Evidence of Vocal and Manual Event

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Research Article

Evidence of Vocal and Manual Event


Files in Auditory Negative Priming
Susanne Mayr, Malte Möller, and Axel Buchner
Department of Experimental Psychology, Heinrich-Heine-Universität Düsseldorf, Germany
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Abstract. Negative priming with auditory as well as with visual stimuli has been shown to involve the retrieval of prime response information as
evidenced by an increase of prime response errors to the probes of ignored repetition trials compared to control trials. We investigated whether
prime response retrieval processes were also present for response modalities other than manual responding. In an auditory four alternative forced
choice task participants either vocally or manually identified a target sound while ignoring a distractor sound. Negative priming was of equal size
in both response modalities. What is more, for both response modalities, there was evidence of increased prime response errors in ignored
repetition trials compared to control trials. The findings suggest that retrieval of event files of the prime episode including prime response
information is a general mechanism underlying the negative priming phenomenon irrespective of stimulus or response modality.

Keywords: audition, negative priming, event file integration, manual/verbal response modality

The impaired responding to a previously ignored stimulus – This dependency of negative priming on a prime-to-probe
which is called the negative priming effect – is an estab- response change is explained by a mechanism retrieving
lished and robust phenomenon in the visual as well as the prime response information that is probe response incompat-
auditory modality (see May, Kane, & Hasher, 1995; Mayr ible in the response change case, but compatible in the
& Buchner, 2007; Tipper, 2001). A model that has been pro- response repetition case. An inhibition model cannot
posed – either as an alternative or in addition to the origi- account for such a Stimulus repetition · Response repetition
nally proposed inhibition model (Tipper, 1985, 2001; interaction. Instead, it implies that negative priming is
Tipper & Cranston, 1985) – is the episodic retrieval model. caused by inhibition of distracting information during the
This model assumes that in ignored repetition trials the prime that persists into the probe and interferes with
probe target functions as a retrieval cue to the prime episode. responding regardless of response repetition or change.
The model predicts negative priming due to retrieval of a Support for the retrieval of prime response information also
do-not-respond tag attached to the previously ignored stim- comes from experiments showing for both the auditory and
ulus (Neill & Valdes, 1992) or due to retrieval of inappropri- the visual modality that the tendency to repeat the prime
ate prime response information (Mayr & Buchner, 2006; response is significantly stronger in ignored repetition than
Rothermund, Wentura, & De Houwer, 2005). Either way, in control trials (Mayr & Buchner, 2006, 2010; Mayr,
the retrieved information conflicts with the requirement to Buchner, & Dentale, 2009). The presence of such a prime
respond to the same stimulus when it becomes the probe tar- response retrieval effect can be explained by automatic
get. Resolving this conflict takes time. retrieval of incompatible prime response information.
A considerable number of studies have reported evi- Retrieval of prime episode information is conceptually
dence in favor of the episodic retrieval model. Some studies equivalent to what has been called event files by Hommel
provided evidence that the size of the negative priming (1998, 2004). A number of recent studies have presented
effect is influenced by variables that usually determine evidence showing that stimulus features together with
memory performance, such as the temporal discriminability response information are integrated into event files, the
of the to-be-retrieved memory trace at the time of retrieval retrieval of which can influence performance. Zmigrod
(Mayr & Buchner, 2006; Neill, Valdes, Terry, & Gorfein, and Hommel (2009, 2010) presented evidence indicative
1992) or the contextual similarity between encoding (prime) of spontaneous integration of auditory features (such as
and retrieval (probe; Chao & Yeh, 2008; Fox & de Fockert, pitch or location) and manual responses into event files by
1998; Neill, 1997; Stolz & Neely, 2001; but see Wong, showing that performance in a discrimination task was better
2000). Another line of evidence showed that prime response when the stimulus and response features of the first of two
information is retrieved in ignored repetition but not in con- consecutive sounds shared all or none of the stimulus and
trol trials. Rothermund et al. (2005) demonstrated that neg- response features of the second sound rather than when only
ative priming depends on a response change. If prime and one of the features overlapped, that is, when a partial repe-
probe responses are identical the typical slowdown of tition occurred. Partial repetition costs are explained by a
ignored repetition trials inverts into a positive priming effect. feature integration process that establishes an episodic trace

