Opponent coding explains why combinations like "reddish green" don't exist

What it is

People can effortlessly describe a "yellowish red" (orange) or a "bluish green" (cyan), yet cannot imagine, let alone name, anything that would count as a "reddish green" or a "yellowish blue" — combinations that simply don't exist perceptually are called forbidden colors. This is not a vocabulary or imagination limit; the perceptual system itself cannot produce this perceptual state.

The counterpart concept is unique hues: red, green, blue, and yellow are perceptually "pure" — they contain no trace of any other hue — and every other color perception can be described as a mixture of these four unique hues in different proportions. Forbidden colors are exactly the combination of "mixing a pair of mutually opposed unique hues at once."

Why it happens

Red and green sit at the two poles of one opponent channel; blue and yellow sit at the two poles of another. An opponent channel outputs a single difference signal, not two independently free-to-be-positive readings — at any moment the red-green channel's output leans toward the red pole, leans toward the green pole, or sits at the neutral point; there is no state where it is simultaneously "a positive red signal and a positive green signal," any more than a single needle can point at two opposite marks on a dial at once. The blue-yellow channel works the same way.

"Reddish green" would require the red-green channel to output "red" and "green" simultaneously, and there is simply no corresponding physical state in this single-difference-code architecture to express that — it is not that the visual system chooses not to produce this perception, it's that the coding scheme means this state doesn't exist anywhere in the space of possible signals. Unique hues are "pure" precisely because red and green (and blue and yellow) are the extreme values on the red-green and blue-yellow axes that carry no admixture of their opposing pole. Orange, by contrast, is naturally perceived because it corresponds to the red-green channel leaning red while the blue-yellow channel simultaneously leans yellow — two different opponent channels each contributing a share without conflict, which is nothing like "one channel outputting both poles at once."

Studying it

The classic method for locating unique hues is the hue-naming / unique hue setting procedure: participants adjust a continuously variable hue wheel until it looks like "pure green with absolutely no trace of red" (or another unique hue), and the specific coordinates of that hue in color space are recorded; repeating this across participants and sessions yields the location of each unique hue along with its individual variability.

Research on forbidden colors more often uses perceptual cancellation experiments: special visual stimulation techniques (such as flickering two opponent colors at an extremely high rate, so the visual system can't switch normally and is effectively forced to try to produce a "mixture" of the two) are used to observe what participants report — most describe intense color rivalry, flicker, or a spatial patchwork rather than a stable "reddish green" blended percept. Such experiments serve as further behavioral support for opponent-process theory.

Methodological note: the coordinate location of unique hues shows clear individual variation and cross-cultural naming differences; there is no single, universally agreed, exact physical wavelength definition, so results are typically reported as a distribution range rather than a single value.

Where it stops holding

• This describes a perceptual limit at the level of opponent channels, not a limit imposed by language or culture on how color categories are carved up. Different languages and cultures divide color categories differently (whether cyan gets its own name, for instance), but that is a higher-level linguistic and cognitive issue, distinct from this more basic perceptual limit on red-green and blue-yellow never mixing.
• The specific wavelength coordinates of unique hues vary between individuals, which does not mean everyone's judgment of "pure red" or "pure green" lands on exactly the same physical wavelength.
• A forbidden color is not "a color with no corresponding physical spectrum." It is entirely possible, physically, to construct a complex spectral stimulus that strongly activates both L and M cones with opposite polarity responses; the issue is not whether such a physical spectrum can be produced, but that the visual system's coding scheme has no way to interpret such a stimulus as "red and green coexisting."
• This describes the opponent-coding structure of normal trichromatic vision. It does not apply to people with color vision deficiencies, whose cone input itself is missing or altered, changing the opponent-channel computation accordingly.

Related

• Same group: A1.05.5 Past the cones, signals are re-coded into red-green, blue-yellow, and black-white opponent channels · A1.05.7 Color afterimages are caused by adaptive reversal in opponent channels · A1.05.8 Interface color discriminability should be designed along opponent channels, not isolated hue
• Nearby: A1.06 Color vision deficiency · A1.05.1 The summed response of three cone types produces color perception
• Search terms: forbidden colors · unique hue · opponent color · hue cancellation