++plato's Color Naturalism

North American Philosophical Publications
PLATO'S COLOR NATURALISM

Author(s): Ekai Txapartegi
Source: History of Philosophy Quarterly, Vol. 28, No. 4 (OCTOBER 2011), pp. 319-337
Published by: University of Illinois Press on behalf of North American Philosophical
Publications
Stable URL: http://www.jstor.org/stable/23032317
Accessed: 21-02-2017 06:20 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
University of Illinois Press, North American Philosophical Publications are collaborating

with JSTOR to digitize, preserve and extend access to History of Philosophy Quarterly
This content downloaded from 132.248.9.8 on Tue, 21 Feb 2017 06:20:40 UTC
All use subject to http://about.jstor.org/terms
History of Philosophy Quarterly
Volume 28, Number 4, October 2011
Ekai Txapartegi
Plato's
withdoctrine on colors
some prevailing is in need of reexamination.
interpretations, Plato conceived In accordance
colors as
phenomenological objects.1 The central tenet of these renditions is that
the act of perception not only yields chromatic sensations but also par
ticipates in creating the qualitative character of the world we perceive.
The natural world would be devoid of any qualitative character outside
the deceptive act of perception. In this article, I propose to replace this
subjectivist reading with a naturalistic one.
In support of the naturalistic interpretation, not only did he write
that colors have natural essences (Cratylus 423e), but he also specu
lated about their precise identity as shown by his definition of colors
as flames that are constantly streaming off from external objects and
are commensurate with the visual ray (Timaeus 67c, Meno 76d). Plato's
colors, then, at least basic colors,2 cannot be purely subjective entities.
They are flames of a sort, objects of the natural world, like any others
(Ierodiakonou 2005, 329). They are also the causal actors that, in coop
eration with the visual flux (Timaeus 45b46a, Meno 76de, Theaetetus
152e), yield color sensations. Therefore, it seems most sensible to think
that, for Plato, at least basic colors cannot be purely subjective entities
and that he assigned to colors a more fundamental nature than their
mere manifestation in perception.
In the first part, I will present and criticize this subjective interpreta
tion of Plato's doctrine on colors. Since other authors have already done
much to discredit its most general aspects (mainly McDowell 1973 and
Burnyeat 1999), I will examine only a particular argument that, as of
yet, has not been addressed. Based on the analogy between the colors and
the letters of the alphabet, Gulley (1962) concluded that Plato thought
of colors as simple and, therefore, as fully knowable in the act of percep
tion. I will argue that Plato could not have accepted this consequence
because, for him, merely perceiving a color was not enough to recognize
its nature.
319
HISTORY OF PHILOSOPHY QUARTERLY
Then, I will frame Plato's doctrine on colors in the current debate

between physicalists and primitivists. My aim is to show, against the
suggestion made by Ierodiakonou (2005,229), that the new naturalistic
conception does not necessarily commit Plato to a physicalist conception
of colors as nonqualitative physical objects. In particular, I will argue that
there is no need for a semantic switch of color terms, from essentially
qualitative to nonessentially qualitative, to grant that Plato conceived
colors as independent dwellers of the natural world. As contemporary
primitivists do, he could have thought of colors as flamesthat is, as
ontologically independent entitiesthat are essentially qualitative.
1. A Colorless World
The dominant interpretation of Plato's doctrine on colors has prefe

a subjectivist interpretation to the naturalistic one. In this section,
tempt to show that the two main ideas that substantiate the subjectivi
interpretation have been wrongly attributed to Plato. First, there i
basis for him to endorse the idea that every causal account of chromat
sensation is committed to a nonqualitative natural world; and, seco
there is no basis for him to endorse the idea that every proper ob
of the sense, such as color, is epistemologically exhausted in the ac
perceiving that sense. Once we leave these two general assumptio
aside, I think that the naturalistic interpretation emerges again as
most sensible interpretation of Plato's conception on colors.
1.1. The Newtonian Model
Plato worked on a causal account of the production of chromatic sensa

tions (Timaeus 45c-d). Real colors, or flames, as he thought of them, are
partially causally responsible for some, if not all, chromatic sensations
('Timaeus 67c-68b). How close is this to the Newtonian model?
As is well known, the Newtonian model for the causal explanation
of chromatic vision combines three mini-theories: "atomism" about its
causes,3 "subjectivism" about its character, and "dispositionalism" about
its attribution.4 The outcome is that the real world is colorless (atomism)
and that the phenomenal character of the colored objects we perceive is
a subjective projection of the mind (subjectivism). Nevertheless, atomism
and subjectivism were not considered obstacles in justifying our ordinary
color attributions because, under normal circumstances, objects gener
ally have the disposition of constantly yielding in us approximately the
same chromatic responses (dispositionalism). This triplet summarizes
the color theory as outlined by Newton.
What is most probably fostering the Newtonian reading of Plato's
doctrine on colors is that any causal account of the production of chro
matic sensations, such as the one presented by Plato, is seen as nat