 2011 Hogrefe Publishing Experimental Psychology 2011; Vol. 58(5):353–360


DOI: 10.1027/1618-3169/a000102
354 S. Mayr et al.: Evidence of Vocal and Manual Event Files

containing links between the simultaneously presented fea- Interestingly, evidence of event file integration is based
tures and the response. These episodes are assumed to be on simple (bi-)manual stimulus-response tasks (e.g.,
retrieved as a whole when at least one of the elements is Hommel, 1998, 2005; Pösse, Waszak, & Hommel, 2006).
encountered again. Event file retrieval is unproblematic or Here, too, it has yet to be shown that event file integration
even beneficial for subsequent responding if all features can also be observed with more complex responses.1
bound together are repeated. However, partial repetitions In the experiment reported here, participants identified
lead to reactivation of misleading information which slows the sounds played at one ear while ignoring the sounds
down responding. The negative priming effect can be con- played at the other ear. Identification responses were given
ceptualized as a partial repetition cost that occurs when either manually or vocally. If the prime response retrieval
retrieval of the prime episode is cued by the repetition of mechanism were a general process that is not restricted to
the prime distractor but the concomitantly retrieved prime very simple prime responses, then we should find a prime
response is incompatible with probe responding (for exam- response retrieval effect with manual but also with vocal
ples in the visual modality, see Frings, 2011; Frings, responding.
Rothermund, & Wentura, 2007; Giesen & Rothermund,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

in press; Rothermund et al., 2005). Note that event files in


This document is copyrighted by the American Psychological Association or one of its allied publishers.

terms of Hommel (1998) originally referred to stimulus


and response features of one object only. The idea that sev-
eral stimuli such as the target and the distractor in a standard
Method
negative priming task can be integrated into one event file
including the prime response exceeds the original event file Participants
concept. This extension of the concept is necessary to
explain negative priming effects that depend on response Data of one participant were eliminated due to excessive
repetition/change as demonstrated by Rothermund et al. error frequencies (> .50). Six participants did not reach the
(2005; Frings et al., 2007) but also to explain prime response learning criterion. The final sample of 61 participants (52
retrieval effects as shown by Mayr and colleagues (2006, females) ranged in age from 19 to 59 years (M = 25.6).2
2009, 2010). Participants were tested individually and were either paid
All studies that have revealed evidence of prime or received course credit.
response retrieval in negative priming used very primitive
manual responses (but see Behrendt et al., 2010, who pres-
ent ERP evidence compatible with episodic retrieval in a Materials
vocal naming task). Participants had to press specific
response buttons to identify target sounds or pictures. Retrie- The stimuli were four synthesized sounds sampled with a
val-based interference may be facilitated by, or perhaps even rate of 44.1 kHz. Sound generation followed the description
be restricted to, situations in which the prime response is by Mondor, Leboe, and Leboe (2005) who successfully
extremely simple and can be retrieved very easily. If so, then demonstrated negative priming with such artificially gener-
we would expect a smaller, or perhaps even no contribution ated sounds. The sounds differed considerably from each
of retrieval-based interference to the negative priming effect other with respect to spectral composition and were easily
with more complex responses such as verbalizations. In discriminable. Each sound was 250 ms long. Participants
addition, as will be explicated more fully in the Discussion heard the sounds over earphones that were plugged into
section, manual responses have special properties which an Apple iMac.
raise the possibility that prime response retrieval might Each experimental trial consisted of a prime and a probe
occur only for a very specific subset of the relevant trials. presentation. For both presentations, a click indicated the ear
To clarify, auditory negative priming has been observed with at which the to-be-attended sound would be presented.
vocalizations before (Banks, Roberts, & Ciranni, 1995), but A simultaneously presented distractor appeared at the other
these effects could have been due mostly, if not exclusively, ear. The attended ear in the prime presentation was ran-
to inhibition (Tipper, 1985, 2001; Tipper & Cranston, 1985) domly assigned; the attended probe sound was presented
or do-not-respond tags attached to the previously ignored to the opposite ear. All participants completed a block of tri-
stimulus (Neill & Valdes, 1992) or both. It thus seemed the- als with a manual and one with a vocal response require-
oretically important to investigate whether evidence for ment. The sequence of blocks was varied between
prime response retrieval could be found with vocalizations, participants. In the manual response block, participants
and if so, to compare the size of this effect to that found with pressed the response key assigned to the attended sound.
simple manual responses. Sound-key associations were randomly assigned for each

1
To our knowledge there is one study in which the influence of manual versus verbal action preparation on identification performance of
masked arrows or spatial words was analyzed (Hommel & Müsseler, 2006). The results can be interpreted as evidence of feature integration
processes taking place during manual and vocal action preparation and visual stimulus identification. However, this paradigm to analyze the
so-called ‘‘action-effect blindness’’ is very different from the standard stimulus-response task usually applied to investigate feature
integration processes.
2
There were three participants older than 40 years. Statistical analyses without the three data sets did not change the pattern of results in any
way.

