Linguistic Processing of Accented Speech Across the Lifespan

Abstract

In most of the world, people have regular exposure to multiple accents. Therefore, learning to quickly process accented speech is a prerequisite to successful communication. In this paper, we examine work on the perception of accented speech across the lifespan, from early infancy to late adulthood. Unfamiliar accents initially impair linguistic processing by infants, children, younger adults, and older adults, but listeners of all ages come to adapt to accented speech. Emergent research also goes beyond these perceptual abilities, by assessing links with production and the relative contributions of linguistic knowledge and general cognitive skills. We conclude by underlining points of convergence across ages, and the gaps left to face in future work.

Full text

REVIEW ARTICLE
published: 08 November 2012
doi: 10.3389/fpsyg.2012.00479
Linguistic processing of accented speech across the
lifespan
Alejandrina Cristia1*†, Amanda Seidl 2, Charlotte Vaughn3, Rachel Schmale4, Ann Bradlow 3and
Caroline Floccia5
1Max Planck Institute for Psycholinguistics, Nijmegen, Netherlands
2Purdue University,West Lafayette, IN, USA
3Northwestern University, Evanston, IL, USA
4North Park University, Chicago, IL, USA
5Plymouth University, Plymouth, UK
Edited by:
Holger Mitterer, Max Planck Institute
for Psycholinguistics, Netherlands
Reviewed by:
Marieke Van Heugten, University of
Toronto, Canada
Katherine White, University of
Waterloo, Canada
*Correspondence:
Alejandrina Cristia, Neurobiology of
Language Department, Max Planck
Institute for Psycholinguistics, PO Box
310, 6500 AH Nijmegen, Netherlands.
e-mail: alecristia@gmail.com
†Present address:
Alejandrina Cristia, Neurobiology of
Language Department, Max Planck
Institute for Psycholinguistics,
Wundtlaan 1, 6525XD Nijmegen,
Netherlands.
In most of the world, people have regular exposure to multiple accents. Therefore, learning
to quickly process accented speech is a prerequisite to successful communication. In this
paper, we examine work on the perception of accented speech across the lifespan, from
early infancy to late adulthood. Unfamiliar accents initially impair linguistic processing by
infants, children, younger adults, and older adults, but listeners of all ages come to adapt
to accented speech. Emergent research also goes beyond these perceptual abilities, by
assessing links with production and the relative contributions of linguistic knowledge and
general cognitive skills. We conclude by underlining points of convergence across ages,
and the gaps left to face in future work.
Keywords: infancy, childhood, aging, accent adaptation, speech perception
INTRODUCTION
Infants, children, and adults may all experience different chal-
lenges in processing unfamiliar accents. Learning how to adapt to
unfamiliar accents is necessary for efficient language processing,
given the variety of accents that surrounds us. Nearly everywhere
in the world, a simple trip to the market will most likely put you
within earshot of dialectal or foreign accents. For instance, a report
of 26 countries by the Organization for Economic Cooperation
and Development (2007) estimated that about 9% of each coun-
try’s population was foreign and thus might speak a language not
spoken in their current country of residence. To take a more spe-
cific example, a census report in the USA documents that 20% of
respondents declared speaking a language other than English at
home, and half of that 20% estimated their own English speak-
ing abilities as below fluent (United States Census Bureau, 2008).
Moreover, these numbers underestimate the likelihood of encoun-
tering an accent different from one’s own, as they do not take into
account variation in within-language accents. In this article, we
review evidence bearing on how we perceive speech in the face of
accent variation, both as our linguistic system develops and after
we have become efficient language processors. To our knowledge,
this is the first review that aims to assemble findings on infant,
child, and adult accent perception. Examining accent perception
across the lifespan allows us to underline points of convergence
and divergence, as well as gaps that remain for future work.
Before we delve into this literature,we clarify a few terminolog-
ical points. First, we use the term linguistic variety as an umbrella
covering different (1) regional or sociolectal varieties of a single
language (i.e., within-language accents), (2) different languages,
and (3) non-native varieties of a language (i.e., foreign accents). In
addition, we define the term accent from the listener’s perspective,
as follows: a talker may be described as accented if his/her speech
diverges from that of the listener’s systematically at the supraseg-
mental and/or segmental level. Consequently, if the listener speaks
a “non-standard” regional variety, and the talker a standard vari-
ety, the latter would still be described as accented, because his/her
speech deviates from that of the listener. Notice that we extend
Wells’ (1982) definition of accent, as deviations along the pho-
netic, phonotactic, phonological, and lexical levels, in order to
encompass the suprasegmental level as well.
In the following four sections, we summarize current literature
on accent perception in young adults, infants, children, and older
adults. Looking throughout all age groups, we identified two cen-
tral themes of research evident in each and every age group. One
theme pertains to initial processing difficulties exhibited when
hearing accented speech; the other to the effects of exposure to the
accent. Each of the following four sections is devoted to research
on one age group, reviewing research on each of the two themes,
and ending with a summary and brief discussion of the particular
contributions of that age group to our understanding of accented
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 1

Cristia et al. Accent perception across the lifespan
speech perception. Research on young adults is presented first
because the majority of research on accent perception has been
carried out on young adults, typically college students. Further-
more, work with other populations usually use young adults as
a reference point. Thus, these data can be viewed as the bench-
mark against which researchers working with younger or older
populations will compare their findings.
ACCENT PERCEPTION IN EARLY ADULTHOOD
In view of the wealth of this work, we do not endeavor to be
comprehensive in the present section, but focus instead on lines
of research that allow comparisons with the literature from other
age groups summarized below. The interested reader is referred
to Floccia et al. (2006) and Samuel and Kraljic (2009) for more
detailed reviews focusing on adult perception of accented speech.
Here, we limit our summaries to a few representative studies
documenting both the initial processing costs and the adaptive
processes involved in recognizing the linguistic content of accented
speech.
INITIAL PROCESSING COSTS
The rich work on accented speech perception that has been car-
ried out on young adults shows that accented speech affects
both accuracy and speed of processing. Listeners are less accu-
rate in transcribing the speech of both foreign accented speakers
(Gass and Varonis, 1984) and within-language accented speakers
(Mason, 1946;Labov and Ash, 1997). Moreover, intelligibility of
both foreign accented speech (Rogers et al., 2004) and regional
accented speech (Clopper and Bradlow, 2008) can be affected by
background noise to a greater extent than speech spoken in the lis-
teners’ own accent. Accented speech is also processed more slowly.
This has been shown in a wide range of tasks, including assessing
whether a wordform is a real word or not (Floccia et al., 2006),
deciding whether a word heard matches one printed on a screen
(Munro and Derwing, 1995;Clarke and Garrett, 2004), making
semantic judgments (Adank and McQueen, 2007), and evaluating
whether a sentence is true or false (Adank et al., 2009).
In fact, some research suggests that delays when processing
speech in an accent that is not one’s own could actually indi-
cate that different mechanisms are recruited, or that they are
relied upon to a different extent when processing accented and
unaccented speech. These differences are sometimes evident when
processing is rendered difficult. For example, Bürki-Cohen et al.
(2001) tested native English listeners on a phoneme detection task,
either in isolation or paired with a secondary linguistic task (decid-
ing whether the item was a noun or a verb). The key question
was whether listeners would in fact recruit lexical information in
their judgments, in which case response times should be lower for
higher frequency words than for lower frequency words. For the
unaccented speech, listeners did not make use of lexical informa-
tion (response times did not vary between high and low-frequency
words), even when the secondary task was added. However, the
secondary task led listeners to rely on lexical information when
processing foreign accented speech.
Interestingly, several top-down factors have been shown to
modulate the processing cost involved in perceiving accented
speech, suggesting that, to a certain extent, a different processing
profile may not be due only to differences in the acoustic signal.
Indeed, the mere expectation that speakers will have an accent
may hinder listeners’ comprehension. For example, Rubin (1992)
found that the same general American “unaccented” speech was
understood less accurately when paired with a photograph of an
Asian face than when it was paired with a Caucasian face. Although
it remains to be explored to what extent such top-down fac-
tors impact lower-level processing, it is clear that the challenges
involved in processing accented speech go beyond the fact that lis-
teners’ usual mechanisms are suboptimal in the face of less familiar
speech patterns.
EFFECTS OF EXPOSURE
In order to have precise control on exposure and accent complex-
ity, a recent line of work has opted to train listeners on artificially
created, novel accents. These studies simulate what it is like to be
exposed to a new regional or foreign accent – or at least to one
feature of a new accent – and thus may help us better understand
how initial adaptation to real accents might occur. For example,
Maye et al. (2008) created an accent where all vowels were shifted
down in the vowel space (i.e., “wetch” became an acceptable pro-
nunciation of the word “witch”). After mere minutes of hearing
the story of the Wizard of Oz spoken in this “accent,” partici-
pants gave more“word” responses on a subsequent lexical decision
task to items that were plausible implementations of real words in
that accent. Most work shows that adult listeners adapt to novel
pronunciations by applying top-down, lexical knowledge (Norris
et al., 2003; see Mitterer and McQueen, 2009, for evidence that
orthographically presented forms may also facilitate adaptation);
but reports also exist for other sources of disambiguation, such as
visual input (Bertelson et al., 2003) and phonotactics (Cutler et al.,
2008).
There is considerable debate surrounding the format of
adaptation. Some argue that changes occur at a prelexical
level of interpretation, since the transfer to untrained lexi-
cal items is cost-less (e.g., Maye et al., 2008). Along a similar
line, Skoruppa and Peperkamp (2011) argue that such adap-
tation crucially recruits phonological knowledge, since it is
constrained by phonological simplicity. In their study, listen-
ers were able to learn an accent in which vowels harmonized
(i.e., vowels in adjacent syllables had to both be rounded – the
French word liqueur/likœ /was pronounced ∗luqueur/ lykœ /,
where the first vowel has become rounded) or disharmonized
(pudeur/pydœ /became ∗pudere/ pydε/, with the second vowel
becoming unrounded). However, listeners failed to encode an
accent that was unnaturally complex (where mid vowels harmo-
nized and high vowels disharmonized). Others argue that adapta-
tion involves altering lexical representations, for example storing
the unusual instantiations without abstracting away talker infor-
mation. Dahan et al. (2008) argue for this interpretation based on
effects in lexical competition. They measured lexical competition
by tracking eye movements to potential visual referents of mini-
mally differing words, such as“back”and “bag,” while one of these
words was presented auditorily. For half of the participants, the
talker in the stimuli raised/æ/to [ε] before/g/; for the other half,she
did not. When hearing the beginning of the unaltered word“back,”
participants in the latter (control) group sometimes looked at the
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 2

