Disambiguating geographical names in historical British censuses and travel writing

Historical texts and statistical reports almost always contain geographical names or toponyms, not co-ordinates, so mapping requires linkage to some form of gazetteer. The historian must decide where places are, what location to associate with a particular toponym, and whether features with the same name represent the same place. This paper presents two complementary case studies.

Firstly, disambiguating parish names appearing since 1801 in UK census reports. A novel analysis of English parish names shows that most were ambiguous, meaning that two or more parishes had the same or very similar names. In most cases, this ambiguity is removed by also specifying county, but complete disambiguation requires that intermediate ‘districts’ also be specified. As the systems of counties and districts have changed substantially over the last 200 years, comprehensive disambiguation in turn required matching to a poly-hierarchic Administrative Unit Ontology (AUO), within which particular parishes maintain their identity while their names and hierarchic position evolved.

Secondly, identifying toponyms in a uniquely large collection covering over three hundred years of historical travel writing with a gazetteer of ‘places’ constructed partly from nineteenth century descriptive gazetteers. Here the individual place references lack hierarchical context, and travellers often inaccurately reported toponyms, so instead the traveller’s route must be followed to identify the sequence of places visited.
Rather than link sources to existing gazetteers, the AUO and the ‘places’ gazetteer were constructed in parallel with the toponym matching, systematically adding variant toponyms from the sources. The final section describes recent work to extend both gazetteers to include Ireland, drawing extensively on census reports, travel writing and also two nineteenth century gazetteers. The end result is a uniquely rich data structure, integrating diverse historical sources around place identifiers which combined to sit behind the website ‘A Vision of Britain through Time’.