Navigazione – Mappa del sito

HomeNumeri43ontologia dei coloriDefault beliefs on colors: the me...

ontologia dei colori

Default beliefs on colors: the methodological value of what we believe to know about colors

Ekai Txapartegi
p. 213-229


In this article I provide some reasons to justify why the preferable ontological account about colors is that which respects the highest number of beliefs contained in the so-called default position, keeping in mind the relative weight of each one of them. The full system of associated beliefs about colors contained in the default position is also offered.

Torna su

Testo integrale


1Given the vast scientific knowledge about colors, it should be uncomplicated to point out their nature. Yet, it is not.

2There is a manifest lack of consensus in this field that casts doubt about the methodology we employ when practicing color ontology. Within all available color theories, such as subjectivism, dispositionalism, relationalism, primitivism, physicalism, etc., how do we know which one is the right one?

  • 1 Color perception is usually pictured as a set of biological mechanisms whose function is to detect (...)

3The naturalistic view requires us to set our preferences based on current scientific knowledge. According to various scientific disciplines, colors are tightly related to the surface spectral reflectances of the objects1. Consequently, many philosophers have embraced a physicalist conception about colors, identifying particular colors with particular types of surface spectral reflectances of the objects.

4In this article I do not want to address this popular choice, physicalism, but rather its underlying methodological assumption, namely, that we should set our ontological preferences based on current scientific knowledge.

5In the first section, I provide some reasons to “rescue” the methodological value of what we believe to know about colors prior to acquiring any scientific knowledge about them. I do not pretend to diminish the methodological value of science in disclosing the real nature of colors. I only want to note the risk of doing bad ontology if we do not insert the scientific knowledge within a wider net of accepted beliefs about colors.

6The idea is that the best ontological theory about colors is that which respects most of the beliefs in the so-called “default position”, given the relative weight of each of them. According to this methodology, all our commonly accepted beliefs about colors are relevant, to a greater or lesser degree, to make our best ontological choice.

7This kind of methodology might not be new for some. What might be new, I believe, is the articulation of the default position that I present in the second section. I will try to introduce, in a systematic way, an exhaustive list of beliefs about colors that we commonly accept prior to engaging in any philosophical dispute about the nature of colors.

1. Why scientific knowledge is not enough

8There are two main reasons to explain the decrease of confidence in the direct ontological results originating from the sciences that study colors.

1.1. One science, various ontologies

9Science is no longer seen as implying only one ontological view about the nature of colors.

10As is known, the blossoming of the natural sciences led most modern philosophers to believe that colors were secondary qualities. Within this view coexists, even today, two color theories: subjectivism and dispositionalism. Subjectivism responds to the question “What are colors?”, while dispositionalism addresses the demand of relocating colors in the objects given that the secondary quality view subtracts colors from the list of categorical properties truly attributable to objects. According to this philosophical tradition, when we attribute colors to objects we are only attributing the disposition of those objects to appear a certain way to a specific type of subjects.

  • 2 Even nowadays, this methodological view continues having many followers. Clyde Hardin, for example, (...)

11For a long time other ontological options were simply rejected as naive because allegedly they were already surpassed by the available scientific knowledge. The philosophical debate about the constitutional nature of colors seemed to remain permanently settled by the scientific image of the physical reality2.

12The advance of certain scientific disciplines such as psychology or biology has finally cracked this consensus. Some newly discovered chromatic facts now point in other directions as well, such as physicalism or the relational theory.

  • 3 Locke, J. Essay, book II, ix, § 8.
  • 4 See Land 1977.
  • 5 Hilbert 1987.

13Consider, for instance, the psychological phenomenon of color constancy. This known phenomenon, that was already mentioned by Locke3, is explained by Land’s Retinex theory4 as the elimination of the illuminant. This explanation that comes from the cognitive or computational psychology suggests that the chromatic vision, in its basics, is “designed” to detect certain physical properties of objects, independent of the atmospheric lighting conditions in which the object is found. Some philosophers have rapidly seen its ontological consequence: if our chromatic system is designed to perceive the surface spectral reflectances of the objects, then colors should be identified with those surface spectral reflectances5.

  • 6 Thompson et. al. 1992; Matthen 1999.

