Olle Engstrand, ‘The IRIS speech data base - a status report’. In O. Engstrand (ed.): Papers from the Swedish Phonetics Conference held in Uppsala, October 17-18, 1986. Reports from Uppsala University, Department of Linguistics (RUUL) 17, 121-126, 1987.

Abstract

To date the IRIS data base contains speech samples from approximately 100 languages and from variants of Swedish spoken by several ethnic minorities residing in the country. The data base is meant to provide an easily accessible reference material for cross-language studies in phonetic typology and the phonetics of Swedish as a second language. "IRIS" is an acronym for "Invandrarröster i Sverige" (Immigrant voices in Sweden). The paper describes the input to the database (language selection, speech samples, informants) and the processing (digitizing and labeling) of the data.