Íslenskt mál og almenn málfræði - 01.01.2009, Side 68

66
Heimir Freyr Viðarsson
This is further supported by a younger example cited by Halldór Halldórsson (1982) from
the early I7th century with the pattern dat-nom. Nominative substitution is, of course,
another possible explanation of the personal uses of skorta where it is used without the
dative. Examples of dative case with angra clearly have the case pattern dat-nom.
However, with harma, lysta and skipta, the theme is either not present or its case marking
is equivocal.
The semantics of alternative verbs involve a bidirectional causal structure, meaning
that it is possible to centralise either the dative or the nominative argument. With the
exclusion of harma and lysta, the verbs in question fulfil this criterion in that it is possible
to focus on the theme, leaving out the experiencer. This is generally not possible with the
ACC-ACC verbs prone to dative substitution assigning the role of experiencer/perceiver.
Although the thematic role assigned to the dative argument of alternating verbs is some-
times said to be experiencer, these verbs do not involve feelings or perception as the psych-
verbs most affected by dative substitution in Modern Icelandic. Again with the exclusion
of harma and lysta, the verbs do not involve feelings or perception, most commonly
referred to as experiencers. Rather, they seem to be benefactives, similar to the role
assigned by the alternating verb henta ‘suit’. It is possible to interpret the meaning of lysta
in the examples cited in chapter 4 as being similar to henta and líka and the dative could
therefore be analysed as a benefactive.
With angra and bíhaga, where the word order also clearly indicates an alternating argu-
ment structure, this analysis is fairly uncontroversial. If the analysis is extended to the
dative contexts with bila, bresta, harma, lysta, skipta, and skorta, it is not necessary to
assume any kind of dative substitution in the language of adults, which is at most sporad-
ically present in texts and apparently without any trace in later Icelandic (until the mid I9th
century). Instead, the dative is used in an underlyingly personal construction, correspond-
ing to a bidirectional argument structure denoting the meaning of a benefactive. If this was
indeed an option in Old Icelandic, the group of alternating verbs in Modern Icelandic may
be the result of a grammaticalisation of such a structure. This Iater restriction in the use of
the dative, now seen as a verb-specific property, would then be related to the increased cor-
relation between dative and the experience-based semantic class (cf. Jóhanna Barðdal
2009b), as well as the change towards a more fixed word order. A large number of struc-
turally ambiguous examples make it very difficult to decide whether this alternation
hypothesis should be preferred over dative substitution, a choice I leave for further
research. The two verbs, harma and lysta, are least convincing as alternating verbs and the
examples of datives may need to be analysed as dative substitution. However, the dat-
nom/nom-dat case marking pattern attested with bresta, klaja and þverra, later replaced
with an acc-acc case frame, does provide independent evidence in favour of a change in
an alternating structure along the lines proposed here.
Heimir Freyr Viðarsson
Faculty ofHumanities — Research group Dutch Linguistics
University ofAmsterdam (UvA)
Spuistraat 210
1012 VT Amsterdam, THE NETHERLANDS
h.f.vidarsson@uva.nl