cross-form conceptual relations between sounds and words: effects on the novelty p3
TRANSCRIPT
www.elsevier.com/locate/cogbrainres
Cognitive Brain Research 18 (2003) 58–64
Research report
Cross-form conceptual relations between sounds and words:
effects on the novelty P3
David Friedmana,*, Yael M. Cycowicza, Isabel Dziobekb
aCognitive Electrophysiology Laboratory, New York State Psychiatric Institute-Unit 6, 1051 Riverside Drive, New York City, NY 10032, USAbNeuroimaging Laboratory, Center for Brain Health, New York University School of Medicine, 560 First Avenue, New York City, NY 10016, USA
Accepted 10 September 2003
Abstract
In order for cross-form conceptual priming to occur, the brain must extract an amodal representation of the presented concept. To
determine whether the novelty P3 would show such cross-form effects, novel, environmental sounds or their verbal equivalents were repeated
two blocks after their first presentation in two cross-form conditions, word–sound (e.g., the word ‘‘pig’’ followed by the sound ‘‘oink’’) or
sound–word. Conceptual repetition engendered an asymmetric reduction in novelty P3 amplitude, i.e., amplitude was reduced in the sound–
word but not in the word–sound condition. The data suggest that the novelty P3 reflects an evaluative stage of processing in which some
semantic information is extracted. However, the lack of amplitude reduction for the word–sound condition implies that, at least at the delays
used here, repetition as a conceptually equivalent sound may have failed to make contact with the initial verbal concept.
D 2003 Elsevier B.V. All rights reserved.
Theme: Neural basis of behavior
Topic: Learning and memory
Keywords: ERPs; Novelty P3; Orienting response; Conceptual priming
1. Introduction
Orienting is a critical aspect of human behavior. The
orienting response [28] is an involuntary shift of attention
that is a fundamental biological mechanism necessary for
survival. The neurocognitive changes engendered by the
orienting response enable the organism to respond appro-
priately to a variety of familiar and unfamiliar environmen-
tal events (see Ref. [19] for a review). The orienting
response functions as a what-is-it detector and involves a
rapid response to new (never experienced before), unex-
pected (out of context) or unpredictable stimuli. The detec-
tion of a new or unexpected event precedes the orienting
response and induces the involuntary capture of attention,
enabling the event to enter consciousness. This permits an
evaluation of the salience of the stimulus. If the event is
deemed significant, this could lead to appropriate behavioral
action. Hence, in addition to sensory analysis, evaluation of
0926-6410/$ - see front matter D 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.cogbrainres.2003.09.002
* Corresponding author. Tel.: +1-212-543-5476; fax: +1-212-543-
6002.
E-mail address: [email protected] (D. Friedman).
a potentially relevant event also appears to involve some
kind of semantic analysis to extract meaning (see Ref. [19]
for examples). A key characteristic of an orienting response
is that it habituates with subsequent presentations of the
stimulus or stimuli that initially elicited it.
Modulations of the novelty P3 component of the event-
related brain potential (ERP) have been interpreted within
the rubric of the orienting response. Hence, the novelty P3
has been the subject of intense scrutiny in situations in
which the unexpectedness and novelty of stimuli have been
shown to engender an orienting response (reviewed in Ref.
[13]). This component, which has a latency to peak of about
320 ms, is elicited by infrequently occurring ‘‘novel’’ or
unexpected stimuli that are intermixed with frequently
occurring events within the context of an ‘‘oddball’’ para-
digm. These stimuli can be considered novel on the basis of
their unusual characteristics, such as weird, unexpected
environmental sounds presented among expected tones, or
on the basis of the context in which those events occur, for
example, infrequent tones presented for the very first time to
the subject [11]. The novelty P3 does not appear to be the
brain event that signals the detection of an unexpected
D. Friedman et al. / Cognitive Brain Research 18 (2003) 58–64 59
event, but rather reflects an evaluation of the novel or
unexpected stimulus. Several studies have demonstrated
habituation at frontal as compared to posterior scalp sites to
be a ubiquitous characteristic of the novelty P3 [4,5,12,16].
