12. When does between-sequence phonological similarity promote irrelevant sound disruption?.txt

Journal of Experimental Psychology:
Learning, Memory, and Cognition
2008, Vol. 34, No. 1, 243–248

Copyright 2008 by the American Psychological Association
0278-7393/08/$12.00 DOI: 10.1037/0278-7393.34.1.243

When Does Between-Sequence Phonological Similarity Promote Irrelevant
Sound Disruption?
John E. Marsh, François Vachon, and Dylan M. Jones

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Cardiff University
Typically, the phonological similarity between to-be-recalled items and TBI auditory stimuli has no
impact if recall in serial order is required. However, in the present study, the authors have shown that the
free recall, but not serial recall, of lists of phonologically related to-be-remembered items was disrupted
by an irrelevant sound stream (end rhymes) sharing similar phonological content. These findings can be
explained by the notion that between-sequence phonological similarity effects emerge when categorycueing processes become an important determinant for recall, such as when shared category information
can be used as a retrieval aid to cue list items or plausible list candidates. In this case, the presence of
categorically similar irrelevant items impairs the retrieval of list items and leads to intrusion error.
Implications of these results for theories of auditory distraction are discussed.
Keywords: auditory distraction, phonological similarity, selective attention, free recall, serial recall

At an empirical level, research on auditory distraction has been
strongly preoccupied by whether similarity of content between
rehearsal of visually presented TBR items and TBI auditory stimuli produces, or amplifies, the disruption in serial recall. This has
been fueled for the most part by the working memory model
(Baddeley, 1986), a theoretical perspective that suggests that such
similarity of content is important (Salamé & Baddeley, 1982; cf.
Larsen, Baddeley, & Andrade, 2000). A host of empirical findings
have now ruled out this interpretation of the irrelevant sound
effect: Similarity of phonological content (either defined by identity, rhyme, or the vowel sounds) between TBR and TBI items
does not increase disruption to serial recall (Buchner, Irmen, &
Erdfelder, 1996; Jones & Macken, 1995; Larsen et al., 2000;
LeCompte & Shaibe, 1997; but see Hughes & Jones, 2005).
Indeed, it is not a necessary prerequisite for irrelevant sound to
contain phonological content to produce an irrelevant sound effect;
tones can produce disruption (Jones & Macken, 1993). In light of
these findings, contemporary accounts, including computational
models, of the irrelevant sound effect dismiss between-sequence
phonological similarity as a disruptive agent (e.g., Jones, 1993;
Neath, 2000).
Like its phonological content, the semantic content of an irrelevant sound does not produce or exacerbate the irrelevant sound
effect in serial recall: Irrelevant speech in a monolingual participant’s native tongue produces no more disruption than when it is
presented in a language foreign to that participant (Jones, Miles, &
Page, 1990). Recently, however, evidence has emerged that
between-sequence semantic similarity—in tasks other than serial
recall— can produce disruption (Beaman, 2004; Neely & LeCompte, 1999). For example, a “semantic irrelevant sound effect”
has been shown in a category-exemplar recall task in which a list
of 10 semantically rich items (nouns) taken from a single semantic
category is presented for free recall. For this task, betweensequence similarity impairs performance (Marsh, Hughes, &
Jones, in press): The free recall of category-exemplars (e.g., avocado for “Fruits”) is disrupted (as reflected in reduced recall) more

There is ample evidence that cognitive tasks involving serial
recall are disrupted by the mere presence of concurrent to-beignored (TBI) sounds that are irrelevant to the focal task (Jones,
1999). This impairment is known as the irrelevant sound effect
(Beaman & Jones, 1997). The degree of disruption of serial recall
by irrelevant sound is related to the amount of acoustic change
between each successive irrelevant item (Jones, Madden, & Miles,
1992). Nonacoustic factors are shown to play a rather limited role
in the irrelevant sound effect (e.g., Neely & LeCompte, 1999). For
instance, the phonological similarity between the to-beremembered (TBR) and the TBI items does not impair serial recall
(e.g., Jones & Macken, 1995). However, these previous studies
that have focused on serial recall have not considered whether
between-sequence phonological similarity effects could be found
with other tasks for which participants can engage a strategy that
may be more susceptible to disruption by the phonological properties of irrelevant information. Here, for the first time, we compare the effect of between-sequence phonological similarity on
free and serial recall: We replicate the null effect on serial recall,
but we also demonstrate a pronounced disruptive effect of
between-sequence phonological similarity on the free recall of lists
of end-rhyming words.

