Attentional and perceptual sources of the auditory attentional blink

Perception amp Psychophysics2005 67 (2) 195-208

A notable example of a limitation to attentional selec-tivity is manifest in the attentional blink (AB) phenome-non When two successive targets among nontargets (orfillers) in a rapidly presented sequence are to be reportedthe second target (usually called the probe) is often missedif it is presented within approximately 500 msec of the firsttarget (eg Broadbent amp Broadbent 1987 RaymondShapiro amp Arnell 1992) Whether the AB reflects a fun-damental limitation shared among some or all modalities isnot certain (for a discussion see eg Arnell amp Larson2002) In the present study we sought to contribute towardcharacterizing the generality of the AB phenomenon byexamining the degree to which the well-established fea-tures of the AB in the visual modality extend also to theauditory modality

Although there is a substantial body of work on visualAB (see McLaughlin Shore amp Klein 2001 for a re-view) a thorough understanding of its auditory counter-part has been hampered by a relative dearth of studiesCertainly the existence of auditory AB would constitutea challenge to theories that explain the AB purely interms of visual mechanisms (eg Shapiro Raymond ampArnell 1994) To date however the few results availablepoint to a relatively more inconsistent effect in the audi-tory modality than in the visual modality Some havegone so far as to claim that there is simply no AB for au-ditory sequences (Potter Chun Banks amp Muckenhoupt1998 see also Chun amp Potter 2001) others posit that theeffect is attenuated in audition relative to vision (egArnell amp Jolicœur 1999 Arnell amp Larson 2002 Soto-Faraco amp Spence 2002) whereas still others have re-ported a marked auditory AB (eg Duncan Martens ampWard 1997 Goddard amp Slawinski 1999) Although onbalance the evidence points to the existence of auditoryAB phenomenamdashand hence to the likelihood that theAB reflects a general limitation on cognitionmdashit is cer-tainly the case that a full functional characterization ofthe auditory AB is some way off (see Mondor 1998 fora discussion see also Arnell 2001)

In the present article interest centers on the role playedby the context in which the probe and the target appearmdashthat is the role played by the presence and nature of thefiller items in determining the magnitude of the blinkSuch interest arises from two sources (1) studies of thevisual AB in which fillers have been implicated as a keyvariable modulating the magnitude of the AB (eg Kawa-

195 Copyright 2005 Psychonomic Society Inc

Parts of this research were reported at the Experimental PsychologySociety (EPS) and the Canadian Society for Brain Behavior and Cogni-tive Sciences (BBCS) joint meeting in July 2000 in Cambridge This workwas supported by a grant from the Fonds de recherche sur la nature et lestechnologies du Queacutebec to ST Part of the research described here wasconducted at the Cardiff School of Psychologyrsquos Human Factor Labora-tory and received financial support from the United Kingdomrsquos Economicand Social Research Council in the form of a grant to DMJ FV receivessupport from the Natural Sciences and Engineering Research Council ofCanada Thanks are due Mary R Jones Pierre Jolicœur and two anony-mous reviewers for their very helpful comments We also thank RobertHughes and Katherine Gueacuterard for critical reading of an earlier draft ofthis article We are grateful to Sophie Lamontagne and Amelia Woodwardfor running the experiments Correspondence should be addressed toS Tremblay School of Psychology Universiteacute Laval Quebec City PQG1K 7P4 Canada (e-mail sebastientremblaypsyulavalca)

Attentional and perceptual sources of the auditory attentional blink

SEacuteBASTIEN TREMBLAY and FRANCcedilOIS VACHONUniversiteacute Laval Quebec City Quebec Canada

and

DYLAN M JONESCardiff University Cardiff Wales

When a rapid succession of auditory stimuli is listened to processing of the second of two succes-sive targets among fillers is often impaired a phenomenon known as the attentional blink (AB) Threeexperiments were conducted to examine the role of filler items in modulating the size of the auditoryAB using a two-alternative forced choice discrimination paradigm In the first experiment dual-streampresentations in which low- and high-pitch items were separated by six semitones were tested A tran-sient deficit in reporting the probe was observed in the presence of fillers that was greater when fillerswere in the same stream as the probe In the absence of a filler there was a residual deficit but thiswas not related to the time lag between the target and the probe In the second and third experimentsin which single-stream presentations were used a typical AB was found in the presence of homoge-neous fillers but heterogeneous fillers tended to produce a greater deficit In the absence of a fillerthere was little or no evidence of a blink The pattern of results suggests that other attentional and per-ceptual factors contribute to the blink

196 TREMBLAY VACHON AND JONES

hara 2003 Ross amp Jolicœur 1999) and (2) the general-izations suggested by a body of work on auditory stream-ing that suggests that fillers play a significant role in iso-lating or incorporating single events within streams (seeBregman 1990 for an overview) This body of workspeaks to the organization of auditory events into streamsof temporally extended auditory objects Evidence isemerging that the principles of auditory perceptual or-ganization play a key role in a range of cognitive phe-nomena particularly those related to the isolation ofevents within sequences For example auditory stream-ing has been shown to modulate cognitive effects suchas memory for pitch (Jones Macken amp Harries 1997)and the auditory suffix effect (Nicholls amp Jones 2002)In each of these cases the perception of particular itemscan be altered by modifying the extent to which theseevents are embedded within streams Given that a keycharacteristic of the AB paradigm is the isolation of twoevents within an auditory sequence it seems not too fan-ciful to suggest that streaming constructs could be ap-plicable here also

Auditory and visual modalities certainly differ in thetemporal coherence of the stimulus sequence The power-ful organizational processes manifest in auditory percep-tion can be construed as a means of deriving stabilityfrom an evanescent stimulus The binding and embed-ding of events into a temporally extended perceptual ob-ject is a key characteristic of audition This may haveparticular relevance in the AB paradigm inasmuch as itseems reasonable to expect that as a result of this promi-nent characteristic the effect of the context in which thetarget and the probe appear will be particularly influen-tial Evidence for context effects in the perception of theorder of simple auditory sequences is compelling Takethe perception of order for two immediately successivetones differing only slightly in pitch When these arepresented in isolation there may be little difficulty indiscerning the order However if two other tones flankthe pitches of interest the order now becomes muchmore difficult to discern Adding yet more flankers sothat they now form a stream distinct from the tone pairsignificantly improves the perception of their order Theflanking stimuli may either capture or partition the tonepair depending on the relation in time and pitch to theflankers (see eg Bregman amp Rudnicky 1975)

The impact of context on the auditory AB is tested inthe present study through the manipulation of fillers thatsurround a target and a probe In the first experiment theeffect of the grouping of targets and fillers within dif-ferent streams was investigated In this experiment thegrouping of target items and fillers was studied in streamsdefined by pitch On each trial low- and high-pitch itemswere presented These could be just a target and a probeof different pitch presented in isolation or a target anda probe surrounded by fillers of the same pitch Suchmanipulations might lead the target or the probe to popout of the sequence thus making that item easier to dis-tinguish and select In Experiments 2 and 3 pitch wasnot used as a means of perceptual organization rather

grouping was promoted by manipulating the heterogene-ity among fillers Here only a single stream was formedand we manipulated the coherence of that stream bychanging its composition Homogeneous fillers (repeateditems) tend to be more coherent than heterogeneous fillers(changing items) The homogeneous fillers are thereforeperceived as a whole and tend to stream apart from thetarget and probe Again the perceptual organization ofthe target and the probe in relation to the fillers shouldaffect the magnitude of the AB