These data further support the hypothesis advanced by
Friedman et al. [13] that the novelty P3 reflects aspects of
the orienting response, and that it is comprised of frontal
and posterior subcomponents, each with different func-
tional characteristics (for a somewhat similar view, see
Refs. [10,31]).
Semantic habituation effects, also a characteristic of the
orienting response [19], have been demonstrated on the
novelty P3. For example, Cycowicz and Friedman [6]
presented novel, environmental sounds, along with fre-
quently occurring tones, in an oddball paradigm, in which
many of the environmental sounds repeated two blocks
after their first presentation. These investigators had sub-
jects rate the familiarity of these sounds, and then, based
upon those ratings, categorized the sounds into familiar and
unfamiliar classes. The key findings were that (1) repetition
of familiar sounds led to a reduction of the novelty
P3 over frontal scalp sites, but no significant difference
occurred over posterior scalp sites, and (2) for unfamiliar
sounds, there was no change with repetition over frontal
locations, but over posterior scalp sites, repetition led to an
enhancement of novelty P3 amplitude. Because novelty P3
amplitude appeared to be modulated by the familiarity of
the sound, this was interpreted to suggest that it must
reflect a fairly late stage of information processing, one that
might be close to the extraction of the sound’s semantic
attributes.
The current study was initiated to shed further light on
the issue of semantic habituation effects on the novelty P3.
To more definitively conclude that the novelty P3 is affected
by the semantic attributes of stimuli would require demon-
stration of cross-form habituation effects. That is, a reduc-
tion in novelty P3 amplitude should be demonstrable when
an environmental sound is repeated as its verbal equivalent
(e.g., the sound ‘‘oink’’ repeated as the word ‘‘pig’’) or
when the verbal equivalent is repeated as the conceptually
equivalent sound (the word ‘‘duck’’ followed by the sound
‘‘quack’’). This is not typical cross-form repetition, as for
example, would be represented by the visually presented
word ‘‘tiger’’ followed by the auditory version of ‘‘tiger,’’
which occurs between modalities. Rather, these kinds of
pairings reflect within-modality, conceptual relations be-
tween the word and the environmental sound (see Ref.
[30] for behavioral and ERP examples during tasks other
than the novelty oddball). That is, in order for habituation
effects to occur in these situations, the brain would have to
have access to an amodal, semantic representation of the
auditory concept. Hence, in the experiment reported here,
cross-form conceptual ‘‘repetition’’ of unique environmental
sounds (both word–sound and sound–word pairings) oc-
curred within the context of a novelty oddball paradigm.
Similar to previous studies from this laboratory, subjects
were not informed about the presence of either the sounds or
their verbal counterparts. The observation of cross-form
ERP reductions to conceptually equivalent word–sound or
sound–word pairings would suggest that environmental
sounds do, indeed, engender access to meaning. This would
be the case because an amodal, semantic, representation
would have had to have been generated in order for this kind
of reduction to take place. Because word–sound and
sound–word pairings are not identical in terms of both their
perceptual and semantic attributes, we expected that habit-
uation effects for these two types of pairings would not be
identical.
2. Materials and methods
2.1. Subjects
Sixteen subjects (nine female) between the ages of 20
and 26 (mean age = 21.7) were recruited by notices posted
within the Columbia Presbyterian Medical Center commu-
nity. All subjects reported themselves to be in good health
and to have no major medical, neurological or psychiatric
problems. All subjects signed informed consent, were
native English speakers and received payment for their
participation.