John E. Marsh, François Vachon, and Dylan M. Jones, School of
Psychology, Cardiff University, Cardiff, United Kingdom.
Dylan M. Jones is also an adjunct professor at the Department of
Psychology, University of Western Australia, Perth, Western Australia,
Australia.
This research was supported by United Kingdom’s Economic and Social
Research Council Grant RES-000-22-1526 awarded to Dylan M. Jones,
Robert W. Hughes, and William J. Macken and by a postdoctoral fellowship from the Fonds québécois de la recherche sur la nature et les technologies awarded to François Vachon. We thank Robert Hughes for a
critical reading of a draft of this article.
Correspondence concerning this article should addressed to John E.
Marsh, School of Psychology, Cardiff University, P.O. Box 901, Cardiff
CF10 3AT, United Kingdom. E-mail: MarshJE@cardiff.ac.uk
243

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

244

BRIEF REPORTS

by semantically related irrelevant category items (that are not
included in the TBR list; e.g., apple) than by categorically unrelated, irrelevant items (e.g., hammer for “Tools”). The presence of
related irrelevant category items also results in more intrusions
relative to a quiet or an “unrelated” condition.
Previous studies that have focused on irrelevant sound effects in
serial and free recall have failed to consider whether differences
between the type of disruption (acoustic or semantic) produced by
irrelevant sound on free recall and serial recall arise from the
process/strategy differences underpinning the two tasks (see Marsh
et al., in press). Contemporary accounts suggest that serial recall is
driven primarily by temporal (order) cues, whereas free recall is
based more on semantic cues (e.g., Howard & Kahana, 2002;
Klein, Addis, & Kahana, 2005). The difference between the findings with serial recall and free recall can thus be explained by
supposing that the acoustic (but not semantic) properties of sound
disrupt the processing of, or representation of, temporal order cues
but have no effect on processing semantic cues (Jones, 1999).
Following the same reasoning for a scenario in which semantic
cueing is used (e.g., for free recall of semantically rich lists), it
would appear that the semantic (but not acoustic) properties of
sound disrupt processing of semantic cues (Marsh et al., in press).
To elaborate, the shared attributes between items of a semantically
categorized list can act as an important facilitative aid: Selfinitiated semantic-category cueing can be used to generate list
items or potential list candidates (Anderson & Bower, 1972; Gronlund & Shiffrin, 1986). However, such category cueing applied to
free recall carries a cost in that it can lead to the generation of
related (but irrelevant) auditory items, thus giving rise to their false
inclusion as responses (Marsh et al., in press).
Importantly, the use of category knowledge and category-cueing
processes to organize recall and generate list candidates is not
exclusive to semantic categories, and the literature is replete with
examples indicating functional similarities between semanticcategory and nonsemantic-category processes. Although randomly
arranged in a study list, participants cluster phonologically similar
(i.e., rhyming) words together at output (e.g., thread, bread, lead;
Long & Allen, 1973). Participants often generate, then falsely
remember, nonpresented critical items to a list of phonological
associates (e.g., cat; fat, that, cab; Sommers & Lewis, 1999). In a
category-exemplar generation task, fewer words are generated and
written corresponding to a graphemic rule (e.g., van. . .; vanish,
vanity, vanguard) when participants are presented with irrelevant
auditory words from the same, as compared with a different,
graphemic category (e.g., dis. . .; Watkins & Allender, 1987).
Collectively, these findings suggest that the similarity in rhyme
category, like semantic category, can be used to cue recall (see
Gupta, Lipinski, & Aktunc, 2005) and that between-sequence
phonological similarity between streams of TBR and TBI items
could produce disruption with the proviso that the recall task is
dominated by processes that utilize long-term memory, such as
category cueing.
In sum, researchers of prior studies that have failed to find any
effect of between-sequence phonological similarity have done so
by way of using serial recall tasks. This leaves open the possibility
that between-sequence phonological similarity effects could be
found in the context of alternative mnemonic tasks. The current
study overcomes the restrictions of previous studies by comparing
the effect of between-sequence phonological similarity in the