One of the key issues within the visual blink literatureis the contribution of perceptual interference by mask-ing The impact of masking in modulating the size of ABhas been tested extensively in the visual modality (seeEnns Visser Kawahara amp Di Lollo 2001 for a review)There is ample evidence that the item following the tar-get (T1 item)1 serves to mask the target and hence isa key determinant of the magnitude of the AB (egChun amp Potter 1995 Raymond Shapiro amp Arnell 1995Seiffert amp Di Lollo 1997) The action of the item fol-lowing the probe (the P1 item) also appears to be a crit-ical aspect of the AB (eg Giesbrecht amp Di Lollo 1998Jolicœur 1999a see also Kawahara Zuvic Enns ampDi Lollo 2003) There is a broad consensus with regardto the effects of masking in the visual AB (see eg Ennset al 2001) but evidence of interference effects bymasking is not yet available in the auditory AB (seeMondor 1998) In vision the absence of target maskingreduces the size of the blink but it does not abolish theeffect whereas the absence of probe masking does ap-pear to eliminate the AB (eg Giesbrecht amp Di Lollo1998 Jolicœur 1999a) In the first experiment of the se-ries that follows rather than exploring specifically therole of T1 and P1 items we compared the effect ofthe presence and absence of the whole filler set There-fore the emphasis in the present study was on investi-gating the role of the context rather than the role of asingle posttarget mask The impact of masking is diffi-cult to dissociate from that of streaming both point tothe contribution of perceptual sources to the AB (see Ar-nell amp Larson 2002 for a related discussion)

That perceptual organization plays a major role inmodulating the AB may be taken to imply that the ABwill be subject to the structural constraints of the partic-ular modality and hence will take different forms in thevisual and the auditory modalities Typically it is claimedthat central rather than modality-specific mechanismsare responsible for the AB (eg Jolicœur amp DellrsquoAcqua1998 see also Ruthruff amp Pashler 2001 for a discus-sion) For example the AB has been interpreted as a pro-cess related to the transfer of events from a sensory orperceptual encoding stage into short-term storage andinterference is assumed to result either from confusion ina short-term memory (STM) overloaded with stimuli(eg retrieval competition theory Shapiro et al 1994)or from a bottleneck arising from the serial nature of thetransfer process leading to a failure of consolidation inSTM (eg Chun amp Potter 1995 see also Jolicœur 1999a1999b) As we already mentioned some approaches

AUDITORY ATTENTIONAL BLINK 197

suggest that the AB is restricted to vision taking place invisual STM (VSTM) where filler items compete with theprobe for retrieval thus inducing the ensuing AB (Shapiroet al 1994) However it is plausible that the auditorymodality is endowed with a mechanism that is equivalentto VSTMmdashnamely the echoic short-term store (egCrowder amp Morton 1969)mdashwithin which confusion mayoccur between the auditory stimuli competing for retrieval(Duncan et al 1997 for a contrary view see Mondor ampTerrio 1998 and Nicholls amp Jones 2002) These notionsof similar but separate modality-based processes are con-sistent with a central process (Jolicœur 1999b) The con-cept of central limitations to STM consolidation does notnecessarily imply that the character of the AB will be uni-versal and independent of modality indeed modality-specific factors can modulate the limits on central pro-cesses (see eg Arnell amp Larson 2002 Shih 2000)

The present experimental series began with a partialreplication and extension of a frequently cited study thatof Duncan et al (1997) in which auditory AB was foundIn Experiment 1 we extended the range of variables usedby Duncan et al by exploring whether the presence or ab-sence of fillers plays a role in determining the magnitudeof the auditory AB

EXPERIMENT 1

Duncan et al (1997) demonstrated a compelling audi-tory AB In addition to replicating that original findingin Experiment 1 we sought to show the extent to whichacoustic context is important to the detection of a probeThe same general procedure and auditory stimuli as thoseused by Duncan et al were employed here In Duncanet alrsquos study targets and fillers were presented in twoconcurrent auditory sequences each of a different pitchOne set of targets (nab or nap) was presented at a lowpitch embedded in a sequence of filler items at the samepitch as the target where fillers consisted of repetitionsof the item guh Another set of targets (cod or cot) waspresented at high pitch again with repetitions of thefiller item guh but this time at the same high pitch of thetarget Coupled to a fast rate of presentation the pitchdifference was of sufficient magnitude that the sequenceswere likely to be perceived as two separate streams2 Theexperiment was designed to distinguish the influence ofthe presence of the fillers independently for the targetand the probe That is fillers could be presented eitherwith both sequences be absent from the sequence con-taining the probe (but present with the target) absentfrom the sequence containing the target (but present withthe probe) or absent from both sequences leaving thetarget and the probe in isolation (see Figure 1)

These manipulations allowed us to examine whetherauditory AB-like effects can be observed even in the ab-sence of any filler According to the existing literatureAB deficits should be eliminated when masks are re-moved but there might be some cost due to the pitch dis-tance between the target and the probe given the re-

quired reallocation of attention from for example a low-pitch target to a high-pitch probe (see eg Mondor ampBregman 1994) At the same time the design allowssome separate estimation of the extent to which the re-spective filler contexts of the target and the probe con-tribute to the magnitude of the blink These comparisonsmay also be informative about the relative roles of targetand probe masking in the auditory AB

MethodParticipants Twenty-four volunteers each of whom reported

normal hearing were recruited from Cardiff University studentsThey received course credit for their participation

Materials There were two concurrent rapid auditory presenta-tions (RAPs) one high pitch and one low pitch The low-pitch se-quence contained a single instance of either nab or nap surroundedby fillers (guh) or no filler The high-pitch sequence contained asingle target either cod or cot with or without fillers First all theitems (nab nap cod cot and guh) were digitally recorded in a malevoice Great care was taken to produce the vowels at an even pitch(using a reference pure tone of 1038 Hz available to the malespeaker at recording) and level All the samples were digitallyedited to 16-bit resolution at a sampling rate of 48 kHz and werecompressed in order to last exactly 150 msec Each filler was a dig-ital copy of an original The high-pitch sequence was created by adigital-processing method (using Sound Designer II software) thatshifted the items (the targets cod and cot and the filler guh) up inpitch by six semitones Importantly these digital transformationsdid not decrease the intelligibility of the individual items

Syllables lasted for 150 msec each and were separated by silentgaps of 100 msec The sequences were presented in such a way thatthe items in the high- and the low-pitch sequences were slightly outof phase with one another One of the two concurrent sequenceschosen at random on each trial began 125 msec before the otherThe first sequence to begin also contained the target presented afterfive filler items or in the case in which no filler was presented afterthe corresponding silent interval The probe was presented follow-ing delays of 125 625 or 1375 msec measured from target onsetto probe onset (stimulus onset asynchrony or SOA) Each sequencecontained 13 items per stream and lasted 3275 msec Within eachsequence the rate of presentation was 250 msec per item (150 msecon 100 msec off )

There were four experimental conditions in Experiment 1 andthey are referred to here by a notation that signifies the relation ofthe target (T) and the probe (P) to the presence of fillers (F) in thesequence (see Figure 1) In the TF PF sequence both of the itemsfor report were embedded in a sequencemdashone of high pitch theother of low pitchmdashof fillers In one condition fillers surroundedonly the target with no filler around the probe (TF P) and in an-other fillers surrounded only the probe with no filler around the tar-get (T PF) In a fourth condition (T P) no filler was present

Experimental design The three independent variables weretreated as repeated measure variables task (single vs dual) targetndashprobe SOA (125 625 or 1375 msec) and filler distribution (TF PF T PF TF P or T P) Each participant took part in twosingle-task conditions (one with instructions to focus their atten-tion on the low-pitch stimuli and one with instructions to focus onthe high-pitch stimuli) and one dual-task condition in which theparticipants were told to report both targets These conditions wereconducted in blocks and there was one experimental block per taskcondition the order of the three blocks being counterbalancedacross participants using a Latin square procedure Within eachblock targetndashprobe SOAs and filler distribution variables were ran-domized There were 96 trials in each block preceded by 24 prac-tice trials It is important to note that the items to be detected in thesingle-task condition corresponded to the probe items from the

