Cross-modal attention capture by affective stimuli: evidence from event-related potentials.

Abstract

The P3 component of the event-related potential (ERP) to an acoustic startle probe is modulated during picture viewing, with reduced P3 amplitude when participants view either pleasant or unpleasant, as opposed to neutral, pictures. We have interpreted this as reflecting capture of attentional resources by affective pictures, with fewer resources available for processing the secondary startle probe. In the present study, we tested this resource allocation hypothesis by presenting either pictures or sounds as foreground stimuli, with the prediction that P3 amplitude in response to secondary startle probes would be reduced for affectively engaging foregrounds regardless of modality. Using dense-array electroencephalography and a source estimation procedure, we observed that P3 amplitude was indeed smaller when startle probes were presented during emotional, as opposed to neutral, stimuli for both sound and picture foregrounds. Source modeling indicated a common frontocentral maximum of P3 modulation by affect. The data support the notion that emotionally arousing stimuli transmodally attract resources, leading to optimized processing of the affective stimuli at the cost of the processing of concurrent stimuli.

Full text

Page 1

Copyright 2007 Psychonomic Society, Inc.?18
The?abrupt?presentation?of?a?brief?intense?acoustic?stim-
ulus?(e.g.,?a?burst?of?white?noise)?elicits?a?reflexive?startle?
response?that?is?modulated?by?emotion.?One?component?
of?the?startle?response—the?reflexive?eyeblink—is?modu-
lated?by?hedonic?valence,?with?larger?eyeblinks?elicited?
when?participants?view?unpleasant?pictures,?and?smaller?
blinks?when?they?process?pleasant?stimuli.?On?the?other?
hand,?the?electrocortical?response?to?the?same?startling?
probe?varies?with?emotional?arousal,?rather?than?with?
hedonic?valence:?The?amplitude?of?the?P3?component?of?
the?event-related?potential?(ERP)?is?smaller?when?partici-
pants?view?either?pleasant?or?unpleasant,?as?opposed?to?
neutral,?pictures?(Cuthbert,?Schupp,?Bradley,?McManis,?
&?Lang,?1998;?Schupp,?Cuthbert,?Bradley,?Birbaumer,?&?
Lang,?1997;?Schupp?et?al.,?2004).?This?pattern?of?results?
has?been?interpreted?as?reflecting?reduced?availability?of?
resources?for?processing?the?startle?probe?in?the?presence?
of?affectively?arousing,?as?opposed?to?neutral,?pictures.?The?
affectively?arousing?foreground?stimuli?effectively?capture?
attentional?resources?from?a?limited-capacity?system.?
This?interpretation?is?consistent?with?reports?of?ampli-
fied?electrocortical?processing?for?affectively?engaging?vi-
sual?stimuli?at?early?(Junghöfer,?Bradley,?Elbert,?&?Lang,?
2001;?Pourtois,?Grandjean,?Sander,?&?Vuilleumier,?2004)?
and?late?(Cuthbert,?Schupp,?Bradley,?Birbaumer,?&?Lang,?
2000;?Keil?et?al.,?2002)?latencies.?Electrocortical?evidence?
for?a?limited-capacity?attention?system?is?also?abundant.?
Researchers?using?P3?as?a?marker?for?attentional?involve-
ment?in?a?primary?task?have?reported?that?the?auditory?ERP?
tends?to?be?diminished?if?the?auditory?stimulus?eliciting?
the?P3?is?accompanied?by?a?task-irrelevant?visual?stimu-
lus,?in?comparison?with?what?is?found?when?the?auditory?
stimulus?is?presented?alone?(Oray,?Lu,?&?Dawson,?2002).?
