Bilingualism: Language and Cognition 12 (1), 2009, 83–95 

C 2008 Cambridge University Press doi:10.1017/S136672890800388X 83

PA N O S AT H A NA S O P O U L O S
Cognitive representation of Bangor University
colour in bilinguals: The case
of Greek blues∗

A number of recent studies demonstrate that bilinguals with languages that differ in grammatical and lexical categories may
shift their cognitive representation of those categories towards that of monolingual speakers of their second language. The
current paper extended that investigation to the domain of colour in Greek–English bilinguals with different levels of
bilingualism, and English monolinguals. Greek differentiates the blue region of colour space into a darker shade called ble
and a lighter shade called ghalazio. Results showed a semantic shift of category prototypes with level of bilingualism and
acculturation, while the way bilinguals judged the perceptual similarity between within- and cross-category stimulus pairs
depended strongly on the availability of the relevant colour terms in semantic memory, and the amount of time spent in the
L2-speaking country. These results suggest that cognition is tightly linked to semantic memory for specific linguistic
categories, and to cultural immersion in the L2-speaking country.

1. Introduction In fact, there is a wealth of recent empirical evidence

to support this view, most notably from the fields
At what point does an oar become a winnowing fan? This
of spatial cognition (Levinson, 1996; Levinson, Kita,
is what the hero in Homer’s Odyssey has to find out, travel-
Haun and Rasch, 2002; Majid, Bowerman, Kita, Haun
ling inland carrying an oar on his shoulder, until he reaches
and Levinson, 2004), grammatical number and object
a place where one of its inhabitants will ask him why he is
classification (Lucy, 1992; Imai and Gentner, 1997; Lucy
carrying a winnowing fan on his shoulder. There, he is to
and Gaskins, 2001, 2003; Imai and Mazuka, 2003), and
fix the oar in the ground and make sacrifices to the gods in
colour categorisation (Davidoff, Davies and Roberson,
order to appease them. Thus, the idea that the way we con-
1999; Roberson, Davies and Davidoff, 2000; Roberson,
ceptualise reality and the world around us is tightly linked
Davidoff, Davies and Shapiro, 2005; Roberson, 2005).
to language and culture is indeed ancient, as this example
If language-specific concepts direct attention to
from classical literature demonstrates. Related to this ob-
certain features of stimuli more than others, then it
servation is the classic issue of the linguistic relativity hy-
is central to the investigation of human cognition to
pothesis: do speakers of different languages think differ-
know how attention is modulated in bilinguals whose
ently as a result of encoding and using different concepts?
languages partition reality in different ways (Green,
A popular misinterpretation of the linguistic relativity
1998; Pavlenko, 2005a). The current paper takes the
hypothesis is to assume that there is a “strong” version,
first steps towards examining bilingual cognition in
which holds that our thought is constrained by our
the domain of colour, thus complementing the recent
language, and a “weak” version, which holds that our
investigation of linguistic relativity and bilingualism in
thought is not limited by our linguistic concepts, but we
a range of other domains (e.g. grammatical number and
pay more attention to those aspects of the world that
object classification: Athanasopoulos, 2006, 2007; Cook,
are encoded in our language than to those that are not
Bassetti, Kasai, Sasaki and Takahashi, 2006; emotion:
(see e.g. Hunt and Agnoli, 1991; Carroll, 2008). This
Pavlenko, 2005b, 2006; time: Boroditsky, 2001; gender:
distinction neither reflects Sapir’s (1921) and Whorf’s
Boroditsky, Schmidt and Phillips, 2003; artefact naming
(1956) original arguments, who never actually claimed
and categorisation: Malt and Sloman, 2003; Ameel,
that language constrains a speaker’s worldview, nor their
Storms, Malt and Sloman, 2005; note also earlier work on
recent revival by Lucy (1992), Levinson (1996), Roberson
person cognition by Hoffman, Lau and Johnson, 1986).
(2005) and others. The only possible version that faithfully
The paper is organised as follows. In the next two
reflects the original Whorfian arguments is the so-called
sections, the issue of linguistic relativity in relation to
“weak” one – that language acts as an attention-directing
colour cognition is discussed, presenting past and recent
mechanism to specific perceptual attributes of reality.
evidence to bear on the debate, as well as the emerging
* I am grateful to Vivian Cook, Debi Roberson, Aneta Pavlenko and issue of linguistic relativity in relation to bilingualism.
Terry Regier. Any faults that remain are entirely my own. Then the paper will examine naming behaviour and
Address for correspondence:
Department of Linguistics and English Language, Bangor University, Bangor, Gwynedd LL57 2DG, UK
panos@bangor.ac.uk
84 Panos Athanasopoulos

categorical perception of colour in bilinguals with way we judge the similarity or the difference between two
languages that differ in their lexical division of the blue items is determined by how our language carves the world
region of colour space. Finally, results are discussed, into nameable parts (Roberson et al., 2000).
focusing on the impact of several variables that may affect These findings do not mean that there are no
bilingual cognition, and that may in turn have more general physiological and perceptual constraints common to all
implications for conceptual representation and for the way humans. For example, there are no attested languages
linguistic categories influence human cognition. which have terms for “blue” and “yellow”, but not for
the intermediate category of “green” (Roberson, 2005).
Nevertheless, language influences colour categorisation
1.1 Linguistic relativity and colour categorisation
by directing our attention to those categories that
The field of colour categorisation has been a classic scien- are encoded in it. For example, a recent study by
tific battleground for the linguistic relativity hypothesis. Roberson et al. (2005) found that speakers of a language
Until the 1970s, the prevalent view was that there were with only five basic colour terms showed differential
no constraints on the way languages encoded colour (see cognitive representation of colour from speakers of
e.g. Gleason, 1961). Consequently, colour categorisation another language which also had five basic colour terms.
was viewed as one of the most important fields where Participants showed categorical perception (judging two
the influence of linguistic diversity on cognition was stimuli to be more similar if they fall within a category
manifested. For example, Brown and Lenneberg (1954) boundary than if they cut across the boundary) for the
observed that Zuni does not separate lexically the colours specific colour areas encoded in their respective language,
“orange” and “yellow”, and instead uses a single term suggesting that even if languages have the same number
to describe them. In a recognition memory task, Zuni of basic colour terms, their precise boundaries in colour
speakers did not distinguish between the two colours as space may differ, and yield differential cognitive patterns.
accurately or as frequently as English speakers did. More recently, the debate has moved beyond the tradi-
In the 1970s, the tide changed in the opposite direction, tional universality vs. relativity views to a more mediating
mainly because of the influential studies by Berlin and Kay position (see e.g. Kay and Regier, 2006). Regier, Kay and
(1969) and Rosch-Heider (1972). Berlin and Kay (1969) Khetarpal (2007), based on a proposal by Jameson and
obtained samples of colour areas and category foci (best D’Andrade (1997), empirically demonstrated that colour
exemplars) from native speakers of 20 different languages, naming is determined by universal perceptual constraints,
including English. The researchers found no differences such that focal colours are indeed linguistically and
between the participants as a function of language perceptually more salient than non-focal colours across
background, concluding that colour categorisation is not cultures. On the other hand, there also exists considerable
language-specific and that the foci of colour terms are cross-linguistic variation at the demarcation of category
universal. Explicit in their account is the assumption boundaries, i.e. linguistic convention influences where
that languages should have between two and eleven speakers of different languages partition colour space.
basic colour terms, the number determined by a specific This view accommodates the existence of BOTH universal
evolutionary path each language may take when adding constraints on colour naming AND the influence of cross-
new colour categories. Rosch-Heider’s (1972) study found linguistic differences on colour cognition (Kay and Regier,
no differences in memory for colour between speakers 2007).
of English and speakers of a language with only two
basic colour terms, Dani. Both groups of speakers showed
1.2 Linguistic relativity and bilingualism
better recognition memory for focal stimuli (i.e. near the
focus of each of the basic categories reported in Berlin Since there is now a range of converging evidence
and Kay, 1969) than non-focal stimuli. These results, to suggest the tight relationship between linguistic
in conjunction with Berlin and Kay’s (1969) findings, concepts and cognition, the issue of bilingualism becomes
seriously undermined the scientific basis of the linguistic increasingly important. For example, in Berlin and Kay’s
relativity hypothesis. (1969) seminal study the majority of participants were
However, subsequent studies by Davidoff et al. (1999) in fact bilingual in English and were living in the US.
and Roberson et al. (2000) did not replicate Rosch- However, if bilingualism affects cognitive representation
Heider’s (1972) results with speakers of English and of categories, it is difficult to interpret the behaviour of
speakers of the Berinmo tribe in Papua New Guinea. bilingual individuals in their study: is their English-like
In fact, the researchers found that linguistic categories behaviour a sign of universality of categories and foci, or
for colour influenced cognitive representation of colour. the result of semantic shift of their first language (L1)-
Both Berinmo and English speakers judged two colours specific categories and foci towards the second language
to be more similar if they shared the same name in their (L2)? (For criticisms of Berlin and Kay, 1969, see Ratner,
respective languages. The researchers concluded that the 1989; Saunders and van Brackel, 1997; Lucy, 1997;
 Bilingual blues 85

