The Inner Maps That May Underpin Flexible Thinking

Intelligence tests are good at producing a number and terrible at explaining what that number reflects. One long-running idea is that a big part of flexible reasoning comes down to organization: not how many facts you can hoard, but how well you structure what you know so you can navigate it and draw new conclusions. A recent brain-imaging study puts a concrete mechanism behind that intuition, linking reasoning ability to the quality of the mental maps the brain builds while learning.

The study, published in Cell Reports by Rebekka M. Tenderra and Stephanie Theves of the Max Planck Institutes for Empirical Aesthetics and for Human Cognitive and Brain Sciences in Germany, focused on what researchers call a cognitive map [1]. The term comes originally from the study of physical navigation, but the concept has expanded. A cognitive map is any internal representation that captures how things relate to one another in a structured space, and the hippocampus, a region deep in the brain, is thought to be its main architect. Crucially, such maps let us infer relationships we never directly experienced, the same trick that underlies a lot of abstract reasoning.

Learning a small world

To watch these maps form, the researchers had participants learn a compact virtual environment while lying in a functional MRI scanner [1]. People explored a circular arena and learned where six objects were located within it. Afterward, their grasp of the layout was tested in a few ways: arranging the objects from a bird's-eye perspective they had never actually seen, and estimating distances between items on a sliding scale. A separate task checked plain item memory, simply recognizing the objects, so the team could distinguish knowing the pieces from knowing how they fit together.

The brain data let the researchers ask a precise question. As people learned the arena, did their hippocampal activity come to encode the space in a genuinely map-like way, with the pattern of neural responses reflecting the true spatial relationships among objects? And did that map quality vary with how people scored on fluid intelligence, the capacity to reason through novel problems rather than lean on accumulated knowledge?

Sharper maps in stronger reasoners

The answer was yes. Participants with higher fluid intelligence showed stronger signs of map-like encoding, particularly in the right hippocampus [1]. Their brains, in effect, were building a cleaner, more spatially faithful internal model of the environment they had just learned. Those with lower scores showed representations that were less spatially consistent, a fuzzier internal map.

What makes the result more than a curiosity is a control the researchers built in. The link between fluid intelligence and map quality held up even after accounting for how well people remembered the individual objects [1]. That distinction is the heart of the finding. It was not that smarter participants simply had better raw memory for the items. It was that they organized what they learned into a more structured whole, capturing the relationships between things rather than just the things themselves. Reasoning ability, on this view, is tied less to storage and more to structure.

That fits a growing perspective in cognitive science: much of what we call intelligence may rest on the brain's talent for arranging information into flexible, navigable maps, whether the territory is a room, a family tree, or a web of abstract concepts. A well-built map lets you take mental shortcuts and reach conclusions you were never explicitly taught, which is close to a working definition of flexible thought.

The limits worth keeping in view

As with most brain-imaging work, the caveats are substantial. The design was cross-sectional and correlational, so it can establish that fluid intelligence and map-like encoding go together, but not that one causes the other [1]. It is entirely possible that some third factor shapes both, or that the relationship runs in a loop. Nobody has shown that deliberately building better maps would raise reasoning scores.

The study also examined spatial learning specifically, a physical arena with objects in it. Whether the same map-like advantage extends to purely abstract relationships, like concepts, categories, or social rules, is a reasonable hypothesis but remains untested here [1]. The sample was a fairly homogeneous group of healthy adults, which limits how broadly the pattern can be generalized, and the study left the contribution of the prefrontal cortex, another key player in reasoning, largely unexplored.

Even so, the finding is a satisfying piece of a larger puzzle. It suggests that a familiar, slippery concept, intelligence, may have a tangible signature in how the brain organizes experience into structure. The mind's quiet work of building and refining internal maps turns out to be deeply physical, echoing other research on how the body and brain jointly construct experience, such as why your heartbeat may flag a mistake before you notice it. For more on how the mind organizes and reasons over information, browse more cognitive science coverage.

Sources

1. Tenderra, R. M., & Theves, S. (2025). Human intelligence relates to neural measures of cognitive map formation. Cell Reports. https://doi.org/10.1016/j.celrep.2025.116033
2. National Library of Medicine (PubMed) record. https://pubmed.ncbi.nlm.nih.gov/40711882/

This article summarizes published research for general informational purposes only and does not constitute professional advice.