Resource?allocation?during?emotional?processing?can?
be?studied?by?using?the?P3?component?of?the?ERP?that?
occurs?in?response?to?the?secondary?startle?probe.?The?P3?
is?a?positive?deflection?of?the?ERP,?which?is?observed?in?
response?to?task-relevant?stimuli,?and?it?is?also?present?in?
response?to?novel?or?infrequent?stimuli?(Kok,?2001;?Lin-
den,?2005).?Extensive?work?with?this?measure?has?yielded?
a?complex?database?suggesting?that?P3?has?several?over-
lapping?subcomponents,?depending?on?aspects?of?the?
stimulus?eliciting?the?P3?component?(e.g.,?a?novel?or?a?rare?
stimulus),?and?on?the?type?of?task?used.?For?the?present?
purpose,?studies?of?so-called?novelty?P3?(also?referred?to?
as?P3a;?see?Simons,?Graham,?Miller,?&?Chen,?2001)?are?
of?particular?interest.?This?component?has?been?related?
to?more?automatic?detection?of?rare?or?novel?and?relevant?
events?and?has?been?linked?to?generator?structures?in?the?
frontal?cortex?(Dien,?Spencer,?&?Donchin,?2003).
We?have?suggested?that?P3?modulation?in?the?context?
of?affective?picture?processing?reflects?motivated?atten-
tion,?in?which?stimuli?that?activate?fundamental?defensive?
or?appetitive?neural?systems?prompt?an?increase?in?orient-
ing?and?attentional?resource?allocation?(Lang,?Bradley,?&?
Cuthbert,?1997).?In?line?with?that?notion,?enhancements?of?
Cross-modal attention capture
by affective stimuli:
Evidence from event-related potentials
AndreAs Keil, MArgAret M. BrAdley, MArKus Junghöfer,
thoMAs russMAnn, WilliAM loWenthAl, And Peter J. lAng
University of Florida, Gainesville, Florida
The?P3?component?of?the?event-related?potential?(ERP)?to?an?acoustic?startle?probe?is?modulated?during?
picture?viewing,?with?reduced?P3?amplitude?when?participants?view?either?pleasant?or?unpleasant,?as?opposed?
to?neutral,?pictures.?We?have?interpreted?this?as?reflecting?capture?of?attentional?resources?by?affective?pictures,?
with?fewer?resources?available?for?processing?the?secondary?startle?probe.?In?the?present?study,?we?tested?this?
resource?allocation?hypothesis?by?presenting?either?pictures?or?sounds?as?foreground?stimuli,?with?the?prediction?
that?P3?amplitude?in?response?to?secondary?startle?probes?would?be?reduced?for?affectively?engaging?foregrounds?
regardless?of?modality.?Using?dense-array?electroencephalography?and?a?source?estimation?procedure,?we?ob-
served?that?P3?amplitude?was?indeed?smaller?when?startle?probes?were?presented?during?emotional,?as?opposed?
to?neutral,?stimuli?for?both?sound?and?picture?foregrounds.?Source?modeling?indicated?a?common?frontocentral?
maximum?of?P3?modulation?by?affect.?The?data?support?the?notion?that?emotionally?arousing?stimuli?transmod-
ally?attract?resources,?leading?to?optimized?processing?of?the?affective?stimuli?at?the?cost?of?the?processing?of?
concurrent?stimuli.?
Cognitive, Affective, & Behavioral Neuroscience
2007, 7 (1), 18-24
A. Keil, andreas.keil@uni-konstanz.de

Page 2

Startle Probe P3? ? ? ? 19
P3?during?affective?picture?viewing?have?been?observed?in?
a?number?of?studies,?suggesting?increased?resource?alloca-
tion?to?emotionally?engaging?stimuli?(Cuthbert?et?al.,?2000;?
Keil?et?al.,?2002).?Conversely,?the?P3?response?to?an?acous-
tic?startle?probe?presented?during?picture?viewing?shows?
the?opposite?pattern,?with?smaller?probe?P3?responses?when?
participants?concurrently?view?emotional,?as?opposed?to?
neutral,?pictures?(Cuthbert?et?al.,?1998;?Schupp?et?al.,?
2004).?That?is,?because?motivationally?relevant?pictures?uti-
lize?greater?attentional?resources?than?do?neutral?pictures,?
fewer?are?available?for?processing?the?startle?probes.?Probe?
P3?amplitude?is?thus?interpreted?as?indexing?resource?al-
location?for?the?processing?of?emotional?stimuli.
In?the?present?study,?we?examined?the?P3?response?to?
secondary?startle?probes?presented?while?participants?
either?viewed?affective?pictures?or?listened?to?affective?
sounds.?In?both?groups?of?participants,?we?measured?the?
