What a "grid-like code" actually measures
A grid-like code is a pattern of brain activity that rises and falls in a regular, repeating way as a person moves through a space—whether that space is a room or a set of related ideas. Researchers infer it by tracking how similar the brain's response is at evenly spaced positions, so the signal looks like a hexagonal grid laid over the space being navigated [1][2]. This is a measurement of a repeating pattern, not a literal picture of a map inside the head.
The key point for this question is that the same kind of repeating signal can appear in different brain regions and different tasks. That makes it a candidate for a general mechanism, but it does not by itself prove that one mechanism is doing all the work.
Navigating meaning uses more than one kind of spatial reference
When people learn the meanings of words and then mentally search for a target concept, brain activity shows a grid-like signal in regions such as entorhinal, prefrontal, and parietal cortex [1][2]. That signal is not static: its strength changes with how close the person is to the goal in the conceptual space, which is what you would expect if the code helps track progress toward a target [2].
But meaning navigation also recruits a second, complementary representation. During goal-directed mental search, the parietal cortex carries egocentric signals—codes that describe where a target sits relative to oneself, like "to my left" or "to my right"—alongside the allocentric, map-like grid signal [2]. The authors suggest these two reference frames work in parallel so the brain can read out where a remembered concept is stored in a flexible way [2]. That is a proposal about function, not a settled finding, but it already argues against a single rigid map: the system appears to combine a world-centered grid with a self-centered direction signal depending on the task.
Smell navigation shows grid-like responses tied to the odor task
In olfactory navigation, people were asked to move from a start odor toward an imagined end odor and then make a decision [3]. During this, grid-like neural responses appeared in olfactory-related brain regions [3]. This is direct evidence that the grid-like pattern is not confined to visual or verbal spaces; it also shows up when the "space" is defined by smells.
What stands out is that the response is tied to the specific demands of following odors—localizing a source and deciding where to go next—rather than to a generic act of thinking about a map [3]. The abstract-level evidence does not let us say the smell task and the meaning task use identical computations, only that both engage grid-like coding in regions suited to their own kind of navigation.
Why the balance of evidence favors a shared toolkit over one fixed map
Putting the findings together, the common thread is a repeating grid-like signal that can be recruited across domains, but the surrounding machinery differs by task. In meaning navigation, the grid signal is accompanied by egocentric direction codes and is modulated by goal proximity [2]; in smell navigation, the grid-like response is embedded in odor-guided route decisions [3]. If one identical map mechanism were running everywhere, we would expect the same supporting codes and the same modulation across tasks, and that is not what these studies describe.
A reasonable reading is that the brain reuses a general spatial coding strategy—periodic grid-like activity—as a building block, then adapts it with task-specific reference frames and goal signals. The evidence is consistent with that view, but it comes from separate paradigms in separate domains, and the semantic studies do not test smell while the olfactory study does not test meaning [1][2][3]. So the honest answer is that the data favor flexible, task-tuned use of a shared code, while leaving open whether the underlying computation is truly the same one.
Sources used in this answer
1
Multiple spatial codes for navigating 2-D semantic spaces
fMRI evidence shows a grid-like code in the brain while people navigate a learned two-dimensional semantic space of word meanings.
2
Mental search of concepts is supported by egocentric vector representations and restructured grid maps
During goal-directed mental search in conceptual space, parietal cortex shows egocentric vector representations alongside a grid-like signal whose strength is modulated by goal proximity.
3
Neurobehavioral Mechanisms of Human Olfactory Navigation
Humans navigating from a start odor toward an imagined end odor show grid-like neural responses in olfactory-related brain regions.
Related research questions
