This study demonstrates how lexical spanning can be used to
explain the various stages in the acquisition of double definiteness in
Norwegian. The approach takes syntactic terminals to consist of submorphemic
elements that are lexicalised by words or morphemes. Work on
Scandinavian DPs has demonstrated that they involve two determiner-like
projections. Within the spanning approach, simple, unmodified structures in
Norwegian are distinguished from modified ones by the fact that one
morpheme spans both projections in the first case, while two morphological
exponents are required to spell them out separately in the second.
Furthermore, it has been argued that the term double definiteness is a
misnomer, and that the two determiners spell out separate subcomponents
of definiteness, Uniqueness and Specificity. For developmental reasons,
Norwegian children start out by lexicalising these features individually. The
suffixal article spells out Specificity, while a phonologically zero determiner
spells out Uniqueness. When adjectives are introduced into this grammar,
the result is a system that only spells out the suffixal article overtly, that is,
an Icelandic type system. This development is followed by a period with a great deal of instability as the grammar tries to determine how to lexicalise
the various terminals in the determiner phrase.
