[Interpretationandmethods] Patrick's post

[Ed Schatz]

Thanks, Patrick, for clarifying why formal models cannot be "tested" in any
serious sense. Of course, the same would go for those laboring in other
traditions, doncha think? Can we really validate/invalidate anything that we
do? The difference, of course, is that others rarely claim the mantle of
science in the same way.

 

My own take is that everything we do-whether we use formal modeling,
ethnography, Ragin's configurational approach, etc.-is an attempt to
represent an aspect of the "real world" that is "out there." To me, a formal
model is as much an attempt at representation of the real world as an
ethnography is (though each holds a different place for the double
hermeneutic of interpretation). So, what separates formal modeling is, as
Patrick indicates, its attention to logical pathways (nicely put!). But, if
I am reading between the lines correctly, Patrick is suggesting that a
tension exists between a logical, formal model and the real world. Is that
right?? Is the world illogical, where the models are logical? Or is a
logical model necessarily a partial representation of messy, complex
reality? I would tend to go with the latter. My sense is that the part of
reality that a formal model tends to capture is usually miniscule and often
trivial. But that doesn't necessarily make formal models fundamentally at
odds with reality!

 

On KKV and "neopositivism": I recognize the limits to variable-oriented
research; all my work to date is much closer in spirit to Ragin's
configurational analysis (though I don't use Ragin's approach explicitly).
Having said that, my own preference is to be agnostic about epistemology as
long as I possibly on the front end of a project. From that starting
position, there is no reason to turn a blind eye to variable-oriented
research that can be thought-provoking. Instead of dismissing a priori what
Pzworski and Lmongi argue about economic development and democracy because
of its (statistical) approach, why not take it, ruminate, ponder, if only to
spit it out as irrelevant or misleading in the end? It seems to me that
variable-oriented research has one key advantage in terms of provoking
thought: it strives for broad generalization. Now, I am not a fan of rash
generalization, of the sort we see all the time, but those venturing such
generalizations bring into the conversation cases that otherwise would just
not be considered together. To go back to my example from earlier, it gives
me (a non-Indianist) a way to think about India! That's a pretty useful
starting point, as long as I recognize that it's just a starting point.

 

 

 

Ed Schatz

Assistant Professor

Political Science

University of Toronto

 <mailto:ed.schatz at utoronto.ca> ed.schatz at utoronto.ca

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://malagigi.cddc.vt.edu/pipermail/interpretationandmethods/attachments/20061006/145da540/attachment-0001.html