Pavlenko, 2005a, 2005b. Berlin and Kay subsequently Athanasopoulos and Kasai (2008) showed that knowledge
acknowledged some of these shortcomings and modified of specific grammatical properties of the L2 was a better
their claims, see e.g. Kay, Berlin and Merrifield, 1991; predictor of bilingual cognitive behaviour than length of
Kay and Berlin, 1997.) stay in the L2 country and age of L2 acquisition were.
The problem is compounded by the fact that the The way bilinguals categorised novel objects as units or
majority of the world’s population is bilingual anyway, substances depended strongly on their performance on an
a large proportion in English (Cook, 1999). “Pure” oral production task specifically assessing grammatical
monolingual populations are more likely to be found in number marking in the L2.
agricultural, usually uneducated communities, or tribes Boroditsky (2001) showed effects of age of L2
inhabiting remote parts of the world (see e.g. the studies acquisition on Chinese–English bilinguals’ perception of
by Rosch-Heider, 1972; Lucy, 1992; Davidoff et al., 1999; time. Younger learners were less likely to follow the L1
Gordon, 2004, and others). However, a serious caveat pattern than older learners (see however Chen, 2007,
in interpreting results from studies of this kind is that and January and Kako, 2007, for criticisms and counter-
the stimuli and tasks used may be unwittingly culturally evidence to Boroditsky’s, 2001, findings). A study of
biased, suited to the cultural background and cognitive grammatical gender and picture similarity judgments in a
demands of university-educated Westerners (Levinson, group of Spanish–German bilinguals revealed a strong
2001). relationship between similarity scores and language
On the other hand, bilingualism still tends to be proficiency, age of acquisition, and length of language
routinely overlooked in many studies in the field. Laws, use (Boroditksy et al., 2003).
Davies and Andrews (1995) found no differences in colour Pavlenko (2006) showed that bilinguals perceive
cognition between Russian and English speakers, despite themselves as different persons when using their different
the fact that Russian partitions the blue region of colour languages as a function of linguistic and cultural
space into two areas varying in lightness. More recently, differences between their languages, the language-
a study by Winawer, Witthoft, Frank, Wu, Wade and learning context, the degree of language emotionality, and
Boroditsky (2007) did find significant differences between the level of language proficiency. Hoffman et al. (1986)
Russian- and English-speaking populations on essentially found that the way Chinese–English bilinguals judged
the same task. In both studies, the Russian participants character traits in people depended on the language they
were bilingual in English, with varying degrees of L2 were addressed in. Addressed in Chinese, the bilinguals’
proficiency, and in many cases living and studying in the judgements were in line with Chinese stereotypes;
L2-speaking country. Yet no methodological measures addressed in English, the bilinguals’ judgements were
were taken to control for the possibility that their very similar to those of English monolinguals and
bilingualism may have influenced the results. An earlier conformed to English stereotypes. Studies by Malt and
study (Andrews, 1994) did show L2 influence on L1 Sloman (2003) and Ameel et al. (2005) showed that
in colour categorisation by Russian–English bilinguals. bilinguals have an integrated naming pattern for artifacts
Therefore, variables such as L2 proficiency, acculturation, like bottles and dishes, influenced by both languages, but
and age of L2 acquisition, to name a few, may be crucial in did not find effects on non-linguistic categorisation. This
determining the way bilinguals operate in cognitive tasks suggests that the way lexical or grammatical categories
that aim to investigate the degree of linguistic influence influence different cognitive or perceptual domains may
on cognition. vary (Roberson et al., 2005).
Indeed recent studies have demonstrated the effects
of these variables on the way bilinguals behave in cross-
1.3 Aims of the current study
cultural cognitive tasks. Cook et al. (2006) showed effects
of acculturation on the way Japanese–English bilinguals The current study extends the issue of linguistic relativity
extended the novel name for a target object or substance to and bilingualism to the domain of colour cognition.
a shape or material alternate. The behaviour of bilinguals Specifically, the aim is to examine whether cognitive
living in the L2-speaking country for more than three years representation of colour will be affected by knowledge
had shifted towards that of monolingual speakers of their of two languages that differ in their coding of the colour
L2, and was significantly different from that of bilinguals space. Earlier work had investigated “semantic shifts” in
who had lived in the L2 country for shorter periods bilinguals, showing L2 influence on use and prototypes
of time. Athanasopoulos (2006, 2007) showed robust of L1 colour terms. For example, Ervin (1961) showed
effects of L2 proficiency on the way Japanese–English that Navajo–English bilinguals used their L1 colour
bilinguals perceived the similarity between countable terms differently from Navajo monolinguals. Caskey-
objects and non-countable substances. Bilingual speakers Sirmons and Hickerson (1977) showed that individuals
shifted with L2 proficiency towards the cognitive pattern from five different language backgrounds who were all
of monolingual speakers of their L2. More recently, bilingual in English had shifted their prototypes (foci)
86 Panos Athanasopoulos

