Íslenskt mál og almenn málfræði - 01.01.2023, Blaðsíða 204

The example in (2a) illustrates an experiencer predicate resulting from abstract
goal semantics. The examples in (2) both involve an N-V complex predicate
where the light verb governs the case of the subject. With a motion verb like a
‘come’ in (2a), the subject is the abstract goal and thus results in a dative experi-
encer (Mohanan and Mohanan 1990). In contrast, an agentive light verb like kar
‘do’ calls for an ergative on the subject (cf. Mohanan 1994; Montaut 2016).
(2)a. nadya=ko kahani yad a-yi
Nadya.f.sg=dat story.f.sg.nom memory come-perf.f.sg
‘Nadya remembered a/the story.’ (lit.: ‘Memory of the story came to Nadya.’)
b. nadya=ne kahani yad k-i
Nadya.f.sg=erg story.f.sg.nom memory do-perf.f.sg
‘Nadya remembered a/the story.’ (lit.: ‘Nadya did memory of the story.’)
The dative-marked argument in (2a) can be conclusively analyzed as functioning
as a subject status via the various subject tests established by Mohanan (1994) for
Hindi. However, there is no evidence for dative subjects in Indo-Aryan prior to
the New-Indo Aryan developments as of about 1100 CE. With respect to this,
data from the Indo-Aryan language Marathi provides an interesting window
onto the path of historical development. Joshi (1993) first noted that it is very dif-
ficult to establish subject vs. object status for the dative arguments of a subclass
of Marathi verbs because the available subject tests (such as agreement patterns,
anaphora resolution, control into predicates, etc.) yield split results: some indi-
cate subjecthood, some do not. The verbs for which it is difficult to establish the
(non)subject status of dative arguments are: 1) nonvolitional transitives (e.g.,
‘find’, ‘like’); 2) passives of simple ditransitives (e.g., the passive of ‘give’); 3) pas-
sives of some causativized ingestives (e.g., causatives of ‘eat’, ‘drink’).
Deo (2003) shows that there are three different historical sources in Sanskrit for
dative subjects in Marathi: 1) intransitive verbs with an oblique goal (such as ‘go’,
‘occur’ or ‘shine’); 2) transitive verbs which exhibited the canonical nominative-
accusative pattern for subjects and objects, but have the lexical semantics of psy-
chological predicates (e.g., ‘think’, ‘perceive’, ‘know’, ‘recall’); 3) transitive verbs
with inanimate causers (such as ‘burn’ or ‘pain’). What unites these three types is
the absence of a prototypical agentive argument and one can chart a historical
development by which goal arguments are reinterpreted as experiencer subjects,
as in (2a). For a deeper look at how this historical change can be ac counted for in
terms of a reorganization of the linking between semantic predicate arguments
and grammatical relations, see Schätzle (2018) and Beck and Butt (2023). They
propose to understand this change in terms of a reanalysis of originally goal/loca-
tive sentient arguments as sentient holders of state, namely, experiencers. They
illustrate this not only with respect to Indo-Aryan and Urdu/Hindi in particular,
but also with respect to the formation of the new experiencer predicate finnast
‘seem’ in Icelandic. This predicate accounts for most of the occurrences of dative
Miriam Butt204