context of both serial recall and free recall tasks. The novel idea
maintained as a working hypothesis for the current study is that the
semantic irrelevant sound effect (e.g., Marsh et al., in press) is a
generic “categorical” effect and that a between-sequence phonological similarity effect could be found for phonological categories
if the use of phonological information is promoted by requiring
free recall. Therefore, the main purpose of the present study was to
investigate, for the first time, whether a between-sequence similarity effect could occur for the free but not serial recall of
categorized material that is defined not semantically but phonologically. For this purpose, rhyme categories were used: The
auditory TBI items were end rhymes (e.g., side, ride, guide. . .)
that were either phonologically similar to the visual TBR items
(e.g., glide, fried, vied. . .) or phonologically dissimilar to the TBR
items (e.g., frock, dock, crock. . .). The effect of between-sequence
phonological similarity was examined through analysis of recall
performance and extra-list intrusions.

Method
Participants
Forty-eight students from Cardiff University participated in a
mixed-factor design in return for course credit. All participants
reported normal hearing and normal or corrected-to-normal vision
and were native English speakers.

Apparatus and Materials
The rhyme-generating program Rhymer (WriteExpress; http://
www.rhymer.com) was used to generate words by entering 30
words each with a different but common end rhyme (e.g., “ide” as
in “side”). The monosyllabic end rhymes generated for each word
were then chosen and scored for frequency with the Thorndike–
Lorge word count (Thorndike & Lorge, 1944) as were the originally entered words. For each end rhyme, the resultant six highest
frequency words were chosen as the TBI auditory items, whereas
the 10 lowest frequency words were chosen as the TBR visual
items.
This method of constructing and selecting materials differed
from previous studies (e.g., Beaman, 2004) for the following
reasons. Items chosen in these prior studies were derived from
category norms: Although the common attribute of an end rhyme
means that end-rhyme words form categories, we were unaware of
any “end-rhyme category norms.” This is unfortunate to the extent
that the dominance relations between the items cannot be determined: We cannot be sure that the highest frequency items are the
first items participants will generate when presented with the end
rhyme as a cue. Logically however, there should be a correlation
between frequency of use in the English language and dominance
level, and because of this, the highest frequency items were chosen
as the irrelevant items. Limits imposed by restricted category sizes
forced omissions of items not scored for frequency in the
Thorndike–Lorge word count, and the removal of homophones
resulted in the use of six words as TBI items and 10 words as the
TBR items in each list.
The items within the TBR and TBI sequences were presented in
a randomized order that was the same for each participant. We
recorded the TBI items in a male voice at an even pitch at a

BRIEF REPORTS

sampling rate of 44.1 kHz in 16-bit resolution using Sound Forge
5.0 (Sonic Inc., Madison, WI). Each TBI item was then digitally
edited to a duration of 500 ms. We presented the TBI items
stereophonically over headphones at approximately 65–70 db(A)
using SuperLab Pro software (Cedrus Corp., San Pedro, CA).

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Design
A mixed design was used with one between-participants factor
and one within-participant factor. The between-participants factor
was “Instruction Type,” and participants were assigned to one of
two levels: free recall and serial recall. The within-participant
factor was “Sound Condition,” of which there were three levels:
(a) irrelevant items phonologically dissimilar to the TBR items, (b)
irrelevant items phonologically similar to the TBR items, and (c)
a quiet control condition.
The 30 TBR lists and 30 irrelevant sound sequences were
divided into two 15-list sets. For each instruction type, half the
participants were randomly assigned to one of the two 15-list sets.
Each 15-list set was further subdivided into five blocks of three
lists. Within each three-list block, the TBR lists were assigned
randomly to one of the three sound conditions. The order of the
three sound conditions within each three-trial block was counterbalanced across participants such that an equal number of participants were assigned to each of the six possible order permutations.
When a TBR list was assigned to the phonologically similar
condition, the sound sequence corresponding to the category represented by that list was presented. When a TBR list was assigned
to the phonologically dissimilar condition, the sound sequence was
randomly chosen from the 15 categories not represented by the
TBR list.