for a range of L1 colour terms towards the English as the best examples (foci) of the two categories. The
prototypes. This led the researchers to claim that “the aim is to calculate precise category boundaries of ble and
worldview of bilinguals, whatever their first language, ghalazio which will be used for the subsequent cognitive
comes to resemble, to some degree, that of monolingual experiment, to compare bilinguals’ naming behaviour to
speakers of their second” (Caskey-Sirmons and Hickerson that of English monolinguals reported in previous studies,
1977, p. 365). A study by Zollinger (1988) showed that and to investigate whether bilinguals exhibit shift of
Japanese children living in Germany used certain L1 category foci on the colour space, as was previously shown
colour terms significantly less frequently than Japanese by Caskey-Sirmons and Hickerson (1977). To this end,
children living in Japan. Jameson and Alvarado (2003) bilinguals will be separated into two groups based on
found that bilingual speakers whose L1 (Vietnamese) L2 proficiency, an Intermediate and an Advanced group,
has fewer colour terms than their L2 (English) tended since it is extremely difficult to find individuals who are
to modify their colour naming behaviour according to the completely monolingual in Greek and who match the
distinctions made in their L2, suggesting that bilinguals bilinguals in socio-educational background at the same
use whichever system is maximally informative. time.1
Although these studies did not investigate non- Experiment 2 will then examine cognitive represen-
linguistic cognitive behaviour, their results suggest that tation of the ble/ghalazio distinction in Greek–English
bilinguals may differ from monolinguals in the way they bilinguals who have not taken part in Experiment 1,
use their colour terms and in the way they represent them by asking them to make perceptual similarity ratings
on colour space. However, no study has systematically of pairs of stimuli that fall within the ble or ghalazio
investigated the consequences of bilingualism on boundary, and pairs of stimuli that cross-cut the boundary.
perception and cognitive representation of colour Bilinguals’ ratings will be compared to those of English
categories. The current study opens up that investigation, monolinguals, and will then be correlated with their level
in order to address a crucial question in the field of L2 of L2 proficiency as measured by a general English
acquisition and bilingualism: does learning a second set of proficiency test, their length of stay in the L2 speaking
categorical divisions lead to cognitive restructuring in the country, the age at which they started to learn English,
mind of the bilingual person and, if so, to what extent, and and the extent to which they use English in their
why (Cook, 1997, 2002, 2003; Green, 1998; Pavlenko, everyday lives. In addition, correlations will be made
1999, 2005a)? The pursuit of this question in the domain with semantic memory for specific colour terms in
of colour is particularly important due to the centrality English and Greek, given that recent evidence suggests
of that domain in the language and thought debate. If a strong relationship between specific L2 proficiency and
categorical perception of colour is intrinsically derived cognitive behaviour (Athanasopoulos and Kasai, 2008).
from language, then the colour domain is an ideal tool to These measures are necessary since it is nearly impossible
test the question of how contrasting categorical aspects to obtain comparable monolingual Greek behaviour,
are represented in the bilingual mind. as mentioned earlier. Furthermore, such measures will
The bilinguals under investigation will be native determine the degree to which bilingualism affects
speakers of Greek who have English as an L2. Greek cognitive representation of colour, by identifying the
divides the blue region of colour space into two distinct potential impact of a range of variables, which previous
regions, a darker shade called ble, and a lighter shade studies have shown may influence bilingual cognition.
called ghalazio (Androulaki, Pestaña, Lillo and Davies,
2001). A similar distinction between two monolexemic
2. Experiment 1: Foci and areas of Greek blues
terms that divide the blue region of colour space has also
been documented in a range of other languages such as
2.1 Method
Japanese (Uchikawa and Boynton, 1987), Turkish (Özgen
and Davies, 1998), and Russian (Davies and Corbett, 2.1.1 Participants
1997). If cognitive representation of colour is affected Participants were 20 native speakers of Greek who were
by the L2, then the blue vs. light blue distinction will separated into two groups based on their proficiency in
become less salient in the cognition of Greek–English English. The group with advanced English proficiency
bilinguals as a result of using a second language that consisted of 10 adult individuals who were tested in
does not mark the contrast between these categories. This the UK and selected on the basis of them attending
would suggest that learning new conceptual dimensions a postgraduate course (MA or PhD) in Linguistics
leads to some (potentially quantifiable) extent of cognitive or English Language at a UK university. This would
reorganisation.
The current study investigates these issues in two 1 This is apparently not the case for native speakers of English, the
experiments. Experiment 1 will obtain naming patterns majority of whom tend to be monolingual even when they have
for chips that lie in the blue area of colour space, as well attained high levels of education.
 Bilingual blues 87

Table 1. Summary of advanced and intermediate level 10PB, and in lightness (Munsell Value) from 2 to 9. These
bilinguals’ details in Experiment 1. were also fully saturated. They were 10 × 20 mm glossy
finish chips individually mounted on 40 mm square pieces
Advanced Intermediate of white card.
N 10 10
Age range 21–31 19–26
2.1.3 Procedure
Each participant was seated at a table next to a window and
Sex 7F/3M 8F/2M
was shown each colour chip one at a time, in random order,
Mean Nation score 85/90 (80–90) 64/90 (60–69)
and asked to indicate its name in their native language.
(range) Each response was recorded by the administrator of the
Mean L2 acquisition 8 years old (5–13) 9 years old (5–13) experiment and gathered for analysis. After completion
age (range) of the naming task, participants were shown the full array
Mean length of stay 30 months (9–48) n/a of Munsell chips. They were asked to indicate which chip
in the L2 country is the best example of ble and which chip is the best
(range) example of ghalazio (counter-balanced within each group,
such that half of the participants were asked to identify
the ble focus first, and half were asked to identify the
ghalazio focus first). Participants received instructions in
ensure that they had high proficiency in English. Their Greek by a Greek native speaker. Each testing session
proficiency was measured by the Nation vocabulary test lasted for about 40 minutes. The order of tasks was
(Nation, 1990), which measures vocabulary at five levels, as follows: first participants conducted the naming and
ranging from the 2,000-word level up to the 10,000-word focus placement tasks (20 minutes), then they completed
level. Their mean score was 85/90 (SD = 3, range 80–90), a personal information questionnaire (5 minutes), and then
i.e. around the 10,000-word level. Their age-range was the Nation test was administered (15 minutes).
21–31 years, 7 female, 3 male. The mean age at which
they started to learn English was 8 years old (SD = 3),
ranging from 5 to 13 years old. Their mean length of stay 2.2 Results
in the UK was 30 months (SD = 14), ranging from 9 to Figure 1 represents the area of Munsell colour space that
48 months. The group with intermediate English was used for naming. It shows the modal naming data
proficiency were tested in Greece and consisted of as well as best exemplar placement for ble and ghalazio
individuals who were attending non-English related by advanced- and intermediate-level bilinguals, alongside
university courses in Greece and had not previously visited naming data and focus placement for blue in English
the UK or any other English-speaking country for more native speakers reported in the studies by Roberson
than one month. Their mean score in the Nation test was et al. (2000, 2005). A chip was completely shaded ble
64 (SD = 4, range = 60–69), i.e. around the 6,000-word or ghalazio on the basis of at least 80% of participants
level. Their age-range was 19–26 years old, 8 female 2 naming it so. Otherwise, boundaries drawn through an
male. Their mean age of L2 acquisition was 9 years old individual chip represent the proportion of participants
(SD = 2, range 5–13). All of the participants reported that named the particular chip ble or ghalazio. Within-
that they had normal colour vision. Table 1 provides a group naming agreement for intermediate bilinguals was
summary of the participants’ details. 77% for ble and 67% for ghalazio, while for advanced
bilinguals naming agreement was 82% for ble and 65%
2.1.2 Materials for ghalazio. The chips with lowest agreement were the
The stimuli used for elicitation of category foci were chips located between Hue levels 5B–5PB at lightness
identical to the ones used by Heider and Olivier (1972) Value 5. These were named ble by 57% of advanced
and Roberson et al. (2000, 2005). These comprised of 160 bilinguals (the rest naming them ghalazio), compared
fully saturated Munsell colour chips varying in hue and to 33% of intermediate bilinguals, who in turn tended
lightness. They were mounted on a sheet of stiff white to use ghalazio to name them. The difference between
cardboard and were arranged to represent hue levels 5 the two groups was specifically located at the 5B/5 chip,
and 10 of ten equally spaced steps around the Munsell which was ambiguous for advanced-level bilinguals (see
circle (Munsell dimension Hue R, YR, Y, YG, G, BG, the top part of Figure 1), but not for intermediate-level
B, PB, P, RP) each at eight lightness levels (Munsell bilinguals, who consistently named it ghalazio (see the
dimension Value 9/, 8/, 7/, 6/, 5/, 4/, 3/, 2/). Since the middle part of Figure 1). Thus, the naming data show that
focus is only on the blue region of colour space, stimuli both groups treat the chips located at 10B/5 and 5PB/5
used for name elicitation were a subset of the full array as ambiguous with regards to their status as members
described above, ranging in Munsell Hue from 5BG to of either category, while 5B/5 is ambiguous only for
88 Panos Athanasopoulos

