What is left is the distinction that I made in my opening post: 
productive labour is that which produces a successful commodity under 
capitalist relations of production. Such labour creates surplus value 
- whatever its specific (use-value producing nature)
First to Ian Hunt's message. I unfortunately lost this (finger 
trouble) whilst I was replying to it. But I remember it as effectively 
agreeing with the above, and giving the examples of domestic labour 
and self-employed traders as exemplifying unproductive labour. Just 
so - because they are not performed (with a caveat about Alan's point 
that de facto capitalist relations may be lurking behind the 
apparently self-employed) under capitalist relations of production, 
don't produce a commodity, etc.. [Of course, here we could cue a 
discussion about 'necessary but unproductive' or 'indirectly 
productive' labour, but that would be a digression.]
Turning to Alan Freeman, he wrote:
> Alan wrote
> >I'm with Michael ... 93% of the way.
And then goes on to argue why.
And later gets to the 7% remaining matters of disagreement, writing:
> >though commodities include services, I would stop short of
> >saying that *all* services are commodities. The dividing line is, for me,
> >whether their principal function is to circulate existing use-values or
> >whether they create a new use-value.
And I need to know why the service of circulating existing use-values 
is not a use-value, and so the labour doing it is not commodity 
producing and productive (just so it takes place under capitalist 
relations of production)?
And so to Chris Arthur, who is:
> with Alan 96% of the way. 
And so presumably agrees with me 89.28% of the way (unless, of 
course, some of what he agrees with Alan about falls into the 7% 
about which Alan disagrees with me ...)
Chris goes on:
> The difficulty arises with services whose use-value appears to be in some
> way purely parasitic. Alan puts this in terms of circulating USE-VALUES.
> However it could be argued this is itself a use value. For example we
> standardly take advertising and estate agents as unproductive but in any
> society it is useful to provide information about the availability and
> performance of products so some sort of analogue for these services will be
> required under socialism.
Now this makes clear that Chris's criterion of the unproductive (and 
indeed the productive) here is a comparison with some rationally and 
humanely organized society. Of course such a comparison is an 
important source of political critique of capitalist society - but it 
is not relevant to a distinction the first (and I think 
uncontroversial) criterion for which is the creation of 
surplus-value.
But he continues:
> As for financial services here the use value seems to me merely that of
> circulating VALUE and I would say that even though these have a commodity
> form they are certainly not products as the industry presumptuously claims
> and the service is purely parasitic on the circulation of value. For
> instance when a bank sells me a bond there is a transfer of ownership
> (something Jurriaan seemed to think was important to the commodity)  but of
> an asset which simply represents a claim to part of produced value. So the
> labour of the bank salesman is unproductive - at least from the point of
> view of capital as a whole.
Well, I have a problem with this notion of 'parasitic' here, if we 
mean draining the life-blood from capitalist commodity production 
and circulation.. Financial 'products' certainly have a use-value - 
they help in the dynamic allocation  of money capital to the sectors 
most likely to be in a position to employ labour most likely to 
generate lots of surplus-value.
Chris continues:
> I think that without the
> distinction between production and circulation of values we could not
> sustain any connection with labour.
I need this unpacked a little for me.
and then
> For example the people employed to
> speculate in currency may or may not create fortunes for their employer and
> bonuses for themselves but the return has no relation to the hours spent.
Again, I don't see why not, if we look at the specific services of 
speculation in bringing to a head monetary crisis, that is a 
use-value, as is helping to establish exchange rates in accordance 
with relative national value-productivity. Of course, this opens up 
massive opportunities for gambling - that is a zero-sum game, in 
which value just changes hands. But is bookmaking  labour employed by 
Ladbrokes unproductive?
And so to Jurriaan's post. He writes:
>In probably 80-90%
>of cases we can agree on where the frontier [between productive and 
>unproductive] should be drawn, as Alan >would point out, so it's not 
>really so quantitatively significant. 
Yet Duncan, in a recent post, thought that the absence of an adequate 
productive/unproductive distinction in NI accounts was one of (the 
few) difficulties with using them for Marxist empirical work. It is 
clear that conventions differ across Country's NI accounts, and also 
that labour not employed under capitalist relations is treated 
inconsistently. Going on the 'Keynes marrying his housekeeper 
reducing National Income' story, I am sure that domestic labour 
doesn't figure in the UK accounts. Last time I looked, it was also 
the case that the money value of  State labour's untraded products 
was based only on the costs of inputs. All of which seems to suggest 
that the problems for Marxist quantitative work are probably 
fragmentary, if not random.
Jurriaan goes on:
> his [Mandel's] "second cut" definition looks at what labour is productive
>for capital as a whole, i.e. what labour adds to the total mass of
>surplus-value (which may not be the same as additions to "wealth",
>understood as the total mass of use-values)
Maybe, but someone will have to explain to me how labour that is 
productive (of sv) for an individual capital systematically fails to 
add to the total mass of surplus-value
Jurriaan goes on to cite an example from Marx:
>In this sense Marx discusses, in Capital Volume 2, the case of e.g.
>the buying and selling agent, whose surplus labour does not create 
>any new value, arguing: "...society does not appropriate by this 
>means any additional product or value.  But the costs of circulation 
>that he represents are reduced...". (Penguin ed., p. 210). 
But as Chris and Alan agree, it is difficult to argue the case that 
trading labour does not produce a (service) commodity; and if it is 
capitalistically employed, it will in general create new value.
Jurriaan concludes:
>The question to ask then is, from the standpoint of capital as 
>whole, do schools, hospitals, hairdressing, cleaning services etc. 
>as necessary elements of social reproduction also add to the total 
>mass of surplus-value (to the social surplus-product) ?
This implicitly trails the red herring of potentially 'necessary but 
unproductive' or 'indirectly productive' labour. Apart from that, the 
answer rests upon the basic criterion - are these labours employed 
under capitalist relations.
Continuing ...
>Does the capitalist society (as distinct 
>from an individual capitalist) appropriate through these activities 
>themselves any additional product or additional commodity value ? 
Again, if the one does, paru passu, so does the other.
And then:
>There are various ways to answer this question, but I've tended to 
>think that the answer is no, that these are necessary maintenance 
>activities which do not themselves expand (add to) the total mass of 
>surplus-value, from the social point of view.
Are not the 'maintenance' workers in a factory, *servicing* the plant 
and machinery typically productive?
And finally:
> I could be wrong about 
>that but I see no general principle that would fully adjudicate the 
>issue or eliminate all controversy. 
Well, this has been the point that I have been posing: that the clear 
criterion of demarcation can only be production relations. The rest 
is empirical contingency of no more import than the fact that some 
labour employed by capital may turn out to not contribute to the 
creation of sv, because the (potential) commodity they produce fails 
(is in excess supply). (But that opens up another can of worms.)
Thank you for your patience.
Comradely greetings,
-
Michael
*===================================*
Michael@mwilliam.u-net.com 
"Books are Weapons" 
Dr Michael Williams 
Department of Economics		                   Home:
Faculty of Humanities and Social Sciences	   26 Glenwood Avenue
De Montfort University			           Southampton
Milton Keynes                                      SO16 3QA
MK7 6HP
tel:+1908 834876			           tel/fax: +1703 768641
fax:+1908 834979
mwilliam@torres.mk.dmu.ac.uk
http://www.mk.dmu.ac.uk/~mwilliam