=Paper=
{{Paper
|id=Vol-1347/paper21
|storemode=property
|title=Electrophysiological correlates idioms comprehension: semantic composition does not follow lexical retrieval
|pdfUrl=https://ceur-ws.org/Vol-1347/paper21.pdf
|volume=Vol-1347
|dblpUrl=https://dblp.org/rec/conf/networds/CanalPVMC15
}}
==Electrophysiological correlates idioms comprehension: semantic composition does not follow lexical retrieval==
https://ceur-ws.org/Vol-1347/paper21.pdf
Electrophysiological correlates of idiom comprehension: semantic
composition does not follow lexical retrieval
Paolo Canala,b, Francesca Pesciarellia, Francesco Vespignanic, Nicola Molinarod,e & Cristina Cacciaria
a Department of Biomedical Sciences, Università degli Studi di Modena e Reggio Emilia, Italy
b NEtS Center for Neurocognition Epistemology and Theorethical Syntax, IUSS, Pavia, Italy
e Ikerbasque, Basque Foundation for Science, Bilbao, 48001, Spain
c Department of Cognition and Formation Sciences, Università degli Studi di Trento, Italy
d BCBL, Basque center on Cognition, Brain and Language, Donostia/San Sebastian, Spain
gling between memory retrieval and semantic
1 Introduction integration processes [e.g., Hoecks & Brower,
2014].
Idiomatic expressions, such as break the ice, are
pervasive in everyday communication. They are 2 The present Study
frequently co-occurring sequences of words with
a conventional meaning that is not derived from We carried out two Experiments in which
word-by-word semantic composition, but rather short and literally plausible idioms (e.g., break
can be retrieved as such from semantic memory. the ice), i.e. having a literal well-formed meaning
Idioms are often read faster compared to literal and a conventional meaning, were embedded in
sentences [e.g., Siyanova-Chanturia et al., 2011] literal or idiomatic contexts. Notably, materials
and also lexical decision times are faster on id- were designed in such way that the sentential
iom related words than on literal related targets context would constrain expectations on the up-
[e.g., Cacciari & Tabossi, 1988]. Recent EEG coming target words to a similar extent across
data further suggest that semantic composition conditions. By doing so we minimized the im-
processes of idiomatic constituents might be not pact of differential sentence constraints, known
fully engaged during comprehension [Rommers to elicit N400 effects, and we carried out a com-
et al, 2013]. Finally brain-imaging studies re- parison between sentences that were semanti-
ported stronger and more widespread activation cally well-formed and for which contextual ex-
of the language network when reading idioms pectations on upcoming words were always ful-
compared to non-idiomatic sentences [Zempleni filled. Experiment 1 used EEG measures as de-
et al., 2007; Lauro et al., 2008; Boulenger et al., pendant variable to investigate the time course of
2009], suggesting that idiom comprehension idioms comprehension and was followed up by
might involve more cognitive resources. From Experiment 2 in which a cross modal priming
these fragmented results, it is not clear yet how paradigm was implemented, in order to confirm
idiomatic semantic processing differs from literal the activation of the literal meaning of the idio-
semantic processing and this might be due to the matic constituents in both types of contexts.
paradoxical nature of idioms [e.g., Libben & Ti- On the basis of the previous ERP literature we
tone, 2008], which seem to be at the same time hypothesized that meaning retrieval processes
amenable of direct memory retrieval and word- would affect the N400 component [e.g., Feder-
by-word compositional analysis. meier, 2007]: more demanding retrieval proc-
The two main questions of the present re- esses should be associated to larger N400 effects.
search thus concern two aspects of idiom The debate about the role of the N400 in seman-
comprehension: one relates to how the meaning tic integration vs. retrieval mechanisms [see se-
of the whole is retrieved and integrated in the mantic unification processes in Hagoort & Van
sentence representation; the second relates to Berkum, 2007] makes it hard to exclude that the
what happens to word-by-word semantic N400 component is not associated with the se-
composition of the literal meanings of the mantic integration of the meaning of the whole;
expression: is it carried out or suspended? To however, given the available evidence on figura-
answer these questions we used EEG measures tive language processing, we could also expect
(with the analysis of Event-Related Potentials an effect on later occurring positivities, previ-
and oscillatory dynamics of Time-Frequency ously associated with metaphor (Late Positive
representations) because of their temporal Complex, LPC) [e.g., Coulson & Van Petten,
precision [e.g., Luck, 2014], and because of the 2002; Lai et al., 2009] or irony (P600) [Regel et
possibility of disentangling between memory
Copyright © by the paper’s authors. Copying permitted for private and academic purposes.
In Vito Pirrelli, Claudia Marzi, Marcello Ferro (eds.): Word Structure and Word Usage. Proceedings of the NetWordS Final
Conference, Pisa, March 30-April 1, 2015, published at http://ceur-ws.org
98
al., 2010] processing, or semantic pragmatic 1a) La maestra aveva notato che Nicola dis-
reanalysis (frontal Post-N400 Positivity) [e.g., turbava i compagni, ma la prima volta chiuse
Van Berkum et al., 2009; Molinaro et al., 2012]. un occhio e continuò la lezione.
