You are on page 1of 16

Journal of Psycholinguistic Research (2021) 50:707–721

https://doi.org/10.1007/s10936-021-09774-3

Effects of Phonetic Training on the Discrimination of Second


Language Sounds by Learners with Naturalistic Access
to the Second Language

Georgios P. Georgiou1 

Accepted: 16 January 2021 / Published online: 29 January 2021


© The Author(s), under exclusive licence to Springer Science+Business Media, LLC part of Springer Nature 2021

Abstract
The present study investigates the effect of high variability phonetic training (HVPT) on
the discrimination of second language (L2) vowel contrasts by adult speakers who live in
a country  where the  L2 is dominant. The same subjects who participated in a previous
discrimination task were trained in the discrimination of four L2  vowel contrasts which
were relatively difficult for this population of learners. Both the post-test and the generali-
zation test showed significant improvement in the discrimination of most vowel contrasts
(both stressed and unstressed). The findings suggest that HVPT may facilitate the forma-
tion of robust L2 phonological representations even for learners who live and are educated
in an L2-dominant environment, dissolving in that way the perceptual confusions which
emerge from first language interference. Finally, important implications are made for the
implementation of HVPT in L2 classrooms.

Keywords  Vowel contrast discrimination · Greek as a second language · High variability


phonetic training

Introduction

Many studies indicate the difficulty of adult learners in acquiring native-like pronuncia-
tion in a second language (L2), which is evident from an easily-detectable foreign accent
in their speech (Georgiou 2019b; Piske, Mackay, and Flege 2001; Bongaerts 1999; Flege
1999a; Long 1990; Goto 1971). This difficulty derives from possible deficient percep-
tion of nonnative sounds when learners do not perceive much dissimilarity between a first
language (L1) sound and its closest L2 sound or two nonnative sounds (Georgiou 2019a;
Georgiou et al. 2020b; Flege 1995; Best 1995).
By contrast, children can learn an L2 (also pronunciation) identically to the native
speakers without much effort (Snow 1987). Further evidence is provided by studies which
underline the differences between learners who have been exposed in the L2 at an early age

* Georgios P. Georgiou
georgiou.georg@unic.ac.cy
1
Department of Languages and Literature, University of Nicosia, Nicosia, Cyprus

13
Vol.:(0123456789)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
708 Journal of Psycholinguistic Research (2021) 50:707–721

and learners who had contact with the L2 at a later stage of their life (e.g., adolescence or
adulthood) (Piske et al. 2002; Mackay et al. 2001; Flege et al. 1995b; Flege and MacKay
2004; Flege et al. 1999a; Yamada 1995). These studies highlight the adult–child differences
in the perception of L2 segments, concluding that children are more rapid and successful
learners. For example, Tsukada et al. (2005) after investigating English vowel perception
and production by Korean adults and children in North America, they found that children
could better discriminate English vowel contrasts than adults. They justified this inconsist-
ency by referring to differences in factors such as the amount of perceived L2 stimuli, the
age of arrival in the L2-dominant country, the L2 use, and other. Moreover, according to
Best and Tyler (2007), it is harder for adults to perceive the L2 sounds because of the great
amount of input they receive in their L1. Therefore, external factors (e.g., L1 experience)
rather than internal ones (e.g., brain plasticity) might explain why adults have more diffi-
culties in acquiring L2 pronunciation in comparison to children.

Phonological Acquisition Models

Two well-known models in the area of speech acquisition are the Speech Learning Model
(SLM) (Flege 1995) and the Perceptual Assimilation model-L2 (PAM-L2) (Best 1995).
These models aim to predict the difficulties of experienced learners of an L2 in the percep-
tion and production of the L2 sounds. Although they differ in their philosophical back-
ground (e.g., SLM claims that listeners create mental representations from speech input,
while PAM suggests that there is a direct perception of speech gestures from the speech
signal), both SLM and PAM support the lifelong learning of L2 sounds; that is, even adults
may be able to refine their native sound inventory as there is not any age barrier which pre-
vents the acquisition of the L2 sounds. In other words, there is not any biological process
related to age that can impede learners from acquiring the sounds of an L2 (Lenneberg
1967). So, according to SLM, new phonetic categories may be formed even in adults,
resulting in better perception and production of the L2 sounds. Similarly, PAM-L2 argues
that under certain circumstances, new phonological categories may be formed for nonna-
tive sounds that constituted bad exemplars of a native sound in the past.
Several sociolinguistic (e.g., age, length of residence in an L2-speaking country, L1-L2
use, etc.) (Lewandowski et al. 2009), maturational (Hyltenstam and Abrahamsson 2008),
psychological (Rota and Reiterer 2009), and idiosyncratic factors (e.g., self-esteem, risk-
taking, etc.) can affect the acquisition of an L2. Also, phonetic training is considered
important for the better discrimination of “difficult” L2 contrasts. Especially, high vari-
ability phonetic training (HVPT) is said to increase the learners’ perceptual skills by many
studies.
HVPT is a type of auditory training which aims at improving the perceptual patterns of
nonnative language learners mostly at the segmental level (Thomson 2018). These learners
are often late learners (i.e., they have late onset of language learning) whose phonological
systems ‘refuse’ to change after year-long exposure to the target language. HVPT relies
on the use of multiple voices instead of the traditional single one. That is, the stimuli of
the training sessions are provided to the participants by several talkers. Also, the stimuli
are embedded in multiple linguistic contexts. Learners listen to the target language stimuli
and they are called to identify the sound they hear, whereas immediate feedback is given
to them. In that way, they pay attention to the acoustic properties that are important to
distinguish these sounds and they ignore the acoustic properties that vary among different
speakers.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 709

