Research: Vernacular Geography

People's sense of place

What is it?

Put simply, vernacular geography is the sense of place that is reflected in ordinary people's language. This is the way that people talk about their world and say where things are within it – the unofficial names and extents of landmarks, streets, neighbourhoods, regions and other geographic features.  Often these places don't have clear, rigid boundaries. Sometimes the same name refers to more than one feature, and sometimes local people use more than one name for the same feature.

Why research it?

Maps, gazetteers and digital geographic data tend to reflect the relatively easy stuff – official names, well-defined administrative and physical boundaries; things that everyone agrees on. This is partly because traditionally it has been expensive and difficult to collect and represent vernacular geographic information.

Through our research in this area, we hope to change this. Ultimately, we would like to be able to efficiently and reliably define, predict, capture, model, evaluate, quality assess, describe and derive products from vernacular geography data.

The practical challenge for vernacular geography research is not merely to understand the names and place concepts that people use, since people seem to manage that by themselves. Ordnance Survey needs to work out how to collect and reflect that understanding   digitally. That way, a non-local person or system (for example, an emergency services call‑centre operator or a web search engine) can interpret or find out what local people might say when they describe where something is.

What's the scope of the work?

In thinking about vernacular geography, it also helps to research the human cognitive aspects of it: how do people understand place and represent their 'sense' of it in their own minds?

This isn't simple. When people describe space they freely use not only area (place) names, which have been the main focus of previous research, but also landmarks, streets, open spaces, water bodies, landforms, fields, woods and many other geographic features. Again, they do not necessarily use the official or current names for those features. The topological relations among the features are also key to understanding people's language descriptions: what they mean by 'by', 'on', 'in', 'opposite', 'at', 'on the way out of' and 'near', among many others. Some named places may be related to each other in a hierarchy (for example, a neighbourhood within a town), but others may overlap, be loosely equivalent to each other, or simply belong to different ways of thinking about a place altogether.

Therefore, we would ideally avoid restricting the scope of vernacular geography merely to area names and extents, which have been the main focus of most research in this area. However, in the short term we might be able to work more quickly towards producing, say, a better gazetteer of place names by focusing particularly on that aspect, so some strands of the research we are doing and sponsoring are more specifically focused than others.

So what's going on?

Some of the things we're currently doing include:

• Sponsoring research at the Universities of Cardiff and Sheffield into the potential for gathering vernacular geography data from the Web, both by automatic means and by surveying people's views about place.
• Reviewing past research, which has taken place in a very wide range of disciplines and over many years, that can help us to understand vernacular geography phenomena, and give us clues as to how to predict and capture them.
• Discovering what some of our customers have done to tackle vernacular geography issues that arise in their work, and finding out their needs and concerns about them.
• Gathering and analysing vernacular data from various sources to try to assess the common task, personal and spatial factors that may predict where and when particular types of names are used by people when describing place, and also where and when the extents of places tend to be vague rather than precise.
• Bringing together the various types of place and name that are already represented in our current products and looking into ways of modelling place computationally (to link it with other information that may be stored in databases and geographical information systems).