The fold is real. The memory isn't.

Mimosa pudica gets called the sensitive plant for good reason: touch a leaflet and it folds shut in well under a second, an actual action potential racing along the rachis to trigger it. I built /touch around that, timed to a real measured propagation speed. What sent me looking in the first place wasn't the fold, though — it was a much bigger claim riding on the same plant, one I'd half-remembered as settled science and wasn't.

In 2014, Monica Gagliano and coauthors published a widely reported Oecologia paper arguing that Mimosa pudica can learn. Their method: drop potted plants repeatedly, a stimulus with no actual consequence, and watch the folding response fade — habituation, the same phenomenon a sea slug or a startled cat shows to a stimulus it learns is harmless. The plants stopped folding after enough drops, then — the striking part — some individuals still remembered not to fold weeks later, with no further drops in between. It was covered as the first real evidence of learning in an organism with no nervous system to learn with, and the framing stuck; I'd absorbed it as a settled fact somewhere along the way without ever checking where it came from.

It isn't settled. A 2018 response in the same journal, from Robert Biegler, went back through Gagliano's own published data and found the crucial control missing: the paper never tested dishabituation — exposing a habituated plant to a different stimulus and checking whether the folding response comes back. That test is what separates real learning from simple exhaustion; a motor system that's just run out of the ions or the turgor pressure it needs to fold again will also stop responding, to any stimulus, and recovers on its own schedule regardless of what caused the fatigue. Biegler's reading of the data he was working from: everything reported is consistent with ordinary fatigue, and nothing in the paper rules that reading out. No independent lab has since published a replication that closes the question either way.

None of that makes Mimosa pudica boring. A separate, less famous 2017 study (Reed-Guy et al., PeerJ) ran a more modest version of the same idea — repeated touches rather than drops, over minutes rather than weeks — and found the opposite direction of change: plants hid measurably longer after each successive touch, about 13.6 seconds more per trial, real sensitization rather than habituation. The same study found individual plants differ enough from each other that which plant you touched explained most of the variation in how long it stayed folded, and that plants running low on light — genuinely energy-stressed — cut their hiding time short, because staying folded means staying in the dark, and a plant that's already starving for light can't spare more of it. That's a real, specific, well-controlled result. It just isn't the same as the wide claim I'd walked in remembering.

The room builds the sensitization finding, compressed for a browser tab's patience, and says in its own prose what it isn't building: no fatigue-versus-learning distinction, no individual variation, no dishabituation test either. The most honest thing about a demo of a contested finding is not building the contested part.

Sources: Gagliano, Renton, Depczynski & Mancuso, Experience Teaches Plants to Learn Faster and Forget Slower in Environments Where It Matters, Oecologia 175(1), 2014; Biegler, Insufficient Evidence for Habituation in Mimosa pudica. Response to Gagliano et al. (2014), Oecologia 186(1), 2018; Reed-Guy, Gehris, Shi & Blumstein, Sensitive Plant (Mimosa pudica) Hiding Time Depends on Individual and State, PeerJ 5:e3598, 2017.