Literature Review

Logan et al. (1991) investigated the effect of training on the discrimination of the English
/r/-/l/ contrast by Japanese speakers of English. During training, the trainees were involved
in a minimal-pair identification task in order to classify the L2 tokens to their L1 pho-
nological categories. Moreover, they were presented with English words which included
the /r/-/l/ sounds; these sounds were naturally produced in multiple phonetic environments.
There were 15 sessions and the stimuli were presented by different talkers (native speak-
ers of General American English) since the authors believed that by ignoring variability,
learners would focus on the phonetic details of the sounds. According to the results, the
trained group of listeners could better discriminate the contrast than the control group (who
did not receive training), while this perceptual ability was retained for many months after
training. Lively et  al. (1993) repeated the above study with the addition of a generaliza-
tion test. In the first experiment, participants were trained in the identification of the /r/-/l/
contrast which was embedded in English words found in several phonetic environments
(initial singleton, initial consonant clusters, and intervocalic positions) and produced by
multiple talkers. Significant differences were found between pre- and post-test and par-
ticipants were able to generalize the novel words that were presented by a familiar and a
nonfamiliar talker. In the second experiment, new subjects participated and English words,
which included the /r/-/l/ contrast, were presented by a single talker in five different pho-
netic contexts. The results indicated that listeners improved from pre- to post-test but they
could not generalize their knowledge to the tokens produced by the new talker. The study
suggests that variability in such training is an important aspect for the improvement of the
L2 learners’ perceptual abilities. Likewise, Bradlow et al. (1997) investigated the effect of
perceptual HVPT on the identification of the /r/-/l/ contrast. The study consisted of three
parts: a pre-test, training sessions, and a post-test, while 11 adult Japanese speakers of Eng-
lish living in Japan participated in it. During the pre-test, the subjects were presented with
English minimal pairs of the /r/-/l/ contrast, which were produced by a native speaker of
American English. The training procedure included 45 sessions and lasted 3–4 weeks. Par-
ticipants were presented with naturally produced exemplars of the /r/-/l/ contrast embedded
in English words; these exemplars were recorded by five American English speakers. The
two generalization tests included minimal pair identification tasks; in the first test, par-
ticipants were presented with novel words spoken by a new speaker and in the second test,
they were presented with novel words spoken by one of the previous talkers. The find-
ings showed that perceptual learning was improved after the completion of training since
significant differences emerged between pre- and post-test. Rato (2014) applied HVPT to
Portuguese learners of English as a foreign language to examine its effect on the discrimi-
nation of the English /i/-/ɪ/, /ε/-/æ/, and /u/-/ʊ/ contrasts. The participants were 34 adult
Portuguese undergraduate students who participated in a pre-test, post-test, and a delayed
post-test; all tests included both vowel identification and vowel contrast discrimination
tasks. The five training sessions included immediate trial-by-trial feedback and cumula-
tive feedback, which was provided at the end of the sessions. The results showed that the
perceptual skills of adult learners were significantly improved after their involvement in
training sessions. Furthermore, the delayed post-test indicated that this knowledge could be
retained two months after the completion of the perceptual training. Finally, a generaliza-
tion test revealed that at least two of the contrasts were robustly learnt. All these results
provide evidence that perceptual learning through HVPT can be achieved within a short
period of time.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
710 Journal of Psycholinguistic Research (2021) 50:707–721

Nishi and Kewley-Port (2007) examined the effect of training sets on the perception
of American English vowels by Japanese learners. The results showed that both the train-
ees and the control group (untrained) improved their vowel identification accuracy by
25% after training, they generalized improvement to untrained words, and they were able
to maintain improvement after three months. Inceoglu (2016) trained American English
native speakers on the perception and production of French nasal vowels. Participants were
divided into three groups: those who received audiovisual training, those who received
audio-only training, and those who did not receive any training; all participants took a pre-
and a post-test. The findings revealed that there were significant differences between the
two groups in both tests but there were no differences between the audiovisual and the
audio-only training. Shinohara and Iverson (2018) compared the effects of identification
and discrimination training on the learning of the /r/-/l/ contrast by Japanese learners of
English. All speakers completed pre-, mid- and post-tests in four different types of training:
identification, auditory discrimination, category discrimination, and /r/–/l/ production. The
results indicated that both the identification and the discrimination tests had similar effects
on the learning of the English consonantal contrast as long as high variability was included
in these types of training.
Most of the studies that employ HVPT for the acquisition of nonnative contrasts rely
on inexperienced learners who recently arrived in the L2-speaking country (e.g., Nishi
and Kewley-Port 2007), learners who learn the nonnative language through controlled
environments only (e.g., Bradlow et  al. 1997; Carlet 2019; Hazan et  al. 2005), or learn-
ers who were trained in entirely new contrasts in a language that do not have experience
with (Pruitt et al. 2006). Iverson et al. (2012) argued that HVPT would be unsuccessful for
learners who have experience in the L2 and live in a country where L2 is dominant since
they receive many array of phonetic experience compared to the input that they can receive
in some hours of laboratory training. The authors used HVPT to train adult experienced
(residents of England) and inexperienced (residents of France) French learners of English
in the acquisition of the English vowels. An important finding of the study was that both
groups managed to learn the English vowels in the same degree, concluding that not only
exposure in the L2 stimuli is responsible for the acquisition of the L2 sounds but also some
aspects of the focused attention to phonetic differences in an identification task (Iverson
et al. 2012).

This Study

The present study aims to investigate the effect of short-term HVPT on the discrimina-
tion of challenging L2 contrasts by adult Egyptian Arabic learners of Greek as an L2. The
selection of Arabic speakers has been made on the basis of the difficulties of this popu-
lation of learners in the discrimination of Greek vowel contrasts. Specifically, Georgiou
(2018) investigated the vowel identification and vowel contrast discrimination abilities of
adult Egyptian Arabic speakers who permanently live in Cyprus and are educated in Cyp-
riot schools. The 15 participants completed an identification task in order to categorize
their L2 vowels to the phonological categories of their L1. Subsequently, they completed
an AXB task regarding two Greek vowel contrasts in both stressed and unstressed condi-
tions (stressed /i/-/e/, stressed /o/-/u/, unstressed /i/-/e/, unstressed /o/-/u/). The results indi-
cated that learners could discriminate three L2 Greek vowel contrasts only in a moderate
degree. Therefore, they could not identify much dissimilarity between the vowels of the

