Kant and the Four Temperaments: How the Humors Became a Grid

In 1798 Kant kept the four temperaments and threw out the fluids underneath them. Wundt and then Eysenck turned what he left behind into two crossed axes, which is how Galen's four names ended up on a modern personality diagram.

Kant taught the same winter lecture course in Königsberg for more than twenty years before he allowed any of it into print. It was a course on anthropology, open to students and to curious townspeople, and it was popular in a way his hard philosophical seminars were not. In 1798, old and nearly finished with his working life, he assembled the material into a book called Anthropology from a Pragmatic Point of View. In the second part, among observations on faces and on national character, he takes up the four temperaments. He does not present them as a curiosity from the past. He presents them as still usable, which is strange coming from a man who had spent forty years taking apart inherited furniture of exactly that kind.

The blood becomes a figure of speech

Kant keeps all four names. He even keeps the blood inside the German words he chooses for them: leichtblütig, light-blooded, for the sanguine; schwerblütig, heavy-blooded, for the melancholic; warmblütig, warm-blooded, for the choleric; kaltblütig, cold-blooded, for the phlegmatic. But he is clear that this is a manner of speaking and not a physiology. The names arrived from an old theory of the fluids, he uses them because they are convenient and because he thinks the four kinds they mark out are real, and the machinery underneath he simply sets aside. The idea that a man is choleric because yellow bile predominates in his body, which is what Galen actually meant by the word, no longer does any work.

That is a bigger move than it looks. For most of two thousand years the temperaments had been medicine. They told a physician what was wrong with you and what to feed you. Kant cuts that tether and leaves four kinds of mind standing on their own. This is the moment the temperaments stop being a diagnosis and start being a psychology.

Feeling on one side, activity on the other

Then he sorts them, and not into a flat list of four. He makes two pairs. Sanguine and melancholic are temperaments of feeling. Choleric and phlegmatic are temperaments of activity. Inside each pair there is a light member and a heavy one. The sanguine feels quickly and briefly, the melancholic slowly and for a long time. The choleric starts hot and burns down fast, the phlegmatic is difficult to start and then difficult to stop.

Two pairs, each with an internal opposition. Written on paper that is very nearly a grid, though Kant would not have drawn it as one. He also insists the four cannot be compounded: you have one temperament, not a mixture of two, a stricter line than the tradition usually took and stricter than most readers of temperament blends would accept today. Still, the raw material for a grid is lying there on the table.

Wundt draws the thing

Wilhelm Wundt, who opened the first laboratory for experimental psychology at Leipzig in 1879, picked the same four names up and set them against two variables. One was how fast a person's feeling changes, quick or slow. The other was how strong the feeling is, strong or weak. Choleric: strong and quick. Melancholic: strong and slow. Sanguine: weak and quick. Phlegmatic: weak and slow.

Wundt was not restoring the humors. He was doing what Kant had made possible, treating the four names as labels for positions rather than for fluids. Two variables, four combinations, four ancient words. It is a two axis grid in everything except the drawing.

Eysenck puts the old names back on the rim

Hans Eysenck, working at the Maudsley Hospital in London in the middle of the twentieth century, built a personality model out of factor analysis of questionnaire answers. Two dimensions came out of the numbers: extraversion against introversion, and neuroticism against emotional stability. Then he did something a more cautious man would have skipped. He published a circle with his two axes crossed at the centre, the trait words around the rim, and in each of the four quadrants an old Greek name. Stable extravert, sanguine. Unstable extravert, choleric. Stable introvert, phlegmatic. Unstable introvert, melancholic. He pointed at Wundt openly as the source of the arrangement.

He was not being decorative. He thought the old sorters had found the same four corners he had.

What actually survived

The line runs Galen, Kant, Wundt, Eysenck. The bile did not make the trip, and neither did the phlegm. What survived was a shape: two independent axes, four quadrants, one name in each corner.

The humors were an answer to why a person is like this. The temperaments were an answer to what they are like. Kant is the place where the two came apart, and only one of them kept going.

The shape is now so ordinary we stop seeing it. Introversion and extraversion crossed with something else, four boxes, four labels: that is the skeleton under a great many modern personality instruments, including some that have never heard of a humor. When you take the test here, you are meeting a vocabulary from the second century sitting on a geometry from the eighteenth, and it is the geometry that did the surviving.