Another result that has been previously reported (The teacher saw Nick was bothering his desk mate
in the ERP literature of idioms processing is the but for the first time she closed an eye (turned a
finding of an involvement of the P300 compo- blind eye) and kept on teaching.)
nent. The P300 is generally associated with cog-
nitive mechanisms of context update [Donchin & 1b) Alla visita oculistica Enrico, prima di leg-
Coles, 1988] or context closure [Verleger, 1988]: gere le lettere indicate sulla lavagna lumi-
Vespignani et al. (2010) found that the brain’s nosa, chiuse un occhio per valutare la mio-
electrical response to the correct idiom constitu- pia.
ent was different if recorded before or after the (At the Ophthalmological visit, before starting to
read the letters on the panel aloud Henry closed an
idiom recognition point (RP, e.g., prendere il
eye in order to evaluate his nearsightedness.)
toro per leRP … corna -- take the bull by theRP …
horns). The match to the correct idiom word was
1c) Giovanni ha rotto gli occhiali durante la
associated with an N400 reduction before recog-
rissa perché ha preso un pugno in un occhio e
nition, but the electrophysiological response led
gli sono caduti a terra.
to a P300 effect after the recognition of the id- (Jack broke his glasses during the fight because got
iom. Such effect would mirror a qualitative a punch in his eye and fell on the ground.)
change in readers’ expectations about upcoming
words, after the expression has been recognized.
We also expected to replicate Rommers et al 3.3 Procedure
(2013) results in the time-frequency domain of In Experiment 1 sentences were presented word-
the EEG. The authors observed a power increase by-word at the centre of the screen
in the upper gamma frequency band after the (SOA=600ms). In Experiment 2, contexts sen-
presentation of the expected target words in lit- tences were auditorily presented via headphones
eral but not in idiomatic contexts, supporting the until the last word of the expression. Targets that
hypothesis that semantic unification mechanisms could be related or unrelated to the literal mean-
are less engaged in idioms comprehension. ing of the last word of the expression were visu-
ally presented at the offset of the audio file.
3 Method
3.1 Participants 4 Results
380 students at Università degli studi di Modena Fig.1 Grand Average ERPs from a pool of 7 fron-
e Reggio Emilia participated to the study set up tal electrodes (AF3, AF4, F3, FZ, F4, FC1, FC2) in
to norm the experimental materials. 32 different which frontal PNP effects are usually reported
students took part in Experiment 1. 42 students (negative voltage is plotted upwards). Idiomatic
volunteered in experiment 2. condition (solid line), Literal condition (dashed
line) and Control condition (dotted line) are com-
3.2 Materials pared at the onset of the last word of the idiomatic
Experiment 1 materials were 90 idiomatic ex-
pressions of similar structure (VP+NP idioms)
embedded in sentences. Idioms were selected for
being highly Familiar and correctly paraphrased.
Three sentential contexts for each expression
were created so that the last word of the expres-
sion was highly predictable in the three contexts
(above 85% cloze probability). ERPs were time-
locked to the presentation of the first word of the
expression (W1), and epochs comprising W1,
W2 and W3 were extracted from the EEG. In
Experiment 2 a subset of 44 idioms was used.
99
expression (e.g., ice). Concerning the second experimental question
related to the composition of individual constitu-
ent words we argue that Experiment 2 showed
Experiment 1 showed that:
that the literal meaning of the last word of the
- No N400 differences emerged between
expression was at least accessed, and confirms
literal and idiomatic context, during the
other evidence supporting the idea that readers
processing of the three constituent words.
process the literal meaning of idiomatic constitu-
- Differences between Idiomatic vs. Literal,
ents (Boulenger, Shtyrov & Pulvermüller, 2012).
and Idiomatic vs. Control conditions
Moreover, the lack of N400 differences across
emerged during the presentation of the last
conditions and word positions, suggests that lexi-
word of the expression (e.g., ice), and oc-
cal retrieval processes similarly occurred in
curred in the 400 to 600 ms time interval.
literal and idiomatic contexts. However, the
- Consistently with Rommers et al (2013)
analysis of the frequency domain replicated
study, the Time-Frequency analysis of the
Rommers et al’s findings of a larger power in-
EEG revealed power differences in the
crease in the high gamma frequency band for
higher gamma frequency band (60-80Hz)
literal compared to idiomatic contexts, which,
between expressions embedded in literal
consistently with their interpretation, could sig-
vs. idiomatic contexts: no power increase
nal that word-by-word composition mechanisms
was associated with the idiomatic condi-
are less engaged in idioms comprehension.
tion.
Conclusions
Experiment 2 showed that: When presented with idiomatic expressions
- Target words related to the literal meaning readers retrieve the literal meaning of the con-
of the idiomatic constituents obtained stituent words. However, word-by-word seman-
faster lexical decision times with respect tic composition mechanisms are idling, and, only
to unrelated targets, regardless of type of at the end of the expression, a seman-
context. tic/pragmatic wrap-up of the idiom is carried out
to update the sentence representation.
