Íslenskt mál og almenn málfræði - 13.07.1981, Side 224

214
Janez Oresnik
the results interpreted as follows. For speakers who can use the short
imp. stuð and spur, as they can use other forms of stuðja and spurja,
the imp. studdu and spurðu pertain to stuðja and spurja, and are not
parallel with keyptu, orktu. For speakers who do not use the short imp.
stuð and spur, or other forms of stuðja and spurja, the imp. studdu and
spurðu are of the same origin as keyptu, orktu, or are preserved due to
the forces that have brought about keyptu, orktu.)
The moral of the present section is an old one: when a morphological
form is investigated linguistically, it should not be observed in isolation,
but together with all the other forms cooccurring in the same inflexional
paradigm. With which other forms a form F cooccurs — or conspicu-
ously does not cooccur — in an inflexional paradigm is as important
as what shape F possesses.
2. Beside the regular imp. sg. ettu, the Icelandic verb etja ‘egg on’ pos-
sesses an irregular imp. sg. attu. The present section discusses the origin
of attu.
The oldest example of the imp. attu that I have seen in print occurs
in Páll Þorkelsson 1902:132, where the author describes the form as
used in the district of Skaftafell, South-Eastern Iceland. My informants
produce this form without hesitation, in response either to the question,
“What is the imp. sg. of etjaT’ or to the question, “Do you use the imp.
sg. attuT’ Thus the imp. attu is at least eighty years old, and is still
used in spoken Icelandic. The origin of this form has never been dis-
cussed in print, as far as I know.
I assume that the imp. attu contains the vowel of the pret. att- of etja,
and that this fact is to be explained in the same way as the correspond-
ing facts of the new imp. orktu, keyptu, etc. of Table (2). However, Jón
Helgason’s explanation adduced in the previous section cannot be
adopted for the imp. attu without change: the 3p. pl. pret. ind. of etja
is öttu, so that Jón Helgason’s explanation predicts the innovation imp.
sg. öttu, which does not exist.
Jón Helgason’s explanation has to be made slightly abstract to ac-
commodate both the new imperatives of Table (2) and the imp. attu.
Speakers must have noticed that, in a great number of verbs (most ian-
and cn-verbs), the part of the imp. sg. that precedes the final u is
identical with the dental stem of the verb. Consequently the rule for the
formation of the long imp. sg. became in such cases (4), q.v. Example: