Place names are the oldest documents in the landscape

Place names are unusually durable. Populations change, languages are replaced, and the names of significant features frequently survive, borrowed by the incoming speakers who no longer understand them.

Which makes toponymy a source of historical information that reaches further back than documents.

The layers in a landscape

In regions with a long history of successive populations, names form strata.

River names are typically the oldest. Large rivers in many regions carry names that cannot be explained from any language currently or recently spoken there, and which are attributed to earlier populations whose language is otherwise unknown. Rivers are prominent, permanent and useful for orientation, which is presumably why their names transfer so readily.

Hills and major landscape features come next in durability. Settlements are less stable. Field names change most readily, though some are surprisingly old.

A single landscape can therefore contain names from three or four different linguistic layers, and identifying which is which is a substantial part of what toponymists do.

What the names record

Vegetation. An enormous number of names describe what grew somewhere — the oak wood, the birch slope, the reedy place. Where the vegetation has since changed entirely, the name is the only evidence that it was ever there, and these names have been used to reconstruct historic woodland distribution.

Animals. Names referencing species that have since been extirpated locally, recording their former presence.

Land use. Names describing what a field was for, how it was worked, or what it produced. Names indicating assarts — land cleared from woodland — mark the historical expansion of cultivation.

Ownership. A personal name attached to a place records who held it, sometimes a thousand years ago, and frequently the only surviving trace of that person.

Boundaries and administration. Names marking meeting places, boundary points and assembly sites.

Events. Occasionally something specific happened, and the name records it.

The mistakes that get baked in

Names are transmitted orally, and they get reanalysed when the original meaning is lost.

A name in a dead language becomes meaningless, and speakers reshape it into something that sounds like words they know. The result looks transparent and is entirely misleading.

This produces a large class of names whose apparent meaning is a folk etymology imposed on an older and unrelated original. Local explanatory stories frequently attach to these, explaining the name through an invented event, which is a genuine folklore genre in its own right.

Tautological names are a related result: a feature name from one language, borrowed and no longer understood, with the equivalent word from the new language appended. There are hills across the world whose names, unpacked, mean hill-hill or river-river-river, each layer added by speakers who did not know the previous one.

Names as claims

Naming is an assertion of relationship, and renaming is an assertion of control.

Colonial administrations renamed extensively, replacing existing names with those of officials, sponsors and distant places. Those names then appeared on maps, in administration and in law, which made them the official reality regardless of what people locally called things.

Restoring earlier names has consequently become a live political matter in many countries. Dual naming, official recognition of Indigenous names, and full replacement have all been used, and the arguments around them are about a great deal more than cartography.

Where a name is restored, it frequently turns out to have been in continuous use locally throughout, which is a useful reminder that official and actual naming can run in parallel for a very long time.

Field names, and why they are disappearing

The smallest-scale names are the most numerous, the most informative and the most vulnerable.

Every field on a traditionally worked farm had a name, known to the household and to neighbours, frequently descriptive and frequently old. They were used daily and rarely written down.

Farm amalgamation, hedgerow removal and mechanisation have eliminated both the fields and the need to refer to them individually. When the last person who used the names dies, they go, and they are usually not recorded anywhere.

Local recording projects in several countries have collected field names before that happened, and the results are among the richest sources available for historic land use.

Reading a map

Anyone can do a version of this. Take a detailed map of somewhere you know, list the names, and look for the elements that recur — the words for hill, valley, water, wood, boundary, farm, and the endings that mark settlement types.

The pattern that emerges is a partial account of who lived there, what the land looked like, what they did with it and where the edges were, written by people who left no other record and who were not writing for anyone.