14A second example where the sciences deviate off of its road toward the secondary quality view is offered by the comparative analysis of diverse chromatic systems. The variability of the chromatic vision among species or among evolutionary phases of the same species, according to the respective pragmatically specific biological functions, seems to support the relational theory of colors. Given the divergences and the resemblances among the chromatic systems, the chromatic vision should not be defined independently of any perceptive system. Therefore, some other philosophers think that the subjectivist theory as well as the objectivist theory should be revised to adopt some other view that is essentially relational6.

15It is relatively easy to find other ontological theories that rest on other color facts. Therefore, it seems unreasonable to assume that scientific knowledge alone could settle the ontological debate about the true nature of colors, unless all color facts end up pointing to a single ontology.

1.2. Changing the subject

16Lets consider that last hypothesis. Imagine that all known color facts end up pointing uniquely to a single ontology. Shall we accept that ontology straight away? I would answer no.

17The underlying methodological assumption for the “yes” answer is that, with regards to ontology, phenomenologically generated beliefs have less methodological value than the scientifically generated ones. But, do they? I will try to show the contrary, that from the methodological point of view, phenomenologically generated color beliefs set up the subject matter and, therefore, they are as necessary as the scientifically generated ones to make our best ontological choice.

18I will take it for granted a relation of dependence between our chromatic phenomenology and most ordinary color beliefs. For instance, it seems clear to me that our conceiving them as properties of objects directly depends on our perception of colors as located in objects. In general, the semantics associated with chromatic concepts seem to be in perfect harmony with the way in which colors appear to us in perception.

19The same concordance occurs with our epistemic or ontological beliefs about colors; they also seem to be consistent with the chromatic phenomenology. For instance, the epistemic belief according to which to know a color it is necessary to have seen it, or the ontological belief according to which objects have at least one true color, are cases of beliefs that are consistent with the chromatic phenomenology from which, most probably, they are originated.

20The fundamental idea is that our most common beliefs about colors find its justification, or part of its justification, in the way we perceive them. It is in this sense that I claim that the chromatic beliefs that we commonly accept depend on our shared chromatic phenomenology.

  • 7 Averill 1985: 301.
  • 8 Hardin 2003: 202.

21However, some philosophers think that all those phenomenologically generated beliefs are disposable. According to them, the phenomenology shows only the perceptual appearance of things while science uncovers how they are in reality. Fortunately, we have already seen what happens if we abandon some of those beliefs. You could end up saying, as a known physicalist did, that «yellow is not a color»7. Or as Hardin, saying that «we need not invoke the colors of common-sense realism at all»8. It is not difficult to find more examples of this sort.

22The problem with taking those beliefs as easily “disposable”, when they collide with the ontological view that is believed to be implied by scientific facts, is that it gives the philosophers the illusion of power to question every part of the attributive semantics associated with our ordinary chromatic concepts. The force of attraction of those “scientific facts” would be so intense that they could be used to impose new orbits of semantics, epistemology and ontology to color concepts.

  • 9 Ross 2001.

23In other words, the unacceptable consequence of assuming the “disposability” of phenomenologically grounded beliefs is the risk that philosophers might change the subject. Because there is no interest in claiming that colors are surface spectral reflectances if that claim was not about “yellow”. With the same spirit, no matter how scientifically justified subjectivism may be, it would lack most of its philosophical interest if it does not address the problem of color location9.

24In this need of constant thematic anchorage, chromatic phenomenology and the beliefs that are extracted from it have proven to be of vital methodological importance. Likewise, scientific knowledge is nothing more than a source of constrictions that any color ontology should respect. Some scientific facts can be utilized as counter-examples for certain metaphysical theories, but cannot be utilized in a more positive way to prove the “truth” of any metaphysical theory.

2. The default position

25In this section, I will present the default position or, said in another manner, what we know, or we believe to know, about colors. By default position, I understand the assembly of ideas that we commonly accept when speaking about colors. This means that the default position does not collect any original ideas; it is limited to those that we normally accept about colors.

26This recollection may seem philosophically sterile. Nevertheless, if the methodology I am advancing is correct, the default position is of great practical importance because, aside from clarifying the thematic environment, and being the starting point to devise any philosophical theory about color, it also offers a measure to be able to judge the best philosophical theory about the constitutional nature of colors.