Experimental Psychology 2011; Vol. 58(5):353–360  2011 Hogrefe Publishing


S. Mayr et al.: Evidence of Vocal and Manual Event Files 355

participant. The response keys were sagittally aligned on a end, in the vocal block, the experimenter immediately coded
response box. Participants pressed the four keys with the the response, using a second monitor outside the partici-
index and middle fingers of both hands. Hand assignment pant’s range of vision behind a movable wall. Participants
– whether the two proximal/distal keys were operated with entered the experiment proper when their responses were
the left/right hand – alternated between participants. In the correct in at least 70% of the preceding 20 training trials.
vocal response block, participants named the attended Participants who did not qualify for the experiment within
sound. The sounds were artificial meaningless sounds. 50 trials were given the choice to quit or to start anew.
Names for the sounds were randomly assigned for each par- In the manual response block, each experimental trial
ticipant. The names comprised four German nouns with began with the presentation of the 20-ms click that indicated
easy detectable initial phonemes: Buchse (socket), Diener the target ear. After a 500-ms click-target interval, the prime
(butler), Tenor (tenor), and Palme (palm). We chose mean- sounds were presented. Three thousand milliseconds after
ingless sounds and randomly associated them with familiar prime onset, the probe click was presented, followed by a
nouns in order to keep the learning requirements of the man- 500-ms click-target interval after which the probe sounds
ual and vocal blocks parallel: Sound-vocal response associ- followed. In prime and probe, participants had to respond
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ations had to be learned in the same way as sound-manual within 2,000 ms after sound onset. Responses faster than
This document is copyrighted by the American Psychological Association or one of its allied publishers.

response associations. 100 ms were counted as false alarms. Three thousand milli-
Ignored repetition trials were constructed as in previous seconds after probe onset, participants were given audio-
experiments (Mayr & Buchner, 2006) by randomly selecting visual feedback about the correctness of their responses. A
three of the four stimuli as prime and probe sounds with the 2,000-ms interval followed before the next prime-probe trial
restriction that the ignored prime had to be identical to the was presented. For the vocal block, the sequence of events
attended probe. Parallel control trials were constructed by was the same with the only difference that participants
replacing the ignored prime with the remaining stimulus. responded vocally.
Within these two types of trials the ignored prime would After every 12th trial, participants received a summary
have been the attended probe stimulus on 50% of the trials, feedback with respect to average speed and accuracy. After
and the prime target would never have been equal to the the final trial of the first block, instructions and training for
probe target. Therefore, attended repetition trials were con- the second response modality block started. After the final
structed by randomly selecting three of the four stimuli as experimental trial, participants were informed about the pur-
prime and probe sounds with the restriction that the attended pose of the experiment. The experiment took 60 min.
prime had to be identical to the attended probe.3 Attended
repetition control trials were constructed by replacing the at-
tended prime with the remaining stimulus. There were 48 Design
unique trials, including 12 trials of each of the four trial
types. Each trial was presented four times, twice in each re- The experimental design comprised a 2 · 2 design with
sponse modality resulting in 192 experimental trials overall. trial type (ignored repetition vs. control) and response
Within the vocal and the manual response block, all trials modality (vocal vs. manual) as within-subject variables.
were presented in a random sequence. Data were collapsed over the sequence of blocks between-
subject variable.4 The dependent variable of greatest interest
was the probe error frequency, accumulated across partici-
Procedure pants, but we also analyzed participants’ average reaction
times and overall probe error rates.
Participants adjusted the loudness of the sounds to a com- The a priori power analysis focused on finding a differ-
fortable level (about 65 dB(A) on average). In each response ence in the probability of prime response retrieval (repre-
block, a training phase comprised trials with a target sound sented by the prr parameter of the prime response
presented to one ear, a distractor sound presented to the retrieval model as described further on) between the ignored
other ear, and a preceding click that indicated the target repetition and control condition for each response modality.
ear. Participants reacted to the target by pressing the associ- The population effect of the difference in the prr parameter
ated response key or by naming the sound. Participants were was estimated to be of size x = .065 (Mayr & Buchner,
given audio-visual feedback about the correctness of each 2006, Experiment 2). Calculations were based upon the esti-
reaction in the training and the experimental trials. To this mate that each participant would contribute approximately

3
We refrained from presenting the attended repetition data and from comparing them with the ignored repetition data for two reasons: First,
attended repetition trials were not hypothesis-relevant but were solely included as filler trials to prevent the development of any prime-to-
probe response strategies (e.g., ‘‘never repeat a prime response’’). Second, attended repetition trials differed from ignored repetition trials
not only with respect to the attentional focus at the repeated location (attended-attended vs. ignored-attended) but also with respect to the
presence of identity-location mismatches (which is a possible determinant of the negative priming effect, see Park & Kanwisher, 1994). The
to-be-attended ear always changed between prime and probe which implies that the repeated stimulus in ignored repetition trials was
presented to the same ear whereas the repeated stimulus in attended repetition trials was presented to different ears. As a consequence,
prime-to-probe identity-location mismatches were absent in ignored repetition trials but present in attended repetition trials. This confound
would have turned any comparison between these trial types ambiguous.
4
A supplementary analysis neither revealed a main effect nor any interaction with the sequence of blocks variable.

