Íslenskt mál og almenn málfræði - 01.01.1982, Page 314

312
Ritdómar
V:CC, VCC (where CC could also stand for a long consonant) — we have, in the
modern Ianguage, the patterns as discussed above where syllable length can be carried
either by the vocalism or the consonantism, never by both, and where a short stressed
syllable — *VC — can never emerge. Thus out of the four syllable shapes we have
now only two: V:C and VCC (or VC:). The historical question is the when, how and
why of the quantity shift.
The shift resulted as Arnason notes, quoting Benediktsson’s (1959) accurate state-
ment, in the „disappearance of the quantity correlation" (p. 122). The ensuing discus-
sion traces the development of the vowel system from about 1200 (after the First
Grammatical Treatise had been written). The evidence of metrical poetry is analysed
in an attempt to decide whether the shift was, to use Árnason’s unsuitably flippant
terminology, an „overnight change" or a conspiracy. The evidence seems to suggest
that the lengthening process extended for a period of almost two centuries before be-
coming stabilised in the language. Thus it was a gradual process in the triple sense
of the word (pp. 149-150) by being phonetically gradual, by spreading gradually across
the community and across the lexicon. In view of this, Árnason wonders if the question
of whether the change affected stressed syllables in monosyllables at the same time
as in polysyllables („overnight") or whether the two sets were affected separately (by
a conspiracy) „may be pointless" (p. 150). This, I think, is hardly true since gradualness
of a change and the context of the change are two, logically independent notions and
should, in theory, be independently decidable. They may not be for lack of conclusive
evidence but this does not mean that their separation is pointless. If this is also Árna-
son’s belief, then we agree on this point, but it was not clear to me that it is.
The shift itself had been prepared, so to speak, by two independent tendencies in
the history of Icelandic. First, Árnason assumes that long vowels may have been pho-
netically shorter before consonantal clusters or long consonants but this shortening was
non-distinctive. Secondly, long vowels started to diphthongise quite early so that long
and short vowels developed qualitative differences; once these differences took over
and became distinctive, the quantitative distinctions could be adjusted or reduced to
a rule-like regularity. In this way, the quantity correlation was eliminated from the
phonology of Icelandic.
All of this is quite plausible, in particular since we have the end products of the
process, i. e. stabilisation of quantity. What I find totally implausible is Árnason’s dis-
cussion of the troublesome clusters /p t k s + j v r/. Árnason once again involves
his idea of syllabification to account for the change; unfortunately, the syllable is as
helpless diachronically as it is synchronically: saying that there was a change to the final-
maximalistic stressed syllable — except in the case where the final-maximalistic prin-
ciples were disregarded — is just one way of describing correspondences, probably as
good as the traditional one. Arnason argues on structural grounds for one type of sylla-
bification but since the metrical evidence requires a different one, he falls back on
the notion of phonetic and phonological syllables and reconsiders the metrical evidence
by claiming that „this syllabification only existed for metrical purposes; in other words
it was imposed on the language only when it was used in poetic performance" (p.
158). But by playing around with terms like the phonetic and the phonological (and
even the metrical?) syllable — which are not defined or related to some specific reality