P3?to?an?acoustic?startle?probe?delivered?during?looking?
or?listening.?To?increase?the?sensitivity?to?emotion-related?
changes?in?topographical?distribution?of?P3,?we?used?a?
dense?sensor?array?(129?channels).?We?compared?ERP?am-
plitudes?in?the?P3?range?for?startle?probe?processing?when?
participants?looked?at?pictures?and?listened?to?sounds,?and?
also?conducted?source?strength?analyses,?using?a?linear?
projection?method,?the?minimum?norm?estimate?(Hauk,?
Keil,?Elbert,?&?Muller,?2002),?to?examine?possible?neural?
sources?of?affective?modulation?of?probe?P3.?
We?expected?to?replicate?our?previous?finding?of?re-
duced?P3?amplitude?during?affective?picture?viewing?
(Cuthbert?et?al.,?1998),?and?further,?to?identify?neural?
structures?underlying?this?effect?using?the?minimum?norm?
source?estimation?approach.?A?more?central?research?ques-
tion?concerns?whether?acoustic?probe?P3?modulation?will?
be?similar?when?the?foreground?modalities?differ.?Several?
studies?have?suggested?that?allocation?of?attentional?re-
sources?may?occur?in?a?modality-specific?manner?under?
some?circumstances?(see?Lipp?&?Neumann,?2004),?fa-
cilitating?subsequent?processing?in?the?same?modality?
(e.g.,?Schmitt,?Postma,?&?De?Haan,?2000).?In?this?type?of?
framework,?attentive?processing?of?motivationally?relevant?
sounds?should?increase?P3?amplitude?elicited?by?the?audi-
tory?startle?probe,?because?resources?would?be?allocated?
to?the?auditory?modality.?Interpretation?of?the?reduction?of?
auditory?startle?probe?P3?for?emotionally?arousing?visual?
foregrounds?in?this?perspective?would?be?due?to?increas-
ing?transfer?to?the?visual?modality?at?the?cost?of?resources?
available?for?auditory?processes.?
In?contrast,?if?motivated?attention?is?allocated?in?a?
supramodal,?central?fashion,?the?amplitude?of?P3?to?an?
acoustic?startle?probe?should?be?similarly?modulated?dur-
ing?processing?of?affective?stimuli,?regardless?of?the?pre-
sentation?modality?of?the?emotional?foreground?(i.e.,?audi-
tory?or?visual).?Here,?we?drew?on?the?concept?of?motivated?
attention?as?a?process?that?enhances?adaptive?behaviors?in?
the?presence?of?stimuli?that?threaten?or?sustain?life?(Lang?
et?al.,?1997).?Thus,?we?expected?similarly?reduced?P3?am-
plitudes?to?the?startle?probe,?regardless?of?whether?partici-
pants?were?viewing?emotional?pictures?or?were?listening?
to?affective?sounds.
METHOD
Participants
A?total?of?42?students?taking?the?General?Psychology?course?at?the?
University?of?Florida?participated?as?part?of?a?course?requirement?
after?giving?informed?consent.?The?study?was?approved?by?the?Uni-
versity?of?Florida?institutional?review?board.?Twenty-three?students?
(10?female,?13?male)?listened?to?sounds,?and?19?students?(8?female,?
3?male,?8?inadvertently?not?coded)?viewed?pictures.
Stimuli and Design
Visual foreground: Looking at pictures.?Sixty?pictures1?were?
selected?from?the?International?Affective?Picture?System?(IAPS;?
Bradley,?Cuthbert,?&?Lang,?2005),?a?collection?of?photographs?that?
vary?in?affect,?to?include?20?pleasant?(mean?pleasure,?arousal?5?7.3,?
5.4),?20?neutral?(mean?pleasure,?arousal?5?4.9,?2.8),?and?20?un-
pleasant?(mean?pleasure,?arousal?5?2.3,?6.9)?pictures.?Unpleasant?
and?pleasant?pictures?were?significantly?more?arousing?than?neutral?
pictures,?and?unpleasant?pictures?prompted?slightly?higher?arousal?
ratings?than?did?pleasant?pictures.?The?screen?was?placed?1.5?m?in?
front?of?the?viewer,?resulting?in?a?picture?presentation?with?a?visual?
angle?of?about?18º?horizontally?and?10º?vertically.?Each?picture?was?
presented?for?1,500?msec,?followed?by?a?3.5-sec?intertrial?interval.?