Cristia et al. Accent perception across the lifespan
picture of “bag,” showing a certain amount of lexical competition
between the two items. When faced with the same situation, the
group exposed to raised /æg/ words looked less to “bag” than the
control group did. In other words, the raised group found /bæk/
less ambiguous between the “bag”and “back” interpretations (i.e.,
if the talker meant to say“ bag,”the syllable should have started with
/bε/). Such artificial phonologies are well-suited to investigate the
question of how we encode accented speech patterns.
However, natural accents are much more complex than those
previously implemented through artificial phonologies; for exam-
ple, the former typically affect multiple phonological levels.
Research is still needed to assess the extent to which learning such
artificial phonologies resembles learning real accents. Thus, this
work cannot currently replace research investigating the poten-
tial benefits of short-term, laboratory-based exposure to a natural
accent, which provides valuable evidence. For example, Clarke and
Garrett (2004) showed that response times for foreign accented
speech in a cross-modal matching task (assessing whether a word
heard and a word printed on the screen are the same) can catch
up with the native accent response times after less than a minute
of exposure (see also Wingstedt and Schulman, 1987, for trans-
ference to production, and Kraljic et al., 2008a, for adaptation in
perception without transfer to production). However, how and
when exactly this adaptation can be brought about in the lab is
not entirely clear.
Bradlow and Bent (2008) provided important evidence con-
cerning when accent adaptation is more likely too ccur. Specifically,
they found that exposure to multiple Chinese-accented speakers
improved adaptation to a novel Chinese-accented speaker to a
larger extent than exposure to a single Chinese-accented speaker
did. Different groups of native English listeners heard sentences
in noise produced either by five unaccented speakers, the same
Chinese-accented speaker who would serve as the test speaker,
one Chinese-accented speaker who was not the test speaker, or
five Chinese-accented speakers, and were asked to transcribe sen-
tences played to them. Following this exposure phase was another
sentence transcription task serving as a test. Participants who
heard one Chinese-accented speaker in training and a different
Chinese-accented speaker at test did not perform any better than
participants who heard unaccented speakers in training. In con-
trast, exposure to multiple Chinese-accented talkers resulted in
adaptation to a novel Chinese-accented talker, at a level equiv-
alent to being trained with the test talker. Thus, it seems that
exposure to multiple talkers of the target foreign accent can be
an effective means of achieving talker-independent adaptation in
adults. Interestingly, this adaptation was accent-dependent rather
than accent-general since training on Chinese-accented English
(whether with one or five talkers of the accent) did not result
in adaptation to another unfamiliar accent (Slovakian-accented
English).
Other work focuses on the effects of long-term exposure. For
example, native speakersof British English show less difficulty pro-
cessing American English speakers’ productions of medial-/t/as
a tap (“ciddy” for “city”) if they have lived in the United States
(Scott and Cutler, 1984). Sumner and Samuel (2009) studied the
processing of r-final words spoken by talkers of rhotic dialects
(General American English/GA), who pronounce the final/r/, or
non-rhotic dialects (New York City English/NYC), who do not
pronounce the final /r/ (“bakuh” for “baker”). This study used
a priming paradigm and tested listeners who were either famil-
iar or not with NYC English. For all participants, a target was
never primed as well if it had a prime of a mismatching dialect,
showing an overall cost of switching dialects from prime to target.
Further, participants with prior exposure to NYC English showed
both form (“slenda”primes “slenda” and “slender”) and semantic
priming (“slenda”primes “thin”) for NYC English primes on NYC
English and GA English targets. However, GA English speakers
did not show semantic priming for NYC English primes (“slenda”
does not prime “thin”), suggesting that experience with the dialect
is necessary for a dialect form to facilitate processing. Exposure
effects are also evident when exposure is mostly through media,
as shown by studies of asymmetrical cross-accent perception. For
example, Impe et al. (2008) found that Netherlandic speakers of
Dutch are much slower in a lexical decision task when processing
words recorded by Belgian Dutch speakers than by fellow Dutch-
men, while Belgian Dutch speakers process both varieties equally
well. Similarly, in Adank et al. (2009), a truth value judgment task
was administered to both Glasgow and London listeners using
spoken stimuli recorded from Glasgow and London talkers. While
Londoners were slower with the speech from Glasgow speakers
than the speech from speakers of their own accent, Glaswegians
were equally fast with both accents. These asymmetries fit with
asymmetries in media exposure of the two accents.
A combination of behavioral and electrophysiological measures
begins to shed light on the prelexical and lexical effects of a lifetime
of exposure to multiple linguistic varieties on the perception of
native contrasts, as well as those present in non-native, but famil-
iar, varieties. This question has been approached using regional
variation in French. The varieties spoken in France have either lost
or are in the process of merging /e/ and /ε/, a contrast that has not
merged in the varieties spoken in Switzerland. Current results indi-
cate that long-term exposure to a variety where a given contrast
is merged (i.e., French as spoken in France) could actually result
in loss of discrimination in one’s own unmerged variety (affecting
Swiss listeners; Brunellière et al., 2009, 2011). Additionally, being
exposed to a variety which has a contrast where one’s native vari-
ety has merged does not suffice to preserve baseline discrimination
(Dufour et al., 2007), nor typical lexical access (Dufour et al., in
press); and discrimination training on the preserved contrast does
not result in normal lexical processing either (Dufour et al., 2010).
Pure exposure, however, is not the only factor affecting accent
processing. In the semantic priming study discussed above, Sum-
ner and Samuel (2009) also reported that participants who were
familiar with the NYC dialect but were not non-rhotic in their own
productions did not retain the priming benefit after a 20–30 min
lag, while participants who were non-rhotic in their own produc-
tion did show long-term priming. Similarly,Kendall and Fridland
(2012) report that individual listeners from various regions of the
USA categorized vowels along an /e/-/ε/ continuum differently
depending on not only the listener’s region of origin, but also
their own production of those sociophonetically marked vowels.
The correlations between production and perception can even be
tracked longitudinally. Evans and Iverson (2007) collected pro-
duction and perception measures over a period of 2 years from
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 3

Cristia et al. Accent perception across the lifespan
students who had moved from northern England to attend college
in southern England. Results showed that students who displayed
more southern English accent features in their own speech also
exhibited better adaptation to southern speech in perception tasks.
Naturally, it is impossible to know whether such correlations
are mediated by the quality of experience in terms of perception
only versus production-perception, or whether the social values
and/or experiential opportunities that accompany a given speech
pattern also play a role (Labov, 2007). That is, one can posit many
explanations: perhaps these correlations are caused by amount of
exposure to the preferred accent (e.g., people who adopt the local
accent are precisely those who spend more time with locals), or by
the social valence attached to accent adaptation (e.g., people who
show incipient traces of adaptation in their production get more
positive reinforcement from locals, which further boosts their per-
ceptual adaptation). In a recent laboratory study, the causal link
between production and ease of processing has been documented:
Adank et al. (2010) evaluated ease of processing (as measured in
terms of the maximal tolerated signal to noise ratios) of a rela-
tively unfamiliar accent after different types of training. In some,
listeners simply had to listen; in others, they were asked to repeat,
or write down the words they heard; in yet others, they were asked
to repeat imitating the accent. Significant improvements were only
evident when the training involved imitation of the novel accent.
Interestingly, our knowledge of regional accents shapes our
perceptual expectations. In fact, merely evoking a specific region –
through the name of a place being written on a response sheet
(Niedzielski, 1999) or through exposure to a stuffed toy that brings
into mind a particular place (Hay and Drager, 2010) – causes
listeners who are familiar with the dialects of both regions to
categorize vowels in accordance with the evoked region’s accent
pattern. The shifting of phoneme categorization boundaries seen
in these studies reflects adaptation to the incoming speech sig-
nal, contingent upon the listener’s knowledge of the patterns of a
particular within-language accent.
SUMMARY
This brief overview of accent perception research in young adults
has allowed us to identify a few key findings, which will be car-
ried over in our review of the developing and older populations.
Accented speech initially perturbs word recognition and/or sen-
tence processing in terms of accuracy (e.g., Gass and Varonis, 1984)
and speed of processing (e.g., Floccia et al., 2006). With exposure in
lab conditions, evidence of adaptation can be found (e.g., Clarke
and Garrett, 2004). A lifelong exposure to a variety of accents
shapes perceptual abilities so that listeners are able to process each
variant equally rapidly (e.g., Sumner and Samuel, 2009), suggest-
ing certain flexibility of the representations or the way the signal
is mapped onto them. Finally, processing ease varies with factors
that go beyond simple exposure (e.g., Kendall and Fridland, 2012).
ACCENT PERCEPTION IN INFANCY AND TODDLERHOOD
Before embarking on this section, we remind readers of the unique
challenges of research with very young children who cannot
respond to explicit instructions. Despite this challenge, there is
one considerable empirical advantage to working with infants and
toddlers: it is much easier to gage, and at times even measure, the
linguistic input participants have had over the course of their rel-
atively short lifespan. As a result, studies on infants and toddlers
can shed unique light on the effects of exposure. It should also be
mentioned that this is a burgeoning research area, and that some
of the questions set out here have already begun to be answered.
PROCESSING COSTS
In accent discrimination studies, infants are familiarized with one
linguistic variety until they accumulate a certain amount of “atten-
tion time.” Then in a subsequent test phase, infants are presented
with two types of trials, one in which the familiarized variety is
presented, and another in which a novel variety is presented. Pre-
sentation of the sound is contingent on the infant orienting toward
the source of the sound, and looking times are measured. Dis-
crimination is surmised to have occurred if infants orient longer
when hearing the novel variety than when the familiarized variety
is being presented. Under these conditions, 5- and 7-month-olds
can only discriminate rhythmically similar linguistic varieties if
they have experience with at least one of the varieties on which
they are tested; that is, they can discriminate their native variety
from an unfamiliar one, but not between two unfamiliar, rhyth-
mically similar linguistic varieties (Nazzi et al., 2000;Butler et al.,
2011).
Preference paradigms skip the familiarization phase to tap
infants’ early preferences for one variety over another,simply mea-
suring infants’ attention while they hear utterances in their own or
an unfamiliar variety. In this paradigm, preference is dependent
on age (younger infants show stronger preferences than younger
ones), and experience (infants with some exposure to the non-
native variety lose their preferences earlier; Kitamura et al., 2006).
A decrease of preference for the native over the non-native variety
has been taken as evidence that infants learn to interpret the unfa-
miliar accents as a variant of the native accent. In other words,
younger infants fail not because they are unable to implement
compensatory strategies to “correct” for the accent, but rather
because they reject the accented speech as a potential implemen-
tation of their native language. However, an alternative cognitive
interpretation is that these early preferences are driven by ease
of processing. For example, it may be that young infants have a
difficult time processing unfamiliar variants, and thus implicitly
dislike the non-native variant (this is a possibility that we dis-
cuss in greater detail in See Concluding Remarks). As they age,
their processing abilities may expand (e.g., vocabulary size, work-
ing memory, selective attention), and thus they may find it easier
to process even the unfamiliar properties of accented speech. This
would occur in the domain of accent processing similar to the
way it occurs in other cognitive domains that involve processing at
more than one level. For example, as working memory increases,
the child is able to process both the social valence of an accent
(Kinzler et al., 2007), and its linguistic content simultaneously.
The two explanations differ only in whether age changes reflect
variation in metalinguistic judgments (i.e., as the infant ages, she
learns that a language may have different accents, and all of them
should be accepted), or variation in executive functions (i.e., as
the infant ages, she becomes better at coping with difficult tasks).
These alternatives could be explored by simplifying the task for
younger infants (for example, presenting a single word, which
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 4

