ArticlePDF Available

Cue-reliance and vowel perceptual patterns of Cypriot Greek children who learn English

Authors:

Abstract

The aim of the present study is to investigate the vowel perceptual patterns of Cypriot Greek children who learn English as a foreign language in Cyprus and their cue-reliance during L2 vowel categorization. To this purpose, 26 native speakers of Cypriot Greek participated in a perceptual assimilation task and an AXB discrimination task. The findings showed that learners assimilated two or more English vowels to a single phonological category of their native language. Furthermore, their discrimination over selected English vowel contrasts ranged from poor to moderate. The small L1 vowel inventory of learners resulted in multiple assimilations of L2 vowels to the same L1 phonological category and the lack of linguistic experience prevented the establishment of novel phonological categories. In addition, the temporal cues of L2 vowels were not employed for their better fitting to the already assimilated L1 phonological categories. The predictions were in general consistent with the assumptions of the Perceptual Assimilation Model-L2.
Artículo recibido el día: 13/06/2019
Artículo aceptado definitivamente el día: 04/11/2019
Estudios de Fonética Experimental, ISSN 1575-5533, XXVIII, pp. 229-253
CUE-RELIANCE AND
VOWEL PERCEPTUAL PATTERNS
OF CYPRIOT GREEK CHILDREN
WHO LEARN ENGLISH
PONDERACIÓN DE RASGOS
Y PATRONES PERCEPTIVOS
DE LAS VOCALES INGLESAS EN NIÑOS
GRECOCHIPRIOTAS QUE APRENDEN INGLÉS
GEORGIOS P. GEORGIOU
People’s Friendship University of Russia (RUDN University)
(Russia)
georgiou.georgos@hotmail.com
230 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 231
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
ABSTRACT
The aim of the present study is to investigate the vowel perceptual patterns of
Cypriot Greek children who learn English as a foreign language in Cyprus and
their cue-reliance during L2 vowel categorization. To this purpose, 26 native
speakers of Cypriot Greek participated in a perceptual assimilation task and an
AXB discrimination task. The findings showed that learners assimilated two or
more English vowels to a single phonological category of their native language.
Furthermore, their discrimination over selected English vowel contrasts ranged
from poor to moderate. The small L1 vowel inventory of learners resulted in
multiple assimilations of L2 vowels to the same L1 phonological category and the
lack of linguistic experience prevented the establishment of novel phonological
categories. In addition, the temporal cues of L2 vowels were not employed for
their better fitting to the already assimilated L1 phonological categories. The
predictions were in general consistent with the assumptions of the Perceptual
Assimilation Model-L2.
Keywords: Cypriot Greek, English, children, perception
RESUMEN
El objetivo del presente estudio es investigar los patrones de percepción vocal de
los niños grecochipriotas que aprenden inglés como lengua extranjera en Chipre y
su ponderación de la señal durante la categorización de las vocales de la L2. Para
este propósito, 26 hablantes nativos de griego chipriota participaron en
experimentos de asimilación y discriminación de tipo AXB. Los resultados
mostraron que los aprendices asimilaron dos o más vocales inglesas a una sola
categoría fonológica de su idioma nativo. Además, su discriminación sobre los
contrastes de las vocales inglesas seleccionadas varió de pobre a moderada. El
pequeño inventario de vocales de la L1 de los aprendices dio lugar a múltiples
asimilaciones de vocales de la L2 a la misma categoría fonológica de la L1 y la
falta de experiencia lingüística impidió el establecimiento de nuevas categorías
fonológicas. Es más, las señales temporales de las vocales de la L2 no se
emplearon para su mejor ajuste en las categorías fonológicas de la L1 ya
asimiladas. En general, las predicciones fueron consistentes con las suposiciones
del Perceptual Assimilation Model-L2.
Palabras clave: chipriota griega, inglés, niños, percepción
232 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
1. INTRODUCTION
It is well documented by the literature that infants are able to discriminate
nonnative sound contrasts in most languages until the first six months of their life
(Georgiou, 2018a; Kuhl et al., 2006; Werker and Tees, 1984). This ability is
gradually lost at a later stage with the development of the first language (L1)
phonological system, which creates speech perception difficulties since learners
filter every second language (L2) sound through their L1 phonological system
(Bosch et al., 2000; Sebastián-Gallés and Soto-Faraco, 1999; Bongaerts et al.,
1997; Best and Strange, 1992; Flege and Eefting, 1987). In other words, they move
from a universal acoustic discrimination to wrapped language-specific perception
(Kuhl, 2000). Therefore, learners with different L1s have different perceptual
difficulties. For instance, the German vowel contrast /ε/-/æ/ is discriminated
differently by Spanish and French learners of German (Flege et al., 1997). Spanish
speakers are able to discriminate the aforementioned German vowel contrast very
well since they perceive the two vowels as being exemplars of two different
phonological categories of their L1 while French speakers perceive these vowels as
being instances of a single French phonological category. Also, phonological
inventory size matters since speakers with small inventories tend to assimilate
nonnative sounds that come from larger sound inventories to a single phonological
category of their L1 (Georgiou, 2019a; Lengeris, 2009; Frieda and Nozawa, 2007;
Hacquard et al., 2007; Fox et al., 1995). For instance, Kartushina and Frauenfelder
(2013) reported that Spanish learners of French did not manage to discriminate
accurately the French /e/-/ε/ vowel contrast since these vowels were perceived as
being similar since Spanish has only the sound /e/ and not /ε/.
The Native Language Magnet (NLM) model was proposed by Kuhl (1992) aims at
investigating speech perception by taking into consideration cognitive aspects.
Specifically, it argues that phonetic representations of speakers’ L1 categories,
which are called “prototypes”, are developed by infants during the first year of
their life (Kuhl, 1992; Kuhl and Iverson, 1995). These prototypes are good
instances of categories and work as “magnets” for other sounds of that categories.
NLM makes as well predictions for L2 learning by adults. It predicts that a learner
might face difficulties in discriminating an L2 sound that is phonetically similar to
an L1 prototype while L2 sounds that are dissimilar to any L1 prototype are
predicted to be discriminated more easily. The Second Language Linguistic
Perception Model (L2LP) was developed by Escudero (2005, 2009) supporting
that the initial learning of L2 acquisition is the result of L1 acquisition since
similarities or differences between the L1-L2 phonemes shape the perception of
the L2 phonemes. The model is based on three scenarios: (a) “new scenario”,
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 233
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
which indicates that two L2 phones have been included to the same L1 category
and therefore learners must split their L1 category or create new phonological
categories, (b) “similar scenario” in which two L2 contrasts are acoustically close
to two different L1 categories; this scenario is less problematic than creating new
categories according to the L2LP, and (c) “subset scenario”, which indicates that a
single L2 sound was perceived as more than one category; this scenario creates
less difficulties for the learners than the “new scenario” since learners do not have
to create new categories (van-Leussen and Escudero, 2015). The Automatic
Selective Perception (ASP) Model (Strange, 2011) suggests that there is not only a
single way of perceiving speech, but different acts of perception that adapt to the
purpose of communication. Thus, according to the ASP, evaluation tasks are
generally carried out taking into account a mode of phonetic perception, which
requires more attention and more cognitive effort. On the other hand, in order to
recognize the forms of words, the phonological mode is used, since speakers
ignore phonetic variations such as speech speed, prosodic structure, or minor
dialectal variations.
The Speech Learning Model (SLM) (Flege, 1995) is a perceptual and production
model that deals with experienced learners of an L2 (that is, learners who live in an
L2-dominant country). It supports that there is a common phonological space for
the L1 and the L2 and that learners form phonological categories from speech
input. If an L1 sound is perceived to be similar with an L2 sound, then merged
categories will be formed and the perception and production of that L2 sound will
be inaccurate. The acquisition of similarity between an L1 and an L2 phone is
fostered by the mechanism of equivalence classification. The SLM considers the
effect of several sociolinguistic factors on the perception and production of L2
sounds such as age of arrival in the L2-speaking country, age that someone started
to learn the L2, L1/L2 use and many others. However, the model does not form
detailed predictions about the perception of speech sounds. Also, the consideration
of time of learning (e.g., early stage) is important in order to make accurate
predictions about the L2 learners’ speech acquisition patterns (Georgiou, 2018b).
