Colour Opponency
On this page
Direct answer
A green afterimage floating on a white wall after half a minute spent staring at a red patch is evidence that no theory of three independent cones can explain — but opponency can. Colour is coded in two stages: three cone types at the receptor level (long-wave cones peaking near 560 nm, medium-wave near 530 nm, short-wave near 420 nm — Young-Helmholtz trichromacy), then opponent channels downstream (Hering): a red-green channel computing long against medium cone output, a blue-yellow channel computing short-wave against the sum of long and medium, and a separate luminance (black-white) channel. Red-green opponent responses appear in retinal ganglion cells and the parvocellular lateral geniculate body; the cortex adds double-opponent cells that encode colour contrast at edges, the substrate of colour constancy. Because each channel is a difference, a reddish-green or a bluish-yellow is physiologically impossible — the two halves of one channel cannot fire together.
What you must remember
- Two-stage theory: trichromatic receptors, opponent post-processing — viva answers must state both, since each alone explains only half the phenomena.
- Channel arithmetic: red-green = L minus M; blue-yellow = S minus (L + M); luminance = L + M weighted — yellow is a computed colour, not a receptor output.
- Afterimages: prolonged red adapts L cones, so a white field (equal output) reads as the opponent, green — adaptation, not imagination.
- Colour constancy: the cortex compares surfaces across an illuminated scene (Land's retinex logic), keeping a tomato red under morning or fluorescent light; colour is a ratio judgement, not a wavelength readout.
- Inheritance patterns: red-green defects are X-linked recessive — about 8% of males, under 1% of females, deuteranomaly the single commonest type; blue-yellow (tritan) defects are rare and typically autosomal dominant, while achromatopsia (no cone function) is autosomal recessive with photophobia and nystagmus.
- Classification: anomalous trichromats have three cones with a shifted pigment; dichromats (protanopes, deuteranopes, tritanopes) lack one cone class entirely.
- Testing conventions: Ishihara pseudoisochromatic plates screen red-green defects; Farnsworth D-15 or 100-hue grades them — and both are standard in Indian driver-licensing and railway recruitment medicals, an occupational hook examiners like.
A screening encounter, worked through
A 19-year-old applying for a railway post fails the Ishihara plates at his medical. First, separate congenital from acquired: lifelong difficulty selecting ripe tomatoes and a brother similarly affected point to X-linked deuteranomaly, whereas a new defect in a diabetic or a patient on digoxin or ethambutol raises acquired blue-yellow or red-green loss — ethambutol is a classic optic nerve culprit at 15 mg/kg-day dosing regimens. Next, test each eye separately and under daylight illumination, since dim lighting and monocular disease both distort plates. If he is deuteranomalous, he is an anomalous trichromat: three cone pigments, the medium-wave one shifted, so he confuses some reds with greens but retains luminance and most discrimination — he is not seeing grey, a point candidates lose marks on. Finally, grade with Farnsworth D-15 for the fitness decision: signal colour recognition depends on brightness and position cues that most anomalous trichromats use well, which is why policy distinguishes mild anomalous trichromacy from dichromacy rather than lumping all "colour blindness" together.
Where students slip
Saying "colour-blind people cannot see colour" is the first slip — a dichromat lives in a two-dimensional colour world with rich yellows and blues, confusing only the axes his missing pigment collapses. The second is mechanism mixing: trichromacy lives in the cones, opponency begins in retinal horizontal and ganglion cells, and crediting "opponent cells in the cones" to Hering scores nothing. Third, students forget why blue-yellow defects are not X-linked: the short-wave cone gene sits on chromosome 7, not the X-linked L/M array produced by recent gene duplication — a genetics-physiology bridge that separates strong answers. The protan note worth adding: protanopes also lose luminance for long wavelengths, so deep red traffic signals look dark — practically relevant and frequently asked.
Frequently asked questions
What are the three opponent channels of colour vision?
Red-green (long minus medium cone signal), blue-yellow (short-wave cone against the sum of long and medium) and a black-white luminance channel carrying summed cone output.
Why do afterimages appear in complementary colours?
Prolonged viewing adapts one half of an opponent channel, so a neutral white surface, by producing equal cone output, tips the channel towards the unadapted opponent colour.
Why is a reddish-green sensation impossible?
Red and green are the two poles of one difference channel, which cannot signal both directions simultaneously — a structural limit, not attention.
Which colour vision defect is most common and how is it inherited?
Deuteranomaly, an X-linked recessive anomalous trichromacy affecting roughly 8% of males and under 1% of females.
Which drugs cause acquired colour vision loss?
Ethambutol classically (red-green with optic neuropathy), digoxin (yellow-green xanthopsia) and viagra-type phosphodiesterase-5 inhibitors (transient blue tinge) — acquired defects that must precede Ishihara interpretation.