color terms : Berlin & Kay

[Arnold M. Zwicky]

On Oct 8, 2004, at 6:51 AM, Steve Kleinedler wrote:

> It's been over 10 years since I've read Berlin and Kay, so this is just
> vague recollection, but I don't think their point was that English
> speakers can't differentiate between light blue and dark blue, we
> obviously can, but that we conceptualize light blue and dark blue as
> blue
> and others conceptualize them as completely different colors.
>
> Compare this with the earlier orange/red discussion we had. We
> conceptualize orange/red differently than languages which have no word
> for
> orange.
>
> Berlin & Kay 1969
> Color Terms: Their Universality and Evolution
>
> is the basic book that lays all of this out and I highly recommend it
> if
> you're going to do any work in this realm, because it's the one text
> that
> anyone reviewing your work will have expected you to read.

an important correction: the title of the book is *Basic* Color Terms.
the idea is that the categorization of experience involves categories
of various sizes, but that in many domains people in a culture share a
*basic* level of categorization, which is the primary one they use for
quick identification of the nature of things and events, for rapid
sorting, for remembering bits of experience, and so on.  (this is not
to say that there won't be culturally relevant supercategories and
subcategories.)  on the whole, basic categories will have relatively
simple names in the relevant language, names that everybody knows,
offers early in lists of words for things in the relevant domain,
mostly agrees on the denotations of, etc.: the *basic vocabulary* or
*basic terms* in that language for that domain.

B&K was about basic color terms, in this sense of "basic".  later work
established that the basic color categories in a culture were
associated with classic categorical perception effects, much like those
demonstrated long ago in phonetics/phonology.  that is, if one group of
people distinguish basic categories A and B but another group treats
instances of A and B as belonging to a single category, then people in
the first group will be much more sensitive to small differences in the
boundary region between A and B, while people in the second group will
treat these differences the same as differences of an equal size within
A or within B.  put another way: people in the first group will
perceive some differences as much bigger than they actually are,
because they refer bits of experience to either A or B.

so, if you have a basic color distinction between BLUE and GREEN,
you'll see small differences in hue in the BLUE/GREEN boundary region
as being really big -- your perception will be biased -- while if you
have just one basic hue GRUE here, these differences will seem no
greater to you than similar hue differences within your BLUE and GREEN
(nonbasic) subcategories.  (you'll probably have such subcategories,
and might well call them something translatable as 'sky grue' and 'leaf
grue'.)

similarly, if you have a basic color distinction between LTBLUE and
DKBLUE, you'll see small differences in the LTBLUE/DKBLUE boundary
region as being really big -- your perception will be biased -- while
if you have just one basic BLUE category here, these differences will
seem no greater to you than similar differences within your LTBLUE 'sky
blue' and DKBLUE 'sea blue' (nonbasic) subcategories.

as it happens, most modern english speakers show categorical perception
effects for BLUE vs. GREEN but not for LTBLUE vs. DKBLUE.  much as they
do for [t] vs. [d] but not for aspirated vs., unaspirated [t].

nobody is claiming that differences in culture/language are associated
with differences in people's visual discriminatory abilities.
categorical perception effects can be suppressed in various ways,
though this takes some work; in normal circumstances you'll be
especially attentive to differences that count as basic category
differences, and much less attentive to differences that do not.

arnold (zwicky at csli.stanford.edu)