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 711

pairs due to the effect of their L1 vowel system. Allophonic variation in Egyptian Arabic
vowels is usually transferred into the L2 and, thus, neighboring Greek vowel categories are
merged (vowel allophony does not occur in Greek). It is reminded that the Greek vowel
inventory includes five vowel qualities /i e a o u/ (Trudgill 2009; Arvaniti 2007) without
any phonemic length distinction; however, stress may enhance the duration of vowels
(Georgiou 2018). The Egyptian Arabic language consists of eight vowel qualities: the short
/i a u/ and the long /iː eː aː oː uː/ (Munro 1993). The short vowels [e] and [o] are claimed
by many studies (e.g., Norlin 1987) as being allophones of the vowel phonemes /i/ and /u/
respectively rather than separate phonemes. For a more detailed comparison between the
Cypriot Greek and Egyptian Arabic vowel systems see, Georgiou (2018).
So, considering the difficulties of Egyptian Arabic learners of Greek regarding the dis-
crimination of specific vowel contrasts, we aim to examine the extent to which phonetic
training improves the L2 learners’ discrimination abilities; we take into account two lan-
guages (Arabic and Greek) of which the interference was not investigated much. Another
novel aspect of the present study is the investigation of the training effect on L2 learners
who receive both naturalistic and classroom L2 input since the majority of the studies that
examine the effect of phonetic training focus on learners who learn an L2 mainly through
classroom controlled environments. Moreover, not enough studies investigated the effect of
phonetic training on stressed vs. unstressed sound contrasts.
In particular, the study seeks to answer the following questions: (a) to what extent is
phonetic training able to improve the discrimination of difficult L2 contrasts by experi-
enced learners who live in an L2-dominant country, (b) what is the impact of phonetic
training on vowel stress (e.g., are stressed contrasts discriminated better than the unstressed
ones or vice-versa after training?), (c) is phonetic training able to create robust phonologi-
cal categories for the L2 vowels after a period of time?
The current study employs a five-session HVPT for each participant which relies on
forced-choice tasks with feedback. A generalization test was applied a month after the
post-test in order to examine if the acquired discrimination ability would carry over to the
discrimination of new words. Concerning the findings of a vast number of studies which
support the positive effect of phonetic training on the discrimination of L2 vowel contrasts,
it is hypothesized that training will improve the discrimination of Greek vowel contrasts
by Egyptian Arabic learners of Greek. Thus, significant differences in the discrimination
scores between pre- and post-test are expected. Yet, it is unknown the impact of HVPT on
the discrimination of stressed versus unstressed contrasts since there is not enough litera-
ture which deals with the effect of perceptual phonetic training on stress. One important
conclusion which was found in the studies of Georgiou (2018) and Georgiou and Themis-
tocleous (2020) indicated that stress is crucial in vowel perception and production since
stressed Greek vowel contrasts were discriminated and produced with different accuracy
than the corresponding unstressed contrasts by Egyptian Arabic speakers.

Methodology

Participants

The same 15 female learners who participated in the discrimination task (henceforth pre-
test) in Georgiou’s (2018) study were trained in the discrimination of L2 Greek contrasts.
The participants also performed a post- and a generalization test. They were adult native

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
712 Journal of Psycholinguistic Research (2021) 50:707–721

speakers of Egyptian Arabic who have been living in Nicosia, Cyprus for 4–5 years. Their
age varied from 18–24  years and they have learnt Greek through schools and universi-
ties. Moreover, they started to learn Greek at the age of 15–20. They were all middle-class
school or university students who were using equally in their daily life both their native
language and Greek. According to their own reports, based on a questionnaire provided to
them, they were poorly speaking Greek (1.9/5) and their understanding competence was
moderate (3.2/5). Also, they stated that they had a healthy hearing. All participants were
paid for their attendance in both training and post-test sessions.
The control group consisted of 15 Cypriot Greek speakers (the same who participated in
Georgiou 2018) with an age range of 18–26 years. They were all students at universities in
Cyprus coming from moderate-income families. The participants said that they have never
faced any language or auditory disorder. The control group performed one session of AXB
discrimination task.

Stimuli

The stimuli of the training session consisted of four trials, AAB, ABB, BBA and BAA (A
is the first vowel of the contrast and B is the second one). Each trial included triads of dis-
syllabic pseudowords for each Greek vowel contrast (four vowel contrasts: stressed /i/-/e/,
unstressed /i/-e/, stressed /o/-/u/, unstressed /o/-/u/) in the frame of ˈsVsa, sVˈsa, ˈVsa, Vˈsa.
For instance, for the /i/-/e/ contrast, the AAB trial included the following triads in four
conditions: 1) /ˈsisa/-/ˈsisa/-/ˈsesa/, 2) /siˈsa/-/siˈsa/-/seˈsa/, 3) /ˈisa/-/ˈisa/-/ˈesa/ and 4) /iˈsa/-/
iˈsa/-/eˈsa/ (following Themistocleous 2017). The frame of the words was the same in every
triad; only vowel differed. Pseudowords were used to eliminate the chance of memorizing
the exact phonetic details of the sounds due to lexical knowledge (Stanovich 2000). The
triads for each L2 contrast were randomly presented through a set of headphones which
were connected to the laptop. The words were naturally produced by four native speakers
of Greek (two males and two females aged 20–32), and they were recorded at a sample of
44.1 kHz in a quiet room.
The stimuli of the post-test were identical to those of the pre-test (see Georgiou 2018)
and to those used in the training sessions (see above). However, in contrast to the training
session stimuli, distractor words were used (as in the pre-test) with the structure of ˈpVsa
and pVˈsa (16 distractors). The stimuli were recorded at a 44.1  kHz sample by an adult
female speaker of Greek (the same talker who recorded the pre-test stimuli).
The stimuli of the generalization test included the aforementioned four trial types for
each Greek vowel contrast in the four conditions. However, new Greek pseudowords in /
ˈCVCV/, /CVˈCV/, /ˈVCV/ and /VˈCV/ context were used and they were presented ran-
domly through the laptop loudspeakers. The words were recorded by a new native speaker
of Greek (female; 26 years old) in a sound-attenuated room. Table 1 shows the words that
were used in the generalization test.

Procedure

Training. The training took place a week after the participation of the L2 learners in
the discrimination task (pre-test) (see Georgiou 2018). It was conducted in a sound-
attenuated room and each participant was trained individually. The training focused on
the investigation of the learners’ vowel contrast discrimination skills using a computer-
assisted HVPT. Specifically, the HVPT was set up in TP-Version 1.0 (Rauber et  al.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 713