198 TREMBLAY VACHON AND JONES

dual-task condition thus making the two conditions comparableThere was always one na word (nab or nap) and one co word (codor cot) item on each trial regardless of the task condition

Procedure The participants had to perform a two-alternativeforced choice discrimination they were told to discriminate betweennab and nap for the na words and between cod and cot for the cowords The participants initiated a trial by a mouse click on the startbutton displayed on the computer screen Presentation of the audi-tory stimuli began after a fixed delay of 250 msec The stimuli werepresented via headphones at approximately 65 dB(A) with a PCcomputer running a Visual Basic 50 program The participantsrsquo re-sponses were typed in following the presentation of each set of stim-

uli using keys labeled appropriately for the targets Under the single-task conditions there was a single response identifying the attendedtarget The participants were instructed to focus on either the na orthe co targets Under dual-task conditions two responses were typedin in either order The participants were encouraged to take time withtheir responses taking care that typing errors were not made

ResultsTarget discrimination In the dual-task condition

the mean target accuracy was 899 A repeated mea-sures analysis of variance (ANOVA) was performed on

Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh GuhNap

Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh GuhCod

Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh GuhNap

Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh Guh GuhCod

Cod

Cod

Nap

Nap

150 msec 100 msec

125 msec

150 msec 100 msec

125 msec

0 msec 3275 msecSOA 125 msec

0 msec 3275 msec

0 msec 3275 msec

0 msec 3275 msec

Low Pitch

High Pitch

TF + PF

TF + P

T + PF

T + P

Figure 1 Schematic diagram illustrating the four dual-stream presentations employed in Experiment 1TF PF TF P T PF and T P (examples at an SOA of 125 msec) T target P probe

AUDITORY ATTENTIONAL BLINK 199

these data with SOA (three levels) and filler distribution(four levels) as factors There were significant effects forboth SOA [F(246) 1039 MSe 001 p 01] andfiller distribution [F(369) 495 MSe 0011 p 01] but the interaction between the two factors was notsignificant ( p 71) Target discrimination was at itslowest level with TF P and TF PF when the targetwas embedded with fillers (see Table 1)

Probe discrimination The data in the single-taskconditions were pooled over targets (co and na as singletargets) and were analyzed only for trials in which thetarget was in the second stream to begin with (the samestream as that for the probe in the dual-task conditionthis precaution was taken in order to make single- anddual-task conditions comparable) Probe accuracy in thedual-task condition was computed only for trials on whicha correct discrimination was made for the target Theprobability of identifying the probe given correct iden-tification of the target is presented in Figure 2 as a func-tion of SOA task and filler distribution A repeatedmeasures ANOVA was carried out on the data with taskSOA and filler distribution conditions The main effectsof task [F(123) 8317 MSe 0038 p 01] andSOA [F(246) 1537 MSe 0019 p 01] were sig-nificant but that of filler distribution did not reach sig-nificance [F(369) 223 MSe 0015 p 09] Theinteraction between SOA and filler distribution was sig-nificant [F(6138) 588 MSe 0017 p 01] butthere was no significant interaction between task andfiller distribution ( p 55)

The interaction between SOA and task which is theempirical signature of the AB was significant [F(246) 812 MSe 0019 p 01] Of particular importancefor the purposes of the present experiment is the signif-icant three-way interaction between task SOA and fillerdistribution [F(6138) 651 MSe 0015 p 01]This significant three-way interaction reinforces the im-pression given by a visual inspection of Figure 2 that thecritical relation between SOA and task differs accordingto filler distribution Here and applicable elsewhere inthis study a stricter alpha level of 0125 was used inorder to compensate for the family-wise error Furtheranalysis of the three-way interaction revealed that theinteraction between task and SOA was significant with

TF PF [F(246) 580 p 01] at T PF [F(246) 1286 p 01] did not reach significance with TF P[F(246) 399 p 03] but was far from significantat T F [F(246) 016 p 79] In the absence offillers the difference between single and dual tasks wassignificant ( p 01) but it did not interact with SOA

Context effects The critical interaction between SOAand task that reflects the time-related nature of AB-likedeficits was not significant for filler distributions inwhich there was no filler in the probe stream (T P andTF P) but was highly significant for distributions inwhich fillers surrounded the probe (T PF and TF PF) In order to test the relative impact of the context onthe dual-task cost in a target stream (absence or presenceof fillers) and a probe stream (absence or presence offillers) the conditions were contrasted with respect toSOA using the difference in performance between sin-gle (control) and dual tasks as the dependent variable(see Figure 3) A repeated measures ANOVA with targetstream (two levels) probe stream (two levels) and SOA(three levels) as factors was carried out on the data Themain effect of target stream was not significant [F(123) 029 MSe 0048 p 60] The effect of SOA was sig-nificant [F(246) 1739 MSe 0029 p 01] butmore important the effect of probe stream was also sig-nificant [F(123) 1496 MSe 0026 p 01] Therewas no significant interaction between target stream andprobe stream [F(123) 166 MSe 0038 p 21]The interaction of target stream and SOA also was notsignificant [F(246) 034 MSe 0033 p 71] norwas the three-way interaction significant [F(246) 061MSe 0026 p 55] However the interaction betweenprobe stream and SOA reached significance [F(246) 503 MSe 0034 p 01] Simple main effects wereperformed on the data in order to pinpoint the source ofthe interaction between SOA and probe stream The dif-ference between absence and presence of fillers in theprobe stream was significant at both 125 msec [t(23) 327 p 01] and 625 msec [t(23) 288 p 01] butwas not significant at the longest SOA [t(23) 071p 48] The effect of SOA was significant when therewere fillers in the probe stream [F(246) 1533 p 01]but was not significant when no filler surrounded theprobe [F(246) 237 p 12]

DiscussionThe results in Duncan et al (1997) in relation to the

auditory AB were replicated with the demonstration ofan SOA-related deficit in identifying the probe (givencorrect identification of the target) within a dual-streampresentation In relation to the manipulation of contextthe results of Experiment 1 may be summarized as fol-lows The context in which targets appear is importantIndeed when there is no filler there is no blink (al-though there is a penalty to be paid for switching be-tween stimuli differing in pitch that is roughly the sameat all SOAs) In some sense therefore context deter-mines the blink However the effect appears to be asym-

Table 1Probability of Target Report (SE) Under a Dual Task

as a Function of Stimulus Onset Asynchrony (SOA) (125 625 and 1375 msec) and Filler Distribution

(TF PF TF P T PF and T P)

SOA

125 msec 625 msec 1375 msec

Filler Distribution p SE p SE p SE

TF PF 906 019 891 022 813 025TF P 912 018 912 018 849 024T PF 912 021 922 018 875 020T P 953 015 943 017 906 019

NotemdashT target P probe F filler

200 TREMBLAY VACHON AND JONES

metric the context provided for the probe plays a morepowerful role than does the context provided for the tar-get (see Figure 3)

The results of Experiment 1 suggest that masking ofthe probe is more important than masking of the targetat least in the auditory AB A greater blink was observedwhen the probe was surrounded by fillers of the samepitch (T PF) than when the target was surrounded byfillers (TF P) Also in line with this finding is the dem-onstration by Mondor (1998) that the auditory AB is sub-stantial even when the stimulus immediately following thetarget is replaced with a silent gap Mondor suggests thatthe survival of the blink in the absence of a T1 item re-flects a bottleneck limitation at the late stage of responseselection (see Jolicœur 1998 1999a 1999b) In supportof the latter view Vachon and Tremblay (in press)showed in a follow-up to Mondor that the auditory ABis eliminated when the stimuli following a probe are re-placed with a silent gap The question remains as towhether the observed time-related deficit reflects a trueAB or some other dual-task cost