Cristia et al. Accent perception across the lifespan
reduces memory load), or by increasing processing demands in
older infants (e.g., through the addition of a concurrent task). If
either of these manipulations reverts previous preference patterns
(disappear in younger infants, reappear in older infants),then pro-
cessing resources (rather than a metalinguistic reinterpretation of
language varieties) would be implicated.
Beyond this issue of interpretation, it is clear that young infants
are sensitive to the acoustic changes introduced by different lin-
guistic varieties. Consequently, one may wonder whether infants
are able to segment wordforms when suprasegmental and sub-
phonemic cues are altered by accented speech. This is impor-
tant because other research suggests infants are very sensitive to
acoustic mismatches in the wordforms they hear (Singh et al.,
2004). For example, infants sometimes fail to recognize words
across very different voices (Houston and Jusczyk, 2000; but see
van Heugten and Johnson, 2012). In a typical segmentation task,
infants are first familiarized with a wordform (e.g., they hear the
word “candle” produced in isolation several times). At test, they
are presented with several passages, only half of which contain
the familiarized wordform, and the presentation of the passages
is contingent on infants’ attention. When the same voice, speak-
ing in the infants’ native accent, is used in familiarization and
test, even 7.5-month-olds reliably prefer passages containing the
familiar wordform (Jusczyk and Aslin, 1995). This is already no
small feat, given that word segmentation is an incredibly com-
plex task, as there is no single reliable cue to word boundaries
in all languages. Interestingly, Polka and Sundara (2012) report
that Quebecois French-learning 8-month-olds are as capable of
segmenting words from unfamiliar European French as they are
from their native variety,but they fail to segment Canadian English
(which is also unfamiliar to them).
These results could suggest that 8-month-olds can already
accommodate for within-language varieties, in a way that does
not extend to an unfamiliar language. However, we believe this
interpretation is too strong in view of the following two sets of
results. First, infants raised in Paris show no behavioral evidence
of segmenting words from their own native variety until several
months later (Nazzi et al., 2006), yet these European French-
learners succeed in segmenting the unfamiliar Canadian French
variety at 8 months (Nazzi et al., 2012). Further, American English-
learning 9-month-olds are able to segment words in Dutch, a
language unfamiliar to them (Houston et al., 2000). Segmenta-
tion in these studies could be resolved through some approximate
pattern-matching that does not require extensive experience with
the specific linguistic variety, and it could be affected by a host of
acoustic and prosodic factors that are difficult to control across
varieties (see e.g., Singh, 2008, for evidence of the impact of non-
linguistic variation in such tasks). For example, Nazzi et al. (2012)
propose that irrelevant prosodic cues (e.g., the quality or degree
of the infant-directed speech and hence its likability) could shape
infants’ performance when not explicitly and carefully controlled.
Cross-accent segmentation studies ask whether infants can rec-
ognize and segment a familiarized word across the native variety
and an accented variety. Specifically, the procedure is identical to
the segmentation studies described above, except that the familiar-
ization stimuli are spoken in one accent, and the test passages in a
different accent. Using this design, Schmale and Seidl (2009) tested
9- and 13-month-olds on their tolerance for a change between
the native and a foreign (Spanish) accent, whereas Schmale et al.
(2010) assessed accommodation across two within-language Eng-
lish accents (North Midland American and Southern Ontario
Canadian) in 9- and 12-month-olds. In both cases, the older group
succeeded where the younger group failed. Naturally, as we pointed
out for the language preference tasks,here the effects of experience
and maturation are confounded. Thus, one possibility is that older
infants are better at segmenting words because most of them have
accumulated more experience with diverse talkers, which allowed
them to develop linguistic strategies to cope with talker variability.
But there are many other alternatives; to give just one example, they
may have also undergone experience-independent advancements
in cognitive skills recruited by the task, such as selective attention
and working memory. These advancements could be clearer when
processing accented speech because infants are not at ceiling in
that task.
Perhaps stronger evidence for accommodation to a novel accent
is provided by word preference studies. In this paradigm, experi-
menters measure toddlers’ preference between two types of trials:
trials with high-frequency wordforms, which are likely to be famil-
iar to them (e.g., bottle); and trials with low-frequency wordforms,
which toddlers have likely never heard (e.g., enzyme). In both
kinds of trials, wordforms are presented auditorily, without any
visual support, and attention is gaged by making the presentation
of the speech contingent on the infant response. Infants as young
as 11 months show a strong preference for high-frequency word-
forms when tested in their familiar language variety (Hallé and de
Boysson-Bardies, 1996). However, 15-month-olds growing up in
Connecticut, USA showed no such preference when tested in an
unfamiliar variety (Jamaican English), but they did when tested
with their native variety, whereas 19-month-olds showed a high-
frequency preference within both native and unfamiliar varieties
(Best et al., 2009). This high-frequency preference paradigm offers
a promising avenue of research, since it could shed light on tod-
dlers’ long-term representations. Nevertheless, one area that needs
to be explored more carefully is what exactly underlies these pref-
erences. The standard interpretation is that toddlers are matching
the wordforms they hear with their own long-term lexical entries.
However, it is also possible that toddlers are evaluating these items
phonologically only, since the contrast between high- and low-
frequency wordforms has typically confounded the frequency of
sounds and phonotactics.
Up to this point, we have discussed how infants and toddlers
recognize wordforms in unfamiliar linguistic varieties. But a key
point in accent accommodation is how we come to realize that two
different pronunciations map onto the same referent. This topic
has been breached using intermodal preference paradigms, where
a wordform is heard at the same time that two pictures are dis-
played in the visual modality. Typically, toddlers will look longer
at the target (the picture matching the wordform heard) than the
competitor (another image on the screen that does not match
the wordform). This matching preference ensues provided that
the wordform is sufficiently similar to the target’s name to prime
this association, that is, even when it is not identical (Swingley and
Aslin,2000). In Mulak et al. (in press),it is reported that 15-month-
olds showed a significant preference for the matching picture
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 5

Cristia et al. Accent perception across the lifespan
when the wordform played was produced in their native vari-
ety (Australian English), but not in an unfamiliar one (Jamaican
English), suggesting that the unfamiliar pronunciation departed
significantly from the one toddlers were accustomed to process-
ing. Similarly,unfamiliar accents may prevent recognition of newly
learned words. Schmale et al. (2011) taught toddlers a new word
by pairing a wordform with a picture. Then they were tested on
their recognition of that word in two subsequent trials involving
changes in language varieties. In one test trial, two pictures were
displayed on the screen while the familiar wordform was provided
(looks to the matching target are expected if children have learned
the word-object association). The second test trial was more cog-
nitively demanding, since a new label was provided, and toddlers
were expected to infer that the correct referent was the competitor.
In this demanding task, 30-month-olds were able to recognize a
newly learned word across Spanish-accented and native English
pronunciations, regardless of which variety was used in training
and test. Contrastingly, 24-month-olds showed significant pref-
erences for the object that matched the label (the trained object,
when the trained label was provided; the novel object otherwise)
when trained with a Spanish-accented talker and tested with a
native English talker, but not when the opposite presentation order
was provided. This order of presentation effect suggested that even
short exposures to the accent could suffice in easing children into
the unfamiliar accent, a possibility that was investigated in a study
reported in the next section.
EFFECTS OF EXPOSURE
White and Aslin (2011) examined the effects of exposure to an
accent on toddlers’accommodation of an unfamiliar variety using
lexical feedback. Specifically, during a training phase, 19-month-
olds saw pictures of highly familiar objects (e.g., block, bottle)
while hearing the vowel in the words associated with that object
consistently produced with an (æ) sound (as “black, battle”). At
test, toddlers evidenced generalization of the consistent sound
change to untrained, highly familiar words. For example, they
looked longer to a picture of a sock (than to a picture of an irrel-
evant item) while hearing the word “sack,” but not when hearing
the word “sick,” showing that the sound reinterpretation was rel-
atively precise. Thus, 19-month-olds can adapt to novel accents
when provided with clear and sufficient evidence.
Other work suggests that toddlers also benefit from more natu-
ralistic exposure to a complex accent Schmale et al. (2012) exposed
toddlers to brief stories with no accompanying visual referent.
Thus, no effort was made to train toddlers on the host of phonetic
changes imposed by a natural Spanish accent. After 2min of expo-
sure to such speech, 24-month-olds were able to recognize a newly
learned word across their native accent and the foreign accent.
Their performance was improved both when the same speaker
was used for pre-exposure and test, and when four different voices
with the same accent, none of whom produced the test stimuli,
told the brief stories. Toddlers’ performance in accommodating
the foreign accent was unaffected by a pre-exposure to one or four
native English speakers, suggesting that the improvement was truly
driven by foreign accent exposure.
These recent training studies suggest that even brief exposure
can reshape infants’ perception of unfamiliar linguistic varieties
of speech. A natural follow-up question is how long-term expo-
sure to multiple varieties affects early development. One intriguing
study suggests that bi-varietal toddlers recognize words better in
the variety that is more widely spoken in their general environ-
ment, even if they have greater exposure to the minority form
(Floccia et al., 2012). Word recognition was assessed in 20-month-
olds growing up in a region where rhoticity was prevalent (e.g.,
“car” pronounced with a final “r” by most of the population).
There were two groups of participants. One was a mono-varietal
group, where both parents produced rhotic variants, as in the local
environment. The other group was bi-varietal, because they were
exposed to the locally predominant rhotic variant outside of the
home, and they were exposed to a non-rhotic variant at home as
one or both parents spoke a non-rhotic variety. Estimation of their
weekly exposure showed that, overall, the bi-varietal children had
more exposure to the non-rhotic variant (through one or both
caregivers) than the rhotic variant (through the rhotic caregiver if
they had one, plus the local context). All toddlers were tested with
three groups of words: control items (lexical items that contained
no “r”; produced by non-rhotic talkers), rhotic items (lexical items
that contained “r” and had been produced by rhotic talkers), and
non-rhotic items (lexical items that contained “r” but had been
produced by non-rhotic talkers, and therefore there was no /r/ in
them). Results revealed that both mono- and bi-varietal toddlers
looked at the target significantly longer for both control and rhotic
items, but neither group lookedab ove chance for non-rhotic items.
Thus, it appears that 20-month-olds are particularly sensitive to
the frequency of forms across the local population. Notice in par-
ticular that the bi-varietal toddlers still performed better with the
rhotic variant, despite their personal larger frequency of exposure
to the non-rhotic forms.
SUMMARY
Before the age of 6 months, infants seem to have learned accent-
specific information about their most commonly encountered
variant that allows them to discriminate and prefer it to unfa-
miliar variants (e.g., Kitamura et al., 2006). Whereas infants’ and
toddlers’ linguistic processing is impeded by an unfamiliar accent
early on, they become better able to recognize words and word-
forms across different accents with further development (due to
changes in maturation and/or experience; e.g., Schmale and Seidl,
2009). Even holding maturation constant, brief exposure enables
toddlers to cope with accent changes (e.g., Schmale et al., 2012).
Finally, toddlers are better able to recognize words produced in
the linguistic variety spoken widely in their local community, even
when they primarily hear another variety at home (Floccia et al.,
2012).
Thus, results reviewed largely coincide with the picture found
in young adults, in that there are initial processing costs when a
novel accent is encountered, which are diminished through brief
exposure. However, results in infants and young adults do not
align with respect to long-term exposure to multiple accents. In
adults, a lifetime of exposure to an accent provides listeners with
the ability to access the same lexical items through both varietal
forms; for example, Sumner and Samuel (2009) document prim-
ing across regional variants. In contrast, the only study carried
out so far with toddlers suggests that they do not store all variants
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 6

