In 1906, zoologist Herbert Spencer Jennings published Behavior of the Lower Organisms, a book that contained a provocative idea: microbes can change their minds.

His subject was a single cell bristling with beating hairs called Stentor. These trumpet-shaped predators are so large fish can eat them and humans can see them, and so brazen they can catch and eat rotifers—proper animals with hundreds of cells and a simple brain. In the microbial galaxy, stentors lie somewhere between Star Destroyer and sarlacc pit.

Jennings decided to annoy it and see what happened. When confronted with a stream of irritating carmine powder expertly aimed at their mouths by his steady hand, Stentor would first bend away, then reverse the beating of its hairs (called cilia) to expel the powder, then contract and finally detach.

He noted that the order of behaviors varied somewhat with different stimuli (he tried other chemicals) and steps were sometimes omitted. “But it remains true,” he wrote, “that under conditions which gradually interfere with the normal activities of the organism, the behavior consists in ‘trying’ successively different reactions, till one is found that affords relief.”

In short, stentors could confront a stimulus with one behavior, and then choose a costlier approach if the irritant persisted. At least for a short while (a period that Jennings declared difficult to determine experimentally and still unresolved), it could “remember” that it had tried one solution without success, and opt for another.

But in 1967, scientists from a different school of animal behavior repeated his experiment and failed to produce the same result. And with that, Jennings’s findings were consigned to the dustbin.

Then about 10 years ago, Jeremy Gunawardena, an associate professor of systems biology at Harvard Medical School, discovered the experiment and its defenestration and decided that it deserved another look. To his surprise, he discovered the 1967 team had not used the correct species of Stentor (being behaviorists who believed variation flowed from the environment and not genes, they might have felt the species didn’t matter). The one they had chosen, Stentor coeruleus, strongly prefers to swim, unlike Jennings’s Stentor roeselii, which prefers to chill poolside.

Gunawardena became fascinated by what replicating the experiment might reveal about what single cells are capable of. After years of dangling the idea fruitlessly at lab meetings, he found undergrad Joseph Dexter and postdoc Sudhakaran Prabakaran were willing to give it a try at night and on weekends—with no funding.

This time, the Harvard team managed to track down the correct species in an English golf course pond, construct their own “Device for Irritating Stentors” (being quantitative biologists, they lacked Jennings’s extreme pipette skills), and discovered something extraordinary.

In their setup, Stentor did not respond to carmine powder the way Jennings described. However, when faced with barrages of 21st-century plastic microbeads, individual Stentor roeseli behaved consistent with Jennings’s description—and in one remarkable way that Jennings did not observe in 1906.

If Stentor really can “decide,” it certainly isn’t the only way the ciliates—the group of shaggy microbes to which Stentor belongs—resemble us. A ciliate operates like an animal at the scale of a single huge cell, and the resemblance can be startling.

For example, some glue bundles of their cilia into structures called cirri and can use them as legs, mouths, paddles or teeth. Euplotes skitters nimbly along surfaces atop cirri like some sort of Close Encounters–class water flea. The cirri are wired by nervelike neurofibrils. If the fibrils are cut, the cirri fall limp.

Some ciliates pack tiny tethered darts they can fire to attack prey, deter predators or simply drop anchor. Others sport tentacles that snag food. Like sea stars, ciliates can regenerate entire bodies within a day or two from shockingly tiny pieces provided those pieces contain both a bit of the cell’s cilia-studded armor and a bit of nucleus, the cell’s genetic heart. Many ciliates divide in the usual way by pinching in two, but some stalked or sessile ciliates push small round larvae into the world through a special birth canal.

One ciliate called Diplodinium lives in the rumen of cows and other hoofed animals, a special environment known to harbor all kinds of strange things, about half of which by mass may be ciliates (think about that next time you see a cow placidly chewing its cud). Diplodinium contains neurofibrils, cirri, musclelike striated contractile fibers called myonemes, a “backbone” made of stacked plates, a mouth, an esophagus that contracts with the help of a ring tethered to its exterior, and an anus. But remember: single cell.

In short, ciliates have taken the biology of the solo cell to its apparent earthly limit. Having something like a noggin in there is less credulity-stretching once you grasp this.

In the new study, published in the journal Current Biology in 2019, the scientists found that Stentor indeed switched behaviors in response to repeated puffs of beads, and the order of operations was generally consistent with Jennings’s description. Detachment was always preceded by contraction, and mathematical analyses revealed cilia alternation or bending were far more likely to appear before contraction than after.

There is something else interesting about their data, which I encourage you to examine for yourself: it sure looks like stentors have personalities. Some repeatedly contracted and relaxed, or bent, contracted, then relaxed, seemingly willing to tolerate irritation—or to live dangerously. These were the optimists.

Some contracted once or just a few times, never to relax again. Others contracted and detached, and that was it. These were the pessimists (or perhaps just the ones with a more recent successful “door dash”).

Some stentors always responded with one or two preferred behaviors, and never with others that they were surely just as biologically capable of performing. One indefatigable individual subjected to 13 bead blasts responded persistently with ciliary alternation or contraction, never bending or detachment.

Does Stentor possess something like agency—a capacity to make decisions? This study and Jennings’ evidence certainly suggest so.

There was a final provocative finding. This team's statistical analysis revealed that the choice between contracting or detaching was consistent the probability of a fair coin toss. In other words, it seemed perfectly random.

There’s only one problem: no known cellular mechanism can produce this result. That head scratcher remains both unreplicated and unexplained.

Perhaps it is time to let go of our preconceived notions of what cells are capable of because they are only cells, and the cells in our own soviet-style bodies are the equivalent of worker bees. The capabilities of wily, gunslinging, free-living cells may well exceed our dim primate imaginations.