the best example of ble one step away from the blue focus,
both in lightness and hue. On the other hand, the majority
of advanced bilinguals tend to shift the ble focus towards
blue. This semantic shift toward the L2 is in line with
Caskey-Sirmons and Hickerson’s (1977) results. The case
of ghalazio is interesting. Advanced bilinguals exhibit
focus shift for this category as well, but in the opposite
direction of the L2 category. Thus, the distance in colour
space between ble and ghalazio foci is similar for both
bilingual groups, with advanced bilinguals shifting their
foci symmetrically. This suggests that bilingual semantic
shift of colour foci need not always occur in the direction
of the L2. In this particular case, advanced bilinguals
shift their prototypes for ble towards blue, but they also
maintain the perceptual distance between ble and ghalazio
foci by shifting the latter category’s prototypes away from
blue. At the same time, there are minimal differences
between the groups in where they place the boundary
between the two categories in the naming task. Does
this mean that despite bilinguals shifting their category
foci they still perceive the distinction between ble and
ghalazio in the same way? Previous research shows
that participants rate two colours to be more similar if
they fall within a category boundary than if they cut
across the boundary. (Roberson et al., 2000, 2005). In
the case of Greek speakers, Experiment 2 will attempt
to investigate cognitive sensitivity to the ble/ghalazio
distinction by asking participants to rate the perceptual
similarity between pairs of stimuli that are within the ble
or ghalazio boundary, and pairs of stimuli that cross-cut
the boundary. The aim is to identify potential changes
in bilinguals’ perception of stimuli corresponding to ble
and ghalazio by correlating cognitive performance with
semantic memory for specific colour terms, general L2
proficiency, length of stay in the L2-speaking country, age
of acquisition, and degree of L2 use. These correlations
will help to elucidate the precise impact of each of these
variables on bilingual cognition.

Figure 1. Distribution of advanced- and intermediate-level 3. Experiment 2: Similarity judgements of Greek

bilingual naming and choices of best exemplar for ble and blues
ghalazio, and English naming distribution for blue adapted
from Roberson et al. (2000, 2005). The three-dimensional 3.1 Method
Munsell system is shown as a two-dimensional projection
3.1.1 Participants
of hue (horizontal axis) against lightness (vertical axis).
Numbers represent the number of individuals choosing an Participants were 30 adult native speakers of Greek who
exemplar as best example of the category. Dots represent had not taken part in the previous experiment (20 female,
the English best example of blue reported in Roberson et al. 10 male, age-range 19–32 years old). They were all
(2000, 2005). bilingual in English and were studying at a UK university.
Their proficiency in English was measured by the Nation
advanced-level bilinguals. Overall, these data allow us vocabulary test (Nation, 1990). Their mean score was
to place the category boundary between ble and ghalazio 75/90 (SD = 9), i.e. around the 8,000-word level. Their
at those particular chips. scores ranged from 61 (around the 6,000-word level) to
Turning now to category foci placement, it is clear from 90 (10,000-word level). Their mean length of stay in the
Figure 1 that the majority of intermediate bilinguals place UK was 33 months (SD = 27 months), ranging from 2 to
 Bilingual blues 89

Table 2. Summary of Greek–English bilinguals’ and Table 3. Mean similarity judgements (and standard
English monolinguals’ details in Experiment 2. deviations) of ble and ghalazio within and cross-
category pairs. Figures are rounded to the nearest two
Greek–English English decimal places.
bilinguals monolinguals
Groups Within-category pairs Cross-category pairs
N 30 22
Age range 19–32 19–28 Greek (n = 30) 4.08 (0.99) 4.34 (1.07)
Sex 20F/10M 18F/4M English (n = 22) 4.08 (0.86) 3.96 (1.22)
Mean Nation score (range) 75/90 (61–90) n/a
Mean L2 acquisition age 7 years old (1–13) n/a
(range) Five pairs of within-category stimuli and three pairs of
Mean length of stay in the 33 months (2–96) n/a cross-category stimuli were created from the ble and
L2 country (range) ghalazio areas obtained in Experiment 1. Pairs were
Mean daily L2 use (range) 9 hours (3–17) n/a constructed so that either both members were within the
same category (e.g. Hue level 10B/Lightness level 4 –
Hue level 10B/Lightness level 2), or the pair cut across
the ble/ghalazio boundary (e.g. Hue level 10B/Lightness
96 months. Their mean age of L2 acquisition was 7 years level 4 – Hue level 10B/Lightness level 6). A full list of
old (SD = 3 years), ranging from 1 to 13 years old. They pairs can be found in the Appendix.
reported that they used English for 9 hours per day on
average (SD = 4), ranging from 3 to 17 hours. 3.1.3 Procedure
In addition to all the measures above, elicited by means Each participant was seated at a table next to a
of a questionnaire, semantic saliency of ble, ghalazio and window and pairs were presented, one at a time, in
blue was measured by asking participants to write down random order. Participants were asked to judge “how
all the colour names they could think, first in one language different or similar these two colours are” using a 10-
and then in the other. To minimise the possibility of point scale where 10 represents maximum dissimilarity
back-translation, participants were given the Nation test to and 1 represents maximum similarity. Each pair was
complete in-between. The order of languages was counter- shown twice, counterbalancing the position of each
balanced across the whole sample. Elicitation of colour individual chip in the pair. All participants were tested
lists is a standard task used to establish the basicness of in the UK and each testing session lasted for about
colour terms in a language (Corbett and Davies, 1995). 40 minutes for bilinguals, and about 15 minutes for
Basic colour terms should be more available in semantic English monolinguals. The order of tasks for bilinguals
memory than non-basic terms, and thus appear higher in was as follows: first participants conducted the similarity
the list. Here, it is used to measure the saliency of blue, ble judgment task (10 minutes), then they completed the
and ghalazio. The aim is to correlate the semantic saliency personal information questionnaire (5 minutes), then the
of each term (i.e. how high it appears in the list) with colour list for one language was elicited (5 minutes),
participants’ similarity judgments. On average, ble was then the Nation test was administered (15 minutes),
placed 6th on the list (SD = 4, range 1–17), ghalazio was and finally the colour list for the other language was
placed 12th (SD = 7, range 2–28), and blue was placed elicited (5 minutes). The English monolinguals were given
5th (SD = 3, range 1–13). the similarity judgement task, followed by the personal
The similarity judgement task was also given to 22 information questionnaire. Instructions were given to all
adult English monolingual participants who were all participants in their native language by a Greek–English
university students. Their mean age was 21 years, age- bilingual speaker.
range 19–28, 18 female, 4 male. Since English generally
uses a single term to refer to the blue area of colour space,
English speakers are expected to treat pairs of stimuli 3.2 Results
that cut across the ble/ghalazio boundary no differently Table 3 shows the mean similarity judgments of cross-
from pairs that fall within each category boundary. All category and within-category pairs of stimuli for the
participants reported that they had normal colour vision. two groups. A 2 (Group: English vs. Greek) × 2 (Pair
Table 2 provides a summary of the participants’ details. type: Within vs. Cross) mixed ANOVA showed that
the main effects of Group and Pair type were not
3.1.2 Materials significant, F(1,50) = 0.513, p > .05, and F(1,50) = 0.448,
The stimuli were individual 10 × 20 mm glossy Munsell p > .05 respectively. The Group × Pair type interaction
chips, mounted on 40 mm square pieces of white card. approached significance, F(1,50) = 3.247, p = .08. This
90 Panos Athanasopoulos