Cristia et al. Accent perception across the lifespan
encountered, nor do they attend more to those that are heard more
frequently, but focus on variants that are more prevalent in their
environment (Floccia et al., 2012).
An important contribution of this literature relates to the
marked developmental changes that have been documented, with
the general pattern being that older age groups succeed in rec-
ognizing wordforms across accents more readily than younger
ones. One interpretation of such findings is that experience
teaches learners to ignore accent-relateddifferences, allowing them
to retrieve the constant phonological and lexical information.
Indeed, this is the viewpoint commonly adopted when document-
ing developmental changes in initial processing costs (e.g., Best
et al., 2009). Another interpretation is that developmental changes
are not tied with an awareness of what kinds of variation are to be
ignored, but rather with changes in cognitive flexibility or exec-
utive control. For instance, one could postulate that advances in
selective attention, memory, and executive control allow toddlers
to more readily process accented speech in terms of both linguistic
and social dimensions. The role of multiple cognitive and linguis-
tic factors in accent processing and perceptual adaptation is a topic
that we revisit in Section “Accent Perception in Late Adulthood”.
A second key contribution of the infant literature pertains to
the mechanisms triggered by exposure to a novel accent. The dis-
appearance of initial costs with exposure can occur in the absence
of a large lexicon (as even 12-month-olds can recognize newly
trained wordforms across accents; Schmale et al., 2011), and the
effects of exposure to an accent are evident even without explicit
lexical training (Schmale et al., 2012). We believe that the contri-
bution of non-lexical factors to accent perception and adaptation
is relatively understudied, as it is generally assumed that accent
accommodation is primarily guided by lexical entries (there are a
handful of exceptions, such as the studies cited above by Bertel-
son et al., 2003 and Cutler et al., 2008). The questions of how and
whether listeners of different ages cope with accentual variation
at phonological levels where lexical feedback is irrelevant (e.g.,
intonation) could be investigated in future work.
ACCENT PERCEPTION IN CHILDHOOD
Accent perception during childhood is less well-documented than
accent perception in early infancy or in adulthood, possibly
because the focus of many studies with children has been on
production. Indeed, numerous questions within this topic have
been studied, such as when children acquire local features of their
native variety (Labov, 1989;Roberts and Labov, 1995;Roberts,
1997a,b, 2005;Jacewicz et al., 2010) or even a new language (Flege,
1999, 2003;DeKeyser, 2000) and to what extent they manage to
acquire an accent in the local variety of a region they move into
(Tagliamonte and Molfenter, 2007). Accent production research
in children suggests an outstanding ability to acquire a new accent
(e.g., Tagliamonte and Molfenter, 2007), which very likely sug-
gests an excellent perceptual flexibility for accent variations. Some
work argues that foreign accent in caregivers is ignored (in order
to acquire the local native accent; Chambers, 2002). One line of
research within perception has thus studied potential differences
in the detection of native and foreign accents.
This line of work builds on research on young adults, who
are highly adept at rating the accentedness of talkers (e.g., Piske
et al., 2001), identifying and discriminating accented talkers (e.g.,
Goggin et al., 1991;Winters et al., 2008;Perrachione et al., 2009),
and categorizing accented talkers according to regional language
background (e.g., van Bezooijen and Gooskens, 1999;Clopper and
Pisoni, 2004a,b; Clopper and Bradlow, 2008). Children have also
been tested on their ability to categorize talkers on the basis of
their accents. Girard et al. (2008) presented French-speaking 5- to
6-year-olds with sentences in two accents and instructed them to
group the speakers into two sets according to their accent. Results
showed that children succeeded in this task only if the unfamil-
iar accent was a foreign accent, and not a within-language accent.
Additional experiments showed that children were able to hear the
within-language accent differences in a simple discrimination task,
albeit to a lesser extent than foreign accent differences. This was
interpreted as evidence that children were more sensitive, or aware
of, unfamiliar foreign accents than of unfamiliar within-language
accents. However, the perceived strength of the accent in the for-
eign accented stimuli was stronger than in the within-language
accent ones in that study, which could have explained the greater
salience of the foreign accented features than the within-language
features. Floccia et al. (2009) addressed this concern by selecting
stimuli spoken in a regional (Irish) accent and a foreign (French)
accent on the basis of similar ratings of accent strength by British
speakers of English. They then presented these items to 5- and 7-
year-olds in a simpler version of the sentence categorization task,
in which children were simply asked tospot the speaker who “spoke
like an alien.” Results were similar to those of Girard et al. (2008) in
the older group of children. Specifically, 7-year-olds were better at
spotting the foreign accent over the within-language accent. Curi-
ously, this same tendency was not statistically significant for the
5-year-olds. One interpretation of these results is that children are
increasingly sensitive to a foreign accent with age and experience.
More generally, this perceptual work demonstrates that children
can perform metalinguistic tasks that require a reliance on accent.
INITIAL PROCESSING COSTS
In Nathan et al. (1998), Londoner children were presented with
isolated words produced in either a London or a Scottish Glaswe-
gian English accent. After hearing a short story spoken in an accent,
4- and 7-year-olds heard words in isolation and were asked to
report what word the talker had said, and to define it (notice
that, since there was no control group who did not hear the story,
we cannot shed light on the effects of exposure and thus do not
include this paper in the next section). Responses were classified
as phonological if the child correctly identified the Glaswegian-
accented word,repeating it in her own London accent, or phonetic
if the child imitated the form in the Glaswegian accent (and thus
produced a form that was not a word in the London accent).
As expected, accuracy in the definition task increased with age:
older children gave significantly more phonological responses than
phonetic ones, showing that they “corrected” the accented input.
Alternatively, younger children gave significantly more phonetic
responses than older ones. Enhanced phonetic sensitivity at 4 years
(as compared to 7 years) might lead to enhanced adaptation abil-
ities, which in turn would lead to an enhanced capacity to learn
a new accent in production. Of course, another interpretation of
these results is that younger children’s remarkable sensitivity to
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 7

Cristia et al. Accent perception across the lifespan
phonetic contrasts prevented them from accessingstored represen-
tations, as if acute phonetic abilities were blocking lexical accom-
modation. In sum, the Nathan et al. (1998) study provides the first
evidence of developmental changes in within-language accent per-
ception in childhood, documenting an improvement in children’s
ability to identify common words across within-language accent
variations between the ages of 4 and 7 years. However, the presen-
tation of isolated words in the absence of context might have been
partially responsible for the younger children’s inability to retrieve
the intended word produced in the unfamiliar accent, whatever
the underlying explanation. Therefore, studies using continuous
speech sequences, or a greater guiding context, may provide a more
realistic view of how children cope with unfamiliar accents in their
everyday life. With this goal in mind, Holtby (2010) presented 9-
and 15-year-olds with foreign accented sentences and asked them
to assess the truth value of the sentence, to rate the intelligibility
of the sentence, and to rate the accent strength of the sentences.
In this study, there was no developmental difference in detecting
a foreign accent, as shown by similar accentedness ratings in the
9- and 15-year-olds and similar intelligibility ratings. Although a
significant improvement in the truth value judgments was found
across ages, it is unclear whether this followed from gains in the
ability to process accented speech or simply the ability to perform
the task well.
EFFECTS OF EXPOSURE
Although no work has experimentally isolated the effects of expo-
sure to natural accents, some research using laboratory learn-
ing of artificial accents suggests that children can learn to map
ambiguous sounds to specific categories when provided with suf-
ficient information. In McQueen et al. (2012), 6- and 12-year-olds
learned to map an ambiguous sound between /f/ and /s/ onto one
of these endpoints after hearing them in the context of unam-
biguous lexical items (such as platypus and giraffe). Other work
suggests that there may be some developmental differences in
the ability to integrate multiple cues in order to perform such
remapping. van Linden and Vroomen (2008) presented 5- and 8-
year-olds with videos where talkers said /aba/ (or /ada/) when the
paired audio was an ambiguous sound between /b/ and /d/, and
videos of /ada/ (or /aba/) with an unambiguous audio. As adults
had in a previous study (Bertelson et al., 2003), 8-year-olds clearly
learned to interpret the ambiguous sound in terms of the visually
presented category. In stark contrast, 5-year-olds showed no such
recalibration.
SUMMARY
Children are remarkably good at spotting accents, although they
may be better with some accents than others (e.g., Floccia et al.,
2009). During childhood, the ability to retrieve meaning from
accented speech improves with age (Nathan et al., 1998). Short-
term exposure to an accent with a single sound change guided
by visual or lexical information clearly shapes children’s per-
ception, although there may be developmental changes in the
ability to profit from bootstrapping information (van Linden and
Vroomen, 2008). Surprisingly, no work has assessed the effects of
long-term exposure to an accent in childhood, although clearly
this must matter in view of effects on production (Tagliamonte
and Molfenter, 2007). It is unclear whether similar mechanisms
are used by children and young adults, or whether adaptation
strategies vary with cognitive and linguistic development. This
seems an unfortunate state of affairs. The empirical and theo-
retical contribution that research with children has made to our
understanding of the development of language production is sim-
ply extraordinary. There is every reason to believe that it will
be just as insightful to evaluate the development of perception,
as in childhood there are dramatic changes in lexical and cog-
nitive development, which could play a role in accented speech
perception.
ACCENT PERCEPTION IN LATE ADULTHOOD
Research is increasingly turning to how older adults cope with
dialectal, foreign, or simply novel accents, a question that is both
theoretically and empirically important. There are several factors
which change with aging that could impact accented speech per-
ception. To begin with, older adults often suffer from age-related
hearing loss (presbycusis), which impairs sensitivity (i.e., loud-
ness), and fine tuning (i.e., spectral resolution). This hearing loss
may render speech perception in general more difficult. It could
potentially decrease the difficulty gap between accented and unac-
cented speech, as it leads listeners to rely on context more. Alter-
natively, it could also increase the challenge of processing unusual
speech patterns, because less information is available. Aging also
impacts cognitive function, including speed of processing, working
memory, long-term memory, and inhibition or cognitive con-
trol (a recent review in Park and Reuter-Lorenz, 2009). Accented
speech imposes both linguistic and cognitive load challenges for
the listener (see Concluding Remarks for further discussion). The
confluence of a reduction in signal strength (due to the listener’s
presbycusis), and a reduction in cognitive function (which may be
more evident in harder tasks; i.e., for accented speech more than
unaccented speech) could lead to significant impairments when
aging adults are faced with unusual accents. Understanding how
older adults process accented speech thus sheds light on the factors
involved in dealing with such variation in the spoken signal, and
could provide us with cues to facilitate communication for older
adults.
INITIAL PROCESSING COSTS
Some previous work has reported that older adults experience
greater difficulty when faced with multiple talkers (Sommers,
1996), fast speech (Janse, 2009), and speech in noise (e.g., Kalikow
et al., 1977) than younger adults. Thus, one could expect older
adults to experience even greater difficulties than younger adults
when presented with accented speech. However, most work focus-
ing on this question confirms that while adults do have greater dif-
ficulty with accented than unaccented speech, the size of this effect
is not significantly greater for older than younger listeners (Burda
et al., 2003;Shah et al., 2005;Ferguson et al., 2010;Gordon-Salant
et al., 2010a). It does not appear that the absence of a difference
across age groups could be due to a ceiling effect (Adank et al.,
2009; put forward a similar argument to explain data collected
from younger listeners). Indeed,while the addition of multi-talker
babble yielded age differences in one study (Gordon-Salant et al.,
2010b), it did not in another (Ferguson et al., 2010).
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 8

Cristia et al. Accent perception across the lifespan
Several factors could explain differences in results like this one.
First, perhaps different studies sample from a diverse population
without considering variables that would structure this variabil-
ity. Subsequent work has more explicitly attempted to control for
differences in hearing status (e.g., Burda et al., 2003;Ferguson
et al., 2010), which is usually found not to interact with accented
speech perception (but see Gordon-Salant et al., 2010b for a more
complex pattern of results; and Janse and Adank, 2012, for con-
trary results). However, there is still some progress to be made in
understanding the effects of cognitive decline, and its contribution
to the diversity of results reported. Based on individual variation
data, Janse and Adank (2012) report that memory subsystems play
a role in accented speech processing. Finally, it may be the case
that different results are partially due to differences in the stimuli,
particularly the quality of the accent under study or the amount
of familiarity with it.
EFFECTS OF EXPOSURE
Work on adaptation to altered speech, such as time-compressed or
noise-vocoded signals, suggests that older adults definitely profit
from short-term exposure to such degraded signals, although the
extent to which they resemble younger adults is still a matter of
debate (there was no difference across groups in Golomb et al.,
2007; in Peelle and Wingfield, 2005, older adults adapted to time-
compressed and noise-vocoded speech similarly to younger adults,
although the former showed smaller maintenance and transfer
effects). More general research on implicit learning and aging sug-
gests that older adults may adopt compensatory strategies, which
are often sufficient, except for particularly challenging tasks (for a
recent review, see Rieckmann and Bäckman, 2009).
Overall, studies on short-term exposure to accented speech
support the view that older adults can adapt to novel speech
patterns after brief exposure, but, as with perceptual learning in
general, some report differences between older and younger listen-
ers (Adank and Janse, 2010 reported that older listeners stopped
adapting more quickly than younger ones); and others find no sig-
nificant differences across age groups (Gordon-Salant et al.,2010).
As argued above, diverse results could be explained by sampling
from a variable population. Janse and colleagues have begun inves-
tigating whether individual variation in accented speech compre-
hension and adaptation correlates with individual variation along
cognitive and linguistic dimensions. For example,Janse and Adank
(2012) report that both measures of selective attention and vocab-
ulary predicted adaptation in a group of older adults. Moreover,
while some indices of executive function can predict adaptation,
they might not be the same in younger and older adults (Jesse and
Janse, 2012). This is clearly a promising avenue of research, which
could inform our understanding of the variety of mechanisms that
are involved in both accent perception and adaptation.
SUMMARY
Although only a handful of studies have been carried out with
older adults, it is clear that this population experiences an ini-
tial cost when processing accented speech,which may be rendered
smaller through exposure. But it is still unclear whether this cost is
greater or smaller than that found in younger adults, and whether
adaptation occurs at the same or a different pace, with the same or
different mechanisms. The greatest contribution of the research
reviewed in this section centers on the study of factors that struc-
ture individual variation in performance. Recent results suggest
that accent perception, at least in this population,is greatly affected
by linguistic (e.g., vocabulary size) and cognitive (e.g., executive
control) dimensions. At this point, however, it remains difficult
to determine to what extent initial processing costs and effects of
short-term exposure are qualitatively different (i.e., recruit a dif-
ferent profile of speech perception mechanisms) in younger and
older adults. As with research in childhood, we feel that this is an
understudied population, and hope future work will shed further
light on these already interesting results.
GENERAL DISCUSSION
GENERAL POINTS OF CONVERGENCE AND DIVERGENCE ACROSS
MULTIPLE AGE GROUPS
Specialists in infancy, childhood, adulthood, and aging have all
investigated initial processing costs, although the level of preci-
sion with which these questions have been investigated varies. The
exact ways in which an accent trumps speech processing have been
best described for younger adults. Accent affects both speed and
accuracy, and may lead to different processing styles in young
adults. As for adaptation, it is clear that individuals of all ages can
learn to adapt to new accents.
In terms of the mechanisms recruited, lexical feedback is clearly
the main source of information that learners have been assumed
to use, and the focus of attention has been on single segmental
changes. However, infants can adapt to new accents when they
are too young to have a large lexicon; and they can do so with-
out a disambiguating lexical context. These facts should inspire
adult researchers to consider other aspects of accent processing.
We predict that accent adaptation, particularly in infancy, can be
triggered by suprasegmental deviations. The presence of such devi-
ations would invite listeners to employ processing schemes that
are robust in the face of uncertainty; for example, they should
allow less strict acoustic matching and combine more cues for
segmentation. We state that such suprasegmental deviations must
be crucial triggers of adaptation because young infants have not
yet established their native phonological inventory, but we expect
that the importance of such “suprasegmental cues to accented-
ness” may decline with age and experience. In contrast, it is to
be expected that lexical factors play an increasingly large role
throughout toddlerhood and later childhood, as lexical growth
allows listeners to detect accents through mismatches between the
original and expected lexical forms. An interesting question, to
which the answer is far from obvious, is whether the influence of
suprasegmental and lexical cues to accentedness could be a stable
predictor of individual differences in accented speech perception
in older adults.
One recurrent question was the format of adaptation. While
some suggest that learners extract prelexical patterns, others favor
lexical storage as the way in which learners capture their newly
gained accent knowledge. We have reviewed evidence that 19-
month-olds exposed to an artificial accent did not accept any
sound change in untrained items, but only mispronunciations
along the lines of the experienced sound change (White and Aslin,
2011). However, this may not indicate that phonemic remapping
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 9

