How have medical professionals measured pain throughout history? Not very well, it turns out.

In one famous test in the late 1940s, two American male researchers performed pain experiments on pregnant women in labor. Whenever a woman had a contraction, they burned her hand with a machine, and asked how the pain of the burn compared to the pain of the contraction. It was a horrific, pointless exercise that risked the women’s health and safety. It was also one in a series of attempts to quantify and measure people’s pain.

Pain is hard to measure because, while it’s a universal human experience, it’s also highly subjective—and also because we don’t really know what it is. The combination of physical and psychological factors that cause pain remain somewhat mysterious, even with recent advances in brain imaging. In fact, the desire to measure pain appears to be a fairly modern concern in Europe and the United States.

“From medieval time onwards, people were much more interested in causing pain than in measuring it,” observes Stephen McMahon, a professor of physiology at King's College London and director of the Wellcome Trust Pain Consortium. “[René] Descartes, who famously described a mechanism for pain, didn’t describe how you measured it, no. And the Greeks, they didn’t even include pain in their basic sensations that they felt people could perceive. So no, I think it’s quite a modern endeavor, trying to evaluate [pain].”

The 19th-century German school of “psychophysics” offered one early inquiry into how to measure pain. Its goal was to study the relationship between stimuli and sensation, and it led scientist Maximilian von Frey to develop a method of identifying what he called Schmerzpunkte, or pain points, with horse hair. Specifically, he’d select hairs of varying stiffness from a horse’s tail and attach them individually to sticks. Then he’d used the stick to press the hair against someone’s skin.

“The stiffer the hair, the more pressure it took for the hair to bend.” McMahon says. “Each would exert a different force before they bent… He used them to test the sensitivity of skin, and you can still find them today; they’re plastic today, they’re not made of horse hair.”

Using this method, Von Frey could record the amount of pressure at which a person started to feel pain from a particular hair. He and others involved in psychophysics also employed other methods of testing skin sensitivity, like hot or cold rods. Their research, McMahon says, “drove the development of a whole number of scales and techniques.”

One group of researchers influenced by these experiments were James Hardy, Helen Goodell and Harold Wolff. In 1940, they announced they’d invented a new device to measure pain thresholds called the “dolorimeter.” It used heat to inflict pain at various levels, and was strong enough to give people second-degree burns—which it did when Hardy and another researcher named Carl Javert tested it on pregnant women in labor several years later.

Hardy and Javert seemed to have little regard for the complaints of the pregnant women they tested the dolorimeter on, writing that one “patient became so hostile that attempts at further measurements were abandoned… [T]his failure to obtain valid measurements was due mainly to an unwillingness on the part of the patient to cooperate.” In their research, they published a graphic picture of what one woman’s hand looked like after they used the dolorimeter on her at its highest setting.

As with Von Frey hairs, you can still buy modern versions of dolorimeters today. But since the 1950s, questionnaires and pain scales have overtaken them as the main way that doctors measure pain in their patients. Questionnaires ask patients a range of queries to get a sense of how they’re feeling and what might be wrong. Pain scales ask patients to rate their pain based on numbers, a series of increasingly distressed-looking cartoon faces or simply by pointing to a place on a straight line to indicate where their discomfort lies on a scale from “no pain” to “worst pain.”

There are also scales for infants and children or adults who are nonverbal, so that doctors can try to assess their pain. For example, the FLACC Behavioral Scale helps doctors look for signs of pain in children between two months and seven years. These signs include things such as a clenched jaw, intense and prolonged crying, kicking legs and rigid activity.

The big shift between Von Frey hairs and dolorimeters on the one hand, and questionnaires and scales on the other, is that instead of inflicting pain on test subjects and then trying to study it, researchers today try to study pain that their patients already have. In McMahon’s work, he says he usually uses a visual analog scale of a straight line to ask his patients to assess their pain. Yet he emphasizes that there is no perfect method.

“The point to realize is that there is no objective measure of pain,” he says. “Pain is a subjective sensation… We can report on it, but we don’t have an objective machine that can assess it. And that’s true today, and it’s always been true.”