The Perceptual Assimilation Model (PAM) (Best, 1995) is one of the most
influential speech models that investigates the sound perceptual patterns of
monolingual listeners with respect to languages that they do not have experience
with. PAM’s framework is based on a direct realism theory where speech signal is
directly picked by the listeners without any process of the acoustical information
(Best, 1995). PAM contends that nonnative sounds might be assimilated to similar
L1 sounds due to the effect of the speakers’ L1 (Bundgaard-Nielsen et al., 2011),
however, perception of within-category L1-L2 differences might be possible if
234 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
listeners are able to perceive much gestural dissimilarities between the L1 and the
L2 sounds (Tyler et al. 2014; Antoniou, 2010). In order to explain better the
perceptual mechanisms, PAM proposed six types of assimilation for nonnative
sounds to the listeners’ L1 phonological system; these assimilation types also
define the degree of discrimination of pairs of nonnative sound contrasts (Best and
Tyler, 2007). In the Two Category (TC) Assimilation, two nonnative speech
sounds are assimilated to two different L1 categories and, thus, the discrimination
of this contrast is predicted to be excellent. The Single Category (SC) Assimilation
signals that a nonnative contrastive sound contrast differs from an ideal L1 sound
and, therefore, the discrimination of its members will be poor. In the Category
Goodness (CG) Assimilation type two nonnative sounds are assimilated to the
same L1 phonological category but one of them constitutes a good exemplar of
that L1 sound while the other constitutes a bad exemplar of it; discrimination is
predicted to be moderate to good. The Uncategorized-Categorized (UC)
Assimilation implies that one nonnative sound is assimilated to an L1 category
while the other falls into the phonological space without being assimilated to any
L1 phonological category. In this case, it is expected a very good discrimination of
the nonnative sounds. In the Uncategorized-Uncategorized (UU) both nonnative
sounds fall into the phonological space; discrimination of these members is
predicted to be poor to very good. Finally, the Non-Assimilable (NA) type signals
that neither of the two nonnative sounds is perceived as a speech sound and
discrimination might be poor to excellent. PAM-L2 (Best and Tyler, 2007) is the
extended version of PAM and deals with the perceptual patterns of L2 learners
(instead of novice learners that PAM deals with). PAM-L2 argues that L2
phonological categories might be assimilated to existing L1 phonological
categories but if an L2 sound is perceived as dissimilar from any L1 sound, then
novel phonological categories might be developed. Also, PAM-L2 supports that
the expansion of vocabulary is associated with the discrimination of more phonetic
distinctions on the basis of lexically relevant categorical differences in the L2
(Antoniou et al., 2012).
The purpose of SLM, PAM, and PAM-L2 is similar: to predict the ability of
learners/listeners in the perception of L2/unfamiliar contrasts. However, SLM is
mainly a production model despite the fact that it makes some general predictions
about the reorganization of speech sounds after perceiving L2 speech input. Both
PAM and PAM-L2 make explicit perceptual predictions, however, PAM’s
predictions are based on inexperienced learners while PAM-L2 focuses on learners
with experience in the L2. So, for developing hypotheses in this study, PAM-L2
was chosen since the learners of the study have some experience with English.
Several studies in the literature have tested the predictions of PAM-L2. Georgiou
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 235
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
(2018) investigated the discrimination of L2 Greek vowel contrasts by Arabic
speakers. The results showed that contrasts that signalled a CG assimilation were
more distinguishable than contrasts of SC assimilation type. The same relationship
of assimilation types as reported in Georgiou (2018) was also reported in
Mahmoud (2013) who investigated the perception of Arabic contrasts by American
English learners of Arabic. By contrast, Sun and van Heuven (2007) studied the
perception of English vowel contrasts by Chinese listeners concluding that the CG
assimilation was less discriminable than the SC assimilation type. Thus, the
predictions of PAM-L2 were not confirmed by this study.
An important issue in speech perception studies is the cue-reliance of L2 learners
during the perception of L2 sounds. Most studies provide evidence about cue-
reliance during the discrimination of L2 contrasts rather during cross-linguistic
categorization of L2 sounds. For example, Kim et al. (2018) examined the changes
in cue-weighting strategies of adult and child Korean learners of English with
respect to English contrastive vowel pairs. The results showed that both
populations of Korean learners used both spectral and duration cues to distinguish
/i/-/ɪ/ but they relied only on duration to discriminate the English /e/-/æ/ contrast.
With respect to cross-linguistic cue-reliance, there is much evidence that L2
speakers tend to assimilate L2 sounds to the spectrally closest phonological
categories of their L1 (see Georgiou, 2018b; Escudero et al., 2014). Nevertheless,
not much studies examined whether speakers rely or not on durational cues during
the categorization of L2 sounds. Lengeris (2009) investigated the perception of
English vowels by both Greek and Japanese speakers to conclude that both
populations of learners relied on both spectral and duration cues during the
classification of L2 vowels to their L1 system.
The purpose of the present study is to investigate the assimilation of English
vowels and the discrimination of English vowel contrasts by children who learn
English as foreign language in Cyprus, and to determine the cue-reliance of L2
learners during vowel categorization. This study is expected to contribute
significantly to the large body of research on cross-language speech perception
since it investigates vowel perceptual patterns of an under-researched population
(children aged 8-12) in a relatively under-researched context (foreign language
instruction rather than L2 immersion). Furthermore, the languages involved, that is
Cypriot Greek and English, have not been investigated much in the literature. The
same participants that took part in Georgiou (2019a) also participated in this study.
Both low and high proficiency learners of English in Georgiou (2019a) had
significant difficulties in perceiving the acoustical differences of specific English
vowel pairs; however, this study does not take into account the learners’
236 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
proficiency level but only children’s general perceptual patterns. Lengeris (2009)
investigated the perception of English vowels by adult Greek speakers to conclude
that many English vowels were assimilated to a single Greek phonological
category, e.g., /iː/ and /ɪ/ were both assimilated to the Greek category /i/, English
/ε/ and /ɜː/ to Greek /e/, English /æ/and /ʌ/ to Greek /a/, English /ɑː/, /ɒ/, and /ɔː/ to
Greek /o/, and English vowels /ʊ/ and /u:/ to the Greek phonological category /u/.
The author added that most of the English contrasts showed low to moderate
discrimination scores.
We selected children instead of adults in order to examine the perceptual abilities
of this population of learners who are believed to acquire a nonnative language
with less effort than adults (Snow, 1987). In any case, children differ to a great
extent from adults in the way they process speech sounds. For example, perception
of synthetic vowels by 3-year-olds relies more on dynamic spectral change
information than adults (Murphy et al., 1989), while perceptions of 5-11 year-olds
are more influenced by the duration of stimulus and the consonantal context
compared to adults (Ohde et al., 1996). This is to say that children’s perceptual
patterns are not yet adult-like (Walley, 2005); they are not segmental but holistic
since they are conducted on the basis of information distributed through the speech
signal. Similarly, De Cara and Goswami (2003) point out that children’s lexical
representations differ from those of adults in that they are holistic, that is, children
do not pay attention on particular phonemic contrasts but in other features such as
prosody when they produce speech. This seems to change only when they get
mature, and gain more linguistic experience. Also, the expansion of lexicon plays
an important role since it forces speakers to give more emphasis on the phonetic
details of the speech signal. In that way, their holistic representations are
restructured and thus smaller segments of sounds such as syllables and finally
phonemes are represented (De Cara and Goswami, 2003).
2. CYPRIOT GREEK AND ENGLISH VOWEL SYSTEMS
The Cypriot Greek vowel system consists of 5 pure vowel phonemes /i e a o u/
without short-long or tense-lax distinctions (Georgiou, 2018b). On the contrary,
English vowel system (Received Pronunciation) consists of 11 monophthongs
which can be either tense /iː ɜː ɑː ɔː / or lax ʊ e æ ʌ ɒ/ (Roach, 2004). Tense
vowels are longer than lax vowels. However, in many contexts (e.g., before
voiceless obstruents) vowel length differences between tense and lax vowels are
neutralized. It is evident that the English vowel inventory is more complex than the
Cypriot Greek one since more acoustic features are needed to describe it.
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 237
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
Specifically, the two vowel systems differ both in “density” (5 vs. 11 vowels) and
in that specific vowels in each of them differ considerably in terms of phonetic
features (for the role of vowel system “density” in speech perception and
production, see Meunier et al., 2003).
The predictions about the assimilation of English vowels and the discrimination of
English vowel contrasts by Cypriot Greek speakers will be developed on the basis
of the PAM-L2 which makes predictions about the perception of contrasts by L2
learners. To this purpose, we have to take into consideration the acoustic properties
of both Cypriot Greek and English vowels in the common phonological space.
Considering the articulatory/phonetic features of the aforementioned languages, we
can conclude that the English vowels /iː/ and /ɪ/ will be assimilated to the Cypriot
Greek phonological category /i/. However, bearing in mind that there is evidence
that L2 learners rely on both spectral and temporal cues even if there are no
temporal distinctions in their native language (Lengeris, 2009) we assume that /ɪ/
will be good exemplar of the Greek /i/ since they match for duration while /iː/ will
be a moderate exemplar of that Greek category. The English vowels /ε/ and /ɜː/ are
expected to be assimilated to the Greek /e/ with the former English vowel to
constitute a good exemplar of that category while the latter to constitute a bad
exemplar of it. The English vowels æ ɑː/ are predicted to be assimilated to the
Greek /a/ due to their close phonetic distance. Nevertheless, only /ʌ/ and /æ/ are
expected to be good exemplars of the Greek /a/ since they have close durations
with that Greek vowel while /ɑː/ might constitute moderate exemplar of it due to
its temporal difference. Also, the English /ɒ/and /ɔː/ are hypothesized to be
assimilated to the Greek /o/ as two different exemplars of that category due to their
temporal difference. Lastly, the English vowels /ʊ/ and /uː/ are expected to be
assimilated to the Greek vowel /u/ both as different exemplars of that Greek
phonological category: the former as a good exemplar of it since they have close
durations and the latter as a moderate exemplar of it.
