An MRI study of the effects of vowel context on fricatives.

Many recent studies have used Magnetic Resonance Imaging (MRI) to obtain the vocal tract shape during speech production. Since acquisition times for magnetic resonance images are still relatively long { from seconds to minutes { the con gurations to be imaged must either be produced by sustainable sounds, or the speech sequence must be uttered repeatedly so that a `moving' image can be built up over successive repetitions (Foldvik et al., 1995). As a result, MRI studies have tended to focus on either vowels (e.g. Baer et al., 1991) or fricatives (e.g. Narayanen et al., 1995).