Milli mála - 01.01.2010, Blaðsíða 303
3.7 Treatment of Icelandic dental fricatives [q], [ð]
13 tokens: Þór, Þórir, Þóra, Þuríður, Þrúður, Ragnheiður, Davíð,
Friðfinnur, Garðar, Guðrún, Hafliði, Guðmundur, Gyrðir
2 choices: [q], [ð] → [s], [z] or [q], [ð] → [t] [d]
Dental fricatives exist in English, but the way they are transliterated
into Japanese is not consistent. ‘Th’ is usually pronounced [s] in
Japanese, but in some of the newer loan words, [t] is used, e.g. ther-
apy [se ra pi i], but aromatherapy [a ro ma te ra pi i]. One might
assume that phonetic similarities would determine the choice of
sounds used in transcription. [T] and [ð] are dental fricatives, and [s]
and [z] are alveolar fricatives.22 The places of articulation are not very
far apart and the two groups are both fricatives. Whereas [t] ([d]) are
alveolar stops, [s] ([z]) seem closer to [T] ([ð]), therefore it could be
assumed that subjects would choose [s] ([z]) for the transliteration.
From the point of view of speech perception as well, [s] ([z]) are clos-
er to [T] ([ð]) than to [t] ([d]). Figure 1 shows Johnson’s perceptual
map of fricatives and [d] in English.23 However, auditory/perceptual
similarity is not always the only factor driving the alternation of a
sound. as Johnson points out, this and that may be mispronounced
[q] – [ð] and [ð] – [z], but not [ð] – [v], though [v] is closer to the orig-
inal sound [ð] on the perceptual map than either [d] or [z].24
This study examines the choices made by Icelandic and Japan -
ese speakers between the fricatives [s] ([z]) and plosives [t] ([d]),
as alternative sounds to the original [T] ([ð]). It also observes
whether the choice is affected by environment, e.g. the types of
vowels that follow.
KaOru uMEZaWa
303
22 It is said that there are two ways of articulating [s] by native speakers of Icelandic, one with
the tip of the tongue nearing the alveolar ridge, and the other with the tip of the tongue
behind the lower front teeth with the result that a stricture is formed by the blade of the
tongue against the alveolar ridge. Kristján Árnason, “atkvæðagerð, hljóðskipun og lengd”,
p. 16; Eiríkur rögnvaldsson, Íslensk hljóðfræði. Kennslukver handa nemendum á
háskólastigi, reykjavík: Málvísindastofnun Háskóla Íslands, 1989, p. 30. (I would like to
express my gratitude to Eiríkur rögnvaldsson and Magnús Sigurðsson for verifying the con-
tent of the Icelandic sources cited.)
23 Keith Johnson, Acoustic & Auditory Phonetics, Oxford: Blackwell Publishing, 2nd edition,
2003, p. 69. (The perceptual map is based on the confusion matrix where the number of
times each sound is mistaken for another is counted. The number of times a sound is con-
fused with another is calculated into the degree of similarity by ratio, and then into the per-
ceptual distance.)
24 Ibid, pp. 59–71.
Milli mála 2011_Milli mála 1-218 6/28/11 1:39 PM Page 303