To appear in Trends in Cognitive Sciences.
Language, thought and color:
Paul Kay1 and Terry Regier2
1International Computer Science Institute, University of California, Berkeley, USA
2University of Chicago, Illinois, USA.
Corresponding authors: Kay, P. (firstname.lastname@example.org); Regier, T. (email@example.com).
PK and TR contributed equally to this work.
The classic issue of color naming and color cognition has
been re-examined in a recent series of articles. Here, we
review these developments, and suggest that they move the
field beyond a familiar rhetoric of ‘nature versus nurture’, or
‘universals versus relativity’, to new concepts and new
The ‘Whorfian’ debate over color naming and color
cognition has been framed by two questions:
(1) Is color naming across languages largely a matter
of arbitrary linguistic convention?
(2) Do cross-language differences in color naming
cause corresponding differences in color cognition?
In the standard rhetoric of the debate, a ‘relativist’
argues that both answers are Yes, and a ‘universalist’
that both are No. However, several recent studies,
when viewed together, undermine these traditional
stances. These studies suggest instead that there are
universal tendencies in color naming (i.e. No to
question 1) but that naming differences across
languages do cause differences in color cognition (i.e.
Yes to question 2). These findings promise to move the
field beyond a conceptually tired oppositional rhetoric,
towards a fresher perspective that suggests several new
questions. Here, we review these recent studies, the
clarification they bring to the debate, and the further
questions they raise.
Color naming varies across languages; however, it has
long been held that this variation is constrained. Berlin
and Kay  found that color categories in 20 languages
were organized around universal ‘focal colors’ – those
colors corresponding principally to the best examples
of English ‘black’, ‘white’, ‘red’, ‘yellow’, ‘green’ and
‘blue’. Moreover, a classic set of studies by Eleanor
Rosch found that these focal colors were also
remembered more accurately than other colors, across
speakers of languages with different color naming
systems (e.g. ). Focal colors seemed to constitute a
universal cognitive basis for both color language and
The ‘relativist’ challenge
Recently, however, Debi Roberson and colleagues [3,4]
failed to replicate Rosch’s results. They compared
speakers of three languages: English, Berinmo, a
language of Papua New Guinea, and Himba, a Bantu
language – and did not find privileged memory,
similarity judgments or paired associates learning in
Berinmo and Himba at the proposed universal foci.
Instead, they found that these cognitive variables were
well predicted by the boundaries of each language’s
color categories. This is a form of ‘categorical
perception’ of color (categorical perception is said to
occur when stimuli that straddle a category boundary
are perceived as more distinct than equivalently spaced
stimuli within a category). Because color term
boundaries vary across languages (see Figure 1a,b),
speakers of different languages apprehend color
differently. Moreover, these linguistic differences
actually seem to cause, rather than merely correlate
with, cognitive differences , confirming and
extending earlier findings by Kay and Kempton. These
results call into question the cognitively privileged
status of the universal focal colors. And they provide a
positive answer to question 2 above: language
differences do cause differences in color cognition.
Roberson and colleagues have gone further,
proposing that universal foci play no central role in
color naming either (question 1). They argue that color
categories are determined at their boundaries by
language, and that best examples of categories are
mere epiphenomena of this process . The one
universal constraint they do acknowledge is ‘grouping
by similarity’ – the very general principle that similar
colors will tend to receive the same name. They also
emphasize that they have studied languages of nonindustrial
societies, suggesting that the Berlin and Kay
results – based mostly on languages of industrialized
societies – are parochial. Lucy has also argued against
universals of color naming . He suggested that Berlin
and Kay’s finding of universality was based on
hopelessly subjective methodology: the data had been
analyzed largely by human inspection, rather than
objective test. If these claims about color naming turn
out to be well founded, the overall picture would be a
clearly ‘relativist’ one: that is, a Yes answer to both of
our framing questions.
Current status of the debate
However, when the above-mentioned (‘relativist’) results
on color cognition are juxtaposed to some recent
(‘universalist’) findings on color naming, the traditional
stances break down. For despite the clear evidence that
To appear in Trends in Cognitive Sciences.
language affects color cognition, there is also new
evidence for color naming universals. Kay and Regier
 conducted the first comprehensive objective tests of
color naming universals – in part in response to the
‘relativist’ claims above – and found strong statistical
evidence of universal tendencies in color naming across
languages of both industrialized and non-industrialized
societies, the latter from the World Color Survey (WCS).
Moreover, there is evidence specifically for universal
focal colors in naming. Regier, Kay and Cook ,
extending earlier work by MacLaury , found that
best examples of color terms in the WCS strongly tend
to cluster near the proposed focal colors (Figure 1c).
This pattern would not be predicted if the only major
universal force in color naming was ‘grouping by
similarity’. Webster and Kay  found that the foci
vary somewhat in placement across languages – but
much less than the variation across speakers within a
language. The overall picture emerging is that color
categories appear to be organized around universal foci
(No to question 1) – and at the same time, differences
in color naming do induce differences in color cognition
(Yes to question 2).
Figure 1. Color categories in (a) English and (b)
Berinmo, shown on the same standard array of colors.
Color cognition varies across languages in accordance
with category boundaries. (Data redrawn from ). (c)
Nevertheless, variation in color naming is constrained
by universal foci . The contour plot shows the
number of best-example choices for color terms across
110 languages from non-industrialized societies, which
cluster near those of English (black dots).
This non-traditional pair of answers to our two main
questions suggests further questions that are currently
under investigation. Most broadly: which aspects of
color cognition shape language, and which are shaped
by it? How do these reciprocal influences work
together? Some initial answers are emerging, as we
What causes universal tendencies in color naming?
