Vowel Length Shifts Between Speech Stances

RM 215-218

The paper looks at the relationships between stance and vowel length. It is hypothesized that in narrative stance, vowel length will be longer in duration than when in conversational stance. Data is drawn from the University of California, Santa Barbara linguistic corpus, with conversations focused on individuals in and around Southern California. Sections of data chosen focused on two parts of dialogic exchange: typical conversation stance and narrative stance. The paper builds on work by James Purvis (1980) and others regarding the nature of cognitive monitoring and how it affects speech patterns and Wiliam Labov (1966/2006) on modes of speech that may represent more accurate-to-life patterns. Tokens chosen for analysis are /a/, /ai/, and /e/. Three of each token in onset or first syllable position is collected for analysis from both conversational and narrative stance. Analysis of tokens then take place by (mean) averaging each token’s length for each speaker in each stance, then the total vowel average time is calculated again for each speaker in each stance. Beyond intra-vowel, intra-speaker averages, interspeaker average is calculated to assess consistency of the vowel length changes between stances. The paper finds that the length of tokens shows an average increase during narrative stance over conversational stance, especially during certain tokens. The paper concludes by motivating further research on the subject to locate the specific animating factor behind this durational shift.