committed to the entire package of colorless causes, spurious chro
sensations, and an indirect justification of our ordinary color
tions. We might, however, question whether this assumption
to Plato. In fact, in this article I try to show that both models m
run parallel to each other, mainly because it is not certain tha
would agree to give atomism the additional thesis of a nonqua
natural world. Although Plato identified colors with flames th
chromatic sensations, it is uncertain that they were conceptu
lacking any qualitative character. He could have conceptualized
as essentially independent of the human perceptual system and
same time, as essentially qualitative. If Plato's atomism were d
from the idea of a nonqualitative natural world, then it would be
to commit him to either subjectivism about phenomenal chara
dispositionalism about color attributions.
1. 2. Heraclitean Ontology
Cornford (1937/1973) pushes a radical subjectivist interpretatio
Platonic account of colors. Following Aristotle (Metaphysics 98
1078bl2-16), he thinks that Timaeus expresses Plato's commit
to the Heraclitean ontology concerning the natural world (178
Theaetetus expresses his endorsement of Protaeoras's theory of pe
tion (49, 58).
If Cornford's subjectivist interpretation is right, the result wou
simply be that there cannot be any perceptual knowledge about th
sible world, something in line with Plato's own standards of ep
justification, but also that the natural world would become un
gible.5 As many commentators have already pointed out, there mu
something wrong with Cornford's reading.6 The author of Timaeu
not have conceived the natural world as completely unintelligib
Additionally, Burnyeat maintains that there are two readings
Theaetetus. In accordance with the reading that he calls "B,"
logue is intended to be a reductio ad absurdum. Plato would be
that the extreme application of the Heraclitean ontology on the n
world, together with Protagoras's theory of perception, collap
the impossibility of any natural language (Burnyeat 1999, 321).
if reading B of the Theaetetus were legitimate, Plato would be far
embracing the Heraclitean ontology as referred to the natural
At a minimum and in order to preserve the fact that we do n
sible qualities, one should put into brackets the total instability
natural world. The natural world in which colors are object-eff
was conceived by Plato as minimally stable and intelligible.
As a matter of fact, however, Cornford's view was not only that,

according to Plato, the natural world is colorless but also that it is
a phenomenological construction. Physical substances are chunks of
sensible qualities (Cornford 1937/1973, 180-81), and since sensible
qualities are created by individual, possibly idiosyncratic, acts of
perception (ibid., 40, 43), every physical substance is then indirectly
constructed in perception. The result of that view is some sort of so
lipsist epistemology, which Cornford also attributes to Plato (ibid., 51).
In sum, Plato would be saying something even stronger than just "the
universe of atoms is colorless": he would be saying, in the first place,
that there is no ontologically independent universe of atoms out there
that could be colored.
The main problem with Cornford's solipsist reading of Plato, how

ever, is that it leads to nihilism. If an efflux is not a substance but a
collection of sensible qualities and those sensible qualities do not exist
before the act of perception, then effluxes cannot exist before the act of
perception. But, now, if effluxes are causally necessary for perception to
occur (Timaeus 45b-d) and there is no efflux before the act of perception,
then no perception can occur. Therefore, without perception, not even
the natural world could exist.
It seems safe to say that Plato never aimed to espouse nihilism about
the natural world. Therefore, it is only natural to dismiss Cornford's
subjectivist interpretation and, instead, stick to the naturalistic one
according to which color flames are minimally stable as well as ontologi
cally independent from human cognition.
1. 3. Protagoras's Theory of Perception

At this point of the argument, it would appear that the naturalistic in
terpretation should be rid of Protagoras's theory of perception as well.
If we adhere to Burnyeat's reading B of the Theaetetus, we might take
for granted that Plato did not adopt Protagoras's theory of perception
because his argumentative purpose when presenting it was precisely to
show that both led to an impossibility of a natural language. But this
inference is not straightforward. Even though the combination of the
Heraclitean ontology and Protagoras's theory of perception could make
any natural language impossible, this argument alone does not tell us
whether we should abandon both (the Heraclitean ontology as well as
Protagoras's theory of perception) or simply eliminate one of the two.
Therefore, as a precautionary step, we still have to consider the hypoth
esis that Plato rejected the extreme stand of the Heraclitean ontology
but accepted Protagoras's theory of perception.
1.3.1. Phenomenological Colors

It is a common procedure to present as evidence in favor of Pro
theory of perception the fact that the correct interpretation of
mixtures of the Timaeus (67c-68d) seems mysterious to us.7
mixture, for example:"amber with black yields green"8 (68c). Fo
time, it has been assumed that the only way to explain the m
with the observational data was to claim that the references of color
concepts are highly relative to cultures.9 In particular, and even though
it is taken as uncontroversial that the color concepts employed by Plato
still refer to hues (Taylor 1928, 484), it has been believed that thei
qualitative extension remains hidden from us. This sort of explanatio
in turn indirectly advances Protagoras's theory of perception becaus
to be so culturally sensitive, color concepts must be constructed out of a
plethora of mere sensations (Timaeus 67c). In fact, the building blocks of
our color concepts are thought to be purely qualitative shades with some
characteristically mental traits10: subjective (Theaetetus 158a), relativ
(156e), private (154a), certain (179c), infallible (157e-160e), and so on
And, coincidentally, that is precisely how Plato describes them when
he exposes the end result of Protagoras's type of perception. Therefore,
colors, as sensible qualities created in the act of perception, are take
to be purely phenomenological shades, exhaustively knowable in the a
of perception. Their nature is thoroughly phenomenological, and ther
is no further objective nature to be discovered.
This analysis of the phenomenological nature of colors is obtained by
trying to explain the puzzling character of some of Plato's color mixture
with Protagoras's theory of perception as it is presented in Theaetetu
The problem is that Plato did not consider his color mixtures as odd
therefore, he did not have to explain them by appealing to Protagoras's
theory of perception. If, in order to explain the alleged cultural relativit
of color concepts, someone is in need of Protagoras's characterization of
color sensations, it is we, not he. Hence, we can accept both, that the color
mixtures are strange to us and that Plato did not endorse Protagoras
theory of perception, in particular the phenomenological conception
colors that is usually associated with it.
1.3.2. Colors and Letters
A more solid argument to support Plato's endorsement of Protagoras's