With respect to the discrimination of the English vowel contrasts: /iː/- /ɪ/, /e/- /ɜː/,
and /ɒ/- /ɔː/, we expect a CG assimilation type with moderate (to good)
discrimination of these contrasts since both members are predicted to be
assimilated to the same Greek phonological category as two different exemplars of
that category. However, for the English vowel contrast /æ/- /ʌ/, it is expected a SC
assimilation since both of its members are predicted to be assimilated to the same
Greek phonological category as equally good exemplars of it. At this point, we do
not assume that learners will form novel phonological categories since they do not
have much experience with the foreign language. Figure 1 illustrates the F1 and F2
238 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
of Cypriot Greek and English vowels as produced by their respective native
speakers (Georgiou, 2019a). Figure 2 and 3 illustrate the durations of Cypriot
Greek and English vowels as produced by their native speakers respectively
(Georgiou, 2019a).
Figure 1. F1 × F2 of Cypriot Greek (CGR) and English (RP) vowels
as produced by their respective native speakers (Georgiou, 2019a).
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 239
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
Figure 2. Mean values (in m/s) and SDs for the durations of the CGR
vowels as produced by young native speakers of CGR (Georgiou,
2019a).
Figure 3. Mean values (in m/s) and SDs for the durations of the RP
vowels as produced by young native speakers of RP (Georgiou,
2019a).
240 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
3. METHODOLOGY
3.1. Participants
Twenty-six subjects aged 8-12 years participated in the study (the same
participants as in Georgiou, 2019a). Specifically, there were six 8-years-olds,
seven 9-years-olds, five 11-years-olds and eight 12-year-olds. We selected only
female speakers in order to minimize gender-based bias that may affect L2 vowel
perception. Their native language was Cypriot Greek as they were born and raised
in Cyprus. The participants were residents of urban areas of Larnaca city, Cyprus
and they originated from moderate-income families. Although children’s
proficiency level differed, further evidence suggests that the effect of proficiency
level on the speech perception of these learners was minimal since learners did not
receive any pronunciation instruction and their teachers were nonnative speakers
of English (Georgiou, 2019a). All of the children started to learn English at the age
of 7 (Mage = 7.3) by attending courses at school and at private institutes. Their
contact with English was mainly though foreign language courses since they were
not practicing their English much in their daily life (e.g., by reading English books,
by watching English movies, etc.). Also, the majority of the children reported that
they often listen to the music, listening mostly to Greek music and to a lesser
extent to English music. None of the children had any knowledge of any foreign
language apart from English. Finally, all subjects reported that they had a normal
hearing and they never had any language disorder.
3.2. Stimuli
One 10-year-old female native speaker of British English produced the stimuli of
the perceptual assimilation task and her productions were recorded via Zoom audio
recorder (44.1 kHz sample). The stimuli of the AXB discrimination task were
recorded again by the same speaker as in the assimilation task (all A, X, B were
produced by the same speaker as well). The stimuli were monosyllabic words
consisting of 11 English (Received Pronunciation) monophthongs (/ɪ ʊ e æ ʌ ɒ iː uː
ɜː ɔː ɑː/) in the frame of /bVd/ (real words). For the discrimination task, only 4
English vowel contrasts were chosen: 1) /iː/- /ɪ/, 2) /e/- /ɜː/, 3) /æ/- /ʌ/, and /ɒ/- /ɔː/;
all of these contrasts were predicted to be difficult to discriminate by the learners.
3.3. Procedure
The perceptual assimilation task was created in PRAAT (Boersma and Weenink,
2018) and the script included orthographical transliteration of the 5 Greek vowels
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 241
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
(e.g., “ι, ε”, α”, ο”,ου”). The participants were seated in front of a laptop
and they listened to the English stimuli at a comfortable level (75dB) via the
headphones that were connected to the laptop’s loudspeakers. Then, they were told
to click on the Greek vowels in the script that are acoustically closer to the English
vowels they heard. After choosing the vowel, they were told to rate it by clicking
on one of the script options (scale 1-5: 1 = very poor, 5 = very good), depending
on how good exemplar of the corresponding L1 vowel the English vowel was. It
was included a 5-point scale (as in Georgiou, 2018b, 2019a) instead of a 7-point
one since we wanted to make it quite simple for the children to read out the
complete list of scale descriptors and thus respond better (Dawes, 2008). In total,
44 tokens were presented to the subjects (11 vowels × 4 repetitions) and they had 5
seconds to assimilate each vowel with an optional 2-minute break at the midpoint.
The task was accomplished in a sound attenuated room.
After the perceptual assimilation task, learners completed an AXB discrimination
task (4 trials: AAB, ABB, BBA, BAA; where A and B represent different English
words having different vowels) (Best et al., 2001) in a sound-attenuated room. The
task was completed on a PC monitor using a PRAAT script (the script displayed
the labels “first” and “third”). The subjects of the study were informed that tokens
A and B were acoustically different vowels and that vowel X was the same as A or
B. The participants were listening (at a listening volume of 75dB) to a triad of
words that contained the target English vowel contrast (e.g., for the /ɜː/-/e/
contrast: /bet/ - /bet/ - /bɜːt/) through headphones that were connected to a laptop.
Then, they were instructed to choose by clicking on the script labels whether the
middle sound was the same as the first or the third sound. They had to discriminate
a total of 64 items (4 trial types × 4 vowel contrasts × 4 repetitions) with an
optional brake after the 32nd item. The interstimulus and intertrial intervals were 2s
and 7s respectively.
4. RESULTS
A vowel was considered as “categorized” if the same category label was selected
on at least 70% of responses (following Georgiou, 2019a, 2018b). The findings
indicated that two or more English vowels were assimilated to a single Cypriot
Greek phonological category. As it can be seen from Table 1, English vowel /ɪ/
was assimilated to the Cypriot Greek phonological category /i/ while English /iː/
was assimilated to the same phonological category. English vowels /e/ and /ɜː/
were both assimilated to the Cypriot Greek /e/. Also, English /ʌ æ ɑː/ were
assimilated to the Greek /a/, and English vowels ɔː/ to the Cypriot Greek /o/.
Lastly, English /ʊ uː/ were assimilated to the Greek /u/. So, within-category
242 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
assimilations emerged for the English vowel contrasts /ɪ/ - /iː/, /e/ - /ɜː/, / - /æ/,
/ʌ/ - /ɑː/, /ɑː/ - /æ/, /ʊ/ - /uː/, and /ɒ/ - /ɔː/.
To investigate further the overall fit of English vowels to the Cypriot Greek
phonological categories, “fit indexes” were developed (see Guion et al., 2000).
The fit indexes combine the identification and goodness of fit rating data into a
single metric. For example, the fit index of the English vowel /ɜː/ (1.4) emerged by
multiplying its proportion of identification (0.7) with its goodness of fit rating (2).
Fit indexes are shown in Table 2. It is assumed that English vowels with high fit
indexes can be accepted as good instances of Cypriot Greek phonological
categories while low fit indexes signify that English vowels are distorted instances
of Cypriot Greek categories. The mean fit index of all English vowels was 3.1 (SD
= 1). When an English vowel fell within 1 SD from the mean fit index of all
English vowels, then it was classified as a “good” instance of a Cypriot Greek
phonological category (2.1 and over). When an English vowel fell within 2 SDs
from the mean fit index of all English vowels, then it was classified as a “fair”
instance of a Cypriot Greek phonological category (1.1-2).
RP vowels
CGR
vowels
/ɪ/
/iː/
/e/
/ɜː/
/ʌ/
/æ/
/ɑː/
/ʊ/
/uː/
/ɒ/
/ɔː/
/i/
95
(2.9)
100
(4.4)
17
(1.5)
/e/
5
(1.8)
83
(3.5)
70
(2)
16
(1.6)
8
(2.5)
/a/
30
(1.6)
84
(4)
92
(4.2)
80
(3.9)
13
(2.2)
/o/
19
(1.6)
24
(1.9)
1
(1.3)
85
(3)
81
(2.7)
/u/
85
(2.9)
100
(4.6)
3
(1.6)
19
(2.5)
Table 1. Percentage of the assimilation of the English vowels to
the Cypriot Greek phonological categories. In brackets, the
goodness of fit ratings are indicated (1= very poor, 5=very good).
The bold cells represent the predicted assimilations while the
blank cells show that the assimilation was below 1% to that vowel
category.
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 243
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
English
vowel
Most common
identification
Proportion of
identification
Fit index
i
1
4.4
Good /i/
ɜː
e
0.7
1.4
Fair /e/
ɑː
a
0.8
3.1
Good /a/
ɔː
o
0.81
2.2
Good /o/
u
1
4.6
Good /u/
ɪ
i
0.95
2.8
Good /i/
e
e
0.83
2.9
Good /e/
æ
a
0.92
3.9
Good /a/
ʌ
a
0.84
3.4
Good /a/
ɒ
o
0.85
2.5
Good /o/
ʊ
u
0.85
2.5
Good /u/
Table 2. Fit indexes emerged for English vowels in terms of Cypriot
Greek categories. Only identifications that were more than 70% are
included.
