MIT CogNet, The Brain Sciences ConnectionFrom the MIT Press, Link to Online Catalog
SPARC Communities
Subscriber : Stanford University Libraries » LOG IN

space

Powered By Google 
Advanced Search

 

Retroflex Consonants in Norwegian: Are They Really? Evidence from EMA and EPG

 Hanne Gram Simonsen, Inger Moen and Steve Cowen
  
 

Abstract:

The term retroflex has a widespread use as a term for describing the articulation of certain speechsounds. The definitions mirror the latin meaning of the term, e.g. "A retroflex articulation is one in which the tip of the tongue is curled up to some extent" (Ladefoged & Maddieson 1996:25).Although this clearly is a term describing tongue configuration, it is in most descriptions "for practical reasons" placed among the terms for place of articulation, with a location somewherebetween the (post)alveolar and palatal region, as indicated for instance by its placement in the IPA table of phonetic symbols.In this paper, we will question how useful and appropriate the termretroflex is in its current usage, exemplified through Norwegian. In Norwegian, there is a distinctive opposition between two sets of coronal consonants where the members of the first set aretraditionally referred to as alveolar, while those in the second set are referred to as retroflex. The articulation of these consonants, with particular focus on the oral stops, have been investigatedthrough the use of Electropalatography (Reading EPG 3) and Electromagnetic Articulography (Carstens Articulograph). The investigation shows that the so-called retroflex phonemes do notalways show a retroflex tongue configuration. There is considerable variation, inter- and intraindividually, both regarding tongue configuration and regarding the contact area in the palate. Whatis distinctive between the two sets of consonants is not whether the articulation is retroflex or not, since the tongue tip is very often not bent upwards or backwards during the articulation of thesounds labelled retroflex. Nor is there a clearly distinctive place of articulation in the roof of the mouth. As indicated also by earlier EPG investigations of Norwegian (Moen & Simonsen 1997;1998) there is overlap concerning the area of contact between the two sets of coronals. The alveolar consonants have both alveolar and postalveolar contact, and for the allegedly retroflex consonantsthe contact area varies widely from alveolar to palatal, varying i.a. with the vowel context. What is distinctive, however, is whether the articulation is apical or laminal: the so-called retroflexconsonants are all apical and the so-called alveolar consonants are all laminal.

 
 


© 2010 The MIT Press
MIT Logo