I have to admit I am torn. I often preach against excessive parsimoniousness in social sciences. In particular, I am no longer a (strong) advocate of the Caltech rules (one argument per paper), because I have seen too many papers, say, claiming that there is a causal relationship between inequality and voting behave, and too many papers claiming the reverse causal relationship with none of the two types of papers ever citing each other. Such stuff is good for CVs, but leads to an inflation of papers with very little cumulative scientific knowledge production.
There is, however, the opposite danger, which is becoming too complex, too holistic. In the struggle of serving to many masters, many MA and PhD students want to squeeze everything into a framework or an approach and often these students get lost in these. While the outcomes are often overly ambitious, occasionaly hilarious, the students take their cues from the scholarly literature.
Let me make examples from two fields: In comparative political economy, very established frameworks or approaches are Peter Hall and David Soskice’s Varieties of Capitalism (VoC) and Gosta Esping-Andersen’s Welfare Regimes. In the field of development, there is Amartya Sen’s Capabilities Approach approach. Now, I don’t want to defecate onto some of social science most sacred institutions. And my rant is not really against the existence of these frameworks: In God’s Pantheon there is enough space for all types of scientific contributions. Nonetheless, I think these frameworks are very challenging if not ‘god-awful’ to teach.
Just to give you an example. Try using this version of the capabilities approach for your MA (and smoke it): https://www.sciencedirect.com/science/article/pii/S1875067211000320
Hard, isn’t it? And its the authors fault. The types of frameworks I have in mind, often sound great on paper and in theory, but are almost impossible to apply, especially in MA or PhD thesis.
First, they are very complex and holistic, covering material from many different disciplines. Take VoC, which effectively combines insights from business studies, economics, political science and sociology. That’s a lot of literature to work through, if you take this seriously.
Second, they often have an enigmatic status in the research logic. Are VoCs or welfare regimes an independent variable, a dependent variable, both or neither? I have seen all of these options applied, almost always without much reflection. For instance, many researchers use regimes as an independent (macro) variable for micro-level data, say on public attitudes, but are these regimes really independent variables in the common sense? Often the clusters or ideal cases are greatly informed by what researchers know about the macro- and micro-level information about these cases.
Third, are these regimes historical configurations, real cases, ideal cases or a mix thereof? The confusion makes empirical and conceptual research hard. Is Sweden a Scandinavian type of welfare state? Is Denmark? This leads to endless debates, often with the outcome that there are as many regime types as there are countries in this world (if not more). Even Esping-Andersen does not seem to be bothered following these discussions any longer. No wonder, the empirical record of such classifications sometimes looks like this: https://media.springernature.com/lw685/springer-static/image/art%3A10.1057%2Fs41267-016-0001-8/MediaObjects/41267_2016_1_Fig1_HTML.gif
Fourth, frameworks and complex approaches are often hard to fill with content and to operationalize. Take Sen’s capabilities approach. What are capabilities exactly? Some scholars have tried to fill the content (e.g. Nancy Fraser). Others have tried to operationalize them (see the UN’s expanding catalogue of SDGs). But, more than often, this leads to even more follow up questions. And, by the way, what are functionings and why do we need such a clumsy word for this?
Fifth, some of these frameworks and approaches have the half-life of fruitcake on a garden table on a hot and wet German summer day, with the wasps already buzzing around. For example: Does anyone still bother what the EU’s Open Method of Coordination was all about? Probably not, but this did not hinder many PhD students 20 odd years ago to write their theses and books about it. Chapeau, by the way, to those who were the first ones to ride the wave, they usually ended up with permanent positions, but, alas, fortune was with those who came thereafter. In such cases, the frameworks (Green or White books on Governance, Aid Effectiveness, Anti-Poverty Strategy Papers…) come from the policy world, and the policy world constantly reinvents itself to constantly refresh its own legitimacy. See William Easterly’s hilarious though cynical view on the proliferation of such policy buzzwords.
Sixth, there is also the related problem of picking very current and politicized topics. Umberto Eco once said that a good topic needs to be at least 500 years old. Now, Umberto Eco was a historian interested in medieval semiotics. In social science we do not have the luxury to wait that long. But as a caveat his advice remains pertinent.
Why I am saying all this? Because I have seen dozens of PhD students filling all the voids, buttering over all the cracks and rationalizing all the inconsistencies of such frameworks and failing miserably. I have seen people writing not one, but two or three PhD theses on VoC simultaneously: one covering the individual- or firm-level analysis of skills, another on the cross-country comparison of macro regimes, a third one on the evolution of government policies. And I have seen dozens if not hundreds of MA term papers and theses pretending to use the capabilities approach while in fact just name-and-concept dropping it and then moving on to something much more feasible.
Again, I am not saying one should exaggerate parsimoniousness. So don’t do this either:
I guess the trick is balance. Albert Einstein allegedly said that one has to do things as simply as possible but not ‘more simple’ than that.
This is why I try to steer MA and PhD students away from overly complex frameworks. There is nothing wrong with frameworks as sets of statements about reality, guiding research, helping narrowing down the field. However, when complex approaches and frameworks try to copy reality and map too broad a landscape, the terrain becomes mushy and full of pit holes. At some point, you need look at the map and to cut and chop it again.
Some talented students may handle the complexity and generality of Sen or Hall and Soskice. That’s good then. But for the majority: Be realistic, look at small aspects of reality and carve out smaller contributions. Otherwise you poor students end up citing these frameworks (which is good for the authors I guess), but will either ignore them or else you will get dragged down by them.