Language is Embodied, Emotional, and Contextualized: Evidence from
Psychophysiological Studies of Bilingual Speakers

Catherine L. Harris, Ph.D.
Department of Psychology, Boston University


Julia Feld, Tufts University


To be presented at: Language, Cognition and Mind July 18-29, 2004, Portsmouth, UK

Bilingual speakers frequently report that use of their first language feels
more emotional than their second language. What is the physiological basis
of these subjective accounts? Personal emotional experiences have long been
assumed to be outside the scope of scientific research, especially those
that are subtle, variable across individuals and dependent on individual
learning experiences. The current paper takes a brain-based perspective on
the subjective experience of using language. Subjective experience can be
studied by comparing individuals' self-reported emotion with physiological
indices. The index used here is skin conductance, which corresponds to
sweating of the palms and finger-tips during emotional arousal. Prior work
has shown that skin conductance amplitudes are greater when bilinguals hear
emotional phrases in their first language compared to analogous phrases in
their second language.

The hypothesis investigated is that language comes to have a distinctive
emotional feel by virtue of being learned, or habitually used, in a
distinctive emotional context. The reason a first language is often
experienced as more emotional than a second language is because the first
language is learned in a context which is the most consistently emotional,
and least abstract, of any in human experience: the childhood context when
emotional regulation systems develop. This hypothesis was tested by
studying Mandarin-English and Spanish-English speakers who had learned their
second language (English) in early childhood, middle-childhood, or the
mid-teen years. Language was learned either via immersion, bilingual
school, or formal instruction. Participants were interviewed about their
language learning context. Skin conductance measures generally reflected
self-reported emotionality (the emotional intensity of a phrase and the
overall intensity of that language) and contexts of learning.
A case where emotional context of learning was reported to be intense even
though the language was learned late occured when L2 learners immigrate as
teenagers. Mandarin-English bilinguals reported English to be highly
emotional even when their first immersion experience was after age 18. Some
of these participants reported English to feel more emotional than Mandarin
because it is more acceptable to express emotional topics among
English-speaking peers in North America, than it is in Taiwan, Singapore or
Hong-Kong. Again, skin conductance amplitudes largely reflected these

These results establish that subjective experience has a quantifiable,
replicable physiological correlate. Findings raise the question of why the
emotional situation surrounding use of language becomes associated with
actual words and phrases. Most language learning theorists assume that
nonlinguistic correlates are stripped away during learning, allowing the
abstraction of linguistic meaning and context-independent grammatical rules.
In contrast to this view, the "emotional context of learning" hypothesis is
consistent with a radical contextualization view: language forms (words,
phrases, grammatical constructions), are always stored with their contexts
of use. Context-independence is the exception, not the norm. Learning more
about the extent to which language is embodied, rather than abstract, has
the potential to shed new light on diverse areas of language study.