Appearance
Historical GIS (HGIS) and modern GIS run on the same software — QGIS, ArcGIS, PostGIS — but HGIS is the discipline of applying GIS to the past, where data is fragmentary, geometry is uncertain, place names and boundaries change, and time is a first-class dimension rather than an afterthought. Understand the distinction so you choose it deliberately: HGIS earns its substantial time cost only when your question is genuinely spatial and your sources can support a map you would defend. This article lays out the trade-offs and the signals for when to commit and when to walk away.
What actually differs between HGIS and modern GIS?
The software is identical; the data and the demands are not. Modern GIS usually starts from current, surveyed, authoritative layers. HGIS starts from a tithe map, a census book and a trade directory, and has to reconstruct geography that no longer exists.
| Dimension | Modern GIS | Historical GIS (HGIS) |
|---|---|---|
| Source data | surveyed, authoritative | fragmentary, derived, biased |
| Time | usually a snapshot | first-class — validity intervals |
| Boundaries | stable, official | shifting, contested, renamed |
| Coordinates | known CRS, GPS | georeferenced, approximate |
| Uncertainty | minor | central; must be modelled |
| Place names | current gazetteers | historical gazetteers, variants |
The recurring HGIS skills — georeferencing, dating boundaries, modelling uncertainty, resolving historical place names — all flow from that left-to-right shift.
When does HGIS genuinely pay off?
Commit to a real HGIS build when several of these hold:
- The question is spatial: "where", "how far", "what was adjacent", "how did the pattern move".
- Your sources let you place things with defensible accuracy.
- You will reuse the layer across multiple analyses or maps.
- Comparison across time-slices matters (e.g. urban growth, redistricting, epidemic spread).
Cholera deaths by parish, enclosure of common land, migration flows, or the changing reach of a market town are all strong HGIS cases. The spatial structure is the argument.
When should you not use HGIS?
GIS is a heavy tool. Walk away when:
- The question is not spatial — a chronology or prosopography needs a timeline or network, not a map.
- Locations are too uncertain to map without implying false precision. A point map of vaguely-described places lies by looking exact.
- A single static reference map already answers the question — then make one good map, not a database.
- The schedule cannot absorb weeks of georeferencing and digitising for a marginal payoff.
Choosing not to build an HGIS is a legitimate, often wise, methodological decision.
How much does the HGIS path actually cost?
Be honest about effort. A modern thematic map from clean data is an afternoon. The HGIS equivalent multiplies because the data is not map-ready:
text
Modern map: load clean layer -> join table -> style (hours)
HGIS layer: georeference 5 sheets ~1 day
digitise features ~2-4 days
date & clean boundaries ~1-2 days
geocode addresses + manual ~variable
model & document uncertainty ongoing
=> weeks, not hoursThis is not a reason to avoid HGIS — it is a reason to scope it and to reuse authoritative historical datasets (GB Historical GIS, NHGIS) before building from scratch.
What skills must you add beyond modern GIS?
If you already know modern GIS, the gap to HGIS is specific and learnable: georeferencing historical maps; modelling time with validity intervals; handling administrative units that split and merge; representing positional and dating uncertainty as explicit attributes; and the source criticism to judge what a fragmentary record will and will not bear. Source criticism is the part no software supplies.
A decision rule you can apply today
Ask three questions. Is the core question spatial? Can the sources support a map you would defend in print? Will the layer be reused or compared across time? Three yeses: build the HGIS. One or no: make a single careful map, or pick a non-spatial method. That rule alone prevents most wasted GIS effort in history projects.
Key Takeaways
- HGIS and modern GIS share software; HGIS differs in data, uncertainty and time-handling, not tools.
- HGIS treats time and uncertainty as first-class; modern GIS usually assumes authoritative current data.
- Commit to HGIS when the question is spatial, sources support a defensible map, and the layer will be reused.
- Avoid HGIS for non-spatial questions, hopelessly uncertain locations, or when one static map suffices.
- Expect weeks, not hours — the data is not map-ready; reuse authoritative historical datasets first.
- The added skills are georeferencing, temporal modelling, boundary change, uncertainty and source criticism.
- Apply the three-question decision rule before committing.
Frequently Asked Questions
What is the difference between HGIS and modern GIS?
They use the same software; HGIS (Historical GIS) is the application of GIS to past places and times, where the data is fragmentary, the geometry uncertain, and time is a first-class dimension. Modern GIS typically assumes current, surveyed, authoritative data.
Is HGIS just modern GIS with old maps?
No. The tools overlap but HGIS adds requirements modern GIS rarely needs: explicit temporal validity, modelled uncertainty, changing administrative units, and reconstruction from incomplete sources. The mindset, not the software, is what differs.
When should I NOT use HGIS for a history project?
Skip a full HGIS build when your question is not spatial, when locations are too uncertain to map honestly, or when a single static reference map answers the question. GIS is a serious time cost; do not adopt it for decoration.
Do HGIS and modern GIS use the same software?
Yes — QGIS, ArcGIS Pro, PostGIS and the same formats serve both. The difference is in data modelling: temporal fields, uncertainty attributes, and historical CRS handling rather than any special HGIS application.
What extra skills does HGIS need beyond modern GIS?
Georeferencing historical maps, modelling time and validity intervals, handling changing boundaries, representing uncertainty, and reconstructing geography from textual and cartographic sources. Source criticism matters as much as software skill.
How long does a small HGIS project take versus a modern map?
A modern thematic map from clean data can be an afternoon. A small HGIS layer — georeferencing a few sheets, digitising, dating boundaries, geocoding — is realistically weeks, because the data does not arrive ready to map.