Unique Hues
Take a moment and think of what you would consider to be green. It should have no hint of blue and no hint of yellow in it. Just pure green. What is curious is that the green that you are picturing is very likely different than the green that I would picture, and by a margin that is wider than any other color.
This is called a unique hue. Unique hues were described by Ewald Hering back in 1870. There are four: red, yellow, green and blue. Hering wrote that these four unique hues can appear without a trace of any other color. In a later book he described them as four outstanding places in the circle of hues: a primary yellow, a primary blue, and the red and the green that are neither bluish nor yellowish.
You will notice that orange is not one of them. In any orange we can see something yellowish and something reddish. Green is the surprising member of the four. Anyone who has mixed paint knows that most greens are made from yellow and blue. Even so, it is possible to find a green that looks like neither.
Hering proposed that any hue can be described by how red or green it is and how yellow or blue it is. Four unique hues are enough to place any other hue.
What I find truly fascinating is that people with normal color vision do not all agree on where those four hues are and by differing amounts of variation.
In a 2004 paper, Rolf Kuehni compared ten experiments in which nearly 600 people had been asked to find their unique hues. All of them had normal color vision. Six of the experiments used light of different wavelengths. If those six are taken as comparable, the widest range for pure green in the paper’s table runs from 487 to 557 nanometers. For pure blue it runs from 458 to 495, and for pure yellow from 544 to 594.1

The six experiments that used light, plotted from Table I of Kuehni (2004).
Those ranges overlap. The implication, in the paper’s words, is that “one observer’s unique blue is another observer’s unique green and vice versa.” In those pooled numbers, the same applies where green meets yellow.
These experiments capture which light or which chip a person identifies as pure. They do not tell us what anyone else sees in their mind, and they do not imply that we see wildly different colors.

Approximate ranges from color-chip experiments, read from Figure 5 of Kuehni (2004).
The second chart shows a different kind of experiment, in which people chose from color chips. Each white curve demonstrates how far apart the choices fell for one hue. The curve for green covers about a quarter of the wheel. The other three are shorter. For every hue except green, the chips produced narrower ranges than the lights did.
Red is missing from the wavelength numbers, and for a good reason. Pure red is not in the spectrum. Hering had noted it himself: his pure red does not correspond to the red of the spectrum, which he called yellowish.23 A 2015 review says the same in modern terms: pure red “does not correspond to a single wavelength.”4 When people are tested with displays or printed palettes instead, the review reports comparable variations in where they put pure red.
The obvious suspect is the hardware. Light is first absorbed by three kinds of cone receptors, sensitive to long, medium and short wavelengths. In one study, two people whose ratios of long to medium cones were measured at 1.15 and 3.79 picked almost the same pure yellow. Their two settings were 576.8 and 574.7 nanometers.5 Green is different. In a study of 14 people, the ratio of cone types predicted where pure green fell, and did not predict pure yellow.6 More often, a 2019 review says, “it has been difficult to show an association between unique hues and sensitivity.”7
The next place to look is the brain. Kuehni’s paper says that how the brain generates the unique hues “remains a complete mystery.”1 A 2005 study by Sophie Wuerger and colleagues concluded that the color mechanisms below the cortex are not the source of these hues, and that higher-order mechanisms “need to be postulated.”8 In 2017, a study that recorded the brain’s electrical activity reported a neural signature of the unique hues 230 milliseconds after a color appeared, at what its authors called “a post-perceptual stage of visual processing.”9 A 2019 review, citing that same study, described a neural substrate for pure blue and yellow as having “proven elusive.”7
Color science uses the four hues anyway. Color appearance models such as CIECAM02 describe a hue as percentages of up to two unique hues, so that an orange midway between red and yellow comes out as 50% red and 50% yellow. The models define those anchors for an average observer.10 Kuehni’s conclusion was that it is not justified, “certainly not in case of green and red,” to assume that an average can be considered representative of people.1
It would be easy to take this too far. Kuehni notes that some people repeat their own picks quite closely, and others vary more. The widest ranges come from pooling experiments that were run in different ways, which is why the “if” matters.1 The Wuerger study also measured the variability between people a different way, as perceptual errors, and reported that it is small. That fits, the study said, with the hypothesis that adult color vision can recalibrate itself from what it has seen.8
Four hues anchor the way color appearance models describe every other hue. The research I have described has not settled where they come from. I think more people should know that they exist.
-
Rolf G. Kuehni (2004). “Variability in Unique Hue Selection: A Surprising Phenomenon.” Color Research & Application, 29(2), 158–162. https://doi.org/10.1002/col.10237 The ranges given here are the extremes of its Table I. In its text the paper gives the green range as 490 to 555 nanometers. ↩
-
Ewald Hering (1878). Zur Lehre vom Lichtsinne. Sechs Mittheilungen an die Kaiserl. Akademie der Wissenschaften in Wien. Vienna: Carl Gerold’s Sohn, pp. 108–109. The passage is in the sixth communication, presented to the academy on May 15, 1874, and was read here in the second, unaltered printing of 1878. On red, his footnote reads: “Dieses reine Roth entspricht nicht dem Spectralroth, welches gelblich ist.” https://archive.org/details/zurlehrevomlich01herigoog ↩
-
Ewald Hering (1920). Grundzüge der Lehre vom Lichtsinn. Berlin: Julius Springer, pp. 41–42 and 274. https://archive.org/details/grundzgederleh00heriuoft The English wording of the four places is adapted from the translation by Leo M. Hurvich and Dorothea Jameson, Outlines of a Theory of the Light Sense (Cambridge, Mass.: Harvard University Press, 1964), as quoted in Kuehni (2004), p. 158. ↩
-
Michael A. Webster (2015). “Individual differences in color vision.” In Handbook of Color Psychology, pp. 197–215. Cambridge University Press. https://doi.org/10.1017/cbo9781107337930.010 ↩
-
David H. Brainard, Austin Roorda, Yasuki Yamauchi, Jack B. Calderone, Andrew Metha, Maureen Neitz, Jay Neitz, David R. Williams and Gerald H. Jacobs (2000). “Functional consequences of the relative numbers of L and M cones.” Journal of the Optical Society of America A, 17(3), 607. https://doi.org/10.1364/josaa.17.000607 ↩
-
Brian P. Schmidt, Phanith Touch, Maureen Neitz and Jay Neitz (2016). “Circuitry to explain how the relative number of L and M cones shapes color experience.” Journal of Vision, 16(8), 18. https://doi.org/10.1167/16.8.18 ↩
-
Kara J. Emery and Michael A. Webster (2019). “Individual differences and their implications for color perception.” Current Opinion in Behavioral Sciences, 30, 28–33. https://doi.org/10.1016/j.cobeha.2019.05.002 ↩
-
Sophie M. Wuerger, Philip Atkinson and Simon Cropper (2005). “The cone inputs to the unique-hue mechanisms.” Vision Research, 45(25-26), 3210–3223. https://doi.org/10.1016/j.visres.2005.06.016 ↩
-
Lewis Forder, Jenny Bosten, Xun He and Anna Franklin (2017). “A neural signature of the unique hues.” Scientific Reports, 7(1), 42364. https://doi.org/10.1038/srep42364 ↩
-
Mark D. Fairchild (2018). “Unique hues and principal hues.” Color Research & Application, 43(6), 804–809. https://doi.org/10.1002/col.22261 ↩