 2011 Hogrefe Publishing Experimental Psychology 2011; Vol. 58(5):353–360


356 S. Mayr et al.: Evidence of Vocal and Manual Event Files

40 utilizable probe responses in the manual as well as the 1,200 1,200


vocal response modality. Assuming desired levels of Ignored repetition
a = b = .05, data had to be collected from a sample of 1,150 Control 1,150
N = 77 participants (Faul, Erdfelder, Lang, & Buchner,
2007). We were able to collect data from N = 61 partici- 1,100 1,100
pants. With the overall N of 2,654 manual and 2,668 vocal

Reaction Time (ms)


responses, the power was somewhat smaller than what we 1,050 1,050
had planned for (1 b = .92).
1,000 1,000

950 950
Results
900 900
First, the negative priming effect was assessed, then the
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

prime response retrieval error was analyzed. Probe reaction


This document is copyrighted by the American Psychological Association or one of its allied publishers.

850 850
times were evaluated for trials with correct prime and probe
responses. Probe errors were evaluated only if they followed
a correct prime response. Vocal response latencies were 800 800
manual vocal
determined offline by the SayWhen-software package Response modality
(Jansen & Watter, 2008). The software detects vocal activity
that passes a specified threshold value in digital recordings 0.2 0.2
and employs additional steps to validate correct onset detec- Ignored repetition
Control
tion. If detected above-threshold activity cannot be verified
as the beginning of an actual response, the trial is flagged
as problematic. Additionally, we flagged all trials with 0.15 0.15
detected response latencies exceeding the range of two stan-
dard deviations above or below the mean of the respective
Error Rates

trial type as problematic. Voice onset in flagged trials was


based on visual inspection of the recorded speech 0.1 0.1
waveform.
There was a negative priming effect in response times as
responses to ignored repetition trials were slowed down
compared to control trials (Figure 1). Furthermore, responses
0.05 0.05
to manual trials were faster than responses to vocal trials.
In the 2 · 2 MANOVA of the reaction time data with
trial type (ignored repetition vs. control) and response
modality (manual vs. vocal) as within-subject variables the
main effects of trial type and of response modality were sta- 0 0
manual vocal
tistically significant, F(1, 60) = 57.05, p < .01, g2 = .49
and F(1, 60) = 91.71, p < .01, g2 = .61, respectively. The Response modality
interaction between both variables was not significant,
F(1, 60) = 1.15, p = .29, g2 = .02. There was a significant Figure 1. Sample average of participants’ mean reaction
negative priming effect for manual, t(60) = 4.96, p < .01, times (upper panel) and error rates (lower panel) as a
dz = 0.64, and for vocal responses, t(60) = 7.18, p < .01, function of response modality (manual vs. vocal) and trial
dz = 0.92. type (ignored repetition vs. control). The error bars depict
An analysis of the error data (Figure 1) revealed only a the standard errors of the means.
significant negative priming effect, F(1, 60) = 11.89,
p < .01, g2 = .17. Neither the main effect of response
modality nor the interaction between both variables was sig- generating a probe response. Participants correctly identify
nificant, F(1, 60) = 0.08, p = .78, g2 < .01 and F(1, 60) = and respond to the probe target with probability ci. Errors
0.38, p = .54, g2 = .01, respectively. Negative priming (1 ci) may have different causes. Participants might expe-
was significant for manual, t(60) = 2.11, p = .04, dz = 0.27, rience a probe stimulus confusion in that they confuse target
as well as for vocal responses, t(60) = 2.99, p < .01, and distractor and respond with the probe distractor with
dz = 0.38. probability psc. If probe stimulus confusion does not domi-
In order to test whether prime response retrieval occurred nate responding (1 psc), then, with probability prr, prime
in the manual as well as the vocal response modality, we response retrieval may occur and lead to incorrect prime
analyzed the error frequencies displayed in Table 1 using responses. Central to our concerns is the size of this prr
the multinomial prime response retrieval model introduced parameter in the ignored repetition (prrIR) compared to the
by Mayr and Buchner (2006; see Figure 2). The model control condition (prrC). If prime response retrieval occurs,
represents the processing stages presumably involved in then the probability of retrieving a prime response in the