Table 1  The words used in the generalization test


Greek vowel ˈCVCV CVˈCV ˈVCV VˈCV
contrast

/i/-/e/ /ˈtino/—/ˈteno/ /ziˈno/—/zeˈno/ /ˈipa/—/ˈepa/ /iˈma/—/eˈma/


/o/-/u/ /ˈrota/—/ˈruta/ /poˈli/—/puˈli/ /ˈota/—/ˈuta/ /oˈra/—/uˈra/

2011) and included vowel contrast discrimination tasks. HVPT was preferred since
many studies (e.g., Iverson et  al. 2012) indicate its effectiveness on the improvement
of the learners’ perceptual skills. Specifically, it included multiple voices (four peo-
ple) and multiple phonetic contexts (stressed, unstressed, initial fricative /s/ context,
initial target vowel context). In this kind of training, learners were listening to a triad
of words through the loudspeaker of a PC and, similarly to AXB tests (A = first vowel,
B = third vowel, X = a vowel that was identical to either A or B) (Best et al. 2001), they
were asked to choose whether the middle vowel (X) was the same as the first (A) or the
third (B) one. Immediate triad by triad feedback followed either if the response was
correct or incorrect (following Thomson 2011) with an optional auditory repetition of
the triad. In the case of a correct response, a tick mark was flashing next to the correct
response, while in the case of an incorrect response, an X mark was appearing next
to the incorrect response. The learner was then advised by the trainer to click on the
“Replay” button to listen again to the triad. The duration of the phonetic training was
about two hours in five different sessions (five different days) for each participant (10 h
in total for each participant). The training sessions were completed within a period
of 1.5–2  weeks. Each participant was trained in a total number of 96 triads (4 con-
trasts × 2 conditions × 4 trial types × 3 repetitions).
Post-Test Several days after training, the participants were called to perform the
post-test. The procedure was the same as in the pre-test (see Georgiou 2018). The test
took place in a quiet room and the participants were tested individually. An AXB dis-
crimination test in a Praat script (Boersma and Weenink 2018) was provided to the
learners through the laptop and they had to choose whether the middle vowel they
heard from the loudspeaker was the same as the first or the third one. Tokens were
heard through a set of headphones and the volume was set at around 75 dB (108 items
in total; 4 contrasts × 2 conditions × 4 trial types × 3 repetitions + 16 distractor words).
The participants could have a five-minute break after the 36th and the 72nd item. The
inter-stimulus interval was one second and the inter-trial interval was five seconds;
participants had five seconds to choose the appropriate answer.
Generalization Test One month after the completion of the post-test, learners par-
ticipated in a generalization test. The purpose of this test was to investigate if learners
were able to generalize new words beyond just the words they were trained with (Bar-
riuso and Hayes-Harb 2018). To this purpose, it was followed the same procedure as in
the post-test. Nevertheless, the stimuli included new pseudo-words which represented
the Greek contrastive vowel pair /i/-/e/ and /o/-/u/ in both stressed and unstressed posi-
tions. The stimuli were heard through a set of headphones at 75  dB. There were 32
items in three repetitions (96 items in total). After the 32nd and the 64th item, partici-
pants could have a five-minute break. No more than five seconds could be spent on the
decision of the answer.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
714 Journal of Psycholinguistic Research (2021) 50:707–721

Table 2  Percentages (%) and Standard Deviations (SD) of Correct (C) and Incorrect (I) responses in the
discrimination of Greek vowel contrasts by the L2 learners in the pre- and the post-test
Vowel contrast Pre-test Post-test
C I SD C I SD

Str. /i/-/e/ 65 (263) 35 (142) 18.06 90 (363) 10 (42) 4.45


Str. /o/-/u/ 61 (246) 39 (159) 19.40 89 (359) 11 (46) 13.71
Unstr. /i/-/e/ 98 (396) 2 (9) 2.82 98 (397) 2 (8) 2.92
Unstr. /o/-/u/ 55 (223) 45 (182) 12.76 91 (367) 9 (38) 8.25

The number of responses are shown in the parenthesis. The pre-test results are taken from the study of
Georgiou (2018)

Table 3  Percentages and Vowel contrast C I SD


Standard Deviations (SD) of
Correct (C) and Incorrect (I)
Str. /i/-/e/ 100 (404) 0 (1) 1.03
responses in the discrimination
of Greek vowel contrasts by the Str. /o/-/u/ 100 (403) 0 (2) 1.03
control group Unstr. /i/-/e/ 99 (402) 1 (3) 2.24
Unstr. /o/-/u/ 100 (405) 0 (0) 0

The number of responses is shown in the parenthesis

Results

It was observed that most of the L2 contrasts could be better discriminated in the post-test
compared to the pre-test. Only the L2 unstressed contrast /i/-/e/ was discriminated with
near-ceiling accuracy both in the pre-test and the post-test (98%). The L2 stressed /i/-/e/
contrast received 90% correct responses, showing a great difference in relation to the pre-
test (pre-test: 65% correct, 35% incorrect responses). The L2 stressed /o/-/u/ contrast por-
trayed 89% correct responses; again, the difference from the pre-test was considerable (pre-
test: 61% correct responses). Moreover, the L2 unstressed vowel contrast /o/-/u/ received
91% correct responses and thus learners demonstrated substantial improvement from pre-
to post-test. Correct (C) and incorrect (I) percentages, numbers and standard deviations
(SD) of responses for both the pre- and the post-tests are shown in Table  2. The control
group, as expected, had excellent performance in the discrimination task as its overall dis-
crimination accuracy ranged from 99 to 100% (see Table 3).
Also, it was observed that the discrimination scores of the generalization test were
slightly lower than the scores of the post-test, but they were higher than the scores of the
pre-test. Specifically, the Greek stressed vowel contrast /i/-/e/ had 79% correct responses
and the stressed /o/-/u/ contrast had 82% correct responses. The unstressed contrast /i/-/e/
maintained its high discrimination score (it was also high both in the pre- and the post-test)
with 91% correct responses. Finally, the unstressed vowel contrast /o/-/u/ demonstrated
85% correct responses. Table 4 shows the percentages, standard deviations (SD), and the
number of correct (C) and incorrect (I) responses of the L2 learners in the generalization
test.
A two-way repeated measures ANOVA was conducted in R to detect the effect of test
and vowel contrast on the discrimination scores; Test (three levels: pre-test, post-test, gen-
eralization test) and Vowel Contrast (four levels: stressed /i/-/e/, stressed /o/-/u/, unstressed

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 715

Table 4  Percentages and Vowel contrast Generalization test


Standard Deviations (SD) of
Correct (C) and Incorrect (I) C I SD
responses in the discrimination
of Greek vowel contrasts by the Str. /i/-/e/ 79 (263) 21 (142) 13.12
L2 learners in the generalization
Str. /o/-/u/ 82 (246) 18 (159) 11.14
test
Unstr. /i/-/e/ 91 (396) 9 (9) 4.89
Unstr. /o/-/u/ 85 (223) 15 (182) 15.6

