Íslenskt mál og almenn málfræði - 01.01.1980, Page 201

On the Dental Accretion
199
The basic variant of the imperative singular is the long imperative.
This assumption must be made in order to account for the fact that the
long imperative has prevailed over all other imperative variants in the
informal language, to a lesser extent in the formal language. By pre-
vailing over the remaining imperative variants the long imperative has
proved that it is more resistant than the remaining imperative variants
against „attacks“ upon itself. Its resistance must be due to the circum-
stance that the long imperative is a better linguistic sign, from the
niorphological point of view, than the remaining imperative variants.* * * 7
The „bettemess“ of the long imperative is in my opinion ascribable to
hs being better than the remaining imperative variants characterised as
a second person singular: the suffixed þú is the most universal Icelandic
2. p. sg. ending, and as such best of all the morphological signs denotes
the second person singular.8
The long imperative, which we have now established as the basic
imperative variant of Icelandic (from the morphological point of view)
on the evidence of its eventual rise above all other imperative variants,
has been the base from which the short imperatives gradually, and at
least optionally, came to be formed in the last thousand years or so in
tives can only be formed from the long imperatives that end in CDu, where C =
any consonant, D = t, d, or Ö\ the long imperatives kallaðu, þegiðu, etc. lack the
C of the formula.
7 This Darwinian reasoning owes much to Willi Mayerthaler’s lectures at the
Linguistic Society of America Summer Institute in Salzburg, 1979.
8 It is questionable whether the long imperative would win the title of the basic
ímperative variant if the contest among the imperative variants were conducted on
the syntactic level. In that case the contestants would include the short impera-
l've + the non-suffixed þú, e. g. kalla þú. However, it never came to a contest
among the imperative variants on the syntactic level: the type exemplified by kalla
þú did not disappear because it was less good a linguistic sign than kallaðu, but
because its integral part, the short imperative kalla, lost the battle against the long
imperative kallaðu on the morphological level. (For the difference between the
morphological and syntactic variants of the 2.p.sg. see footnote 3 above.) — That
linguistic changes can occur within one component only of the grammar, is com-
monplace: remember by way of example that sound laws can operate blindly and
create highly irregular morphological forms. The psychologically real, pronounced
compartmentalisation of grammars into components has been known for a long
time, and has now led Wolfgang U. Dressler to posit his polyzentristische Theorie
des Sprachsystems, see Dressler 1977:60 and passim.