|
Abstract:
In contrast to English, overt morphological case marking is
omnipresent in German. It is considered to be an important
processing feature. Behavioral and electrophysiological measures
were used to investigate the processing ofmorphological structural
and inherent case in sentence comprehension. The central
experimental manipulation consisted of a variation of the object
case encoded in the complement structure of two sets of verbs.
Accusative (structural) and dative case (inherent) were used in a
theory-driven approach.
In a cross-modal naming experiment participants listened to a
sentence preamble and named a sentence-completing participle verb
(1a & 1b, ISI 0 ms). In a counterbalanced 2x2 design the object
and the verb case requirements could either be congruent or
incongruent. Grammatical acceptability judgments served as
secondary task. The experiment revealed a case congruency effect.
The effect was replicated without the secondary task. This
post-lexical sentence integration effect demonstrated that
processing of the critical case feature takes place fast enough to
influence naming of the critical word itself with mean RTs in the
500 to 600 ms range.
Characteristics and temporal course were further investigated
using event-related brain potentials (ERP). As in the RT
experiments overtly accusative- or dative-marked personal pronouns
preceded the case assigning verbs in OSV word order (2a & 2b).
In a counterbalanced 2 x2 design half of the sentences were
syntactically correct (incorrect). An electroencephalogram (EEG)
was recorded while participants read 200 sentences in a
word-by-word SVP setting at 500ms a word. Grammatical acceptability
judgments were used to ensure sentence reading.
The ERPs for the verb participles showed more negative going
deflections of the ERP wave forms in the 480 to 780 ms time range
after onset of the critical verb. A late positivity was obtained
for incongruently presented verbs featuring structural case.
To test specificities of two different sources of information an
OSV to VSO word order variation was employed (3a & 3b).The
procedure was similar to the previous experiment. Both
syntactically incorrect conditions elicited more negative going ERP
wave forms in the 300 to 480 ms time range after stimulus onset,
that were broadly distributed and non-lateralized. Violations of
structural accusative case resulted in a larger negativity than
dative. There was no reliable P600.
The data revealed the time window of the processing of case
information marked on pronouns in VSO word order. The difference in
magnitude of the violation effect in both case conditions indicated
that the violation is more severe for structural case.
In sum, the present study revealed processing characteristics of
structural and inherent case along with its time course. In
particular, RT and ERP results converged. Fundamental differences
in the processes of structurally determined versus inherently
determined case in German were demonstrated. The case information
coded in the verb's complement structure controls processing. In
addition, the theoretical linguistic concept of structural versus
inherent case was supported.
Material examples:
(1a) Wen (*Wem) hat Maria ermordet?
Whom (*To whom) has Maria murdered?
(1b) Wem (*Wen) hat Maria zugenickt?
To whom (*Whom) has Maria nodded?
(2a) Er wusste, wen (*wem) Nina gesehen hat, bevor sie ging.
He knew, whom (*to whom) Nina seen has, before she left.
(2b) Er wusste, wem (*wen) Nina geholfen hat, bevor sie ging.
He knew, to whom (*whom) Nina helped has, before she left.
(3a) Klaus wusste, sehen wuerde Nina ihn (*ihm) erst, wenn diese
ging.
Klaus knew, see would Nina him (*him) not until, when she left.
(3b) Klaus wusste, helfen wuerde Nina ihm (*ihn) erst, wenn diese
ging.
Klaus knew, help would Nina him (*him) not until, when she
left.
|
