[Interpretationandmethods] What is The Right Way to do Interpretive Social Science?

[Patrick Thaddeus Jackson]

On Jul 7, 2008, at 5:28 PM, WJKELLPRO at aol.com wrote:

> So, this is what I would advise grad students to consider as a  
> proper interpretive social science method.  That is, establish the  
> relevant facts, interpret these for the meanings that the actors  
> understood for themselves, and check for meanings of which they may  
> not be aware, or may not state.  Then appraise the mix for its  
> degree of rationality under the circumstances, with a view to  
> measuring its degree of human rationality.


Bill's suggestions provide an interesting roadmap for doing a certain  
kind of normative evaluation of social action, by comparing what  
people do with an idealized standard of human rationality. Setting  
aside for the moment any question about Polanyi (since I am in no way  
an expert on Polanyi) and dealing just with the recommendation itself,  
I must disagree with Bill's advice on two counts. First, whatever else  
this is, it isn't explanatory social science. Second, the proposed  
technique runs into some perhaps insurmountable operational problems,  
particularly when it comes to "establishing the relevant facts" and  
describing "meanings of which they [the actors] might not be aware, or  
may not state."

My biggest disagreement here is with the notion that social science is  
supposed to be about issuing normative appraisals of actions. That  
strikes me as a deeply problematic conflation of two different  
intellectual operations, explanation and evaluation. Explanation is  
about determining how and why things happen, and as such says nothing  
whatsoever about the moral and ethical status of of those happenings;  
evaluation is about rendering a judgement on some happening or  
condition, and is logically independent of any particular explanation  
of how or why we have that situation or occurrence. Both of these  
operations are important intellectual functions, but they aren't the  
same thing -- they're logically disjoint.

This is apparent from Bill's two illustrative sketches of studies.  
Bill asks "why George Bush (GB), acting as Commander-in-Chief, ordered  
the preemptive strike on Iraq, which has since resulted in the  
prolonged Iraq War" but the subsequent elaboration doesn't actually  
answer that question! Instead, it answers the parallel question "was  
it rational for GB to have ordered the preemptive strike on Iraq?" (I  
don't want to get into the rest of the causal chain that Bill posits,  
except to say that it is far from clear that the preemptive strike on  
Iraq was the cause of the prolonged Iraq War; I'd personally cite  
something like the reconfiguration of US foreign policy identity as  
the main cause, but there are other takes on this question. The point  
is that the social-scientific jury is very much out on this issue.)  
Bill also asks about "the ways in which the ideas and attitudes of  
accountants and IT practitioners enable or constrain the involvement  
of accountants in IT governance" but the subsequent elaboration  
doesn't answer this question, but the parallel question "how rational  
is the use that IT practitioners make of accountants?"

I don't want to be misunderstood here. I'm all for normative and  
ethical evaluation of action. But let's be clear that such an  
evaluation is not the same thing as an explanation of why that action  
took place. Explanations are, by design, ethically neutral -- which is  
not to say that the terms and notions used in an explanation don't  
have ethical commitments built into them. They obviously do, since  
it's not possible to have a commitment that doesn't have ethical  
content at some level (which was half of Weber's point, many years  
back). But an explanation is not an evaluation, because the use of  
social-scientific concepts to explain something is logically distinct  
from the use of those concepts to evaluate something (which was the  
other half of Weber's point).

That said: if one is going to do ethical evaluation, I don't think  
that Bill's proposal is the best way to go about doing it. Bill's  
procedure requires the analyst to first "establish the relevant  
facts," which he illustrates through techniques of data-collection  
which basically take first-hand and other participant accounts as  
though they were true and accurate (such as consulting the 9/11  
commission report). This makes no sense to me, since what that data  
will likely yield is not any analytically rigorous facts but instead a  
particular, perhaps peculiar, political and discursive construction of  
"the facts." That's interesting stuff, but to my mind not particularly  
useful for figuring out whether the actions undertaken were in any  
sense rational ones. In fact, I think that would be very interesting  
data to use in an analysis of the social production of facts, which  
could be part of an explanation of how and why something  
happened . . . but that would tell me nothing whatsoever about whether  
that thing that happened was rationally justified. For "rational" I  
would need to know the situation as it was, not the situation as it  
was later constructed to have been.

But where I really get lost is in Bill's suggestion that the analyst  
then proceed to elucidate "meanings of which they [the actors] might  
not be aware, or may not state." That either looks like a license to  
introduce whatever the analyst believes that the actors should have  
concluded, or a warrant to import some universal theory about  
interests or motivations or whatever as a supplement to what the  
actors actually said and thought. That's a door that I'm never  
comfortable opening, since once we do then a) there is no reason to  
take the experienced world of the actors at all seriously, since we  
can just replace it with what we already know to be the case; and b)  
we are then placed in the unenviable position of having to  
transcendentally justify our preferred universal framework. Marx  
couldn't do it, Habermas hasn't been able to do it, the rational  
choice people can't do it . . . so either there's a solution that they  
haven't thought of yet, or one abandons the attempt altogether and  
just rests on faith in the normative rightness of one's account of  
motives and interests and the human person. Either way, the whole  
exercise is called into question because we are now tossed into a  
debate about the ultimate status of the theoretical apparatus that the  
analyst is using to pierce through the actors' partial account of  
things to get to  what is "really going on" -- and that's not a debate  
that is likely to end any time soon, if at all.

Personally, I think a much more consistent way to go about ethical  
evaluation is to elaborate some ethical rules and then see if actions  
were carried out in accordance with them. Why muddy the analytical  
waters with efforts to accurately describe the state of mind of the  
actors, or to ascertain what it "rationally" ought to have been? Even  
if such a  description were possible -- and there are good reasons,   
deriving from Wittgenstein, to doubt that it is even in principle  
possible to describe a subjective mental state -- that would only give  
us a sense of what people thought. This would neither explain their  
actions nor give us much purchase on ethically evaluating them (to say  
nothing of their consequences -- that would require a social- 
scientific explanation of the results of some action, and certainly  
couldn't rest content with some actor's evaluation of the results of  
their action).

To sum up: I can't agree that Bill's proposal is the best way to think  
about "interpretive social science" or even that it constitutes a  
"proper interpretive social science method." I think Bill is sketching  
a procedure for doing ethical evaluation of social action -- and it's  
not a procedure that I would recommend, given certain operational  
difficulties that it presents.

I stick by my four-part suggestion for a good social-scientific  
research proposal, interpretive or otherwise: 1) research question; 2)  
review of scholarly literature to situate the study; 3) initial guess  
at an answer to the research question; 4) methodological plan for  
evaluating the initial guess and answering the question.

PTJ
===
Patrick Thaddeus Jackson
Director, General Education Program, American University
Editor-in-Chief, Journal of International Relations and Development
http://profptj.blogspot.com | http://www.kittenboo.com
calendar: http://ical.mac.com/onyxdr/Patrick

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://listserv.cddc.vt.edu/pipermail/interpretationandmethods/attachments/20080707/dc6caa34/attachment-0001.html