The fit index provided that both members of the following within-category
contrasts were perceived as good instances of the Cypriot Greek phonological
category they were assimilated to: /ɪ/ - /iː/, /ʌ/ - /æ/, /ʌ/ - /ɑː/, /ɑː/ - /æ/, /ʊ/ - /uː/,
and /ɒ/ - /ɔː/. On the contrary, the members of the English contrast /e/ - /ɜː/ were
assimilated as good and fair instances of the Cypriot Greek /e/ category
respectively. However, fit indexes provided only a general picture of how do
English vowels fitted to the learners’ L1 phonological system with the majority of
them to have been perceived as “good” instances of Cypriot Greek categories. In
order to determine the exact assimilation pattern of within-category assimilations
(which can be either SC or CG), paired-sample t-tests were used for the goodness
of fit ratings of vowels that were assimilated to the same Cypriot Greek
phonological category. If the vowels’ goodness of fit ratings do not differ, then
learners are not able to perceive the phonetic differences between the two sounds;
in this case, a SC assimilation will occur. By contrast, if the vowels’ goodness of
fit ratings differ, then learners are able to perceive the sounds’ phonetic distance; a
CG assimilation will take place. The results showed significant differences for the
goodness of fit ratings of /ɪ/ - /iː/, [t(25) = 2.51, p=0.01], /e/ - /ɜː/ [t(25) = - 2.41,
p=0.01], and /ʊ/ - /uː/ [t(25) = 4.07, p=0.001]; all these contrasts signaled a CG
assimilation type. The other contrasts, namely, /ʌ/ - /æ/, /ʌ/ - /ɑː/, /ɑː/ - /æ/, and /ɒ/
- /ɔː/ signaled a SC assimilation since their goodness of fit ratings did not differ.
Table 3 shows the discrimination of the 4 English vowel contrasts under
investigation. In general, the discrimination patterns of Cypriot Greek learners of
244 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
English ranged from poor to moderate; poor for /æ/ - /ʌ/ and /ɒ/ - /ɔː/ contrasts and
moderate for /ɪ/ - /iː/ and /e/ - /ɜː/ contrasts. Specifically, the two CG contrasts
could be discriminated almost with the same manner but more accurately (68 and
70%) than the SC contrasts (55 and 52%).
Vowel contrast
Type
Correct (%)
SD
/ɪ/ - /iː/
CG
68
1.2
/e/ - /ɜː/
CG
70
2.1
/æ/ - /ʌ/
SC
52
1.3
/ɒ/ - /ɔː/
SC
55
1.6
Table 3. Percentages of correct responses and SDs with respect to
the discrimination of English vowel contrasts by Cypriot Greek
learners of English.
A one-way repeated measures ANOVA was carried out in order to examine the
effect of vowel contrasts on the discrimination scores with Discrimination Score as
the dependent variable (percentages of correct responses) and Contrast as the
within-subjects factor (4 vowel contrasts). The results showed a significant effect
of Contrast on the Discrimination Score [F(3, 72) = 5.69, p = 0.001, η2 = 0.13].
Further checks with the Bonferroni correction showed significant differences
between the /ɪ/ - /iː/ (M=68%) and /æ/ - /ʌ/ (M=52%) (p=0.001), /ɪ/ - /iː/ (M=68%)
and /ɒ/ - /ɔː/ (M=55) (p=0.001), /e/ - /ɜː/ (M = 70%) and /æ/ - /ʌ/ (M=52%)
(p=0.001), and /e/ - /ɜː/ (M = 70%) and /ɒ/ - /ɔː/ (M=52%) (p=0.001). Therefore,
the CG contrasts had significantly higher scores than the SC contrasts.
For determining the relationship between identification and discrimination scores,
Pearson r correlation tests were used. The sum of percentages of assimilation for
two contrastive vowels (e.g., /ɪ/ and /iː/) by each individual were the one factor
while the other factor included the percentages of correct discrimination responses
for the same vowel contrast (e.g., /ɪ/ - /iː/) by each individual. The findings showed
that there was a positive correlation between the identification and discrimination
scores for /ɪ/ - /iː/ (r = 0.52, p = 0.01), /æ/ - / (r = 0.34, p = 0.03), and /ɒ/ - /ɔː/ (r
= 0.39, p = 0.001), but a non-positive correlation for /e/ - /ɜː/ (r = 0.55, p = 0.13).
Therefore, for the majority of the contrastive pairs, the findings indicated that
individuals who were more consistent in their cross-language classification of
vowels were more apt to discriminate them accurately.
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 245
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
5. DISCUSSION
The present study investigated the assimilation of English vowels to the Cypriot
Greek phonological categories and the discrimination of English vowel contrasts
by Cypriot Greek children who learn English as a foreign language. In general, our
initial assimilation predictions were confirmed since they assimilated two or more
English vowels to a single phonological category of their native language: English
/ɪ iː/ were assimilated to Greek /i/, English /e ɜː/ to Greek /e/, English æ ɑː/ to
Greek /a/, English /ʊ uː/ to Greek /u/ and English /ɒ ɔː/ to Greek /o/. These
findings agree to a great degree with the results found by Lengeris (2009) who
investigated perceptual patterns of English vowels by Greek adults. This is to say
that perceptual patterns did not differ a lot between adult and children learners of
English since both populations of learners had similar English vowel assimilation
patterns in the two studies.
Also, as expected, learners did not establish new phonological categories for the
L2 vowels (but merged L1-L2 categories) since they still have limited linguistic
experience in the L2. It is widely showed in several cross-linguistic studies (e.g.,
Strange, 1995; Escudero, 2009) that speech perception is shaped by experience in
the fine-grained acoustics of a specific language environment. Escudero (2009)
pointed out that speech perception depends on language-specific acoustical
mappings and it is not performed solely by our general auditory system. It has to
be considered that the children of this study were attuned to the acoustic features
of their native language while the acquisition of speech stimuli in the L2 was very
limited since they were receiving English input mainly through the classroom. In
terms of the L2LP, this would be a “new scenario” since children assimilated many
L2 categories to a single L1 category and thus will face many perceptual
difficulties. In general, the findings of this study provide further evidence that
learners with an L1 vowel system that differs significantly in terms of phonetic
features with the L2 vowel system and which is smaller in size than the vowel
system of the L2 make multiple assimilations to a single L1 phonological category,
struggling to perceive much phonetic difference between two acoustically different
L2 sounds. Meunier et al. (2003) observed an effect of perception when the L1 of
the learners differed in formants but not in size. However, we cannot be sure
whether the perceptions of this study were influenced by the low density of the
Greek vowel system or by the phonetic differences between the vowel systems of
the two languages.
The most difficult vowel contrasts for the learners were the /ʌ/ - /æ/, /ʌ/ - /ɑː/, /ɑː/
- /æ/, and /ɒ/ - /ɔː/ since they could not perceive their between phonetic differences
246 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
as an influence from their already formed L1 phonetic units; we prompt that none
of these vowels is present in the Greek phonological system. According to the
PAM-L2, learners might be able to develop novel phonological categories for
these sounds in the future, and thus, discriminate their phonetic differences, under
specific circumstances. For example, phonetic training is believed to be efficient
for the improvement of both perceptual and production skills of the learners with
respect to L2 phones (Best and Tyler, 2007).
It is remarkable that contrast members that were longer were rated as better
exemplars of a Cypriot Greek phonological category than contrast members that
were shorter. For instance, the English /iː/ was judged as a significantly better
exemplar of the Greek /i/ (M=4.4) than the English /ɪ/ (M=2.9) despite the fact that
the English /ɪ/ has closest duration with that Greek vowel than the English /iː/. This
is inconsistent with our initial predictions that a temporal match between an L1 and
an L2 vowel would make the L2 vowel fit better on a spectrally close L1
phonological category. Perhaps, learners rely more on spectral cues during speech
perception, a conclusion that is in contrast with the findings reported in Lengeris
(2009). Also, we do not exclude that the articulation of long vowels might aid
listeners have a clearer picture of their acoustic features driving them in that way
to recognize these vowels as good exemplars of native phonological categories.
The findings are also offered for the evaluation of PAM-L2 discrimination
predictions. According to the results, the contrasts that signaled a CG assimilation
type could be discriminated with a moderate manner (~70%) while contrasts that
signaled a SC assimilation type had a poor discrimination (<60%). Also, the
statistical analysis showed significant differences for the discrimination scores of
CG and SC assimilations; CG assimilations could be discriminated more
accurately than the SC assimilations. These findings coincide with the hypotheses
of the PAM-L2 that predicts a moderate (to good) discrimination of the CG
contrasts and a poor one for the SC contrasts; this relationship can be depicted on
the following formula: CG > SC. This relationship is also evident in the results of
other studies such as Georgiou (2018b) and Mahmoud (2013). Thus, both within-
category assimilations, namely CG and SC may differ in terms of discrimination
accuracy depending on the perceived phonetic similarity of the L2 sounds that
were assimilated to the same L1 phonological category. Another important point is
that the cross-linguistic assimilation of L2 vowels to the learners’ L1 phonological
categories was related with their ability to discriminate L2 vowel contrasts.
