Preaspiration in the Nordic Languages: Synchronic and Diachronic ...
Preaspiration in the Nordic Languages: Synchronic and Diachronic ...
Preaspiration in the Nordic Languages: Synchronic and Diachronic ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
turn <strong>in</strong>to ChV sequences <strong>in</strong> which V is lost. The distributional character<br />
of h-sounds may turn out to be quite different <strong>in</strong> such cases, s<strong>in</strong>ce <strong>the</strong><br />
h-sound will occur not only after voiceless stops. How will this affect <strong>the</strong><br />
phonologist’s <strong>in</strong>terpretation of h-sounds?<br />
Insights <strong>in</strong>to this question may be obta<strong>in</strong>ed from, for example,<br />
Bowden <strong>and</strong> Hajek’s (1999) analysis of Taba, an Austronesian language<br />
spoken <strong>in</strong> Indonesia. In Taba, an aspiration that follows a stop is most<br />
appropriately viewed as an <strong>in</strong>dependent segment <strong>and</strong> not as a property of<br />
<strong>the</strong> stop. This is unusual, <strong>and</strong> its reason<strong>in</strong>g may be that <strong>the</strong> h-sound <strong>in</strong><br />
Taba can occur <strong>in</strong> various postconsonantal positions, due largely to <strong>the</strong><br />
prefix<strong>in</strong>g of verbs with simple consonants. Apparently, <strong>the</strong>se prefixes<br />
orig<strong>in</strong>ally had a CV structure, but <strong>the</strong> vowel has been lost. Thus <strong>in</strong> Taba<br />
we f<strong>in</strong>d words such as [mhonas] ‘sick,’ [nhik] ‘bat,’ [khan] ‘I go’ <strong>and</strong><br />
even [hhan] ‘you (pl.) go’ (Bowden <strong>and</strong> Hajek’s transcriptions). Given<br />
this distribution for h-sounds, a phonological <strong>in</strong>terpretation by which a<br />
sequence of a stop <strong>and</strong> an h-sound build a cluster of two separate phonological<br />
units is viable <strong>and</strong> appropriate.<br />
We now come to <strong>the</strong> question of whe<strong>the</strong>r preaspiration is reversed<br />
postaspiration. Thrá<strong>in</strong>sson (1978) argues that s<strong>in</strong>ce preaspiration <strong>in</strong> Icel<strong>and</strong>ic<br />
has normal segment length (i.e., is not that different from, for example,<br />
[s] or [n] <strong>in</strong> its typical duration), it is “not simply <strong>the</strong> <strong>in</strong>verse of<br />
postaspiration.” For this reason, Thrá<strong>in</strong>sson suggests that preaspiration <strong>in</strong><br />
Icel<strong>and</strong>ic should be given “segment status.” Consequently, Thrá<strong>in</strong>sson<br />
opposes <strong>the</strong> view that preaspiration is a “component” or a “phonetic<br />
feature” of <strong>the</strong> stop that follows. However, <strong>the</strong> question itself is imprecise.<br />
Are we ask<strong>in</strong>g whe<strong>the</strong>r preaspiration is an exact phonetic mirror<br />
image of postaspiration? Or whe<strong>the</strong>r preaspiration <strong>and</strong> postaspiration are<br />
phonological correspondences? Or both?<br />
Let us consider <strong>the</strong> phonological aspects. As to whe<strong>the</strong>r preaspirations<br />
are components or features of <strong>the</strong> follow<strong>in</strong>g stop, this is only relevant<br />
if we have a strictly segmental view of phonology. In a less segmental<br />
view, <strong>the</strong> question of whe<strong>the</strong>r or not preaspiration “belongs to”<br />
<strong>the</strong> stop is less of an issue. <strong>Preaspiration</strong>s are used as one of <strong>the</strong> phonetic<br />
correlates that signal <strong>the</strong> difference between two k<strong>in</strong>ds of vowel + stop<br />
sequences. As for phonotactics, it has been argued that preaspiration <strong>and</strong><br />
– 15 –