The tongue map that came from a misread graph

In 1901 the German scientist David Hänig measured how sensitive different edges of the tongue were to different tastes. He found small differences: some areas needed a little less of a substance before a person noticed it.

In 1942 the Harvard psychologist Edwin Boring retold the results in a history of psychology, with a graph of relative sensitivity. Later writers read that graph as meaning each zone could detect only one taste, and the familiar map, with sweet at the tip and bitter at the back, entered textbooks for decades.

Researchers re-examined the question in the 1970s and found that all areas of the tongue that have taste buds can detect all the basic tastes. Work on the molecules of taste, including a 2006 review in Nature by Charles Zuker's group, showed that taste receptors for each basic taste are spread across the tongue. Umami, a savoury taste first described by the Japanese chemist Kikunae Ikeda in 1908, is now widely accepted as a fifth basic taste.

The map survived because it was simple, easy to draw and copied from one textbook to the next. Few people tested it themselves, though a drop of salt water on the tip of the tongue shows the problem quickly.

What is true is that sensitivity varies slightly from place to place, and the centre of the tongue has few taste buds. That is a long way from separate zones.

The AI angle: AI models learn from text that copies other text, so an error repeated in many places can look like settled fact. Look for the original source, and when you can, test a claim yourself.

The idea behind this caseLesson: Check what AI tells you →

1. 1Ask whether a claim traces back to an original study, or only to other books copying each other.
2. 2Try a simple test yourself when one is possible.
3. 3Watch for small real differences that get exaggerated into sharp categories.
4. 4When an AI tool states a classic fact, check a recent scientific source, since old errors are common in training text.