Why Frisian [ɪ] is really [i̙] or [e̝]

Abstract

Yn dit artikel ûndersykje wy de akoestyske posysje fan it Fryske heech-middel foarlike lûd dat meast as <ɪ> transkribearre wurdt en tusken de kardinale fokalen [i, e, ʉ, ɵ] pleatst wurdt. Op basis fan in grutskalich korpus mei foarlêzen en spontaan Frysk brûke wy Bayesiaanske mixed-effects-modellen om út te sykjen yn hokker diminsjes [ɪ] signifikant ferskilt fan dy oare lûden, en foar hoefier’t dy ferskillen strike mei in transkripsje fan <ɪ>. De resultaten litte sjen dat de akoestyske posysje fan [ɪ] wat F1–F2 oangiet kompatibel is mei de posysje neffens [i] en [e] op ’e offisjele kaart fan ’e IPA. Wy konkludearje dat de Fryske [ɪ] akoestysk sjoen liket op [i̙] of [e̝], en dat in transkripsje fan <ɪ> notaasje-technysk passend is en in adekwate werjefte jout fan de wurklike fonetyk fan it lûd yn kwestje.

This paper investigates the acoustic position of the Frisian high-mid front vowel, traditionally transcribed as <ɪ> and located somewhere between the cardinal vowels [i, e, ʉ, ɵ]. Working with a large-scale corpus of read and spontaneously spoken Frisian, we used Bayesian mixed-effects models to find out in which dimensions [ɪ] differs significantly from these other vowels, and whether these differences are compatible with a transcription of <ɪ>. Results show that the [ɪ]'s acoustic position in F1–F2 space is compatible with its position relative to [i] and [e] on the official IPA chart. We conclude that Frisian [ɪ] is acoustically similar to [i̙] or [e̝], and that its transcription as <ɪ> is therefore not only notationally convenient, but also adequately representative of this vowel’s actual phonetics.