Rit (Vísindafélag Íslendinga) - 01.06.1970, Page 372

37°
diíference can play an important part in the phonemic
structure of a word in the borrowing language. The same
applies to dialectal differences. Einar Haugen defines, in my
view with some oversimplification, a loanword as ‘a phoneme-
by-phoneme reproduction of the original’ (1956:762). This
does of course not mean that a given phoneme of the lending
language is always reproduced by the corresponding phoneme
of the borrowing language. However, this is, I think, the main
rule. But it must be admitted that we sometimes come across
exceptions to this rule. I shall take only one example. Old
English ea is in Old Icelandic manifested as a, not as ia, later
ja, as could be expected. The word harri ‘a chieftain, leader’
is fairly frequent in Old Icelandic poetry (Egill, Sighvatr,
Arnórr, etc.). This word is generally considered to be borrowed
from OE hearra, and I think this is the only possibility. We
could have expected *hjarri, which would have been in com-
plete accordance with the Old Icelandic linguistic system.
The same applies to stallari ‘a king’s marshal’, which also
occurs in Sighvatr’s poetry, and is borrowed from OE steallere.
In the latter case an association with the word stallr ‘a stable’
might have taken place. (However, borrowing from Anglian
is also possible.)
I think we must therefore modify the above-mentioned
main rule in such a way that in the borrowing language we
must expect to find, not necessarily the phoneme correspond-
ing to that of the lending language, but sometimes, instead,
a phonetically less closely related phoneme. Yet, some inter-
fering factors must not be forgotten.
The first interfering factor is what I want to call an ana-
logical phoneme substitution. I mentioned above that the
word keisari could not be a Middle Low German loanword
because of its age in Old Icelandic. It can hardly be borrowed
from any language other than Old Saxon, where it is késar.
According to the main rule mentioned above, we should
expect the word to have, not a diphthong, but a long é in
Old Icelandic. This, however, is not the case. We know that