Interface of Linguistic and Visual Information During Audience Design

Area of Science:

• Psycholinguistics
• Cognitive Science
• Communication Studies

Background:

• Speakers adapt their language based on addressee needs (audience design).
• The specific contextual cues (visual, linguistic) driving this adaptation are not fully understood.
• Previous research has explored audience design, but the interplay between visual and linguistic context needs further investigation.

Purpose of the Study:

• To investigate how speakers use visual and linguistic context during audience design.
• To determine if speakers adjust referring expressions when addressees lack access to prior linguistic information.
• To examine the role of contextual cues in mapping referring expressions to referents.

Main Methods:

• Speakers described referents in an experiment.
• Manipulated addressee's access to the linguistic context (prior descriptions).
• Varied the visual distinctiveness of referents (e.g., uniform vs. plain clothes).

Main Results:

• Speakers used less specific descriptions (e.g., "man") and more basic-level terms when addressees missed linguistic context, but only if visual cues were also ambiguous.
• Speakers reduced pronoun use when addressees lacked shared linguistic antecedents for anaphoric reference.
• This indicates flexible use of contextual information for successful reference.

Conclusions:

• Speakers dynamically integrate visual and linguistic context to design their utterances for addressees.
• Audience design involves flexibly relying on available cues, prioritizing non-visual ones when linguistic information is inaccessible.
• Speakers demonstrate cooperative communication by avoiding ambiguous pronouns when antecedents are not shared.