Some authors claim that the character of a true AB isrevealed in its relationship with SOA In contrast to thepresent results in many studies there has been little im-pairment in reporting the probe when it was temporally

adjacent to the target (an effect often referred to as lag 1sparing) as compared with when there was one or morefiller items between the target and the probe Howeverthis U-shaped appearance of the relationship betweenAB and SOA is not universal In some studies this rela-tionship has been linear as in the present case with asignificant deficit in probe detection even when the tar-get and the probe were adjacent (eg Arnell amp Jolicœur1999) In the visual case lag 1 sparing is more likely tooccur when the target and the probe are presented in thesame spatial location whereas failure to report the lag 1probe seems to occur when there is a location switch be-tween the target and the probe (eg Visser Bischof ampDi Lollo 1999 but see Shih 2000) The same reasoningmay apply in audition to pitch distance in the case ofdual-stream presentations such as the one used here (seeMondor amp Bregman 1994) Thus one possibility is thatthe poorer performance with a probe at T1 (ie ab-sence of lag 1 sparing) is related to the cost of having toreallocate attention to a different frequency region

The same reallocation cost of having to switch atten-tion from the frequency region of the target to that of theprobe could also be responsible for the residual time-independent deficit observed in T P presentations Ifwe construe reallocation of attention not merely in terms

Figure 2 Results from Experiment 1 probability of probe report given correct re-port of the target as a function of task SOA and filler distribution (TF PF TF P T PF and T P) Error bars represent standard errors

AUDITORY ATTENTIONAL BLINK 201

of task requirements (eg Potter et al 1998) but also interms of a locus of operation (as when the target and theprobe are presented in different spatial locations seeVisser et al 1999) we gain some renewed insight intothe effects of context Studies with single targets revealthat detection is impaired (mostly slowed reaction time)if attention is cued away from the target Such a cost inreallocating attention to the appropriate locus has beenobserved with visual and auditory stimuli cuing spatiallocation (eg Spence amp Driver 1994) as well as withauditory stimuli cuing frequency (eg Mondor amp Breg-man 1994 see also Woods Alain Diaz Rhodes ampOgawa 2001) The penalty of this form of switching isreflected in the small deficit not related to time observedin the T P condition in Experiment 1

It is plausible that the pitch-switching3 cost is addi-tional to the AB as is suggested by the greater deficitobserved in the presence of fillers (see Chun amp Potter2001 for a discussion on the issue of additivity) Thatpitch switching may have been partly responsible for theAB found in Experiment 1 does not rule out the possi-bility of finding an effect in a single stream There issome evidence although limited of auditory AB withina single-stream presentation (eg Arnell amp Jolicœur1999 Experiment 4 Arnell amp Larson 2002 Soto-Faracoamp Spence 2002) and in the next experiment of the pres-ent series we attempted to provide further evidence thatan auditory AB can be observedmdashthis time using single-stream presentations A key objective of the followingexperiments was to further characterize the effect of con-text by exploring the impact of the nature of the fillers onthe auditory AB

EXPERIMENT 2

In Experiment 2 we followed a procedure similar tothat employed in Experiment 1 and used the same dis-crimination task with the same target items However in-stead of dual-stream presentations a single-stream pre-sentation was used in which there was no overlap in timebetween successive stimuli Here the effect of contextwas manipulated by the homogeneity of the filler se-quence the items being either repeated or varying InExperiment 2 the SOA was 150 msec and there werelag 1 lag 2 and lag 4 conditions so that potentiallylag 1 sparing could be observed Also the role of pitchswitching was minimized by the use of a single streamrather than a dual stream

One of the key manipulations in Experiment 2 was ofthe composition of the filler sequences these could be aheterogeneous set (guh gah gih) or a homogeneous setconsisting of repetition of the same filler (guh) This wasdone with the aim of trying to understand the role of per-ceptual organization and in particular the grouping ofsounds into streams Experiment 1 pointed to the role ofthe context in which a probe appeared One interpreta-tion would be that the role of fillers is to mask the probeitem and make its recognition more difficultmdashhence thelack of a blink when the fillers are removed (for exam-ple in the T P condition in Experiment 1)

From the standpoint of perceptual organization (espe-cially in the auditory domain) the similarity of the filleritems determines the likelihood of their grouping witheach other There is more than ample evidence that stream-ing depends on the coherence of a sequence and we may

Target stream Probe stream

SOA (msec) SOA (msec)

Dua

l-tas

k co

st (

)

No FillerFillers

No FillerFillers

35

30

25

20

15

10

5

0

35

30

25

20

15

10

5

0125 625 1375 125 625 1375

Figure 3 Results from Experiment 1 dual-task cost as a function of SOA target streamand probe stream This cost corresponds to the difference in performance between singleand dual tasks Error bars represent standard errors

202 TREMBLAY VACHON AND JONES

expect that the nature of grouping depends on the degreeof change in a sequence It is expected that repeatedfillers will tend to be grouped together and make the tar-get and the probe stream apart from that sequence Apossible consequence of the latter streaming outcome isthat target discrimination will be facilitated On the otherhand heterogeneity among fillers should lower the prob-ability that filler items cohere and stream out of targetitems Hence the magnitude of the auditory AB mightbe larger in the presence of changing fillers than in thepresence of repeated fillers However such an increasedAB may also be the result of more items (changing fillers)competing for limited memory or processing resourcesor causing more confusion in some auditory STM

Just as in Experiment 1 the design in Experiment 2 in-corporated a control condition in which there was nofiller Without any filler a single-stream presentation inwhich all items are played at the same pitch (thereby cir-cumventing a possible pitch-switching cost) should notproduce any AB-like deficit

MethodParticipants Twenty-four students volunteered to take part in

this experiment Each reported normal hearing They received ei-ther a small honorarium or course credit for their participation

Materials The stimuli were the same as those employed in Ex-periment 1 with the following exceptions All the items were pre-sented at the same pitch (low voice at 1038 Hz) the items werewithin a single sequence and in the changing-filler condition therewere three different filler items presented in a fixed order (guhgah and gih) There was no gap between successive items Theprobe was presented following one of four delays of 150 msec (ad-jacent or lag 1) 300 msec (lag 2) 600 msec (lag 4) or 1350 msec(lag 9) These were measured from target onset to probe onset Se-quences with fillers were all made up of 17 items in total there weresix items preceding the target item and that item was then followedby 10 items (including one probe and nine fillers) There was eitherno filler (lag 1) or one (lag 2) three (lag 4) or eight (lag 9) fillersbetween the target and the probe

Experimental design There were three repeated measures vari-ables task (single vs dual) targetndashprobe SOA (150 300 600 or1350 msec) and homogeneity of fillers (no filler repeated filleror changing fillers) Each participant took part in two single-taskconditions (one with instructions to attend to the na word and theother to attend to the co word) and the dual-task condition (with in-structions to attend to both targets) There was one experimentalblock per task condition and the order of the three blocks was coun-terbalanced across participants using a Latin square design Withineach block SOA and homogeneity conditions were randomizedfrom trial to trial There were 96 trials in each block preceded by24 practice trials

Procedure The procedure was the same as that in Experiment 1

Results and DiscussionTarget discrimination On average the target was re-

ported correctly on 961 of the trials in the dual-task con-dition An ANOVA with SOA (four levels) and homo-geneity (three levels) as repeated measures factors wascarried out on the data None of the effects was significant[SOA F(369) 149 MSe 0014 p 24 homogene-ity F(246) 091 MSe 0016 p 41 and SOA ho-mogeneity F(6138) 199 MSe 0011 p 07]

