|
Abstract:
In a series of investigations we have found a significant
modulation of the N100 restricted to left anterior regions,
associated with strategically unexpected Case and simple
NP/relative clause ambiguity resolutions in German. The modulation
is apparent on the disambiguating element, not at the point of
ambiguity. Further more, this modulation is evident for only those
conditions which in previous studies have shown reaction time
effects in a self-paced reading paradigm.
The N100 (a negative going deflection in the averaged
electro-encephalgraphic record peaking 100ms post stimulus onset)
is understood to be a marker of signal detection. It has a broad
bilateral distribution and is strongest over those cortical areas
associated with the stimulus processing, i.e., occipital regions
for visual stimulus, central temporal for auditory. Since it
reflects the detection of a stimulus the N100 is not expected to be
contrastive with respect to cognitive variables.
We report modulation of the N100 in the following
conditions:
On the verb for number disambiguation of Case ambiguities as in:
(1a) Welche Botschafterin besuchten die Richter
which(amb) ambassador(sg) visited(pl) the judge(pl)
(1b) Welche Botschafterin besuchte die Richter
which(amb) ambassador(sg) visited(sg) the judge(pl)
(1c) Der Professor fragte welche Botschafterin die Richter
besuchten
The Professor asked which(amb) ambassador(sg) visited(pl) the
judge(pl)
(1d) Der Professor fragte welche Botschafterin die Richter
besuchte
The Professor asked which(amb) ambassador(sg) visited(sg) the
judge(pl)
On the Case marked determiner immediately following the Case
ambiguous argument in embedded clauses as in:
(2a) Der Professor fragte welche Botschafterin den Richter
besuchte
The Professor asked which(amb) ambassador(sg) visited(sg) theacc
judge(sg)
(2b) Der Professor fragte welche Botschafterin der Richter
besuchte
The Professor asked which(amb) ambassador(sg) visited(sg) thenom
judge(sg)
And on the noun immediately following the determiner/relative
pronoun as in:
3a) Der Professor wusste dass die Managerin die Arbeiterinnen
gesehen hat
The professor knows that the manager has ssen the worker
(see for comparison Das ist die Managerin die die Arbeiterinnen
gesehen hat
that is the manager that the worker seen had )
No N100 modulation is found on the Case marked determiner
immediately following the Case ambiguous argument in matrix clauses
a in:
(4a) Welche Botschafterin besuchte den Richter
which(amb) ambassador(sg) visited(sg) the(acc) judge(sg)
(4b) Welche Botschafterin besuchte der Richter
which(amb) ambassador(sg) visited(sg) the(nom) judge(sg)
Nor is there N100 modulation on the noun following a determiner
in matrix constructions as in:
(5a) Welche Botschafterin besuchten die Arbeiterinnen
which ambassador visited the worker
We argue that this contrastive behavior of the N100 in left
anterior cortex is evidence of a grammatically driven strategic
deployment of language cortex in anticipation of relevant incoming
stimuli. The typically non-contrastive behavior of the N100 attests
to the fact that only the barest of cognitive processing takes
place in the first 100ms after stimulus presentation. We suggest
that the explanation for our observed effect lies in the parser
committing to a categorical decision before the relevant stimuli
has been received. Thus cognitive processes which dependent upon
properties of the as yet undetected stimuli are already in train
before the stimuli occurs. The modulation of the N100 then reflects
the match between the expected and the actual grammatical
features.
We discuss our findings with respect to expectation effects in
parsing theory and in the context of the relation between
electrophysiological measurements and reaction time results.
|
