The Difference Between Showing Data and Explaining Geography

A map can show a pattern without explaining it.

That sounds obvious, yet the distinction is easy to lose because spatial patterns feel explanatory: a cluster appears near a road, high values line a border, or service gaps seem concentrated in one region. The visual arrangement suggests a mechanism before any mechanism has been tested.

Showing geography and explaining geography are different jobs.

Description asks where

Descriptive maps are essential.

They tell us where observations occur, how values vary across areas, where facilities are concentrated and which places look different from their neighbours. That is not “mere visualisation”, because good description can reveal errors, generate hypotheses and identify questions that would otherwise remain invisible in a table.

But a descriptive pattern is still a pattern.

A map of reported incidents cannot, by itself, tell us whether the cluster exists because risk is higher, population is denser, reporting is better, policing is more intensive or several mechanisms overlap.

The map earns the right to say where before it earns the right to say why.

Explanation needs a mechanism

To explain a geographic pattern, we need a relationship that could plausibly produce it.

Why are clinics concentrated along one corridor? Perhaps population is concentrated there, road access matters, historical investment followed an administrative policy, or the dataset covers only one provider network.

The map can help compare these possibilities, but the existence of spatial coincidence does not choose among them.

This is where other evidence enters: temporal sequence, domain knowledge, additional variables, field information, statistical analysis or policy history.

Explanation is a larger claim than visual association.

Maps are especially persuasive when variables share space

Two layers that overlap can feel causally connected.

Deforestation appears near roads, disease near water bodies, high rents near transit, and incidents near borders.

Sometimes the relationship is real, sometimes both variables are driven by a third factor, and sometimes the overlap is expected simply because most people and infrastructure occupy the same dense parts of the landscape.

The visual power of co-location is precisely why the map needs interpretive restraint.

A useful phrase such as “spatially associated with” can preserve the observation without pretending the mechanism has been established.

Explanation can require leaving the map

There is a temptation to keep adding layers until the map explains itself.

Roads, population, income, elevation, administrative boundaries, land use—eventually the screen becomes a stack of possible causes.

At some point, a chart, model, table or qualitative account may be better suited to the question.

Geographic explanation is not obligated to remain cartographic. The map can identify the spatial structure that needs explaining, while another method tests the mechanism.

That is a strength, not a failure of mapping.

Titles reveal which job the map thinks it is doing

Compare:

“Reported cholera cases by district, 2026”
“Why cholera is concentrated in eastern districts”

The first title describes, while the second promises explanation.

If the map only contains case counts, the second title overreaches even if the pattern is striking.

This is one reason titles are analytical, not cosmetic. They define the level of claim the evidence must support.

Good maps can make the boundary explicit

An exploratory map might say: “Reported incidents and road access” and invite investigation. A publication might annotate an observed association while linking to the analysis that tests it. A decision-support map might show the factors explicitly built into a model and distinguish them from contextual layers.

The important thing is to know whether the map is offering evidence, hypothesis or explanation.

Showing data well is already valuable, while explanation becomes trustworthy when the map stops pretending that visual proximity alone has done the causal work.