For discussion on emergence, reality, objectivity, etc:
64 thoughts on “Purple Doesn’t Exist?”
Leave a Reply
You must be logged in to post a comment.
The Skeptical Zone
"I beseech you, in the bowels of Christ, think it possible that you may be mistaken."
Colors do not exist. Not sure how you can even talk about them in good faith. What’s next, talking horses and gay pride parades touting the reality of rainbows?
Interesting! That tells us that our visual system can be tricked — that not everything that shows up in visual consciousness can be mapped neatly onto physical properties.
But does that mean that purple doesn’t exist? Only if we’ve already decided that physics is the ultimate arbiter of what counts as real or as existing. And why in deuces should we decide that?
Who, or what, is “tricking” our visual system?
“…that not everything that shows up in visual consciousness can be mapped neatly onto physical properties. ”
Now that would be a neat trick! Not sure why you find that surprising though.
I don’t find it surprising. I just find it very interesting.
Because that seems to be the default position here at TSZ. And I don’t think the question is why we should decide that, but rather how we should decide that. If the reason we see magenta isn’t physical, what is it?
Surely the reason is physical and can therefore be reduced to physics.
Off to a great start though!
Is magenta aka purple real?
How can science answer this question?
If you have never experienced either magenta or purple you might have reason to doubt whether they are real.
Anyone here who has never experienced either magenta or purple?
For the rest of you, what will you do if “consensus science” declares that neither magenta or purple is real? Will you cease experiencing those colors?
It isn’t a science question. It’s either a philosophy question or a person opinion question. And I lean toward it being a personal opinion question.
Is it purple, or is it gold?
You get violet in a rainbow because “violet light” is twice the frequency of red. Being twice the frequency, red cones start firing again, but the blue ones continue to do so since it’s still close to blue, hence you end up with “purple,” or “violet.”
That could well be the evolutionary cause for the “purple” color that we see, because it happens to be meaningful that both blue and red are firing and not much else (it is rather ambiguous, though, since how are we to know if it’s “violet light” doing it, or mixes of blue and red?). We aren’t especially sensitive to violet, however, which is one reason that the sky is blue, not violet (I think that there’s more blue light as well, though, because I think that a greater range of frequencies gives us “blue”).
Glen Davidson
Let me see if I have this straight. There are wavelengths of light that do not correspond to red, green, or blue. Yet only red, green and blue cones evolved.
Not quite. All the cones respond to the whole spectrum. Just differentially.
No, birds have tetrachromacy, and most mammals have cones most sensitive in only two regions, green and blue. The mantis shrimp differentiates between 16 [I changed this from 18 to what many sources say, 16, but source in my next post says 12] different frequency regions. Mammals are thought to have lost two color cone types, due to being largely nocturnal early on. Old world monkeys, apes, and humans have re-evolved “red,” and new world monkeys are a mixed bag, most with just two cone “colors” but some with three.
And I don’t know that one can say that there are wavelengths of light that don’t correspond to red, green, or blue, because something like “yellow” simply corresponds to red and green. It’s not either-or.
Glen Davidson
Primates are unusual in being trichromatic mammals, but many vertebrates and some invertebrates can see parts of the spectrum that we can’t. And cephalopods can see polarized lights. Many animals can detect electrical fields, though the only mammal that can do so is the platypus.
The diversity of sensory and cognitive abilities of animal life is astonishing, and if anything it should call into doubt what we mean by “objective”, ” real”, etc. In some respects I think that our vocabulary for doing metaphysics is still too anthropocentric — it has not really caught up with biology.
I took the trouble to look up mantis shrimp, and got an article that deflates the visual prowess of them, here. Twelve different photoreceptors is what it says for them, but their frequency discrimination isn’t all that fantastic anyhow. I always did sort of suspect that they had so many different receptors because they don’t have the brains to make use of differential signals like we do.
Glen Davidson
I sure hope you all aren’t trying to convince me that what really exists is observer dependent! What would science look like if that was the case?
Does the color purple or magenta really exist?
If they exist, is their existence an objective fact?
If they exist, is their existence a scientific fact?
Kantian Naturalist,
Many animals living in water can detect electric fields, hence the well-known mammal species that can lives (and importantly, hunts) in water. Air is too poor a conductor to make electroreception worthwhile there–I think.
However, the echidna likely evolved from the platypus not so long ago, and apparently has some electroreceptive capability. I don’t suppose that it uses it in air, though, probably more likely in moist ground or some such thing.
Glen Davidson
Color constancy, possibly a more impressive feat of vision. See the red and green pictures, especially, where the “blue” square in the red picture is the same color as the “red” square in the green picture.
