Electro-physiology of language production: What are the questions and the challenges?

Chairs:   Marina Laganaro¹ &  F.-Xavier Alario²

¹ Université de Genève

² Aix-Marseille Université & CNRS

Abstract:

The investigation of psycholinguistic processes using electro-physiological methods has a long
tradition. In most of these studies, however, participants are requested to understand or decode
language (e.g. reading or listening to utterances). By comparison, language production research is
considerably underdeveloped. Such an underdevelopment is, partly, related to the artifacts induced
by overt speaking on the signals measured by imaging techniques, which led to a rather indirect
investigation of language production through the use of meta-linguistic paradigms.
Recently, this situation has started to change. An unexpected number of publications reporting EEG
studies using overt language production paradigms have been published. While this new trend can
benefit from the well established understanding of the electro-physiology of language
comprehension, the specificities of language production may call for specific research strategies.
The current symposium will examine whether this new thread of research is better described as
phasic transient response or as a more sustained tonic trend. More importantly, we shall discuss the
questions and challenges that face the electro-physiological investigation of language production
processes both from a methodological point of view and from a theoretical perspective.

Talk 1:

The use of electroencephalography in language production research: a review

Ingrid Christoffels
Universiteit Leiden

Speech production was long avoided electrophysiological experiments
because it was suspected that artifacts caused by muscle activity of
overt speech would lead to a bad signal-to-noise ratio in the
measurements. Speech production has been assessed with implicit speech
production tasks, such implicit or delayed naming. However, covert
speech is likely to involve qualitatively different processes than overt
speech. Recently, the number of papers that assessed overt speech using
electroencephalography (EEG) has been rising. There is an increasing
interest and demand for overt speech research within the field of
cognitive neuroscience of language. In this presentation we will review
available results of overt speech production involving EEG measurements,
such as picture naming, Stroop naming, and reading aloud. Although there
are potential problems and results should be interpreted with care.
Nevertheless, our review indicated that overt speech production can be
successfully studied using electrophysiological measures, for instance,
event-related brain potentials (ERPs). This new approach may provide
exciting new insight to studying speech production.

Talk 2:

Tracking the Time-course of Spoken Word Production with Event-Related Potentials

Phillip Holcomb
Tufts University

For over four decades Cognitive Neuroscientists has used event-related
potentials (ERPs) to augment the temporal resolution of traditional
behavioral measures such as reaction time in what is now a substantial
body of work aimed at providing a better understanding of perceptual,
cognitive and even linguistic processes. Used in this way ERPs offer the
advantage of allowing the investigator to explore both the fine-grained
time-course of information processing, but also to gain more direct
insight into the brain systems involved in processing. Such studies have
yielded important new insights into a host of mental and neural
phenomena. However, until recently one area of traditional language use
that has not been studied nearly as often using ERP techniques is
language production. The primary reason for this is that movement of the
articulators during speech production results in substantial electrical
artifact that makes examination of the comparatively small changes in
electrical brain activity very difficult. I will discuss this limitation
and its ramifications, as well as review a number of recent studies
that have attempted to circumvent speech artifacts during ERP language
production experiments.
   

Talk 3:

Towards a component-free, correlative approach to event-related potentials acquired during overt speech tasks: A more natural context to test language production.

Guillaume Thierry
Bangor University

In several recent event-related brain potential studies of language production, participants were
asked to produce overt speech while their EEG was being recorded. Whilst motor artefacts
engendered by the contraction of the jaw and that of facial muscles during speech are clearly
incompatible with EEG recording, the signal before speech onset seems relatively unaffected
(indeed it is unlikely that different experimental conditions would lead to systematic differential
effects on EEG signals prior to speech initiation). I will present data from two studies which have
used such a paradigm and show how relating ERP signals with voice reaction times can lead the
way to component free mental chronometry.
 

Talk 4:

Comprehensive spatio-temporal analysis of event-related potentials

Christophe Michel
Université de Genève

In this talk I will explain ERP analysis methods that are based on the temporal variations of the scalp electric field, rather than on changes of amplitudes or latencies at certain electrodes. The spatial analysis techniques have the advantage of being reference-independent and thus non-ambiguous and that they do not require a pre-selection of electrodes or time periods of interest. The basic aim of the spatial analysis approach is to look for significant topographic changes of the electric field across time or between conditions, because differences in topography directly indicate differences in the underlying generators. I will illustrate the method with ERPs in language comprehension as well as language production tasks.