An Experimental Owl: The domain of knowledge on experimental urbanism

Frameworks for understanding urban experimentation, in their broadest sense, are a simplification of reality (Batty & Torrens, 2005), theories of how real-world systems operate. Yet they are invariably underpinned by an ontological framework, either tacit or explicit, that provides a descriptive structure for our knowledge relating to current experimental practice. The idea of an ontology as a simplified description of reality is supported by the mapping of “typologies”, “concepts”, “relationships”, “structures” and “rules”. At best it can become a formal, and standardised description of these concepts. Ideally such ontological frameworks are developed collaboratively so shared knowledge achieves clarity through common vocabularies and definitions. Indeed, the actual collaborative task of building / diagramming an ontology is as much about constructing a shared conceptual model of knowledge and understanding (Eskridge & Hoffman, 2012) as it is a strategy about organising and handling information. This is increasingly important in the digitisation of experimental urbanism project data, where clarity in meaning and semantics relating to the use of language is necessary to support information sharing and better interoperability between data handling software without the loss of meaning in the data (Sayogo et al, 2014).

We are producing a contemporary ontology of the structure of urban experiments; mapping the relationships between the range of legal structures, organisational governance, internal / external scrutiny, public / private funding and the use of assets; to construct a theoretical framework from contemporary practice. This will be based on a mix of collaborative practitioner reflection with domain experts, data mining open and closed organisational resources, with verification through the experimental urbanism professional network. We aim to identify the specific characteristics of the domain, generate the relationships table, and validate the draft ontology, “(u)tilizing the heuristic approach …. will make the ontological variation … more transparent” (Dunning, 2017 p35), being honest about where our ideas on urban experiments have come from practitioner engagement and project specific experience from within the industry sector. We have initially defined the general (class) and specifics (sub-class) of experimental urbanism and described the relationship between the characteristics and associated attributes as a “triple” - a clear binary relationship that includes a “subject” (the object described by the triplet), a “predicate” (defines the data within the object being given a value), and an “object” (the value). Within the coding classification, this approach allows for the development of an RDF (resource description language) file, and the relational structure of the information leads to appropriate and functional data mining algorithms based on the structure of the ontology within standardised RDF and OWL (web ontology language, a syntax language developed by the W3C Web Ontology Working Group as the way to describe the semantic metadata) (Nebot & Berlanga, 2012). These RDF / OWL triples are machine readable (Barati et al 2017) and allow for data mining the patterns of relationships as much as the specific characteristics.

The approach is borne out of a frustration around the excluding technical language and repeating / confusing practice in experimental urbanism allied with the implicit political bias in many academic and policy-based research. In this work, we are abstracting and simplifying the underlying structures, though “there are … dangers in being theoretically polygamous … (but can be) commensurate under (a) critical realist ontology” (Aalbers & Christophers 2014, pp 423-424). We suggest the relevance of a grounding in critical realism, being interested in finding an explanation for social systems, processes and structures, for the scope of work on experimental urbanism, of knowledge derived from in-depth case studies and practice-based experiences (Lawson, 2006).