A Map Should Say What It Cannot Tell You

Maps are persuasive partly because they remove friction. A table asks the reader to compare rows, and a methodological note asks them to follow qualifications, whereas a map compresses the result into colour, position and shape. The pattern arrives before the caveat.

That is useful. It is also why a map should say what it cannot tell you.

A limitation is not an apology attached to a weak map; it is part of the claim, telling the reader where the evidence stops.

Every map invites conclusions beyond its literal contents

A map of reported disease cases may be read as a map of disease incidence. A map of health facilities may be read as a map of access. A map of protected areas may be read as a map of effective conservation. None of those inferences is absurd. They are simply stronger than what the visible features alone establish.

The problem is that maps rarely announce the gap between the measured thing and the inferred thing.

A precise point does not reveal whether the source knew the location precisely, just as a filled district does not reveal variation inside it. Nor does a smooth interpolated surface reveal where observations were sparse, or a current basemap reveal that the thematic layer is ten years old.

Without an explicit boundary around the claim, visual completeness is easily mistaken for evidential completeness.

The most useful limitation changes interpretation

Weak limitation notes are generic: “Data may contain errors.” “For informational purposes only.” “Coverage may vary.” They protect nobody because they do not tell the reader what to do differently.

A useful limitation is specific enough to alter interpretation.

“Facilities shown are those reported by participating agencies” tells the reader that absence from the map is not proof of absence on the ground.

“Travel times exclude seasonal road closures” identifies a mechanism that could make accessibility worse than shown.

“District rates should not be interpreted as neighbourhood risk” establishes the scale at which the result stops being meaningful.

“Locations are geocoded to settlement centroids, not facility entrances” prevents false precision.

Each statement narrows the claim. That is exactly what a good limitation should do.

Not every uncertainty can be turned into a number

Some limitations are quantifiable: positional uncertainty may be expressed in metres, a survey estimate may have a confidence interval, and a model may provide an error range.

Others are structural. Reporting may be incomplete in unknown ways. A boundary may be politically disputed. A classification may depend on a contested definition. A data source may omit informal activity that was never measured.

Forcing those issues into a single confidence score can create another form of false precision.

Plain language remains one of the strongest cartographic tools. A short sentence can often communicate a boundary of knowledge more honestly than a numeric badge whose meaning few readers understand.

Placement is part of the ethics

A limitation hidden in a metadata page technically exists but may not function editorially.

The map's strongest claims are visible immediately. If the qualification requires three clicks, a download and a methods appendix, the information is not being given comparable weight.

That does not mean every map needs a wall of warnings. Instead, the most consequential limitation can appear close to the relevant claim—in the subtitle, legend, popup, source note or a short publication note—while fuller documentation remains available elsewhere.

The test is practical: could a reasonable reader encounter the visual conclusion without encountering the qualification that materially changes it?

If yes, the map is asking the caveat to do work it has been designed not to do.

Limitations can improve the map itself

Writing the limitation before publication often exposes a deeper problem.

If the note becomes “population counts are shown without accounting for population size”, perhaps the map needs rates instead. If it says “areas with no observations are indistinguishable from zero”, the styling should change. If it says “the facility list is incomplete and coverage varies by district”, the title may need to describe reported facilities rather than facilities generally.

In that sense, limitation writing is not merely documentation, but a form of editorial QA.

The act of stating what the map cannot tell you forces the author to ask whether the map is claiming too much in the first place.

A trustworthy map has visible edges

The ideal map is not one that appears to contain the whole world. It is one that helps the reader understand the world through a bounded piece of evidence.

That requires confidence where the evidence is strong and restraint where it is not.

The most credible publications often feel slightly less complete because they preserve distinctions that a cleaner visual could hide: observed versus modelled, zero versus missing, current versus historical, precise versus approximate, available versus accessible.

Those distinctions do not weaken the map; they tell the reader what kind of knowledge it actually offers.

A map should say what it cannot tell you because otherwise its silence becomes part of the argument.