Probe discrimination The probability of reportingthe probe as a function of task and SOA is displayed inFigure 4 for the no-filler repeated-filler and changing-filler conditions The means were calculated using onlythose trials in which report of the target was correct Thedata were submitted to a repeated measures ANOVAwith task (two levels) SOA (four levels) and homogene-ity (three levels) as independent variables The analysisrevealed a main effect of SOA [F(369) 797 MSe 001 p 01] and a main effect of homogeneity [F(246) 952 MSe 0014 p 01] but the main effect of taskwas not significant [F(123) 309 MSe 0033 p 09] The interaction of task and homogeneity was signif-icant [F(246) 410 MSe 0008 p 02] whereas theinteraction of SOA and homogeneity was not significant[F(6138) 150 MSe 0012 p 21] The critical inter-action between SOA and task [F(369) 1065 MSe 001 p 01] and the three-way interaction [F(6138) 233 MSe 0008 p 04] were significant

Figure 4 Results from Experiment 2 probability of probe re-port given correct report of the target as a function of task SOAand homogeneity of fillers (no filler repeated and changing)Error bars represent standard errors

AUDITORY ATTENTIONAL BLINK 203

The homogeneity of f illers seems to modulate theinteraction of task and SOA A closer look at Figure 4suggests the occurrence of a typical AB in the presenceof fillers and the lag 1 sparing appears to be slightlygreater in the presence of repeated fillers Tests of sim-ple effects were performed on the data in order to de-compose the three-way interaction In the absence offillers the two-way interaction of task and SOA did notreach significance [F(369) 240 p 08] In thepresence of repeated fillers task interacted with SOA[F(369) 820 p 01] Further analysis revealed thefollowing pattern The effect of task was not significantat the shortest SOA or lag 1 [t(23) 014 p 89] butwas significant at the 300-msec SOA [t(23) 356 p 01] and then did not reach significance at 600 msec[t(23) 023 p 82] and 1350 msec [t(23) 106p 30] In the case of changing fillers again the inter-action of task and SOA was significant [F(369) 482p 01 and the pattern of results was similar to the latterThe effect of task was marginally significant at 150 msec[t(23) 249 p 02] significant at the following SOAof 300 msec [t(23) 295 p 01] but not significantthereafter ( ps 20)

Homogeneity of fillers The relative impact of het-erogeneous and homogeneous fillers on the dual-taskcost was tested using the difference in performance be-tween single and dual tasks as the dependent variable forthe repeated- and changing-filler conditions at the SOAsthat revealed an AB-like deficit (ie 150 and 300 msec)A repeated measures ANOVA with homogeneity (twolevels repeated vs changing fillers) and SOA (two lev-els 150 and 300 msec) as factors was carried out on thedata (see Figure 5) The main effect of homogeneity wasnot significant [F(123) 041 MSe 0029 p 53]The effect of SOA was significant [F(123) 543 MSe

0034 p 03] and so was the interaction of homogene-ity and SOA [F(123) 464 MSe 0019 p 04]Decomposition of the interaction revealed the followingpattern The difference between repeated and changingfillers was marginally significant at 150 msec [t(23) 193 p 07] but was far from significant at 300 msec[t(23) 085 p 41] As for the effect of SOA thedifference between 150 and 300 msec was significant inthe presence of repeated fillers [t(23) 300 p 01]but was not significant with changing fillers [t(23) 059 p 56] The latter result reflects the observedlag 1 sparing

The results of Experiment 2 provide further evidencethat AB-like deficits within the auditory modality arenot restricted to dual-stream presentations (see also egArnell amp Jolicœur 1999 Arnell amp Larson 2002) Bypresenting the various filler conditions (no filler re-peated filler and changing fillers) randomly from trial totrial the possibility that the AB effects observed here(and also in Experiment 3) were the result of preparatorystrategy differences is ruled out The presence of fillersseems to be a necessary condition for the blink to occurand the perceptual organization of the fillers appears tohave a differential effect on the blink response Therewas no residual dual-task cost in the absence of filleritems this time within a single-stream presentationWithout a change of pitch between the target and theprobe there was very little difference in discriminationperformance between single- and dual-task conditionsBoth repeated- and changing-filler presentations seemto show the U-shaped function of task and SOA the so-called signature of the true AB (although the sparing oflag 1 appears to have been larger for the repeated-fillerpresentations) The magnitude of the blink was reducedalbeit only slightly when the auditory sequence was ho-mogeneous Fillers in the heterogeneous condition werephonologically similar at their onset which may have di-minished the degree of streaming between homogeneousand heterogeneous filler streams In Experiment 3 an at-tempt was made to enhance the effect of heterogeneityon the size of the AB by increasing the degree of changebetween fillers in the changing-filler presentations

EXPERIMENT 3

The design and procedure in Experiment 3 were thesame as those employed in Experiment 2 except for thestimuli used as fillers In Experiment 2 given that the threesyllables were presented in fixed order and shared a simi-lar onset (guh gah and gih) the degree of change amongthem can arguably be regarded as moderate and hence in-sufficiently heterogeneous to break up the coherence of theauditory stream In the present experiment we used a setof six items phonologically dissimilar at both onset andoffset and the presentation order of these items withinfiller sequences was randomized rather than arranged inthe same fixed and predictable order thereby diminishingthe coherence of the filler sequences still further

SOA (msec)

Dua

l-tas

k co

st (

)

RepeatedChanging

20

15

10

5

0150 300

Figure 5 Results from Experiment 2 dual-task cost as a func-tion of SOA (150 and 300 msec) and homogeneity (repeated vschanging fillers) This cost corresponds to the difference in per-formance between single and dual tasks Error bars representstandard errors

204 TREMBLAY VACHON AND JONES

MethodParticipants Twenty-four students volunteered to participate in

this experiment in exchange for a small honorarium Each reportednormal hearing

Materials As in Experiment 2 all the items were presented atthe same pitch within a single sequence In the changing-filler con-dition there were six different fillers (guh kev del rum tic andbow) presented in a random order In the repeated-filler conditionone of the six items was presented repeatedly The filler to be re-peated was chosen randomly with the restriction that all six werepresented at least twice in the experimental session The construc-tion of the individual items and the item sequences were the sameas those in Experiment 2 in terms of both procedure and param-eters As in the previous experiment the probe was presented fol-lowing one of four delays of 150 msec corresponding to lag 1 24 or 9 The experimental design and the procedure were the sameas those described in the previous experiments

Results and DiscussionTarget discrimination On average performance at

target discrimination in the dual-task condition was 842A repeated measures ANOVA with SOA (four levels)and homogeneity (three levels) as factors showed a sig-nificant effect of homogeneity [F(246) 607 MSe 0024 p 01] but neither the effect of SOA [F(369) 245 MSe 0013 p 07] nor the interaction betweenhomogeneity and SOA [F(6138) 038 MSe 002p 82] was significant Post hoc comparisons (Bonfer-roni t tests alpha 01) were carried out on the main ef-fect of homogeneity Performance in the presence ofchanging fillers (800) was significantly worse thanperformance in the presence of repeated fillers (846)and in the absence of fillers (878)

Probe discrimination The probability of reportingthe probe (given a correct report of the target) as a functionof task (two levels) SOA (four levels) and homogeneity(three levels) was analyzed with a three-way repeatedmeasures ANOVA The data are displayed in Figure 6All three main effects were significant [task F(123) 1205 MSe 0051 p 01 SOA F(369) 1143MSe 0018 p 01 homogeneity F(246) 4109MSe 0019 p 01] The interaction between task andhomogeneity was significant [F(246) 495 MSe 0016 p 05] as was that between SOA and homo-geneity [F(6138) 616 MSe 0012 p 01] Im-portantly the interaction between SOA and task was sig-nificant [F(369) 1240 MSe 0011 p 01]