Cristia et al. Accent perception across the lifespan
is already perfect at this young age. For example, van Linden and
Vroomen (2008) suggest that additional experience helps learn-
ers become more informed listeners, allowing them to integrate
multimodal information.
Furthermore, the work carried out with older adults suggests
that certain cognitive skills (such as selective attention) could play
a major role in adaptation. It is crucial to extend this insight to
other populations, and particularly to infancy, toddlerhood, and
childhood, where cognitive skills are in constant development. A
good approach would consider the linguistic factors discussed in
the previous paragraphs, as well as precise cognitive constructs,
in the description of the system underlying perception of deviant
speech patterns, rapid short-term adaptation, or gain of novel
processing strategies, and long-term changes in a multi-varietal
context.
Finally, while hearing status has not been found to impact
processing in older adults, it has not been explored in any other
age group. Necessarily, having a distorted or smaller signal could
have a much greater impact in infancy and childhood, and inter-
act in more complex ways with cognitive skills than it does in
older adults. Additionally, future work should examine special
populations, such as autistic spectrum disorders (ASDs),Williams
Syndrome, and Specific Language Impairment (SLI). Individuals
with ASDs are characterized by deficits in social interaction and
suprasegmental processing, SLI is characterized by deficits primar-
ily in language processing, and children with Williams Syndrome
show a good ability to navigate social territory, yet low IQ scores.
This work could shed unique light on the influence of certain
social, cognitive, and linguistic factors on accented speech per-
ception, in addition to making steps toward the study of speech
perception by all language users, and not only normative ones.
One aspect of accented speech perception that we have not
covered pertains to links between speech and social evaluation.
Listeners engage in social processing during the course of speech
recognition, since processing indexical information is not inde-
pendent from processing of linguistic information (e.g., Mullenix
and Pisoni, 1990). An unfamiliar accent has been found to trigger
negative social evaluations in infants (Kinzler et al., 2007), chil-
dren (Kinzler et al., 2009, 2011), and young adults (Preston, 2003;
Bresnahan et al., 2002;Lev-Ari and Keysar, 2010; of course, accents
that differ from one’s native accent can also have positive or neutral
connotations, e.g., van Bezooijen, 2005). Even if the listener has
no previous experience with an accent, and therefore cannot eval-
uate it based on its social valence, speech streams that are harder
to process are not as well-liked (Alter and Oppenheimer, 2009).
Since accented speech is more difficult to process, it could be rou-
tinely disliked at an implicit level. In the confluence of this added
linguistic and social processing, accented speech must also differ
from non-accented speech in terms of the cognitive load imposed
on the listener,a factor that, at least theoretically, may be separable
from the complexity of accent marking on the signal (Mattys et al.,
2009). Furthermore, this cognitive load mayhave the greatest effect
on populations who have difficulty following explicit instructions
(e.g., infants and toddlers) and those with executive control issues
(e.g., a subset of the aging population), but it may affect all listen-
ers to a smaller or greater extent. A fuller discussion of the social
perception of accent across development awaits a separate review.
OPEN QUESTIONS
While the body of work on accent perception carried out in the
past 40 years is noteworthy and solid, there are many areas that
necessitate further exploration. Our review of the literature also
revealed two recurring “beliefs” not mentioned above, but which
are worthy of exploration because they permeate the literature.
Ultimately,the most relevant evidence to support these “beliefs”is
currently lacking, so we address them here in the hope of inspiring
the community to revisit them.
PROCESSING FOREIGN AND WITHIN-LANGUAGE ACCENTS IS
FUNDAMENTALLY DIFFERENT
In the Introduction, we merely stated that we would use “linguistic
variety” as an umbrella term. We viewed this umbrella as necessary
for both conceptual and empirical reasons. At a conceptual level, it
is impossible to draw stable, non-arbitrary boundaries between (1)
different languages; (2) different dialects of the same language; and
(3) non-native, dialectal, sociolectal accents. For example, among
linguists, it is often said that “a language is a dialect with an army
and a navy” (Magner, 1974). This phrase captures the fact that
variation occurs along a continuum, whereas hard boundaries
are derived from political, social, and historical reasons, rather
than true linguistic distance between two linguistic varieties. At
an empirical level, it is not rare to find two “languages” that are
closer to each other (in terms of mutual intelligibility and ease of
processing) than two “dialects” of the same language. One often
cited example is Dutch and German, intuitively conceived as two
languages in spite of the fact that they are fairly mutually intelligi-
ble; in contrast, Taishanese and Pinghua (“dialects” of Cantonese,
itself a “dialect”of Chinese) are not mutually intelligible.
Similarly, it is impossible to describe foreign accent as always
being more distinct from the listener’s native language than an
unfamiliar within-language accent. A common misconception is
that dialects differ only in sound instantiation, but not supraseg-
mentally, whereas two languages will be different both at the
segmental and the suprasegmental level. This is simply not the
case. For instance, Dutch is rhythmically more similar to Stan-
dard Southern British (SSB) than Glaswegian English (White
et al., 2012); therefore, it could be easier for SSB-learning infants
to segment words in Dutch or Dutch-accented English than in
Glaswegian English. Naturally, SSB adults will have a harder time
understanding a Dutch speaker than a fellow Glaswegian, given
the smaller lexical overlap with the former variety. In other words,
it is not always the case that dialects are closer to each other than
languages. Moreover, the degree to which processing an unfamiliar
within-language accent resembles processing an unfamiliar for-
eign accent at any given age is an empirical matter and probably
depends on the dimension of focus.
And yet, it is extremely common to find statements concluding
that non-native accents are fundamentally different (e.g., Major,
2007;Adank et al., 2009;Munro et al., 2010;Weber and Poellmann,
2010;Goslin et al., 2012). In one sense, the statement is trivially
true: people who are speaking a language that is not their native
one will almost certainly be under greater cognitive and motor
strain than those speaking their native language, which will cause
a host of paralinguistic acoustic changes (such as disfluencies and
increased variability in target production) above and beyond the
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 10

Cristia et al. Accent perception across the lifespan
“interesting” linguistic deviations. But in another sense, the state-
ment is simply misleading. We would like to invite the research
community to remove from the table the arbitrary distinctions
language/dialect/foreign accent, and replacethem by (psycho) lin-
guistic distance metrics. These metrics should consider not only
rhythmic, intonational, phonetic, phonological, and lexical simi-
larity between the specific speech samples being used; but also the
additional cognitive and/or social costs brought about by those
changes, and which are dependent on certain characteristics of the
listener group under study, ranging from degree of experience with
the variety to social values associated with it. Naturally, this ren-
ders our descriptive task considerably more difficult, since these
distance metrics will necessarily vary across ages, linguistic back-
grounds, and possibly even tasks. Nonetheless, we will hopefully
gain in explanatory power, as those distance metrics will be avail-
able to listeners either as cognitive constructs, or as ready-made
distinctions on the signal, whereas the arbitrary “accent” bins are
likely not. A first step along this direction could make use of the
masses of knowledge already accumulated, regressing infant,child,
and adult perceptual measures on a host of possible measures cal-
culated from the actual stimuli used. Once candidate measures
are thus identified, they can be more neatly investigated in ad hoc
experimental studies.
The question of how to draw lines between linguistic varieties
is relevant for another line of research. It has been repeatedly
reported that bilingual speakers develop more flexible cognitive
and linguistic systems (Kovacs and Mehler, 2009;Bialystok, 2010;
Sebastián-Gallés, 2010). If the line between accents, dialects, and
languages is difficult to draw,does this mean that bi-accentual/bi-
dialectal children will also experience similar cognitive gains? The
degree of similarity in cognitive gains exhibited by bilingual lis-
teners and those with exposure to two within-language accents is
likely dependent on the similarity in their life experiences. That is,
children who are routinely exposed to two variants, but grow up
to speak only one variant may experience lower internal conflict
and require lower strength of inhibition than those who come to
speak two variants. This becomes particularly relevant when con-
sidering an infant who is trying to learn the sound system of the
language(s) she hears, and may not know whether she is hearing
one or multiple varieties of the same or different languages (for
recent summaries on infant bilingual acquisition, see Sebastián-
Gallés, 2010;Werker, 2012). For example, one may argue that
it should be more difficult to tease apart Spanish from Catalan,
which are very similar at the phonological level, than native Eng-
lish from a heavily French-accented English, since French differs
from English even at the rhythmic level. Thus, perhaps the degree
of perceptual difference between the varieties might also be a factor,
even when the child grows up to speak only one variety. Although
we cannot attempt it here, we hope a future review will explore
similarities and differences between bi-accentual/bi-dialectal and
bilingual acquisition.
PROCESSING DIFFERENT ACCENTS AND DIFFERENT VOICES IS
FUNDAMENTALLY THE SAME
Even encountering novel talkers within one’s own accent group
presents the perception system with massive inter-speaker varia-
tion, which has a processing cost. Listening to speech by multiple
talkers as compared to one talker results in slower reaction times
and disrupted accuracy on many tasks, a phenomenon that has
been called the talker interference effect (Creel and Bregman,
2011). For example, listeners are slower to respond in a word
monitoring task when there are multiple talkers than when there
is only one talker (e.g., Nusbaum and Morin, 1992 for words;
Assmann et al., 1981 for vowels), and this slowing is affected
by working memory load (Nusbaum and Morin, 1992). Like-
wise, when given a set of utterances, listeners are slower and
less accurate at naming a word spoken in noise if the utterances
are spoken by a mix of talkers instead of one talker (e.g., Creel-
man, 1957;Mullennix et al., 1989;Sommers et al., 1994). Finally,
listeners recall fewer words from a list spoken by multiple talk-
ers as compared to a list spoken by one talker (Martin et al.,
1989, but see Goldinger et al., 1991 and Nygaard et al., 1994
for evidence that inter-stimulus-interval modulates this effect).
To a certain extent, the talker interference effect is due to top-
down biases, since it emerges when the listeners expect to hear
two voices, even if the signal from “both voices” is acousti-
cally identical (Magnuson and Nusbaum, 2007, using synthetic
speech).
Behavioral measures of multi-talker speech perception reveal
not only a processing cost, but also phenomena that can be com-
pared to the short-term perceptual adaptation effects and long-
term effects of exposure noted above. Akin to accent adaptation,
repeated exposure to a given talker aids speech processing in a
variety of tasks. First, word recognition under difficult processing
conditions or in the presence of noise is enhanced when the lis-
tener has some experience with the talker (Nygaard et al., 1994;
Nygaard and Pisoni, 1998). Recognizing spoken-words in noise is
facilitated even when the familiarization phase only involves lip-
reading (Rosenblum et al., 2007). Second,voice familiarity appears
to increase memory for words or sentences,though results vary for
different task types (see Luce and Lyons, 1998;Goh, 2005, for dis-
cussions). Recognition memory for words is more accurate when
the voice is the same at exposure and at test, and this same voice
priming can last for up to a week (Goldinger, 1996). Similarly, with
a continuous recognition memory paradigm, listeners are better
at recognizing whether a given word was previously heard if the
second presentation of the word is in the same voice as the first
presentation rather than in a different voice (Palmeri et al., 1993;
Bradlow et al., 1995; see also Martin et al., 1989;Goldinger et al.,
1991 for similar effects of talker variability/consistency on serial
recall). The effects of voice consistency appear to be more robust in
explicit memory tasks than in implicit memory tasks (Goldinger,
1996;Luce and Lyons, 1998; but see Schacter and Church, 1992
and Church and Schacter, 1994 for evidence of voice effects in
implicit tasks).
Additionally, in many perceptual adaptation paradigms, listen-
ers are exposed to a single talker with a quirky pronunciation,
and tested on the same voice used in the exposure phase. There-
fore, unless follow-up experiments are carried out, it usually is
impossible to be sure that results reflect adaptation to very spe-
cific features, or rather more general adaptation processes. In fact,
when the question has been addressed, it has been found that the
level of specificity varies across studies. Indeed, in some studies,
listeners interpret the remapping as being situation-specific (e.g.,
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 11