2.1. Moore’s Experiment

27I propose to start off with what we are most secure with, our most basic beliefs about colors.

28Consider Moore’s popular experiment. The eminent professor approaches the audience, shows a red object and asks what they know given the perception that they are having of that red object. There are at least four ideas, relating to colors, that the hypothetical members of the audience know, or believe to know, when faced with that scene:

(1) we know that we perceive a color,

(2) we know how to identify a given color,

(3) if we consider that the circumstances are adequate, we are able to attribute that color to that object and,

(4) if we consider that the circumstances are not adequate, we would know how to identify better suited circumstances to carry out that attribution.

29This seems to be the minimum that we know (or that we believe to know) about colors, given Moore’s simple experiment. Evidently, the four statements, (1)-(4), refer to a type of practical knowledge that governs the guidelines of normal behavior relating to colors. That is to say, we have not yet introduced any genuinely philosophical or scientific consideration of said “knowledge”. We have simply shown that we know how to identify a color and to attribute it to the object that it corresponds in a conscious manner.

30According to (1), when we perceive a color we know that we perceive a color. At least, some times, when we perceive a color, we can know that we perceive a color.

31Evidently (1), as it is formulated, does not imply the strongest thesis, one that affirms the need to know what a given color is in order to perceive it. In the case of the color red, for example, it is reasonable to think that a red object can be perceived without knowing that it is a red object. It could be that the lighting is not adequate and we think that it is an orange object when, in reality, it is a red object. Other times, however, we are simply not conscious of our chromatic perceptions. For instance, if we are asked what color is the book that we just left on the table, it is possible that we would not know the answer in spite of the fact that we had it in our hands five minutes ago.

32There is also another excessive interpretation of (1) according to which it is not possible to perceive a red object without first understanding the concept “red” or without knowing that red is a color. The belief contained in (1) does not imply any of the two ideas. In fact, it does not seem reasonable to state that language or reflection are necessary conditions for the normal operation of chromatic vision. It seems unreasonable to assume, for only philosophical reasons, that all subjects without language or without reflexive capacity lack chromatic vision or that this linguistic or reflexive deficiency prevents them from perceiving the red color, just as we perceive it.

33Nevertheless, apart from whether or not reflection is necessary for chromatic vision what (1) affirms, simply, is that in human beings chromatic vision can combine perfectly with reflection. If we reflect on our perceptive behavior we know when we are perceiving a color and (1) is limited to that affirmation.

34On the other hand, according to (2) and (3), we normally possess sufficient ability to identify colors and to attribute them to objects. If, just as it occurs normally, we have the appropriate chromatic capacities and if, through socialization, we have acquired the linguistic capacity that is required in the management of the standard chromatic terms, which also occurs normally, then we know how to indicate a determined color (for example, “‘red’ is this →”) and we also know, how to attribute a color to an object (for example, “this → is red”). At least such is the case with the hues that we are used to, like red, green, yellow, blue, white and black. In fact, neither the ostensive definition nor the ordinary attribution of colors should be considered as a type of privileged knowledge. Normally, anyone can indicate the color red in a determined scene and anyone can say what objects are the red objects (if found in the adequate conditions for it) without any problem.

  • 10 Myin 2001.

35In relation to (2) we can even affirm, given that currently linguistic and chromatic incapacities are scientifically detectable10, that we know that the majority of subjects satisfy the two aforementioned conditions. That is to say, the majority of people possess the appropriate linguistic and chromatic capacities to identify colors and to attribute them to objects. From this fact we can infer, against hypotheses such as the inverted spectrum, that given the same circumstances, the normal subjects situated in front of a red object will basically perceive the same quality and, therefore, will attribute the same property as when we say “this is red”. In conclusion, we believe

(5) that our chromatic concepts basically share the same color denotation.

36And, relating to (3), we can complete that idea by saying that if Moore presented us with a second object, and the circumstances remained adequate, we would be able to know whether or not the second object shared the same color with the first object:

(6) In normal circumstances, ordinary perception is sufficient to know whether or not two objects share the same color.

