Íslenskt mál og almenn málfræði - 01.01.1992, Page 96

94
Höskuldur Þráinsson og Kristján Árnason
126% (!) of the relevant age group (Guðfinnsson 1946:99). BG admits
that a part of the reason for this extremely high percentage is the fact
that sometimes a few younger children were included in the testing so
he ended up with a number which was higher than the total number of
children in the targeted age group in the relevant district.
Being somewhat restricted by his methodology and (lack of) equip-
ment, BG classified the pronunciation of each speaker into one of three
categories with respect to the dialectal feature under discussion: pure,
mixed or absent. Thus one speaker would be classified as having “pure
[xv] pronunciation” (cf. (3c) below), another as having “mixed [xv]
pronunciation” and a third as having no [xv] pronunciation at all. BG
then computed the percentage of each category for each district he vis-
ited and thus determined the geographical boundaries of the features
(cf. Guðfinnsson 1947:16-30; see also Dahlstedt 1958a, 1958b).
In RIN we have used somewhat diíferent methods. First, we divide
our subjects into 5 age groups: 1.12-20; 2.21-45; 3.46-55; 4.56-70;
5. over 70 years of age. We have concentrated somewhat on two of
the age groups, namely teenagers (12-20) and “middle aged” people
(46-55). The reason for this is the fact that we wanted to compare our
results to BG’s. Our youngest group corresponds to the age group he
tested but our middle aged group actually includes a sample of the
subjects he tested in the 1940s.
We have also tried to use a more varied testing method but most
of the data reported on is based on two methods: 1. the discussion
and naming of objects shown on pictures; and 2. reading of special
texts. We recorded every session on tape, of course, and our research
assistants and we ourselves have spent thousands of hours listening to
these tapes.
We2 have visited every inhabited part of Iceland (even islands off
2 When we (Höskuldur and Kristján) are discussing data collection and observations
“out in the field” in this paper, we frequently say that “we” observed something as if we
had both been present. This is not always accurate since different teams were involved
in the data collection, sometimes not even including either one of us. Much of the data
from Eastem Iceland was collected by Kristján and Zophonías Torfason, for instance,