> WS doesn't seem to have picked it up. That was my point.
> Thanks for allowing me to clarify it. I was criticizing WS's
> application of the "Popper [falsification] test," not Popper himself.
I do not think that my interpretation of falsificationism was out of range. In fact, some social theories are subjected to empirical tests in which you set two alternative explanations and run a test and then see which one is not supported by the data. I'm trying to do that as much as possible and I can give you references to some of my published works that do that, for whatever it is worth. So I do not think that I missed much here.
I do realize, of course, that social science is not all about hypothesis testing, and a big part of it is "idiographic explanation," or what Clifford Geertz calls "thick description" - a detailed empirical accounts of some aspect of social life which also have some analytical structure- albeit they are not really theories in the sense of being generalizable. I think much of Marx's work can fit that model. The main value of thick description is not its predictive power, but the generation of reliable (i.e. analytically useful) social facts. This is a nontrivial affair in social sciences where facts are not as directly observable as in natural sciences.
So with that in mind we may say that while LTV - and much of Marx analysis - was indeed very useful as an "idiographic explanation" or a "thick description" of the 19th century capitalism in England - treating it as a nomothetic theory or science with predictive power is unwarranted, to say the least. But that is, however, how it is treated.
So if someone insists on treating Marx's work that way, the fundamental question is 'What did it predict?' And the only honest answer to that question is "Not much." And this gets us precisely into Lakatos and his concept of scientific research programme (also mentioned by jks) - which in essence is an attempt to save theories that fail empirical tests. Lakatos is quite specific in identifying modern Marxism as such an attempt, but I also argue that much of economic theorizing is also a scientific research programme (see my paper "The Death Knell of the Utilitarianism" in the December 2000 issue of _Voluntas_). I may also add that economics is not in alone here, sociology is just as bad.
So to summarize, I have no quarrel with doing social science by compiling empirically reach and analytically informed descriptions of particular historical phenomena. There is indeed a great value in that effort - the establishment of sociological facts. But treating such descriptions as if there were nomothetic science - a set of abstract propositions that apply to a broad range of empirical phenomena and thus allow making predictions - is really a misapplication, if not outright abuse, of scientific methodology. I think it is intolerable when the right does with the neo-classical economics, and it is equally intolerable when the left does it with Marxist economics.
Another thought - methinks the main purpose of the LTV in Marx's writings was to refute the populist and utopian socialist notion that capitalists somehow do not pay fair wages. One implication of that notion is that capitalism is OK as long capitalists pay a "fair" wage. Marx rejected that moralistic interpretation by arguing that the wages being paid are "fair" within a particular system of property relations (which he showed by the means of the LTV), so if the workers get the short end of the stick it is because they are screwed by the system itself rather than a bunch of Scrooges. So the way to go is to reform the system instead merely demanding "fair" wages. If memory serves, this is consistent with Leszek Kolakowski's interpretation, who I think has a good understanding of the issue and is a first rate thinker despite his later "indiscretions."
However, if you treat the LTV as a nomothetic proposition, you end up with something like "wages under capitalism are determined by the socially necessary cost of the reproduction of labor power." That strikes as a rather trivial tautology unless you come up with a method of determining the socially necessary cost of reproduction of labor power, which is a rather murky area of interactions between power, demographics, technology, culture, international relations etc. So if at the end of the day we do manage to develop a theory of socially necessary costs of reproducing labor power, the sheer magnitude of that opus will reduce the LTV to a rather trivial footnote. It is like saying "To fly to the Moon, one needs a rocket." Without the science of rocketry, the proposition is really a tautology since "rocket" in this context can be defined only as a "device that can fly to the Moon." With the science of rocketry, however, all the excitement is in designing and constructing the thing, and then saying "btw, you can use it to fly to the Moon" which is not tautological but still trivial.
Wojtek