In other words, it really isn’t just a matter of the number of signals coming from each cone that determines color, it all shifts according to context, too. This keeps colors from shifting to the red end of the spectrum as the light reddens toward evening.
Glen Davidson
I have seen many of these types of things and love them. they make a good creationist point.
We don’t see the natural world as it really is.
We only see a edited replay. We are not looking out the windopw, our eyes, but instead in the basement watching a video repeat.
Our sight being not so great it invents things not actually there like some colours.
It is our soul watching a memory of whjat our eyes recorded. So we are just watching our memory. This is also why we can see things not there as when drugged or mental problems. its just the memory that is interfered with and not our eyes.
We are immaterial souls meshed to a material memory machine called the brain/mind.
Mung,
Each receptor has a spectrum of firing across the continuum of wavelengths, most strongly centred on R, G or B and tailing off to nothing either side, but not sharply. These spectra overlap. I think we perceive approx 7 rainbow colours because of the combination of 3 peaks and 4 ‘shoulder regions’ across the visible range. We detect a transition in perception as the wavelength moves from one semi-discrete zone to another, because it’s harder to resolve differences in intensity in some regions of this ‘hilly’ spectrum than others. The wavelength changes continuously, our perception in jumps.
There is also only a limited wavelength range in which proteins can be conformationally triggered by light, hence the relative narrowness of the ‘visible spectrum’. It would be quite useful to see infra-red, so I’d suspect that its absence is a mechanistic constraint.
Magenta is my favorite example.
Thanks for posting.
Yellow is interesting too, of course, because as far as the brain is concerned, it doesn’t matter whether the red and green receptors are stimulated by a pure yellow frequency or by a pair of pure red and green sources.
And of course there are a tetrachromats….
Mung:
Wavelengths really exist. In that limited sense, color really exists. Buut the perception of color, the experience of color, is entirely an artifact of eye and brain behavior. Colors can be induced by playing with flickering and movement of monochromatic light.
No one really knows how this works, but I worked with a teacher in college who tried to research it. I’ve seen intense reds and blues and greens induced by a monochromatic sodium vapor light. The color sensation are induced by movement an the rate of alternation of light and dark. We were hacking the retina’s digital to analog color encoder.
If no one really knows how colours work, then by what definition was the light in the experiment monochromatic?
Wavelength. It’s monochromatic physically – sodium vapour produces electromagnetic radiation at a single wavelength in the visible range (actually two, but very close together), and it’s in the part of the visible spectrum we normally call “yellow”.
But what is true is that even with a total yellow filter, you can still “perceive” other colours – in other words, the visual system makes inferences about what “colour” objects are from more than the simple intensity and wavelength of light falling on the retina.
Contrast plays a huge role, but so do expectations and associations and the application of generalising rules about what lighting conditions produce what shifts in hue as well as intensity in normal conditions.
It is an objective fact that certain objects are reliably described by independent trichromat observers as being “magenta” or “purple” and that such objects have in common the property of stimulating red and blue receptors in the retina simultaneously.
It is also an objective fact that certain objects are reliably described by independent trichromat observers as being yellow, and that such objects have in comon the property of stimulating the red and green receptors in the retina simultaneously.
However, in the latter case, such objects may do so either by reflecting/emitting red and green light simultaneously, or by reflecting/emitting yellow light only, or, most commonly, a spectrum of frequencies between red and yellow. We can’t tell by looking – we’d have to actually put the light through a prism to find out which.
Is color:
– a property of light (any wavelength)
– a property of how objects reflect light
– a property of the light reflectance of the object in context as represented by a “normal” observer
– a property of the experience of that normal observer (the dreaded qualia)
– a property of the relation between a normal observer, the reflectance conditions, and the object
– a behavioral disposition of a normal observer
– something else?
Many a philosophy thesis and paper in conceptual analysis of color.
And there is also the separate scientific question of explaining vision processing in animals, of course.
Yet the experiment consisted in “movement an the rate of alternation of light and dark.” It’s common experience that play of shadows and any movement in light is change of colours, so it’s not properly monochromatic.
And movement. Movement is, in one of its aspects, change of how light is reflected, and change of light is change of colour.
“Chroma” (also called “hue”) refers to the wavelength, not the intensity. So a scene lit solely by sodium light will reach your retina as different intensities of yellow-wavelength light. Where it is low intensity or absent, it will look dark; where it is high intensity, it will look bright. After a while, it won’t look “yellow” at all – it will look like shades of grey, but what will be entering your eye will still only be light in a narrow band of yellow (defined as wavelength).