theory of perception comes from his preventive statements regarding
the extent of obtainable knowledge about colors. According to Plato,
it is hopeless for humans to try to dig into the exact composition of
the colors in nature because "no human being could possess either of
these [knowledge and power sufficient to blend the many into one and
to resolve the one into many], whether at the present time or at any
time in the future"11 (Timaeus 68d). One natural way of explaining this
epistemological prophecy is by appealing, as subjectivists do, to the
phenomenological nature of colors.12 No one will ever know their exact
composition because, apart from their appearance, there is nothing in
them to be known. Colors are subjective sensations that are episte
mologically exhausted in the act of perception. And although Plato's
endorsement of Protagoras's theory of perception is not the only way of
explaining his prophecy, some commentators, like Gulley (1962), assure
that this is the case. According to this interpretation, Plato concluded
that the nature of colors would never be known because he assented to
Protagoras's theory of perception for sensibles and to the idea, usually
associated with it, that colors are purely phenomenological entities.
Gulley supports his interpretation with an analogy between the
colors and the letters of the alphabet. He claims that Plato illustrated
the simplicity of sensible qualities comparing them with the letters of
the alphabet (Theaetetus 201d-202d). In the same way that simple and
unintelligible letters are used to construe meaningful linguistic struc
tures, the mind uses simple and unintelligible chromatic sensations to
make sense of the complex and intelligible reality (Gulley 1962, 99).
Therefore, colors are sensible qualities, which from the epistemological
point of view are simple. Since Plato considered reality, which is complex,
as the object of episteme, then, according to him, there could not be any
real knowledge about colors.
One strategy, deployed by McDowell against this conclusion, is to
deny that Plato took complexes as the only knowable entities. It is an
overstatement to affirm that every act of perception results in knowledge
(Theaetetus 184b-187a), but it would also be too farfetched to affirm that
every epistemic act is directed toward complexes (Theaetetus 202d-206c).
In fact, the so-called dream theory could be summarized in the thesis
that simple and complex entities cannot be distinguished in relation to
their cognizability (McDowell 1973, 240).
Unfortunately, this strategy is not satisfactory for a naturalistic in
terpretation. If colors are simple (and weakly knowable)13 then they will
be as the letters of the foreign language (simple and weakly knowable):
the only "knowledge" available would be perceptual recognition. Thus,
"this is red" would be analogous to "that was an R-sound." One could only
recognize colors to the same extent that one recognizes familiar sounds
while listening to some foreign language. Restraining our critics to the
idea that simple entities are not knowable does not make colors natu
ralizable. On the contrary, a naturalistic interpretation is committed to
show that colors, for Plato, had a complex nature not fully recogniz
in ordinary perception.
A more promising strategy for the naturalist interpretation
show that the repeated identification of colors with their corre
ing flames (Timaeus 67c, Meno 76c) implies that colors are not th
of as simply phenomenological entities. In this sense, I suppose,
misled Gulley to take colors as simple phenomenological entities
common confusion of taking the proper objects of the senses as if t
were the proper objects of perception.
As it is well known, colors, for Plato, are the proper objects of vi
in contrast with other senses (hearing, touch, taste, etc.). This s
means that we cannot recognize "white" by smelling (touching, e
Yet colors, for Plato, are not the proper objects of perception; they
also the objects of the soul since there can be a better way of kn
what "white" is than merely by looking at white objects. For ins
we could one day discover that the particles of "white" are smaller t
visual particles (Timaeus 67c-e), even though looking at white ob
will never give us that knowledge. For this reason, "white" is no
proper object of perception, if there is such a thing as the proper o
of perception. Something being the proper object of the senses does
make it the proper object of perception. In particular, the visual
nition of "white" does not automatically exclude the possibility o
soul's acquiring a deeper knowledge of its nature, as it does in th
of the letters of the alphabet.14 The analogy, therefore, does not w
Then, what could better explain the Platonic prophecy that onl
gods will ever be familiar with the real nature of colors? Accord
the naturalistic interpretation, the answer does not lie in the al
phenomenological nature of colors but rather in the size of the part
that constitute their real nature. Plato is quite explicit about this
he says that only gods possess the fundamental knowledge abou
laws concerning the composition of the basic elements (Timaeus
His caveat can also be taken as epistemological advice: anyone
wants to increase her knowledge about the nature of each color s
construct direct access to the world of tiny particles that constitute t
It denotes a highly pessimistic attitude concerning the prospects
human access toward the fundamental particles, but it also exp
why Plato restricts himself merely to the task of providing a po
model of what might have been the causal process from which c
emerge in nature, without labeling his description of that process as
actual "explanation." He was aware that this part of the Timaeu
sheer speculation based on conjectures and not on direct observa
But the very fact that he dared to speculate about it indicates that he
thought of the actual explanation as perfectly intelligible to humans.
In conclusion, if Plato conceived of colors as complexes, he could not
have accepted that colors are created and epistemologically exhausted
in the act of perception.15 Instead, it seems more reasonable to suppose
that, according to Plato, colors are flames composed by particles that are
too minute in size to be observed directly but are minimally stable and
intelligible, as well as ontologically independent of human perception.
2. Varieties of Color Naturalism
In agreement with the naturalistic interpretation of Plato's doctrin

colors are flames and flames are complexes. Colors as such cannot
exhaustively knowable in the act of perception; their essential natu
is independent of the subjects perceiving them16 and, were they
cessible, could be correctly identified. Nevertheless, the naturalist
rendition opens the possibility of two different epistemic routes
colors, one perceptual and the other "scientific." The first route leads t
the perceived color (color-as-a-sensation), the second to its deep natu
(color-as-in-nature). The question is, if the above is true, which of
two epistemic routes should be preferred. Those who endorse Plat
with the Democritean ontology for the natural world, in which col
in nature are nonqualitative, think that he would have preferred the
scientific route while dismissing perception as deceptive in this case.
Supporting this view, reference can be drawn from Plato's remarks on
the fact that only gods can grasp the deep nature of colors. Most prob
ably, Plato thought that ordinary perception is not enough to identify
correctly the real nature of colors.
In this section, I will frame Plato in a contemporary debate about the
nature of colors. My aim is to show that the physicalist reading that is
based on the Democritean ontology was not the only option available
to him. In particular, I want to stress that, even if Plato accepted the
fact that colors have some identifiable material nature, he could have
thought of them as qualitative outside the act of perception. The idea
is that the naturalistic interpretation of Plato is compatible with at
least two contemporary philosophical views, physicalism (on some of its
forms) and primitivism, since both views share the fundamental idea
that colors have identifiable essential natures (independent of their
perceiving subjects) that cause chromatic sensations.17
2. 1. Physicalism
The primary assertion of physicalism about colors is that colors are phys
ical properties of physical objects (surfaces, volumes, spaces, or sources of
light). That is to say, they are complex, categorical, and probably m

