also holds when knowledge of the two grammars is consolidated by early
acquisition.
This finding is particularly remarkable as the entire experiment was con-
ducted in English by experimenters who do not speak Welsh. Consequently,
bilingual participants were in an all-in-English context. In other words par-
ticipants in this study were in as close a monolingual ‘mode’ as we could
generate from external cues (Grosjean, 1998, 2001). This means that we cre-
ated the most adverse conditions for any activation of the non-target language
to take place: along the continuum between a monolingual and bilingual
context/mode, the monolingual mode, where only one language is being
spoken, is the most unlikely environment for activation of the non-target
language to occur. Despite this, our data indicate co-activation of Welsh
syntax during the processing of English sentences. We therefore propose that
highly fluent, early Welsh-English bilinguals have syntactic expectations that
are compatible with both their languages even when they are tested in a
monolingual context. Clearly, situation/context influences the amount of
interaction between the two languages, in language processing as well as
production (see, e.g. Marian, 2009). Our study reveals presence of such inter-
action in the most adverse situation possible, indicating that no environment
can eliminate cross-linguistic influence.
An important aspect of this study is that we were able to detect the
effects of a syntactic violation before the violation actually occurred, and
based on the modulation elicited by generic executive processes that are not
exclusive to language processing, namely decision making and response
withholding indexed by N2 amplitude across numerous experimental con-
texts. To our knowledge, among studies examining grammatical violations
in monolinguals or bilinguals (e.g. Friederici et al., 1996; Gunter et al., 1999;
Hahne, 2001; Weber-Fox & Neville, 1996) this is the first study which has
focused on speakers’ anticipation of a syntactic violation rather than the
brain’s response to the violation itself. As expected, the electrophysiological
response to the anticipation of a syntactic violation differs from the classical
response to the occurrence of the violation.
Our design and findings hold further implications that could poten-
tially call for a re-evaluation of previous findings on syntactic violations:
the absence of components linked to the processing of syntactic violations
in monolinguals (LAN, P600) from the ERPs of bilingual participants (e.g.
Hahne, 2001; Hahne & Friederici, 2001; Weber-Fox & Neville, 1996) could
be interpreted in at least two possible ways. Specifically, it could indicate
either that the bilingual participants are not sensitive to the violations (i.e.
they have difficulties in detecting violations) or that their brain is more
tolerant towards the violations. The consistent use of late bilinguals and
bilinguals with low proficiency in the target language in the relevant
studies favours the first interpretation: participants with incomplete
knowledge of the target grammar might have difficulties in detecting
228 Part 5: The Bilingual Brain