Cristia et al. Accent perception across the lifespan
the talker had a pen in her mouth, hence she does not normally
talk funny; Kraljic et al., 2008b); in others, listeners interpret it
as being talker-specific (a quirky /s/ in the speaker’s non-native
language is expected to remain quirky in the speaker’s native
language; Reinisch et al., 2012); and yet other studies document
talker-general patterns of adaptation (retuning to stop voicing, but
not fricative place, is generalized across talkers; Kraljic and Samuel,
2007).
Thus, based on the similarity of behavioral evidence of ini-
tial processing costs and perceptual adaptation effects, one could
posit the hypothesis that the same underlying mechanisms gov-
ern processing of talker and accent variation (e.g., Nygaard and
Pisoni, 1998). In some cases, the parallel between voice and
accent normalization is transformed into an explicit assumption
of how speech processing should proceed. In most infant/toddler
accent perception studies, it is assumed that children should
come to ignore talker accent just as they ignore voice (see e.g.,
Best et al., 2009). For example, a similar developmental time-
line has been posited for the two in word segmentation tasks
(see Schmale and Seidl, 2009;Schmale et al., 2011, for further
discussion).
Although the parallels between processing talker and accent
variation are remarkable, further work is needed before conclud-
ing that this stems from their involving the same mechanisms.
To begin with, some of the changes imposed on the signal by a
talker’s voice are due entirely to their physical properties, which
are automatically undone by our auditory system with little need
for linguistic representations (see e.g., Lotto et al., 1997, for a sum-
mary of animal research on talker normalization). It is unlikely
that differences between linguistic varieties can be undone through
such universal and innate mechanisms. Additionally, while all chil-
dren have some exposure to talkers with different voices, not all
children have exposure to multiple accents. Thus, infants have pos-
itive evidence of the kinds of additional transformations that are
required to deal with multiple talkers, but may not have developed
robust remapping mechanisms for different accents. Thus, it is an
empirical matter as to what extent the mechanisms recruited, at
any given age and for a given task, are overlapping for talker and
accent variation.
CONCLUDING REMARKS
This review reveals some points of convergence of research on
accent perception across the lifespan. Throughout the lifespan,
online measures have provided evidence that an accent can ini-
tially impair linguistic processing, but further experience allows for
rapid adaptation. Admittedly, obtaining a full picture of the devel-
opment of accented speech perception from infancy to adulthood
is impossible at present, especially given the major methodolog-
ical and theoretical differences that exist across research with
infants, children, and adults. In this quest, it will be necessary to
develop appropriately controlled stimuli, and to establish which
behavioral and brain measures are comparable across popula-
tions. Ultimately, it would benefit researchers to employ com-
parable tasks that can be implemented across the lifespan. This
type of methodological innovation would allow researchers to
more reliably identify specific developmental changes in accent
perception.
The second set of roadblocks can be argued to relate to theo-
retical factors. First, it is likely impossible,and arguably unnatural,
to design tasks which isolate a single dimension of interest, such
as the effect of linguistic deviations while controlling for social
and cognitive effects. Additionally, when assessing populations as
different as infants and adults, we simply cannot assume that the
linguistic system is organized in the same way. For example, infants
may not parse speech into phone-sized categories that are limited
to the native inventory, assigning them symbolic labels. Certainly,
these theoretical challenges complicate the interpretation of find-
ings within the realm of infant and child research, and render it
more difficult to draw inferences from research carried out at many
ages across the lifespan.
Nonetheless, individuals in all age groups grapple with accented
speech. Therefore, research on accented speech perception makes
a unique contribution to our understanding of ecologically valid
language processing. Over this backdrop, the contribution of the
present article relied on the comparison of research carried out
at different points of the lifespan. This comparison both uncov-
ered the aspects of linguistic processing that are common to all
human perceivers and underlined which aspects can vary across
individuals and populations.
REFERENCES
Adank, P., Evans, B., Stuart-Smith, J.,
and Scotti, S. (2009). Comprehen-
sion of familiar and unfamiliar
native accents under adverse listen-
ing conditions. J. Exp. Psychol. Hum.
Percept. Perform. 35, 520–529.
Adank, P., Hagoort, P., and Bekkering,
H. (2010). Imitation improves lan-
guage comprehension. Psychol. Sci.
21, 1903–1909.
Adank,P.,and Janse,E. (2010). Compre-
hension of a novel accent by young
and older listeners. Psychol. Aging
25, 736–740.
Adank, P., and McQueen, J. M. (2007).
“The effect of an unfamiliar regional
accent on the speed of word pro-
cessing,” in Proceedings of the XVIth
International Congress of Phonetic
Sciences, Saarbrücken, 1925–1928.
Alter, A. L., and Oppenheimer, D. M.
(2009). Uniting the tribes of fluency
to form a metacognitive nation. Pers.
Soc. Psychol. Rev. 13, 219–235.
Assmann, P. F., Nearey, T. M., and
Hogan, J. T. (1981). Vowel
identification: orthographic,
perceptual, and acoustic aspects. J.
Acoust. Soc. Am. 71, 975–989.
Bertelson, P., Vroomen, J., and de
Gelder, B. (2003). Visual recalibra-
tion of auditory speech identifica-
tion: a McGurk after effect. Psychol.
Sci. 14, 592–597.
Best, C. T., Tyler, M. D., Good-
ing, T. N., Orlando, C. B., and
Quann, C. A. (2009). Development
of phonological constancy: Toddlers’
perception of native- and Jamaican-
accented words. Psychol. Sci. 20,
539–542.
Bialystok, E. (2010). Global-local
and trail-making tasks by mono-
lingual and bilingual children:
beyond inhibition. Dev. Psychol. 46,
93–105.
Bradlow, A. R., and Bent, T. (2008).
Perceptual adaptation to non-native
speech. Cognition 106, 707–729.
Bradlow, A. R., Nygaard, L. C., and
Pisoni, D. B. (1995). “On the con-
tribution of instance-specific char-
acteristics to speech perception,” in
Levels in Speech Communication:
Relations and Interactions, eds C.
Sorin, J. Mariani, H. Meloni, and J.
Schoentagen (Amsterdam: Elsevier),
13–25.
Bresnahan, M., Ohashi, R., Nebashi,
R., Liu, W., and Shearman, S.
(2002). Attitudinal and affective
response toward accented English.
Lang. Commun. 22, 171–185.
Brunellière, A., Dufour, S., and Nguyen,
N. (2011). Regional differences in
the listener’s phonemic inventory
affect semantic processing: a mis-
match negativity (MMN) study.
Brain Lang. 117, 45–51.
Brunellière, A., Dufour, S., Nguyen,
N., and Frauenfelder, U. H. (2009).
Behavioral and electrophysiological
evidence for the impact of regional
variation on phoneme perception.
Cognition 111, 390–396.
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 12