2.2. Stimuli
The stimuli were pure tones, environmental sounds and
their verbal equivalents. The pitches of the pure tones
were 1000 Hz (high tone) and 700 Hz (low tone), and
their duration was 336 ms. In a separate normative study,
65 undergraduates were asked to write down the name of
each of 180 sounds taken from sound effects compact
disks. Of these 180 sounds, 64 unique sound concepts
with the highest name agreements (i.e., the percentage of
the 65 subjects producing the same name) were retained
for the experiment. These sounds and their verbal equiv-
alents were then categorized into four lists, with 16 sounds
per list. There were no reliable differences in name
agreement (which varied from 83% to 91%), or intensity
(matched using the sound and word envelope character-
istics) among the four lists. Mean durations and ranges
were, for the sounds, M = 881 ms, range = 377–1000 ms;
for the words, these values were M = 709 ms, range = 384–
1000 ms. Additionally, the four sets of words and four sets
of sounds did not differ in intensity. Two of the lists were
assigned to the novelty oddball task (see below) to the
sound–word or word–sound repetition conditions, and
the other two were assigned as foils for the recognition
task. These list assignments were counterbalanced across
subjects. The sounds came from several categories, such
as animal, human, musical instrument and artificial
or machine sounds. All stimuli were delivered at 75 dB
SPL.
e Brain Research 18 (2003) 58–64
2.3. Procedure
The experiment consisted of three phases. The first was a
standard auditory oddball task, the second was a novelty
oddball task and the third was a recognition memory test
phase. The data resulting from the recognition phase will not
be reported.
2.3.1. Standard oddball task
Subjects were presented with high and low pure tones in
random order. One tone was presented 88% of the time and
was designated the standard and the other tone was pre-
sented 12% of the time and was designated the target. There
were two blocks of 100 trials each with an inter-stimulus
interval of 1500 ms.
2.3.2. Novelty oddball task
Following the standard oddball task, subjects were pre-
sented with 10 blocks of 80 trials each, again with inter-
stimulus intervals of 1500 ms. Novel stimuli (10%) were
randomly intermixed with standard (80%) and target (10%)
tones. The novels were either environmental sounds or their
verbal equivalents. Subjects were not informed of the
occurrence of the novel stimuli.
There were 32 different non-tonal, environmental sounds
and 32 verbal equivalents. Hence, 16 of each type of novel
event was repeated in its cross-form counterpart (i.e.,
sound1–word2; word1–sound2; these are hereafter referred
to as prime–repeat pairings). Each block contained eight
novel events. In the first two blocks, all of the novel events
were new, while in the rest of the blocks only half were new.
The eight sounds and eight verbal equivalents that did not
repeat (labeled ‘‘unique’’) comprised half of the novel items
in the first two and the last two blocks (four in each block).
These items were taken from environmental sound stimuli
that did not pass the name agreement criterion and are not
analyzed in the data reported below. Conceptual repetition
of the novel stimuli occurred two blocks after their initial
presentation, such that, for example, the novel stimuli
initially presented in the first block were repeated in the
third block. The repetitions were always cross-form. Stimuli
were randomized separately for each subject, with the
restrictions that a target or a novel could not occur as the
first or the last stimulus, and that two targets or novels could
not be presented sequentially.
For both tasks, subjects were instructed to press a button
(emphasizing speed) with the thumb of one hand as soon as
they heard the target tone. Hand of response and target tone
were counterbalanced across subjects. Reaction times be-
tween 200 and 1600 ms were accepted as correct responses.
2.4. EEG recordings
EEG (5-s time constant; 50 Hz upper cutoff; 200 Hz
digitization rate) was recorded (Sensorium amplifiers) con-
tinuously using an Electrocap (Electrocap International)
D. Friedman et al. / Cognitiv60
from 62 extended 10–20 system placements [21], including
left and right mastoids, all referred to nosetip. Vertical EOG
was recorded bipolarly from electrodes placed on the
supraorbital and infraorbital ridges of the right eye, and
horizontal EOG was recorded bipolarly from electrodes
placed on the outer canthi of the two eyes. Trials containing
eye movement artifact were corrected off-line using the
procedure developed by Gratton et al. [14]. Trials were
epoched off-line with 100 ms pre- and 1400 ms post-
stimulus periods.
2.5. Data analyses
The SPSS Version 10 for Windows (repeated-measures
ANOVA) computer program was used for all analyses. The
Greenhouse–Geisser epsilon (e) correction [15] was used
where appropriate. Post hoc testing was performed using the
Tukey Honestly Significant Difference test. Uncorrected
degrees of freedom are reported along with the epsilon
value; the P values reflect the epsilon correction.
