Íslenskt mál og almenn málfræði - 01.01.2016, Qupperneq 123

summary
‘On the dating of diphthongization in Old Icelandic’
Keywords: vowel system, sound change, diphthongization, Old Icelandic, manuscript spelling
The vowel system of early Old Icelandic was characterized by correlations of short and
long vowels with the same quality. Subsequently, short and long vowels were affected by
divergent quality shifts. Short vowels became more centralized than before and long non-
high vowels developed close offglides, é > [e], æ (> [æ]) > [a], ó > [o] and á (> [ɔ]) >
[a]. As a result, when Icelandic later underwent a quantity shift, whereby the distribution
of long and short vowels came to be determined by context, the distinctive function of
quantity was transferred to the new qualitiatve differences.
The centralization of short vowels apparently began in the late 12th century, affecting
front vowels first and back vowels only later, towards the mid 13th century. It is usually
assumed that the change in short vowels, the evidence for which mainly comes from
spelling of unstressed vowels, took place considerably earlier than the diphthongization of
long vowels. The paper presents evidence suggesting that the qualitative changes in long
and short vowels occurred at about the same time. This comes in the form of diphthongal
spellings of the front vowels é and æ, e.g. “ei” and “æi”, in manuscripts from as early as
around 1200, accompanied by a much larger number of monophthongal spellings, e.g. “e”
and “æ” for ei, interpretable as reverse spellings. The diphthongization of front long vow-
els must therefore have begun in the late 12th century. The earliest evidence for the diph-
thongization of á (and indirectly ó) is the spelling change “va” (for vá) > “vo” which occurs
in original dated documents from the first half of the 14th century. It is argued that this
does not preclude dating the diphthongization of the long back vowels to around the mid
13th century.
For some, the view that diphthongization of long vowels took place relatively late
appears to have received support from the belief that in poetry from the 14th to the 16th
century æ frequently rhymes with é [jɛː] (a variant which developed alongside and eventu-
ally replaced the falling diphthong [e] from é), because such rhymes would be impossible
had æ become [a]. A proposal is put forward to explain this rhyme-type — the frequency
of which has been greatly overestimated — in terms of a variant pronunciation of æ as a
sequence of the semi-vowel [j] and the vowel [æː] or [ɛː] (i.e. [jæː] or [jɛː]), for which there
is independent evidence from spelling in (primarily) 14th century manuscripts.
Aðalsteinn Hákonarson
Íslensku- og menningardeild Háskóla Íslands
Árnagarði við Suðurgötu
IS-101 Reykjavík, ÍSLAND
adh3@hi.is
Aldur tvíhljóðunar í forníslensku 123