Learners who were more consistent in the assimilation of two contrastive L2
vowels to their L1 system were more accurate in the discrimination of these
vowels. Therefore, it was detected a relationship between assimilation and
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 247
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
discrimination scores proposing that assimilation tests may be predictive of the
learners’ ability to discriminate L2 sound contrastive pairs.
6. CONCLUSIONS
This study indicated the interference of the Cypriot Greek vowel system in the
perception of the English vowels by Cypriot Greek children who learn English as a
foreign language in Cyprus. It was found that children relied on spectral cues in
order to assimilate L2 vowels to their L1 phonological system. Furthermore, it is
suggested that small size vowel inventories pose a degree of difficulty with respect
to the perception of nonnative sounds. The absence of much experience in the L2
is a significant factor that prevents the formation of new phonetic categories for L2
sounds that are dissimilar from any L1 sound. However, more factors that affect
L2 vowel perception have to be taken into consideration in future studies such as
learners’ proficiency level, vocabulary size in the L2, context of the stimuli
provided (e.g., preceding or following consonant), etc. Also, it is assumed that
children might not have the same metalinguistic capacity due to age-related
differences and therefore they might differ as well in speech perception; a future
perceptual study should take into consideration children’s metalinguistic skills.
Furthermore, the uncovering of these difficulties can offer significant feedback for
speech pedagogy. For example, educators can focus on the acoustical differences
of the aforementioned English vowel contrasts (e.g., by using minimal pairs) in
order to make students perceive these differences (Georgiou, 2019b). There is
evidence that minimal pair training sessions improve learners’ L2 pronunciation
accuracy (Haghighi and Rahimy, 2017; Tuan, 2010). Also, teaching material could
adapt to the needs of these learners and include activities that will enhance the
perception of the phonetic details of the English vowels. For instance, more
emphasis can be given on a vowel that is less perceivable than other vowels.
248 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
7. REFERENCES
ANTONIOU, M. (2010): One Head, Two Languages: Speech Production and
Perception in GreekEnglish Bilinguals, PhD dissertation, University of
Western Sydney.
ANTONIOU, M.; M. D. TYLER and C. T. BEST (2012): «Two ways to listen: Do L2-
dominant bilinguals perceive stop voicing according to language mode?»,
Journal of Phonetics, 40, pp. 582-594.
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 Methodological
Issues, Baltimore, York Press, pp. 171-204.
BEST, C. T. and W. STRANGE (1992): «Effects of phonological and phonetic
factors on cross-language perception of approximants», Journal of
Phonetics, 20(3), pp. 305-330.
BEST, C. T. and M. D. TYLER (2007): «Nonnative and second-language speech
perception: Commonalities and complementarities», in O.-S. Bohn and M.
J. Munro (eds.): Second Language Speech Learning: The Role of Language
Experience in Speech Perception and Production, Amsterdam, John
Benjamins, pp. 13-34.
BOERSMA, P. and D. WEENINK (2018). Praat: Doing Phonetics by Computer
[Computer software], http://www.fon.hum.uva.nl/praat/
BONGAERTS, T.; C. VAN SUMMEREN, B. PLANKEN and E. SCHILS (1997): «Age
and ultimate attainment in the pronunciation of a foreign language, Studies
in Second Language Acquisition, 19(4), pp. 447-465.
BOSCH, L.; A. COSTA and N. SEBASTIÁN-GALLÉS (2000): «First and second
language vowel perception in early bilinguals», European Journal of
Cognitive Psychology 12(2), pp. 189-222.
BUNDGAARD-NIELSEN, R. L.; C. T. BEST and M. D. TYLER (2011): «Vocabulary
size matters: The assimilation of second-language Australian English
vowels to first-language Japanese vowel categories», Applied
Psycholinguistics, 32(1), pp. 51-67.
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 249
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
DAWES, J. G. (2008): «Do data characteristics change according to the number of
scale points used? An experiment using 5 point, 7 point and 10 point
scales», International Journal of Market Research, 50(1), pp. 61-77.
DE CARA, B. and U. GOSWAMI (2003): «Phonological neighbourhood density:
Effects in a rhyme awareness task in five-year-old children», Journal of
Child Language, 30, pp. 695-710.
ESCUDERO, P. (2005): Linguistic Perception and Second-Language Acquisition:
Explaining the Attainment of Optimal Phonological Categorization, PhD
dissertation, Universiteit Utrecht.
ESCUDERO, P. (2009): «Linguistic Perception of ‘similar’ L2 sounds», in P.
Boersma and P. S. Hamann (eds.): Phonology in Perception, Berlin,
Mouton de Gruyter, pp. 151-190.
ESCUDERO, P.; B. SISINNI and M. GRIMALDI (2014): «The effect of vowel
inventory and acoustic properties in Salento Italian learners of Southern
British English vowels», The Journal of the Acoustical Society of America,
135(3), pp. 1577-1584.
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, Baltimore, York Press,
pp. 233-277.
FLEGE, J. E. and W. EEFTING (1987): «Production and perception of English stops
by native Spanish speakers», Journal of Phonetics, 15(1), pp. 67-83.
FLEGE, J. E.; E. M. FRIEDA and T. NOZAWA (1997): «Amount of native-language
(L1) use affects the pronunciation of an L2», Journal of Phonetics, 25(2),
pp. 169-186.
FOX, R. A.; J. E. FLEGE and M. J. MUNRO (1995): «The perception of English and
Spanish vowels by native English and Spanish listeners: A
multidimensional scaling analysis», Journal of the Acoustical Society of
America, 97(4), pp. 2540-2551.
250 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
FRIEDA, E. M. and T. NOZAWA (2007): «You are what you eat phonetically: The
effect of linguistic experience on the perception of foreign vowels», in O.-
S. Bohn and M. J. Munro (eds.): Language Experience in Second Language
Learning: In honor of James Emile Flege, Amsterdam, John Benjamins, pp.
79-96.
GEORGIOU, G. P. (2018a): Perception and Production of Greek Vowels by
Egyptian Arabic Learners of Greek as a Second Language, Phd
dissertation, University of Cyprus.
GEORGIOU, G. P. (2018b): «Discrimination of L2 Greek vowel contrasts: Evidence
from learners with Arabic L1 background», Speech Communication, 102,
pp. 68-77.
GEORGIOU, G. P. (2019a): «‘Bit’ and ‘beat’ are heard as the same: Mapping the
vowel perceptual patterns of Greek-English bilingual children», Language
Sciences, 72, pp. 1-12.
GEORGIOU, G. P. (2019b). «EFL teachers’ cognitions about pronunciation in
Cyprus», Journal of Multilingual and Multicultural Development, 40(6),
pp. 538-550.
GUION, S. G.; J. E. FLEGE, R. AKAHANE-YAMADA, and J. C. PRUITT (2000): «An
investigation of current models of second language speech perception: The
case of Japanese adults’ perception of English consonants», Journal of the
Acoustical Society of America, 107, pp. 2711-2724.
HACQUARD, V.; M. WALTER and A. MARANTZ (2007): «The effects of
inventoryon vowel perception in French and Spanish: An MEG study»,
Brain and Language, 100(3), pp. 295-300.
HAGHIGHI, M. and R. RAHIMY (2017): «The effect of L2 minimal pairs practice on
Iranian intermediate EFL learners’ pronunciation accuracy», International
Journal of Research in English Education, 2, pp. 42-48.
KARTUSHINA N. and U. H. FRAUENFELDER (2013): «On the role of L1 speech
production in L2 perception: Evidence from Spanish learners of French», in
Proceedings of the 14th InterSpeech Conference, Lyon, pp. 2118-2122.
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 251
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
KIM, D.; M. CLAYARDS and H. GOAD (2018): «A longitudinal study of individual
differences in the acquisition of new vowel contrasts», Journal of
Phonetics, 67, pp. 1-20.
KUHL, P. (1992): «Psychoacoustics and speech perception: internal standards,
perceptual anchors, and prototypes», in L. Werner, and E. W. Rubel (eds.):
Developmental Psychoacoustics, Washington DC, American Psychological
Association, pp. 293-332.
KUHL, P. K. (2000): «A new view of language acquisition», in Proceedings of the
National Academy of Sciences colloquium, Auditory Neuroscience:
Development, Transduction, and Integration, Irvine, Arnold and Mabel
Beckman Center, pp. 11850-11857.
KUHL, P. K.; E. STEVENS, A. HAYASHI, T. DEGUCHI, S. KIRITANI and P. IVERSON
(2006): «Infants show a facilitation effect for native language phonetic
perception between 6 and 12 months», Developmental Science, 9, pp. F13-
F21.