Experimental Psychology 2011; Vol. 58(5):353–360  2011 Hogrefe Publishing


S. Mayr et al.: Evidence of Vocal and Manual Event Files 357

Table 1. Accumulated absolute frequencies of correct probe responses and of the different types of probe errors as a
function of response modality (manual vs. vocal) and trial type (ignored repetition vs. control)
Manual responses Vocal responses
Ignored repetition Control Ignored repetition Control
Correct probe target responses 1,213 1,281 1,264 1,303
Incorrect probe distractor responses 44 48 33 32
Incorrect prime target responses 39 4 26 3
Other incorrect responsesa 12 13 3 4
Note. aIgnored repetition trials: Incorrect responses using the key that was assigned to the non-presented stimulus.
Control trials: Incorrect prime distractor responses.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ignored repetition condition, prrIR, is expected to be larger the restricted model assuming that prrIR equals prrC had
than prrC, the same probability in the control condition. to be rejected for both modalities.
Thus, if prrIR > prrC and if, in addition, the goodness-
of-fit test of the restricted model assuming prrIR = prrC
leads to a significant misfit, then this is evidence in favor
of the assumption that prime response retrieval occurs and Discussion
contributes to the negative priming effect.
The unrestricted multinomial model displayed in Figure 2 The experiment revealed a negative priming effect in reac-
fitted the frequency data in both response modalities per- tion times and overall errors for the manual as well as the
fectly. The prime response retrieval parameters prrIR and vocal response modality. Whereas auditory negative priming
prrC are illustrated in Figure 3. We tested the goodness-of- has been demonstrated repeatedly with manual responses,
fit of the model with the restriction that prrIR = prrC. The there has been only one report of auditory negative priming
restricted model neither fit the data in the manual, with vocal responses (Banks et al., 1995). Our data nicely
G2(1) = 15.25, p < .01, x = . 076, nor the vocal response replicate their finding. Our data further show that the size
modality, G2(1) = 6.62, p = .01, x = .050. In other words, of the negative priming effect is independent of whether

Correct Probe
Target
ciIR Response
1 1
prr
Ignored Incorrect Probe Ignored repetition
Repetition Distractor prr
Control
pscIR Response
1 - ciIR 0.8 0.8
Incorrect Prime
prrIR Target
Probability Estimate

1 - pscIR Response

1 - prrIR Other Incorrect 0.6 0.6


Response

Correct Probe 0.4 0.4


Target
ciC Response

Incorrect Probe
Control
Distractor 0.2 0.2
pscC Response
1 - ciC
Incorrect Prime
prrC Target
1 - pscC Response 0 0
manual vocal
1 - prrC Other Incorrect Response modality
Response

Figure 3. Probability estimates for the model parameters


Figure 2. Multinomial processing tree model (‘‘prime representing the probability of prime response retrieval as
response retrieval model’’) for analyzing the probe a function of response modality (manual vs. vocal) and
reactions in the trial type condition ‘‘ignored repetition’’ trial type (ignored repetition vs. control). The error bars
(above) and ‘‘control’’ (below). For details see text. depict the standard errors of the probability estimates.

 2011 Hogrefe Publishing Experimental Psychology 2011; Vol. 58(5):353–360


358 S. Mayr et al.: Evidence of Vocal and Manual Event Files

participants respond vocally or manually. To our knowledge, in the auditory modality seem to come from the same distri-
this is the first systematic comparison of auditory negative bution, whereas effect sizes in the visual modality seem to
priming between response modalities (Tipper, MacQueen, come from a different distribution with a lower mean.
& Brehaut, 1988, for a similar finding in the visual In other words, stimulus modality seems to have a large ef-
modality). fect on the retrieval mechanism, whereas response modality
More important, however, was whether there would be does not seem to have an effect.
evidence of prime response retrieval when participants Given the essential differences between manual and
responded vocally. This was indeed the case. For both the vocal responding, the fact that the prime response retrieval
manual and the vocal response modalities, prime response effects were present (and of similar size) in both response
errors were more likely for ignored repetition compared to modalities sheds additional light on the underlying mecha-
control trials. In other words, for both response modalities nism. First, manual responding per se is of spatial nature
there was evidence of probe-triggered retrieval of prime epi- in that manual responses are spatially defined (e.g., a spe-
sodes that includes prime response information.5 This find- cific finger of the left or right hand). Furthermore, the audi-
ing indicates that prime response retrieval processes are not tory stimulation here as well as in similar experiments was
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

specific to the peculiarities of simple (manual) responses. of spatial nature, too, in that the target sounds were pre-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Taken together with the fact that the prime response retrieval sented at the left or right ear. Therefore, in addition to the
effect was found with artificial sound material (Mondor critical trial type manipulation a standard auditory negative
et al., 2005) instead of environmental sounds that were com- priming experiment also includes hidden spatial stimulus-
monly used in our previous experiments, this suggests that response compatibility manipulations (in the sense of Simon
prime response retrieval is a general phenomenon. A prob- & Rudell, 1967). For example, prime and probe presenta-
lem when comparing the size of the prime response retrieval tions can be stimulus-response compatible in that the loca-
effect with manual as opposed to vocal responses is that a tion of the target sound coincides with the response hand
direct statistical comparison between the sample effect size (left-left or right-right). They might also be stimulus-
measures representing these effects is not possible. How- response incompatible in that the location of the target sound
ever, we may assess the sample effect sizes observed here is opposite to the response hand (left-right or right-left).
relative to the distribution of all sample effect sizes obtained In addition to the stimulus-response compatibility variations,
so far with our prime response retrieval model in several manual responding in experiments like the one reported here
auditory negative priming experiments, all of which required is complicated by hand shifts and repetitions depending on
manual responding (Mayr & Buchner, 2006, 2010; Mayr whether or not the response hand changes between prime
et al., 2009). These previous sample effect sizes all were and probe. What is more, stimulus-response compatibility
within the interval (.045 < x < .075) with a mean of as well as prime-to-probe hand shifts or repetitions might
x = .060. This shows that the sample effect sizes of the also have a qualitative influence on prime response retrieval
prime response retrieval effect in the present experiments effects (for a comparable argumentation in visual negative
(x = .076 and .050 for the manual and vocal response priming, see Gibbons, 2009; Gibbons & Stahl, 2008).
modalities, respectively) are well within the distribution of Conceivably, prime response retrieval might take place pri-
the effect sizes that are typical of prime response retrieval marily, if not exclusively, in compatible probe trials without
effects in the auditory modality. What is more, they are gen- a prime-to-probe hand shift. Due to the stimulus-response
erally much larger than the prime response retrieval effect in compatibility these probe trials would lead to automatic acti-
the visual modality, which was found to be as small as vation of the correct probe response hand that had also been
x = .018 (Mayr & Buchner, 2006, Experiment 3). We thus the correct prime response hand. In ignored repetition trials,
conclude that the effect sizes of manual and vocal responses the retrieved prime response activation (of the same hand,