37Finally, according to (4), in the event that some controversy arises around the color of a determined object, we would know how to find the way to settle that controversy. At least, in cases where the objects are not especially problematic (such as fire, iridescent objects, holograms, etc.) and have a single color. In fact, most of the time it seems simple, at least from the practical point of view, to determine which are the most appropriate circumstances to know the color of a determined object. This is such in everyday environments as it is in highly especalized and exclusive ones such as, for example, industry, commerce or certain scientific areas. The usual way to settle controversies is by rigidification, i.e. the specification of subjects and circumstances that are considered “normal”, according to a determined standard pragmatically established. That is to say, according to the commonly used method, first they tend to exclude the circumstances that they don’t consider normal (situations with poor lighting or with a medium that distorts perception, etc.) and, later, if more precision is required, they continue proofing circumstances against others. The more precision the result requires, the more precise the definition of the circumstances and the subjects considered as “normal” must also be. Therefore, I take it that the notion of normality that should be accepted in the default position is a gradual and pragmatically determined concept.

38I will call the assembly of beliefs (1-6) “epistemological beliefs”. This is because they are beliefs that we acquire or we justify based on ordinary perception just as Moore’s experiment shows – except for (5) that, in spite of the fact that it is a common supposition, it appears to have been confirmed scientifically-.

39In fact, we have already accepted that ordinary perception (along with socialization) seems to be sufficient to know how to identify a color, (2), and to attribute a color to an object, (3).

40Now, is ordinary perception also necessary to carry out those simple operations? Common sense responds affirmatively, that is, if nobody ever perceived the color “red”, then nobody would be able to indicate any property of an object as “red” neither would he be able to know which objects are red and which are not. Therefore, the seventh epistemological idea that is included in the default position’s list is:

(7) if we did not have the chromatic capacities that we possess, we would not have any of the previous beliefs (1-6).

2.2. Phenomenological beliefs

41Up to this point we have identified the beliefs about color that are justified by mere perception of any colored object. Nevertheless, in the act of ordinary perception certain “phenomenological” beliefs are also acquired that determine, in some way, the semantics that we usually associate with our chromatic concepts. To identify those other phenomenal beliefs, we imagine, now, that Moore is asking us what we know about the quality that we have identified as “red”. According to the “phenomenal” aspect of what we see, we believe that the reference of the chromatic concepts such as “red”…

  • 11 The default position is surely more demanding: colors are visible, necessarily: «Colours are visib (...)

(8) … is visible11, since it is defined precisely by its phenomenal aspect, that is, by how it appears to subjects of our type.

(9) … is located externally, that is, it is ostensively defined by pointing out a certain area of our visual field.

(10) … is exclusive, that is, a “red” spot cannot be at the same time, and in the same area, as another color.

(11) … is universal, that is, several things can be “red” at the same time.

(12) … is vague, that is to say, the assembly of phenomenal aspects that are described as “red”, as well as the extension of “red” objects, do not have a precise extension; and, finally.

(13) … is included in a complex phenomenal or qualitative structure where, for example, it is derived that, necessarily:

– “red” is more similar to “orange” than to “blue”;

  • 12 If we accept the results of Crane and Piantanida’s experiment (1983), it is most reasonable to thi (...)

– “red” and “green” are opposite12;

– “violet” is binary, since it is composed of as much “red” as it is “blue” etc.

  • 13 Matthen also mentions other types of “structural” properties such as categoricity or affective cap (...)

– etc13.

42I have now presented the most basic and accepted chromatic beliefs, with barely any controversy. Nevertheless, among these “common” beliefs and the most sophisticated beliefs of elaborate philosophical theories another series of beliefs can also be identified, though they are commonsensical they can be qualified as quasi-philosophical.

2.3. Quasi-philosophical beliefs

43We have now established the default position as the “prephilosophical” or the “commonsensical” position. Additionally, there are some beliefs with certain philosophical spirit that, in my opinion, should also be considered as components of the default position. I believe that color phenomenology, aside from determining the characteristics of the phenomenal aspect and its identification criterion, also indicates the quasi-philosophical or ontological characteristics of colors.

44For example, according to the default position, any color attributed to an object has certain ontological characteristics that make it possible for “red” to be…

(14) … simple (vs complex), that is to say, the qualitative nature of colors is captured through perception without need of subsequent analysis of another type (scientific, logical, ethnographic…).

(15) … categorical (vs dispositional), that is to say, when we perceive a color we see that color as a categorical property of the object.