Procedure
Participants were seated at a viewing distance of approximately
60 cm from a PC monitor on which TBR items were displayed in
a central position. Lists of TBR items appeared in lowercase black
72-point Times New Roman font one word at a time against a
white background. Each word—whether TBR or TBI—was presented for 500 ms with no interstimulus interval. The end of the list
was notified by the appearance of a red “WAIT” sign that was
displayed for 9 s before a green “RECALL” sign was shown to
indicate the end of the retention period and start of the retrieval
period. Each of the TBI items was presented three times in the
retention period.
Participants were seated in individual cubicles. They were informed that they would be presented with fifteen ten-word lists and
that each list would be presented one word at a time on the
computer monitor from which they were asked to memorize as
many words as possible and write the words they remembered
down on their recall sheets when a “RECALL” cue appeared on
the screen.
On presentation of the RECALL cue, participants given free
recall instructions were instructed to try and remember as many of
the words in any order, whereas participants given serial recall
instructions were instructed to remember the words according to
their original order of presentation. Participants from the serial
recall group were also told that they could leave gaps in their recall
protocols if necessary. Participants in this group were also in-

245

structed to guess the original position if they had an item available
to them for recall but could not remember its position. Recall
sheets contained 15 columns of 10 rows each: Participants given
serial recall instructions were provided with specially prepared
recall sheets with serial positions marked on them; the free-recallinstructed group had the same recall sheets but without the serial
positions marked.
Participants were informed that they would have 30 s to retrieve
as much as they could of the list and that after this time a tone
would sound to signal the beginning of the next list (some 5 s
later). Participants were instructed to ignore any sound that they
heard through the headphones and were told that they would not be
tested on its content at any point in the experiment. On completion
of the experiment participants were instructed to return to the
beginning of their recall sheets and mark next to an item recalled
a confidence rating as to whether they thought that item had been
visually presented (1 ⫽ guess, 10 ⫽ absolutely certain; see Smith,
Ward, Tindell, Sifonis, & Wilkenfeld, 2000).

Results
Correct Responses
The raw serial recall data were scored according to a strict serial
recall criterion— recalled items were only scored as correct if they
corresponded to their presentation position—whereas a free recall
criterion was employed to score the raw free recall data: Recalled
items were scored as correct regardless of serial position. On
inspection of Figure 1, there is an irrelevant sound effect in both
free and serial recall, whereby recall is impaired with the presentation of TBI items. However, only free recall showed a betweensequence phonological similarity effect, that is, lower performance
in the phonologically similar compared with the phonologically
dissimilar condition. A 2 ⫻ 3 mixed analysis of variance
(ANOVA) was carried out on the overall proportion of correct
recall, with instruction type (free vs. serial recall) as a betweenparticipants factor and sound condition (quiet, phonologically dissimilar, and phonologically similar) as a within-participant factor.
The analysis confirmed a significant main effect of sound condition, F(2, 92) ⫽ 32.32, p ⬍ .001, d ⫽ 1.68, but no main effect of
instruction type, F(1, 46) ⬍ 1.00, p ⫽ .66, d ⫽ 0.13. Of particular
importance for the purpose of the present study is the significant
interaction between instruction type and sound condition, F(2,
92) ⫽ 3.13, p ⬍ .05, d ⫽ 0.52. This significant interaction arose
because phonologically similar irrelevant items disrupted recall
compared with phonologically dissimilar items in free recall,
t(23) ⫽ 3.94, p ⬍ .001, d ⫽ 1.64, but not in serial recall, t(23) ⬍
1.00, p ⫽ .53, d ⫽ 0.30.