The number of responses is shown in the parenthesis

/i/-/e/, unstressed /o/-/u/) were the two within-subject factors since all subjects had under-
gone the two conditions. Discrimination Score was the dependent variable (% of cor-
rect responses).The analysis showed that there was a significant main effect of Test [F(1,
15) = 35.43, p < 0.05, η2 = 0.1], Vowel Contrast [F(3, 45) = 15.97, p < 0.05, η2 = 0.23] and
Test × Vowel Contrast interaction [F(3, 45) = 17.11, p < 0.05, η2 = 0.17]. In order to identify
specific differences across the tests and the vowel contrasts, further post-hoc tests were car-
ried out. The Bonferroni correction showed that there were significant differences for the
stressed /i/-/e/ between pre- and post-test (p < 0.05), between pre-test and generalization
tests (p < 0.05) and between post- and generalization tests (p < 0.05); this yields that the
stressed /i/-/e/ contrast could be better discriminated in the post- (M = 90%) than in the pre-
test (M = 65%) but worse in the generalization test (M = 79%) compared to the post-test.
For the unstressed /i/-/e/ contrast, there were no significant differences between the three
tests. Regarding the stressed /o/-/u/ contrast, significant differences between pre- and post-
test (p < 0.05) and between pre-test and generalization test (p < 0.05) were observed. So,
this contrast could be better discriminated in the post-test (M = 89%) and the generalization
test (M = 82%) than in the pre-test (M = 62%). Also, for the unstressed /o/-/u/ contrast, sig-
nificant differences were found between pre- and post-test and between pre-test (p < 0.05)
and generalization test (p < 0.05); thus, the contrast had better discriminability in the post-
test (M = 91%) and the generalization test (M = 85%) than in the pre-test (M = 55%).
Furthermore, the Bonferroni post-hoc test showed that there were significant differ-
ences between the unstressed /i/-/e/ and the stressed /i/-/e/ (p < 0.05), the unstressed /i/-
/e/ and the stressed /o/-/u/ (p < 0.05), and the unstressed /i/-/e/ and the unstressed /o/-/u/
(p < 0.05) in the pre-test. Therefore, there was better discrimination of the unstressed /i/-
/e/ contrast (M = 98%) in comparison to the other contrasts (M = 65%, 61%, 55%) which
were discriminated almost in the same manner. In the post-test, there were no significant
differences between any of the contrasts. In the generalization test, significant differences
were found only between the stressed /i/-/e/ and the unstressed /i/-/e/ contrast (p < 0.05);
the unstressed /i/-/e/ was more accurately discriminated (M = 90%) than the stressed /i/-/e/
contrast (M = 79%).
Independent sample t-tests were used for the comparison of the discrimination scores of
each vowel contrast in the pre-test, post-test, and generalization test with the discrimination
scores of the control group. Specifically, there were significant differences for the stressed
/i/-/e/ [t(14) = 2.76, p < 0.05), the stressed /o/-/u/ [t(14) = 3.42, p < 0.05) and the unstressed
/o/-/u/ [t(14) = 2.89, p < 0.05) between the scores of the experimental group in the pre-
test and the scores of the control group; these contrasts were discriminated less well by
the experimental group compared to the control group. Also, significant differences were
observed only for the stressed /o/-/u/ [t(14) = 2.83, p < 0.05) between the productions of the
experimental group in the post-test and the productions of the control group; this contrast

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
716 Journal of Psycholinguistic Research (2021) 50:707–721

was better discriminated by the control group. Finally, significant differences emerged for
the stressed /i/-/e/ [t(14) = 2.94, p < 0.05), the stressed /o/-/u/ [t(14) = 3.79, p < 0.05), and
the unstressed /o/-/u/ [t(14) = 3.1, p < 0.05) between the scores of the experimental group
in the generalization test and the scores of the control group; the discrimination accuracy
of the experimental group was lower than the discrimination accuracy of the control group.

Discussion

The initial hypothesis predicted that perceptual training would be beneficial for the better
discrimination of the Greek contrasts by the L2 learners. Specifically, significant differ-
ences for the L2 contrast discrimination scores between pre- and post-test were predicted.
According to the findings, there were more correct responses for all contrasts after the
implementation of phonetic training; significant differences were found for the discrimi-
nation scores of three out of four L2 contrasts between pre- and post-test, while learners
could discriminate three out of four L2 contrasts with the same accuracy as the control
group. The discrimination of the unstressed contrast /i/-/e/ was not differentiated much
after the intervention. Furthermore, the generalization test showed that at least three out of
four Greek vowel contrasts could be discriminated with the same accuracy as in the post-
test and with more accuracy than in the pre-test. However, the discrimination accuracy of
these three contrasts in the generalization test differed from the discrimination accuracy of
the control group.
In general, the computer-assisted phonetic training was proved effective for the bet-
ter discrimination of L2 Greek vowel contrasts by adult Egyptian Arabic learners of
Greek since there was a shift from pre- to post-test and learners could maintain rela-
tively high scores in the generalization test. This is compatible with a vast number of
studies which indicate the significance of such training in the improvement of the L2
learners’ perceptual patterns. An important fact is that even short-term HVPT (five
sessions) managed to bring positive results regarding the discrimination of L2 con-
trasts which seemed to be problematic for the L2 learners. HVPT changed the atten-
tional weights and made learners focus more on stimuli aspects that discriminate pho-
nological categories in different talkers and different phonetic environments without
considering the variability. In general, the discrimination scores in both the post- and
the generalization tests were high ranging from 79 to 98% (average 88.1%). The dis-
crimination scores of the post- and the generalization tests in this study were higher
compared to the scores found in other studies. For example, in the study of Iverson
et  al. (2012), the inexperienced French learners of English discriminated the English
vowels 77% on average in the post-test. Similarly, in the study of Kartushina and Frau-
enfelder (2013) the training group, which consisted of French monolingual speakers,
discriminated Danish vowel contrasts with less than 70%. For the justification of the
high discrimination scores of the L2 learners in this study, we have to take into consid-
eration the linguistic environment. In the aforementioned studies, learners were receiv-
ing nonnative input mainly through classroom environments or they did not have any
knowledge in the nonnative language. In this study, learners gained linguistic input
through classroom settings since they learnt Greek through schools/universities but
they also receive naturalistic input since they live in a country where the L2 is domi-
nant. So, we do not rule out that since learners also have naturalistic access to the L2
stimuli, they might further be assisted with additional acoustic cues which are still