properties of objects. Colors are also thought to be constant prop
(with the possible exception of iridescent objects) and related to
Basically, the most general way of being a physicalist is to claim
colors are those physical propertiescomplex, categorical, and, prob
monadicof objects, constant and related to light, which normally c
human subjects to perceive a given object as being of a particular
However, but in a very similar spirit, chromatic primitivism holds
colors are primitive, sui generis properties, nonreducible to physical
more basic properties and not dependent on the perceptual respo
any subject. Ontologically, the most notable difference between thes
views is that primitivism rejects the reduction of colors to nonq
tive physical properties. The physicalist reading commits Plato
allegedly Democritean nonqualitative atomism, while the primi
reading does not. So, what reading should we favor?
There is an important epistemological difference as well. Primitiv
usually accepts the thesis of perceptual availability (Johnston 1
138), while physicalism, contrarily, holds that common percepti
not enough. Thus, according to physicalism, but not to primitivism,
support of science is needed to identify the real nature of colors
determine whether two distinct objects share the same color.
In the following paragraphs, I aim to use the criteria on whi
physicalists and primitivists disagree to determine which of th
comes closest to Plato's doctrine on colors. The first criterion consists of
knowing whether colors can be reduced to basic physical properties. It
seems that colors, for Plato, are completely reducible to flames.18 They
would, thus, be monadic, complex, constant, and closely related to the
fluxes emanating from objects. Therefore, if Plato identifies each color
with the figure, combination, and proportion of its fundamental par
ticles, that seems to be a typically physicalist reduction. In fact, when
Plato describes "whiteness," he seems to be speaking of that complex,
categorical, monadic, and constant something that objects give off and
that, in normal circumstances, causes them to be seen as white. As it
appears, the mere attempt to show the genesis of the essential nature
of colors seems enough to identify his doctrine as physicalist.
It so happens, however, that the kind 01 material objects colors are
would be special and, in a sense, problematic. The particles identified
with colors are not perceived as such. We are unable to see them as
particles and, therefore, unable to discern their shape, composition, or
proportions. This might simply be because they are too small for us. But,
in that case, one can only wonder whether these tiny objects that are
identified with colors are themselves colored (qualitative). It is possible
that, just as, in order for wood to be white, one would need, besides the
whiteness itself, the participation of the visual flux, in the same way, for
those particles to be white, the active participation of the visual flux is
also necessary. Obviously, this would indirectly support the thesis that
all white objects are relative to some perceiver, eventually even the most
fundamental particles as well, if they were qualitative at all. However, as
McDowell claims (1973,183), there is nothing in Theaetetus to indicate
that colors themselves must be colored.
Following the same opinion, some commentators are committed to the

view that, in the hands of Greek philosophers, the concept "whiteness"
dramatically changed its meaning. Ierodiakonou, for instance, instead
of postulating a Heraclitean ontology that lacks any substance, thinks
that Plato was working within the Democritean ontological framework
of minimally stable and intelligible colorless atoms. As a consequence,
the natural world cannot be colored because, ontologically, everything is
ultimately composed by colorless atomsat least not if we understand
color in its ordinary qualitative sense (Ierodiakonou 2005, 229). She
reads the definition of colors as flames as evidence that Plato tried to
change the ordinary semantics of color concepts. After the switch, color
terms refer only to the nonqualitative geometrical figures that consti
tute the object-effluences that cause chromatic sensations. With that
semantic move, similar in effect to the dispositional move promoted by
Isaac Newton, the atomic world could be legitimately referred to as col
ored againeven though its aspect is free of any qualitative ascription.
Thus, in Plato's precise use, color terms would no longer point to any
qualitative aspect characteristically attached to chromatic sensations.
The qualitative aspect would not be an essential part of their meaning
because Plato, according to Ierodiakonou (ibid.), in identifying colors with
flames, was saying that, once we use color terms properly, they strictly
and uniquely refer to some yet unknown specific nonqualitative flames.
2.2. Some Problems with the Physicalist Reading

