Back to the Language Demos

Icelandic presents several interesting features related to consonant voicing.


Although there are differing views on the phonological status of preaspiration in Icelandic, it can essentially be seen as a mirror-image of the more familiar (post)-aspiration. Not surprisingly, preaspiration does not occur in initial position. Please note that the short unaspirated sounds may be weakly post-aspirated. Since consonant length is also relevant in Icelandic, in medial position the following three-way pattern involving voiceless consonants can be found:

emu WebApp
emu WebApp
emu WebApp
emu WebApp
emu WebApp
emu WebApp
Notes on the sonagrams:
The length of preaspiration plus consonantal closure is about the same as the length of consonantal closure in the long consonants. Long vowel plus short consonant is about the same length as short vowel plus long consonant. In initial position "normal" postaspiration can occur: "keppa". /r/ in word-final position may be virtually voiceless: "dökkur". For a similar set of examples from SoWL click here.

Additional Examples

The following words give examples of voiceless approximants and nasals occurring in contexts that are unfamiliar for German or English speakers.

  1. Voiceless lateral preceeding final voiceless plosive. It is possible to find surface contrasts for voicing in laterals and nasals in such positions. The voiceless variants may be related to the occurrence of pre-aspirated plosives in Icelandic, i.e the voicelessness may be a kind of mirror-image of the devoicing that can occur in laterals following syllable-initial voiceless plosives in English and German.

    emu WebApp

  2. Voiceless lateral in final position. Voiced laterals do also occur in this position.
    emu WebApp

  3. Voiceless nasal following preaspirated plosive; thus a very long stretch of voicelessness in this word.

    emu WebApp

  4. The voiced trill /r/ is followed by complete articulatory closure, and then a voiceless nasal release.

    emu WebApp

  5. Another example of a very long voiceless phase (/kst/), followed by a final voiceless /l/.

    emu WebApp

  6. Voiceless /r/ in initial position.

    emu WebApp

Audio Source:
Institute of Phonetics and Speech Processing, LMU Munich.