5
An alternative explanation for the increase of prime response errors in ignored repetition trials is also conceivable: The retrieved prime
distractor might trigger a state of response uncertainty, most probably due to probe-incompatible nonresponse information attached to it (in
terms of Neill’s, 1992 model). Possibly, this state of response uncertainty allows that hypothesized residual response activation from the
prime response breaks through and dominates probe responding. We thank Henning Gibbons for suggesting this alternative explanation. In
contrast to the prime response retrieval model, this explanation does not necessarily (but might) comprise the assumption of retrieved event
files (including prime response information). There is empirical evidence that does not support an explanation along the lines of assuming
residual response activation: We could show (Mayr & Buchner, 2010, Experiment 1) that neither the negative priming effect nor the prime
response retrieval effect was increased although participants had every reason to be biased to repeat responses from prime to probe. In this
experiment, there was one group that was exposed to 50% attended repetition trials beside ignored repetition and control trials, another
group was exposed to only 25% attended repetition trials, a third group was confronted with ignored repetition and control trials only (0%
attended repetition trials). The more frequently participants had to repeat responses, the more likely they should have developed a tendency
to repeat prime responses in the probe. In other words, in the 50% group as opposed to the 0% group, participants should have developed a
very liberal criterion to repeat responses because they benefitted very frequently from repeating a prime response. Therefore, if there were
such a thing as response uncertainty caused by the distractor-to-target repetition in ignored repetition trials, participants in the 50% group
should have been particularly susceptible to simply let go if there was a remaining prime response activation, as a consequence of which the
prime response retrieval effect should have been particularly large in this group. However, compared to the other groups, the 50% group did
not show an increase in the prime response retrieval effect (nor did they show an increase in the negative priming effect). Consequently, an
event-file interpretation is to be preferred to a ‘‘residual-activation-and-repetition-under-uncertainty’’ concept. Alternatively, it is also
possible that erroneous responses with the former prime response are sometimes caused by retrieval of event files including response
information and sometimes caused by residual prime response activation breaking through in a moment of response uncertainty.

Experimental Psychology 2011; Vol. 58(5):353–360  2011 Hogrefe Publishing


S. Mayr et al.: Evidence of Vocal and Manual Event Files 359

albeit of a different response finger) might boost the already target, distractor, context, and responses as evidenced here is
increased hand activation above a critical level necessary for conceptually equivalent to the idea of event files (Hommel,
response execution. A prime response error would occur. 1998). Zmigrod and Hommel (2009, 2010) demonstrated
Due to the small number of errors participants usually com- that auditory features are integrated into auditory event files
mit in our tasks we had to abstain from systematically ana- and that these event files are retrieved from memory and
lyzing prime response retrieval effects with respect to influence subsequent performance. Our data show that the
stimulus-response compatibility and hand shift variations integration and retrieval of event files takes place in auditory
in previous studies. However, the current experiment re- negative priming, and, what is more, that event file coding is
vealed a prime response retrieval effect with vocal responses not limited to manual responses but also takes place with vo-
which are unequivocally of nonspatial nature. We can there- cal responding. In other words, event files have a much
fore be sure that the response retrieval processes do not crit- broader significance than what we have assumed before.
ically depend on spatial stimulus-response compatibility or
prime-to-probe hand shifts.
There is another essential difference between manual and
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Acknowledgments
vocal responses. Manual choice tasks require the activation
This document is copyrighted by the American Psychological Association or one of its allied publishers.