  • 14 As nothing has been indicated about the intuitions of the default position on animal chromatic vis (...)

(16) … monadic (vs relational), that is to say, the usual semantics of chromatic concepts is attributive and attributes color to objects, not to the relations between objects and subjects. What’s more, color attributions are considered, in principle, as universally valid. In spite of the fact that we think that there are people, or other types of subjects, that cannot perceive some colors that we attribute to objects, we cannot infer, from that belief, that our attributions of color are not valid14.

45Therefore, from (16) we also know that,

(17) … colors are properties of objects. That is to say, the true value of chromatic attribution to objects does not vary according to the chromatic capacities of the subjects or the type of light used in each case. And, that

  • 15 Except mirrors and perfect voids (Averill 1985).

(18) … each object has, at least, one true color15. That is to say, there is at least one attributive statement of color that is true of colored objects. Therefore, there can be attributive statements that are false but that not all attributive statements are false. Moreover, we normally think that attributive statements are true if they comply with the following causal condition:

(19) … we see that A is red because A is red.

46In this sense, according to my interpretation, causal relation is a substantial part of the default position. On the other hand, to accept this causal statement (19) implies that we also accept the following four beliefs about colors:

(20) Colors have causal power. That is to say, the color of an object is causally responsible for our chromatic experiences of said object.

(21) One color produces, ceteris paribus, one chromatic experience. A same object, in the same circumstances, tends to produce the same chromatic experiences. In this sense, based on the variety of chromatic experiences, it is inferred that there are several causes and, therefore, various colors.

(22) Each color causes our chromatic experiences in a contingent way. That is to say, we can be wrong about all our attributions of color given that the true color of each object could have been another and our chromatic experiences could have been others.

(23) The reference of the chromatic concept coincides on both sides of the causal statement. That is to say, “red” does not have two different references when we say “I see that x is red because x is red”.

47These 23 epistemological, phenomenal and ontological beliefs exhaust, in my understaning, what we recognize as our common sense view about colors, the default position.

3. Properties of the default position

48The default position covers what we generally know, or believe to know, about colors. However, as I said, its value is mostly practical. It can be utilized to contrast the relative reliability of any philosophical endeavor towards the nature of colors. Next, I would like to introduce four properties that, according to my understanding, the beliefs associated with the default position hold and explain how they should be used.

3.1. Intrinsic Value

  • 16 Watkins 2002: cap. I.

49The fundamental characteristic of the ideas contained in the default position is that they have an intrinsic value that the philosophically elaborated ideas lack. As opposed to the beliefs contained in philosophical theories, the beliefs that constitute the default position do not have to defend themselves unless they are explicitly attacked by an argument. Therefore, if any theory intends to surpass the default position, as is the case with subjectivism or physicalism, the weight of the test will fall on the theory16.

50This intrinsic value is the distinctive feature of the default position. That is why it also functions as the point of departure for any philosophical proposal about the true nature of colors as well as the cornerstone with which to compare our ontological choices (see the appendix).

51I am claiming that the default position functions as the starting point as much as the cornerstone with which they should be contrasted, and constantly compares the different metaphysics alternatives, whatever those may be. This means that the preferable ontological theory will be the one that, being equivalent in the other aspects, has the least number of renunciations (also keeping in mind the specific weight of each one) with respect to the default position.

3.2. Non-essential

  • 17 Johnston 1992; Boghossian and Velleman 1991: 85.
  • 18 Watkins 2002, cap. I; Spackman 2001: 265s.

52The default position is composed of an assembly of commonly accepted ideas about colors. Some add to the default position a listing of restrictions17 or unacceptable rejections18.

53In my understanding, however, it is preferable that there are not any ideas about colors whose renunciation, by itself or in principle, is taken as unacceptable. It is not sensible to rule out a theory simply because it has renounced one of the intuitions that we have regarding colors, no matter how consolidated that intuition seems to be. This is another way of saying that changing the subject is not as easy as to abandon one of the 23 beliefs.

54Throughout the process that has to be carried out to truely know what colors are, probably some renunciations must be made and I believe that it is a bad strategy to say beforehand which are the untouchable ideas.

