Rit (Vísindafélag Íslendinga) - 01.06.1970, Síða 106

104
But that the change u>o (and i>e) was, nevertheless, an
aspect of the same historical process as the change e>i (by
z-umlaut or otherwise) is proved beyond doubt by the fact
that when we compare the conditions for the change e > i with
those governing the subphonemic variation [u]: [o], we ob-
serve that they form part of the inverse of the conditions
for the change u>o; that is, they are among the conditions
under which u remained unchanged. Thus, u remained not
only before an i in the following syllable (e.g. ON burþr ‘birth’,
OHG gi-burt, OE ge-byrd<*burði-), but also before a nasal +
a consonant or a consonant +j, even when an a followed in
the next syllable (cf., e.g., OHG joh ‘yoke’ vs. hund ‘hundred’ <
*juka vs. *xunda, ON lofa ‘to praise’ vs. dylja ‘to hide’ < *lobðn
vs. *duljan), as well as before an u in the following syllable
(e.g. ist plur. pret. OHG ON lugum ‘we lied’, etc.); in these
four positíons, precisely, the change of e>i took place. In
other words, the same set of environments was among the
conditioning factors for the change both of e and of u, in a
positive and negative way, respectively (see Kurylowicz 1952:
52, H. Benediktsson 1967:182).
Thus, these two changes—the so-called a-umlaut and the
change of e> i—belong together as two instances of a single
process, the Common North and West Germanic reorganiza-
tion of short-vowel relationships, and should be separated
from the umlaut proper (that is, in particular, the z-umlaut).
The z-umlaut, which also took place in North and West Ger-
manic, was, in its first stage, the development of new (front)
variants of all nonfront vowels, short and long, or in other
words, a paradigmatic split of each nonfront vowel into a
front and a nonfront variant. In the second stage, the new
distinction was phonemicized as shown in §§ 1-2. The a-um-
laut, on the other hand, was part of a process (which may be
called a tendency, since it never reached its final stage) which
consisted in the abolishing of a distinctive opposition and in
the establishing, in its place, of an automatic distribution of
subphonemic variants. That is, umlaut proper was the crea-