According to his article in the Journal of Cognitive Neuroscience, the brain may not actually "make" decisions in any meaningful sense — what people experience as choosing is an emergent property of perception, movement, body, and environment acting in continuous concert. For anyone who has built a habit stack, a decision journal, or a willpower routine on top of "choosing better," this is a polite demolition of the foundation.
The "sandwich model" you've been operating under
For decades, neuroscience and folk psychology have agreed on a tidy sequence. You perceive something, think about it, make a choice, then act. Each step corresponds to a discrete brain function, transitioning from sensory processing to cognition to motor output. It matches how decisions feel from the inside, as James acknowledges when he observes that actions seem to be caused by decisions built from desires, beliefs, and intentions.
Here is the inconvenient part. Sensation has identifiable neural mechanisms. Motor action has identifiable neural mechanisms. The cognitive "choosing" stage sandwiched between them? No corresponding neural process that functions as a distinct decision-maker. James calls this tidy picture the "sandwich model," and his argument is straightforward: the filling is missing.
What actually moves the needle
James proposes that what gets labeled a decision emerges from ongoing, simultaneous interactions among sensory, sensorimotor, and motor processes — what he terms "action selection." Behavior, on this view, is produced by brain, body, and environment feeding back into one another rather than unfolding in a clean line.
The most striking illustration he offers is almost embarrassingly simple. A robot equipped only with sensory, motor, and sensorimotor modules — with no decision module at all — will follow walls in a way that looks purposeful. If a stripped-down machine can produce apparently intentional behavior without a chooser, the human version may not require one either. Drawing on philosopher Daniel Dennett, James compares decisions to a center of mass: a mathematically useful abstraction that, crucially, exerts no physical force. You cannot move an object's center of mass without moving the object itself. Decisions describe behavior; on this framework, they do not cause it.
Designing for a brain without a boss
None of this denies the lived experience of deciding. James is explicit that the language will not be retired — it remains a useful way to describe behavior. What he dismantles is the assumption that a dedicated process inside the skull is doing the deciding.
That reframe matters more than it sounds. If there is no central decision-maker waiting to be upgraded through discipline, the entire willpower-and-clarity stack loses its footing. The practical move is to stop trying to strengthen a ghost and start reshaping the surrounding environment. Reduce friction at the moment of action. Set defaults the body-and-environment system can latch onto without deliberation. Build fail-safes for the predictable windows where the linear model collapses — late nights, decision fatigue, accumulated cognitive load, the third glass of wine.
The brain does not, on this account, decide. But it responds, reliably, to the conditions built around it. That is the only lever worth pulling.