Steve writes, in response to Paul:
> Agreed. My point is that this analysis is only possible if you
> "de-couple" the value-productivity impact of training from its
> value-input (which then lets you begin with a hypothesis like "suppose
> training [doubles/triples/sextuples] productivity"). This is feasible
> with Hilferding's use-value/exchange-value analysis, but not with the
> Sweezy/Meek approach.
I still don't see this. IMO, the things that have to be
"de-coupled" are (a) the physical productivity increase due
to training, and (b) the labour-time input to the production
of skill -- but then, as I tried to show with the typing
example, there's really no temptation to couple these in the
first place. That is, there's no paradox in the idea that a
training process that doubles a worker's physical
productivity (output per hour of direct labour) might
involve (say) a mere one percent increase in the total
labour-time transmitted to the product (both directly and
indirectly via the training) per clock hour. There _would_
be a paradox if one "read" the doubling of physical
productivity as a doubling of _value_ productivity; but
Paul's argument, with which Steve says he agrees, implies
that is the wrong way to think about things. (I.e., if the
training were taken as doubling the workers' value
productivity, it would appear to create a source of extra
surplus value independently of the machanism Paul specified,
namely a reduction in the value of labour power just as
per 'ordinary' technical progress.)
Allin.