The physicalist program links perceived colors to real colors in a merely
contingent way. The general criticism against the physicalist reading is
that chromatic concepts can no longer be regarded as experiential con
cepts. Reducing the chromatic vocabulary to another nonchromatic one,
where each color can be thoroughly described using purely geometrical
vocabulary, makes it no longer essential that each color appears in a
particular way to human subjects under normal conditions. This has a
disturbing consequence; classical Greeks might not have known what
eruthron or alourgon really meant. They could not be sure whether the
extension they associated with alourgon-objects is correct, since there
could be objects they saw as being alourgon that, in fact, were not
alourgon, and vice versa. That is to say, it could have been the
they were encountering pairs of metamer objects (Hilbert 19
same appearance but not the same physical composition. Ac
in order to know the true color of each object, it would not
merely to look at the object in daylight, but rather it wou
sary to await its scientific verification. Only experts would kn
they discover it, what alourgon or any other color concept mea
which objects they can be predicated veridically. All these co
clearly violate basic beliefs associated with the ordinary sem
chromatic concepts. Would Plato have been prepared to acc
consequences? He might have been. In contemporary philoso
have taken this point of view. Searle (2004, 121s), for insta
that colors are only contingently associated with their aspe
(1985) holds that colors need not be essentially qualitative19
(1997) propounds disjunctivism so as to avoid, in an ad hoc m
physicalist consequence that there is no way of knowing wh
objects are the same color before going through a laborator
unreasonable to assume that Plato also could have adopted an
philosophical views.
Nonetheless, there are some additional reasons to think t
physicalist reading of Plato's doctrine on colors cannot be
reading. Ontologically, it is questionable whether Plato's ontolog
natural world was influenced by Democritus and, if so, to what
There is no textual evidence to support the view that, accordin
perception creates not only sensations but also qualitative a
their similarities. When he says that gold is yellow (Timaeu
instance, he does not suggest that it is only so indirectly. In fac
ment with the definition of colors as flames, Republic (507d-e)
states that objects are colored before the act of perception
man, he is even clearer: the similarities we see are relations of
objects, and perception simply captures those objective sim
Therefore, epistemologically, chromatic vision only creates
tions; it represents what is already there. Thus, the main diffi
the physicalist reading is that it undermines the epistemic
Plato bestows on perception. Instead of detaching us from t
world, chromatic vision puts us in contact with it. The De
type of physicalism is compromised in its stance that we
colors directly (strictly speaking, we do not even see objects di
But, could this be true of Plato? It is one thing to say that
does not provide real knowledge. That reading of Plato seem
It is another thing, however, to say that for him chromatic pe
was deceptive. Plato never hints the latter, except perhaps
presenting Protagoras's theory of perception. Quite the con
epistemological value that Plato systematically bestows on perception

suggests that he has a qualitative natural world in mind.
Before turning to explore the hypothesis of a primitivist Plato, how
ever, it is worth explaining briefly why Plato could not be a physicalist of
the representationalist sort.23 On the specific thesis of representational
ism, perceived color can be reduced exhaustively to the basic physical
property represented by our visual system as color. Thus, the relation
between perceived color and physical color is not contingent, as originally
proposed by physicalism, but necessary, since it postulates an identity
relation between both of them. Whiteness determines how we see; and,
since the white color we perceive is nothing but a representational or
functional access to whiteness (that particular physical property that
is found in some objects), what we see is nothing but whiteness. This
proposal partially agrees with the Platonic one. For instance, in order
to define whiteness (the physical property), it is not necessary to invoke
the color white (the perceived color). Whiteness presents itself precisely
as that which causes something to be seen (to be represented) as white.
Thus, the identity of whiteness is defined by its function of generating
white color (when interacting causally with the visual flux), although
it may be supposed that its identity could also be characterized inde
pendently by describing its nature (the particular form, combination,
and proportion of its basic particles). The congruence of both views
seems remarkable. However, this could not be Plato's stand. According
to him, white color exists as an entity ontologically (although not caus
ally) independent from whiteness. In this sense, the white color is not
reducible to whiteness; thus, there can be no identity relation between
both of them as is postulated by physicalist representationalism.
2. 3. Primitivism
If one prefers neither to alter the semantics of chromatic concepts nor

to reduce the qualitative aspect of the white color to whitenessa
flameeliminating it from the natural world, one can still appeal to
the primitivist option. The distinctive thesis of primitivism holds that
the identity of colors, even if independent of the subjects who perceive
them, involves in an essential way their phenomenal or qualitative
aspect. Plato could have thought, for instance, that chromatic essences,
the flames, describe their deep nature but that, in turn, this nature is
essentially phenomenal, that is, it essentially contains some qualita
tive character. Whiteness could itself contain the qualitative character
or aspect that we attribute to white objects. This position would avoid
some of the problems associated with the physicalist or representation
alist construals, for which the qualitative aspect does not figure in the
fundamental descriptions of objects. It would not support that part of
the Democritean ontology nor would it question the epistemic

perception for capturing the traits of the natural world. Howe
main achievement of this primitivist reading would be semant
one could avoid postulating a dramatic change in the meaning
terms, from essentially qualitative to nonessentially qualitativ
In fact, reading the Cratylus (423e-24a) it seems clear that
did not aim to revise the meaning of color terms so as to mak
qualitative aspect nonessential. He explicitly stated that color
essences and that those essences express the nature of colors.
concepts were not essentially qualitative, as Ierodiakonou (20
gests, Plato would have said that painting colors could not express
real nature. However, in this same passage, he insists that it is pr
the painters who best express their nature. The essence of colo
best expresses their nature, is qualitative. Therefore, the ph
reading has to explain why Plato would write, if he thought of co
nonqualitative, that painters best express their nature.
This view is also in accordance with Struycken's explanation
puzzling color mixtures of Timaeus. We have already mentioned t
color mixtures of Timaeus have been thought inexplicable for a lo
Struycken (2003) has shown one way to surpass the semantic i
bility by postulating a strong link between ordinary color concep
their material base. Color terms are indirectly used to refer to pa
objects and processes. I think that this explanation is best und
if one accepts that Plato could have only used those color terms in
a way if he took those particular objects and processes to be es
qualitative, in the sense that the description of their qualitative c
plays an important role in the explanation of their change.
2.4. Some Problems with the Primitivist Reading

