When you think about a social hierarchy ("who is above whom?"), a career path ("moving forward"), or a value comparison ("this option is close to that one"), you are using spatial metaphors. This is not just a linguistic quirk. A growing body of neuroscientific evidence suggests that the brain literally maps abstract, non-spatial information using the exact same neural machinery it evolved for physical navigation.
In a comprehensive review, Wenyan WU (The cognitive map and its intrinsic mechanisms) synthesizes how the "cognitive map"—originally discovered as a mental representation of physical space—serves as a general-purpose framework for relational reasoning across diverse domains. Instead of building separate, specialized modules for every type of abstract knowledge, the brain projects non-spatial dimensions onto low-dimensional geometric coordinate systems.
Constructing Concept Spaces
How does the brain turn an abstract concept into a map? It extracts the underlying relational dimensions of an experience and assigns them coordinates. For instance, Stephanie Theves et al. demonstrated in The Hippocampus Maps Concept Space, Not Feature Space that the hippocampus selectively organizes information based on dimensions relevant to a learned concept, rather than mapping every physical feature of an object. In this concept space, the relative distances between entities directly reflect their conceptual similarity.
This mapping capability allows humans to navigate entirely artificial spaces. Doerte Kuhrt (An immersive first-person navigation task for abstract knowledge acquisition) showed that humans can learn to navigate abstract "quantity spaces" using the same cognitive mechanisms and exhibiting the same behavioral biases found in real-world navigation.
Even complex, multi-dimensional emotional states can be mapped onto continuous cortical gradients. Giada Lettieri et al. (Emotionotopy in the human right temporo-parietal cortex) revealed that the right temporo-parietal cortex organizes affective experiences into an "emotionotopy" defined by overlapping spatial gradients of polarity, complexity, and intensity.
Grid Codes for Abstract Navigation
In physical navigation, grid cells in the entorhinal cortex fire in a strikingly regular hexagonal pattern to provide a metric coordinate system for space. Why hexagonal? Computational models by Fabio Anselmi et al. (A computational model for grid maps in neural populations) and Ben Sorscher et al. (A unified theory for the origin of grid cells through the lens of pattern formation) demonstrate that hexagonal symmetry is the mathematically optimal way to encode continuous information with minimal variance under basic biological constraints, such as non-negative firing rates.
Remarkably, this optimal hexagonal code is recruited when we navigate non-spatial decisions. For example, Alexander Nitsch et al. (Grid-like entorhinal representation of an abstract value space during prospective decision making) discovered that the entorhinal cortex utilizes grid-like neural codes to map trajectories through a two-dimensional abstract value space during decision-making.
Similarly, Irina Barnaveli et al. (Hippocampal-entorhinal cognitive maps and cortical motor system represent action plans and their outcomes) showed that the entorhinal cortex deploys hexadirectional grid-like signals to organize abstract action-outcome relationships into a structured relational space. To capture this computationally, Haruo Hosoya (A Cognitive Model for Learning Abstract Relational Structures from Memory-based Decision-Making Tasks) developed a model showing how these hexagonal activity patterns naturally manifest when extracting relational structures like hierarchies from non-spatial tasks.
Temporal Sequencing and Phase Precession
Navigation is not just about knowing where you are in a static space; it is about tracking when events occur along a path. In the spatial domain, the hippocampus uses "phase precession"—where neurons fire at progressively earlier points relative to the local theta oscillation cycle—to compress sequential spatial positions into rapid neural sequences.
The same mechanism structures non-spatial sequences. Leila Reddy et al. (Theta-phase dependent neuronal coding during sequence learning in human single neurons) recorded from single neurons in human patients and found that theta-phase precession organizes abstract sequences of learned images, demonstrating that the hippocampus acts as a general-purpose generator of cognitive maps for temporal sequences.
This temporal compression allows the brain to multiplex multiple types of information. Babak Shahbaba et al. (Hippocampal ensembles represent sequential relationships among an extended sequence of nonspatial events) showed that during non-spatial odor-sequence tasks, hippocampal ensembles compress information about event identity, timing, and correctness into rapid theta cycles. This sequential organization is critical for resolving ambiguity; Kara L. Agster et al. (The Hippocampus and Disambiguation of Overlapping Sequences) demonstrated that a functional hippocampus is strictly required to disambiguate overlapping non-spatial sequences that share common middle elements.
Generalization and Geometric Distortions
Because abstract knowledge is mapped onto a geometric framework, it inherits both the strengths and the vulnerabilities of physical spatial maps.
On the positive side, geometric maps enable powerful, zero-shot inference. As Ila R Fiete (Mental navigation in the primate entorhinal cortex) notes, cognitive maps in the entorhinal cortex support zero-shot generalization to novel paths between landmarks, proving that the underlying representation is a structured relational model rather than a collection of isolated associations. This allows the brain to map complex, branching storylines; Branka Milivojevic et al. (Coding of Event Nodes and Narrative Context in the Hippocampus) found that the hippocampus acts as a high-level narrative mapper, separating overlapping storylines into distinct neural representations while tracking shared elements.
However, this spatial reuse also means that abstract memories are subject to geometric distortions. In physical space, Jacob L. S. Bellmund et al. (Deforming the metric of cognitive maps distorts memory) demonstrated that environmental boundaries (like trapezoidal walls) physically warp the neural metric of space, causing predictable errors in distance perception and memory. A major open question in the field is whether similar boundary constraints and geometric warping distort our mental maps of social hierarchies, value spaces, and abstract conceptual domains.
Key Foundational Work
Coding of Event Nodes and Narrative Context in the Hippocampus (2016) — contributed 1 concept elements to this direction.
The Hippocampus and Disambiguation of Overlapping Sequences (2002) — contributed 1 concept elements to this direction.