Table 4. Correlation matrix showing Pearson’s r for the eight variables.

L2 Length Age of L2 Amount of Saliency Saliency of Saliency

proficiency of stay acquisition L2 use of ble ghalazio of blue

CPI −.358∗ −.399∗ .005 .037 −.421∗∗ .149 .401∗

L2 proficiency .604∗∗ .174 .259 .014 .367∗ −.426∗∗
Length of stay −.345∗ .181 −.079 .153 −.354∗
Age of L2 acquisition .026 .082 .377∗ −.259
Amount of L2 use −.026 .177 −.004
Saliency of ble .042 −.228
Saliency of ghalazio .029

∗
p < .05, ∗∗ p < .01

Table 5. Pearson’s r for partial correlations between CPI and each independent variable, controlling for all the other
variables.

L2 Length Age of L2 Amount of Saliency Saliency of Saliency

proficiency of stay acquisition L2 use of ble ghalazio of blue

CPI −.178 −.389∗ −.241 .138 −.528∗∗ −.046 .410∗

∗
p < .05, ∗∗ p < .01

nearly significant interaction does not readily reveal (Howell, 2002), the multiple regression analysis was
whether Greek speakers differ from English speakers not performed.2 Instead, CPI was correlated with
in their cognitive representation of the blue area of each independent variable, partialling out all the other
colour space. However, we now turn to examine whether variables. Table 5 gives a summary of these correlations.
the cognitive behaviour of Greek speakers is modulated As Table 5 shows, the strongest significant correlation
by any of the variables that were measured and was obtained for semantic saliency of ble. This means
which previous research has shown may affect the way that the further down ble appears on the colour list
bilinguals perform in non-linguistic similarity judgment (and thus the greater its number on the list), the less
tasks. bilinguals distinguish between within- and cross-category
pairs. There was also a moderate correlation with semantic
saliency of blue, such that the higher blue appears in each
3.2.1 Identifying the variables that may predict
participant’s colour list (and thus the smaller its number
bilingual behaviour
on the list), the less distinction is made between within-
In order to identify the variables that may affect bilingual
and cross-category pairs. Finally, the weakest significant
cognition, the data were initially going to be analysed by
correlation was obtained for length of stay in the UK,
a multiple regression analysis. The independent variables
such that the more bilinguals have stayed in the L2
were general L2 proficiency as measured by the Nation
country, the less they distinguish between within- and
test, length of stay in the L2 country, age of L2 acquisition,
cross-category pairs. None of the rest of the variables
amount of L2 use, and semantic saliency of ble, blue and
correlated significantly with CPI. To sum up, these results
ghalazio, as measured by the position of each term on each
show that when controlling for a range of variables that
participant’s list of colour terms. The dependent variable,
may influence bilingual cognition, semantic saliency of
hereafter called the Categorical Perception Index (CPI),
specific colour terms is most strongly correlated with
was calculated by subtracting each participant’s mean
bilingual cognition, with length of stay having a weaker
similarity judgment score for within-category pairs from
but significant effect.
their mean similarity judgement score for cross-category
pairs. Table 4 shows the correlation matrix between all
variables. 2 Indeed multicollinearity is a problem that comes as no surprise in a
As Table 4 shows, several independent variables study of bilingualism. It is very difficult to dissociate variables such as
significantly correlate with each other, as well as age of L2 acquisition, L2 proficiency, degree of L2 use, acculturation,
with CPI. Because of this multi-collinearity problem and so on (see e.g. Athanasopoulos, 2007).
 Bilingual blues 91

3.2.2 Dissecting memory and length of stay stay group (N = 14, mean CPI = .59), with length of stay
Since memory for specific colour terms and length of stay ranging between 2 and 22 months (again the sizes of the
in the L2 country were statistically significant predictors groups could not be exactly the same for similar reasons
of CPI, further analyses were carried out in order to to those mentioned earlier). A one-way ANOVA showed
examine whether there is a discernible threshold in a significant main effect of group, F(2,49) = 4.269,
memory and/or length of stay at which participants shift p < .05. Post-hoc Bonferroni tests showed that the only
their representations, or whether these shift progressively. significant difference was between the Short stay group
To this end, bilinguals were split into two groups based and the English monolinguals (p < .05). This means that
on their memory for the relevant colour terms and their once individuals have spent 24 months in the L2-speaking
length of stay in the UK. Using CPI as the dependent country their colour cognition begins to shift towards the
variable, their behaviour was compared against each other L2 pattern.
and against that of the English monolingual group (mean
CPI = −.12, see Table 3) in separate one-way ANOVAs.
4. Discussion
Memory for ble and blue were examined first.
Bilinguals were evenly distributed into High and Low Empirical evidence is accruing to suggest that bilingual
groups, based on the position of each term in each speakers with languages that differ in their lexical
participant’s list. For ble, the split yielded a High or grammatical concepts and categories differ from
group (N = 15, mean CPI = .57), who placed the term monolingual speakers of their L1, and shift towards
between first and fourth place, and a Low group (N = 15, monolingual speakers of their L2, in their cognitive
mean CPI = −.03), who placed the term between fifth representation of those categories. The present study
and seventeenth place. A one-way ANOVA showed a addressed the issue by asking whether, and to what degree,
significant main effect of Group, F(2,49) = 4.212, p < .05. L2 colour categories influence the representation of L1
Post-hoc Bonferroni tests showed that the only significant colour categories on the colour space, and the cognitive
difference was between the High group and the English discrimination of L1 colour categories in a similarity
monolinguals (p < .05). This means that those bilinguals judgment task.
that tended to place ble in the first four places on their The data from naming and prototype placement
list distinguished significantly more between within- and (Experiment 1) showed that the majority of advanced-
cross-category pairs than the English monolinguals did. level bilinguals shift the focus of one of their L1
Thus, cognitive behaviour begins to shift towards the L2 categories towards the L2 category focus, as did the
pattern once ble is placed consistently below fourth place earlier study by Caskey-Sirmons and Hickerson (1977).
on the list. However, they also shift the focus of the other L1 category
The groups that were formed on the basis of memory away from the L2 focus, thus maintaining the perceptual
for blue were a High group (N = 17, mean CPI = −.07), distance between their native category foci. Furthermore,
who placed the term between first and fourth place there were minimal differences between advanced and
(because many participants placed the term fourth it was intermediate bilinguals in naming stimuli ble or ghalazio.
impossible for the groups in this case to have exactly The data from the similarity judgment task (Experiment 2)
the same number of participants; this division is as showed that straightforward comparisons do not provide
close to even as possible), and a Low group (N = 13, a conclusive answer as to whether English and Greek
mean CPI = .71), who placed the term between fifth and speakers differ in their cognitive representation of the blue
thirteenth place. A one-way ANOVA showed a significant area of colour space. However, a rigorous examination
main effect of Group, F(2,49) = 6.221, p < .01. Post- of a range of variables that previous studies have shown
hoc Bonferroni tests showed that the Low group differed to influence bilingual cognition revealed that semantic
significantly both from the High group (p < .05) and the memory for specific colour terms strongly correlated
English monolingual group (p < .01), while there was with the way Greek–English bilinguals perceived the
no significant difference between the High group and the ble/ghalazio distinction, while there was a weaker but
English group (p > .05). This means that those bilinguals significant effect of length of stay in the L2 country.
who tended to place blue between fifth and thirteenth Thus, although bilinguals are able to observe the
place on the list distinguished significantly more between perceptual distance between their native category foci on
within- and cross-category pairs than the High group and the colour space, and “know” which stimuli correspond
the English monolinguals did. Thus, cognitive behaviour to their native colour terms, their underlying perception
shifts completely towards the L2 pattern once blue is of the difference between these categories appears to have
placed consistently above fifth place on the list. been distorted as a function of the saliency of specific
For length of stay in the L2 country, the split yielded colour terms in semantic memory, and the amount of
a Long stay group (N = 16, mean CPI = −.02), who had time spent in the L2 country. Further analyses revealed
stayed in the UK between 24 and 96 months, and a Short specific thresholds in semantic memory and length of stay
92 Panos Athanasopoulos