This position, however, does not completely break free from earlie
For example, primitivism does not guarantee that the relation
chromatic essence and its appearance among normal human s
under normal conditions, would be noncontingent. It may be
that objects have a given chromatic appearance but that human
are unable to perceive it or that they perceive it in a different way
if the relation is merely contingent and we have no epistemic rou
the knowledge of colors other than that of sight, how can we kno
we perceive the true color of objects? Normal perception does n
to underwrite the truthfulness of our chromatic ascriptions. That
again imply associating another semantics with those concepts
The only guarantee available to Plato to avoid postulating a
gent relation would be his faith that the gods would not have
us with eyes to see properties that are not really as we perceive them.
In that case, if the relation between colors and our perceptual access
to them would be of a divine necessity, then another equally important
problem would arise: that of species chauvinism. That is to say, if we
think that our chromatic response is truthful but we sense that other
animal species could have chromatic perceptions of the environment that
differ from ours, we must conclude that those other species misperceive
the chromatic composition of the environment. Normally, no philosopher
would want to commit herself to the empirical claim that any chromatic
visual system differing from ours systematically deceives those who use
it.24 Would Plato have been prepared to assume for himself these two
options: the guarantee of religion or species chauvinism? It is possible.
In sum, I believe that, from an ontological point of view, if we com
pare Plato's position with contemporary views, we do not have enough
textual evidence to decide whether he would be more sympathetic with
primitivism or with physicalism. To decide this, it would be crucial to
know whether he held colors to be essentially qualitative. Given his
distinction between whiteness and white color (Theaetetus 156e, 182b)
and the identification between whiteness and a flame (Timaeus 67c,
Meno 76d; see also McDowell 1973, 139-40), many have thought that
the answer would be negative. However, we cannot determine whether
he would have been prepared to accept the consequences of that answer
or whether he would have been inclined to explore other philosophical
paths, such as representationalism or primitivism. Epistemologically,
things are even more obscure. On the one hand, it seems that Plato
would have been closer to physicalism since, as he states, perception is
not enough to enable us to know the nature of a color. On the other hand,
the physicalist reading diminishes the epistemological value that Plato
never ceases to bestow on perception. Finally, I think that semantically
Plato would have been closer to primitivism since, as we have seen, color
terms as Plato used them never cease to be essentially qualitative.
3. Conclusion
Plato's doctrine on colors has been received, probably with Democritus's,

as the herald of Newton's modern conception. Apart from observatio
that his color science was not well developed and even if his color co
cepts were strange to us, his general conception of colors was taken to b
basically Newtonian. In particular, Plato was often considered to stan
for subjectivism about the nature of colors and relativism about the
attributions. Building on that tradition, it has also become a commo
practice among his commentators to use some genuinely modern theorie
such as sense-data or the veil of perception to describe his doctrine
colors. Nevertheless, this type of subjectivist interpretation has alre

been severely undermined. For instance, McDowell (1973, 143) a
that the language of sense-data is alien to Plato, and Burnyeat (
questioned whether Theaetetus necessarily commits Plato to eithe
Heraclitean ontology or to Protagoras's epistemology. Finally, Struyc
(2003) has directed our attention to the essential link that Plato d
in Timaeus between the origin of colors and their underlying na
processes. In the same spirit, I have argued, against Gulley (1
94-107), that Plato's analogy between colors and letters does not i
that colors are simple and, therefore, exhaustively knowable in the a
seeing them. I think that all these considerations open up the possibi
of a naturalistic interpretation of Plato's doctrine on colors, as explo
in this article. Colors partly furnish the natural world. However
relevantly, I have also shown that this naturalistic view does not
sarily commit Plato to the idea that the world itself is nonqualit
before the act of perception.
University of the Basque Country
NOTES
1. Some key examples of this rendition: "[D]ifferences in colour corres

to differences in the result of the encounter between this flame and the
light issuing from the eye" (Taylor 1928, 480); "The external thing '
white'; its surface is 'saturated with whiteness'. This last statement
difficult; the object is described as affected by the act of sight and a
colour. The meaning might be that the 'flame' or light belonging to the
cannot until this moment be called 'colour' or 'white'.. . . Plato's poin
these properties, whatever they are, are always changing, however
and that they are not the qualities I perceivemy sense-objectsa
should not be called 'black' or 'white'" (Cornford 1935/1973,50-51; "[A]ny
belongs to a physical object only in relation to human observers.. . . a
the external object is not white prior to perception, it becomes 'saturated
whiteness when it is perceived and becomes a 'white thing'" (Yolton 1
32); "The object of perception [according to Plato] is a phenomenal o
is created through the interaction of a sense organ and the external
(Modrak 2006, 140).
2. Plato neither defined nor discussed basic colors, i.e., a minimum
essary set from which one gets, by mixing them, all the rest. However,
commentators (with the exception of Taylor 1928) take it that, in the Ti
67c-68d, he is mixing "basic colors." They suspiciously disagree, thoug
Plato's list of basic colors. Struycken (2003, 279) cuts the list in tw
and white. Theophrastus includes bright as well. Ierodiakonou (2005, 227)

adds red, thereby enlarging Theophrastus's list to four. Levidis also ascribes
to Plato with a list of four, but his fourth (2002) is not Theophrastus's bright
but ochre.
3. I present Newton's view as atomist for simplicity of exposition. Histori

cally, he was considered a proponent of the corpuscular theory of light, even
though some authors have questioned this in favor of the wave hypothesis. See,
for instance, Whittaker (1931, xi), Cantor (1983, 31), or Hall (1993,163-67).
4. For a more extensive philosophical analysis of the Newtonian model,
see Hacker 1987.
5. "As Geoffrey Lloyd has pointed out, Cornford's reading, in effect, collapse
Timaeus' position into a Parmenidean-style dismissal of the physical world
unintelligible" (Kjeller Johansen 2004, 162).
6. Gulley, for instance, thinks that in Cratylus and Theaetetus Plato
accepted the Heraclitean theory of flux for sensibles (1962, 73-76) withou
embracing the extreme stand of total ontological instability (ibid., 27). Plat
must have acknowledged that sensibles were minimally stable, at least, stabl
enough to be perceivable and nameable (ibid., 96). Modrak seems to share th
same view. She thinks that, by not ascribing to Plato the extreme Heraclitea
ontology, the natural world, as he conceived it, becomes intelligible again (2006,
140). This is also in line with McDowell's favored interpretation. According t
him, Theaetetus is, in fact, endorsing the use of the verb is as applied to sensibl
qualities (1973,125-27,133), and a necessary condition for this is a minimally
stable natural world.
7. Taylor thinks that Plato is "simply indicating resemblances of colour