The three-way interaction of task SOA and homo-geneity of fillers was also significant [F(6138) 330MSe 0012 p 01] The composition of the three-wayinteraction was revealed by further analysis on the data(alpha 01) Again the two-way interaction of task andSOA was not significant in the absence of fillers [F(369) 019 p 90] The same interaction was significant inthe repeated-filler condition [F(369) 280 p 05]and also in the changing-filler condition [F(369) 1332 p 01] The pattern of significance in relation tothe effect of task at each SOA was different whether thefillers were repeated or changing (Bonferroni t testsalpha 01) With changing fillers there was a signifi-cant deficit at the first two SOAs (150 and 300 msec)

but this was far from significant thereafter whereas asignificant difference was observed only at 150 msec inthe repeated condition

Homogeneity of fillers As in Experiment 2 a re-peated measures ANOVA with homogeneity (two levelsrepeated vs changing fillers) and SOA (two levels 150or 300 msec) as factors was performed on the dual-taskcost in order to test the impact of the heterogeneity of thefillers on the blink A visual inspection of the data suggeststhat heterogeneous fillers were more disruptive than ho-mogeneous fillers at both SOAs (see Figure 7) Bothmain effects were significant [homogeneity F(123) 1244 MSe 0027 p 01 SOA F(123) 740MSe 0030 p 01] The interaction of homogeneityand SOA was not significant [F(13) 038 MSe 0033 p 54]

As was the case in Experiment 2 there was a markeddifference between single and dual tasks in terms ofprobe report and importantly that difference was relatedto SOA The latter finding provides strong evidence forthe existence of the AB in the auditory modality Our at-tempt to increase the difference between repeated- andchanging-filler conditions in relation to the size of theblink was successful In Experiment 2 a small set ofchanging fillers with a low degree of heterogeneity yielded

Figure 6 Results from Experiment 3 probability of probe re-port given correct report of the target as a function of task SOAand homogeneity of fillers (no filler repeated and changing)Error bars represent standard errors

AUDITORY ATTENTIONAL BLINK 205

a slightly greater blink than repeated fillers did whereasin the present experiment heterogeneous fillers pro-duced a deficit of significantly greater magnitude to thatproduced by repeated fillers Overall accuracy in Exper-iment 3 was lower than that observed in Experiment 2One possible reason for the latter result is that using adesign in which filler conditions are mixed led to carry-over effects of heterogeneous trials onto repeated trialsA rather puzzling result is that we did not replicate thesparing effect at the shorter SOA observed in Experi-ment 2 These results require further exploration

GENERAL DISCUSSION

A substantial deficit in reporting the second of two au-ditory targets was observed in both dual-stream and single-stream auditory sequences In all three experiments itwas demonstrated that the magnitude of these effects de-pends critically on the context provided by the items thatsurround the stimuli to be reported Experiment 1 con-tributed to an understanding of those contextual factorsin a setting that involved switching attention betweenstreams Presenting the probe embedded with fillers ofthe same pitch increased the AB appreciably Experi-ments 2 and 3 served to clarify the role played by the na-ture of fillers within a single-stream presentation Whetherthe fillers were homogeneous or heterogeneous had asignificant influence on the magnitude of the blink Gen-erally lag 1 and lag 2 probes are susceptible to the effectsof the coherence of the filler sequence When the fillerswere identical (homogeneous conditions in Experiments 2and 3) or phonologically similar and predictable (het-erogeneous condition in Experiment 2) the deficit wassmall but when the fillers were phonologically hetero-geneous and random (heterogeneous condition in Ex-periment 3) the deficit was large These effects of con-

text suggest that perceptual organization contributes tothe overall auditory AB Certainly there are similaritieswith the visual modality but it is not clear whether thesame factors modulate the effect or indeed whetherthere is a general sensitivity to context

Implications for Theoretical Accounts of the ABMost recent models of the AB cleave along the distinc-

tion between retrieval competition (eg Raymond et al1995 Shapiro et al 1994) and bottleneck type of process-ing (eg Chun amp Potter 1995 Jolicœur amp DellrsquoAcqua1998) In short the bottleneck interpretation assumesthat target stimuli are processed in a serial fashion4

While resources are occupied with the target there is aperiod during which no or fewer attentional resources areavailable for processing those stimuli that closely followthe target According to the other type of explanationthe retrieval competition model each item is processedto some degree but only a few items are transferred toSTM Items may be admitted into STM if they match apreset template of the target or the probe and if they aretemporally contiguous to the target or the probe Thistype of interpretation is similar to theoretical accountsin the STM literature that assume that interference iscaused by a similarity of content between target stimuli(eg to-be-remembered items) and nontargets (eg irrel-evant sound see Jones amp Tremblay 2000 for a review)

Bottleneck models are in agreement that the AB arisesfrom attentional limitations but differ with respect to thestage at which to place the bottleneck Jolicœurrsquos (1998see also Jolicœur amp DellrsquoAcqua 1998) central interfer-ence theory places the bottleneck at a stage that of short-term consolidation (STC) later than that at which it is lo-cated in the two-stage model (Chun amp Potter 1995) Akey assumption specific to the central interference the-ory is that STC requires central processing Thereforewhile the target stimulus goes through STC any otheroperations (such as probe STC or task switch) that re-quire central processing have to wait (see Jolicœur 1998)Another possible source of the AB deficit is that task orlocation switching makes use of central resources thuspostponing the central processing required by STC of theprobe (see McLaughlin et al 2001) With a similar ar-gument the central interference theory can account forthe observed cost of pitch switching

Processes other than retrieval competition and con-solidation may also produce AB-like effects Some ver-sions of the blink paradigm may impose different pro-cessing requirements on the target and the probe Theblink may therefore be a product of switching from onetype of activity to another and this switch may be whollyor partly responsible for the loss of efficiency in report-ing the probe In all three experiments of the presentstudy (as in Duncan et al 1997) there was no task switchbetween the target and the probe both the target and theprobe required a qualitatively similar two-alternativeforced choice discrimination The same task and sametarget set size were used for the target and the probeTherefore the auditory AB observed here as well as that

SOA (msec)

Dua

l-tas

k co

st (

)

RepeatedChanging

35

30

25

20

15

10

5

0150 300

Figure 7 Results from Experiment 3 dual-task cost as a func-tion of SOA (150 and 300 msec) and homogeneity (repeated vschanging fillers) This cost corresponds to the difference in per-formance between single and dual tasks Error bars representstandard errors

206 TREMBLAY VACHON AND JONES

found by Duncan et al is unlikely to be caused by anamodal task switch (as claimed by Potter et al 1998)However the possibility remains that there was some costattributable to a switch of target set (eg co for the targetfollowed by na for the probe see Arnell amp Larson 2002)

In the present study we explored the impact of thecontext in which a target and a probe are presented on theauditory AB effects Context effects were tested throughthe manipulation of coherence across streams (Experi-ment 1 pitch segregation) and within a stream (Experi-ments 2 and 3 heterogeneity of fillers) Segregation bypitch (eg in T PF and TF P) leads to inserting theprobe and the target in different but coherent groups andheterogeneity is concerned with placing them into onegroup that is more or less coherent Whether it is throughstream segregation or within-stream coherence our re-sults revealed that perceptual organization plays a majorrole in modulating the auditory AB

Pitch segregation Experiment 1 A logic of percep-tual organization and grouping can be applied to accountfor the results of Experiment 1 on the basis of the princi-ple of similarity by frequency between target and filleritems (as opposed to similarity by repetition or phonol-ogy among fillers) Indeed one explanation is that audi-tory sequences in which the probe was surrounded byfillers of the same pitch (TF PF and T PF) yieldeda greater AB by making the probe harder to distinguishfrom the other items in that sequence On the contrarysequences in which the probe differed from the otheritems in pitch (TF P and T P) induced the probe topop out of the sequence and hence substantially re-duced the size of the AB