5 Discussion
Concerning the question related to how the
meaning of the whole idiom is integrated in the
sentence representation, our results suggest that
integration mechanisms occur only upon presen-
tation of the last constituent word, when the
idiomatic expression has very likely been recog-
nized. On the last constituent, ERP differences
between idiomatic and literal contexts emerged
between 400 and 600 ms in frontal electrodes.
The timing and scalp distribution of the effect
suggest that it affected a positive component (the
frontal Post-N400 Positivity) occurring soon af-
ter the peak of the N400. These results could be
accommodated elaborating the framework pro-
posed by the Retrieval-Integration hypothesis
[Hoecks & Brower, 2014], which holds that se-
mantic - pragmatic integration processes are re-
flected in P600 like positivities. One possible
interpretation is that the observed frontal positive
shift might be part of a larger family of positive
components reflecting the engagement of a se-
mantic/pragmatic wrap-up mechanism that is
performed at end of the expression to assign a
full interpretation to the incoming input.
100
potential study. Memory & Cognition, 30(6),
Reference 958–968.
Siyanova-Chanturia, A., Conklin, K., & Schmitt, Lai, V. T., Curran, T., & Menn, L. (2009). Com-
N. (2011). Adding more fuel to the fire: An prehending conventional and novel metaphors:
eye-tracking study of idiom processing by na- An ERP study. Brain Research, 1284, 145–
tive and non-native speakers. Second Lan- 155.
guage Research, 27(2), 251–272.
Regel, S., Gunter, T. C., & Friederici, A. D.
Cacciari, C., & Tabossi, P. (1988). The compre- (2010). Isn’t It Ironic? An Electrophysiologi-
hension of idioms. Journal of Memory and cal Exploration of Figurative Language Proc-
Language, 27(6), 668–683. essing. Journal of Cognitive Neuroscience,
23(2), 277–293.
Zempleni, M.-Z., Haverkort, M., Renken, R., &
A. Stowe, L. (2007). Evidence for bilateral in- Berkum, J. J. A. V., Holleman, B., Nieuwland,
volvement in idiom comprehension: An fMRI M., Otten, M., & Murre, J. (2009). Right or
study. NeuroImage, 34(3), 1280–1291. Wrong? The Brain’s Fast Response to Morally
Objectionable Statements. Psychological Sci-
Lauro, L. J. R., Tettamanti, M., Cappa, S. F., &
ence, 20(9), 1092–1099.
Papagno, C. (2008). Idiom Comprehension: A
Prefrontal Task? Cerebral Cortex, 18(1), 162– Molinaro, N., Carreiras, M., & Duñabeitia, J. A.
170. (2012). Semantic combinatorial processing of
non-anomalous expressions. NeuroImage,
Boulenger, V., Hauk, O., & Pulvermüller, F.
59(4), 3488–3501.
(2009). Grasping Ideas with the Motor Sys-
tem: Semantic Somatotopy in Idiom Compre- Donchin, E., & Coles, M. G. H. (1988). Is the
hension. Cerebral Cortex, 19(8), 1905–1914. P300 component a manifestation of context
updating? Behavioral and Brain Sciences,
Rommers, J., Dijkstra, T., & Bastiaansen, M.
11(03), 357–374.
(2012). Context-dependent Semantic Process-
ing in the Human Brain: Evidence from Idiom Verleger, R. (1988).Event-related potentials and
Comprehension. Journal of Cognitive Neuro- cognition: A critique of the context-updating
science, 25(5), 762–776. hypothesis and an alternative interpretation of
the P300.Behavioral and Brain Sciences,11,
Libben, M. R., & Titone, D. A. (2008). The
343–427.
multidetermined nature of idiom processing.
Memory & Cognition, 36(6),1103–1121. Vespignani, F., Canal, P., Molinaro, N., Fonda,
S., & Cacciari, C. (2009). Predictive Mecha-
Luck, S. J. (2014). An Introduction to the Event-
nisms in Idiom Comprehension. Journal of
Related Potential Technique. MIT Press.
Cognitive Neuroscience, 22(8), 1682–1700.
Hoeks, J. C. J. and Brouwer, H. (2014).
Boulenger, V., Shtyrov, Y., & Pulvermüller, F.
Electrophysiological Research on
(2012). When do you grasp the idea? MEG
Conversation and Discourse Processing. In:
evidence for instantaneous idiom understand-
Holtgraves, T. (Ed.), Oxford Handbook of
ing. NeuroImage, 59(4), 3502–3513.
Language and Social Psychology, pp. 365-
386. New York: Oxford University Press.
Hagoort, P., & Berkum, J. van. (2007). Beyond
the sentence given. Philosophical Transac-
tions of the Royal Society B: Biological Sci-
ences, 362(1481), 801–811.
Federmeier, K. D. (2007). Thinking ahead: The
role and roots of prediction in language com-
prehension. Psychophysiology, 44(4), 491–
505.
Coulson, S., & Petten, C. V. (2002). Conceptual
integration and metaphor: An event-related
101