sensations themselves" (1928, 485), while according to Cornford, Plato is
"speaking of compound colours as if they were pigments" (1937/1973, 276).
8. Timaeus 68c: Huppou Se |ieA.avi npaaiov.
9. Taylor 1928, 479; Cornford 1937/1973, 276. For arguments against this
interpretation, see Struycken 2003.
10. "From this mundane starting-point [Protagoras's theory of perception] it
is inferred that the colour white, for example, is not inherent in the object, not
a feature which characterizes the object in itself; white exists only in relation
to a given observer, 'between' their eye and the object (154a), and it is therefore
private to them (154a; cf. 161d, 166c), something of which only they can be
aware" (Burnyeat 1999, 326).
11. Timaeus 68d: (...) Kai Suvatoq, avBpdmcov 6s oi)6elq ouSerepa toutcov
iKctvoc; oi)T ectti vOv ours eiq atiSiq tote eaxai.
12. If Plato had thought of colors as mere appearances, he would not have
said that no one would ever know their nature, as he didrather the opposite,
namely, that each person, suited with a properly functioning color perception
system, will readily know his or her exact nature. (I would like to thank the
referee of the journal for this observation.)
13. Here, "weakly knowable" simply means able to recognize an ins

it when it appears. "He will know it, in the sense of knowing whatspe
what letterit is" McDowell 1973,159-60.
14. McDowell also shows that Plato does not use the word perceptio
ply propositional content in it (1973,118,160,192, and 196). Like child
could perceive "red" even if one does not know "what red is." Another e
is that we can see "Michael's mom" even if we do not know that the woman we
are seeing is, in fact, "Michael's mom."
15. Ieradiakonou also thinks that the theory of perception outlined in The
aetetus "is attributed to Protagoras' secret doctrine and should not be taken to
be Plato's own" (2005, 231).
16. Although metaphysically they depend on their teleological function in
relation to the act of perceiving.
17. For a mapping of physicalism, see Txapartegi 2004; of primitivism, Byrne
and Hilbert 2007.
18. Colors would not be properties (in a strict sense, they would be objects),
even though categorically ascribable to seen objects, in the same way as fire is
ascribable to the object burning.
19. And, therefore, essentially qualitative concepts cannot be color concepts:
"Yellow is not a color" (Averill 1985, 301).
20. See, for instance, Taylor (1928,489) or Gage (1993,12). I also think that
Theophrastus (60) is revealing about this specific aspect.
21. "Likenesses which the senses can grasp are available in nature to those
real existents which are in themselves easy to understand, so that when some
one asks for an account of these existents one has no trouble at allone can
simply indicate the sensible likeness and dispense with any account in words
"(...) toIc; |iev tu)v ovxcov paSkoc; Kara^aOdv a'ia0r|Tcu Tivec; 6|ioioTr|T<; ntcpuxaaiv
ac; ouSev xaAendv Sr|\ouv, otav auxtijv Tiq (3ou\r)0fj T(I> \oyov (...)" Statesman
285d-e. See also The Republic 508 c-d.
22. Frede is explicit attributing to Plato a version of the veil of perception
"What we perceive, strictly speaking, are just the proper objects of the differ
ent senses, e.g. colours in the case of sight (184e7 ff.). Thus, strictly speaking,
we do not even perceive the object of which we come to believe that it is red"
(1999, 382). Kjeller Johansen also draws an explicit parallelism with Newton
"In other words, it seems that the mode in which the dunamis is reported is the
perceptible quality associated with the physical affection which the dunami
brings about rather than the dunamis itself. ... If so, an extensional view o
perception seems implied. Similarly, you might be said to see light refracted at
a certain wavelength when you see red" (2004,171). But the clearest is Modrak:
"We do not have direct access to physical objects" (2006,141).
23. Two recent formulations of this position are Tye 2000 and Jackson 2007.
24. A possible way out could exist if Plato was prepared to assume that
each species detects different combinations of elementary particles. Since Plato
openly states his ignorance on the natural bases to be identified with each color,
it could be the case that something so complex emanates from colored objects
that every animal species only manages to detect in those colored objects what
is of interest to itself, this being in each case a potentially different color. Ac
cordingly, objects would not have only one real color, but a possible range of
true colors (and qualitative aspects associated with them). In that case, he could
accept chromatic diversity while rejecting species chauvinism.
REFERENCES
Averill, E. W. 1985. "Color and the Anthropocentric Problem." Jour

losophy 82: 281-304.
Benson, H., ed. 2006. A Companion to Plato. Oxford: Blackwell Publ
Burnyeat, M. F. 1999. "Knowledge Is Perception." In Fine, ed., 320-5
Byrne, A., and Hilbert, D., eds. 1997. Readings on Color. Cambridge
Press.
. 2007. "Color Primitivism." Erkenntnis 66: 73-105.
Cantor, G. N. 1983. Optics after Newton. Manchester, UK: Manchester Univer

sity Press.
Cornford, F. M. 1935/1973. Pla to's Theory of Knowledge. Oxford: Oxford Uni
versity Press.
. 1937. Plato's Cosmology. New York: Bobbs-Merrill.
Fine, G., ed. 1999. Plato: Metaphysics and Epistemology. Oxford: Oxford Uni
versity Press.
Frede, M. 1999. "Observations on Perception in Plato's Later Dialogues." In
Fine ed., 377-83.
Gage, J. 1993. Colour and Culture. London: Thames and Hudson.
Gulley, N. 1962. Plato's Theory of Knowledge. London: Methuen & Co.
Hacker, P. M. S. 1987. Appearance and Reality. New York: Blackwell.
Hall, A. R. 1993. All Was Light. Oxford: Oxford University Press.
Hilbert, D. 1987. Color and Color Perception. Stanford, CA: CSLI.
Ierodiakonou, K. 2005. "Plato's Theory of Colours in the Timaeus." Rhizai 2:
219-33.
Jackson, F. 2007. "Colour for Representationalists." Erkenntnis 66: 169-8