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 717

unknown in the present study (i.e., there might be something else that helps learn-
ers pay attention to the acoustic cues of the vowel contrasts). Also, the results of the
study diverge from the hypothesis of Iverson et al. (2012) that HVPT might have very
little effect on experienced learners who live in a country where the L2 is exclusively
spoken, since HPTV had a very positive impact on the discrimination accuracy of the
learners in this study. Therefore, HVPT can still be successful for L2 learners who live
in an L2-speaking country and who may have reached a stage of fossilization in their
L2 learning progress.
In this study, although learners improved satisfactorily from pre- to post-test, they
were not able to reach a discrimination peak in the generalization test (i.e., to have
discrimination accuracy that is similar to that of the native speakers). For instance,
the correct discrimination percentages of the three “difficult” L2 contrasts varied from
89%-91% in the post-test, but in the generalization test these percentages ranged from
79 to 86%; note that the control group of Greek native speakers performed discrimina-
tion scores ranging from 99 to 100%. Similarly, some studies, such as those of Hattori
and Iverson (2008) and Heeren and Schouten (2008), indicate that learners could not
benefit a lot from the phonetic training with respect to the perception of segmental
contrasts.
The present study also aimed to investigate the effect of HVPT on vowel stress.
Training managed to improve the discrimination of the stressed /i/-/e/ contrast and bring
it close to the score of the unstressed one (which was not difficult for the L2 learners).
Also, the scores of the stressed and unstressed /o/-/u/ contrasts, which were only mod-
erately discriminated by the learners in the pre-test, were also equally improved in the
post-test. Nevertheless, the discrimination of the difficult stressed /i/-/e/ contrast differed
from the discrimination of the corresponding unstressed one in the generalization test.
This did not happen with the stressed vs. unstressed /o/-/u/ contrasts which were equally
discriminated in the generalization test. From the aforementioned results, two main con-
clusions can be drawn. First, HVPT helped learners also focus on stress-related cues
and improve in general their discrimination accuracies in both stressed and unstressed
vowel contrasts. Second, the effect of HVPT was not the same for the stressed and the
unstressed /i/-/e/ contrasts since their discrimination scores differed in the generaliza-
tion test. Thus, HVPT may have different effects in stressed vs. unstressed sound con-
trasts and, therefore, both sets of contrasts should be included in such training.
Another important aspect is that perceptual training could be beneficial for the per-
ception of the L2 contrasts even for adult L2 learners. They were able to refine their L1
phonological system because they better perceived the L2 vowels after the intervention.
This is compatible with the assumptions of both SLM and PAM-L2, which support that
there is a continuous refinement of the L1 phonological system and that L1 language
mechanisms remain intact over the lifespan being able to be adapted also in the L2
learning, correspondingly. However, it was crucial to investigate if these phonetic repre-
sentations would remain stable after a period of time. The generalization test confirmed
that the phonetic representations were relatively robust since discrimination accuracy
was still significantly higher than that of the pre-test at least a month after the com-
pletion of the training sessions. Some studies (e.g., Evans and Martin-Alvarez 2016)
underline that age and matureness correlate positively with the better effect of HVPT
on vowel perception; implying that adults benefit more than children regarding the dis-
crimination of segmental contrasts and the generalization of new knowledge through
HVPT. So, this could justify the rapid reweighting of acoustic cues by the adult learners
after phonetic training.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
718 Journal of Psycholinguistic Research (2021) 50:707–721

Teaching Implications

The results of the present study can provide feedback for educators of Greek as an L2 in
order to solve vowel contrast perceptual problems of Arabic learners. Accurate perception
and production of the unfamiliar sounds might enhance the learners’ competence in the
L2 and develop their communicative skills. Pennington (1998b) argues that phonetic train-
ing is beneficial for L2 learners if it is offered in isolation with other skills; so training
should be provided out of the aspects of vocabulary, grammar, or pragmatics helping learn-
ers focus on the phonological elements of challenging L2 contrasts. HVPT is an easy pro-
cedure which can be applied either into or outside the classroom helping learners develop
metalinguistic skills and increase their perceptual abilities by perceiving most of the acous-
tic invariants of the Greek vowels. Since there is a positive effect of HVPT on the vowel
perceptual patterns of Arabic learners of Greek, we encourage the use of this kind of train-
ing by educators as part of L2 learning. An extra motivation for the use of HVPT in the
instruction of the L2 sounds is its low cost and, thus, schools will not have to invest a lot of
money for this purpose. Also, the free software offered by Rauber et al. (2011), which was
also employed in the present study, does not demand explicit phonetic knowledge and it is
easy to be configured even by nonacademics. The specific program supports the creation of
both identification and discrimination tests (e.g., AXB, AX) as well as training perceptual
sessions with feedback. Another advantage of such a method is that it reduces significantly
the burden of the educator (Barriuso et al. 2018) since after the given instructions, learners
can use the training computer application on their own as automatic feedback is provided
to them according to their response.

Conclusions

The study showed that short-term HVPT can improve the discrimination of L2 sounds with
this improvement to be maintained to some extent after a period of time. The results sug-
gest that learners with naturalistic exposure to the L2 can also benefit from phonetic train-
ing; apart from learners who learn an L2 through classrooms as it has been shown in other
studies. Also, HVPT may have a different impact on stressed vs. unstressed contrasts and
thus stress should be taken into consideration. Finally, the study proposes that educators
can employ phonetic training for the instruction of pronunciation as it is a powerful and an
easy-to-use tool.

Limitations

It has to be taken into account that more training sessions have to occur in order for the L2
learners to have long-term changes regarding the representation of the L2 phonological
categories. Although the generalization test showed that the L2 learners could still discrim-
inate well the majority of the vowel contrasts, the discrimination accuracy of the contrasts
was slightly decreased in relation to the post-test and the control group discriminations. So,
it is unknown if the learners would discriminate the vowel contrasts with high accuracy
after months or years. Future studies could also examine the effect of Greek vowel contrast
discrimination training by Arabic learners who differ in age, amount of L1-L2 use, length

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 719

of residence, L2 vocabulary size (cf. Georgiou et al. 2020a) etc. or listeners who have indi-
vidual differences since some studies (e.g., Ingvalson et al. 2013) highlight that the potency
of HVPT relies on the perception of sensitivity that each individual listener has on the
trained contrast.

Compliance with Ethical Standards 

Conflict of interest  I have no conflicts of interest to declare.

Ethical Approval  The study’s protocol was approved by the Ethics Committee of the Department of Educa-
tion, University of Cyprus.

