Page 34 - Decoding Culture
P. 34
THE WAY WE WERE 27
reliable test materials; in its assumption that propositional form
was the ideal means for expressing scientific knowledge; and in its
emphasis on maximizing precision of evidence. Taken together,
these three legitimized an essentially empiricist research practice.
Concern to establish an adequate observation language was
crucial in the development and application of the rhetoric of vari
ables and indicators. This is not the place to rehearse the complex
and chequered history of that discourse, nor to chart its analytic
deficiencies; these issues have been well illuminated by Pawson
(1989) . Of more immediate relevance here is the tendency of vari
able analysis to encourage, even to legitimize, ontologies which
define the social world in a manner most amenable to that
approach. Thus, if behaviour is to be understood in terms of rela
tions between variables, processes of social action are likely to be
conceptualized as a concatenation of measured values scored by
individuals on specified dimensions. To speak of effects, then,
within such an epistemology, is to demand some form of 'before'
and 'after' exposure measurement on selected scales, most com
monly conducted with the naturalized individual as the primary
focus.
Now add to this the supposition that the ideal form for scientific
knowledge is propositional. Although the case for such a claim is
firmly rooted in the deductivist view of scientific inquiry, in
research practice propositional form often survived without a con
comitant emphasis on deductive structure. The rigorous
requirements of deductive-propositional methodologies were ame
liorated by a slide toward the less demanding inductive
propositional form (Tudor, 1982: 166-168), toward what Willer and
Willer (1973) call 'systematic empiricism'. Here, allegiance to the
hypothetico-deductive model of scientific method is superficial.
Propositions are no longer statements deduced from theory and
subjected to test; they are, rather, empirical generalizations,
Copyrighted Material