Bottleneck models can account for the results of Ex-periment 1 with the assumption that by popping out theprobe suffers less perceptual interference while waitingfor STC The retrieval competition model can also ex-plain the same pattern of results by assuming that the al-location of attentional resources is based on the similar-ity of targets with preset templates As the similaritybetween filler and target items increases the amount ofresources devoted to fillers temporally contiguous withthe targets also increases In Experiment 1 the probeand its surrounding fillers were presented at the samepitch that increased similarity between the probe and thefiller items resulted in more competition for retrievalamong the items admitted in STMmdashhence the greaterAB observed in the TF PF and T PF conditions Ac-cording to the retrieval competition model it is not clearwhy the probendashfiller similarity produced greater AB ef-fects than the targetndashfiller similarity did

Heterogeneity of fillers Experiments 2 and 3 Thereis very little research on the impact of placing targetsamong heterogeneous fillers on the AB In demonstrat-ing that the processing of color information is suscepti-ble to the AB Ross and Jolicœur (1999) provided datathat speak to the issue of heterogeneity Within a rapidserial visual presentation of color stimuli a probe thathad to be detected was immune from the AB when the

fillers were homogeneous and of a different color (egperformance at detecting a red probe among blue fillerswas very high at any lag) However when the probe wasembedded in fillers of heterogeneous colors a signifi-cant AB was observed The authors concluded that ef-fectiveness of masking is reduced when targets pop outof the filler stream (see also Arnell amp Jolicœur 1999)

The effect of heterogeneity on target detection hasbeen demonstrated on many occasions in visual searchstudies (see eg Bauer Jolicœur amp Cowan 1996 Dun-can 1980 Hoffman 1978 Pashler 1987 Treisman ampGelade 1980) Homogeneity among fillers usually makesthem easier to suppress when a target is searched for Inother words heterogeneous fillers make target detectionless likely The results from a variety of experimentsusing auditory analogues of the visual search paradigmprovide evidence that similar context effects are observedin auditory selection (eg Botte Drake Brochard ampMcAdams 1997 Mondor Zatorre amp Terrio 1998 seealso Cusack amp Carlyon 2003) Mondor and his colleagues(Mondor amp Terrio 1998 Mondor et al 1998) proposeda model of auditory selection attention based on the vi-sual selection theory of Duncan and Humphreys (1989)Stimuli that match a preset template are more likely to beselected as targets and other stimuli are rejected Howeverselection is constrained by the outcome of the perceptualorganization of stimuli into streams at some preattentivestage There is ample evidence that selection of informa-tion is preceded by preattentive perceptual organization

By enhancing the heterogeneity among the changingfillers (Experiment 3) we obtained a greater effect ofheterogeneity than that observed in Experiment 2 Thephonological similarity among heterogeneous f illersused in Experiment 2 may have promoted grouping al-though to a lesser extent than repeated fillers did Thatheterogeneous fillers provoke a greater auditory AB canbe ascribed to the effects of grouping Both the retrievalcompetition theory and the bottleneck models can ac-commodate the construct of grouping The developmentof the retrieval competition theory (eg Shapiro et al1994) has its roots in Duncan and Humphreysrsquos (1989)theory of visual selection It is assumed that competitionbetween target and filler items is influenced by princi-ples of grouping Homogeneous fillers will likely tend tobe grouped together and rejected as a whole (which in-cludes items following targets acting as masks) One canmake the assumption that as a consequence of the easierrejection of fillers as a group there is less competitionfor retrieval and the AB is therefore greatly reducedBottleneck models can be taken to posit that homogeneityamong fillers makes the target and the probe pop out ofthe sequence Hence the target is processed more effi-ciently and the delay before the processing of the probeis shorter In addition if the probe suffers less perceptualinterference by masking the probability of overwritingduring the delay is reduced

Another possibility that could account for the effectof heterogeneity observed in Experiments 2 and 3 is the

AUDITORY ATTENTIONAL BLINK 207

concept of stimulus uncertainty Research on the impactof auditory masking upon target identification has shownthat a mask is generally more disruptive with a high de-gree of stimulus uncertainty (see eg Durlach et al2003) Predictability of the frequency or location of theitem acting as a mask has been shown to influence the de-gree of masking efficiency (see Hawkins amp Presson1986) Therefore the effect of heterogeneity could alsobe explained by the fact that the identity of the maskswas unpredictable in the heterogeneous streams (espe-cially in Experiment 3)

ConclusionIn conclusion there is mounting evidence that more

than one factor is responsible for AB interference Or-ganization of stimuli in terms of spatial location (egSpence Ranson amp Driver 2000) pitch (results of Ex-periment 1) and heterogeneity (Experiments 2 and 3)has been shown to affect the AB The present study pro-vides evidence that there is a so-called conventional ABin the auditory modality and that its effect can be addi-tive to other effects such as pitch switching A promis-ing line of research might be to apply the principles andtheories of auditory selection (eg Mondor amp Terrio1998) to the study of the auditory AB

REFERENCES

Arnell K M (2001) Cross-modal interactions in dual-task para-digms In K Shapiro (Ed) The limits of attention Temporal con-straints in human information processing (pp 141-177) Oxford Ox-ford University Press

Arnell K M amp Jolicœur P (1999) The attentional blink acrossstimulus modalities Evidence for central processing limitationsJournal of Experimental Psychology Human Perception amp Perfor-mance 25 630-648

Arnell K M amp Larson J M (2002) Cross-modality attentionalblinks without preparatory task-set switching Psychonomic Bulletinamp Review 9 497-506

Bauer B Jolicœur P amp Cowan W B (1996) Distractor hetero-geneity versus linear separability in colour visual search Perception25 1281-1294

Botte M-C Drake C Brochard R amp McAdams S (1997) Per-ceptual attenuation of nonfocused auditory streams Perception ampPsychophysics 59 419-425

Bregman A S (1978) Auditory streaming is cumulative Journal ofExperimental Psychology Human Perception amp Performance 4380-387

Bregman A S (1990) Auditory scene analysis The perceptual orga-nization of sound Cambridge MA MIT Press

Bregman A S amp Rudnicky A I (1975) Auditory segregationStream or streams Journal of Experimental Psychology HumanPerception amp Performance 1 263-267

Broadbent D E amp Broadbent M H P (1987) From detection toidentification Response to multiple targets in rapid serial visual pre-sentation Perception amp Psychophysics 42 105-113

Chun M M amp Potter M C (1995) A two-stage model for multipletarget detection in rapid serial visual presentation Journal of Experi-mental Psychology Human Perception amp Performance 21 109-127

Chun M M amp Potter M C (2001) The attentional blink and taskswitching within and across modalities In K Shapiro (Ed) The lim-its of attention Temporal constraints in human information process-ing (pp 20-35) Oxford Oxford University Press

Crowder R G amp Morton J (1969) Precategorical acoustic storage(PAS) Perception amp Psychophysics 5 365-373

Cusack R amp Carlyon R P (2003) Perceptual asymmetries in audi-tion Journal of Experimental Psychology Human Perception amp Per-formance 29 713-725

Duncan J (1980) The locus of interference in perception of simulta-neous stimuli Psychological Review 87 272-300

Duncan J amp Humphreys G W (1989) Visual search and stimulussimilarity Psychological Review 96 433-458

Duncan J Martens S amp Ward R (1997) Restricted attentionalcapacity within but not between sensory modalities Nature 387808-810