Informed Consent  The participants were informed on paper about the goals of the study and their rights.
They gave their written consent for participating to the experiments in accordance to the Declaration of
Helsinki.

References
Arvaniti, A. (2007). Greek phonetics: The state of the art. Journal of Greek Linguistics, 8, 97–208.
Barriuso, A. T., & Hayes-Harb, R. (2018). High variability phonetic training as a bridge from research to
practice. The CATESOL Journal, 30(1), 177–194.
Best, C. T. (1995). A direct realist view of cross-language speech perception: New directions in research and
theory. In W. Strange (Ed.), Speech perception and linguistic experience: Theoretical and methodo-
logical issues (pp. 171–204). Baltimore: York Press.
Best, C. T., McRoberts, G. W., & Goodell, E. (2001). Discrimination of non-native consonant contrasts
varying in perceptual assimilation to the listener’s native phonological system. Journal of the Acousti-
cal Society of America, 109(2), 775–794.
Best, C. T., & Tyler, M. (2007). Non-native and second-language speech perception: Commonalities and
complementarities. In O.-S. Bohn & M. J. Munro (Eds.), Second language speech learning: In honor
of James Emil Flege (pp. 13–34). Amsterdam/Philadelphia: John Benjamins.
Boersma, P., & Weenink, D. (2018). Praat: Doing phonetics by computer [Computer program]. Retrieved
from https://​www.​fon.​hum.​uva.​nl/​praat/.
Bongaerts, T. (1999). Ultimate attainment in L2 pronunciation: The case of very advanced late learners.
In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis (pp. 133–160).
Mawhaw: LEA.
Bradlow, A. R., Pisoni, D. B., Akahane-Yamada, R., & Tohkura, Y. (1997). Training Japanese listeners to
identify English /r/ and /l/: IV. some effects of perceptual learning on speech production. Journal of the
Acoustical Society of America, 101(4), 2299–2310.
Carlet, A. (2019). Different high variability procedures for training L2 vowels and consonants. In Proceed-
ings of the ICPhS, Melbourne, Australia, August 2019.
Evans, B., & Martín-Alvarez, L. (2016). Age-related differences in second-language learning? A compari-
son of high and low variability perceptual training for the acquisition of English /ɪ/-/ i/ by Spanish
adults and children. Paper presented at New Sounds. International symposium on the acquisition of
Second Language Speech. Aarhus: Aarhus University.
Flege, J. E. (1995). Second language speech learning: Theory, findings and problems. In W. Strange (Ed.),
Speech perception and linguistic experience: Theoretical and methodological issues (pp. 233–277).
Baltimore: York Press.
Flege, J. E., MacKay, I. R. A., & Meador, D. (1999). Native Italian speakers’ perception and production of
English vowels. The Journal of the Acoustical Society of America, 106, 2973–2987.
Flege, J. E., Munro, M. J., & MacKay, I. R. A. (1995). The effect of age of second language learning on the
production of English consonants. Speech Communication, 16(1), 1–26.
Flege, J. E., & MacKay, I. R. A. (2004). Perceiving vowels in a second language. Studies in Second Lan-
guage Acquisition, 26, 1–34.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
720 Journal of Psycholinguistic Research (2021) 50:707–721

Georgiou, G. P. (2018). Discrimination of L2 Greek vowel contrasts: Evidence from learners with Arabic
L1 background. Speech Communication, 102, 68–77.
Georgiou, G. P. (2019a). Cue-reliance and vowel perceptual patterns of Cypriot Greek children who learn
English. Journal of Experimental Phonetics, 28, 229–253.
Georgiou, G. P. (2019b). EFL teachers’ cognitions about pronunciation in Cyprus. Journal of Multilingual
and Multicultural Development, 40(6), 538–550.
Georgiou, G. P., Perfilieva, N., Denisenko, V., & Novoskaspaya, N. (2020b). Perceptual realization of Greek
consonants by Russian monolingual speakers. Speech Communication, 125, 7–14.
Georgiou, G. P., Perfilieva, N., & Tenizi, M. (2020a). Vocabulary size leads to better attunement to L2
phonetic differences: Clues from Russian learners of English. Language Learning and Development,
16(4), 382–398.
Georgiou, G. P., & Themistocleous, C. (2020). Vowel learning in diglossic settings: Evidence from Arabic-
Greek learners. International Journal of Bilingualism. https://​doi.​org/​10.​1177/​13670​06920​945396.
Goto, H. (1971). Auditory perception by normal Japanese adults of the sounds ‘l’ and ‘r.’ Neuropsychologia,
9(3), 317–323.
Hattori, K., & Iverson, P. (2008). English /r/–/l/ pronunciation training for Japanese speakers. Journal of the
Acoustical Society of America, 123, 3327.
Hazan, V., Sennema, A., Iba, M., & Faulkner, A. (2005). Effect of audiovisual perceptual training on the
perception and production of consonants by Japanese learners of English. Speech Communication,
47(3), 360–378.
Heeren, W., & Schouten, M. E. H. (2008). Perceptual development of phoneme contrasts: How sensitivity
changes along acoustic dimensions that contrast phoneme categories. Journal of the Acoustical Society
of America, 124, 2291–2302.
Hyltenstam, K., & Abrahamsson, N. (2008). Maturational constraints in SLA. In C. J. Doughty & M. H.
Long (Eds.), The handbook of second language acquisition (pp. 539–588). Oxford: Blackwell Publish-
ing Ltd.
Inceoglu, S. (2016). Effects of perceptual training on second language vowel perception and production.
Applied Psycholinguistics, 37(5), 1175–1199.
Ingvalson, E. M., Barr, A. M., & Wong, P. C. (2013). Poorer phonetic perceivers show greater benefit in
phonetic-phonological speech learning. Journal of Speech, Language, and Hearing Research, 56(3),
1045–1050.
Iverson, P., Pinet, M., & Evans, B. (2012). Auditory training for experienced and inexperienced second-
language learners: Native French speakers learning English vowels. Applied Psycholinguistics, 33(1),
145–160.
Kartushina, N., & Frauenfelder, U. H. (2013). On the role of L1 speech production in L2 perception: Evi-
dence from Spanish learners of French. Proceedings of the 14th Interspeech conference (pp. 2118–
2122). Lyon.
Lenneberg, E. (1967). Biological foundations of language. New York: Wiley.
Lewandowski, N., Dogil, G., & Reiterer, S. (2009). Sociolinguistic factors in language proficiency: Phonetic
convergence as a signature of pronunciation talent. In G. Dogil & S. Reitener (Eds.), Language talent
and brain activity (pp. 257–278). Berlin: Walter de Gruyter.
Lively, S. E., Logan, J. S., & Pisoni, D. B. (1993). Training Japanese listeners to identify English /r/and/l/.
II: The role of phonetic environment and talker variability in learning new perceptual categories. The
Journal of the Acoustical Society of America, 94(3), 1242–1255.
Logan, J. S., Lively, S. E., & Pisoni, D. B. (1991). Training Japanese listeners to identify English /r/and/l/: A
first report. The Journal of the Acoustical Society of America, 89(2), 874–886.
Long, M. (1990). The least a second language acquisition theory needs to explain. University of Hawai’i
Working Papers in ESL, 9(1), 59–75.
MacKay, I. R. A., Flege, J. E., Piske, T., & Schirru, C. (2001). Category restructuring during second-lan-
guage (L2) speech acquisition. The Journal of the Acoustical Society of America, 110, 516–528.
Munro, J. M. (1993). Productions of english vowels by native speakers of arabic: Acoustic measurements
and accentedness ratings. Language and Speech 36(1), 39–66.
Nishi, K., & Kewley-Port, D. (2007). Training Japanese listeners to perceive American English vowels:
Influence of training sets. Journal of Speech Language and Hearing Research, 50(6), 1496–1509.
Norlin, K. (1987). A phonetic study of emphasis and vowels in Egyptian Arabic (Doctoral dissertation, Uni-
versity of Lund). Retrieved from http://​journ​als.​lub.​lu.​se/​index.​php/​LWPL/​artic​le/​view/​16952/​15331
Pennington, M. C. (1998). The teachability of phonology in adulthood: A re-examination. International
Review of Applied Linguistics, 36, 323–341.
Piske, T., Flege, J. E., MacKay, I. R. A., & Meador, D. (2002). The production of English vowels by fluent
early and late Italian-English bilinguals. Phonetica, 59(1), 49–71.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Psycholinguistic Research (2021) 50:707–721 721

