ABSTRACT
This chapter contains a rich description of two cases that made remarkable progress toward sustainability with strategies that qualify, in my view, as co-evolutionary. The analytical and interpretive content of this section is deliberately kept to a minimum to avoid colonizing the readers’ perceptions. They should rather have the chance to develop their own unfettered interpretation of the raw data presented therein. My own angle on this material is presented in subsequent chapters. This strategy, along with rigorous naturalistic methods as described in Box 3.1, is intended to make my argument as transparent as possible, to enable the reader to comprehend what I see and why I see it, and lastly to support the trustworthiness of the conclusions drawn. Methodology in a box
The conclusion's credibility, transferability, dependability and confirmability also depend on the suitability of the methods employed. A thorough description of the methodology and methods of the study that led to this book is available at www.coevolution.info. For readers who cannot take this long route, however, it is necessary to present at least the main ideas and techniques that governed the research process.
In adopting a constructivist position à la Lincoln and Guba (1985), I subscribe to the notion that what appears as fact always has interpretive slack that needs to be narrowed in a dialogue between researchers and the people of the setting they study. The statements in this dialogue, however, must not be blindly taken at their face value because they might be purposefully or unconsciously distorted. This is why I call my approach with Kincheloe ‘critical constructivism’ (1993). This calls for a systematic and reiterative confrontation, also known as triangulation, of interview data with the literature and archival data. This approach allows researchers to learn and to adjust their working hypothesis as well as their research design as they go – in comparison with positivist investigations that can only furnish support or falsification for a preconceived and fixed hypothesis through a static procedure. A constructivist, naturalistic or hermeneutic enquiry is therefore much more flexible but takes on the burden of making the choices that guided the selection, analysis and interpretation of data as transparent as possible.
Literature: The study of relevant literature, combined with my personal experience, can be seen as the entry point into a hermeneutic spiral. I explored the existing body of knowledge in a variety of academic fields such as sustainable communities, learning organizations, radical innovation, social innovation, creative cities, alternative dispute resolution, participation techniques, pragmatism, the history and theory of urban planning, science and technology studies and others.
Archives: The electronic archive of the provincial newspaper Het Belang van Limburg and the personal archive of an employee of the Flemish Department of Transportation provided the majority of archival data for Hasselt. In Fürstenfeldbruck, I studied materials from the archive of the newspaper Fürstenfeldbrucker Tagblatt and from Brucker Land's own archive. Access to the latter, however, was slightly constrained as proprietary and financial data were off limits.
Interviews: Existing contacts in both cases helped me to start a snowball sampling process. The total number of suggested interviewees climbed asymptotically to roughly 30 people in each case, which gave me the liberty to pursue a more and more purposive sampling method (Erlandson et al, 1993). Overall, I conducted 15 interviews in Hasselt and 11 in Fürstenfeldbruck, averaging 52 minutes each. Besides these official interviews, I also conducted numerous informal conversations with citizens, bus drivers, shopkeepers, etc.
All interviews were transcribed verbatim, which resulted in roughly 104,000 words of written text. This information, as well as the archival data and literature excerpts, was ‘unitized [into] the smallest pieces of information that may stand alone’ (Erlandson et al, 1993, p117), which ranged from 10 to 200 words. Each unit was printed on a separate index card, making a total of 2800 cards at the end of this process. The content analysis was performed through several rounds of sorting and re-sorting of all index cards. The purpose of this process was not to check whether the data fit into a preconceived pattern but to find out what pattern the data themselves suggested.
A cornerstone in the attempt to preserve transparency throughout the research process was a ‘reflexive journal’ (Erlandson et al, 1993, pp143–145) where I recorded all current findings, tentative interpretations, strategic decisions and so forth. I wrote 18 entries in this journal, each between one and two pages long, during the fieldwork phase, which helped me to trace the path of my deliberations, detect detours, shortcuts, gaps and unexplored territories, to ‘park’ ideas and to account for any adjustment of the emergent research design. The reflexive journal is available upon request and thus allows me to forestall any potential allegation of arbitrariness.
