To take two points slightly out of their original order:
> However, I think Alan doesn't stick to his nostrum 
> (with which I agree) that the 'type' of labour (ie the nature of the 
> use-value of the commodity it produces) is *not* relevant to whether 
> that labour is productive or not.
That isn't quite what I said, but since I said quite a lot, I guess some
could get missed. 
I said that there are two distinct contexts in which the question arises: 
(a) Labour that does produce use-value but not exchange-value (servants,
state labour, etc). First part of TSV I, Murray's category (1)
(b) Labour that does not produce a use-value at all (banking, perhaps
security, etc). Last part of TSV I, Murray's categories (2), (3) and (4).
In case (a) the type of labour doesn't matter (services versus material
production). What has to be established, is whether it is 'exchanged
against capital'.
In case (b) the function of the labour is important, since it matters
whether it produces a use-value or not.
So I think I do stick to the principle ('nostrum'), but apply it (as I
always said) in the context for which it is intended. As does Marx.
> I can see why such labour, 
> either all or in part, is wasteful from the point of view of some 
> hypothetical society rationally and democratically  organised to 
> allocate resources so as to meet human wants and needs. The use-value 
> of the commodities it produces are humanly undesirable (at least in 
> excess) - but then so are the use-values of junk food, cars unsafe at 
> any speed, products that cause the death of 50% of their consumers 
> when used as recmommended (tobacco products), etc. But the labour 
> that produces these commodities is not, thereby, unproductive.
This is the core of the question, and all I can comment is that for
me it is a correct summary of the difference. And that's as far as we can
go. I don't expect to convince you, and I don't expect you to convince 
me. What I do think is that we have clarified the difference and here it 
is: 
view (a)[you] Security guards and Bank workers are productive because 
they create a new use-value 
view (b)[me and Marx]: they are not productive, because they don't 
create a new use-value. 
You're entitled to your view, I'm entitled to mine, Marx is entitled
to his. I don't think the question is whether Marx was 'right' and it 
is probably forlorn to think that this list can settle that question. 
It can only be decided in relation to a much wider enquiry. What is 
relevant and useful are the following questions:
(b) what in fact was Marx's view?
(a) was it coherent/consistent?
Juriaan is toying with the idea that Marx wasn't consistent. I'm not 
clear why he thinks this but I hazard a guess that if one believes 
the capitalists can simultaneously pay the same sum out of both 
revenue and capital, no-one else is going to appear particularly 
consistent. 
I have heard no evidence that Marx is inconsistent or incoherent. I've
heard a lot of disagreement with him, which I have no problem with. 
As long as disagreement doesn't pass into dissing.
As far as I can see Marx is very coherent and sticks to the principle
that if it's productive, it must make both exchange value and 
new use-value:
case (a) above are people who make use-values but don't make exchange
value because they don't sell the stuff to a third party (a capitalist)
but direct to the user.
case (b) above are people who genuinely don't create any new use-value
in the first place.
Marx  sticks to these two (compatbile) ideas throughout, and applies 
them with tediously ruthless consistency. One can disagree with it, 
and many do, but I can't see that it's inconsistent.
Marx has one view, and you have another. I don't have a problem with 
that, as long as it is clear which is his and which is yours.
What do we do? It is very straightforward. We show how the various possible 
views translate themselves into data, so that subsequent students who wish 
to investigate theories in various frameworks - including students who want 
to investigate theories in Marx's framework - have at their disposal data 
that is properly constructed to reflect the theoretical presuppositions 
of the framework concerned, instead of trying to test Marx on data he 
would have doubts about using in the toilet.
Where you (or others who share your views) differ, we both agree to
produce variants of the same data, showing the effect of these
differences, so that the reader is left free to judge for her or 
himself. Economics is too important to be left to the economists.
I note that there are a number of people who agree both with myself on
this, and I propose that we work together to produce transformed
accounts. Since Marx is dead, we can't work with him yet, and so I restrict 
my appeal to the living and propose that we take responsibility for 
our own views, our own results, our own problemsm and our own errors.
And that's it!
Alan