Information Technology Reference
In-Depth Information
mostly have been interpreted as re
ecting some kind of conceptual priming
between words and non-linguistic stimuli,
ect linguistically
mediated effects. For instance, watching a picture of a bird or listening to a birdsong
might automatically activate the verbal label
they may also re
. Therefore, the conceptual
priming cannot be taken to be purely non-linguistic because of the implicit naming
induced by the processing of the stimulus. Such conceptual priming might imply at
least
“
bird
”
language, generation of auditory scenes, and mental
imaging, at various
associative (non speci
c) cortex area levels. This might probably activate large
neural/glial networks using long-distance synchronies, which could be investigated
by a synchronous EEG activity measurement (Lachaux et al.
1999
).
The aim of our
ö
n et al.
2010
) was to attempt
to reduce as far as possible the likelihood that a labelling process of this kind takes
place. To this end, we worked with a speci
first conceptual priming study (Sch
c class of sounds called
“
abstract
sounds
, which physical sources cannot be easily recognized, meaning that verbal
labelling is less likely to take place (Merer et al.
2011
). We then conducted con-
ceptual priming tests using word/sound pairs with different levels of congruence
between the prime and the target. Subjects had to decide whether or not the prime
and the target matched. In the
”
first experiment, a written word was presented
visually before the abstract sound, and in the second experiment, the order of
presentation was reversed. Results showed that participants were able to assess the
relationship between the prime and the target in both presentation orders (sound/
word vs. word/sound), showing low inter-subject variability and good consistency.
The presentation of a word reduced the variability of the interpretations of the
abstract sound and led to a consensus between subjects in spite of the fact that
the sound sources were not easily recognizable. Electrophysiological data showed
the occurrence of an enhanced negativity in the 250
600-ms latency range in
response to unrelated as compared to related targets in both experiments and the
presence of a more fronto-central distribution in response to word targets and a
more centro-parietal distribution in response to sound targets.
In a subsequent study (Aramaki et al.
2010
b), we avoided the use of words as
primes or targets. Conceptual priming was therefore studied using impact sounds
(also used in the categorization experiment previously presented), as both primes and
targets. As described in Sect.
4.5.1
, these impact sounds were quali
-
ed as either
typical or ambiguous with respect to a material category depending on their score in
the categorization experiment. 3
of congruence were investigated through various
combinations of typical and ambiguous sounds as prime and target: related, ambig-
uous and unrelated. The priming effects induced in these conditions were compared
with those observed with linguistic sounds (spoken words) in the same group of
participants. Results showed that N400-like components were also activated in a
sound
°
ect a search for meaning that
is not restricted to linguistic meaning. Moreover, ambiguous targets also elicited
larger N400-like components than related targets for both linguistic and non-lin-
guistic sounds. These
-
sound design. This component may therefore re
findings showed the existence of similar relationships in the
processing of semiotics of both non-linguistic and linguistic target sounds. This study
clearly means that it is possible to draw up a real language for non-linguistic sounds.
