness and information status

Abstract

The introduction of a new discourse referent with a descriptive noun phrase involves the introduction of a new concept and the assignment of a referent to that concept. Concepts can be inferred from previous context, and thus be pre-activated (e.g., gym activates the concept of the noun trainer), or can be discourse-new. The function of the definite article is to signal unique identifiability of the referent, and the function of the indefinite article is to assert the existence of a set introduced by the descriptive content. We tested to what extent concept pre-activation and the function of the definite and indefinite article affect referent activation during retrieval and integration as well as referent activation at the sentence level. In Experiment 1, a visual world eye tracking experiment, we found that inferred referents of definite noun phrases were more accessible at subsequent pronoun resolution than inferred referents of indefinite noun phrases. No effects of definiteness were observed for referents with brand-new concepts. In Experiment 2, recording event-related potentials at the noun phrase itself, referents with pre-activated concepts were accessed and integrated more easily than referents with brand-new concepts. Furthermore, definite and indefinite articles yielded differently large frontal negativities. We discuss our results within a Dual-Process Activation Model, which distinguishes two processes in referent management: concept activation and referent activation. Our data suggest that these processes not only affect noun phrase processing but also trigger specific pragmatic inferences at the sentence level.