Intrusions
For both free and serial recall, two types of intrusion were
distinguished. A related-item intrusion was a response that
matched one of the six high-frequency items presented as irrelevant items in the phonologically similar condition. Such responses
were scored as similar-item intrusions even when those exemplars
had not been presented on a given trial (i.e., in the quiet and
phonologically dissimilar condition), thus providing a baseline
against which to assess such intrusions in the phonologically

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

246

BRIEF REPORTS

Figure 1. Proportion of correct recall in free and serial recall as a function of irrelevant sound condition. Error
bars represent 95% within-participant confidence intervals calculated separately for each instruction type.

similar condition. Other-item intrusions refer to items that were
not presented at all during the experiment. Figure 2 shows for both
recall tasks the mean number of each type of intrusion (pooled
across all trials) for each sound condition. Generally, related-item
intrusions were as common as other-item intrusions in every sound
condition, except in the phonologically similar condition of the
free recall task wherein related-item intrusions were compellingly
more prevalent.
A 2 (Instruction Type) ⫻ 2 (Intrusion Type) ⫻ 3 (Sound
Condition) mixed ANOVA performed on the mean number of
intrusions revealed significant main effects of intrusion type, F(1,
46) ⫽ 4.36, p ⬍ .05, d ⫽ 0.62, and sound condition, F(2, 92) ⫽
9.37, p ⬍ .001, d ⫽ 0.90, whereas the main effect of instruction
type was not significant, F(1, 46) ⫽ 1.98, p ⫽ .17, d ⫽ 0.42. All
two-way interactions were significant: Instruction Type ⫻ Intrusion Type, F(1, 46) ⫽ 5.42, p ⬍ .03, d ⫽ 0.69; Instruction Type ⫻
Sound Condition, F(2, 92) ⫽ 5.20, p ⬍ .01, d ⫽ 0.67; and
Intrusion Type ⫻ Sound Condition, F(2, 92) ⫽ 6.90, p ⬍ .01, d ⫽

0.78. The three-way interaction approached significance, F(2,
92) ⫽ 2.89, p ⫽ .06, d ⫽ 0.50.
Intrusions were also submitted to a 2 (Intrusion Type) ⫻ 3
(Sound Condition) repeated measures ANOVA separately for each
instruction type. In serial recall, none of the effects were significant (all Fs ⬍ 1.00; ds ⬍ 0.30). In free recall, the main effects of
intrusion type, F(1, 23) ⫽ 20.61, p ⬍ .001, d ⫽ 1.89, and sound
condition, F(2, 46) ⫽ 13.31, p ⬍ .001, d ⫽ 1.52, were significant.
Of particular interest is the significant interaction between the two
factors, F(2, 46) ⫽ 8.41, p ⫽ .001, d ⫽ 1.21. Further analysis of
this interaction (with an adjusted alpha level of .01) showed that
related-item intrusions were more common than other-item intrusions in the phonologically similar condition, t(23) ⫽ 4.16, p ⬍
.001, d ⫽ 1.73, but not in the quiet condition, t(23) ⫽ 2.45, p ⬍
.03, d ⫽ 1.02, or the phonologically dissimilar condition, t(23) ⫽
1.03, p ⫽ .313, d ⫽ 0.43. Simple effect tests also revealed that
related-item intrusions were more common in the phonologically
similar condition as compared with the quiet condition, t(23) ⫽

Figure 2. Mean total number of related- and other-item intrusions in free and serial recall as a function of
irrelevant sound condition. Error bars represent 95% within-participant confidence intervals calculated separately for each instruction type.