Piske, T., MacKay, I. R. A., & Flege, J. E. (2001). Factors affecting degree of foreign accent in an L2: A
review. Journal of Phonetics, 29(2), 191–215.
Pruitt, J. S., Jenkins, J. J., & Strange, W. (2006). Training the perception of Hindi dental and retroflex stops
by native speakers of American English and Japanese. Journal of the Acoustical Society of America,
119(3), 1684–1696.
Rato, A. (2014). Effects of perceptual training on the identification of English vowels by native speakers of
European Portuguese. Concordia Working Papers in Applied Linguistics, 5, 529–546.
Rauber, A., Rato, A., Kluge, D., Santos, G. (2011). TP-S (Version 1.0) [Application software]. Retrieved
from http://​www.​worken.​com.​br/​siste​mas/​tp-s/
Rota, G., & Reiterer, S. (2009). Cognitive aspects of pronunciation talent. In G. Dogil & S. Reitener (Eds.),
Language talent and brain activity (pp. 67–96). Berlin: Walter de Gruyter.
Shinohara, Y., & Iverson, P. (2018). High variability identification and discrimination training for Japanese
speakers learning English /r/-/l/. Journal of Phonetics, 66, 242–251.
Snow, C. E. (1987). Second language learners’ format definitions: An oral language correlate of school lit-
eracy. Los Angeles: University of California, Center for Language Education and Research.
Stanovich, K. E. (2000). Progress in understanding reading. New York: Guilford.
Themistocleous, C. (2017). Modern Greek vowels and the nature of acoustic gradience. Phonetica, 74(3),
157–172.
Thomson, I. R. (2011). Computer assisted pronunciation training: Targeting second language vowel percep-
tion improves pronunciation. Calico Journal, 28(3), 744–765.
Trudgill, P. (2009). Greek dialect vowel systems, vowel dispersion theory, and sociolinguistic typology.
Journal of Greek Linguistics, 9(1), 80–97.
Thomson, I. R. (2018). High variability [pronunciation] training (HVPT). Journal of Second Language Pro-
nunciation, 4(2), 208–231.
Tsukada, K., Birdsong, D., Bialystok, E., Mack, M., Sung, H., & Flege, J. E. (2005). A developmental study
of English vowel production and perception by native Korean adults and children. Journal of Phonet-
ics, 33, 263–290.
Yamada, R. A. (1995). Age and acquisition of second language speech sounds perception of American Eng-
lish /r/ and /l/ by native speakers of Japanese. In W. Strange (Ed.), Speech perception and linguistic
experience: Issues in cross-language research (pp. 305–320). Timonium, MD: York Press.

Publisher’s Note  Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center
GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers
and authorised users (“Users”), for small-scale personal, non-commercial use provided that all
copyright, trade and service marks and other proprietary notices are maintained. By accessing,
sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of
use (“Terms”). For these purposes, Springer Nature considers academic use (by researchers and
students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and
conditions, a relevant site licence or a personal subscription. These Terms will prevail over any
conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription (to
the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of
the Creative Commons license used will apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may
also use these personal data internally within ResearchGate and Springer Nature and as agreed share
it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not otherwise
disclose your personal data outside the ResearchGate or the Springer Nature group of companies
unless we have your permission as detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial
use, it is important to note that Users may not:

1. use such content for the purpose of providing other users with access on a regular or large scale
basis or as a means to circumvent access control;
2. use such content where to do so would be considered a criminal or statutory offence in any
jurisdiction, or gives rise to civil liability, or is otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association
unless explicitly agreed to by Springer Nature in writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a
systematic database of Springer Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a
product or service that creates revenue, royalties, rent or income from our content or its inclusion as
part of a paid for service or for other commercial gain. Springer Nature journal content cannot be
used for inter-library loans and librarians may not upload Springer Nature journal content on a large
scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not
obligated to publish any information or content on this website and may remove it or features or
functionality at our sole discretion, at any time with or without notice. Springer Nature may revoke
this licence to you at any time and remove access to any copies of the Springer Nature journal content
which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or
guarantees to Users, either express or implied with respect to the Springer nature journal content and
all parties disclaim and waive any implied warranties or warranties imposed by law, including
merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published
by Springer Nature that may be licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a
regular basis or in any other manner not expressly permitted by these Terms, please contact Springer
Nature at

onlineservice@springernature.com

You might also like