The impact of implicit scalar alternatives on the products of language comprehension: Evidence from recognition memory.

Speakers often leave parts of their message unarticulated and rely on their comprehenders to make inferences about the intended meaning of their message. One way in which comprehenders can recover a speaker's implicit meaning is to consider alternatives to what the speaker said. While we know that alternatives are activated and affect initial processing, less is known about how those alternatives ultimately become integrated into comprehenders' representations of the message in a more long-term durable format. This paper reports two experiments that use a recognition task to probe whether implicit scalar alternatives remain active during an experimental session (Experiment 1) or after a 24-hour delay (Experiment 2). The results show that implicit scalar alternatives seem to be maintained in memory when probed during the experimental session but not after a 24-hour delay. These results suggest that alternatives that are relevant to the discourse are stored in long-term memory but may not be accessible when probing linguistic representations that have been encoded as conceptual meaning representations.