Johnston, M. 1997. "How to Speak of the Colors." In Byrne and Hilbert,
137-76.
Kjeller Johansen, T. 2004. Plato's Natural Philosophy. Cambridge: Ca

University Press.
Levidis, A. 2002. "Why Did Plato Not Suffer of Color Blindness?" In
Ancient Greece, edited by M. A. Tiverios et al., 9-22. Thessaloniki: Ari
University of Thessaloniki.
McDowell, J. 1973. Plato: Theaetetus. Oxford: Oxford University Press.
Modrak, D. 2006. "Plato: A Theory of Perception or a Nod to Sensat
Benson, ed., 133-45.
Penner, T. 2006. "The Forms and the Sciences in Socrates and Plato." In B
ed., 165-83.
Plato. 1900. "Theaetetus" and "Statesman." In Platonis opera, vol. 1, e

J. Burnet, St I.142a-210d and St II.257a-311c. Oxford: Clarendon Pr
. 1902. "Timaeus." In Platonis opera, vol. 4, edited by J. Burnet, St
92c. Oxford: Clarendon Press.
. 1961. Plato's Statesman. Translated by J. B. Skemp. Princeton, NJ:

Princeton University Press.
. 1973. Plato: Theaetetus. Translated by J. H. McDowell. Oxford: Oxford
University Press.
. 1997. Plato's Timaeus. Translated by D. Zeyl. Indianapolis, IN: Hackett
Publishing Company.
Searle, J. 2004. Mind. Oxford: Oxford University Press.
Smart, J. J. C. 1997. "On Some Criticisms of a Physicalist Theory of Colors." In
Byrne and Hilbert, eds., 1-11.
Struycken, P. 2003. "Colour Mixtures according to Democritus and Plato."
Mnemosyne 61: 273-305.
Taylor, A. E. 1928. A Commentary on Plato's Timaeus, Oxford: Oxford Univer
sity Press.
Txapartegi, E. 2004. Color y representacion. Bilbo: UPV/EHU University Press.

. 2009. "Real Colors and Chromatic Forms." Elenchos 30 (2): 337-45.
Tye, M. 2000. Consciousness, Color, and Content. Cambridge, MA: MIT Press.
Whittaker, E. T. 1931. "Introduction to Optics!' In I Newton, I. Optics, ix-xxv.
London: Bell & Sons.
Yolton, J.W. 1949. "The Ontological Status of Sense-Data in Plato's Theory of

Perception." Review of Metaphysics 3 (1): 21-58.

++plato's Color Naturalism

Uploaded by

Copyright:

Available Formats

++plato's Color Naturalism

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

++plato's Color Naturalism

Uploaded by

Copyright:

Available Formats

North American Philosophical Publications

PLATO'S COLOR NATURALISM

University of Illinois Press, North American Philosophical Publications are collaborating

PLATO'S COLOR NATURALISM

Then, I will frame Plato's doctrine on colors in the current debate

The dominant interpretation of Plato's doctrine on colors has prefe

1.1. The Newtonian Model

Plato worked on a causal account of the production of chromatic sensa

matic sensations, such as the one presented by Plato, is seen as nat

As a matter of fact, however, Cornford's view was not only that,

The main problem with Cornford's solipsist reading of Plato, how

1. 3. Protagoras's Theory of Perception

1.3.1. Phenomenological Colors

A more solid argument to support Plato's endorsement of Protagoras's

2. Varieties of Color Naturalism

In agreement with the naturalistic interpretation of Plato's doctrin

light). That is to say, they are complex, categorical, and probably m

Following the same opinion, some commentators are committed to the

2.2. Some Problems with the Physicalist Reading

epistemological value that Plato systematically bestows on perception

If one prefers neither to alter the semantics of chromatic concepts nor

the Democritean ontology nor would it question the epistemic

2.4. Some Problems with the Primitivist Reading

Plato's doctrine on colors has been received, probably with Democritus's,

colors. Nevertheless, this type of subjectivist interpretation has alre

University of the Basque Country

1. Some key examples of this rendition: "[D]ifferences in colour corres

and white. Theophrastus includes bright as well. Ierodiakonou (2005, 227)

3. I present Newton's view as atomist for simplicity of exposition. Histori

7. Taylor thinks that Plato is "simply indicating resemblances of colour

13. Here, "weakly knowable" simply means able to recognize an ins

Averill, E. W. 1985. "Color and the Anthropocentric Problem." Jour

. 2007. "Color Primitivism." Erkenntnis 66: 73-105.

Cantor, G. N. 1983. Optics after Newton. Manchester, UK: Manchester Univer

Jackson, F. 2007. "Colour for Representationalists." Erkenntnis 66: 169-8

Kjeller Johansen, T. 2004. Plato's Natural Philosophy. Cambridge: Ca

Plato. 1900. "Theaetetus" and "Statesman." In Platonis opera, vol. 1, e

. 1961. Plato's Statesman. Translated by J. B. Skemp. Princeton, NJ:

Txapartegi, E. 2004. Color y representacion. Bilbo: UPV/EHU University Press.

Yolton, J.W. 1949. "The Ontological Status of Sense-Data in Plato's Theory of

You might also like