Íslenskt mál og almenn málfræði - 01.01.1985, Síða 70
68
Bergljót Baldursdóttir
still considering that English, the competing language, lacks a gramma-
tical category of gender, this can be regarded as surprising resistance.
Moreover, the majority of the words which contain gender errors are
pronouns or adjectives. These words are different from the Icelandic
nouns in the way they do not have explicit concept of gender in them
but can have the various genders depending on the noun with which
they have to agree. In other words, the genders of these words are not
semantically based as are the genders of the nouns. The nouns them-
selves are very rarely assigned wrong gender. In 1982 only one word of
the „gender errors onIy“ category (see Table 1) is a noun. Similarly, in
1984 only one noun has only a gender error, the other nouns have also
either case or number errors. In 1982 the only word Baldur assigns
gender according to the English natural genders is in sentence 11 the
word bátur (masculine) which he calls bátið (neuter). On the other
hand, the results of the analysis of the code-switching data show that
most of the borrowed words are nouns. In 1982 46% of the code switch-
ing were single-noun switches, and in 1984 they were 68% of the code-
switching. When these words are compared to an Icelandic equivalent
word, it is seen that 81% of these nouns have Icelandic counterparts
which have genders which clash with the notion of the natural genders
in English, that is these words are masculine or feminine objects. This
means therefore that the great amount of single-noun switches, of
which most are potential gender errors, lowers the number of possible
gender errors. That is, these results suggest that when the Icelandic
word demands a serious breaking of Baldur’s sense of natural genders, it
can trigger off code-switching.
4.2 The contact
Although it can be said that Baldur shows a tendency to use the
words in the noun-phrase in a constant uninflected form, as is often
the case in English, and in general his Icelandic can be said to be more
like English than is the Icelandic spoken in Iceland, the changes can-
not be said to be overtly influenced by English or the contact itself. If
that was the case, one might expect that Baldur would frequently leave
e.g. adjectives uninflected or use them with a favoured ending. But
contrary to what might have been expected, there are not many inap-
propriately inflected adjectives in the data and inappropriately in-
flected adjectives are fewer in 1984 (8.9%) than in 1982 (18.8%).