Abstract
The present work describes the building of a conversational corpus from the Metropolitan Area of Guadalajara. This corpus contains more than 40 hours of audio recordings of spontaneous conversations from 114 speakers. The methodology for data collection is described here, followed by a brief study of the discourse marker o sea. From the perspective of usage based phonology (Bybee 2001), we test if the duration of the marker is related to its functions in discourse. We extracted a 100 tokens of o sea produced by 10 speakers and we analyzed them according to their function. We then measured the duration of these tokens using Praat (Boersma y Weenink 2016). An ANOVA test shows that there is a significant difference on duration among the functions o sea has in discourse.Metrics
Metrics Loading ...