Cristia et al. Accent perception across the lifespan
Burda, A. N., Scherz, J. A., Hageman, C.
F., and Edwards, H. T. (2003). Age
and understanding speakers with
Spanish or Taiwanese accents. Per-
cept. Mot. Skills 97, 11–20.
Bürki-Cohen, J., Miller, J. L., and
Eimas, P. D. (2001). Perceiving non-
native speech. Lang. Speech 44,
149–169.
Butler, J., Floccia, C., Goslin, J., and
Panneton, R. (2011). Infants’ dis-
crimination of familiar and unfa-
miliar accents in speech. Infancy 16,
392–417.
Chambers, J. K. (2002). Dynamics of
dialect convergence. J. Socioling. 6,
117–130.
Church, B., and Schacter, D. L. (1994).
Perceptual specificity of auditory
priming: memory for voice intona-
tion and fundamental frequency. J.
Exp. Psychol. Learn. Mem. Cogn. 20,
521–533.
Clarke, C. M., and Garrett, M. F.
(2004). Rapid adaptation to foreign-
accented English. J. Acoust. Soc. Am.
116, 3647–3658.
Clopper, C. G., and Bradlow, A.
(2008). Perception of dialect vari-
ation in noise intelligibility and
classification. Lang. Speech 51,
175–198.
Clopper, C. G., and Pisoni, D. B.
(2004a). Homebodies and army
brats: some effects of early linguis-
tic experience and residential history
on dialect categorization. Lang. Var.
Change 16, 31–48.
Clopper, C. G., and Pisoni, D. B.
(2004b). Some acoustic cues for the
perceptual categorization of Amer-
ican English regional dialects. J.
Phon. 32, 111–140.
Creel, S. C., and Bregman, M. R. (2011).
How talker identity relates to lan-
guage processing. Lang. Linguist.
Compass 5, 190–204.
Creelman, C. D. (1957). The case of
unknown talker. J. Acoust. Soc. Am.
29, 655.
Cutler, A., McQueen, J. M., Butter-
field, S., and Norris, D. (2008).
“Prelexically-driven perceptual
retuning of phoneme boundaries,”
in Proceedings of Interspeech 2008,
Brisbane, Australia, article 2056, eds
J. Fletcher, D. Loakes, M. Wagner,
and R. Goecke (Adelaide: Causal
Productions). [CD-ROM].
Dahan, D., Drucker, S. J., and Scarbor-
ough, R. A. (2008). Talker adapta-
tion in speech perception: adjust-
ing the signal or the representations?
Cognition 108, 710–718.
DeKeyser, R. (2000). The robustness
of critical period effects in second
language acquisition. Stud. Second
Lang. Acquisition 22, 499–533.
Dufour, S., Brunellière, A., and Nguyen,
N. (in press). To what extent do
we hear phonemic contrasts in a
non-native regional variety? Track-
ing the dynamics of perceptual pro-
cessing with EEG. J. Psycholinguist.
Res. doi:10.1007/s10936-012-9212-8
Dufour, S., Nguyen, N., and Frauen-
felder, U. H. (2007). The perception
of phonemic contrasts in a non-
native dialect. J. Acoust. Soc. Am.121,
EL131–EL136.
Dufour, S., Nguyen, N., and Frauen-
felder, U. H. (2010). Does training
on a phonemic contrast absent in
the listener’s dialect influence word
recognition? J. Acoust. Soc. Am. 128,
EL43–EL48.
Evans,B. G., and Iverson, P. (2007). Plas-
ticity in vowel perception and pro-
duction: a study of accent change in
young adults. J. Acoust. Soc. Am. 121,
3814–3826.
Ferguson, S. H., Jongman, A., Sereno, J.
A., and Keum, K. A. (2010). Intel-
ligibility of foreign-accented speech
for older adults with and without
hearing loss. J. Am. Acad. Audiol. 21,
153–162.
Flege, J. E. (1999). “Age of learning and
second language speech,” in Second
Language Acquisition and the Critical
Period Hypothesis, ed. D. Birdsong
(Mahwah, NJ: Laurence Erlbaum),
101–132.
Flege, J.E. (2003). “Assessingconstraints
on second-language segmental pro-
duction and perception,” in Phonet-
ics and Phonology in Language Com-
prehension and Production, Differ-
ences and Similarities, eds A. Meyer
and N. Schiller (Berlin: Mouton de
Gruyter), 319–355.
Floccia, C., Butler, J., Girard, F., and
Goslin, J. (2009). Categorization of
regional and foreign accent in 5-
to 7-year-old British children. Int. J.
Behav. Dev. 33, 366–375.
Floccia, C., Delle Luche, C., Dur-
rant, S., Butler, J., and Goslin,
J. (2012). Parent or community?
Where do 20-month-olds exposed
to two accents acquire their repre-
sentation of words? Cognition 124,
95–100.
Floccia, C., Goslin, J., Girard, F., and
Konopczynski, G. (2006). Does a
regional accent perturb speech pro-
cessing? J. Exp. Psychol. Hum. Per-
cept. Perform. 32, 1276–1293.
Gass, S., and Varonis, E. (1984). The
effect of familiarity on the com-
prehensibility of nonnative speech.
Lang. Learn. 34, 66–85.
Girard, F., Floccia, C., and Goslin, J.
(2008). Perception and awareness of
accents in young children. Br. J. Dev.
Psychol. 26, 409–433.
Goggin, J. P., Thompson, C. P., Strube,
G., and Simental, L. R. (1991).
The role of language familiarity in
voice identification. Mem. Cognit.
19, 448–458.
Goh, W. D. (2005). Talker variability
and recognition memory: instance-
specific and voice-specific effects. J.
Exp. Psychol. Learn. Mem. Cogn. 31,
40–53.
Goldinger, S. D. (1996). Words and
voices: episodic traces in spoken
word identification and recogni-
tion memory. J. Exp. Psychol. Learn.
Mem. Cogn. 22, 1166–1183.
Goldinger, S. D., Pisoni, D. B., and
Logan, J. S. (1991). On the nature
of talker variability effects on recall
of spoken word lists. J. Exp.
Psychol. Learn. Mem. Cogn. 17,
152–162.
Golomb, J. D., Peelle, J. E., and Wing-
field, A. (2007). Effects of stim-
ulus variability and adult aging
on adaptation to time-compressed
speech. J. Acoust. Soc. Am. 121,
1701–1708.
Gordon-Salant, S., Yeni-Komshian, G.
H., and Fitzgibbons, P. J. (2010a).
Recognition of accented English in
quiet by younger normal-hearing
listeners and older listeners with
normal-hearing and hearing loss. J.
Acoust. Soc. Am. 128, 444–455.
Gordon-Salant, S., Yeni-Komshian, G.
H., and Fitzgibbons, P. J. (2010b).
Recognition of accented English in
quiet and noise by younger and older
listeners. J. Acoust. Soc. Am. 128,
444–455.
Gordon-Salant, S., Yeni-Komshian, G.
H., Fitzgibbons, P. J., and Schur-
man, J. (2010). Short-term adapta-
tion to accented English by younger
and older adults. J. Acoust. Soc. Am.
128, 200–204.
Goslin, J., Duffy, H., and Floccia, C.
(2012). An ERP investigation of
regional and foreign accent process-
ing. Brain Lang. 122, 92–102.
Hallé, P., and de Boysson-Bardies, B.
(1996). The format of representation
of recognized words in infants’ early
receptive lexicon. Infant Behav. Dev.
19, 465–483.
Hay, J., and Drager, K. (2010). Stuffed
toys and speech perception. Linguis-
tics 48, 865–892.
Holtby, A. K. (2010). The Effect of Age
and Exposure in the Development of
L2 Accent Perception (PMC Working
Paper series, 10-01). Edmonton, CA:
Prairie Metropolis Centre.
Houston, D. M., and Jusczyk, P. W.
(2000). The role of talker-specific
information in word segmentation
by infants. J. Exp. Psychol. Hum.
Percept. Perform. 26, 1570–1582.
Houston, D.M., Jusczyk, P.W., Kuijpers,
C., Coolen, R., and Cutler,A. (2000).
Both Dutch- and English-learning 9-
month-olds segment Dutch words
from fluent speech. Psychon. Bull.
Rev. 7, 504–509.
Impe, L., Geeraerts, D., and Speelman,
D. (2008). Mutual intelligibility of
standard and regional Dutch lan-
guage varieties. Int. J. Humanit. Arts
Comput. 2, 101–117.
Jacewicz, E., Allen Fox, R., and Salmons,
J. (2010). Regional dialect variation
in the vowel systems of typically
developing children. J. Speech Lang.
Hear. Res. 54, 448–470.
Janse, E. (2009). Processing of fast
speech by elderly listeners. J. Acoust.
Soc. Am. 125, 2361–2373.
Janse, E., and Adank, P. (2012). Pre-
dicting foreign-accent adaptation in
older adults. Q. J. Exp. Psychol. 65,
1563–1585.
Jesse, A., and Janse, E. (2012). Audiovi-
sual benefit for recognition of speech
presented with single-talker noise in
older listeners. Lang. Cogn. Process.
27, 1167–1191.
Jusczyk, P. W., and Aslin, R. N. (1995).
Infants’ detection of the sound pat-
terns of words in fluent speech.
Cogn. Psychol. 29, 1–23.
Kalikow, D. N., Stevens, K. N., and
Elliott, L. L. (1977). Development of
a test of speech intelligibility in noise
using sentence materials with con-
trolled word predictability. J. Acoust.
Soc. Am. 61, 1337–1351.
Kendall, T., and Fridland, V. (2012).
Variation in perception and produc-
tion of mid front vowels in the U.S.
Southern Vowel Shift. J. Phon. 40,
289–306.
Kinzler, K. D., Corriveau, K. H.,
and Harris, P. L. (2011). Chil-
dren’s selective trust in native-
accented speakers. Dev. Sci. 14,
106–111.
Kinzler, K. D., Dupoux, E., and Spelke,
E. S. (2007). The native language
of social cognition. Proc. Natl. Acad.
Sci. U.S.A. 104, 12577–12580.
Kinzler, K. D., Shutts, K., DeJesus, J.,
and Spelke, E. S. (2009). Accent
trumps race in guiding children’s
social preferences. Soc. Cogn. 27,
623–634.
Kitamura, C., Panneton, R., Deihl, M.,
and Notley, A. (2006). “Attuning to
the native dialect: when more means
less,” in Proceedings of the 11th Aus-
tralasian International Conference on
Speech Science & Technology, eds P.
Warren and C. I. Watson (Auck-
land: Australian Speech Science &
Technology Association), 124–129.
Kovacs, A. M., and Mehler, J. (2009).
Cognitive gains in 7-month-old
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 13