BRIEF REPORTS

4.37, p ⬍ .001, d ⫽ 1.82, and the phonologically dissimilar
condition, t(23) ⫽ 4.84, p ⬍ .001, d ⫽ 2.02. Analysis of the
confidence ratings suggested modality confusion: Participants in
the phonologically similar condition of the free recall group tended
to respond that the related-item intrusions they made were visually
presented when in fact they were presented auditorily.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Discussion
In the present study, we established for the first time a betweensequence phonological similarity effect using free recall as the
focal task and replicated the finding that such an effect does not
occur in the context of serial recall (cf. Jones & Macken, 1995).
This irrelevant sound phonological similarity effect can be captured when strategies/processes, such as category cueing (that are
specific to the free recall of categorized material), are used. The
results of this experiment conceptually replicate those reported by
Marsh et al. (in press) but with the use of categorical information
that is phonologically rather than semantically defined. This indicates that there is no difference in the action of categorically
related irrelevant items whether they are semantically or phonologically determined. In this regard, these effects are better termed
categorical irrelevant sound effects than semantic irrelevant sound
effects.
Before discussing the theoretical implications of the present
findings, it should be noted that the current tasks diverged from
classic short-term memory tasks in at least three ways. First, the
present experiment used list lengths (10 items) that are longer than
typically adopted for classic short-term memory research (five to
eight items). Second, the current tasks involved a retention interval
between study and test phases of the task. Delaying recall—such as
with a retention interval—and using supraspan TBR lists of phonologically similar items (as compared with phonologically dissimilar items) can promote the use of phonological similarity as a
facilitative means for recall (Farrell, 2006). Manipulation of factors—such as list length, the presence or absence of a retention
interval, and within-stream phonological similarity—would deserve future scrutiny to allow a more direct comparison with
previous studies. A third difference between the current research
and previous studies that have used irrelevant sound with serial
recall (e.g., Jones & Macken, 1995) is that prior studies have
tended to use a closed set of either phonologically similar TBR
items, TBI items, or both. Because a closed set of rhyming items
decreases the purported use of category cueing in serial recall (e.g.,
Gupta et al., 2005), it was possible that an effect of betweensequence phonological similarity would emerge with serial recall
provided that an open set of rhyming items was used. However, in
the current study, we used an open set and still failed to obtain a
between-sequence similarity effect with serial recall.

Theoretical Implications
There is theoretical controversy concerning the most adequate
account of the irrelevant sound effect (Cowan, 1995; Hughes &
Jones, 2005; Neath, 2000; Norris, Baddeley, & Page, 2004) and
associated phenomena of auditory distraction (Buchner, Rothermund, Wentura, & Mehl, 2004; Norris et al., 2004). However, the
results reported here do appear to contradict some specific accounts of the irrelevant sound effect with serial recall that are

247

heavily reliant on the concept of interference-by-content. For example, although in its initial formulation the working memory
account (Salamé & Baddeley, 1982; cf. Larsen et al., 2000) predicted an additional disruptive effect attributable to betweensequence phonological similarity of content (albeit erroneously;
see Jones & Macken, 1995), proponents of this account have
argued that phonological encoding is abandoned in favor of a
semantic encoding strategy when list length exceeds span (five to
six items; Salamé & Baddeley, 1986). Thus, a between-sequence
phonological similarity effect is not predicted by the working
memory account with lists of the length used in the current study.
Similarly, the feature model (Neath, 2000) does not predict any
form of between-sequence similarity effect due to the specifics of
the feature adoption process that drives the irrelevant sound effect
with serial recall.
The interference-by-process account (e.g., Hughes & Jones, 2005),
although in need of more precise formulation, is the only approach
that predicts that the degree and type of disruption produced by
irrelevant sound are a joint product of the nature of the irrelevant
sound and the prevailing cognitive processes. The account correctly
predicts the following: Tasks that require order processing are much
more vulnerable to disruption via the acoustic-based order processing
of the irrelevant sound than tasks that do not (Beaman & Jones, 1997;
Hughes, Vachon, & Jones, in press); tasks that require memory for
serial order are not impaired by the semantic or phonological properties of irrelevant sound (Marsh et al., in press; the current study);
and tasks that require memory for semantic- or phonological-category
information, regardless of its presentation order, are disrupted by,
respectively, similar semantic or phonological information conveyed
by irrelevant sound but not its acoustic attributes (Marsh et al., in
press; the current study). One candidate process for explaining the
disruption within this interference-by-process framework is source
monitoring: TBI items that are categorically related to TBR items may
be recalled (resulting in the compelling related-item intrusion rate in
the phonologically similar condition) in the free recall condition
because of a failure to attribute the source of the activation of the
auditory item to the correct external modality (visual) in which it was
experienced (see Marsh et al., in press). Contemporary models of free
recall (e.g., Howard & Kahana, 2002) could explain the recall of TBI
items by assuming that they become part of the list temporal context
and are accordingly transmitted to the participant’s output protocols.