So “monochromatic” means light of only one wavelength (or a very narrow band of wavelengths). Strictly speaking, greyscale is not “monochrome” – it’s just that the spectral composition of the light coming from all parts of the scene is the same, only the intensity differing.
Yes, that probably plays a role too, good point. The interesting thing about colour perception is that there is no one-to-one mapping between our experience of colour, and the wavelengths of light in the visible spectrum.
When our eyes receive red from the red end of the rainbow. we will tend to see the object it emanates from as “red” – but if the object is emitting very little red, and the rest of the scene is emitting bright light, we may see it as “dark brown” or even “black”. Ditto with the blue end.
And it’s why you can project a slide onto a white wall, even in a quite well-lit room, and still perceive large areas of the slide as “black”! But the same applies to colours as well – a yellow spot on a dark grey (i.e. “black”) background will look yellow: turn up the intensity of the grey up to bright white, and your yellow spot will look dark brown or black.
And it doesn’t matter whether the yellow spot is made red and green beams (which it usually is, in a slide projector) or a yellow one. The effect will be the same.
And of course magenta can only be produced by a combination of red and blue beams – and it will only look magenta on a certain background. Turn up the background again and it will look deep purple or black.
These colours aren’t “illusions” – they are what we see. It’s just that our labels do not map on to the electromagnetic spectrum in the way that you might expect.
It depends on what the meaning of “is” is.
People disagree over what they mean by “exists”.
Purple is a property. There are long standing arguments between realism, platonism and nominalism for properties. And I actually don’t think it matters.
However, my own personal opinion is that colors exist as the reflectivity of objects. However, this doesn’t account all of the ways that we use color words.
BruceS,
I don’t work in philosophy of color, so I don’t know if this is a crazy or stupid position, but here goes: I think of colors as complex relational properties that we experience as intrinsic properties*. Colors involve at least the following: photon frequency, surface reflectance (when relevant), cones in the retina (when relevant), dynamics of the visual processing system. And, f you think that the kinds of categories you have influence what you perceive, then that too — we know that different cultures vary by how many color-categories they have.
It is true colors are intrinsic properties of the objects described by ordinary experience (in “the life-world”, as Husserl puts it — the world as described from the perspective from those who live in it). It is also true that colors aren’t intrinsic properties of the objects described by fundamental physics.
These two facts still allow for the thought that we can locate colors in the scientific image if we conceive of colors as complex relational properties that require certain kinds of complex biological and neurological systems in order for those properties to be fully instantiated**.
* = while being agnostic about the metaphysics of properties.
** = in contrast to Sellars, who thought of colors in the scientific image as intrinsic properties of the animal’s sensory consciousness. I accept that Sellars was right in thinking that the ontology of the sciences will give us reason to revise our life-worldly categories. But I think that he was wrong in the case of color, and I think that what he gets wrong about color is a symptom of a more pervasive and systematic error in his thinking about how the ontological priority of science has authority over the life-world. The ultimate task of science is not to eliminate the life-world (as Alex Rosenberg thinks) but to explain it.
Thanks for posting that, Richard. Interesting stuff.
Great accent, too!
ETA: Hey, I wonder to what extent accents are analogous to colors….
Neil Rickert,
Hmmm
I still have to make my way through the details of all the philosophical positions that are there. Of course, philosophers keep adding new ones, as in this Nautilus interview: Color is a dance between your brain and the world . (She has a book on this as well.)
I have learned that the order of words is very important. Maybe that should not be a surprise when it comes to philosophy . So “represented property”: that is safe for physicalists. “Property of representation”: watch out, that way be the dragons of non-physical qualia. I’m not sure if “intrinsic properties” is open to this; a good physicalist would want a causal/functional network to cover “intrinsic”. Although I am not sure if you care about that in any event
In what I’ve read so far, I have not seen anyone match the philosophical chutzpah of Jesse Prinz, who relates specific colors to specific patterns of firing over time for neural subpopulations (“gamma vecctorwaves”). I have not read Churchland in detail on this, but I don’t think even he gets that specific.
Nautilus has a whole issue on color.
KN,
Turns out that some humans can, also:
Haidinger’s brush
KN:
Scientists are well aware of the subjectivity of color perception, and are usually careful to specify wavelengths or spectra when describing illumination or reflected light.
