Íslenskt mál og almenn málfræði - 01.01.1983, Side 119

Learning about -ari
117
tic data I have collected from three Icelandic preschool children over
a period of eighteen months, plus other examples contributed by my
students, there are some, but relatively few, spontaneous uses of -ari
as an innovative derivational suffix compared to uses of other proces-
ses, especially compounding.3 Yet compounding is a raíher minor pro-
cess in the Icelandic experimental data. Clark and Hecht (1982:15)
make a similar observation as regards naturalistic data from English,
when they point out that zero derivation is a common process in chil-
dren’s spontaneous instrument coinages, but absent in their experi-
mental data.
It is not too surprising that many children favored -erj-ari as a deriva-
tional device in this experiment to the exclusion of other processes
which they might use spontaneously since -er/-ari was repeatedly mo-
delled to them throughout the task. This modelling happened every
time we asked the child to give the verb base of a unfamiliar agent
or instrument noun formed with -er/-ari, in all twenty times per child.
In a related English study (Hecht, Clark and Mulford 1981) where
-er derived nouns were only one of several possible responses which
we modelled, the proportion of -er nouns that the children produced
relative to other types of responses was lower than in either the English
or Icelandic version of the current study. It can be argued that a child
who has not yet learned that -er/-ari is an agentive and/or instrumental
suffix, that is, who has not yet made the analysis of -er j-ari nouns
into verb base plus derivational suffix, will not be able to apply the
suffix to verbs to create new nouns, even with intensive modelling.
It is far from clear, however, how such modelling may influence children
who recognize the suffix but may not yet be using it much, if at all,
themselves. While intensive modelling of a form may or may not have
a longterm effect on children’s spontaneous productions (Whitehurst,
Ironsmith and Goldfein 1974), its influence on the responses elicited
in a particular experimental context may be substantial and must be
keptinmind.
3 Among the examples of spontaneous uses of -ari are both agents and instruments.
For example, one girl (age 3;10), after turning on the light (að kveikja) announced:
„Mamma, ég er kveikjari. Ég er alltaf að kveikja." ‘Mamma, I’m a ligther. I’m always
turning on the light.’ (I thank Margrét Pálsdóttir for this example.) In another case,
a boy (2;9) who had momentarily forgotten the conventional word for ,,toaster“
(brauðrist, which he knew), called it ristari (<að rista ‘to toast’).