3. Results
The data resulting from the standard oddball task were
entirely consistent with previous data recorded from this and
other laboratories. Because they are uninformative with
respect to the novelty P3, they will not be reported.
During the novelty oddball task, subjects were very
accurate in detecting the target (99%) with mean reaction
time of 431 ms, and very occasionally had false alarms to
standard (0.08%) and novel (0.9%) stimuli.
The range of the mean number (F S.D.) of the trials
entering the four depicted waveforms of Fig. 1 varied,
respectively, from 15.4 (0.6) to 15.7 (1.0). The figure
presents the grand mean ERPs associated with first and
second presentations of sounds and words in the two
prime–repeat conditions. The figure depicts the data at the
electrode locations that were analyzed in the ANOVAs
described below. The left three columns of Fig. 1 depict
the sound–word and the right three columns the word–
sound waveforms. The figure suggests that reduction in the
amplitude of the novelty P3 is larger for the sound1–word2
combination of prime and repeat.
A second positivity is also present in the ERPs depicted
in Fig. 1. This component has been observed previously [7]
and is most likely synonymous with the P3b. Because no a
priori predictions were made for this component and its
relation to conceptual repetition, these data will not be
discussed further.
To assess effects on the novelty P3 statistically, planned
comparisons were performed. Novelty P3 amplitude was
measured as an averaged voltage between 280 and 380 ms,
based on inspection of the grand mean ERPs. To capture any
lateralized and/or anterior/posterior effects, the following
electrode sites were considered: on the left, F1, FC1, C1,
Fig. 1. Grand mean ERPs averaged across the 16 subjects for the two cross-form conditions of the experiment: (left) ERPs in the sound1–word2 prime– repeat
pairing recorded at left, midline and right scalp sites; (right) ERPs in the word1–sound2 prime– repeat pairing also recorded at left, midline and right scalp
sites. Arrows mark stimulus onset. The SP and CSD maps for the two prime– repeat pairings are depicted below the corresponding display of waveforms. The
maps are based on the difference in novelty P3 amplitude (measured from 280 to 380 ms) between the first and second presentations in each prime– repeat
condition. The left two maps represent the sound1 minus word2 difference ERPs, and the second two, the word1 minus sound2 difference ERPs. Filled dots
represent the 62 scalp electrode locations. Shading reflects negativity; unshaded areas reflect positivity. SP D= 0.30 AV; CSD D = 0.02 AV/cm2.
Table 1
Results of the sound1–word2 and word1–sound2 ANOVAs
df F P value Epsilon (e)
Sound1–word2
Prime– repeat (P) 1.15 2.843 0.112
Lateral (L) 2.30 28.098 0.000 0.730
Anterior–posterior (A) 4.60 7.014 0.004 0.459
P�L 2.30 3.451 0.048 0.944
P�A 4.60 0.107 0.817 0.333
L�A 8.120 4.404 0.006 0.418
P�L�A 8.120 0.583 0.611 0.335
Word1–sound2
Prime– repeat (P) 1.15 1.385 0.258
Lateral (L) 2.30 21.528 0.000 0.691
Anterior–posterior (A) 4.60 8.886 0.003 0.366
P�L 2.30 0.318 0.701 0.871
P�A 4.60 0.296 0.707 0.418
L�A 8.120 5.614 0.001 0.479
P�L�A 8.120 0.634 0.627 0.460
Italicized F ratios are significant.
D. Friedman et al. / Cognitive Brain Research 18 (2003) 58–64 61
CP1 and P1; at the midline, Fz, FCz, Cz, CPz and Pz; and
on the right, F2, FC2, C2, CP2 and P2. These locations were
chosen based on our previous work, which showed a highly
focal midline scalp distribution and an anteriorly oriented
gradient for the novelty P3 (reviewed in Ref. [13]).