Shorter?presentation?times?in?comparison?with?those?for?sounds?(see?
below)?were?used?to?reduce?the?amount?of?exploratory?eye?move-
ments?during?picture?viewing,?which?would?have?resulted?in?loss?of?
trials?for?subsequent?analyses.?To?compensate?for?total?study?time,?
each?set?of?60?pictures?was?presented?four?times.?Startle?probes?were?
95?dB(A)?bursts?of?white?noise,?delivered?1,100?msec?after?picture?
onset?on?20?trials?for?each?picture?valence.?The?interprobe?interval?
varied?from?9?to?22?sec,?with?a?mean?of?15?sec.
Auditory foreground: Listening to sounds.?Sixty?6-sec?
sounds2?were?selected?from?the?International?Affective?Digitized?
Sounds?(IADS;?Bradley?&?Lang,?1999),?a?collection?of?acoustic?
stimuli?that?vary?in?affect,?to?include?20?pleasant?(mean?pleasure,?
arousal?5?7.0,?5.9),?20?neutral?(mean?pleasure,?arousal?5?5.0,?4.7),?
and?20?unpleasant?(mean?pleasure,?arousal?5?2.4,?7.5)?sounds.?As?
with?the?picture?stimuli,?the?unpleasant?and?pleasant?sounds?were?
significantly?more?arousing?than?the?neutral?sounds,?and?the?un-
pleasant?sounds?prompted?slightly?higher?arousal?ratings?than?did?the?
pleasant?sounds.?Each?sound?stimulus?was?presented?twice?during?
the?experiment,?with?a?duration?of?6?sec?and?an?intertrial?interval?that?
varied?from?8?to?18?sec.?The?acoustic?startle?probes?were?presented?
on?20?trials?for?each?sound?valence?at?3–4.5?sec?after?sound?onset.
Electrophysiological Recordings
EEG?was?recorded?continuously?from?129?electrodes?using?an?
Electrical?Geodesics?(EGI)?EEG?system?and?digitized?at?a?rate?of?
250?Hz,?using?Cz?as?a?recording?reference.?Impedances?were?kept?
below?50?kΩ,?as?recommended?for?the?Electrical?Geodesics?high?
input?impedance?amplifiers.?A?subset?of?EGI?net?electrodes?located?
at?the?outer?canthi?as?well?as?above?and?below?the?right?eye?was?used?
to?determine?horizontal?and?vertical?electrooculogram?(EOG).?All?
channels?were?preprocessed?online?by?means?of?0.1-Hz?high-pass?
and?100-Hz?low-pass?filtering.
Procedure
Data reduction and analysis.?Epochs?of?600?msec?(150?msec?
prestartle,?450?msec?poststartle)?were?obtained?for?each?startle?probe?
from?the?continuously?recorded?EEG,?relative?to?probe?onset.?The?
mean?voltage?of?a?120-msec?segment?preceding?startle?probe?onset?
was?subtracted?as?the?baseline.?In?a?first?step,?data?were?low-pass?
filtered?at?a?frequency?of?50?Hz?(24?dB/octave)?and?then?submit-
ted?to?the?procedure?proposed?by?Junghöfer,?Elbert,?Tucker,?and?
Rockstroh?(2000),?in?which?statistical?parameters?exclude?channels?
and?trials?contaminated?with?artifacts.?Because?the?startle?probe?reli-
ably?elicited?eyeblinks?from?60?to?120?msec?poststimulus,?an?interval?
ranging?from?190?to?450?msec?poststartle?was?used?for?artifact?cor-

Page 3

20? ? ? ? Keil et al.
rection.?This?procedure?resulted?in?rejection?of?trials?contaminated?
with?artifacts?(including?ocular?artifacts)?outside?the?critical?blink?
interval.?Artifacts?other?than?blinks?lying?within?the?startle?interval?
were?evaluated?by?visual?inspection,?and?the?respective?trials?were?
rejected.?Recording?artifacts?were?first?detected?using?the?record-
ing?reference?(i.e.,?Cz).?Subsequently,?global?artifacts?were?detected?