KUHL, P. K. and P. IVERSON (1995): «Linguistic experience and the perceptual
mag- net effect», in W. Strange (ed.): Speech Perception and Linguistic
Experience: Issues in Cross-Language Research, York Press, Baltimore,
pp.121-154.
LENGERIS, A. (2009): «Perceptual assimilation and L2 learning: Evidence from the
perception of Southern British English vowels by native speakers of Greek
and Japanese», Phonetica, 66, pp. 169-187.
MAHMOUD, M. (2013): «Discrimination of Arabic contrasts by American
learners», SSLLT, 3(2), pp. 261-292.
MEUNIER, C.; C. FRENCK-MESTRE, T. LELEKOV-BOISSARD and M. LE BESNERAIS
(2003): «Production and perception of foreign vowels: Does the density of
the system play a role?», in Proceedings of the 15th International Congress
of Phonetic Sciences, pp. 723-726
MURPHY, W. D.; S. L. SHEA, and R. N. ASLIN (1989): «Identification of vowels in
‘vowelless’ syllables by 3-year-olds», Perception & Psychophysics, 46, pp.
375-383.
252 G. P. Georgiou
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
OHDE, R. N.; K. L. HALEY and C. W. MCMAHON (1996): «A developmental study
of vowel perception from brief synthetic consonant-vowel syllables»,
Journal of the Acoustical Society of America, 100, pp. 3813-3824.
ROACH, P. (2004): «British English: Received Pronunciation», Journal of the
International Phonetic Association», 34(2), pp. 239-245.
SEBASTIÁN-GALLÉS, N. and S. SOTO-FARACO (1999): «On-line processing of
native and non-native phonemic contrasts in early bilinguals», Cognition,
72(2), pp. 111-123.
SNOW, C. E. (1987): «Relevance of the notion of a critical period to language
acquisition», in M. H. Bornstein (ed.): Sensitive Periods in Development:
Interdisciplinary Perspective, Hillsdale, Lawrence Erlbaum, pp. 183-209.
STRANGE, W. (1995): «Cross-language studies of speech perception a historical
review», in W. Strange (ed.): Speech Perception and Linguistic Experience,
Timonium, York Press, pp. 3-48.
STRANGE, W. (2011) «Automatic selective perception (ASP) of first and second
language speech: a working model», Journal of Phonetics, 39, pp. 456-466.
SUN, L. and J. VAN HEUVEN (2007): «Perceptual assimilation of English vowels by
Chinese listeners. Can native-language interference be predicted?», in B.
Los and M. van Koppen (eds.): Linguistics in the Netherlands 2007,
Amsterdam, John Benjamins, pp. 150-161.
TUAN, L. T. (2010): «Teaching English discrete sounds through minimal pairs»,
Journal of Language Teaching and Research, 1(5), pp. 540-561.
TYLER, M. D.; C. T. BEST, A. FABER and A. G. LEVITT (2014): «Perceptual
assimilation and discrimination of non-native vowel contrasts», Phonetica,
71(1), pp. 4-21.
VAN LEUSSEN, J.-W. and P. ESCUDERO (2015): «Learning to perceive and
recognize a second language: The L2LP model revised», Frontiers in
Psychology, 6, pp. 1-12.
Cue-reliance and vowel perceptual patterns of Cypriot Greek children… 253
EFE, ISSN 1575-5533, XXVIII, pp. 229-253
WALLEY, A. C. (2005): «Speech perception in childhood», in D. B. Pisoni and R.
E. Remez (eds.), Handbook of Speech Perception, Oxford, Blackwell, pp.
449-468.
WERKER, J. and R. TEES (1984): «Cross-language speech perception: Evidence for
perceptual reorganization during the first year of life», Infant Behavior and
Development, 7, pp. 49-63.
... Also, it seeks to find whether perceptual training has an impact on the learners' productions. The motivation of training Greek speakers on the acquisition of English vowels emerges from the difficulties of these speakers in discriminating L2 English vowel contrasts as reported in the studies of Georgiou (2019aGeorgiou ( , 2019b for Greek children and Lengeris (2009) for Greek adults. These studies indicate a clear interference of the learners' L1 in the perception of English vowels, suggesting that several English vowels are assimilated to a single Greek phonological category due to the smaller size of the Greek vowel system compared to the English system. ...
... These studies indicate a clear interference of the learners' L1 in the perception of English vowels, suggesting that several English vowels are assimilated to a single Greek phonological category due to the smaller size of the Greek vowel system compared to the English system. Specifically, the Greek vowel system consists of 5 vowels /i e a o u/, while the Received Pronunciation (RP) vowel system, which is usually taught in English as a foreign language classrooms in Cyprus, includes 11 monophthongs which can be considered either lax /ɪ ʊ e ae ʌ ɒ/ or tense /iː uː ɜː ɔː ɑː/ (Georgiou, 2019a(Georgiou, , 2019b. We assume that phonetic training will be able to improve to some extent the identification of vowels. ...
... The stimuli of the perceptual training included monosyllabic real English words produced by 6 new adult British English speakers (3 males and 3 females). These words included the 11 English monophthongs which were divided into 5 different clusters of 2 or 3 vowels: /ɪ iː/ (e.g., sit, seat), /e ɜː/ (e.g., ten, turn), /ae ʌ ɑː/ (e.g., pat, putt, part), /ɒ ɔː/ (e.g., cot, caught), and /uː ʊ/ (e.g., fool, full) based on the English vowel assimilation patterns of Cypriot Greek speakers as reported in Georgiou (2019aGeorgiou ( , 2019b. Specifically, it was found that the aforementioned English vowel clusters were mapped in the Cypriot Greek phonological categories /i/, /e/, /a/, /o/, and /u/ respectively. ...
Article
The purpose of this study is to investigate the effect of auditory perceptual phonetic training on the identification and production of English vowels by Cypriot Greek children and adults. Another two groups of Cypriot Greek child and adult speakers served as controls. The trained groups participated in the pretest, training, and posttest phase, while the controls completed only the pretest and posttest phase. The results showed that perceptual training improved identification accuracy, with children showing greater gains than adults. Although the performance of adults was poorer than the performance of children, their phonological system did undergo substantial alteration through perceptual phonetic training as they significantly improved their identifications in the posttest. Also, the results support a common mental space for the speech perception and production domains since the perceptually-oriented training affected the learners’ productions. However, transfer of improvements in production was observed only to some extent in children and not in adults, suggesting that training has an impact mostly on the trained modality and that some production improvement after perceptual training might be more evident in younger learners.
... So, nonnative speakers often produce accented speech in the L2 which is easily detected by the native speakers of the target language (Georgiou, 2019b). These difficulties are attributed to the differences between the first language (L1) and the L2 phonological systems since the learners' L1 interferes in the L2 (Best & Strange, 1992;Flege et al., 1997;Flege & Eefting, 1987;Georgiou, 2019aGeorgiou, , 2019c. For instance, adult Estonian learners of Spanish who permanently lived in Spain had no problem producing the Spanish vowels /i e o u/ since they are present in their L1 vowel system, but it was challenging for them to produce with a native-like manner the Spanish low vowel /a/ because it is absent from their native vowel system; they produced it closely to the Estonian /ɑ/ (Leppik & Lippus, 2014). ...
Article
The present study aims to investigate the relationship between perceived cross-linguistic similarity and second language (L2) production. To this purpose, Egyptian Arabic learners of Greek in Cyprus who took part in a previous cross-linguistic perceptual study, completed a production test with respect to the Cypriot Greek vowels. The findings showed that perceived cross-linguistic similarity was linked with L2 production since along with the consideration of first language (L1)-L2 acoustic differences, it predicted most of the L2 vowel productions. Also, many L2 vowels were considerably longer than the corresponding L1 vowels. This can be interpreted as an L1 transfer since Egyptian Arabic vowels are longer in duration than the Cypriot Greek vowels. An interesting finding was that the production of the L2 vowels had only partial overlap with the productions of the L1 vowels, a finding that provides support for the hypotheses of the Speech Learning Model.
... Many studies indicate the difficulty of adult learners in acquiring native-like pronunciation 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 perception 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). ...
Article
Full-text available
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 generalization test showed significant improvement in the discrimination of most vowel contrasts (both stressed and unstressed). The findings suggest that HVPT may facilitate the formation 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.
... These difficulties are attributed to the fact that specific phones may not contrast in the L1 phoneme repertoires of the learners, making new L2 contrasts difficult to perceive and produce. Examples of contrasts may be the vowels found in 'bet-bat' for Dutch speakers (Boersma, 2005), 'rocket-locket' for Japanese speakers (Aoyama et al., 2004), and 'beat-bit' for Spanish (Flege et al., 1997), Portuguese (Rauber et al., 2005), and Greek speakers (Georgiou, 2019a(Georgiou, , 2019c. ...