The two ANOVAs that contrasted the factors of Lateral
position on the scalp (left, midline, right), Anterior/Posterior
dimension (frontal, fronto-central, central, centro-parietal
and parietal) and Cross-Form Presentation (first, second—
e.g., sound1–word2) were performed separately on the
sound1–word2 and word1–sound2 prime–repeat condi-
tions. The results of these ANOVAs are presented in Table
1. The sound1–word2 data yielded reliable main effects of
Lateral position and Anterior/Posterior dimension. These
were modulated, however, by a reliable Lateral�Anterior/
Posterior interaction. Post hoc testing indicated that the
novelty P3 was largest at fronto-central scalp sites, with
this gradient steepest at the midline series of electrode sites.
Importantly, the Lateral�Cross-Form Presentation interac-
tion was also reliable. Post hoc testing indicated that,
although the differences between sound1 and word2 were
all reliable at left, midline and right scalp regions, the largest
reduction occurred at the midline (2.6 AV).For the word1–sound2 ANOVA, the Lateral and Ante-
rior/Posterior main effects were, again, reliable, and also
interacted. The Lateral�Anterior/Posterior reflected the
same phenomenon as in the sound1–word2 ANOVA. By
contrast with the sound1–word2 ANOVA, none of the main
or interaction effects involving the Cross-Form Presentation
factor was significant in the word1–sound2 ANOVA.
D. Friedman et al. / Cognitive Brain Research 18 (2003) 58–6462
Fig. 1 also depicts the surface potential (SP) and current
source density (CSD) scalp distributions (computed using a
third-order spherical spline interpolation algorithm [24]) of
the differences in novelty P3 amplitude between the first and
second presentations in the two cross-form conditions. The
left two maps represent the sound1 minus word2 difference
waveforms, and the right two maps, the word1 minus
sound2 difference ERPs. In support of the analysis de-
scribed above, a prefrontal positivity is present in the maps
for sound1–word2 prime–repeat subtraction (in both SP
and CSD distributions) but not in the word1–sound2
subtraction.
To determine if these differences were statistically reli-
able, the current source density difference mean data were
normalized using the root mean square method [20,25]. To be
compatible with the raw data analyses described above, the
waveforms were measured between 280 and 380 ms. The
data from all 62 scalp electrodes were used. The resultant
Condition (sound1–word2, word1–sound2)�Electrode
Location (62 levels) ANOVA on the normalized data returned
a reliable Condition�Electrode Location interaction
[F(61,915) = 4.11, P < 0.001, e= 0.11], indicating that the
scalp distributions of the two subtraction waveforms differed
reliably. Post hoc testing indicated that reliable differences
were obtained primarily over the frontal areas of the scalp.
1 We thank an anonymous reviewer for pointing this out to us.
4. Discussion
One way to ensure that sounds induce access to ‘‘mean-
ing’’ is to demonstrate that an environmental sound can
prime its verbal referent, or that the verbal representation
can prime its sound counterpart. Hence, in order for habit-
uation to have occurred, the brain would have had to have
access to an amodal representation of the concept, and this
would have had to have been available at least by approx-
imately 300 ms or so (the peak of the novelty P3). The
major finding was that reduction in novelty P3 amplitude
occurred only in the case in which a sound concept was
followed by its verbal conceptual equivalent (e.g., ‘‘oink’’
followed by the word ‘‘pig’’). By contrast, the converse,
‘‘pig’’ followed by the sound ‘‘oink,’’ did not lead to a
reduction in the magnitude of the novelty P3. This analysis
based on the raw data was also supported by the topographic
data. This latter analysis indicated that only in the former
situation was a prefrontal positivity present in the prime
minus repeat difference waveforms (see Fig. 1).
One possibility for the lack of reduction in the word–
sound prime–repeat condition is that, much like pictures
(e.g., Ref. [23]), sound concepts require processing on two
levels: perceptual and semantic. Verbal labels, on the other
hand, bear no perceptual relation to their sound equivalents.
Hence, when a verbal label is the conceptual repeat, as in the
sound1–word2 condition (one assumes), the verbal concept
contacts the semantic representation of the previously pre-
sented sound, engendering a reduction in amplitude. How-
ever, when a sound concept is the repeat, as in the word1–
sound2 pairing, both perceptual and semantic analyses must
ensue, leading to a greater amount of processing and hence
no reduction in novelty P3 amplitude.
