Íslenskt mál og almenn málfræði - 01.01.1985, Page 62

60
Bergljót Baldnrsdóttir
There are 42 words (76.4%) altogether in 1982 which have an inappro-
priate case. 10 of these 42 have more than one grammatical category in-
correct (case, number or gender). In 1984, 62 words (79.5%) have an in-
appropriate case ending, of which 14 have more than one category in-
correct. Only one word of all the data has all three categories incorrect.
In 1982, the errors which involve gender are altogether 17 (30.9%).
Thereof are 7 which have just inappropriate gender and 10 which in-
volve gender and case, or gender and number. In 1984, the overall
gender errors are 19 (24.4%). Thereof 12 which have only inappropri-
ate gender.
Considering the amount of errors, the case category appears to be the
weakest category, that is weaker than gender and number. These results
are in line with the results Dorian (1981:418) found in her studies of
East-Sutherland Gaelic. She found case to be weaker category than
gender. The number distinction seems to be the strongest and very
rarely are there number errors unless there are also case or gender errors
(see Table 1). There does not seem to be much difference between the
two timepoints, neither in amount of errors nor in their distribution
across categories. (If the total number of errors per minute is calculated,
in 1982 there are 1.13 errors per minute, and in 1984 1.15 errors per
minute.)
3.1 Case infleclions
On the other hand there is a difference between the two timepoints in
the distribution of case errors across the four lcelandic cases. In 1982
the use of case is more widespread across the four cases although the
nominative is used the most (47.5%) then the accusative (25%) and
third the dative (22.5%). In 1984 the nominative and accusative forms
are used most of the time; 58.7% of the errors are characterized by that
a nominative form is used when the other cases should have been used,
and 38% of the errors involve the use of an accusative form when other
case forms should have been used. Only 3.2% of the errors involve the
use of the dative. The table below shows the overall distribution of case
errors over the four Icelandic cases.
Already in 1982 Baldur showed some tendency to over-use the nomi-
native and accusative forms, but in 1984 he clearly seems to be aban-
doning the use of the Icelandic genitive and dative and using the