Durlach N I Mason C R Kidd G Jr Arbogast T L Col-burn H S amp Shinn-Cunningham B G (2003) Note on infor-mational masking Journal of the Acoustical Society of America 1132984-2987

Enns J T Visser T A W Kawahara J amp Di Lollo V (2001) Vi-sual masking and task switching in the attentional blink In K Shapiro(Ed) The limits of attention Temporal constraints in human infor-mation processing (pp 65-81) Oxford Oxford University Press

Giesbrecht B amp Di Lollo V (1998) Beyond the attentional blinkVisual masking by object substitution Journal of Experimental Psy-chology Human Perception amp Performance 24 1454-1466

Goddard K M amp Slawinski E B (1999) Modality specific atten-tional mechanisms can govern the attentional blink Canadian Acous-tics 27 98-99

Hawkins H amp Presson J (1986) Auditory information processingIn K R Boff L Kaufman amp J P Thomas (Eds) Handbook of per-ception and human performance Vol II Cognitive processes andperformance (pp 1-64) Oxford Wiley

Hoffman J E (1978) Search through a sequentially presented visualdisplay Perception amp Psychophysics 23 1-11

Jolicœur P (1998) Modulation of the attentional blink by on-line re-sponse selection Evidence from speeded and unspeeded Task1 deci-sions Memory amp Cognition 26 1014-1032

Jolicœur P (1999a) Concurrent response-selection demands modu-late the attentional blink Journal of Experimental Psychology HumanPerception amp Performance 25 1097-1113

Jolicœur P (1999b) Restricted attentional capacity between sensorymodalities Psychonomic Bulletin amp Review 6 87-92

Jolicœur P amp DellrsquoAcqua R (1998) The demonstration of short-term consolidation Cognitive Psychology 36 138-202

Jolicœur P DellrsquoAcqua R amp Crebolder J M (2001) The at-tentional blink bottleneck In K Shapiro (Ed) The limits of atten-tion Temporal constraints in human information processing (pp 82-100) Oxford Oxford University Press

Jones D M Macken W J amp Harries C (1997) Disruption ofshort-term recognition memory for tones Streaming or interferenceQuarterly Journal of Experimental Psychology 50A 337-357

Jones D M amp Tremblay S (2000) Interference in memory by pro-cess or content A reply to Neath (2000) Psychonomic Bulletin ampReview 7 550-558

Kawahara J-I (2003) Mere presence of distractors Another deter-mining factor for the attentional blink Japanese Psychological Re-search 3 150-151

Kawahara J-I Zuvic S M Enns J T amp Di Lollo V (2003)Task switching mediates the attentional blink even without backwardmasking Perception amp Psychophysics 65 339-351

McLaughlin E N Shore D I amp Klein R M (2001) The atten-tional blink is immune to masking-induced data limits QuarterlyJournal of Experimental Psychology 54A 169-196

Mondor T A (1998) A transient processing deficit following selectionof an auditory target Psychonomic Bulletin amp Review 5 305-311

Mondor T A amp Bregman A S (1994) Allocating attention to fre-quency regions Perception amp Psychophysics 56 268-276

Mondor T A amp Terrio N A (1998) Mechanisms of perceptual or-ganization and auditory selective attention The role of pattern struc-ture Journal of Experimental Psychology Human Perception amp Per-formance 24 1628-1641

Mondor T A Zatorre R J amp Terrio N A (1998) Constraints onthe selection of auditory information Journal of Experimental Psy-chology Human Perception amp Performance 24 66-79

Nicholls A P amp Jones D M (2002) Capturing the suffix Cogni-

208 TREMBLAY VACHON AND JONES

tive streaming in immediate serial recall Journal of ExperimentalPsychology Learning Memory amp Cognition 28 12-28

Pashler H (1987) Targetndashdistractor discriminability in visual searchPerception amp Psychophysics 41 285-292

Potter M C Chun M M Banks B S amp Muckenhoupt M(1998) Two attentional deficits in serial target search The visual at-tentional blink and an amodal task-switch deficit Journal of Exper-imental Psychology Learning Memory amp Cognition 24 979-992

Raymond J E Shapiro K L amp Arnell K M (1992) Temporarysuppression of visual processing in an RSVP task An attentionalblink Journal of Experimental Psychology Human Perception ampPerformance 18 849-860

Raymond J E Shapiro K L amp Arnell K M (1995) Similaritydetermines the attentional blink Journal of Experimental Psychol-ogy Human Perception amp Performance 21 653-662

Rogers W L amp Bregman A S (1998) Cumulation of the tendencyto segregate auditory streams Resetting by changes in location andloudness Perception amp Psychophysics 60 1216-1227

Ross N E amp Jolicœur P (1999) Attentional blink for color Jour-nal of Experimental Psychology Human Perception amp Performance25 1483-1494

Ruthruff E amp Pashler H E (2001) Perceptual and central inter-ference in dual-task performance In K Shapiro (Ed) The limits ofattention Temporal constraints in human information processing(pp 100-123) Oxford Oxford University Press

Seiffert A E amp Di Lollo V (1997) Low-level masking in the at-tentional blink Journal of Experimental Psychology Human Per-ception amp Performance 23 1061-1073

Shapiro K L Raymond J E amp Arnell K M (1994) Attention tovisual pattern information produces the attentional blink in rapid se-rial visual presentation Journal of Experimental Psychology HumanPerception amp Performance 20 357-371

Shih S-I (2000) Recall of two visual targets embedded in RSVPstreams of distractors depends on their temporal and spatial relation-ship Perception amp Psychophysics 62 1348-1355

Soto-Faraco S amp Spence C J (2002) Modality-specific auditoryand visual temporal processing deficits Quarterly Journal of Exper-imental Psychology 55A 23-40

Spence C J amp Driver J (1994) Covert spatial orienting in auditionExogenous and endogenous mechanisms Journal of ExperimentalPsychology Human Perception amp Performance 20 555-574

Spence C [J] Ranson J amp Driver J (2000) Cross-modal selec-tive attention On the difficulty of ignoring sounds at the locus of vi-sual attention Perception amp Psychophysics 62 410-424

Treisman A amp Gelade G (1980) A feature integration theory of at-tention Cognitive Psychology 12 97-136

Vachon F amp Tremblay S (in press) Auditory attentional blinkMasking the second target is necessary delayed masking is suffi-cient Canadian Journal of Experimental Psychology

Visser T A W Bischof W F amp Di Lollo V (1999) Attentionalswitching in spatial and non-spatial domains Evidence from the at-tentional blink Psychological Bulletin 125 458-469

Woods D L Alain C Diaz R Rhodes D amp Ogawa K H(2001) Location and frequency cues in auditory selection attentionJournal of Experimental Psychology Human Perception amp Perfor-mance 27 65-74

NOTES

1 The notation used here and elsewhere in this article indicates therelation of the target (T) and the probe (P) to the serial position of theirmask in the sequence For example the notation P1 signifies thatthere is an item acting as a mask immediately following the probe

2 In the AB literature dual-stream presentation does not necessarilyrefer to the percept of two separate streams since it would prejudge thenature of the perceptual organization Also given the short duration ofRAP streams may not have enough time to build up fully (see Bregman1978 Rogers amp Bregman 1998)

3 Here and elsewhere in this article we refer to such terms as pitchswitching or switch pitch to indicate that attention has to be reallocatedfrom the frequency region of the target to that of the probe

4 The bottleneck model proposed by Jolicœur and DellrsquoAcqua (1998see also Jolicœur DellrsquoAcqua amp Crebolder 2001) to account for ABphenomena assumes limited capacity and serial processing but the au-thors do not exclude parallel processing as a possible explanation of theeffects

(Manuscript received October 7 2003revision accepted for publication April 27 2004)