Rit (Vísindafélag Íslendinga) - 01.06.1970, Side 142

140
independent one. Whether this principle is false, as Einar
Haugen maintains, or true cannot, in my opinion, be decided
at the present state of research. In my view, both alternatives
have to be kept open, as potential working hypotheses. At
any rate, this principle has never been falsified in the strictest
sense, in that, to the best of my knowledge, there is not a single
case where an external (or ‘extra-linguistic’) cause has been
proved to underlie a specific regular (immanent) phonological
change (where ‘extra-linguistic’ of course refers to factors
outside language but within the language community, i.e.,
does not include language contact). On the other hand,
external causes can well be posited alongside internal
factors, but this by no means renders the latter negligible.
On the contrary, assuming that external causes have to be
posited—and in view of the communicative and social function
of language, this is, a priori at least, by no means an unattractive
idea—the analysis of the internal, systematic, factors of change
would certainly be a necessary prerequisite for a successful
interpretation of the external factors, and can (or rather, must)
therefore be undertaken independently of, and even in advance
of, the analysis of the external factors. Thus, for instance, if it
is accepted, in the Germanic Consonant Shift, that the un-
voicing ofPIE b, d, g, gw is a direct systemic consequence ofthe
tensing of the PIE p, t, k, kw, it makes little sense to seek for
an external cause of the unvoicing independently of the tens-
ing. Likewise, we may be unable, in general, to answer the
question why an identical system changes differently in
diíferent geographical areas. But this is no excuse for refraining
from exploring the internal factors of change. Thus, nobody
would surely deny the systematicity of the Germanic Conso-
nant Shift, even though we cannot explain why it took place
only in this particular sub-area of Indo-European. Therefore,
the essential first step—which is of independent significance,
regardless of any following step—is an internal explanation,
in the sense of a reduction to a simpler internal pattern. In
the analysis of the development of the Icelandic vowel system,