For instance, a positive peak of activity (often referred to as P1 or P100)
will always follow the presentation of a visual stimulus over occipital regions
of a scalp. Its latency will essentially always be in the range of 90 to 110 ms
after stimulus onset. Auditory stimuli with a sharp rising onset will essen-
tially always elicit an N1 or N100 peaking at central electrodes (top of the
head). Interestingly, P1 and N1 are susceptible of modulation by automatic
difference detection, thought to be pre-attentive: if any physical property of
a highly repeated stimulus changes, these peaks will be modulated, a phe-
nomenon called mismatch negativity (MMN, see Näätänen, 1992, for a
review of MMN findings in the auditory domain). Many other peaks with
more or less subtle functional interpretation have been described so far.
Among them, the N2 is thought to be an index of response inhibition and
interference (N2 ref), the P3 indexes shifts in attention (spontaneous or trig-
gered by a low probability target stimulus (e.g. Donchin & Coles, 1988), the
N400 famously indexes semantic integration effort/difficulty (e.g. Kutas &
Hillyard, 1980), and the P600 indexes stimulus re-evaluation (e.g. Osterout
& Holcomb, 1992).
ERPs in the study of language production in bilinguals
As explained above, ERPs require high-precision time-locking to the
onset of a stimulus presented to the participant and reflect the cognitive
process which is elicited by the stimulus. The high temporal resolution of
ERPs allows us to investigate the different stages of language comprehen-
sion and production on a time-scale which is compatible with that of
neural function, that is the millisecond range. Although the number of ERP
studies on bilingual language processing has increased significantly in the
past 10 years, only a limited number of studies have examined speech pro-
duction in bilinguals (e.g. Chauncey et al., 2009; Christoffels et al., 2007;
Ganushchak & Schiller, 2009; Guo & Peng, 2006, 2007; Hoshino & Thierry,
2011; Jackson et al., 2001; Misra et al., 2012; Rodriguez-Fornells et al., 2005;
Strijkers et al., 2010; Verhoef et al., 2009a, 2009b).
Even though studying language comprehension with ERPs comes with
its own challenges (e.g. controlled delivery and standardization of auditory
stimuli, imposition of word-by-word reading in visual sentence processing),
it has been particularly challenging to study speech production, because
muscle activity involved in speech articulation contaminates the EEG and
override the signals originating in the brain. To circumvent this technical
difficulty, some ERP studies on bilingual speech production have used go/
no-go paradigms in combination with implicit picture naming (e.g. Guo &
Peng, 2007; Rodriguez-Fornells et al., 2005). In this paradigm, bilinguals are
typically asked to decide whether the initial letter of the name of the pre-
sented picture is a consonant or vowel and to make a button-press if the
initial letter is a consonant and withhold their response if it is a vowel (or
208 Part 5: The Bilingual Brain