In one scenario, subjects might intentionally label the
presented sound with its verbal equivalent, e.g., ‘‘pig’’ when
presented with the sound ‘‘oink,’’ or ask the question ‘‘Is
that the sound a pig makes?’’1 However, when the verbal
word ‘‘pig’’ is presented, participants would most likely not
say to themselves, ‘‘What sound does a pig make?’’ On this
view, in the case where the sound was presented first, the
verbal concept would have been activated, making contact
with the actual label when the conceptually equivalent word
was subsequently presented. On the other hand, when the
word was presented first, the equivalent sound concept (e.g.,
‘‘oink’’) would most likely not have been activated. There-
fore, when the conceptually equivalent sound was subse-
quently presented, there would have been a ‘‘mismatch’’ in
activated features (semantic and perceptual for the word–
sound pairing, respectively), thus failing to engender an
amplitude reduction in the novelty P3.
An alternative, but not necessarily contradictory, hypoth-
esis is that the sound in the word1–sound2 pairing does not
contain sufficient information to contact the stored repre-
sentation of the previously heard verbal label. This could be
due to the presumed slowed processing required by the
necessity to incorporate both the perceptual and semantic
characteristics inherent in the sound (see above). Or it could
be due to the possibility that the short durations of the
current sounds did not enable sufficient semantic informa-
tion to be extracted. In either case, the end result would be
the same—a failure to activate the stored representation of
the verbal concept.
Some evidence for these different viewpoints comes
from behavioral studies of environmental sound priming,
although the underlying mechanisms involved may be quite
different from those recruited in the novelty oddball para-
digm presented here. For example, Chiu and Schacter [3]
failed to find cross-form priming in a condition in which
sound labels were studied and sound fragments (the first
second of a 5-s sound) were used as test items. When the
verbal label and the sound were presented together during
the study phase, robust priming by the sound fragment at
test was observed. Similarly, Stuart and Jones [29] reported
that identification of an environmental sound was facilitated
by the repeated presentation of the same sound but not by
the previous presentation of its verbal label.
Although the database on non-linguistic sounds is not
nearly as large as that for verbal stimuli, there is some
evidence suggesting similar processing mechanisms for both
sounds and words. For example, sound familiarity is one
determinant of sound identification, in that commonly heard
sounds are easier to identify than less commonly heard
D. Friedman et al. / Cognitive Brain Research 18 (2003) 58–64 63
sounds [1], an effect that is similar to word frequency. Stuart
and Jones [29] demonstrated an analogy to the word
repetition priming effect using environmental sound stimuli
by demonstrating that the identification of environmental
sounds was facilitated by previous exposure to the same
sound (see also below). Furthermore, Van Petten and
Rheinfelder [30], who used meaningful non-speech sounds
and conceptually related words, demonstrated cross-form
priming of both words and sounds regardless of whether the
non-speech sound or its verbal conceptual equivalent served
as prime or target.
However, although repetition of conceptually similar
items is common to the current novelty oddball investiga-
tion and the sound priming studies mentioned above, there
are key elements that distinguish the two: (1) in the
current investigation, cross-form conceptual repetition oc-
curred at intervals longer than those typically employed in
semantic priming studies (respectively, approximately after
a 3-min lag contrasted with only a few sec); and (2) cross-
form conceptual repetitions occurred within the context of
a novelty oddball task (not a typical semantic priming
paradigm), in which subjects had to detect the infrequent
occurrence of a target event and were not expecting (at
least initially) environmental sounds and their verbal
counterparts. Typically, in behavioral studies of semantic
priming, facilitation in reaction time is short-lived relative
to repetition priming, surviving prime–target intervals of
only a few seconds [9]. Furthermore, the major effect of
semantic priming in ERP studies is to reduce the ampli-
tude of the N400 component [26], an ERP waveform
peaking at approximately 400 ms, about 100 ms later than
the novelty P3 (see Ref. [22]). Together, these findings
suggest that the reductions in amplitude of the novelty P3
observed here reflect an earlier and fundamentally different
mechanism than those observed in semantic priming
studies.
