What the Brain Is Doing When We’re Unaware of Our Actions

Hal Blumenfeld, MD, PhD, Mark Loughridge and Michele Williams Professor of Neurology at Yale School of Medicine (YSM), got out of the pool one afternoon and couldn't find his watch. He was certain he'd left it at the side where he entered. He found it on the opposite end. He'd moved it himself while swimming, with no memory of doing so.

"I had no idea that I had done that," Blumenfeld says. "But I clearly did."

A new YSM study, published in PNAS Nexus and led by Blumenfeld and David S. Jin, PhD, a former PhD student in the Yale Interdepartmental Neuroscience Program, is the first to examine the brain signals that determine whether we're conscious of our own actions as we perform them and resolves a debate that has persisted for more than 130 years.

In the late 19th century, psychologists William James, MD, and Wilhelm Wundt, MD, staked out opposing theories of action awareness. James believed it comes after an action, through sensory feedback. Wundt argued it arises from the brain's act of planning and initiating movement, before any sensation occurs. The debate went unresolved because no one had a method to test it.

But Jin, more than a century later, had an idea. His solution was to adapt the sliding block puzzle game Rush Hour, in which players move toy cars and trucks around to create a path for one main car to exit the puzzle.

For the study, Jin had participants play the game while simultaneously watching background videos they were told to memorize. Periodically, the game paused and asked them to identify their last move and rate their confidence. Correct answers with high confidence were deemed "aware." Incorrect answers with low confidence were labeled "unaware."

"If you correctly reported what you were doing, but felt no agency in it, would it be true awareness?" asks Jin, now a postdoctoral fellow at University of Alabama at Birmingham. "When you drive from point A to point B and get out of your car and think, ‘Wait, how did I get here?’—I wanted to recapitulate that feeling in a behavioral task."

The researchers measured brain activity with electroencephalography (EEG) across 67 participants as they completed the task. They found that brain signals differed between aware and unaware moves both before and after the action. A motor-planning signal called the pre-movement positivity was stronger in aware trials. A sensory processing signal called the N140, tied to awareness of bodily sensation, was also enhanced. Neither alone was sufficient.

"It turns out James and Wundt were both right," says Blumenfeld, senior author of the study. "Both the volition- and the perception-related signals are bigger when we're aware of what we do."

The study also found a third factor neither theorist had anticipated: As participants worked deeper into the session, pupil diameter shrank—a proxy for dwindling alertness—and awareness declined in lockstep.

"As the task went on, participants got a little more bored, a little more tired, a little more distracted," Blumenfeld explains. "Their pupils got smaller and smaller, and they became less and less aware of what they were doing."

For Blumenfeld, the multiple signals reflect something fundamental. "Consciousness is so important that it makes sense in evolutionary terms that there's going to be redundancy. Many things are influencing how aware or unaware we are of what we do."

The findings also reframe unawareness not as a failure but as a feature. "To be aware of your actions all the time would really suck," Jin says. "Perhaps it is unawareness of action that allows for daily life to proceed smoothly. Imagine being a musician and having to think out every note moment by moment."

Jin's interest in the subject is partly personal. As a person with epilepsy, he has experienced firsthand what it's like to hold conversations and play piano during a seizure, with no memory of any of it afterward. "The ability to perform all these complex actions and yet not be able to describe them afterwards is just incredibly fascinating," he says.

The same brain signals disrupted in unaware participants are also diminished in Parkinson's disease patients, with similar disruptions appearing in schizophrenia and following stroke—conditions where impaired awareness of one's own actions has real consequences for diagnosis, rehabilitation, and legal questions of intent.

Thousands of studies have examined perceptual awareness, but the equivalent question for action has gone largely unasked. "It basically opens up the whole second half of the field," Blumenfeld says. "It essentially doubles the amount of research that can be done now."

The lab's next step is functional magnetic resonance imaging (fMRI), which will offer better spatial resolution to untangle some of the study's remaining puzzles, including how signals deep in the brain not reachable by surface EEG may contribute to awareness.

Reflecting on how much the results surprised him, Jin says, "When I first began, I thought I'd be calling for Iron Man. One clean signal that could explain awareness of action. But as I went through all the signals we were observing, it was really like Avengers Assemble: Signals across a number of different cognitive domains, all showing up at once."