Concluding Remarks
In sum, the phonological similarity effect of irrelevant sound
appears to be driven by the nature of the primary task: Generally,
it is not found if the primary task is serial recall but is indeed found
when category-cueing processes are likely to be engaged. Whenever possible, participants will use preexisting long-term memory
structures (associations/conceptual knowledge) and natural language habits to learn, mediate, and facilitate recall of TBR material
(Bousfield, 1953; Ebbinghaus, 1964). In serial recall, the role of
long-term knowledge in task-performance is constrained because
the TBR sequences are, by design, stripped of any grammatical–
semantic structure that could support serial order (this is precisely
what makes serial recall difficult; see Macken & Jones, 2003).
With categorized TBR information and free recall, generic processes involving long-term memory (e.g., category cueing) can
operate, but these can become vulnerable to disruption from irrel-

BRIEF REPORTS

248

evant sound that activates competing long-term memory representations. It is entirely plausible that other ways of organizing material and generating retrieval candidates (such as by first letter;
Hicks & Young, 1973) will be disrupted by TBI items (such as
those starting with the same initial letter) but that these will
represent the same generic effect. To conclude, any complete
account of the effects of auditory distraction— be it theoretically or
computationally based—must take into consideration the interdependency of the nature of the focal task and the nature of the
potentially distracting irrelevant sound.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

References
Anderson, J. R., & Bower, G. H. (1972). Recognition and retrieval processes in free recall. Psychological Review, 79, 97–123.
Baddeley, A. D. (1986). Working memory. Oxford, England: Oxford University Press.
Beaman, C. P. (2004). The irrelevant sound phenomenon revisited: What
role for working memory capacity? Journal of Experimental Psychology: Learning, Memory, and Cognition, 30, 1106 –1118.
Beaman, C. P., & Jones, D. M. (1997). Role of serial order in the irrelevant
speech effect: Tests of the changing-state hypothesis. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23, 459 – 471.
Bousfield, W. A. (1953). The occurrence of clustering in the recall of randomly arranged associates. Journal of General Psychology, 49, 229 –240.
Buchner, A., Irmen, L., & Erdfelder, E. (1996). On the irrelevance of
semantic information for the “irrelevant” speech effect. Quarterly Journal of Experimental Psychology: Human Experimental Psychology,
49(A), 765–779.
Buchner, A., Rothermund, K., Wentura, D., & Mehl, B. (2004). Valence of
distractor words increases the effects of irrelevant speech on serial recall.
Memory & Cognition, 32, 722–731.
Cowan, N. (1995). Attention and memory: An integrated framework.
Oxford, England: Oxford University Press.
Ebbinghaus, H. (1964). Memory: A contribution to experimental psychology. New York: Dover.
Farrell, S. (2006). Mixed-list phonological similarity effects in delayed
serial recall. Journal of Memory and Language, 55, 587– 600.
Gronlund, S. D., & Shiffrin, R. M. (1986). Retrieval strategies in recall of
natural categories and categorized lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 12, 550 –561.
Gupta, P., Lipinski, J., & Aktunc, E. (2005). Re-examining the phonological similarity effect in immediate serial recall: The roles of type of
similarity, category cueing, and item recall. Memory & Cognition, 33,
1001–1016.
Hicks, R. E., & Young, R. K. (1973). Input and output variables in recall
of categorized and uncategorized lists. Journal of General Psychology,
89, 109 –114.
Howard, M. W., & Kahana, M. J. (2002). When does semantic similarity
help episodic retrieval? Journal of Memory and Language, 46, 85–98.
Hughes, R. W., & Jones, D. M. (2005). The impact of order incongruence
between a task-irrelevant auditory sequence and a task-relevant visual
sequence. Journal of Experimental Psychology: Human Perception and
Performance, 31, 316 –327.
Hughes, R. W., Vachon, F., & Jones, D. M. (in press). Disruption of
short-term memory by changing and deviant sounds: Support for a
duplex-mechanism account of auditory distraction. Journal of Experimental Psychology: Learning, Memory, and Cognition.
Jones, D. M. (1993). Objects, streams, and threads of auditory attention. In
A. D. Baddeley & L. Weiskrantz (Eds.), Attention: Selection, awareness,
and control (pp. 87–104). Oxford, England: Clarendon Press.
Jones, D. M. (1999). The 1997 BPS Broadbent Lecture: The cognitive

