Theory Building (aaag-theory-building)

An Annals paper is judged on the geographic argument, not only the result. The contribution must do conceptual work with geography's own apparatus — space, place, scale, region, territory, landscape, environment-society relations, spatial process — and be portable enough that other geographers can use it.

When to trigger

• Moving from "we found X" to "X means Y for how geographers understand Z"
• A reviewer said the paper is "descriptive," "atheoretical," or "a case study without a concept"
• Deciding which geographic concept the argument advances or revises

Build the argument with a geographic concept

Pick the concept the paper actually moves and make it load-bearing:

• Scale — the result holds/changes across scales; the argument is about cross-scale processes or the MAUP, not a single resolution.
• Place / region — what is specific vs. generalizable about the setting; what the case is a case of.
• Space-time / spatial process — diffusion, dependence, clustering, mobility as the mechanism, not a backdrop.
• Environment-society — the coupling itself is the object; neither side is mere context.
• Spatial data / representation — for Methods, the conceptual stakes of how space is measured and modeled.

Theory across the four areas (judged on its own terms)

The portability test (Annals-specific)

State the contribution as a sentence another geographer could import: "___ is best understood as ___ rather than ___, which means [a different subfield] should expect ___." If the contribution cannot be lifted out of your case, it is a finding, not yet a geographic argument.

From observation to claim

• Name the mechanism / concept, not just the pattern ("clustering" is a pattern; "why and at what scale the process clusters" is the argument).
• State scope conditions: where the argument holds and where it should break — geography rewards explicit scale and place boundaries.
• Distinguish your claim from the nearest rival concept (link to aaag-literature-positioning).

Worked micro-example (illustrative)

A finding: "flood-relief funds reached richer neighborhoods faster in City X."

Pattern only: "relief was spatially unequal" (a map, not a concept)
Geographic argument: relief speed is governed by administrative SCALE mismatch — funds routed through
  units that do not match the geography of need, so the same policy produces uneven relief by design
Mechanism: scalar mismatch between funding units and exposure units
Scope conditions: holds where exposure crosses administrative boundaries; breaks where they coincide
Portability: any geographer studying service delivery can ask whether the governing unit matches the
  geography of the problem — the concept lifts out of floods

The pattern becomes a portable claim about scalar mismatch, not a description of one city's floods.

Checklist

• The argument names a geographic concept and does conceptual work with it
• Mechanism stated, not just pattern; scope conditions explicit
• Portability sentence written (another subfield could import it)
• Theory matched to the area's tradition and judged on its own terms
• The strongest rival concept is named and distinguished

Anti-patterns

• Descriptive mapping with no concept ("here is the spatial distribution of X")
• Borrowing a non-geographic theory wholesale with no spatial/place/scale work
• A Nature-Society argument that theorizes only the social or only the biophysical side
• Claiming generality from one case with no scope conditions

Output format

【Geographic concept】scale / place / space-time / environment-society / representation
【Mechanism】the process the argument turns on
【Scope conditions】where it holds / breaks
【Portability】the import sentence
【Next】aaag-research-design

Supplementary resources

• ../../resources/exemplars/library.md — papers whose contribution is a portable geographic concept
• ../../resources/official-source-map.md — area definitions and scope