Rit (Vísindafélag Íslendinga) - 01.06.1970, Page 137

J35
tion and frequency of phonemes already in existence deserve
equal attention in historical phonology.
Hreinn Benediktsson: In my report I discussed some of the
main principles of historical phonology and illustrated them
with individual problems from Germanic and in particular
Nordic linguistics. Most of these problems had been treated
earlier, by myself or others. But for some—such as the so-
called a-umlaut and related changes (§3.2), the relationship
between the i- and the a-umlaut (§4.3), or the z’-umlaut of 0
(§5.2)—attempts towards new solutions are made in the
report. The remarks put forth in the discussion call for the
following comments:
(1) Einar Haugen’s discussion of the analysis of the
development of Icelandic vowel quantity in the positions
represented by examples like sá, sú, sjá, búa (§3.5) centers
around the earliest stage of this development. In analysing
this stage we must not forget that, in addition to scribal
orthography and skaldic prosody, there is a third source of
evidence, the First Grammatical Treatise. In the treatise
there occurs, among others, a form like se (‘look’, imper.), the
vowel of which all ‘the behavior of Icelandic speakers in the
centuries that have passed’ would indicate to have been long.
Nevertheless, in the treatise, this example serves to illustrate
the distinction between short e and long é (se þo ‘look thou’
vs. séþo ‘(they) sewed’); in other words, in the position where
‘the contrast of quantity was neutralized,’ the vowel was
identified, not as long, but as short.
Scribal orthography, though on the whole intimately linked
with linguistic structure, bears different, more or less close,
relationships to the different levels of this structure. Thus, I
tried to explain some aspects of the observable fiuctuation in
the frequency of the superscript accent mark in words of the
type á, nú, fé, ná by relating it to the subphonemic (phonetic)
level (H. Benediktsson 1967/68:45-46). Morphophonemic and
morphemic factors are frequent, though they cannot serve as
an explanation in this case, since we have spellings such as