Based on a variety of evidence, the putative neural
mechanism(s) underlying the novelty P3 appears to be
localized to the prefrontal cortex. These studies demonstrate
that the novelty P3 is reduced in patients with dorsolateral
prefrontal lesions when compared to normal controls [8,16],
or patients with lesions elsewhere within the brain [18].
Additional evidence for a prefrontal cortical contribution to
this component comes from intracranial ERP investigations
[2], current source density mapping studies of surface
recorded activity [11,27] and event-related fMRI studies
(see Refs. [17,13] for a review of the latter).
Friedman et al. [13] have argued that the novelty P3 does
not reflect the detection of the deviant event per se, but
rather processes that are activated after the deviant event has
been detected, for example, bringing the event to conscious-
ness for evaluation of its salience and appropriate action. In
the current data, the reduction of the novelty P3 when the
initially presented sound concept is repeated as a verbal
label is consistent with this hypothesis, as processes sub-
sumed under orienting should be called upon to a lesser
extent once a representation of the event has been formed.
The environmental sounds used here are most likely repre-
sented in semantic memory networks similar to those
observed for words and pictures. Although the data are very
limited, some support for this contention has been provided
by experiments in which similar environmental sounds are
embedded within semantic priming and repetition priming
tasks as described earlier for both behavioral [3,29] and ERP
[30] investigations. However, a more complete neurocogni-
tive picture and more definitive conclusions will emerge
only with additional research effort.
Acknowledgements
The research reported here was supported in part by grant
HD14959 (DF) and HD37193 (YMC) from NIH and by the
New York State Department of Mental Hygiene. We thank
Charles L. Brown III for computer programming and
technical assistance, Mr. Martin Duff for technical assistance
and all of the volunteers for generously giving their time.
References
[1] J.A. Ballas, Common factors in the identification of an assortment of
brief everyday sounds, Journal of Experimental Psychology: Human
Perception and Performance 19 (1993) 250–267.
[2] P. Baudena, E. Halgren, G. Heit, J.M. Clarke, Intracerebral potentials
to rare target and distractor auditory and visual stimuli: III. Frontal
cortex, Electroencephalography and Clinical Neurophysiology 94
(1995) 251–264.
[3] C. Chiu, D. Schacter, Auditory priming for nonverbal information:
implicit and explicit memory for environmental sounds, Conscious-
ness and Cognition 4 (1995) 440–458.
[4] E. Courchesne, S.A. Hillyard, R. Galambos, Stimulus novelty, task
relevance, and the visual evoked potential in man, Electroencephalog-
raphy and clinical Neurophysiology 39 (1975) 131–143.
[5] Y.M. Cycowicz, D. Friedman, A developmental study of the effect of
temporal order on the ERPs elicited by novel environmental sounds,
Electroencephalography and clinical Neurophysiology 103 (1997)
304–318.
[6] Y.M. Cycowicz, D. Friedman, Effect of sound familiarity on the
event-related potentials elicited by novel environmental sounds, Brain
and Cognition 36 (1998) 30–51.
[7] Y.M. Cycowicz, D. Friedman, M. Rothstein, An ERP developmental
study of repetition priming by auditory novel stimuli, Psychophysi-
ology 33 (1996) 680–690.
[8] K.R. Daffner, M.M. Mesulam, L.F.M. Scinto, D. Acar, V. Calvo, R.
Fausti, A. Chabrerie, B. Kennedy, P. Holcomb, The central role of the
prefrontal cortex in directing attention to novel events, Brain 123
(2000) 927–939.
[9] G.L. Dannenbring, K. Briand, Semantic priming and the word repe-
tition effect in a lexical decision task, Canadian Journal of Psychology
36 (1982) 435–444.
[10] C. Escera, K. Alho, I. Winkler, R. Naatanen, Neural mechanisms of
involuntary attention to acoustic novelty and change, Journal of Cog-
nitive Neuroscience 10 (1998) 590–604.
[11] M. Fabiani, D. Friedman, Changes in brain activity patterns in aging:
the novelty oddball, Psychophysiology 32 (1995) 579–594.