55I propose, therefore, to start as if there were no untouchable beliefs about colors. Not even the most intuitive ideas – such as, that colors are necessarily visible or that they are causally responsible for our chromatic experiences… – should be considered as untouchable at any price; at least, not at the starting point. In that sense, I consider it more practical, methodologically, to leave the possibility open to renounce any belief no matter how settled it may seem.

56Changing the subject should be identified, not by the renunciation of single ideas, but by the overall evaluation of the proposal. Likewise, bad philosophical accounts should be identified comparing the quantity and quality of renunciations of each account.

3.3. Flawed

57Another metaphysically relevant consideration that can be extracted from the analysis of the 23 beliefs is that the default position is not internally consistent. Therefore, it is not a philosophically satisfactory position; it cannot be automatically transformed into our best ontological account about colors.

58There is an apparent paradox here. If the default position also constitutes the goal, what sense is there in making the effort to devise a theory whose aim is to be as similar as it can be to the default position? If the default position shows the starting point as well as the ideal which we want to reach, the highest degree possible, how can we explain not remaining in the default position?

  • 19 At least, the following chromatic facts: nonexistent intentional objects, color constancy, the dep (...)
  • 20 At least, the following mental experiments: the inverted spectrum (in its different versions), the (...)

59There are basically two reasons that push us to continue looking for a better color ontology. The first reason is that the system of 23 beliefs should be expanded to integrate known chromatic facts19 as well as the most intuitive results of the mental experiments related to colors20.

60The second reason, which is more important, is that the default position, in its entirety and in its expanded version –with facts and chromatic intuitions-, is not a consistent system of beliefs. There are ideas that cannot be accepted jointly.

61Perhaps the most important collision and that of most philosophical penetration is between belief (6), that affirms the sufficiency of ordinary perception to know if two objects share the same color, and belief (22) according to which, given the contingency of chromatic perception, our ordinary chromatic attributions can be false and, therefore, ordinary perception would not be sufficient to know if two objects share the same color.

62In those conflicting cases, which of the incompatible beliefs should be rejected? I think we should abandon those that have the least specific weight. But here the problem arises of how to measure the specific weight of each belief.

3.4. Relative Weight

63How do we measure the relative weight of the beliefs contained in the default position? There are several commonsensical ways of measuring them.

64(A) More specific weight is alloted to the beliefs that contain more dependent beliefs. For example, the belief about the localization of colors in objects (16) has comparatively more specific weight because all our ordinary attributive speech depends on it. The causality belief (19) also has a heavy specific weight because, at least another four beliefs (20)-(23), depend on it.

65(B) Beliefs that are not “attractive” have less specific weight. For example, the belief that colors are “vague” is not that attractive and, perhaps this is why it is considered somewhat positive to devise a philosophical theory that endows precision to chromatic attributions. In that sense, to be “vague” is a characteristic of colors that has little specific weight.

66(C) The beliefs that remain relativized if we accept the fact that animal cromatic vision diverges from us also have less specific weight, because in that case what common sense dictates can differ. In fact, we can accept that some of the 23 beliefs that constitute the default position are not as commonly “acceptable” as they seem and that they should be relativized to our chromatic vision. For example, the principle of the true color can be understood in a strong manner (inter-species) or a weaker manner (anthropological). The same thing also occurs with beliefs (16) and (21). These intuitions are not as strong any more and, therefore, it is not easy to determine whether or not those beliefs should be integrated in the default position. Therefore, they are beliefs that, according to the definition of chromatic vision that we adopt, acquire more or less specific weight and thus it becomes easier to renounce some of them.

4. Conclusion

67Throughout the article I have offered the list of beliefs contained in the default position and some of their properties.

68I have argued that the default position constitutes the starting point and the cornerstone of contrast for any philosophical theory that intends to show what colors are. Therefore, if the methodology that I propose is accepted, anyone that intends to advance in the investigation on the constitutional nature of colors should commit herself to respect the majority of the 23 beliefs that have been presented here, given the relative weight of each of them.

69The importance of the default position resides in the explicit commitment with the ideas that it contains. That commitment converts the default position into the legislator of the theory of colors, since it offers the measure with which the best philosophical theory about the constitutional nature of colors is judged. The best theory will be the one that, being equivalent in all other aspects, has made the least number of renunciations (keeping in mind the specific weight of each one) with respect to the default position. For those that are still interested in finding the constitutional nature of colors, the best metaphysics theory will be the one that has the most support (and the best quality) along with the least counterweight in this network of beliefs known as the default position.