using?the?average?reference,?and?distinct?sensors?from?particular?tri-
als?were?removed?interactively,?on?the?basis?of?the?distribution?of?
their?mean?amplitude,?standard?deviation,?and?maximum?slope.?Data?
at?eliminated?electrodes?were?replaced?with?a?statistically?weighted?
spherical?spline?interpolation?from?the?full?channel?set.?The?mean?
number?of?approximated?channels?across?conditions?and?subjects?
was?9.?With?respect?to?the?spatial?arrangement?of?the?approximated?
sensors,?it?was?ensured?that?the?rejected?sensors?were?not?located?
within?one?region?of?the?scalp,?because?that?would?make?interpola-
tion?for?this?area?invalid.?Spherical?spline?interpolation?was?used?
throughout,?both?for?approximation?of?sensors?and?for?illustration?
of?voltage?maps?(Junghöfer,?Elbert,?Leiderer,?Berg,?&?Rockstroh,?
1997).
Single?epochs?with?excessive?eye?movements?and?eyeblinks?or?
more?than?20?channels?containing?artifacts?in?the?time?interval?of?
interest?were?discarded.?The?validity?of?this?procedure?was?further?
tested?by?visually?inspection?of?the?vertical?and?horizontal?EOG?
as?computed?from?a?subset?of?the?electrodes?that?were?part?of?the?
electrode?net.?Subsequently,?data?were?arithmetically?transformed?
to?the?average?reference,?which?was?used?for?all?analyses.?After?arti-
fact?correction,?an?average?of?74%?of?the?trials?were?retained?in?the?
analyses,?with?a?mean?of?14.9?(range,?11–18)?trials?for?the?pleasant?
foregrounds,?14.5?(range,?12–17)?for?the?neutral?foregrounds,?and?
13.9?(range,?10–19)?for?the?unpleasant?foregrounds.?No?difference?
in?trial?number?was?observed?between?foreground?modalities?or?af-
fective?categories?(all?Fs?,?2.0,?ps?.?.40).
Scoring of P3 amplitude: Regional mean analysis.?For?a?
first?statistical?analysis?of?the?P3?component,?we?scored?the?mean?
voltage?in?a?time?window?ranging?from?260?to?340?msec?for?each?
participant?and?condition?in?each?experiment.?This?latency?range?is?
consistent?with?the?majority?of?studies?of?P3?using?infrequent,?sa-
lient?stimuli?(see,?e.g.,?Opitz,?Mecklinger,?Friederici,?&?von?Cramon,?
1999)?such?as?the?startle?probes?in?the?present?study.?Given?the?low?
trial?count,?we?did?not?execute?a?more?fine-grained?temporal?analy-
sis,?but?relied?on?the?robustness?of?the?P3?wave?in?the?indicated?time?
range.?As?can?be?seen?in?Figures?1?and?2,?the?P3?in?the?latency?range?
of?interest?was?most?pronounced?at?sensors?surrounding?site?Pz.?
Thus,?we?selected?Pz?as?well?as?eight?immediate?neighbor?electrodes?
to?form?a?regional?mean?as?a?measure?for?overall?effects?of?condi-
tions?on?P3?amplitude.?
These?values?were?determined?for?each?foreground?content?and?
participant?and?were?submitted?to?repeated?measures?ANOVAs.?Sep-
arate?ANOVAs?with?one?within-subjects?factor?of?hedonic?valence?
were?conducted?for?visual?foregrounds?(pleasant,?neutral,?unpleas-
ant)?and?acoustic?foregrounds?(pleasant,?neutral,?unpleasant?tone).?
Where?appropriate,?degrees?of?freedom?were?corrected?by?means?
of?the?Greenhouse–Geisser?method,?and?corrected?p?values?are?re-
ported?together?with?uncorrected?degrees?of?freedom.?Significant?
effects?were?followed?by?contrast?analysis?examining?the?presence?
of?the?expected?quadratic?trend?in?the?voltage?data—that?is,?the?find-
ing?that?startle?P3?amplitude?for?emotional?(pleasant?or?unpleasant)?
foregrounds?was?less?than?for?neutral?pictures.
