Research

Vernacular Geography

Vernacular geography is concerned with the geography as used by you and me in everyday language. Place names such as "the city centre", "South of France", "around the shops" or specific names such as Lakeside in Cardiff, Wienerwald around Vienna or Karst in the South of Slovenia are examples for vernacular geography. Up to date these names can not be found in gazetteers, i.e. catalogues describing geographic names, their type and location. The focus of our research here in Cardiff is the representation of such places for the purpose of geographic information retrieval. In an Ordnance Survey funded project we create methods to capture, manage, analyse and describe vernacular geography for the use in search engines, decision support systems and any kind of services that deal with geographic information.

Children and Space

The influental work of Jean Paul Piaget motivated a number of researchers to investigate childrens conceptualization of the world. Recent work in developmental psychology suggests that children build theories in a way scientists do (see Meltzoff and Gopnik 1997). We can assume that the concepts found in children are less spoiled by other metacognitve processes, such as language and logical thinking and that research in developmental psychology gives pointers to the basic building blocks/elements that can be found in adults reasoning about space and geography.

I am specifically interested in how people booststrap knowledge about space. When and how do they build concepts and what drives our learning. Observations of a continous reality lead to the definition of (discrete) concepts related to space. Context, experience and cognitive processes drive the distinction of concepts leading to finer granuled conceptualizations of the surrounding environment.

Formal Descriptions

Formal descriptions help to understand complexity in research. Complexity is the interaction of simple parts. My task as a researcher is to identify simple parts and then study their interaction. If a research problem can't be solved over many years then because it has not been decomposed well enough or the wrong abstraction steps have been taken in the study of interaction.

Formal descriptions can help to come up with a viable model of how people organize and cognize their (spatial) environments. I recently used agent based modeling approaches and algebraic specifications. Currently I focus on probability theory and statistical approaches to model vernacular geography.