Keywords
vowel-shift relativity problem, benchmark limitations in normalized data, midpoint-only vowel analysis, normalization-dependent shift detection
Abstract
It’s difficult to say objectively whether a vowel is shifted:
- Relationship to other vowels (which may also be shifting)
- Benchmarks don’t work because of normalization and dialect differences
Compare to ANAE Benchmarks:
More processing involved:
- remove stopwords
- remove outliers
- remove diphthongs
- remove presonorants
- isolating midpoints
- specific normalization procedure
- remove unstressed vowels
Seven steps = 5040 combinations
Original Publication Citation
Joseph A. Stanley. “Order of Operations in Sociophonetic Data Analysis.” New Ways of Analyzing Variation 49. Online. October 19–24, 2021.
BYU ScholarsArchive Citation
Stanley, Joseph A., "Order of Operations in Sociophonetic Data Analysis" (2021). Faculty Publications. 7974.
https://scholarsarchive.byu.edu/facpub/7974
Document Type
Presentation
Publication Date
2021
Publisher
New Ways of Analyzing Variation
Language
English
College
Humanities
Department
Linguistics
Copyright Use Information
https://lib.byu.edu/about/copyright/