Word frequency and cognitive effort in turns-at-talk: turn structure affects processing load in natural conversation

Area of Science:

• Psycholinguistics
• Corpus Linguistics
• Conversation Analysis

Background:

• Word frequency impacts psycholinguistic processes.
• The role of word frequency in conversational turns-at-talk has been under-explored.
• Conversational turns are central to social interaction.

Purpose of the Study:

• Investigate word frequency distributions within conversational turns.
• Examine the relationship between word frequency patterns and cognitive processing effort in speakers and recipients.
• Test the hypothesis that word frequency in turns correlates with processing effort.

Main Methods:

• Applied corpus-linguistic methods to conversational data from the British National Corpus (BNC) and Freiburg Multimodal Interaction Corpus (FreMIC).
• Utilized pupil size measures from FreMIC to quantify concurrent processing costs in speakers and recipients.
• Analyzed word frequency distributions within turns-at-talk and correlated them with pupil dilation data.

Main Results:

• Turns-at-talk exhibit a regular word frequency pattern: high-frequency words initially, mid-range medially, and low-frequency finally.
• Speaker pupil size increases during a turn, peaking towards the end.
• A decrease in word frequency within turns is inversely correlated with increased speaker pupil size, indicating heightened processing effort.

Conclusions:

• The observed pattern suggests an informational climax within turns, progressing from high-frequency, low-information to low-frequency, high-information words.
• This distribution of processing load across turns enhances conversational efficiency, aids content prediction, and facilitates smooth turn transitions.
• Findings have implications for speech processing, turn structure, and information packaging theories in conversational interaction.