Paradigms and Sand Castles : Theory Building and Research Design in Comparative Politics (2003)
Geddes begins by giving us reasons to be doing what we are doing as political scientists. Her own reason is that being ‘’Confronted by compelling and exiting events in the world scholars quickly turn their attention to trying to understand them’’. Page 12 To Geddes, the aim is to understand the world around her and sound methodological ground helps us do this. Geddes, who as further reading of this book will show, likes a metaphor and likens politics to paths in a jungle. Some of which do not take us where we want to go.
Geddes clearly and helpfully defines paradigms in this book. She calls them ‘a dominant understanding of a particular phenomenon at a particular time’ page 17
There is a thought that quantitative data is of higher evidential value but to counter this Geddes refers to analysis that inferred far greater claims than the data could support and therefore quantitative answers are very vulnerable to bias and subjectivity. For Geddes this can therefore be no different than anecdotal evidence if done badly. Page 24
One reason Geddes thinks paradigms fail is because a broad review of the evidence was not conducted in the first place. This means I will need to do a thorough literature review to formulate the theory and come down on a side of the hypothesis. I need to be spanning as much evidence as possible other wise the paradigm may fail or at least be suspect.
When using the data collected some analysis while not ‘ignoring’ evidence, use it selectively to support of refute theories. I cannot do this as it leads to the eventual failure of the paradigm one day although Geddes says this could be decades later.
This ‘general inattention’ (Geddes employs a careful and kind quote here) just leads to a slowing down of formulating a true working theory and our understanding she highlighted at the beginning. Page 29
Geddes also takes time to persuade us that our explanations need to be whitled down to the bear minimum and not be complex and difficult to understand. Any simplification is not necessarily a good one. Some thing to remember perhaps when writing up a conclusion. Page 33
Alexander Dawson 