where bilingual cognition begins to shift toward that of concepts are not represented in the form of clearly defined
monolingual speakers of the L2. While it is not claimed separate entities, characterised by a set of stable features.
that these thresholds may apply generally in all cases, they Rather, the evidence suggests that they are dynamic and
may be useful in predicting behaviour in future similar flexible, susceptible to influence by a range of variables
studies. In any case, these findings add an important new that characterise the bilingual person, and tightly linked
dimension to investigations of bilingual cognition, which to the semantic saliency of specific L1 and L2 linguistic
thus far have rarely looked at the link between specific categories.
L2 knowledge and cognition (but see Athanasopoulos This tight link between specific linguistic knowledge
and Kasai, 2008), and certainly have not looked at the and cognitive behaviour demonstrates that bilingualism
link between specific L1 knowledge and cognition. The offers a unique insight into the study of the relationship
current findings show that it is not just acquisition of novel between language and thought. Not only can we
linguistic categories that may influence the way bilinguals investigate cognitive differences and similarities between
behave in cognitive tasks, but also attrition of the original groups based on their language background, but we
L1 term in memory. can also explore the precise linguistic elements, as
The current findings also show a weak but significant well as a range of other variables, that may drive
effect of length of stay in the L2 country. The basis of these cognitive differences between speakers of different
that effect is likely to be fundamentally a matter of use languages. This methodological advantage is particularly
of linguistic categories. The more bilinguals stay in the important in the context of cross-cultural research, given
L2 country, the less opportunity they have to use their L1 that comparisons between seemingly “monolingual”
colour terms to refer to their visual environment, either speakers of different languages may mask important
because their L1 colour categories are not visually as cross-linguistic differences, or lead to conflicting results
salient as in their native country, and/or because they are between studies that do not take into account their
not systematically attended to by native speakers of their participants’ bilingualism.
L2. This would neatly explain both the strong effect of Since the current study is the first to directly address
memory for specific colour terms, and the weaker effect of cognitive representation of colour in bilinguals, there
length of stay, suggesting an indirect relationship between are several issues that warrant further examination. A
the two. question arises as to why semantic saliency of ghalazio did
Whatever the case may be, the current study sheds light not correlate significantly with the categorical perception
on an important question in the field, namely whether behaviour of bilinguals. One explanation for this finding
it is language per se or non-linguistic cultural aspects may be offered by Berlin and Kay’s (1969) theory of
that drive the cognitive effect. The current finding of basic colour terms. According to the theory, basic colour
both influence of stay in the L2 country AND semantic terms should be monolexemic, not subsumed within the
saliency of specific colour terms shows that focusing the meaning of any other colour term, and used frequently
debate on linguistic OR cultural influence may obscure and consistently. Ghalazio satisfies the first criterion, but
an interestingly complex situation. It appears that BOTH fails to fulfil the other two criteria. When asked at the
culture and language may influence the way bilinguals end of the experiment whether they considered ble and
perceive categorical divisions, but to different degrees and ghalazio to be different colours, most participants pointed
in different ways. out that they considered ghalazio to be a kind of ble.
These findings, in conjunction with previous recent Furthermore, although ghalazio was used with reasonably
findings, have important implications for conceptual good agreement amongst speakers in Experiment 1, the
representation in bilinguals (Francis, 1999; Pavlenko, colour lists obtained for Experiment 2 showed that it
1999; De Groot, 2000; Cook, 2002) as well as was placed, on average, outside of the top eleven colour
broader implications for the conceptualisation component terms (recall that eleven was the maximal number of basic
of speech production models (Levelt, 1989; Levelt, colour terms that Berlin and Kay, 1969, reported), and its
Roelofs and Meyer, 1999). Recent developmental findings position on the list varied widely amongst participants. All
(Roberson, Davidoff, Davies and Shapiro, 2004) show that this evidence converges to suggest that only semantically
both Himba and English children initially make similarity salient linguistic properties may influence cognition, and
judgements on perceptual grounds only. However, that secondary and less salient terms may not.
after acquisition of their respective colour vocabulary, An additional explanation may be that ble, but not
similarity judgements shift towards language-specific ghalazio, is the translation equivalent of blue, sharing
patterns. The current study suggests that conceptual not only semantic elements but also morpho-phonological
representation is not permanently fixed by the L1, but properties. Certainly, research using the semantic priming
changes with the acquisition of specific L2 properties and paradigm has shown that a target word is processed faster
attrition of pre-existing L1 properties. This provides direct when it is preceded by a semantically related prime than an
empirical support to Pavlenko’s (1999) argument that unrelated word, not only when the target and prime come
 Bilingual blues 93