70This is an illustration of how to use the default position. In this case I compare various versions of physicalism. The outcome is that we should prefer the representationalist type of physicalism because it clearly makes less drop-offs than the other three alternatives.

Default beliefs













is necessary to

identify the color of a given object


judge whether 2 objects share the same color


is sufficient to

identify the color of a given object


judge whether 2 objects share the same color


We know which are the normal circumstances to make color attributions


Our color concepts share same denotation


Colors are





mutually exclusive






The color space represents their structural properties


Color properties (are)

simple (vs complex)


categorical (vs dispositional)


monadic (vs relacional)




allow at least one true attribution


responsable of our color experiences


causally efficient






refer to the same property








Torna su


Armstrong, D.M.

– 1987, Smart and the Secondary Qualities, in A. Byrne and D. Hilbert (eds.), Readings on Color, Cambridge (Ma.), MIT Press, 1997: 33-46

Arstilla, V.

– 2003, True Colours, False Theories, “Australasian Journal of Philosophy”, 81: 41-50

Averill, E.W.

– 1985, Color and the Anthropocentric Problem, “Journal of Philosophy”, 82: 281-304

Boghossian, P.A. and Velleman, J.D.

– 1991, Physicalist Theories of Color, “Philosophical Review”, 100: 67-106

Clark, A.

– 1993, Sensory Qualities, Oxford, Oxford University Press

Crane, H.D. and Piantanida, T.P.

– 1983, On Seeing Reddish Green and Yellowish Blue, “Science”, 221: 1078-1080

Eccles, J.C.

– 1984, The Human Mistery, London, Routledge

Hardin, C.L.

– 1983, Colors, Normal Observers and Standard Conditions, “Journal of Philosophy”, 80: 806-813

– 2003, A Spectral Reflectance Doth Not a Color Make, “Journal of Philosophy”, 100: 191-202

Hilbert, D.

– 1987, Color and Color Perception, Stanford, CSLI

Johnston, M.

– 1992, How to Speak of the Colors, in A. Byrne and D. Hilbert (eds.), Readings on Color, Cambridge (Ma.), MIT Press, 1997: 137-176

Land, E.H.

– 1977, The Retinex Theory of Color Vision, “Scientific American”, 237: 108-128

Matthen, M.

– 1999, The Disunity of Color, “Philosophical Review”, 108: 47-84

Myin, E.

– 2001, Constrained Inversions of Sensations, “Philosophica”, 68: 31-41

Ross, P.

– 1999, The Appearance and Nature of Color, “Southern Journal of Philosophy”, 37: 227-252

– 2000, The Relativity of Color, “Synthese”, 123: 105-129

– 2001, The Location Problem for Color Subjetivism, “Consciousness and Cognition”, 10: 42-58

Smart, J.J.C.

– 1975, On Some Criticism of a Physicalist Theory of Colors, in A. Byrne and D. Hilbert (eds.), Reading on Color, Cambridge (Ma.), MIT Press, 1997: 1-11

Spackman, J.

– 2001, Color, Relativism, and Realism, “Philosophical Studies”, 108: 249-287

Strawson, G.

– 1979, Perception and its Objects, in G. MacDonald (ed.), Perception and Identity, London, MacMillan

– 1989, Red and “Red”, “Synthese”, 78: 193-232

Stroud, B.

– 2000, The Quest for Reality, Oxford, Oxford University Press

Thompson, E., Palacios, A. and Varela, F.J.

– 1992, Ways of Coloring, “Behavioural and Brain Sciences”, 15: 1-26

Tye, M.

– 2000, Consciousness, Color and Content, Cambridge (Ma.), MIT Press

Watkins, M.

– 2002, Rediscovering Colors, London, Kluwer

Zeki, S.

– 1983, Color Coding in the Cerebral Cortex: The reaction of Cells in Monkey Visual Cortex to Wavelengths and Colours, “Neuroscience”, 9: 741-765

Torna su


1 Color perception is usually pictured as a set of biological mechanisms whose function is to detect – under different types of light – variations on the spectral reflectance of the surfaces of objects.

