Íslenskt mál og almenn málfræði - 01.01.2023, Blaðsíða 187

the form-meaning mappings which have been observed by linguists. Indeed, the
results of the studies reported on in the dissertation show a rich use of these
mappings in comprehension and production, implying that non-deterministic
patterns have to be integrated more clearly into broader linguistic argumentation.
In fact, these mappings appear in the semantic basis of the well-documented
robust non-default productivity of the Icelandic datives. This is apparent with
novel verbs, where dative direct objects are produced with verbs containing caused
motion semantics, such as dánlóda ‘download’, pósta ‘post’ and dömpa ‘dump’
(Maling 2002, Barðdal 2008, Jónsson and Thórarinsdóttir 2020, S.L. Sig urð -
ardóttir 2023). Despite this, dative indirect objects are surprisingly rare with
novel verbs in the language and non-nominative subjects almost non-existent.
Non-nominative subject productivity still is well documented in the context of
Dative Substitution, or Dative Sickness, a form of subject case variation (or on -
going language change) where dative typically replaces accusative with experi-
encer subjects (Svavarsdóttir 1982, Jónsson and Eythórsson 2005, Barðdal 2011a
and 2011b, Thráinsson 2013 and Nowenstein 2017 i.a.) as in (4):
(4) Hana langar í epli. → Henni langar í epli.
her.ACC wants in apple. her.DAT wants in apple.
‘She wants an apple.’
Dative Substitution (DS) is a relatively stable but stigmatized variant which spread
at the end of the 19th century (but see Viðarsson 2022 for examples from Old
Icelandic) and has been extensively studied diachronically and in a series of large
surveys (e.g. Svavarsdóttir 1982, Jónsson 1997–1998, Jónsson and Eythórsson
2003 and 2005, Barðdal 2011a and 2011b, Thráinsson 2013 and Nowenstein
2017). Studies furthermore indicate that children’s rate of DS is associated with
their parents’ socio-economic status (Svavarsdóttir 1982 and Jónsson and Ey thórs -
son 2005). More recently, it has been emphasized that intra-speaker variation in
subject case marking is widespread and grammatically conditioned (Nowen stein
2012, 2014 and 2017, Svavarsdóttir 2013, Ingason 2015, Nowen stein and Ingason
2021) as is discussed in greater detail in Chapter 5 of the dissertation.
But how do children acquire this variable case marking system? As has been
mentioned, the acquisition of case in Icelandic is understudied. Previous re search,
most prominently H.Þ. Sigurðardóttir’s (2002) MA-thesis which drew on exper-
iments and corpora, has focused on production, showing that children acquire
the structural/default/most frequent case marking first, that is nominative on
subjects and accusative on objects. This happens early in acquisition, with the
two-word stage already including case marked arguments (Sigurjónsdóttir 2005).5
Dative objects appear from around age two, with indirect objects being acquired
earlier, and dative/non-nominative subjects appearing last, around age three.
Overgeneralizations are in line with this developmental path, as well as re search
on the acquisition of case in other Germanic languages (Schütze 1997, Eisenbeiss
Project rationale and core ideas 187
5 But the arguments during the one-word stage mostly appear in the nominative.