from the same language, but also when the prime comes the behaviour of native speakers of different languages
from one language and the target from the other language whilst ignoring the fact that they are bilingual.
of bilingual participants (Altarriba, 1992; De Groot,
2002). This finding shows that semantic representations Appendix A: Stimulus pairs used in Experiment 2
are shared between the bilingual’s languages. In this Munsell designations are expressed in the form of
particular case, the similarities between ble and blue not Hue/Lightness. The stimuli were fully saturated.
only in meaning but also in form could be the reason why Within-category:
bilingual cognition is influenced by these specific terms 10B/2–10B/4, 5PB/2–5PB/4, 5B/6–5B/8,
and not by ghalazio. Further research could investigate 10B/6–10B/8, 5PB/6–5PB/8
this matter more closely by comparing cognitive Cross-category:
performance for basic colour terms that are similar in 5B/4–5B/6, 10B/4–10B/6, 5PB/4–5PB/6
form with cognitive performance for basic colour terms
that do not share any morpho-phonological properties.
Future research may also address the issue of References
typological distance between the languages of bilinguals.
Altarriba, J. (1992). The representation of translation equivalents
The differences between English and Greek in how
in bilingual memory. In R. Harris (ed.), Cognitive
they partition the colour space into categories are
processing in bilinguals, pp. 157–174. Amsterdam:
subtler than differences between English and languages Elsevier.
with considerably less basic colour terms, say Himba Ameel, E., Storms, G., Malt, B. & Sloman, S. (2005). How
or Berinmo. A cross-cultural investigation of native bilinguals solve the naming problem. Journal of Memory
speakers of these languages who are bilingual in English and Language, 53, 60–80.
would determine whether typological distance between Andrews, D. (1994). The Russian color categories sinij and
languages might also play a role in the degree to which goluboj: An experimental analysis of their interpretation
bilinguals shift their cognition towards the L2. in the standard and émigré languages. Journal of Slavic
Linguistics, 2, 9–28.
Androulaki, A., Pestaña, N. G., Lillo, J. & Davies, I. R. [L.]
(2001). Greek colour terms: Evidence for a natural category
5. Conclusion boundary within blue. Perception #30, ECVP Abstract
Supplement.
The current study examined cognitive representation of Athanasopoulos, P. (2006). Effects of the grammatical
colour categories in bilinguals with languages that differ in representation of number on cognition in bilinguals.
the way they code the colour space. Results showed a shift Bilingualism: Language and Cognition, 9, 89–96.
in focus placement with level of bilingualism, but minimal Athanasopoulos, P. (2007). Interaction between grammatical
differences in the extension of colour terms on colour categories and cognition in bilinguals: The role
space. On the other hand, the way bilinguals perceived of proficiency, cultural immersion, and language of
the distinction between their native colour categories instruction. Language and Cognitive Processes, 22, 689–
depended on the availability of specific L1 and L2 colour 699.
terms in semantic memory, suggesting that semantic Athanasopoulos, P. & Kasai, C. (2008). Language and thought in
bilinguals: The case of grammatical number and nonverbal
saliency as well as similarities in surface linguistic form
classification preferences. Applied Psycholinguistics, 29,
may account for bilingual cognitive behaviour in this
105–123.
case. The implications of these findings for conceptual Berlin, B. & Kay, P. (1969). Basic colour terms: Their
representation were discussed, suggesting that concepts universality and evolution Berkeley, CA: University of
in the human mind are not stable and fixed but flexible California Press.
and changing, susceptible to both linguistic and cultural Boroditsky, L. (2001). Does language shape thought? English
influence. and Mandarin speakers’ conceptions of time. Cognitive
Finally, the current study also demonstrated that Psychology, 43, 1–22.
bilingualism is an ideal testing ground for the linguistic Boroditsky, L., Schmidt, L. & Phillips, W. (2003). Sex, syntax,
relativity hypothesis (as suggested by Hunt and Agnoli, and semantics. In Gentner & Goldin-Meadow (eds.),
1991, but never empirically realised until recently). In pp. 61–79.
Brown, R. & Lenneberg, E. (1954). A study in language and
the increasing absence of pure monolingual populations,
cognition. Journal of Abnormal and Social Psychology,
bilingualism may not simply be the “next best thing”,
49, 454–462.
but may offer a genuine methodological advantage in that Carroll, D. W. (2008). Psychology of language (5th edn.).
we are able to correlate a range of linguistic and non- Belmont, CA: ThomsonWadsworth.
linguistic variables with cognitive behaviour. This type Caskey-Sirmons, L. A. & Hickerson, N. P. (1977). Semantic
of analysis can reveal important interactions and patterns shift and bilingualism: Variation in the colour terms of five
that would not be readily revealed by directly comparing languages. Anthropological Linguistics, 19, 358–367.
94 Panos Athanasopoulos

