Show Less
Restricted access

Trends in Phonetics and Phonology

Studies from German-speaking Europe

Edited By Adrian Leemann, Marie-José Kolly, Stephan Schmid and Volker Dellwo

This volume was inspired by the 9th edition of the Phonetik & Phonologie conference, held in Zurich in October 2013. It includes state of the art research on phonetics and phonology in various languages and from interdisciplinary contributors. The volume is structured into the following eight sections: segmentals, suprasegmentals, articulation in spoken and sign language, perception, phonology, crowdsourcing phonetic data, second language speech, and arts (with inevitable overlap between these areas).
Show Summary Details
Restricted access

Transforming acoustic vowel data: A comparison of methods, using multi-dimensional scaling



This paper makes three contributions to the discussion of vowel transformation in acoustic phonetics. First, it compares a selection of transformation routines, using three performance tests: Linear Discriminant Analysis of vowel categories, inter-speaker vowel space area homogenization, and inter-speaker vowel space overlap. Second, the complementary functions of psychoacoustic transformation and vowel-extrinsic normalization proper are discussed and demonstrated. And, third, multi-dimensional scaling is proposed as a tool that can be useful in the assessment of multi-dimensional (or multi-parametric) results common in the testing of different methodologies, but also in other experimental work.


Acoustic phonetics, vowel analysis, psychoacoustics, vowel normalization, multi-dimensional scaling

*   Corresponding author:, Tel: +49 863 2166, Fax: +49 863 2167

a   Otto-Friedrich-Universität Bamberg, Lehrstuhl für Englische Sprachwissenschaft einschließlich Sprachgeschichte, An der Universität 9, 96045 Bamberg, Germany ← 35 | 36 →

1.0   Introduction

One of the main problems in acoustic vowel analysis is that there is no one-to-one relationship of formant patterns and phonemic vowel categories. In other words, similar formant patterns produced by different speakers may be perceived as different vowels. On the other hand, different formant patterns may be perceived as the same vowels (Hindle, 1978, p. 162). Thus, gross formant center frequencies are meaningful only within the acoustic vowel space generated by a particular speaker. The main reason given for this phenomenon is the different vocal tract lengths particularly of men, women, and children (Fant, 1966,...

You are not authenticated to view the full text of this chapter or article.

This site requires a subscription or purchase to access the full text of books or journals.

Do you have any questions? Contact us.

Or login to access all content.