Do hierarchical organizations in cognitive maps produce systematic distortions in distance and direction judgments?

Hierarchy makes some spatial judgments harder, not all of them

The most direct test comes from a virtual environment built as a building inside a courtyard, with objects scattered through both. Participants memorized the layout, then answered questions about spatial relationships while being scanned. Trials that required integrating information across the two subspaces produced less accurate answers and longer response times than trials that stayed within one subspace [1]. That pattern is the signature of a mind that has split the world into parts and has to do extra work to stitch them back together.

This is a cost, not a uniform warp. The study measured accuracy and speed on memory questions, so it tells us that cross-boundary reasoning is effortful and error-prone; it does not by itself say that every distance is overestimated or every direction bent by a fixed amount [1]. The brain data point the same way: differences between integration and non-integration trials appeared in scene-responsive and medial temporal regions, and those differences tracked the behavioral cost in the retrosplenial complex, occipital place area, and hippocampus [1]. Multivoxel pattern analyses further suggested that these regions carry representations reflecting the nested structure of the space [1].

Direction errors and stretched distances come from different sources

Directional judgments are where hierarchical structure was first tied to systematic error. Hirtle and Jonides reported that people make systematic mistakes when judging directions, and presented this as evidence that cognitive maps are organized hierarchically [2]. The abstract does not specify the size or direction of those errors, so the safe reading is that hierarchy is associated with predictable directional bias, not that any particular angle is misjudged by a known amount.

Distance estimates can be distorted too, but the demonstrated causes are features of the route or the terrain rather than nesting. In sketch maps of Northern Europe, cities separated by the sea were placed farther apart than they should be; on a university campus, landmarks separated by more turns along a route were likewise overestimated [3]. These are metric errors in an external drawing, which the authors treat as a window onto the underlying mental representation [3]. The important point for the hierarchy question is that these distortions arose without any manipulation of nested regions, so they show that distance bias is real and systematic while leaving open how much hierarchy contributes.

What you were doing at a place changes how far away it feels

A separate line of work argues that distance estimates shift with the context in which a place is retrieved. In a series of interview experiments, participants rated landmarks by how often they visited them or by how important the activity there was, then gave verbal distance estimates and route descriptions; the results were read as evidence that action representations get pulled into the judgment [4]. In further experiments, blindfolded participants learned routes from verbal descriptions timed to a metronome and were physically moved through turns, and their later recall was consistent with network-map style explanations involving attention and motor programs [4].

This matters for the hierarchy question because it shows distance judgments are not a single stable readout. If the same pair of places can be judged farther or nearer depending on what you did there, then any hierarchical distortion has to be measured against a baseline that is itself flexible [4]. The thesis work does not test nested spaces, so it cannot confirm or rule out a hierarchy effect; it widens the list of things that systematically bend distance.

The hierarchy explanation is plausible but not yet pinned down

Putting the strands together: people do seem to represent nested environments as separate subspaces that must be actively integrated, and that integration is measurably harder [1]. Hierarchical organization has also been linked to systematic directional errors [2]. But the distance distortions that have been cleanly demonstrated come from gaps and turns in the layout [3] or from the action context of a place [4], not from crossing a region boundary. No supplied study directly compares distance estimates for two pairs of landmarks that are equally far apart but differ in whether a region boundary lies between them.

So the honest answer is a qualified yes. Hierarchy produces a reliable processing cost at boundaries and has been associated with directional bias, while metric distortion is well established but attributed to other causes in the studies available here [1][2][3]. Whether crossing into a new subspace stretches perceived distance the way a sea or a winding route does remains an open question that the current evidence suggests but does not settle.

Sources used in this answer

Related research questions