Roman Empire, AD 117

What you are looking at

A single polygon showing the territory under Roman control at the death of Trajan in AD 117, the largest extent the empire is recorded as having held. On this atlas it does one job above all others: it is the frame the Dacia material is read against. The province whose names, roads and frontiers the Corpus Chartarum Daciae reconstructs is the northern salient of this shape.

How to read this layer

Read the edge as a claim about reach, not about a border. Rome administered provinces, garrisoned roads and collected tribute; it did not maintain a surveyed boundary line of the kind a nineteenth-century treaty produces. The outline here is a modern cartographic convenience, drawn so that an extent can be compared with other extents. Where it runs through desert, steppe or mountain it is at its least meaningful, because control in those places was a function of routes and seasons rather than of ground.

The layer is a fill, so overlapping it with the modern boundary layer will show you two incompatible ideas of a border at once. That comparison is worth making deliberately and worth not mistaking for agreement.

Historical scope

The layer is registered for AD 106–271, which is not the life of the AD 117 extent. Those dates are the life of the province of Dacia: 106, when Trajan completed its conquest, and 271, the conventional date for its abandonment under Aurelian. The geometry is a single moment; the envelope is the window in which that moment is useful context for this corpus. An atlas that showed the AD 117 shape across the whole imperial period would be asserting a stability that the second and third centuries did not have.

Sources and evidence

The geometry is the Ancient World Mapping Center’s roman_empire_ce_117_extent, produced at the University of North Carolina at Chapel Hill and derived from the Barrington Atlas of the Greek and Roman World together with OpenStreetMap base data. It is redistributed here under ODbL 1.0.

This project has not redrawn or adjusted the geometry. What it has done is decide where the layer sits in the Atlas taxonomy - historical, category territories-boundaries - and that decision is the claim it makes here.

Reconstruction and uncertainty

Three uncertainties travel with this polygon and none of them are visible in it.

The eastern conquests of 114–117 - Armenia, Mesopotamia, Assyria - were held briefly and abandoned almost immediately by Hadrian. A map of AD 117 includes them; a map of AD 120 does not. The extent shown here is therefore the single most short-lived configuration the empire ever had, which is exactly why it is the one usually drawn.

The frontier zones were gradients. In North Africa, on the Rhine and Danube, and along the desert margins, the transition from provincial administration to client relationship to no relationship at all happened over distances the line cannot express.

The client kingdoms are in or out by editorial convention rather than by evidence, and different atlases place them differently.

Editorial decisions

The layer is classified as historical rather than context. It would be tempting to treat it as neutral background - it is used as background - but it is a reconstruction of a past state and carries a scholarly argument, and the taxonomy reserves context for material that asserts nothing about the past.

It is not a member of any collection’s default composition except Roman geography, where it is the whole point. In the Dacia collection it is available and off, because the province is the subject there and the empire is the frame.

Data and downloads

The asset is served as GeoJSON from the content-addressed geo release, so a given URL always returns the same bytes. Attribution must travel with any reuse: Ancient World Mapping Center; derived from the Barrington Atlas and OpenStreetMap (ODbL).