In the brain's sports league, neurons are the players, synapses are the plays, and for about a century neuroscience has mostly studied the team during weird indoor drills with no crowd, no weather, and a coach yelling, "Press this button when the dot appears." Useful? Absolutely. But the new Simons Collaboration on Ecological Neuroscience, or SCENE, is asking a very pub-quiz kind of question: what if the brain's best moves only show up when the game is actually being played?
That is the core idea behind Angelaki and colleagues' new Neuron article introducing SCENE, a 10-year collaborative effort to study how brains interact with the world, not just with simplified lab tasks wearing tiny fake mustaches (DOI: 10.1016/j.neuron.2026.04.036). This is a roadmap, not a "we found one magic brain blob" paper.
The Answer Is Not "More Dots On Screens"
Classic neuroscience loves control. Fair enough. To understand a circuit, you simplify the situation until the brain has fewer places to hide. Show a flash. Play a tone. Track a lever press. Repeat until the graduate student starts seeing stimuli in their dreams.
SCENE argues that something gets lost when the world is shaved down that far. Real behavior is messy because the world offers possibilities. A chair is a place to sit, a thing to climb on, or the object you awkwardly inspect when you enter the wrong conference room. Ecological psychology calls these action possibilities "affordances," a term associated with James J. Gibson. In plain English: your brain asks not only "What is that?" but "What can I do with it?"
A doorway means one thing to a running adult, another to a crawling infant, and something else entirely to a person carrying a sofa. Same doorway. Different body. Different goals. Different brain calculations. The environment is not just wallpaper behind cognition. It is part of the puzzle board.
Brains Are Not Floating Quiz Machines
SCENE wants to combine neural recordings, behavior tracking, computational models, virtual reality, cross-species experiments, and machine learning to test how brains represent affordances. Not vibes. Formal hypotheses.
Researchers have called for "human computational ethology," using cameras, wearables, and models to quantify real-world behavior without turning people into button-pressing houseplants (Mobbs et al., 2021, PMCID: PMC8769712). Others argue that neuroscience should invest in species and behaviors where natural variation is the point, not an inconvenience to be swept under the lab rug (Jourjine and Hoekstra, 2021, DOI: 10.1016/j.neuron.2021.02.002).
Decision neuroscience has been nudging this way too, formalizing planning and information search in tasks that look more like what animals face when they move, choose, and gamble on incomplete information (Hunt et al., 2021, DOI: 10.1038/s41593-021-00866-w).
Why This Is More Than Fancy Lab Decor
The promise is not simply "make experiments look more like real life." That would be like adding pub wallpaper to a math exam and calling it social science. The deeper point is that perception, cognition, and action may be inseparable in ways old tasks miss.
Consider navigation. A bat, a mouse, a human, and an AI agent do not face the same world, even if they occupy the same physical space. The bat hears surfaces. The mouse smells trails. The human sees a shortcut and then spends seven minutes overthinking whether it is socially acceptable to cut through a hotel lobby. Each agent has different sensors, bodies, goals, and affordances. SCENE wants to find general principles across those differences.
That matters for artificial intelligence too. Today's AI systems can recognize objects with terrifying competence, then fail at commonsense interaction like a tourist trying to open a pull door by pushing it with moral conviction. If brains encode the world partly in terms of usable possibilities, better theories of affordance could help machines understand action, not just labels.
The Catch, Because Science Is Rude Like That
Naturalistic neuroscience is harder than tidy lab neuroscience. The data are bigger. The behavior is less repeatable. The statistics get spicy. If a mouse runs, pauses, sniffs, turns, rears, and then changes its mind, what "behavior" was that? One move? Five moves? A tiny rodent side quest?
Work on naturalistic animal behavior emphasizes that behavior unfolds across multiple timescales, with structure, variability, and context all tangled together (Mazzucato, 2022, PMCID: PMC9259028). A pause is not always just a pause. Sometimes it is the mouse version of "wait, why did I come into this room?"
So SCENE's bet is ambitious: keep the rigor, bring back the world. Study the brain as an organ for doing, not merely detecting.
If SCENE succeeds, the payoff could be a richer map of intelligence across animals, humans, and machines. Not just where signals go, but why they matter for an organism trying to move through a world full of doors, cliffs, snacks, rivals, tools, shortcuts, and chairs that may or may not be safe to stand on during trivia night.
References
Angelaki DE, Batista A, Fitzgerald T, Kominsky JF, Lengyel M, Mathis A, Mathis MW, Moss CF, Niell CM, Noel JP, Pitkow X, Rothkopf CA, Savin C, Stachenfeld K, Suthana N, Tolias A, Ulanovsky N, Wolpert DM, Wong A, Zimmermann J. The Simons Collaboration on Ecological Neuroscience: Studying how the brain interacts with the world. Neuron. 2026. https://doi.org/10.1016/j.neuron.2026.04.036
Mobbs D, Wise T, Suthana N, Guzman N, Kriegeskorte N, Leibo JZ. Promises and challenges of human computational ethology. Neuron. 2021;109(14):2224-2238. https://doi.org/10.1016/j.neuron.2021.05.021
Jourjine N, Hoekstra HE. Expanding evolutionary neuroscience: insights from comparing variation in behavior. Neuron. 2021;109(7):1084-1099. https://doi.org/10.1016/j.neuron.2021.02.002
Hunt LT, Daw ND, Kaanders P, MacIver MA, Mugan U, Procyk E, Redish AD, Russo E, Scholl J, Stachenfeld K, Wilson CRE, Kolling N. Formalizing planning and information search in naturalistic decision-making. Nature Neuroscience. 2021;24(8):1051-1064. https://doi.org/10.1038/s41593-021-00866-w
Mazzucato L. Neural mechanisms underlying the temporal organization of naturalistic animal behavior. eLife. 2022;11:e76577. https://doi.org/10.7554/eLife.76577
Disclaimer: The image accompanying this article is for illustrative purposes only and does not depict actual experimental results, data, or biological mechanisms.