Cristia et al. Accent perception across the lifespan
bilingual infants. Proc. Natl. Acad.
Sci. U.S.A. 106, 6556–6560.
Kraljic, T., Brennan, S., and Samuel, A.
G. (2008a). Accommodating varia-
tion: dialects, idiolects, and speech
processing. Cognition 107, 51–81.
Kraljic, T., Samuel,A. G., and Brennan,
S. E. (2008b). First impressions and
last resorts: how listeners adjust to
speaker variability. Psychol. Sci. 19,
332–338.
Kraljic, T., and Samuel, A. G. (2007).
Perceptual adjustments to multiple
speakers. J. Mem. Lang. 56, 1–15.
Labov, W. (1989). The child as a lin-
guistic historian. Lang. Var. Change
1, 85–97.
Labov, W. (2007). Transmission and
diffusion. Language 83, 344–387.
Labov, W., and Ash, S. (1997). “Under-
standing Birmingham,” in Language
Variety in the South Revisited, eds
C. Bernstein, T. Nunnally, and R.
Sabino (Tuscaloosa: University of
Alabama Press), 508–573.
Lev-Ari, S., and Keysar, B. (2010). Why
don’t we believe non-native speak-
ers? The influence of accent on
credibility. J. Exp. Soc. Psychol. 46,
1093–1096.
Lotto, A. J., Kluender, K. R., and Holt, L.
L. (1997). “Animal models of speech
perception phenomena,” in Chicago
Linguistic Society, Vol. 33, eds K.
Singer, R. Eggert, and G. Ander-
son (Chicago, IL: Chicago Linguistic
Society), 357–367.
Luce, P. A., and Lyons, E. A. (1998).
Specificity of memory representa-
tions for spoken words. Mem. Cog-
nit. 26, 708–715.
Magner, T. F. (1974). The study of for-
eign languages in China. Mod. Lang.
J. 58, 384–391.
Magnuson, J. S., and Nusbaum, H.
C. (2007). Acoustic differences, lis-
tener expectations, and the percep-
tual accommodation of talker vari-
ability. J. Exp. Psychol. Hum. Percept.
Perform. 33, 391–409.
Major, R. C. (2007). Identifying a for-
eign accent in an unfamiliar lan-
guage. Stud. Second Lang. Acquisition
29, 539–556.
Martin, C. S., Mullennix, J. W., Pisoni,
D. B., and Summers, W. V. (1989).
Effects of talker variability on recall
of spoken word lists. J. Exp. Psychol.
Learn. Mem. Cogn. 15, 676–684.
Mason, H. M. (1946). Understandabil-
ity of speech in noise as affected
by region of origin of speaker and
listener. Speech Monogr. 13, 54–68.
Mattys, S. L., Brooks, J., and Cooke, M.
(2009). Recognizing speech under a
processing load: dissociating ener-
getic from informational factors.
Cogn. Psychol. 59, 203–243.
Maye, J., Aslin, R. N., and Tanenhaus,
M. K. (2008). The weckud wetch of
the wast: lexical adaptation to a novel
accent. Cogn. Sci. 32, 543–562.
McQueen,J. M., Tyler,M. D.,and Cutler,
A. (2012). Lexical retuning of chil-
dren’s speech perception: evidence
for knowledge about words’ com-
ponent sounds. Lang. Learn. Dev. 8,
317–339.
Mitterer, H., and McQueen, J. M.
(2009). Foreign subtitles help
but native-language subtitles
harm foreign speech percep-
tion. PLoS ONE 4, e7785.
doi:10.1371/journal.pone.0007785
Mulak, K. E., Best, C. T., Tyler, M.
D., Kitamura, C., and Irwin, J. (in
press). Development of phonolog-
ical constancy: 19-month-olds, but
not 15-month-olds, identify famil-
iar words spoken in a non-native
regional accent. Child Dev.
Mullenix, J. W., and Pisoni, D. B.
(1990). Stimulus variability and
processing dependencies in speech
perception. Percept. Psychophys. 47,
379–390.
Mullennix, J. W., Pisoni, D. B., and Mar-
tin, C. S. (1989). Some effects of
talker variability on spoken word
recognition. J. Acoust. Soc. Am. 85,
365–378.
Munro,M. J., and Derwing, T. G. (1995).
Processing time, accent and com-
prehensibility in the perception of
native and foreign-accented speech.
Lang. Speech 38, 289–306.
Munro, M. J., Derwing, T. M., and
Burgess, C. S. (2010). Detec-
tion of nonnative speaker status
from content-maskedspee ch.Speech
Commun. 52, 626–637.
Nathan, L., Wells, B., and Donlan, C.
(1998). Children’s comprehension
of unfamiliar regional accents: a pre-
liminary investigation. J. Child Lang.
25, 343–365.
Nazzi, T., Goyet, L., Sundara, M., and
Polka, L. (2012). Différenceslinguis-
tiques et dialectales dans la mise en
place des procédures de segmenta-
tion de la parole. Enfance 1,127–146.
Nazzi, T., Iakimova, G., Bertoncini,
J., Fredonie, S., and Alcantara, C.
(2006). Early segmentation of fluent
speech by infants acquiring French:
emerging evidence for crosslinguis-
tic differences. J. Mem. Lang. 54,
283–299.
Nazzi, T., Jusczyk, P. W.,and Johnson, E.
K. (2000). Language discrimination
by English-learning 5-month-olds.
J. Mem. Lang. 43, 1–19.
Niedzielski, N. (1999). The effect of
social information on the perception
of sociolinguistic variables. J. Lang.
Soc. Psychol. 18, 62–85.
Norris, D., McQueen, J. M., and Cut-
ler, A. (2003). Perceptual learn-
ing in speech. Cogn. Psychol. 47,
204–238.
Nusbaum, H. C., and Morin, T. M.
(1992). “Paying attention to differ-
ences among talkers,” in Speech Per-
ception, Speech Production, and Lin-
guistic Structure, eds Y. Tohkura, Y.
Sagisaka, and E. Vatikiotis-Bateson
(Tokyo: OHM), 113–134.
Nygaard, L. C., and Pisoni, D. B. (1998).
Talker-specific learning in speech
perception. Percept. Psychophys. 60,
355–376.
Nygaard, L. C., Sommers, M. S., and
Pisoni, D. B. (1994). Speech percep-
tion as a talker-contingent process.
Psychol. Sci. 5, 42–46.
Organization for Economic Cooper-
ation and Development. (2007).
International Migration Outlook.
Paris: OECD.
Palmeri, T. J., Goldinger, S. D., and
Pisoni, D.B. (1993). Episodic encod-
ing of voice attributes and recogni-
tion memory for spoken words. J.
Exp. Psychol. Learn. Mem. Cogn. 19,
309–328.
Park, D. C., and Reuter-Lorenz, P.
(2009). The adaptive brain: aging
and neurocognitive scaffolding.
Annu. Rev. Psychol. 60, 173–196.
Peelle, J. E., and Wingfield, A. (2005).
Dissociations in perceptual learn-
ing revealed by adult age differences
in adaptation to time-compressed
speech. J. Exp. Psychol. Hum. Percept.
Perform. 31, 1315–1330.
Perrachione, T. K., Pierrehumbert, J.
B., and Wong, C. M. (2009).
Differential neural contributions
to native- and foreign-language
talker identification. J. Exp. Psy-
chol. Hum. Percept. Perform. 35,
1950–1960.
Piske, T., MacKay, I. R. A., and Flege, J.
E. (2001). Factors affecting degree of
foreign accent in an L2: a review. J.
Phon. 29, 191–215.
Polka, L., and Sundara, M. (2012).
Word segmentation in monolingual
infants acquiring Canadian Eng-
lish and Canadian French: native
language, cross-dialect, and cross-
language comparisons. Infancy 17,
198–232.
Preston, D. R. (2003). “Language with
an attitude,” in Handbook of Lan-
guage Variation and Change, eds J.
K. Chambers, P. Trudgill, and N.
Schilling-Estes (Malden, MA: Black-
well Publishing), 39–66.
Reinisch, E., Weber, A., and Mitterer,
H. (2012). Listeners retune phoneme
categories across languages. J. Exp.
Psychol. Hum. Percept. Perform.
doi:10.1037/a0027979
Rieckmann,A., and Bäckman, L. (2009).
Implicit learning in aging: extant
patterns and new directions. Neu-
ropsychol. Rev. 19, 490–503.
Roberts, J. (1997a). Acquisition of vari-
able rules: a study of (-t, d) deletion
in preschool children. J. Child Lang.
24, 351–372.
Roberts, J. (1997b). Hitting a moving
target: acquisition of sound change
in progress by Philadelphia children.
Lang. Var. Change 9, 249–266.
Roberts, J. (2005). “Acquisition of
sociolinguistic variation,” in Clini-
cal Sociolinguistics, ed. M. J. Ball
(Malden, MA: Blackwell Publish-
ing), 153–164.
Roberts, J.,and Labov,W.(1995). Learn-
ing to talk Philadelphian: acqui-
sition of short a by preschool
children. Lang. Var. Change 7,
101–112.
Rogers, C. L., Dalby, J., and Nishi,
K. (2004). Effects of noise and
proficiency level on intelligibility
of Chinese-accented English. Lang.
Speech 47, 139–154.
Rosenblum, L. D., Miller, R. M., and
Sanchez, K. (2007). Lip-read me
now, hear me better later. Psychol.
Sci. 18, 392–396.
Rubin, D. L. (1992). Nonlanguage
factors affecting undergraduates’
judgments of non-native English
speaking teaching assistants. Res.
Higher Educ. 33, 511–531.
Samuel, A. G., and Kraljic, T. (2009).
Perceptual learning for speech.
Atten. Percept. Psychophys. 71,
1207–1218.
Schacter, D. L., and Church, B. (1992).
Auditory priming: implicit and
explicit memory for words and
voices. J. Exp. Psychol. Learn. Mem.
Cogn. 18, 915–930.
Schmale, R., Cristià, A., and Seidl,
A. (2012). Toddlers recog-
nize words in an unfamiliar
accent after brief exposure.
Dev. Sci. doi:10.1111/j.1467-
7687.2012.01175.x
Schmale, R., Cristià, A., Seidl, A., and
Johnson, E. K. (2010). Infants’ word
segmentation across dialects. Infancy
15, 650–662.
Schmale, R., Hollich, G., and Seidl,
A. (2011). Contending with foreign
accent in early word learning. J.
Child Lang. 38, 1096–1108.
Schmale, R., and Seidl, A. (2009).
Accommodating variability in voice
and foreign accent: flexibility of early
word representations. Dev. Sci. 12,
583–601.
Scott, D. R., and Cutler, A. (1984). Seg-
mental phonology and the percep-
tion of syntactic structure. J. Verb.
Learn. Verb. Behav. 23, 450–466.
Frontiers in Psychology | Cognition November 2012 | Volume 3 | Article 479 | 14

Cristia et al. Accent perception across the lifespan
Sebastián-Gallés, N. (2010). Bilingual
language acquisition: where does
the difference lie? Hum. Dev. 53,
245–255.
Shah, A. P., Schmidt, B. T., Goral, M.,
and Obler, L. K. (2005). “Age effects
in processing bilinguals’ accented
speech,” in ISB4: Proceedings of
the 4th International Symposium on
Bilingualism, eds J. Cohen, K. T.
McAlister, K. Rolstad, and J. Mac-
Swan (Somerville, MA: Cascadilla
Press), 2115–2121.
Singh, L. (2008). Influences of high and
low variability on infant word recog-
nition. Cognition 106, 833–870.
Singh, L., Morgan, J. L., and White, K.
S. (2004). Preference and processing:
the role of speech affect in early spo-
ken word recognition. J. Mem. Lang.
51, 173–189.
Skoruppa, K., and Peperkamp, S.
(2011). Adaptation to novel accents:
feature-based learning of context
sensitive phonological regularities.
Cogn. Sci. 35, 348–366.
Sommers, M. S. (1996). The struc-
tural organization of the men-
tal lexicon and its contribution
to age-related declines in spoken-
word recognition. Psychol. Aging 11,
233–341.
Sommers, M. S., Nygaard, L. C.,
and Pisoni, D. B. (1994). Stimulus
variability and spoken word recogni-
tion: I. Effects of variability in speak-
ing rate and overall amplitude. J.
Acoust. Soc. Am. 96, 1314–1324.
Sumner,M., and Samuel, A. (2009). The
effect of experience on the percep-
tion and representation of dialect
variants. J. Mem. Lang. 60, 487–501.
Swingley, D., and Aslin, R. N. (2000).
Spoken word recognition and lex-
ical representation in very young
children. Cognition 76, 147–166.
Tagliamonte, S. A., and Molfenter,
S. (2007). How’d you get that
accent? Acquiring a second dialect
of the same language. Lang. Soc. 36,
649–675.
United States Census Bureau.
(2008). The American Com-
munity Survey. Available at:
http://www.census.gov/acs/www/
van Bezooijen, R. (2005). Approxi-
mant/r/in Dutch: routes and feel-
ings. Speech Commun. 47, 15–31.
van Bezooijen, R., and Gooskens, C.
(1999). Identification of language
varieties: the contribution of dif-
ferent linguistic levels. J. Lang. Soc.
Psychol. 18, 31–48.
van Heugten, M., and Johnson, E. K.
(2012). Infants exposed to fluent
natural speech succeed at cross-
gender word recognition. J. Speech
Lang. Hear. Res. 55, 554–560.
van Linden, S., and Vroomen, J.
(2008). Audiovisual speech recali-
bration in children. J. Child Lang. 35,
809–822.
Weber, A., and Poellmann, K. (2010).
“Identifying foreign speakers with an
unfamiliar accent or in an unfamil-
iar language,” in New Sounds 2010:
Sixth International Symposium on
the Acquisition of Second Language
Speech (Poznan: Adam Mickiewicz
University), 536–541.
Wells, J. C. (1982). Accents of English:
An introduction. Melbourne: Cam-
bridge University Press.
Werker, J. F. (2012). Perceptual foun-
dations of bilingual acquisition in
infancy. Ann. N. Y. Acad. Sci. 1251,
50–61.
White, K. S., and Aslin, R. N. (2011).
Adaptation to novel accents by tod-
dlers. Dev. Sci. 14, 372–384.
White, L., Mattys, S. L., and Wiget, L.
(2012). Language categorization by
adults is based on sensitivity to dura-
tional cues, not rhythm class. J. Mem.
Lang. 66, 665–679.
Wingstedt, M., and Schulman, R.
(1987). “Comprehension of foreign
accents,” in Phonologica 1984: Pro-
ceedings of the fifth International
Phonology Meeting, ed. W. Dressler
(Cambridge: Cambridge University
Press), 339–345.
Winters,S. J.,Levi, S. V.,and Pisoni,D. B.
(2008). Identification and discrim-
ination of bilingual talkers across
languages. J. Acoust. Soc. Am. 123,
4524–4538.
Conflict of Interest Statement: The
authors declare that the research was
conducted in the absence of any com-
mercial or financial relationships that
could be construed as a potential con-
flict of interest.
Received: 31 May 2012; accepted: 17
October 2012; published online: 08
November 2012.
Citation: Cristia A, Seidl A, Vaughn C,
Schmale R, Bradlow A and Floccia C
(2012) Linguistic processing of accented
speech across the lifespan. Front. Psychol-
ogy 3:479. doi: 10.3389/fpsyg.2012.00479
This article was submitted to Frontiers
in Cognition, a specialty of Frontiers in
Psychology.
Copyright © 2012 Cristia, Seidl , Vaughn,
Schmale, Bradlow and Floccia . This is an
open-access article distributed under the
terms of the Creative Commons Attribu-
tion License, which permits use, distrib-
ution and reproduction in other forums,
provided the original authors and source
are credited and subject to any copy-
right notices concerning any third-party
graphics etc.
www.frontiersin.org November 2012 | Volume 3 | Article 479 | 15