Summary: A new theoretical framework at the intersection of neurophilosophy and cognitive neuroscience challenges the long-held idea that the brain makes decisions through a single, centralized cognitive stage between perception and action. Instead, the framework argues that what we call “decisions” are emergent descriptions of continuous, circular interactions among sensory, sensorimotor, and motor systems rather than discrete neural computations localized in a specific brain region.
Key Facts
- Sandwich model questioned: The traditional view that perception → decision → action proceeds as a linear sequence with a central decision-making stage does not match current evidence about how the brain is organized.
- No localized decision module: While the brain contains specialized sensory and motor circuits, there is no clear, distinct neural structure that performs decision-making in the same way those sensory and motor systems do.
- Action selection, not central control: Behavior is better explained as action selection produced by interacting sensory, sensorimotor, and motor processes operating together rather than by a singular top-down controller.
- Decisions as abstract descriptions: From a physicalist standpoint, decisions are nonphysical, explanatory constructs—useful for describing behavior but not physical causes of movement, analogous to describing an object by its center of mass.
- Cartesian Theater problem: Positing an internal homunculus or central overseer leads to an infinite regress: explaining the brain by imagining a smaller brain inside it simply shifts the problem.
- Robot example: Simple robots built from only sensory and motor modules can show apparently goal-directed behaviors (for example, wall-following) without any internal decision-making system, illustrating how apparent choices can emerge from interactions with the environment.
- Methodological shift recommended: The framework calls for experiments informed by embodied cognition and ecological psychology that capture continuous sensorimotor loops tying brain, body, and environment.
Source: Indiana University
Indiana University Professor Tom James contends there is a mismatch between our folk-psychological sense of how decisions feel and the brain’s actual mechanisms. In everyday language and many scientific models, a choice feels like a distinct mental event that causes actions. That intuition supports research methods that assume a linear causal chain from perception to cognition to movement. But James argues this “sandwich model” does not square with neural organization: sensory and motor functions have clear neural substrates, whereas the putative cognitive decision stage does not.
Rather than treating decisions as an internal command, James proposes treating behavior as the result of sensorimotor mechanisms that continuously integrate sensory input with motor dynamics. He uses a physicalist perspective—drawing on philosophers who emphasize that only physical processes can cause physical effects—to argue that abstract entities such as “decisions” cannot themselves exert causal force. Instead, decisions are convenient, high-level descriptions of patterns produced by underlying sensorimotor activity.
To clarify this point James offers several analogies. One compares a decision to a center of mass: a useful mathematical and descriptive concept that does not itself push an object. Another compares the term “the university” to the actual meetings, phone calls, and actions of individuals; the higher-level term summarizes activity but hides the underlying mechanisms. A striking empirical illustration comes from robotics: a simple robot equipped only with basic sensors and motors can follow walls and behave in ways observers interpret as intentional, even though the robot has no internal planner or strategy module. This example demonstrates how complex-looking behavior can emerge from straightforward sensorimotor loops interacting with the environment.
James also reiterates the philosophical objection known as the Cartesian Theater: explaining cognition by positing a central observer or controller inside the brain just replaces the problem with a smaller version, producing an infinite regress and offering no mechanistic explanation.
An experimental path forward
Given these arguments, James recommends a shift in experimental practice. He urges cognitive neuroscience to adopt paradigms that preserve embodied interaction and ecological validity—settings in which participants actively sense and act, continuously updating sensorimotor mappings—so researchers can reveal the circular, distributed processes that generate behavior. In his own lab he draws on embodied cognition and ecological psychology to design such studies, aiming to map the dynamics that have previously been obscured by linear models.
Key Questions Answered:
A: Language and everyday explanation compress complex sensorimotor dynamics into simple terms. Calling something a “choice” is a useful description of behavior, but the feeling of making a choice can arise from the integrated, ongoing interaction of senses and muscles rather than from a localized controller.
A: It is the problematic idea that a central observer inside the brain makes decisions. This model leads to an infinite regress—each internal observer would require another observer inside it—and fails to provide a real mechanistic explanation.
A: The robot displays apparently strategic behavior using only simple sensing and movement rules. Its example shows that complex, goal-like actions can emerge from interactions between sensors, motors, and the environment without any internal decision module.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The referenced journal paper was reviewed in full.
- Additional explanatory context was added by staff.
About this neuroscience research news
Author: Elizabeth Rosdeitcher
Source: Indiana University
Contact: Elizabeth Rosdeitcher – Indiana University
Image: The image is credited to Neuroscience News
Original Research: Closed access. “Sensorimotor Mechanisms of Decisions and Actions” by Thomas W. James. Journal of Cognitive Neuroscience. DOI: 10.1162/JOCN.a.2484
Abstract
Sensorimotor Mechanisms of Decisions and Actions
Decisions are commonly treated as intermediaries between perception and action, but the extent to which that assumption shapes cognitive neuroscience varies. This perspective argues that decision processes do not themselves cause actions; rather, sensorimotor mechanisms underlie behavior. The article outlines variations on causal models linking decisions and actions, defends the claim that sensorimotor processes drive action, and proposes experimental approaches—especially more ecological testing environments where participants actively update sensorimotor routines—that can reveal those mechanisms.