Why Taste Cannot Be Indexed Like RGB or Wavelength

. By Erik Yang

Diagram comparing colour and sound, each mapped by a few variables, with flavour drawn as a complex system

Essay

Human vision relies on 3 types of cone photoreceptors. All respond to the same physical property: wavelength of light. Trichromacy means any perceivable color can be reproduced by mixing three primaries. The mapping from stimulus to percept is linear, additive, and universal across nearly all observers.

Auditory pitch maps to a single physical dimension: frequency of air pressure waves. One stimulus property. One perceptual axis.

Now taste.

You have ~25 bitter receptor genes alone, each encoding multiple receptor variants. The TAS2R38 gene creates a 100 to 1,000-fold sensitivity gap between tasters and non-tasters (Kim et al., 2003, Science). One person’s “intensely bitter” is another person’s “tasteless.” Same molecule.

Then smell, which drives most of what we call flavor. ~400 receptor classes responding to ~5,000 chemical parameters. Non-linear. Non-additive.

Structurally different molecules can smell identical. Identical molecules can smell opposite depending on your receptor gene repertoire (Doty, 2025, Chemical Senses).

But the deepest problem is not biological. It is cognitive.

Linda Bartoshuk showed that when a supertaster and a non-taster both rate a stimulus “7 out of 9,” they are describing completely different experiences. The scale itself cannot capture the difference. She called it the El Greco Fallacy: a painter with astigmatism stretches his figures but sees them as normal, because he views the world and the canvas through the same distorted lens (Bartoshuk et al., 2002).

Paul Rozin spent decades showing that food preference is not a sensory readout. It is a cognitive construction shaped by learning, memory, culture, and identity. Different experience. Not because of different tongues. Because of different minds (Rozin et al., 1999).

Color has a universal coordinate system because 3 receptor types map to 1 physical dimension.

Taste has none because hundreds of receptor types map to thousands of chemical dimensions, filtered through personal biology, processed through learned cognition, and expressed through culturally-bound language.

The question is not how to standardize taste. It is how to model the variation back to its high dimensional nature.

Related

Another essay, and the science behind it.

Book demo