Article
Full-text available
The present study aims to examine the Cypriot Greek consonant perceptual patterns of Russian speakers who permanently live in Cyprus and do not have experience with Greek, and to determine the effect of naturalistic exposure in an unknown language on speech perception. The experimental group consisted of 16 Russian adult speakers who were living permanently in Cyprus and had no knowledge of Greek, and the control group consisted of 16 Russian adult speakers who were living permanently in Russia, and who did not have knowledge of Greek either. Both populations completed an assimilation test and an AXB discrimination test. The results showed that there were no important differences for the assimilation patterns of both Russian populations, while the discrimination accuracy over Cypriot Greek consonantal contrasts was either similar between the two groups or better for the Russian population living in Russia. Thus, exposure to naturalistic second language stimuli without knowledge of the second language does not contribute to the attunement to phonetic invariants that characterize the second language phones.
... Acquiring the sounds of a second (L2) or a foreign language might be challenging for listeners. This difficulty emerges from the fact that nonnative sounds pass through filtering from the listeners' first language (L1) before being mapped into existing phonological categories of their native language (Best, 1995;Best and Tyler, 2007;Escudero, 2009;Georgiou, 2020aGeorgiou, , 2019bFlege, 1995). For example, although Japanese learners of English are familiar with conversational English, they struggle to perceive the acoustical differences between the English contrastive consonants /r/-/l/ (Miyawaki et al., 1975). ...
Article
Nonnative sound perception might be challenging for adult listeners since they attune from a very young age to the phonological aspects of their native language and, thus, every nonnative sound is filtered through their first language. The present study investigates the perception of Greek consonants in both Consonant-Vowel (CV) and Vowel-Consonant (VC) syllable context by Russian monolingual speakers. To this purpose, 16 Russian speakers aged 19-26 who are students in a University in Moscow and have no knowledge of Greek participated in the study. The participants were tested in an assimilation task in which they were asked to assimilate Greek consonants to their native language phonological categories and in an AXB discrimination task to examine their ability in discriminating nonnative consonantal contrasts. The predictions of the study were developed using the framework of the Perceptual Assimilation Model (PAM). In the assimilation task, the Russian speakers assimilated, as expected, the Greek consonants [t d g x] to the very similar Russian phonological categories [t d g x]. The Greek consonants [θ ð] were merged with the Russian categories [f z] respectively, while the Greek [ç] fell into the Russian phonological category [x j ]. By contrast, the Greek consonant [ɟ] was not perceived as an instance of any phonological category of the listeners' first language. With respect to the discrimination task, the discrimination scores of the Greek contrasts [θ]-[t], [ð]-[d], [ɟ]-[g], and [ç]-[x] ranged from very good to excellent. The findings indicate that Russian speakers assimilated consonants that are not present in their first language in different phonological categories than those proposed by the literature, and that the feature of consonant palatalization that is present in Russian, assisted the listeners to more consistently identify the Greek palatal fricatives. Moreover, the assimilation of Greek consonants differed in some cases for CV vs. VC context and the discrimination of one Greek contrast was different in CV vs. VC context. In general, the results demonstrate that acoustic features found in the speakers' first language interfered with the perception of nonnative consonants, and that the assimilation of nonnative fricatives (i.e., [θ ð]) to Russian fricative categories (i.e., [f z]) can be explained by the conformation of the speakers' native language to the faithfulness constraint, which led the speakers to preserving the continuant feature.
... It is usually difficult for adult speakers to accurately perceive the acoustical information of non-native phones due to the effect of their first language (L1) (Best, 1995;Flege, 1995;Best and Tyler, 2007;Escudero, 2009;Georgiou, 2018;2019a2019b. A very well cited example is the discrimination of the English consonantal contrast /r/-/l/ by Japanese learners of English. ...
Article
Full-text available
The purpose of this study is to investigate the discriminability of two different assimilation types, the Uncategorized-Categorized (UC) and the Uncategorized-Uncategorized assimilation (UU) (Best and Tyler, 2007), as reflected in the discrimination accuracy and reaction times towards non-native contrasts by Russian speakers. The discriminability of these assimilation types varies in the literature. To this purpose, the same Russian speakers who evaluated Greek consonantal contrasts as UC and UU types in an assimilation test of a previous study completed an AXB discrimination test in this study to detect the discriminability of these assimilation types. The findings demonstrated that most of the UU non-overlapping (UU-N) types, and specifically those with focalized-focalized responses, were more accurately discriminated and had faster RTs than the UC non-overlapping (UC-N) type. However, one UU-N type with clustered-clustered responses did not differ in terms of discrimination accuracy and reaction times with the UC-N type. It is suggested that despite having the same overlapping parameters (non-overlapping), UU types might be more discriminable than UC types with respect to consonants. Also, similarity of uncategorized phones with other assimilated phones (e.g., focalized, clustered, dispersed) might shape the UC-UU type relationship. Finally, it is assumed that the discriminability of UC-UU types might be consonant-specific.
... If an L2 sound or a pair of L2 sounds is similar to an L1 sound, then learners might face difficulties in both perceiving and producing correctly the L2 sounds. There are few studies in regard to the phonological patterns of Greek speakers who learn a nonnative language (e.g., Georgiou 2019aGeorgiou , 2019b, and similarly few studies related with L2 errors of foreign learners of Greek. For example, Anastiasiadi-Simeonidi et al. (2010) recorded the phonetic/ phonological errors of Russian speakers who learn Greek as an L2, identifying problems with the production of sounds that are not present in their L1. ...
Article
Full-text available
The present study aims to identify the most important language errors in the speech of a second language (L2) Greek speaker, and explain these errors on the basis of the speaker’s first language (L1) interference, the speakers’ sociolinguistic background, and the effect of language mechanisms. The study employs the data of a semi-structured interview from a Swiss German speaker who permanently lives in Cyprus for 13 years and speaks Greek as an L2. The recordings of the interview were transcripted to identify possible errors in the phonetic/phonological, morphological, syntactical, and semantic level. The results showed errors in every language level indicating a clear interference of the speaker’s L1 with the L2, an overgeneralization of grammatical rules, and an interference of the local dialect in the L2 productions.
Chapter
The present study examines how contrastive L2 English plosives on syllable-level are perceived by Cypriot Greek speakers and how the use of L2 affects these perceptions. The participants were 30 university students with sufficient knowledge of English as an L2. They completed two forced-choice discrimination tasks in which they had to discriminate three different types of English plosives contrasted in voicing in both closed (CVC) and open syllables (VC); plosive position was also taken into consideration. The findings indicated that participants yielded poor to moderate discrimination accuracy over the L2 plosive contrasts under investigation. Also, they performed significantly better in the open-syllable task, while they were proceeding to more errors in terms of plosives in the coda position in closed syllables. L2 use did not have an effect on the discrimination of contrastive English plosives. Together, the data suggest that syllables have an important role in the segmental perception of spoken L2 English, while L1 transfer and universality of phone perception on a syllabic level may be influential factors.
Article
Several speech models have been formed in the past aiming to predict the abilities of nonnative listeners or learners in perceiving and producing speech sounds. The present paper proposes a new model for speech perception, the Universal Perceptual Model of Second Language (henceforth, UPM). UPM assumes that second language phone acquisition is strongly affected by the speakers' native language but still the window of phone learning is open due to the universality of speech sounds. Also, it supports that second language phones are initially activated as disoriented phonetic units. In this paper, we provide some initial insights into the predictability of the model. UPM uses degrees of overlap and chance criteria to form its predictions. We recruited Cypriot Greek novice learners of Italian who participated in two psychoacoustic tasks in which they classified and discriminated Italian vowels, respectively. The findings demonstrated that the degree of overlap between two nonnative phones may be a good predictor of the speakers' discrimination accuracy over these phones. UPM might be a useful model which aims to better explain speech perception mechanisms and patterns of speech acquisition.
Article
Previous research has shown that an increased second language (L2) vocabulary size leads to better attunement to the cues required to distinguish L2 contrastive phones. This has been the central tenet of the vocabulary-tuning model (vocab) on the basis of evidence by Japanese learners of English in Australia. We aim to test the validity of the aforementioned hypothesis by extending the research for learners with a different first language (L1) background and learners who do not have naturalistic access to the L2 input (i.e., learn the L2 through a controlled foreign classroom setting). To this purpose, 28 Russian speakers, who were learning English in Russia at the time, participated in two psychoacoustic tests in which they were asked to assimilate L2 vowels to their L1 phonological system and discriminate vowel contrasts respectively. The participants were divided into two groups according to their vocabulary size in English; comprising the small vocabulary (SV) and the high vocabulary (HV) groups. The results showed that the HV group demonstrated similar assimilation scores to the SV group. However, the HV group was able to perceive within-category differences and more accurately discriminate specific pairs of English vowel contrasts in comparison to the SV group. The findings are partially consistent with the central hypothesis of the Perceptual Assimilation Model-L2 and the vocab model as the expansion of L2 vocabulary was linked with better attunement to phonetic differences in the L2. Another important finding is that a more developed vocabulary results in fine-tuning to L2 phonetic differences, even in a restricted L2 learning setting [work supported by the "RUDN University Program 5-100"].
