Why do we feel in control when we move? A groundbreaking study from the University of Tokyo reveals that our sense of agency doesn’t just happen — it’s built step by step as we learn new motor skills through trial and error.
Key Points at a Glance
- University of Tokyo researchers explored how we develop a sense of agency (SoA) when learning new movements.
- Trial-and-error learning, or motor exploration, plays a crucial role in forming this sense of self-control.
- Participants used a data glove to map finger movements to cursor movements on a screen.
- Only active exploration — not passive imitation — enhanced participants’ feelings of agency.
- Findings may advance rehabilitation, VR, and brain-machine interface design.
What does it really mean to feel in control of your actions? Every time you reach for your coffee or type on a keyboard, a complex internal process tells you: this is me doing this. This sensation — known as the sense of agency (SoA) — is something most of us take for granted, yet scientists are still unraveling the mystery of how it emerges, especially when we’re faced with unfamiliar tasks.
A team of researchers at the University of Tokyo has now taken a significant step in decoding this enigma. Their recent study, led by Assistant Professor Takumi Tanaka, offers a deep look into how the sense of agency forms when we acquire new motor skills — and it all begins with a leap into the unknown.
At the heart of the study is the concept of motor exploration — the often clumsy, trial-and-error phase we all go through when learning something new, from playing piano to regaining movement after an injury. The researchers hypothesized that it’s this exploratory phase, not mere repetition, that helps our brain construct the internal models needed to feel in control.
To test this, they designed an experiment using a data glove that translates finger movements into cursor movements on a computer screen. Participants had to figure out, on their own, how their gestures mapped to cursor actions. In the early stages, participants mainly relied on whether their hand and the cursor moved simultaneously — temporal synchrony — to judge whether they were in control. But with enough practice, something shifted. Participants began to internalize the hand-to-cursor mapping itself. They no longer just saw matching timing; they felt that the cursor was an extension of their intention.
This is a pivotal finding. It suggests that the brain’s comparator model — the idea that we feel agency when our predictions match actual feedback — is not a fixed mechanism we’re born with. Rather, it’s something we build as we explore, fail, adapt, and refine our movements. Those who achieved higher accuracy in the task also reported a stronger sense of agency, reinforcing the connection between motor mastery and self-perception.
To further test their theory, the team ran a second experiment. This time, instead of free exploration, participants were instructed to imitate specific hand gestures that moved the cursor to predefined targets. The result? No significant enhancement in their sense of agency. Without the freedom to explore and discover the rules themselves — like how bending the index finger shifts the cursor right — participants didn’t form the same connection. This points to the importance of what the researchers call “structural representations”: the mental models that emerge only when we engage with the mechanics behind our actions.
The implications of this research stretch far beyond the lab. By illuminating how we develop agency through motor exploration, the study could inform new therapies for stroke patients relearning how to move, more intuitive virtual reality systems, and smarter brain-machine interfaces where users can feel connected to artificial limbs or devices.
In essence, the sense of agency is not just a passive experience but an active construction — one shaped by curiosity, mistakes, and the body’s dialogue with the world. In learning to move, we learn to be.
Source: University of Tokyo
