Íslenskt mál og almenn málfræði - 01.01.1985, Síða 69

Language Shift in an Icelandic Child
67
of the time put forward by the preposition only, that is the preposition
and an uninflected noun. For example sentence (4):
(4) þar kannski er svona eldspýtur með brennisteinar (M.nom.pl.)
there maybe kind of matches with brimstone (sulphur)
Normally, the noun should be in dative brennisteinum as that is the
case the preposition með (‘with’) governs in this context. The fact that
in Icelandic the meaning is brought about by both preposition and in-
flection suggests that one of these is redundant. Moreover, other words
in the phrase than the noun carry inflections as well. That is, if the
noun stands in e.g. accusative the rest of the noun-phrase, i.e. adjec-
tives, adverbs, numerals, pronouns etc. need to stand in accusative as
well. The English noun-phrase is a good reminder of the redundancy of
these inflections, where adjectives, adverbs etc. never have to agree with
nouns and for example words like the demonstrative pronoun this only
has to agree with the number of the noun (this or these). It is therefore
not surprising to find out that Baldur has a tendency to use most words
in the phrase with one favoured suffix he seems to have chosen, and
leave the preposition to express the meaning. For example, sentence
(5):
(5) til þess að hita þessi litla gaffalar
to heat up these little forks
The demonstrative pronoun is in masculine/feminine, nominative,
singular, the adjective in feminine nominative singular and the noun,
although not perfectly formed, in masculine nominative plural. The
sentence should be til þess að hita þessa litlu gaffla, where all three
words are in masculine, accusative, plural.
4.1.3 Gender
The disagreement of the words in the noun-phrase and Baldur’s ten-
dency to use the words with single favoured suffix (ending), does not
affect the case category only. The gender category is also affected. Ac-
cording to Table 3, there are not many words which have gender errors
only. In 1982, they are 12.8% of the data and in 1984 they are 15.4%.
Considering the redundancy of the grammatical gender category and