BruceS,
Thank for that linked interview, BruceS! I found Chirimuuta’s views quite compelling. Methodologically speaking, I couldn’t agree more with this:
I would say that we need to understand how our life-worldly encounters with others and with objects (which is to say, not just subjective in the narrow sense but also the whole subjective-intersubjective-objective nexus of agential experience, including both inference-grounding discursive practices and reference-grounding sensorimotor skills) can be explained in terms of both the causal transactions between subagential cognitive mechanisms and other dynamical systems comprising bodies and environments and also the causal mechanisms that underpin the evolutionary and cultural history that led from apes to the Enlightenment.
Neil,
That can’t be right. A laser beam reflected from a white surface looks red.
There’s a paper by a Kripkean on color that I read a few years back and remember liking quite a bit. I’ll see if I can find a link, but it’d be good if I could remember the author’s name…..
OK, I found it. It’s “Kripke on Color Words and the Primary/Secondary Quality Distinction” by Mario Gómez-Torrente, and it’s in Saul Kripke (Alan Berger, ed.), which I think I found free on the net somewhere at one time or another.
She blogged at Brains if you want more details. M-P gets a cite!
(Strictly speaking, the Brain blog posts are about colour, not color as in Nautilus.)
I think it’s fun that she chose to revive the name “adverbialism”, a theory on its death bed as I read the description in SEP.
I’ve often wondered what my cat sees when she watches TV. Cats are dichromats, but TV screens are designed for trichromats, so it would seem to follow that the colors should be off kilter when a cat looks at something on TV versus the same thing in real life.
I think her views could be used as an example of how philosophical agent-level analysis differs from scientific analysis of the sub-agent processes.
As a neuroscientist and philosopher, she understands both approaches.
She chose to reject a standard philosophical position, representationALism, which explains color at the agent level as a property of objects as we represent them. Instead, she talks of color “as a property of interactions that perceivers have with objects”.
Now a representationalist (in the philosopher sense) would agree that our perceptual mechanisms are involved in how the object’s property is represented. But that would be part of the sub-agent causal description of the property, not a part of the agent-level description of how to think about the concept of color.
Paper is there with a quick search for title. But at 36 pages, a bit more than I want to attempt right now.
Didn’t see the book on quick internet search or on binsearch.info.
BruceS,
Prolly found it on scribd. They’re getting a bit quicker to take stuff down there, though.
Yes, her approach is very exciting. I see it as putting to real work the general methodology I get from Michael Wheeler and Owen Flanagan, about how to put agential and subagential explanations into productive interplay.
I think that standard phenomenology would be direct realism about colors: we ordinarily take colors to be intrinsic properties of objects (for normal observers under standard conditions). It would take a more sophisticated phenomenologist to describe all the ways in which direct realism fails to be true even of our experience — e.g. color constancy, the Dress That Broke the Internet, etc.
Right! Though here too there are going to be subtle debates between enactivists and cognitivists as to whether subagential mechanisms involve representations, and even more subtle debates between cognitivists and other embodied-embedded folks about how to theorize representations. At the level of debate between Fodor, Churchland, Clark, Wheeler, Varela, Chemero, we are hashing out a lot of different ways of understanding what subagential mechanisms might be.
Translating Chirimuuta’s views into Sellarsian terms, color is an intrinsic property of objects within the manifest image but a relational property of objects and biological cognitive systems in the scientific image. In turn, understanding how color is a relational property of objects and biocognitive systems allows us to understand puzzles that the manifest image concept of concept cannot resolve, such as the Dress That Broke the Internet or how we can perceive magenta. Personally, I’m deeply gratified to hear her say that, because I’ve long thought that it is true but never worked out any of the details.
Bruce,
Yes, and it’s important to remember that the representations need not be veridical — they can be vericidal. 🙂
The lines in the Müller-Lyer illusion are being represented; it’s just that their relative length isn’t being represented correctly.
I love “vericidal”. It’s a word we’ve needed for a long time.
Lizzie,
It’s a keeper, isn’t it?
Credit goes to Bruce for his fortuitous typo.
“Typo”, yah, right, that’s the ticket, just a typo.
The honest answer is that it is one of those words I read a lot but never sounded out and so I just didn’t think about how to spell it correctly. But from now it will be a typo ’cause I sure know how to spell it now.
“Not being represented correctly” is an interesting evaluation.
It one sense, of course, it is right to say that; we (as agents!) see the lines incorrectly as different lengths. But it another sense, we* see them correctly.
I mean in the sense that the representation process as a whole is correct, because it is driven by prior beliefs that this context, as presented by input signals from lower levels, means the lines are of different lengths. This is a correct process because it aligns with the regularities of our world. (At least our Western world with lots of square corners).
—————
* This may be equivocating on “we” since the next paragraph is not about the agent level.