Source Space Analyses
To?achieve?a?signal?representation?reflecting?the?high?spatial?
resolution?of?the?dense?array?ERP?data,?we?transformed?the?in-
dividual?ERPs?into?a?four-shell?spherical?source?space,?using?the?
minimum?(L2)?norm?estimate?(MNE)?distributed?source?algorithm?
(Hämäläinen?&?Ilmoniemi,?1984;?Hauk?et?al.,?2002).?This?procedure?
does?not?allow?precise?anatomical?localization?but,?in?comparison?
with?the?use?of?voltage?date,?markedly?improves?spatial?specific-
ity?(Hauk,?2004).?To?this?end,?the?voltage?data?were?projected?to?a?
source?space?consisting?of?655?evenly?distributed?dipoles?with?three?
orthogonal?orientations?at?each?dipole?location.?This?is?achieved?by?
multiplying?the?pseudoinverse?of?the?lead?field?matrix?L?with?the?
data?matrix?D?(Hauk,?2004).?The?L2?(minimum)?norm?provides?a?
mathematically?unique?solution?for?this?equation?by?minimizing?the?
squared?source?current.?Dipole?strength?at?a?given?dipole?site?was?
obtained?as?the?square?root?of?the?sum?of?squared?L2?values?for?each?
of?the?three?orientations,?for?each?time?point?of?the?data.
Statistical parametric mapping.?To?enhance?spatial?resolution?
of?the?statistical?analysis?conducted?for?regional?means,?we?identi-
fied?dipole?sites?showing?the?hypothesized?quadratic?trend?for?dipole?
strength?(pleasant?5?unpleasant?,?neutral?foreground)?by?conduct-
ing?trend?tests?at?each?dipole?location.?Mean?P3?dipole?density?values?
in?the?selected?time?window?were?subject?to?trend?analyses?for?the?
expected?quadratic?trend,?resulting?in?an?F?value?for?each?dipole?site.?
Because?of?the?problems?associated?with?multiple?significance?test-
Table 1
Regional Voltage Means (With Standard Deviations) at Site Pz
and Eight Immediate Neighbor Electrodes in a Window
From 260 to 340 msec Following Presentation
of an Acoustic Startle Probe
Pleasant
M SD
Pictures3.06?3.304.91
Sounds ?4.66?2.46?5.83
Neutral

Unpleasant
M
3.27
4.31
MSD
3.66
3.12
SD
3.19
2.61???
4
2
0
–2
–4
4
2
0
–2
–4
4
2
0
–2
–4
Fz
Cz
Pz
Pleasant
Neutral
Unpleasant
–100 0100200 300400 msec
Figure 1. Startle probe ERPs at three midline electrodes for
the viewing of pleasant, neutral, or unpleasant pictures. ERPs
are shown at three midline electrodes for pleasant, neutral, and
unpleasant foreground pictures. The gray shaded region indicates
the time segment used for statistical analysis. Averages reflect 19
participants.

Page 4

Startle Probe P3? ? ? ? 21
ing,?p?values?for?this?analysis?were?determined?according?to?the?per-
mutation?method?proposed?by?Blair?and?Karniski?(e.g.,?Blair,?Mor-
ris,?Frith,?Perrett,?&?Dolan,?1999).?Permutation?methods?are?based?
on?a?test?distribution?that?is?generated?from?the?actual?data?set?by?
calculating?the?test?statistic?of?interest?for?data?randomly?shuffled?
across?experimental?conditions.?Hence,?a?test?distribution?emerges?
that?contains?realizations?of?the?statistic?under?the?assumption?that?
there?is?no?effect?of?conditions.?The?present?permutation?distributions?
were?based?on?8,000?draws?for?entire?topographies?(i.e.,?one?draw?
taking?into?account?655?dipole?locations)?and?the?greatest F value?of?
each?draw?(i.e.,?of?each?topography?of?permutation-based F values)?
entered?the?test?distribution.?For?the?present?data?set,?this?procedure?
resulted?in?critical F values?of?4.9?(p?,?.05)?and?6.8?(p?,?.01).