and execution of a certain hand/finger response while at the


The research reported in this article was supported by a grant
same time inhibition of the response hand/finger alterna-
from the Deutsche Forschungsgemeinschaft (Ma 2610/2-2).
tive(s) is possible, if not inevitable. EEG studies have shown
that the activation of a manual response indexed by a nega-
tivity over the contralateral motor cortex prior to the contrac-
tion of the response agonist is accompanied by inhibition References
over the ipsilateral motor cortex (Meynier, Burle, Possamai,
Vidal, & Hasbroucq, 2009; Vidal, Grapperon, Bonnet, & Banks, W. P., Roberts, D., & Ciranni, M. (1995). Negative
Hasbroucq, 2003). Based on these findings, manual re- priming in auditory attention. Journal of Experimental
Psychology: Human Perception and Performance, 21,
sponse preparation and execution seems to go along with 1354–1361.
a combination of activation (of the correct response) and Behrendt, J., Gibbons, H., Schrobsdorff, H., Ihrke, M.,
inhibition (of the incorrect response) processes. An increase Herrmann, J. M., & Hasselhorn, M. (2010). Event-related
in prime response errors taking place in ignored repetition brain potential correlates of identity negative priming from
trials might therefore be ambiguous. First, prime response overlapping pictures. Psychophysiology, 47, 921–930.
errors could be the consequence of retrieved prime response Chao, H. F., & Yeh, Y. Y. (2008). Attentional demand and
activation. This comes closest to our original understanding memory retrieval in negative priming. Psychological
Research/Psychologische Forschung, 72, 249–260.
of this mechanism. Alternatively, prime response errors Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007).
could also be caused indirectly by retrieval of the inhibited G*Power 3: A flexible statistical power analysis program for
motor reaction to the prime distractor that, in turn, makes the the social, behavioral, and biomedical sciences. Behavior
execution of all alternative responses (including the prime Research Methods, 39, 175–191.
target response) more likely. Third, retrieval of activation Fox, E., & de Fockert, J. W. (1998). Negative priming depends
and inhibition could take place simultaneously. on prime-probe similarity: Evidence for episodic retrieval.
Psychonomic Bulletin & Review, 5, 107–113.
In contrast to response execution of one finger that can Frings, C. (2011). On the decay of distractor-response episodes.
take place in parallel with motor suppression of the other Experimental Psychology, 58, 125–131. doi: 10.1027/1618-
fingers, there is only one vocal tract that can either produce 3169/a000077
a verbal utterance or remain silent. Consequently, when a Frings, C., Rothermund, K., & Wentura, D. (2007). Distractor
vocal response is executed, simultaneous motor suppression repetitions retrieve previous responses to targets. Quarterly
of alternative vocal responses (e.g., the vocalization of the Journal of Experimental Psychology, 60, 1367–1377.
Gibbons, H. (2009). Functional brain-electrical correlates of
prime distractor name) seems to be impossible. This is a negative priming in the flanker task: Evidence for episodic
favorable property of the vocal response modality with retrieval. Psychophysiology, 46, 807–817.
respect to our concerns: The fact that in a vocalization task Gibbons, H., & Stahl, J. (2008). Early activity in the lateralized
participants more frequently respond with the former prime readiness potential suggests prime-response retrieval as a
response to the probes in ignored repetition trials implies source of negative priming. Experimental Psychology, 55,
that they must have retrieved prime response activation. 164–172.
In a nutshell, with manual responding, a prime response Giesen, C., & Rothermund, K. (in press). Affective matching
moderates S-R binding. Cognition & Emotion.
retrieval effect could still be mediated by retrieved inhibition Hommel, B. (1998). Event files: Evidence for automatic
of the prime distractor motor response, but with vocal integration of stimulus-response episodes. Visual Cognition,
responding the effect must be due to retrieval of prime 5, 183–216.
response activation. Hommel, B. (2004). Event files: Feature binding in and across
The conclusions that can be drawn from the reported perception and action. Trends in Cognitive Sciences, 8, 494–
experiment are not restricted to the negative priming phe- 500.
Hommel, B. (2005). How much attention does an event file
nomenon. Instead, the findings deliver some broader need? Journal of Experimental Psychology: Human Percep-
insights into the existence and significance of feature and tion and Performance, 31, 1067–1082.
response binding into event files in auditory perception Hommel, B., & Müsseler, J. (2006). Action-feature integration
and working memory. Retrieval of prime episodes including blinds to feature-overlapping perceptual events: Evidence

