Performing an action and watching someone else perform the same action trigger the same set of neurons — a real "motor resonance." But it got hyped into "the cells that built civilization," then dragged back by science. This page sorts out what it is from what it isn't.
In the 1990s, Rizzolatti's team in Parma, Italy, was studying the premotor cortex of macaques (an area called F5 that plans actions like grasping). They recorded neurons that fired when the monkey reached for food itself — no surprise, they're motor neurons. The surprise: when the monkey sat perfectly still and merely watched the experimenter reach for the same thing, the same neurons fired again. As if "mirroring," hence mirror neurons.
You can't record single neurons in humans the way you can in monkeys (barring rare clinical windows), but brain imaging found a set of regions with the same "active for both doing and watching" property, collectively the mirror system (or action-observation network). Two leads bookend an action from "plan" to "perception":
Three fairly solid functions, all still at the level of action, not mystical "mind-reading":
Motor resonance — watching another's action, your motor system faintly "rehearses along," possibly helping you predict their next move faster (a ball-sport athlete reading an opponent's wind-up to start early is thought to relate to this). Imitation and learning — aligning "what I see" directly to "how I'd do it" is a natural interface for imitation. Part of understanding an action — recognizing "he's grasping," not "he's pushing." Note it's all understanding at the action level, not "why he's grasping, or what's in his mind."
From the late '90s into the 2000s, mirror neurons were crowned a master key: one voice called them "the cells that built civilization," claiming empathy, language, cultural transmission, theory of mind all rode on them, and proposing autism as a "broken mirror." A stirring story the evidence didn't keep up with.
The sober corrections: ① mirroring ≠ theory of mind — what actually tracks others' beliefs is a separate mentalizing network (medial PFC + temporoparietal junction), different regions entirely. ② mirroring ≠ empathy — empathy has its own affective line through the anterior insula/cingulate. ③ the mirror property is likely learned by association (you almost always see an action while doing it, tying "seeing" to "doing"), not necessarily an innate mind-reading organ (Heyes's associative account). ④ the broken-mirror theory is weakly supported and has largely receded. Hickok's The Myth of Mirror Neurons catalogs these over-readings.
Bottom line: the mirror firing is real and useful — but it's one piece of the social puzzle, not the master key.
Remember this page in one line: the brain reads others by at least two different routes. The mirror system (IFG + inferior parietal lobule) runs "direct resonance" — mapping observed actions into your own motor system; fast, automatic, bodily. The mentalizing network (medial PFC + temporoparietal junction) runs "inference" — slower, more abstract, handling invisible beliefs and intentions. Stop treating mirror neurons as "the seat of empathy/mind-reading": they handle actions; mind-reading is the job of the net next door.
Topic 9 The social brain & theory of mind · future issues on the motor system / imitation learning
Mirror neuron · Rizzolatti · the Parma discovery · Hickok · The Myth of Mirror Neurons · Heyes · associative learning account