Sometimes the simplest questions are the hardest to answer. How, for example, do you decide to wiggle your fingers? A lot is known about the muscles and neural structures involved—the puppet and the strings, as it were—but what about the puppeteer?
How humans develop the ability to willfully make things happen still remains mysterious. In a recent study, we tried to catch infants in the act of discovering their ability to influence the world. As we reported last September in the Proceedings of the National Academy of Sciences USA, we identified these aha! moments and the events surrounding them, revealing for the first time how agency forms.
For more than 50 years researchers have used a very simple method to investigate learning in infancy. They place a baby into a crib with a mobile suspended above it. Then a scientist ties one end of a string to the mobile and the other to the infant’s foot. If the baby moves, the toy will, too. Over multiple sessions, scientists can observe as the infants learn and recall a simple cause-and-effect interaction: kick a foot, and the mobile moves.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
We used that setup to identify the moment when babies first realize they can control the mobile’s motion. We worked with 16 infants who were three to four months of age, employing motion-capture technology to measure the movements of both infant and mobile in three-dimensional space. As in past such experiments, infants kicked significantly more when their foot was tethered to the mobile than when it was not. But did they know that their movements were propelling the mobile?
One clue came when an experimenter pulled the string to make the mobile move instead of letting the baby do it. Infants moved less in that situation than when the mobile was stationary. This finding rules out the idea that babies simply kick in excitement when they see the mobile moving. In fact, our data indicated that it was the coordinated movement of the foot and the mobile that prompted a baby’s activity.
Further, when we removed the string, the babies kept on kicking at the highest rate they had reached while tied to the mobile, suggesting that they expected the toy to respond. Some infants were visibly frustrated when that did not happen.
At some point while they were tethered, then, the infants must have figured out that they were in control. To pinpoint this aha! moment, we developed an algorithm to search for spikes in the rate of foot movement during tethering. For some infants, we found a sudden burst of activity coupled with dramatic changes in the rate of movement, including pauses and abrupt increases or decreases in speed. Such fluctuations and shifts are typical signatures of complex systems—from stock markets to brain activity—that are on the verge of change. We believe those initial spikes in activity coincided with those babies’ aha! discovery.
Our observations also revealed a pattern: in the first minute or so of tethering, each time the mobile responded to the infants’ movements, they froze and waited for the mobile to stop before kicking again. The mobile’s unfamiliar and unexpected movement triggered a strange, dynamic dance on the babies’ part: move, pause, move, pause. We suspect that they were, in a sense, running their own experiment: “If I do this (kick), I see that (the mobile moves),” and, conversely, “if I do not kick, I do not see the mobile move.”
Notably, the infants did not all behave in the same way. In fact, one child showed no such signs of discovering how her behavior might affect the mobile, even though she doubled her activity while interacting with it. This approach could therefore help us understand and predict individual paths of motor and cognitive development for both healthy infants and those at risk of developmental delays.
But what does this experiment tell us about the origins of agency? Quite a lot, we would argue. Several years ago one of us (Kelso) proposed that the birth of agency is a dynamic, self-organizing process and, together with the late physicist Armin Fuchs, developed a model of it. To break that down: “Dynamic” means that patterns and relationships evolve over time. “Self-organization” refers to the fact that many complex systems in nature organize themselves into specific forms without any instructions. In these cases, to return to our opening question, there is no concrete puppeteer pulling the strings. Rather patterns emerge and change spontaneously in systems that are open to exchanges of energy, matter and information with their surroundings.
Our theory is that when an organism (here, a baby) and its environment (the mobile) interact, they form a self-organizing dynamic system. Goal-directed action emerges spontaneously when the organism realizes that its movements cause the world to change.
Our findings align beautifully with our theory. The babies’ initial movements consisted of squirming and thrusting without discernible purpose or direction. But once tethered to the mobile, the more intensely they moved, the more their attention was drawn to the effect their kicking had (both the feeling of the string tugging and the sight of the mobile responding). When babies’ attention to their relationship with the toy reached a critical level, they realized they could make the mobile spin. Spontaneous movements became purposeful action. At that point of transformation, we observed a burst of foot activity and tight coordination between the infant and the mobile.
Historically, the entire issue of purpose and agency in living things—and, dare one say, “free will”—has been clouded in philosophical debate and controversy. Many arguments land at one of two extremes: either there must be an inner director—a self that makes decisions—or free will does not actually exist, because the environment and circumstances predetermine a person’s behavior. But our infant study emphasizes how understanding the relationship between an organism and its environment is essential to uncovering the origins of directed behavior. As our model proposes, the experience of agency emerges only when an organism (the baby) senses it is coupled to its environment (in this case, the mobile setup). In this way of thinking, the interaction and relationship between the two are crucial for purpose to arise.