2 Even nowadays, this methodological view continues having many followers. Clyde Hardin, for example, obtained the prestigious “Johnsonian Prize” of philosophy (1986) for justifying his metaphysics proposal in the revised scientific knowledge on chromatic vision. Austen Clark recommends to the philosopher of colors to put himself in the place of the scientists (1993: ix). On the other hand, scientists often do more philosophy than one would expect, as is the case, among others, of the Nobel Prize winner of physiology, John Eccles (1984), or of the neurophysiologist Zeki (1983), all actual defenders of the subjectivist theory which is in line with the methodological criterion just mentioned. The austere scientific image of physical reality as well as certain scientific facts (for example, the metamer phenomenon, the structural properties contained in the chromatic space or the non-representational colors [Arstilla 2003]) make subjectivism appear as the best philosophical theory about the true nature of colors. All in all, it can be said that even nowadays the «scientifically enlightened common sense» (Stroud 2000: 9) remains and that the secondary quality view seems to be its natural consequence.

3 Locke, J. Essay, book II, ix, § 8.

4 See Land 1977.

5 Hilbert 1987.

6 Thompson et. al. 1992; Matthen 1999.

7 Averill 1985: 301.

8 Hardin 2003: 202.

9 Ross 2001.

10 Myin 2001.

11 The default position is surely more demanding: colors are visible, necessarily: «Colours are visibilia or they are nothing» (Strawson 1979: 56).

12 If we accept the results of Crane and Piantanida’s experiment (1983), it is most reasonable to think that the “necessity” that we normally attribute to these chromatic oppositions is empirical (and as such, they are not semantic, logical, or grammatical).

13 Matthen also mentions other types of “structural” properties such as categoricity or affective capacity of colors (Matthen 1999: 66s). However, given that the description of these properties is quite controversial, and as it’s not clear that “categoricity” and “affective capacity” are structural and necessary properties of colors, in future I will leave them out.

14 As nothing has been indicated about the intuitions of the default position on animal chromatic vision we will push this question aside, although we will have to analyze the question of universal validity with respect to the related chromatic attributions from the chromatic visions of other animals.

15 Except mirrors and perfect voids (Averill 1985).

16 Watkins 2002: cap. I.

17 Johnston 1992; Boghossian and Velleman 1991: 85.

18 Watkins 2002, cap. I; Spackman 2001: 265s.

19 At least, the following chromatic facts: nonexistent intentional objects, color constancy, the dependence of perceived color on the circumstances of perception (distance, medium, preceptor, environment, type of light, etc.), matamer and mondrian pairs, colors dependent on the environment, colors without representational content, colors of other animal species (and their respective chromatic spaces), the asymmetry of chromatic space, the dependence of the chromatic space on the perceptive system, the chromatic variability among humans with normal vision, the diverse causes of color (reflection, refraction, dispersion, … of the light), the diverse functions of chromatic vision, etc.

20 At least, the following mental experiments: the inverted spectrum (in its different versions), the superscientific Mary (better known as “the knowledge argument”), Fred (who can perceive a new color), Nagel’s bat (whose chromatic perception is inaccessible to us), Eliza (closed in a room with colors that fluctuate spontaneously), the transparent world (where objects only have color if they get within a certain distance of the eyes), colors that act differently depending on the context in which they are found (the color green turns to orange under a yellow light, etc.), etc.

Torna su

Per citare questo articolo

Notizia bibliografica

Ekai Txapartegi, «Default beliefs on colors: the methodological value of what we believe to know about colors»Rivista di estetica, 43 | 2010, 213-229.

Notizia bibliografica digitale

Ekai Txapartegi, «Default beliefs on colors: the methodological value of what we believe to know about colors»Rivista di estetica [Online], 43 | 2010, online dal 30 novembre 2015, consultato il 15 juin 2024. URL:; DOI:

Torna su

Diritti d’autore


Solamente il testo è utilizzabile con licenza CC BY-NC-ND 4.0. Salvo diversa indicazione, per tutti agli altri elementi (illustrazioni, allegati importati) la copia non è autorizzata ("Tutti i diritti riservati").

Torna su
Cerca su OpenEdition Search

Sarai reindirizzato su OpenEdition Search