psychology of auditory distraction. British Journal of Psychology, 90,
167–187.
Jones, D. M., & Macken, W. J. (1993). Irrelevant tones produce an
irrelevant speech effect: Implications for coding in phonological memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 1008 –1018.
Jones, D. M., & Macken, W. J. (1995). Phonological similarity in the
irrelevant speech effect: Within- or between-stream similarity. Journal
of Experimental Psychology: Learning, Memory, and Cognition, 21,
103–115.
Jones, D. M., Madden, C., & Miles, C. (1992). Privileged access by
irrelevant speech to short-term memory: The role of changing state.
Quarterly Journal of Experimental Psychology: Human Experimental
Psychology, 44(A), 645– 669.
Jones, D. M., Miles, C., & Page, J. (1990). Disruption of proofreading by
irrelevant speech: Effects of attention, arousal, or memory? Applied
Cognitive Psychology, 4, 89 –108.
Klein, K. A., Addis, K. M., & Kahana, M. J. (2005). A comparative
analysis of serial and free recall. Memory & Cognition, 33, 833– 839.
Larsen, J. D., Baddeley, A. D., & Andrade, J. (2000). Phonological
similarity and irrelevant speech: Implications for models of short-term
verbal memory. Memory, 8, 145–157.
LeCompte, D. C., & Shaibe, D. M. (1997). On the irrelevance of phonology to the irrelevant speech effect. Quarterly Journal of Experimental
Psychology: Human Experimental Psychology, 50(A), 100 –118.
Long, D., & Allen, G. A. (1973). Relative effects of acoustic and semantic
relatedness on clustering in free-recall. Bulletin of the Psychonomic
Society, 1, 316 –318.
Macken, W. J., & Jones, D. M. (2003). Reification of phonological storage.
Quarterly Journal of Experimental Psychology: Human Experimental
Psychology, 56(A), 1279 –1288.
Marsh, J. E., Hughes, R. W., & Jones, D. M. (in press). Semantic auditory
distraction: A process-based approach. Journal of Memory and Language.
Neath, I. (2000). Modeling the effects of irrelevant speech on memory.
Psychonomic Bulletin & Review, 7, 403– 423.
Neely, C. B., & LeCompte, D. C. (1999). The importance of semantic
similarity to the irrelevant speech effect. Memory & Cognition, 27,
37– 44.
Norris, D., Baddeley, A. D., & Page, M. P. A. (2004). Retroactive effects
of irrelevant speech on serial recall from short-term memory. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 30, 1093–
1105.
Salamé, P., & Baddeley, A. D. (1982). Disruption of short-term memory by
unattended speech: Implications for the structure of working memory.
Journal of Verbal Learning and Verbal Behavior, 21, 150 –164.
Salamé, P., & Baddeley, A. D. (1986). Phonological factors in STM:
Similarity and the unattended speech effect. Bulletin of the Psychonomic
Society, 24, 263–265.
Smith, S. M., Ward, T. B., Tindell, D. R., Sifonis, C. M., & Wilkenfeld,
M. J. (2000). Category structure and created memories. Memory &
Cognition, 28, 386 –395.
Sommers, M. S., & Lewis, B. P. (1999). Who really lives next door:
Creating false memories with phonological neighbours. Journal of Memory and Language, 40, 83–108.
Thorndike, E. L., & Lorge, I. (1944). The teacher’s word book of 30,000
words. New York: Columbia University, Teachers College.
Watkins, M. J., & Allender, L. E. (1987). Inhibiting word generation with
word presentations. Journal of Experimental Psychology: Learning,
Memory, and Cognition, 13, 564 –568.

Received March 5, 2007
Revision received August 15, 2007
Accepted August 16, 2007 䡲