RESULTS
ERP Scalp Voltages: Regional Means
The?grand?mean?ERPs?to?the?acoustic?startle?probe?
are?illustrated?when?startle?probes?were?presented?in?the?
contexts?of?viewing?pictures?(Figure?1)?and?of?listening?
to?sounds?(Figure?2).?Startle?probes?elicited?reliable?late?
positivity?in?the?P3?range,?reaching?maximum?amplitudes?
from?260?to?340?msec,?which?is?consistent?with?the?find-
ings?of?earlier?studies?using?abrupt?and?salient?auditory?
stimuli?(Opitz?et?al.,?1999;?Simons?et?al.,?2001).?Regional?
means?in?the?time?region?of?interest?are?listed?in?Table?1.?
The?grand?mean?ERP?also?shows?a?small?P50?and?subse-
quent?N1?component?in?response?to?the?startle?probe,?which?
were?not?a?focus?of?the?present?investigation.?A?negativity?
in?the?temporal?vicinity?of?P3?that?has?been?reported?by,?
for?example,?Spencer,?Dien,?and?Donchin?(1999)?was?not?
observed?in?the?present?data?set.?
Visual foreground: Looking at pictures.?Regional?
means?showed?the?expected?pattern?of?differences?related?
to?affective?picture?content?[F(2,36)?5?7.0,?p?,?.01].?
Viewing?pleasant?or?unpleasant?pictures?resulted?in?signif-
icantly?smaller?P3?amplitude?than?did?viewing?neutral?pic-
tures,?resulting?in?a?significant?quadratic?trend?[F(1,18)?5?
17.4,?p?,?.01]?(see?Figure?1).
Auditory foreground: Listening to sounds.?Par-
alleling?the?results?for?affective?pictures,?the?amplitude?
of?the?acoustic?probe?P3?was?similarly?sensitive?to?fore-
ground?sound?content?[F(2,44)?5?7.2,?p?,?.01].?Listening?
to?pleasant?or?unpleasant?sounds?resulted?in?significantly?
smaller?P3?amplitude?than?did?listening?to?neutral?sounds,?
which?was?again?reflected?in?a?significant?quadratic?trend?
[F(1,22)?5?15.2,?p?,?.01]?(see?Figure?2).?
Statistical Parametric Mapping
Figure?3?illustrates?the?mean?dipole?density?maps?in?the?
P3?time?range?during?processing?of?emotional?(pleasant?or?
unpleasant)?or?neutral?sounds?(bottom)?or?pictures?(top).?
Paralleling?the?ERP?voltage?results,?an?attenuation?in?dipole?
density?is?obtained?in?the?probe?P3?time?segment?for?the?
processing?of?affective?stimuli?in?the?foreground—whether?
pictures?or?sounds.?Source?activity?in?this?time?range?in-
dicated?a?contribution?of?frontocentral?as?well?as?parietal?
structures?when?the?probe?was?presented?during?listening?
to?sounds.?A?stronger?contribution?of?frontal?regions?to?the?
source?activity?in?the?P3?range?was?observed?for?probe?pro-
cessing?during?the?viewing?of?pictures,?with?frontocentral?
and?parietal?regions?also?contributing?to?the?overall?signal.?
In?a?next?step,?we?combined?the?minimum?norm?estima-
tion?approach?with?permutation?tests?of?quadratic?trends?at?
each?of?655?dipole?locations?to?enhance?spatial?resolution?
of?the?statistical?analyses.?Trend?analyses?(in?which?pleas-
ant?and?unpleasant?stimuli?prompted?less?probe?P3?dipole?
activity)?were?conducted?at?each?dipole?site,?resulting?
in?topographical?maps?illustrating?significant?quadratic?
trends?(pleasant?5?unpleasant?,?neutral)?present?in?the?
respective?time?range?and?foreground?modality.?
Visual foregrounds.?Dipole?sites?showing?significant?
(ppermut?,?.05)?quadratic?trends?indicating?reduced?probe?
Pleasant
Neutral
Unpleasant
A
B
µV
13
–8
0 448 msec
4
2
0
–2
–4
4
2
0
–2
–4
4
2
0
–2
–4
Fz
Cz
Pz
Pleasant
Neutral
Unpleasant
–100 0100200300400 msec
Figure 2. (A) Event-related potential in response to startle
probes presented during listening to pleasant, neutral, or un-
pleasant sounds. Frontal electrodes are shown at the top of the fig-
ure. Waveforms reflect means across 23 participants. (B) Startle
probe ERPs at three midline electrodes for listening to pleasant,
neutral, or unpleasant sounds. The gray shaded region indicates
the time segment used for statistical analysis. Averages reflect 23
participants.