Chen, J-Y. (2007). Do Chinese and English speakers think about Howell, D. C. (2002). Statistical methods for psychology (5th
time differently? Failure of replicating Boroditsky (2001). edn.). Pacific Grove, CA: Duxbury Press.
Cognition, 104, 427–436. Hunt, E. & Agnoli, F. (1991). The Whorfian hypothesis: A
Cook, V. J. (1997). The consequences of bilingualism for cognitive psychology perspective. Psychological Review,
cognitive processing. In A. M. B. De Groot & J. F. 98, 377–389.
Kroll (eds.), Tutorials in bilingualism: Psycholinguistic Imai, M. & Gentner, D. (1997). A crosslinguistic study of early
perspectives, pp. 279–300. Hillsdale, NJ: Lawrence word meaning: Universal ontology and linguistic influence.
Erlbaum. Cognition, 62, 169–200.
Cook, V. J. (1999). Going beyond the native speaker in language Imai, M. & Mazuka, R. (2003). Re-evaluating linguistic
teaching. TESOL Quarterly, 33, 185–209. relativity: Language-specific categories and the role of
Cook, V. J. (2002). Background to the L2 user. In Cook (ed.) universal ontological knowledge in the construal of
(2002a), pp. 1–31. individuation. In Gentner & Goldin-Meadow (eds.),
Cook V. J. (ed.) (2002a). Portraits of the L2 user, pp. 1–31. pp. 429–464.
Clevedon: Multilingual Matters. Jameson, K. A. & Alvarado, N. (2003). Differences in color
Cook, V. J. (2003). The changing L1 in the L2 user’s mind. In naming and color salience in Vietnamese and English.
V. J. Cook. (ed.), Effects of the second language on the first, Color Research & Application, 28, 113–138.
pp. 1–18. Clevedon: Multilingual Matters. Jameson, K. A. & D’Andrade, R. (1997). It’s not really red,
Cook, V. J., Bassetti, B., Kasai, C., Sasaki, M. & Takahashi, green, yellow, blue: An inquiry into color space. In Hardin
J.A. (2006). Do bilinguals have different concepts? The & Maffi (eds.), pp. 295–319.
case of shape and material in Japanese L2 users of January, D. & Kako, E. (2007). Re-evaluating evidence for
English. International Journal of Bilingualism, 10, 137– linguistic relativity: Reply to Boroditsky (2001). Cognition,
152. 104, 417–426.
Corbett, G. G. & Davies, I. R. L. (1995). Linguistic and Kay, P. & Berlin, B. (1997). Science = Imperialism: Response to
behavioral measures for ranking basic colour terms. Studies Saunders and van Brackel. Behavioral and Brain Sciences,
in Language, 19, 301–357. 20, 196–201.
Davidoff, J., Davies, I. [R. L.] & Roberson, D. (1999). Colour Kay, P., Berlin, B. & Merrifield, W. (1991). Biocultural
categories in a stone-age tribe. Nature, 398, 203–204. implications of systems of color naming. Journal of
Davies, I. R. L. & Corbett, G. G. (1997). A cross-cultural study Linguistic Anthropology, 1, 12–25.
of colour-grouping: Evidence for weak linguistic relativity. Kay, P. & Regier, T. (2006). Language, thought, and color:
British Journal of Psychology, 88, 493–517. Recent developments. Trends in Cognitive Sciences, 10,
De Groot, A. M. B. (2000). On the source and nature of semantic 51–54.
and conceptual knowledge. Bilingualism: Language and Kay, P. & Regier, T. (2007). Color naming universals: The case
Cognition, 3, 7–9. of Berinmo. Cognition, 102, 289–298.
De Groot, A. M. B. (2002). Lexical representation and lexical Laws, G., Davies, I. [R. L.] & Andrews, C. (1995). Linguistic
processing in the L2 user. In Cook (ed.) (2002a), pp. 32–63. structure and non-linguistic cognition: English and Russian
Ervin, S. (1961). Semantic shift in bilingualism. American blues compared. Language and Cognitive Processes, 10,
Journal of Psychology, 74, 233–241. 59–94.
Francis, W. S. (1999). Cognitive integration of language Levelt, W. J. M. (1989). Speaking: From intention to articulation.
and memory in bilinguals: Semantic representation. Cambridge MA: MIT Press.
Psychological Bulletin, 125, 193–222. Levelt, W. J. M., Roelofs, A. & Meyer, A. S. (1999). A theory of
Gentner, D. & Goldin-Meadow, S. (eds.) (2003). Language in lexical access in speech production. Behavioral and Brain
mind: Advances in the study of language and thought. Sciences, 22, 1–75.
Cambridge, MA: MIT Press. Levinson, S. C. (1996). Relativity in spatial conception and
Gleason, H. A. (1961). An introduction to descriptive linguistics. description. In J. J. Gumperz & S. C. Levinson (eds.),
New York: Holt, Rinehart and Winston. Rethinking linguistic relativity, pp. 177–202. Cambridge:
Gordon, P. (2004). Numerical cognition without words: Cambridge University Press.
Evidence from Amazonia. Science, 306, 496–499. Levinson, S. C. (2001). Yeli Dnye and the theory of basic color
Green, D. W. (1998). Bilingualism and thought. Psychologica terms. Journal of Linguistic Anthropology, 10, 3–55.
Belgica, 38, 253–278. Levinson, S. C., Kita, S., Haun, D. B. M. & Rasch, B. H. (2002).
Hardin, C. L. & Maffi, L. (eds.) (1997). Color categories in Returning the tables: Language affects spatial reasoning.
thought and language. Cambridge: Cambridge University Cognition, 84, 155–188.
Press. Lucy, J. A. (1992). Grammatical categories and cognition:
Heider, E. R. & Olivier, D. C. (1972). The structure of the colour A case study of the linguistic relativity hypothesis.
space in naming and memory for two languages. Cognitive Cambridge: Cambridge University Press.
Psychology, 3, 337–354. Lucy, J. A. (1997). The linguistics of “color”. In Hardin & Maffi
Hoffman, C., Lau, I. & Johnson, D. R. (1986). The (eds.), pp. 320–346.
linguistic relativity of person cognition: An English– Lucy, J. A. & Gaskins, S. (2001). Grammatical categories
Chinese comparison. Journal of Personality and Social and the development of classification preferences: A
Psychology, 51, 1097–1105. comparative approach. In M. Bowerman & S. C. Levinson
 Bilingual blues 95

(eds.), Language acquisition and conceptual development, Roberson, D., Davidoff, J., Davies, I. R. L. & Shapiro, L. (2004).
pp. 257–283. Cambridge: Cambridge University Press. The development of color categories in two languages: A
Lucy, J. A. & Gaskins, S. (2003). Interaction of language longitudinal study. Journal of Experimental Psychology:
type and referent type in the development of nonverbal General, 133, 554–571.
classification preferences. In Gentner & Goldin-Meadow Roberson, D., Davidoff, J., Davies, I. R. L. & Shapiro, L. (2005).
(eds.), pp. 465–492. Colour categories: Evidence for the cultural relativity
Majid, A., Bowerman, M., Kita, S., Haun, D. B. M. & Levinson, hypothesis. Cognitive Psychology, 50, 378–411.
S. C. (2004). Can language restructure cognition? The case Roberson, D., Davies, I. [R. L.] & Davidoff, J. (2000).
for space. Trends in Cognitive Sciences, 8, 108–114. Colour categories are not universal: Replications and new
Malt, B. C. & Sloman, S. A. (2003). Linguistic diversity evidence from a Stone-age culture. Journal of Experimental
and object naming by non-native speakers of English. Psychology: General, 129, 369–398.
Bilingualism: Language and Cognition, 6, 46–67. Rosch-Heider, E. (1972). Universals in colour naming and mem-
Nation, P. (1990). Teaching and learning vocabulary. New York: ory. Journal of Experimental Psychology: General, 93, 10–
Newbury House/Harper Row. 20.
Özgen, E. & Davies, I. R. L. (1998). Turkish color terms: Tests Sapir, E. (1921). Language: An introduction to the study of
of Berlin and Kay’s theory of color universals and linguistic speech. New York: Harcourt, Brace & World, Inc.
relativity. Linguistics, 36, 919–956. Saunders, B. A. C. & van Brackel, J. (1997). Are there non-trivial
Pavlenko, A. (1999). New approaches to concepts in bilingual constraints on colour categorisation? Behavioral and Brain
memory. Bilingualism: Language and Cognition, 2, 209– Sciences, 20, 167–178.
230. Uchikawa, K. & Boynton, R. M. (1987). Categorical color
Pavlenko, A. (2005a). Bilingualism and thought. In A. M. B. perception of Japanese observers: Comparison with that
De Groot & J. F. Kroll (eds.), Handbook of bilingualism: of Americans. Vision Research, 27, 1825–1833.
Psycholinguistic approaches, pp. 433–453. Oxford: Oxford Whorf, B. L. (1956). Language, thought, and reality: Selected
University Press. writings of Benjamin Lee Whorf (J. B. Carroll, ed.).
Pavlenko, A. (2005b). Emotions and multilingualism. Cambridge, MA: MIT Press.
Cambridge: Cambridge University Press. Winawer, J., Witthoft, N., Frank, M. C., Wu, L., Wade,
Pavlenko, A. (2006). Bilingual selves. In A. Pavlenko (ed.), A. R. & Boroditsky, L. (2007). Russian blues reveal
Bilingual minds: Emotional experience, expression and effects of language on colour discrimination. Proceedings
representation, pp. 1–33. Clevedon: Multilingual Matters. of the National Academy of Sciences, 104, 7780–
Ratner, C. (1989). A sociohistorical critique of naturalistic 7785.
theories of color perception. Journal of Mind and Behavior, Zollinger, H. (1988). Categorical color perception: Influence of
10, 361–373. cultural factors on the differentiation of primary and derived
Regier, T., Kay, P. & Khetarpal, N. (2007). Color naming reflects basic color terms in color naming by Japanese children.
optimal partitions of colour space. Proceedings of the Vision Research, 28, 1379–1382.
National Academy of Sciences, 104, 1436–1441.
Roberson, D. (2005). Color categories are culturally diverse in Received: March 28, 2007
cognition as well as in language. Cross-Cultural Research, Final revision received: October 5, 2007
39, 56–71. Accepted: November 1, 2007
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.