Abstract
Before he wrote the recent book The Brain from Inside Out, the neuroscientist György Buzsáki previewed some of the arguments in a paper written 20 years ago ("The brain-cognitive behavior problem: a retrospective"), now finally published. The principal focus of the paper is the relationship between neuroscience and psychology. The direction in which that research had proceeded, and continues now, is, in his view, fundamentally misguided. Building on the critique, Buzsáki presents arguments for an "inside-out" approach, wherein the study of neurobiological objects has primacy over using psychological concepts to study the brain, and should, in fact, give rise to them. We argue that he is too pessimistic, and actually not quite right, about how the relation between cognition and neuroscience can be studied. Second, we are not in agreement with the normative recommendation of how to proceed: a predominantly brain first, implementation-driven research agenda. Finally, we raise concerns about the philosophical underpinning of the research program he advances. Buzsáki's perspective merits careful examination, and we suggest that it can be linked in a productive way to ongoing research, aligning his inside-out approach with current work that yields convincing accounts of mind and brain.
Introduction: The Mapping Problem
For decades, neuroscience has struggled with a fundamental "mapping problem": how do we connect the physical "hardware" of the brain—neurons, synapses, and oscillations—to the "software" of the mind, such as memory, language, and emotion? This gap between biological implementation and cognitive function remains the central hurdle in understanding the human experience.
Recently, prominent neuroscientist György Buzsáki argued that neuroscience is "fundamentally misguided" because it relies on "outside-in" psychological concepts—terms like "attention" or "memory" inherited from philosophy—to study the brain. He proposes an "inside-out" approach, where we start with neural mechanisms and let them define the categories of behavior. In this commentary, David Poeppel and Federico Adolfi push back, addressing whether we can actually "do" neuroscience without psychological concepts and arguing that Buzsáki’s proposed "epistemological primacy of the hardware" is both practically impossible and theoretically flawed.
The "Inside-Out" Challenge and the Implementation Sandwich
Buzsáki’s "inside-out" framework suggests that we should abandon pre-existing psychological taxonomies and instead build a new vocabulary based on what the brain actually does. However, Poeppel and Adolfi identify a recurring pattern in such research programs, which they term the "implementation sandwich."
Even when researchers claim to be working from the hardware up, they inevitably "sandwich" their neural data between psychological concepts to make it meaningful. For instance, describing a specific hippocampal firing pattern as "searching for stored items" relies on the very psychological concepts the "inside-out" approach seeks to replace. The authors argue that one cannot escape the necessity of behavioral theory to interpret neural signals; without a functional context, hardware descriptions remain inert.
The Granularity Mismatch and Hardware Limits
A significant barrier to the "brain-first" agenda is the "granularity mismatch." There is a fundamental gap between the "primitives" of neurobiology (e.g., a dendritic spine or a local field potential) and the "primitives" of psychology (e.g., a phoneme in linguistics or a belief in social cognition). Poeppel and Adolfi contend that we cannot simply "build up" from one to the other through induction alone. Instead, progress requires intermediate, computationally explicit theories that bridge these levels of analysis.
Furthermore, the authors point out that "hardware is not destiny." They cite the example of the C. elegans connectome—a fully mapped nervous system of 302 neurons. Despite having a perfect "wiring diagram," scientists have yet to fully explain how this hardware produces the worm's complex behaviors. This suggests that a complete description of the implementation level does not automatically yield an understanding of the algorithmic or functional levels.
Toward an Alignment Problem
Rather than viewing psychology as a "pre-scientific" or "folk" discipline to be discarded (eliminativism), Poeppel and Adolfi argue that modern cognitive science has developed highly granular, data-supported units—such as "intonation contours" or "syntactic structures"—that are as scientifically "real" as neurons.
The authors propose that the relationship between brain and mind should be treated as an "alignment problem." In this view, psychological theory and neurobiology are equal partners. Progress is made through Abduction (inference to the best explanation), where formal, algorithmic models of human faculties are created and then tested to see how they might be instantiated in neural hardware. This "mapping" or "alignment" requires both fields to refine their concepts until they meet in the middle.
Conclusion
The "brain-first" agenda, while provocative, may be overly pessimistic about the current "outside-in" approach. Poeppel and Adolfi conclude that while linking brain and behavior is exceptionally difficult, the use of theory to guide brain research is a productive and necessary strategy. The "inside-out" approach is a false dichotomy; true understanding will come from the rigorous alignment of sophisticated psychological models with detailed neurobiological data, rather than the hope that answers will simply emerge from the hardware alone.
Summary updated