Variation isn't that hard: Morphosyntactic choice does not predict production difficulty.

Area of Science:

• Linguistics
• Psycholinguistics
• Cognitive Science

Background:

• Grammatical variability is common in spoken language.
• Previous theories suggested this variability might indicate cognitive load or production difficulty for speakers.
• The relationship between production difficulty and grammatical complexity requires empirical investigation.

Purpose of the Study:

• To investigate the correlation between grammatical variability and production difficulty in spoken American English.
• To test the hypothesis that variable linguistic contexts increase cognitive planning demands for speakers.
• To challenge the notion that grammatical variability is inherently sub-optimal for speech production.

Main Methods:

• Analysis of a sub-sample from the Switchboard Corpus of American English (285 transcripts, 34 speakers).
• Measurement of production difficulty using filled pauses (e.g., 'um', 'uh') and unfilled pauses (speech planning time).
• Statistical examination of 20 morphosyntactic variables in relation to production difficulty metrics.

Main Results:

• No positive correlation was found between the presence of variable contexts and the measured metrics of production difficulty.
• Considering 20 morphosyntactic variables collectively (N=6,268) did not reveal a positive effect on production difficulty.
• Variable contexts in speech do not appear to correlate with measurable production difficulties.

Conclusions:

• The findings challenge the view that grammatical variability imposes an additional cognitive burden on speakers.
• Grammatical variability may not be sub-optimal for speech production as previously theorized.
• Spoken language production does not necessarily become more difficult in linguistically variable contexts.