Several explanations for universals in color naming
have been proposed. Kuehni  posits
neurophysiological support for the cardinal colors red,
yellow, green and blue. Lindsey and Brown 
proposed that languages spoken near the equator tend
to lack separate terms for green and blue because
excessive exposure to ultraviolet radiation from
sunlight yellows the lenses of people living in this
region. However, this theory has been challenged
[13,14]. Shepard  suggested that the major
phenomenal hue axes, especially blue–yellow, derive
from evolutionary tuning to the predominant sources of
natural illumination. Yendrikhovskij  also showed
that the sources of color naming universals could
reside in evolutionary tuning to the most frequently
occurring colors in the environment. Jameson and
D’Andrade  argued that the universal focal colors
are salience maxima in color space and that universals
of color naming flow from a process that partitions
color space in a way that maximizes information. Steels
and Belpaeme  emphasize the role of inter-speaker
communication, with evidence from simulations of
interacting agents. In short, there is no lack of
explanations for universals of color naming, some
mutually consistent and others not.
What causes categorical perception of color, and is it really
It has been widely assumed that language is the cause
of color categorical perception. This is suggested
because – as we have seen – named category
boundaries vary across languages, and categorical
perception varies with them. However, Franklin and
Davies  have found startling evidence of categorical
perception at some of these same boundaries in prelinguistic
infants and toddlers in several language
groups. Thus, some categorical color distinctions
apparently exist before language, and could then be
reinforced, modulated or eliminated by learning a
Much of the evidence for categorical ‘perception’ of
color comes from tasks that involve memory; hence it
could be that the category effects stem from memory
rather than perception. Recently, however, Franklin et
al.  found that both adults and infants respond
categorically in a visual search task that minimizes the
involvement of memory. They concluded that the effect
was probably truly perceptual. This is a tentative
conclusion that deserves further investigation. The
perceptual status of ‘categorical perception’ of color is
currently an object of study, as is its status with
respect to innateness, learning and unlearning.
To appear in Trends in Cognitive Sciences.
The debate over color naming and cognition can be
clarified by discarding the traditional ‘universals versus
relativity’ framing, which collapses important
distinctions. There are universal constraints on color
naming, but at the same time, differences in color
naming across languages cause differences in color
cognition and/or perception. The source of the
universal constraints is not firmly established.
However, it appears that it can be said that nature
proposes and nurture disposes. Finally, ‘categorical
perception’ of color might well be perception sensu
stricto, but the jury is still out.
1 Berlin, B. and Kay, P. (1969) Basic Color Terms: Their
Universality and Evolution, University of California Press
2 Heider, E.A. (1972) Probabilities, sampling and ethnographic
method: the case of Dani colour names. Man 7, 448–466
3 Roberson, D. et al. (2000) Color categories are not universal:
replications and new evidence from a stone age culture. J. Exp.
Psychol. Gen. 129, 369–398
4 Roberson, D. et al. (2005) Color categories: evidence for the
cultural relativity hypothesis. Cogn. Psychol. 50, 378–411
5 Roberson, D. and Davidoff, J. (2000) The categorical
perception of colors and facial expressions: the effect of verbal
interference. Mem. Cogn. 28, 977–986
6 Lucy, J.A. (1997) The linguistics of color. In Color Categories in
Thought and Language (Hardin, C.I. and Maffi, L., eds), pp.
320–346, Cambridge University Press
7 Kay, P. and Regier, T. (2003) Resolving the question of color
naming universals. Proc. Natl. Acad. Sci. U. S. A. 100, 9085–
8 Regier, T. et al. (2005). Focal colors are universal after all. Proc.
Natl. Acad. Sci. U. S. A. 102, 8386–8391
9 MacLaury, R.E. (1997) Color and Cognition in Mesoamerica:
Constructing Categories as Vantages, University of Texas Press
10 Webster, M.A. and Kay, P. Individual and population
differences in focal colors. In The Anthropology of Color
(MacLaury, R.E. et al. eds), John Benjamins (in press)
11 Kuehni, R.G. (2005) Focal color variability and unique hue
variability. J. Cogn. Cult. 5, 409–426
12 Lindsey, D.T. and Brown, A.M. (2002) Color naming and the
phototoxic effects of sunlight on the eye. Psychol. Sci. 13, 506–
13 Hardy, J.L. et al. (2005) Color naming, lens aging, and grue:
what the optics of the aging eye can teach us about color
language. Psychol. Sci. 16, 321–327
14 Regier, T. and Kay, P. (2004) Color naming and sunlight.
Commentary on Lindsey and Brown (2002). Psychol. Sci. 15,
15 Shepard, R.N. (1992) The perceptual organization of colors. In
The Adapted Mind (Barkow, J. et al., eds), pp. 495–532, Oxford
16 Yendrikhovskij, S.N. (2001) Computing color categories from
statistics of natural images. J. Imaging Sci. Technol. 45, 409–
17 Jameson, K. and D’Andrade, R.G. (1997) It’s not really Red,
Green, Yellow, Blue: an inquiry into cognitive color space. In
Color Categorization in Thought and Language (Hardin, C.L.
and Maffi, L. eds), pp. 295–319, Cambridge University Press.
18 Steels, L. and Belpaeme, T. Coordinating perceptual categories
through language: a case study for colour. Behav. Brain Sci. (in
19 Franklin, A. and Davies, I.R.L. (2004) New evidence for infant
color categories. Br. J. Dev. Psychol. 22, 349–377
20 Franklin, A. et al. (2005) The nature of infant color
categorization: evidence from eye movements on a target
detection task. J. Exp. Child Psychol. 91, 227–248
To appear in Trends in Cognitive Sciences. Language, thought and color: recent developmentsPosted: 13 Agustus 2010 in colour language
To appear in Trends in Cognitive Sciences.