Page 5

22? ? ? ? Keil et al.
P3?reactivity?for?viewing?emotional?(pleasant?or?unpleas-
ant),?as?opposed?to?neutral,?pictures?are?shown?in?Fig-
ure?4?(top?panel).?The?effects?were?widespread,?and?most?
pronounced?in?the?left?frontotemporal?area?as?well?as?in?
frontocentral?areas.?
Auditory foregrounds.?The?dipole?sites?showing?
significant?quadratic?trends?that?indicated?less?probe?
P3?activity?for?listening?to?affective,?as?opposed?to?neu-
tral,?stimuli?were?very?similar?to?the?maps?obtained?for?
visual?processing.?Effects?were?again?widespread,?with?
a?predominance?in?bilateral?frontotemporal?regions?and?
frontocentral?areas.
DISCUSSION
Affective?modulation?of?the?P3?component?in?response?
to?an?acoustic?startle?probe?was?observed?in?the?present?
study,?regardless?of?whether?people?were?viewing?pictures?
or?listening?to?sounds?in?the?foreground.?In?each?case,?the?
amplitude?of?the?P3?elicited?by?a?concurrently?delivered?
acoustic?startle?probe?was?attenuated?for?the?processing?of?
affective,?as?opposed?to?looking?at?or?listening?to?neutral,?
stimuli.?Source?localization?analyses?suggested?that?these?
differences?were?mediated?mainly?by?differential?activity?
in?frontocentral?as?well?as?temporoparietal?areas.?
These?data?are?strong?evidence?against?a?specific?
? modality-directed?attentional?effect?of?P3?modulation?
during?affective?picture?viewing.?Our?findings?are?in?
line?with?a?more?general?resource?allocation?account,?in?
which?stimuli?that?activate?defensive?or?appetitive?moti-
vational?systems?engage?greater?allocation?of?attentional?
resources,?irrespective?of?modality.?The?interpretation?of?
P3?amplitude?as?reflecting?differential?attention?or?orient-
ing?is?consistent?with?the?assumption?that?P3?amplitude?
reflects?attentional?capacity?allocated?to?categorization?of?
task-relevant?or?salient?events?(Linden,?2005).?Although?
the?interpretation?of?P3?findings?still?remains?controver-
sial,?ERP?amplitudes?in?the?P3?range?generally?tend?to?be?
larger?for?intense?(Sugg?&?Polich,?1995)?or?task-relevant?
(Polich?&?Kok,?1995)?stimuli?and?are?modulated?by?the?
significance?of?the?stimulus?for?an?observer,?even?in?the?
absence?of?a?task?(Keil?et?al.,?2002).?
We?have?suggested?that?emotional?stimuli?prompt?“mo-
tivated?attention,”?in?which?the?activation?of?neural?circuits?
mediating?appetite?and?defense?is?associated?with?the?cap-
ture?of?attentional?resources?and?the?enhanced?perceptual?
processing?of?motivationally?relevant?stimuli?(Lang?et?al.,?
Visual Foreground
Sound Foreground
PleasantNeutralUnpleasant
PleasantNeutralUnpleasant
nA/cm2
120
100
80
60
40
nA/cm2
120
100
80
60
40
Figure 3. Grand mean topographical distribution of the minimum
norm estimate dipole density for the startle probe P3 during viewing of
affective pictures and listening to affective sounds. A top view is shown.
Darker shades represent greater dipole density.
Visual
Foreground
Auditory
Foreground
ppermut
.05 .01
Figure 4. Statistical topographical maps showing dipole sites
with a significant (ppermut .05) quadratic contrast, reflecting
enhanced probe P3 amplitude for neutral, as opposed to pleas-
ant and unpleasant, contents, shown separately for pictures and
sounds. Significance thresholds were determined using permuta-
tion distributions (see the Method section). Dipole sites in dark
gray represent a significance level of ppermut .05. Only dipole
sites shown in black reached the .01 significance criterion.