Slowdowns in scalar implicature processing: Isolating the intention-reading costs in the Bott & Noveck task

Area of Science:

• Psycholinguistics
• Cognitive Science
• Pragmatics

Background:

• Underinformative sentences (e.g., "Some cats are mammals") are true under a semantic reading but false under a pragmatic reading.
• Pragmatic interpretations of such sentences lead to longer response times in truth evaluation tasks.
• Current theories often attribute these delays to the cognitive effort of deriving scalar implicatures.

Purpose of the Study:

• To investigate whether response time delays in processing underinformative sentences are partly due to participants adjusting to the speaker's informative intention.
• To explore the role of speaker informativeness in cognitive processing costs.

Main Methods:

• Experiment 1: Replicated Bott & Noveck's (2004) laboratory task in a web-based format.
• Experiment 2: Assessed changes in response times to underinformative sentences over an experimental session.
• Experiment 3: Examined response times when participants encountered one versus two "speakers" producing underinformative items.

Main Results:

• Response times to pragmatic interpretations of underinformative sentences initially decreased over the session but stabilized.
• These findings challenge the notion that implicature derivation is a consistent source of processing effort.
• Introducing a second speaker significantly increased response latencies to underinformative sentences, particularly after the introduction.

Conclusions:

• A portion of the processing cost associated with scalar implicatures may arise from considering the speaker's informative intentions.
• Cognitive adjustments to perceived speaker informativeness contribute to response time differences.
• The number of speakers influences the processing of pragmatic meaning in underinformative statements.