File spoon-archives/bhaskar.archive/bhaskar_2000/bhaskar.0006, message 223


Date: Mon, 26 Jun 2000 12:29:52 -0500 (CDT)
From: viren viven murthy <vvmurthy-AT-midway.uchicago.edu>
Subject: Re: BHA: Explanatory Power


Dear Jonathan,

Thanks for your post.  It helped clarify some parts of the current
debate.  I have a few general questions, some of which are only
tangentially related to the present issue.  In your post, you explained
that a "theory" is linked to "generative mechanisms" and "structures".  I
have always been puzzled at why Bhaskar used both of these terms, since
he sometimes defines structures as generative mechanisms.  I am not sure
what the difference between generative mechanisms and structures are.
Margaret Archer seems to try to clarify some of these issues in her
Bhaskarian critique of Giddens, but I am still confused about what a
structure is on the critical realist account.  It seems that the idea of
structure is intimately linked to the idea of explanatory power.  

Your comments helped me to understand Collin's position.  In your terms,
one could say that Collin argues that we will not be able to make a
judgment about which theory "holds water" until we have some idea of the
structures that exist.  It seems that you are committed to
this position
as well, since you affirm epistemological relativism;  the ontological
variables ground your choices about theories.  This is why, if we were in
an ontological
vacuum, we could not escape relativism, since we would be confined to the
epistemological realm, which you have claimed is relative.

This is perhaps why Bhaskar does not make a judgement for or against
historical materialism.  He awaits more ontological evidence.

This brings us back to the notions of existential intransitivity.  You
mention that one of the criteria for judging explanatory power is
consistency with evidence.  I would add, (as I think Collin or Tobin have
already mentioned) that the theories consistency with itself is also a
variable.  But here the theory is considered as an ontological object
(intransitive).  Hence the minute we consider two theories ontological
vacuum is already somewhat filled.  We can perhaps rule out some theories
because they do not make sense(we could, of course, be wrong and find out
later, that they do).

Best,

Viren

On Mon, 26 Jun 2000, Jonathan Pratschke wrote:

> Hi
> 
> Apologies for lagging behind in this discussion, I wanted to reply to Colin and Doug but didn't get a chance over the weekend. Colin suggests that theory adjudication proceeds case by case: 'give me a concrete example of a complex social object and two theories of it and I will attempt to show you which one I prefer'. He argues that to prioritise empirical, or any other, criteria prior to engagement with actual theories of specific objects of inquiry is fundamentally mistaken. Instead, we should accept that rational, conventional, consensual, aesthetic and other criteria will at times take precedence over the empirical in various combinations, depending on the object of inquiry.
> 
> As I am in disagreement with this position, I will respond by providing an alternative formulation, as Colin requested. I will construct my account using the concept of 'explanatory power', which plays a key role in Bhaskar's account of CR in PoN, RR etc. For example, Bhaskar argues in PoN (p. 53) that historical materialism "can only be justified by its fruitfulness in generating projects encapsulating research programmes capable of generating sequences of theories, progressively richer in explanatory power".
> 
> Although ontological realism provides a priori grounds for being suspicious of reductionist or empiricist explanations, I am not primarily interested in this form of adjudication. Nor am I concerned with decision-making per se, which is the focus of some of Colin's posts. For example, he provides the examples of a committee decision, the decision to go to the cinema etc. These examples obscure the role of 'explanatory power' in theory adjudication by focusing on non-scientific forms of decision-making.
> 
> In contrast, I believe that the concept of 'explanatory power' as it was developed within CR *must* prioritise empirical criteria in relation to theory adjudication, although in special circumstances other criteria may be considered as surrogates. This does not fall into the trap of foundationalism, because 'explanatory power' is relative and dependent upon existing knowledge and technical capacities. I will try to explain my position in more detail below.
> 
> For CR, a theory provides a description of generative mechanisms and structures which 'explain' a phenomenon in a given context by stating its causes. 'Explanation' in CR has a more precise definition than in other philosophies of science, because of its explicit linkage with 'causes' and 'reasons'. The issue of explanatory power arises in debates between and within research programmes: in the clash between rival perspectives, 'explanatory power' reflects whether theories can accommodate existing evidence without sacrificing their logical coherence or other desirable properties. In the development of individual research programmes, 'explanatory power' captures the same process over time, e.g. by forcing the researcher 'back to the drawing board' if a theory does not hold water. It is important to stress that 'explanatory power' assesses primarily whether a theory is consistent with the evidence and only subsequently whether it achieves this consistency at the expense of logical or other contradictions.
> 
> 'Explanatory power' is thus a mainly empirical criterion of theory adjudication. It is distinct from the positivistic notion of explanatory success because it accepts that the effects of mechanisms are often not immediately observable, that they may interact, cancel each other out etc. and that they must therefore be studied indirectly and in specific contexts. But theories must *explain* something, and their success in doing so is judged by their consistency with the evidence. If a theoretical approach appeals to a countervailing mechanism in order to account for the available evidence, then its explanatory power depends on the new theory which incorporates the countervailing mechanism. We must assess the explanatory power of this new (combined) theory in order to complete the adjudication. By grafting on an ad hoc structure, a research programme runs the risk of generating logical and other contradictions. Lakatos provides a number of rich descriptions of how researchers attempt to 'paper over' the emerging cracks in their theories as a result of ad hoc additions...
> 
> If the available evidence is not sufficient to distinguish between rival theories, then a number of alternative strategies are available. Obviously, the priority is to collect additional information, to design new instruments and to develop more refined concepts in order to facilitate theory adjudication. If the technical precision of available instrumentation does not allow for theory adjudication, then non-empirical criteria must be used as surrogates, e.g. parsimony, elegance, personal preference. Note however, that these criteria are subordinate to empirical criteria, which have simply failed to separate the rival theories, and in the future empirical evidence may provide grounds for making a different choice.
> 
> Just as theories are dependent upon auxiliary assumptions (observational theory, concepts used to collect data, instrumentation), explanatory power is also dependent upon assumptions. For example, a theory may appear to have higher explanatory power than another for the sole reason that the object of inquiry was mis-interpreted or mis-recorded. Thus a primarily-empirical concept of 'explanatory power' is consistent with epistemological relativism and is in no sense foundationalist. Judgements about 'high' and 'low' power are relative and conditional upon auxiliary assumptions and instrumentation which change over time.
> 
> This account of theory adjudication does not tie researchers' hands, as it may be rational to continue working on a theory which appears currently to have low explanatory power, if one is convinced of the possibility of demonstrating its power in the future. The importance of this account of science derives from its opposition to the use of non-empirical criteria to decide scientific debates. If my opponent insists that his or her theory is preferable precisely because of its internal contradictions, challenging presentation, 'difficult' terminology, extensive quotations in Greek etc., then I can respond, at least in part, at the philosophical level. In most cases where social power influences research, insistence on empirical criteria is enormously positive. I would accept, as Colin argues, that we, as researchers, may sometimes resort to authority in defending our theories. However, the important point is that we cannot appeal to CR for support in doing so. Should another researcher come up with a more powerful explanation, THEY can appeal to CR against us!
> 
> To return to the (rather unsatisfactory) example of the coffee, for anybody other than Colin to accept his statement 'this is a cup of coffee' requires an assumption about 'instrumentation': that the instrument correctly records the phenomenon of interest and related evidence. This assumption may be tested directly in some circumstances (i.e. I could take a plane and check out Colin's desk for myself), but if we are using it to assess the truth of his statement, we must merely assume that the instrument works. This has nothing to do with faith, and its rationality is entirely relative.
> 
> Apologies for the length of this message & thanks in advance for any comments
> 
> Jonathan Pratschke
> 
> 
> 



     --- from list bhaskar-AT-lists.village.virginia.edu ---

   

Driftline Main Page

 

Display software: ArchTracker © Malgosia Askanas, 2000-2005