To: bhaskar-AT-jefferson.village.virginia.edu
Date: Thu, 7 Aug 1997 13:48:41 GMT+0200
Subject: Re: BHA: Non-experimental science (was "What must the ...")


Hi again, still harping on the theme of epistemological relativism
....

Suppose we have two theories T1 and T2 about racism.  T1 says that
racism has been brought about in society as a result of the
configuration and reproduction of particular sets of social relations.
T2, on the other hand, asserts that racism is inevitable because 
=91race' as a genetic fact really does exist and determines the
differential powers and capabilities held by different groups of
people within society.

T1 presupposes a supra-individual world in which social relations of
ethnicity and class are the key explanatory elements , whereas T2
presupposes an ontology of atomistic individuals with pre-determined
genetic makeup constituting the basis of its explanatory framework.

Now Bhaskar's offers the following guidelines for choosing between
competing theories on the basis of the principle of judgemental
rationality. He argues that we can still sustain the notion of a
rational choice between theories;

 For we can allow that a theory T1 is preferable to theory T2, even if
 they are incommensurable, provided that T1 can explain under its
 description almost all of the phenomena T2 can explain under its
 description plus some significant phenomena that T2 cannot explain.
 The possibility of saying this depends on the explicit recognition of
 a philosophical ontology .... For to say of two theories that they
 conflict, clash, or are in competition, presupposes that there is
 something - a domain of the real - over which they clash (Bhaskar
 1989b: 19).

Now, my question is, whilst it is an article of faith for all realists
that there must be some philosophical ontology underlying competing
theories about events or processes, given the weak discriminatory
potential of Bhaskar's epistemology (ie. his stance of epistemic
relativism) how are we to employ the structured philosophical ontology
that he has derived from experimental practice in making our choice
between T1 and T2? After all, even competing accounts of the
experimental practice of physics do not simply differ at the level of
their theoretical descriptions of the world, but also in their
accounts of its ontological make-up (as demonstrated by Kuhn and
Feyerabend). 

This is exactly the case in the example of racism (above) where we
have two competing ontologies; an ontology of atomistic individuals in
T2, and; an ontology of supra-individual social relations in T1. 

Any judgement as to the =91ideological' nature of T2, however, would
have to presuppose the =91scientificity' of T1 - a move which cannot be
sustained on the basis of a stance of epistemological relativism. My
reasoning is as follows ....

According to RB a system of beliefs may be characterised as
=91ideological' if, and only if, three types of criteria are satisfied.
These criteria he calls =91critical', =91explanatory' and =91categorial',
respectively, and his procedure might be enumerated after the
following fashion:

1)	One must possess a theory (T1) which explains most, or the most
significant, phenomena under its own descriptions, explained by the
ideological theory (T2). It must in addition be able to explain a
significant set of phenomena not explained by the ideological theory.

2) 	This theory (T1) should also be able to explain the reproduction
of the =91ideology' in question (T2), that is the reasons for its
continued acceptance by social agents. If possible, it should also
specify the limits of the proposed ideology and the conditions for its
transformation. Specifically, it should explain these limits in terms
of some generative structure, or set of relations, which it is capable
of accounting for, but which is altogether absent from, or obscured
within, the proposed ideology.

3)	The proposed ideological theory (T2) must be unable to satisfy a
criteria of scientificity (this material is not quoted verbatim, for
the original see PON, 1989, p 67/8).

My conclusion, on having scrutinised the above conditions, is that for
us to declare an opposing theory =91ideological' we must already be in
possession of a theory that is =91scientific'. But is this anything more
than begging the question? That is:

=91How do we know T2 is ideological?' .... =91It explains less than T1
which is scientific' =91Well, then how do we come to explain the
scientificity of T1?' .... =91It explains more than T2 which is
ideological'.

Before I am accused of caricature let me add that for RB the ability
to explain the conditions of the reproduction of some set of
ideological beliefs would assign to T1 a specific type of cognitive
superiority in that this form of explanation possesses an ontological
depth lacking in any ideology (ibid. p 68).

In terms of this fuller picture the above question as to what it is
that makes a =91scientific' theory un-ideological would now invoke the
answer that;
 =91it can explain more than the ideological theory, and demonstrate its
 necessity', and what's more , =91it must have ontological depth'. 
In other words, as I understand it, it must make reference to the
causal powers of some underlying structure or generative mechanism.

But my point is that there will always be more than one theory whose
claim it is to do this much. In the example of racism this generative
mechanism could be either the micro-structures of genetic constitution
or the structures of social relations. Both T2 and T1 possess
ontological depth so the issue cannot be one of ontological depth but
rather of how to determine the superiority of one theory relative to
another. We cannot escape the issue of epistemological judgements and
in this regard Bhaskar's articulation of =91judgemental rationality' (in
conjunction with epistemic relativism) does not, in my opinion,
provide a sufficient basis for asserting the generalisability of his
structured ontology, or for that matter, even a defence of the
derivation of this ontology.

I think that at least a part of the problem with Bhaskar's weak
epistemology has to do with his failure to elaborate a theory of truth
consistent with the boldness of his ontological derivations (at least
within his earlier works of RTS and PON).

Bhaskar has rightly, to my mind, taken on board the basic insight of
post-structural linguistics and post-analytical philosophy that we can
only ever speak of our knowledge of the ontological entities of
reality in the terms of some language. But this recognition (that we
can only have access to reality through discourse) does not, and
should not, entail the adoption of epistemological relativism. Nor
ontological relativism. It simply forces us to recognise that our
knowledge is always open to correction and revision. It is
spatio-temporally specific and fallible. 

Yet, the acknowledgement that there can be no final truth does not
make the concept of truth itself any less important. It is only on the
basis of a theory of truth that we are provided with the normative
rationale for posing questions about the degree to which a discourse
can be said to convey information about, or knowledge of, the real
world. We need a stronger concept of truth, and a stronger epistemic
stance than that on offer within RTS and PON, in order to combat the
frivolities of postmodernism generally, and specifically (here in
South Africa), the reproduction of an ethic of political opportunism
under the guise of cultural relativism.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Colm Allan                                                Tel: (0461) 318 361
Department of Sociology                           Fax: (0461) 25570
and Industrial Sociology                            e-mail scca-AT-warthog.ru.ac.za
Rhodes University
Grahamstown, 6140
South Africa
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


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