 2011 Hogrefe Publishing Experimental Psychology 2011; Vol. 58(5):353–360


360 S. Mayr et al.: Evidence of Vocal and Manual Event Files

from manual and vocal actions. The Quarterly Journal of Stolz, J. A., & Neely, J. H. (2001). Taking a bright view of
Experimental Psychology, 59, 509–523. negative priming in the light of dim stimuli: Further evidence
Jansen, P. A., & Watter, S. (2008). SayWhen: An automated for memory confusion during episodic retrieval. Canadian
method for high-accuracy speech onset detection. Behavior Journal of Experimental Psychology, 55, 219–230.
Research Methods, 40, 744–751. Tipper, S. P. (1985). The negative priming effect: Inhibitory
May, C. P., Kane, M. J., & Hasher, L. (1995). Determinants of priming by ignored objects. Quarterly Journal of Experi-
negative priming. Psychological Bulletin, 118, 35–54. mental Psychology: Human Experimental Psychology, 37A,
Mayr, S., & Buchner, A. (2006). Evidence for episodic retrieval 571–590.
of inadequate prime responses in auditory negative priming. Tipper, S. P. (2001). Does negative priming reflect inhibitory
Journal of Experimental Psychology: Human Perception and mechanisms? A review and integration of conflicting views.
Performance, 32, 932–943. Quarterly Journal of Experimental Psychology: Human
Mayr, S., & Buchner, A. (2007). Negative priming as a memory Experimental Psychology, 54A, 321–343.
phenomenon: A review of 20 years of negative priming Tipper, S. P., & Cranston, M. (1985). Selective attention and
research. Zeitschrift fur Psychologie/Journal of Psychology, priming: Inhibitory and facilitatory effects of ignored primes.
215, 35–51. Quarterly Journal of Experimental Psychology: Human
Mayr, S., & Buchner, A. (2010). Episodic retrieval processes take Experimental Psychology, 37A, 591–611.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

place automatically in auditory negative priming. European Tipper, S. P., MacQueen, G. M., & Brehaut, J. C. (1988).
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Journal of Cognitive Psychology, 22, 1192–1221. Negative priming between response modalities: Evidence for
Mayr, S., Buchner, A., & Dentale, S. (2009). Prime retrieval of the central locus of inhibition in selective attention. Percep-
motor responses in negative priming. Journal of Experimen- tion & Psychophysics, 43, 45–52.
tal Psychology: Human Perception and Performance, 35, Vidal, F., Grapperon, J., Bonnet, M., & Hasbroucq, T. (2003).
408–423. The nature of unilateral motor commands in between hand
Meynier, C., Burle, B., Possamai, C.-A., Vidal, F., & Hasbroucq, choice tasks as revealed by surface Laplacian estimation.
T. (2009). Neural inhibition and interhemispheric connec- Psychophysiology, 40, 796–805.
tions in two-choice reaction time: A Laplacian ERP study. Wong, K. F. E. (2000). Dissociative prime-probe contextual
Psychophysiology, 46, 726–730. similarity effects on negative priming and repetition priming:
Mondor, T. A., Leboe, J. P., & Leboe, L. C. (2005). The role of A challenge to episodic retrieval as a unified account of
selection in generating auditory negative priming. Psycho- negative priming. Journal of Experimental Psychology:
nomic Bulletin & Review, 12, 289–294. Learning, Memory, and Cognition, 26, 1411–1422.
Neill, W. T. (1997). Episodic retrieval in negative priming and Zmigrod, S., & Hommel, B. (2009). Auditory event files:
repetition priming. Journal of Experimental Psychology: Integrating auditory perception and action planning.
Learning, Memory, and Cognition, 23, 1291–3105. Attention, Perception, & Psychophysics, 71, 352–362.
Neill, W. T., & Valdes, L. A. (1992). Persistence of negative Zmigrod, S., & Hommel, B. (2010). Temporal dynamics of
priming: Steady state or decay? Journal of Experimental unimodal and multimodal feature binding. Attention,
Psychology: Learning, Memory, and Cognition, 18, 565– Perception, & Psychophysics, 72, 142–152.
576.
Neill, W. T., Valdes, L. A., Terry, K. M., & Gorfein, D. S. (1992).
Persistence of negative priming: II. Evidence for episodic
trace retrieval. Journal of Experimental Psychology: Learn- Received June 30, 2010
ing, Memory, and Cognition, 18, 993–1000. Revision received October 22, 2010
Park, J., & Kanwisher, N. (1994). Negative priming for spatial Accepted October 24, 2010
locations: Identity mismatching, not distractor inhibition. Published online February 11, 2011
Journal of Experimental Psychology: Human Perception and
Performance, 20, 613–623.
Pösse, B., Waszak, F., & Hommel, B. (2006). Do stimulus- Dr. Susanne Mayr
response bindings survive a task switch? European Journal
of Cognitive Psychology, 18, 640–651. Department of Experimental Psychology
Rothermund, K., Wentura, D., & De Houwer, J. (2005). Retrieval Institut für Experimentelle Psychologie
of incidental stimulus-response associations as a source of Heinrich-Heine-Universität Düsseldorf
negative priming. Journal of Experimental Psychology: 40204 Düsseldorf
Learning, Memory, and Cognition, 31, 482–495. Germany
Simon, J. R., & Rudell, A. P. (1967). Auditory S-R compatibility: Tel. +49 211 811-2270
The effect of an irrelevant cue on information processing. E-mail susanne.mayr@hhu.de
Journal of Applied Psychology, 51, 300–304.

Experimental Psychology 2011; Vol. 58(5):353–360  2011 Hogrefe Publishing

You might also like