[12] D. Friedman, G.V. Simpson, ERP amplitude and scalp distribution
to target and novel events: effects of temporal order in young,
D. Friedman et al. / Cognitive Brain Research 18 (2003) 58–6464
middle-aged and older adults, Cognitive Brain Research 2 (1994)
49–63.
[13] D. Friedman, Y.M. Cycowicz, H. Gaeta, The novelty P3: an event-
related brain potential (ERP) sign of the brain’s evaluation of novelty,
Neuroscience and Biobehavioural Reviews 25 (2001) 355–373.
[14] G. Gratton, M.G.H. Coles, E. Donchin, A new method for off-line
removal of ocular artifact, Electroencephalography and Clinical Neu-
rophysiology 55 (1983) 468–484.
[15] J.R. Jennings, C.C. Wood, The q-adjustment procedure for repeated
measures analyses of variance, Psychophysiology 13 (1976) 277–278.
[16] R.T. Knight, Decreased response to novel stimuli after prefrontal
lesions in man, Electroencephalography and Clinical Neurophysiol-
ogy 59 (1984) 9–20.
[17] R.T. Knight, Attention, temporal processing and the human prefrontal
syndrome, Eighth International Conference on Cognitive Neuro-
science, Porquerolles Island, France, 2002.
[18] R.T. Knight, D. Scabini, D.L. Woods, C.C. Calyworth, Contributions
of temporal–parietal junction to the human auditory P3, Brain Re-
search 502 (1989) 109–116.
[19] R. Lynn, Attention, Arousal and the Orientation Reaction, Pergamon,
London, 1966, p. 118.
[20] G. McCarthy, C.C. Wood, Scalp distributions of event-related po-
tentials: an ambiguity associated with analysis of variance models,
Electroencephalography and Clinical Neurophysiology 62 (1985)
203–208.
[21] M.R. Nuwer, D. Lehmann, F. Lopes da Silva, S. Matsuoka, W. Su-
therling, J.F. Vibert, IFCN guidelines for topographic and frequency
analysis of EEGs and EPs. Report of an IFCN committee. Interna-
tional Federation of Clinical Neurophysiology, Electroencephalogra-
phy and Clinical Neurophysiology 91 (1994) 1–5.
[22] B. Opitz, A. Mecklinger, A.D. Friederici, D.Y. von Cramon, The
functional neuroanatomy of novelty processing: integrating ERP
and fMRI results, Cerebral Cortex 9 (1999) 379–391.
[23] A. Paivio, Mental Representations: A Dual Coding Approach, Oxford
Univ. Press, New York, 1986.
[24] F. Perrin, J. Pernier, O. Bertrand, J.F. Echallier, Spherical splines for
scalp potential and current density mapping, Electroencephalography
and Clinical Neurophysiology 72 (1989) 184–187.
[25] D.S. Ruchkin, R. Johnson Jr., D. Friedman, Scaling is necessary
when making comparisons between shapes of event-related potential
topographies: a reply to Haig et al., Psychophysiology 36 (1999)
832–834.
[26] M.D. Rugg, The effects of semantic priming and word repetition on
event-related potentials, Psychophysiology 22 (1985) 642–647.
[27] E. Schroger, M.-H. Giard, C. Wolff, Auditory distraction: event-re-
lated potential and behavioral indices, Clinical Neurophysiology 111
(2000) 1450–1460.
[28] E.N. Sokolov, The orienting response, and future directions of its
development, Pavlovian Journal of Biological Science 25 (1990)
142–150.
[29] G.P. Stuart, D.M. Jones, Priming the identification of environmental
sounds, The Quarterly Journal of Experimental Psychology 48A
(1995) 741–761.
[30] C. Van-Petten, H. Rheinfelder, Conceptual relationships between spo-
ken words and environmental sounds: event-related brain potential
measures, Neuropsychologia 33 (1995) 485–508.
[31] E. Yago, C. Escera, K. Alho, M.H. Giard, J.M. Serra-Grabulosa,
Spatiotemporal dynamics of the auditory novelty-P3 event-related
brain potential, Cognitive Brain Research 16 (2003) 383–390.