Íslenskt mál og almenn málfræði - 01.01.2023, Page 198

5. Acquiring and shaping case marking variation
In Chapter 5, I assess how general models of language acquisition account for the
learning of variation and how acquisition in turn shapes variation. This is done
with a detailed study of Dative Substitution (DS), the best known instance of
case variation in Icelandic. DS is a stigmatized but widespread instance of gram-
matically conditioned morphosyntactic variation which consists of dative being
extended to historically accusative experiencer subjects, as well as to two verbs
with historically nominative experiencer subjects. For children currently acquir-
ing Icelandic, it is fair to assume that all DS verbs appear with variable case mark-
ing in their language environment.9 But how do children acquire variation? Can
the learning process itself shape the directionality of change?
One of the fundamental questions within developmental sociolinguistics, and
language acquisition research more broadly, has to do with children’s reaction to
variability in their input, or primary linguistic data (e.g. Labov 1989, Yang 2002,
Hudson Kam and Newport 2005, Smith et al. 2009, Cournane and Pérez-Leroux
2020). Repetti-Ludlow and MacKenzie (2022) describe that at a glance, the liter-
ature points to a paradox: Children are both expected to diverge from and match
their caregivers’ speech. Despite children’s tendency to go beyond the input, with
e.g. overgeneralizations, we do expect children to learn their caregivers’ dialect,
and they have in fact been known to match the rates of variation found in their
environment (Labov 1989, Smith et al. 2009, Johnson and White 2020). When a
closer look is taken at the growing body of literature targeting children’s acquisi-
tion of variation, it becomes clear that the paradox is in fact attested in reality and
that both regularization and matching occur, but under different circumstances.
Indeed, the nature of the developmental path can depend on a number of factors
such as the learner’s age and the amount and consistency of exposure to different
dialects, but also the variable type (language domain), the complexity (or exis-
tence) of the conditioning factors and the social salience of the variable (e.g.
Smith et al. 2009, Hendricks et al. 2018).
The main questions of the chapter are therefore whether children acquire the
widespread DS intra-speaker variation found in the input or regularize it, and
whether they match their caregivers’ variant rate. The associations between the
dative and experiencer subjects in comprehension and production (Chapters 2
and 3) are put in the context of DS, with the corpus results in Chapter 4 addi-
tionally providing a context where variation is distributionally expected. Follow -
ing this, I describe the results of a study where DS was investigated in 101 chil-
dren aged 3–13 and their caregivers (80 dyads) by using forced-choice tasks and
grammaticality judgments across multiple items as a proxy for case use. The con-
ditioning factors investigated, the Person-Specific Retention (Nowenstein and
Ingason 2021) and less documented syncretism effects, are both instances of how
Iris Edda Nowenstein198
9 With hlakka ‘look forward to’ and kvíða ‘be anxious about’ (which historically take
nominative subjects) appearing with accusative and dative subjects.