I worry a lot about discourse forms as traps as well as tools, for they are
subtler in their way than the actual 'saids' that are shaped to these
forms, and they are harder to change, to innovate, to critique ... they are
meta-tools, and so also meta-traps.
Dichotomies, in my experience, are of extremely restricted usefulness in
general theoretical discussions, whether qual vs. quant, narrative vs
expository argument, or nominal vs. real. Each of these dyads merely
indexes the very large spaces of possibilities on many dimensions that
address their concerns. There are neither just two sexes, nor just two
genders; all category systems make _multiple_ contrasts on multiple
dimensions, otherwise they would be pretty useless in practice. Adversarial
genres of argumentation are the discursive meta-trap that attracts us
unwisely toward merely symbolic dichotomizations.
"Qualitative" and "Quantitative" are mere placeholders in both the rather
empty debates over research method and in the more significant issues of
concrete research practices. I think most of us do know the core of this
issue: that one constellation of philosophical assumptions, master
discourses about research methods, and concrete practices (varying greatly
from one discipline to another) sought and more or less achieved in
academic circles for a period of time a political hegemony that sought to
exclude other approaches. That dominance has been successfully challenged,
but the war has left scars and unstable mismatches among theory, practice,
power, politics, etc. in the complex academic web. The radical critique of
the Old School threatens very deep ideologies with long histories in
European cultures, and with impliations far beyond the academy. Most
academics are not willing to embrace the most radical alternatives proposed
by the critique. There has been a certain, re-equilibration, in the network
(to use an old but brief metaphor), an assimilation-accomodation that has
created the 'received wisdom' that the Old School methods and the New
School ones can live in harmony, and indeed complement one another, need
one another. (Think Darwinism and theology, or more generally Science and
Religion for many people.)
When Gary Shank points to the deeper philosophical differences between the
Schools, he is not, presumably denying that both quantitative and
qualitative differences are useful and related, even complementary, kinds
of meanings to make (surely Peirce thought so). And when David Dirlam and
others argue that quantitative analysis ought to have a useful place in the
New School, they are not, presumably, denying that they were in many ways
fundamentally misused under the theoretical and philosophical assumptions
defining the Old School. The problem is not with quantity but with its
fetishism in the service of a fantasy of universal, context-independent
accounts of meaning-based systems. A fantasy with historical roots (often
merely contingent ones) in many kinds of interests from those of patriarchy
to those of capitalism and technocracy.
What we should all be seriously thinking about, I believe, is just what new
paradigms for the use of quantitative analysis and its integration with
qualitative analysis are consistent with the general theoretical,
philosophical, and political stance of the New School?
A large part of natural science deals with differences that can be measured
on a scale of nearly continuous variation, as opposed to qualitative
differences that index differences of type or category, the continuously
variable vs. the discretely different. Quantitative analysis of a genuinely
metric kind can deconstruct invidious uses of cultural categories like
race, gender, or 'sexual orientation'. It can also describe the
potentialities in the dynamics of interacting constituents of complex
systems for novel, unpredictable, emergent self-organizing behavior. What
is called 'quantitative' method in the human sciences is much more often in
fact qualitative in these terms: it is merely counting the number of
members of a category; it is still grounded in a discrete typology with no
continuous or quasi-continuous (cf. fuzzy categories) topology. Non-linear
quantitative mathematics, and many areas of topology of dynamic systems
(e.g. Thom's theorems), show quite precisely how categorial difference
arises from continua (including material continua).
While it may be true that a large part of the meaning resources of natural
languages are essentially typological, or categorial, in their principles
of operation, it is not true in general that meaning is always a matter of
qualitative differences and categories. Many aspects of spatial, temporal,
motor, visual, and tactile-sensory meaning systems are fundamentally
topological or continuum-based. Mathematics itself is the bridge between
these two aspects of semiosis. We are not going to account for
meaning-based human (and ecological) activity without looking at
quantitative difference. But what is taught in the social sciences as
quantitative methods is mostly unknown in the curricula of physicists and
chemists; it is mostly qualitative category-based analysis masquerading as
quantitative, not because it is genuinely needed (sometimes even
inferential statistics do serve a valid purpose), but because it apes the
prestige of the physical sciences. (Biomedical science represents a complex
borderline case.) What is called 'quantitative' gives genuine quantitative
analysis a bad name; it seeks to make generalizations about matters that,
being essentially and complexly context-dependent, can only come to
ideologically-driven pseudo-conclusions, or to banal statements that are of
no practical use in any context.
The Old School quantitative methods should be rejected, except where there
are overwhelmingly persuasive theoretical reasons to use them. What we need
very much indeed is to develop new paradigms for the use of new kinds of
genuinely quantitative analysis, with new philosophical justifications, new
links to theoretical discourse, new forms of integration with other modes
of analysis.
We need a new _integrity_ for the use of quantitative methods. JAY.
[to be continued]
---------------------------
JAY L. LEMKE
CITY UNIVERSITY OF NEW YORK
JLLBC who-is-at CUNYVM.CUNY.EDU
---------------------------