Chapter
Full-text available
The aim of our research is to understand how speech learning changes over the life span and to explain why "earlier is better" as far as learning to pronounce a second language (L2) is concerned. An assumption we make is that the phonetic systems used in the production and perception of vowels and consonants remain adaptiive over the life span, and that phonetic systems reorganize in response to sounds encountered in an L2 through the addition of new phonetic categories, or through the modification of old ones. The chapter is organized in the following way. Several general hypotheses concerning the cause of foreign accent in L2 speech production are summarized in the introductory section. In the next section, a model of L2 speech learning that aims to account for age-related changes in L2 pronunciation is presented. The next three sections present summaries of empirical research dealing with the production and perception of L2 vowels, word-initial consonants, and word-final consonants. The final section discusses questions of general theoretical interest, with special attention to a featural (as opposed to a segmental) level of analysis. Although nonsegmental (i.e., prosodic) dimensions are an important source of foreign accent, the present chapter focuses on phoneme-sized units of speech. Although many different languages are learned as an L2, the focus is on the acquisition of English.
Article
Full-text available
This study explores how individuals’ second language cue weighting strategies change over time and across different contrasts. The study investigates the developmental changes in perceptual cue weighting of two English vowel contrasts (/i/-/ɪ/ and /ɛ/-/æ/) by adult and child Korean learners of English during their first year of immersion in Canada. Longitudinal results revealed that adult learners had an initial advantage in L2 perceptual acquisition over children at least for the /i/-/ɪ/ contrast, but after one year some children showed greater improvements especially on the more difficult /ɛ/-/æ/ contrast. Both groups of Korean learners showed different acquisition patterns between the two vowel contrasts: they used both spectral and duration cues to distinguish /i/-/ɪ/ but generally only duration to distinguish /ɛ/-/æ/. By examining cue weights over time, this study partially confirmed the hypothesized developmental stages for the acquisition of L2 vowels first proposed by Escudero (2000) for Spanish learners of English. However, some unpredicted patterns were also identified. Most importantly, the longitudinal results suggest that individual differences in cue weighting are not merely random variability in the learner’s response patterns, but are systematically associated with the developmental trajectories of individual learners and those trajectories vary according to vowel contrast.
Article
It has been previously proposed that small native vowel inventories impede the acquisition of foreign language vowels that come from larger inventories. The present study aims to investigate how the already-formed Greek phonological system is modified after the perception of the English vowels in a non-naturalistic environment and to what extent Greek speakers are able to discriminate challenging English vowel contrasts. Also, it aims to examine if learning experience enhances sensitivity to acoustical information of the foreign language vowels. The study relies on the theoretical framework of the Perceptual Assimilation Model-Second Language (PAM-L2). For developing predictions about the perception of the English vowels by Greek speakers, the vowel spaces of both Greek and English native speakers were investigated; 3 native speakers of Greek and 3 native speakers of English produced vowels in their native language. 26 Greek learners of English as a foreign language in Cyprus with an age that varied from 8 to 12 years participated in the perceptual study. Learners were divided into two groups according to their proficiency level in English. The participants took an identification test in order to determine how English vowels were assimilated to the listeners' native phonological categories and then they were tested in an AXB discrimination task in order to investigate to what extent they are able to discriminate English vowel contrasts. The results showed that several pairs of two or more English vowels were assimilated to a single Greek phonological category. Furthermore, the discrimination test showed poor to moderate discrimination accuracy for both groups regarding the English vowel contrasts, yet differences in the discrimination accuracy of the contrasts between the novice and the more advanced group of learners were minimal; only the English /e/-/ɜː/ contrast was discriminated slightly better by the advanced learners. Thus, learners with a higher proficiency level did not generally perceive the English vowels better than learners with a lower proficiency level, signifying that perception of foreign language vowels is not merely a matter of amount of exposure to the foreign stimuli. Conclusions are drawn about the interference of native Greek with the learning of English vowels and the acquisition of the foreign language stimuli in a classroom environment.
Article
Neglecting pronunciation in the instruction of foreign languages is quite common both worldwide [Derwing, T. M., and M. J. Murno, 2005. "Second Language Accent and Pronunciation Teaching: A Research-Based Approach." TESOL Quarterly 39 (3): 379-397] and in Greek-Cypriot classrooms [Kyprianou, M. 2015. "Teaching and Learning English Pronunciation in Cyprus: Survey, Analysis and Challenges." PhD diss., University of Cyprus]. The present study aims to investigate cognitions, challenges and practices of various age groups of English as a Foreign Language (EFL) teachers in Cyprus in relation to the teaching of pronunciation in an effort to interpret the low interest that pronunciation receives into classroom. To this purpose, 98 Greek-Cypriot EFL teachers were divided into three groups according to their age and they were called to complete an online survey. The results indicated that the older age groups reported different cognitions in relation to the youngest group; e.g. they believed that other teaching aspects are more important than pronunciation. Also, the three groups faced many similar challenges in the teaching of pronunciation such as limited time and lack of training, and they were using pronunciation practices almost in the same extent. Implications are drawn regarding the development of a stronger EFL pronunciation teaching in the Greek-Cypriot public schools. ARTICLE HISTORY
Article
The present study investigates the assimilation of the Cypriot Greek (CGR) vowels to the phonological categories of Egyptian Arabic (EA) as well as the discrimination of 2 stressed and 2 unstressed Cypriot Greek vowel contrasts by native speakers of Egyptian Arabic. It also intends to test the discriminability of the assimilation types according to the predictions of the Perceptual Assimilation Model (PAM). 15 adult female immigrants, who permanently live in Cyprus for 4-5 years and were taught Greek in formal educational settings, completed a task for the assimilation of the second language (L2) vowels to the phonological space of their native language; followed by an AXB test for the detection of the discrimination accuracy over second language vowel contrasts. The discrimination scores of the second language learners were compared with the ones of the native speakers of Cypriot Greek, who constituted the control group. The results indicated a Category Goodness (CG) difference assimilation type for 3 of the vowel contrasts (stressed /i/-/e/, stressed /o/-/u/ and unstressed /o/-/u/) and a Two Category (TC) assimilation for one of the contrasts (unstressed /i/-/e/) as a consequence of the effect of their native language. A moderate discrimination for the contrasts which reported a Category Goodness difference assimilation type was observed while the Two Category assimilation type reported an excellent discrimination by the second language learners. The results agree with the Perceptual Assimilation Model's assumptions about the predictions of the aforementioned assimilation types, also showing a success of the model in the prediction of the perception patterns of experienced second language learners. Finally, the role of stress on the perception of second language contrasts was important since the stressed contrast /i/-/e/ was discriminated in a different manner in comparison to the corresponding unstressed one.
Thesis
The purpose of this thesis is the investigation of the perception and production of the Cypriot Greek vowels by Egyptian Arab learners of Greek as a second language (L2). The particular group of adult learners has been taught Greek through formal education settings (schools, universities) living as well permanently in a country where Greek is dominant. Moreover, the study aims to show the effect of pedagogical intervention (vowel instruction/training) on the perception and production of Greek vowels by adult L2 learners. The thesis employs the theoretical hypotheses of two speech models: the Speech Learning Model (SLM) and the Perceptual Assimilation Model (PAM). The present study constitutes the first cross-linguistic study which examines the perception and production of Greek phones by learners with Arabic first language (L1) background. The results showed a difficulty in the discrimination of the Greek contrastive vowels /i/-/e/ (only in stressed position) and /ο/-/u/ as well as inaccurate production of several Greek vowels. These results corroborate previous findings in the literature, demonstrating that L2 vowel perception and production are subject to L1 influence. Within the thesis, it was identified that the relationship between perception and production was not straightforward since the excellent perception of the L2 vowels did not report a corresponding accurate production. Although PAM’s hypotheses are based on naïve L2 learners, they were also confirmed for the experimental group of this study, which consisted of experienced L2 learners. Moreover, SLM managed to correctly predict the L2 vowel productions, but it did not draw clear predictions about their perception. In addition, the improvements observed in both perception and production of the Greek vowels by the L2 learners after the implementation of vowel training and instruction, indicate first, the importance of the usage of such practices during the learning of the L2 vowels, and second, the possibility of L2 phone acquisition even by adult learners. Linguistic findings can be also interpreted on a pedagogical basis as they can be employed for the preparation of the appropriate vowel training and instruction, with emphasis on the particular vowels or vowel pairs that seemed problematic in both perception and production. Since pedagogical intervention improved both the perception and production of the L2 vowels, it is recommended the implementation of specific phonetic and phonological awareness exercises and activities in the L2 curriculum in formal pedagogical settings.
Article
This article reports on second language perception of non-native contrasts. The study specifically tests the perceptual assimilation model (PAM) by examining American learners’ ability to discriminate Arabic contrasts. Twenty two native American speakers enrolled in a university level Arabic language program took part in a forced choice AXB discrimination task. Results of the study provide partial evidence for PAM. Only two-category contrasts followed straightforwardly from PAM; discrimination results of category-goodness difference and both uncategorizable contrasts yielded partial support, while results of uncategorized versus categorized contrast discrimination provided counter-evidence to PAM.