I've found the 'assumptions' discussion disturbing. Just as
we are reaching clarity, understanding and even agreement
there seems to be a lot of heat whose source is unclear. I
hope my response is taken, as it is intended, as a positive
attempt to explore reasons and suggest responses.
Jerry [3369] replies to Andrew [3386] thus:
"I am going to have to put on my moderator hat: Your
response to Paul C was way out-of-line."
I don't think Jerry really was wearing his moderator hat or
he would have said this when Andrew actually did respond to
Paul. My take is that he, Jerry, was offended by Andrew's
polemical style, which as a listmember he is perfectly
entitled to say. As a fellow listmember I'd like to discover
why he was offended. I hope this doesn't turn out too much
like probing a sore tooth.
I think the problem lies more in the nature of the
differences than the style adopted. Sparks, if not flames,
have been blowing in more than one direction. In [3314]
Jerry writes:
Do you want your model -- oops, illustration -- to say
anything at all on a *theoretical* level about
accumulation and the FRP? If your answer is no, then
the whole exercise is useless except possibly as a
critique of Okishio. If your answer is yes, then you
have to construct an illustration in which accumulation
is at least *theoretically* possible.
I thought this was polemical excess and didn't respond. I've
reconsidered. There is an underlying reason for the heat and
if it's starting to flame we should find the fire.
Although I'm not convinced etiquette is the real problem, it
could be useful to clarify this general question. I suggest
that when any of us feel a remark is objectionable we should
try and frame a regulatory principle that could avoid the
problem in future. The evidence is that everyone on the list
hears criticism and adapts their style accordingly. Tapping
this reservoir of goodwill requires more than an occasional
remonstrance. It is a job for all of us on an equal footing
to formulate agreed rules of conduct.
In this spirit, here's my thoughts about some guidelines
relating to my own concerns about the 'assumptions' debate.
I haven't found this debate very productive. I suspect two
discussions are treading on each other's toes. I feel we
need a constructive disengagement so that each discussion
can be held in its own separate space.
First off, I think [3314] trivialised our position. I know
Jerry doesn't rate the distinction between a model and an
illustration - but we do. I think it an important principle
of debate - maybe the most important - that no position
supported by rational argument be ruled out of order. I feel
[3314] violates this principle. To me it says a distinction
between model and illustration is meaningless or dishonest.
Ridicule has its uses, and if Andrew's was a new distinction
that had just surfaced it might have sharpened the issue.
But this question has been chewed over and the differences
are quite clear. No-one insists list members agree with our
distinction but I think they should accept that we hold it.
It is possible to respect a view without endorsing it. We
don't have to respect everything, but a collective needs to
respect the constructions which its members need to express
their views or they won't be able to express them.
I try to apply this - perhaps unsuccessfully - to others
such as Gil and Paul Cockshott with whom I clearly disagree
on certain topics. We had some sharp exchanges and through
those I hope I have understood their views enough to ensure
I don't interfere with discussions they might want to hold
on the basis of these views. I'm not making any claims to
sainthood: I'm trying to enunciate a principle that can take
us forward in future.
Second. I was disturbed at the phrase:
"useless except possibly as a critique of Okishio".
My purpose happens to *be* a critique of Okishio. I'd prefer
it if the moderator didn't treat this as a candidate dustbin
for the terminally disfunctional.
We're having an intense debate about the critique of Okishio
with wide and focussed participation. I'm carefully not
saying anything about its intrinsic merit: the point is that
some of us are interested in it.
I like clear debate, and the odd polemical flourish does
clarify things, but there's a crucial gap between saying a
view is misguided, erroneous or contradictory, and implying
it is not worth expressing. I hope I've never said any view
on OPE is useless and if so I was wrong and have no objection
to being asked to admit it. I also beg pardon.
This stated, I suspect there is a deeper problem concerning
the 'relative autonomy' of debates based on differing
assumptions. Existence is not enough. OPE needs space for
discussions which *do* adopt simplifying assumptions that
vary from discussion to discussion and I think it dangerous
to introduce any principle which might inadvertently limit
this. What I hear in the assumptions debate - not just from
Jerry - risks crossing this line.
I am in no doubt that Jerry considers the assumption V = 0
to be wrong. The fact that I haven't replied to him does
*not* mean that I haven't heard him. I hear it quite clearly
and I've considered it very carefully.
But there is a clear discussion going on between people who
don't share Jerry's objections. They seem to have reached a
tentative agreement to discuss on the basis of a number of
assumptions which include some he objects to.
If there was no such agreement, discussion on ground rules
would be both healthy and inevitable. But it seems to me
once a discussion is under way on the basis of some general
limiting assumptions, we should be prudent about disrupting
this general agreement.
I would find this debate easier to follow, more focussed and
less demanding of our time, if it could proceed on the basis
of agreements already reached, whether wrong or right.
We need more room for the suspension of disbelief. We need
the kind of 'what if' argument that says 'OK, I don't agree
but let's explore the consequences anyhow'. I think Duncan
has been particularly thoughtful in introducing this mode of
discussion into OPE and could all learn from him. I find
this quite absent from normal academic debate and I think it
needs to be cherished here.
OK those are my concerns: how can we ensure Jerry's concerns
are met? For me the first step is to establish what they are.
I can only try and state how I perceive them. Since I'm sure
I'll make mistakes, author's corrections to the Freeman
interpretation of Levy are welcome. This is:
1)we can only model reality if our assumptions are all
realistic.
To this I have few objections but it is spilling over into:
2)we can only discuss reality if our assumptions are all
realistic.
This only follows from (1) if we also assert
3)the only way to discuss reality is to build a model of
it.
Problems are pretty well bound to arise if (3) is injected
into discussions that violate (2). The implicit conclusion
is that these discussions are not legitimate. Some brusque
responses to this idea are only to be expected, especially
if it comes from the list moderator.
I don't object to discussing this view but it is not prudent
to insist on it. It confuses the hell out of debates where
the other guys just want to simplify a few things so they
can chew the fat. It's not wise to demand every discussion
leads to a model, even if you think no other outcome can be
useful. If I think something not very useful is going on my
usual practice is to stay out of the party until the noise
gets too loud.
I feel the issue of models - and Jerry is by no means the
only person who wants to use them - is sufficiently distinct
to warrant a separate thread. But I think this has got mixed
up with other discussions which would profit from being
disengaged from it.
I share blame for this. When I wrote the 'essay on Okishio',
I gave the wrong impression. I wasn't trying to formulate
the position that a theorem cannot be discusssed unless its
assumptions are valid. I was trying to deduce the *domain
of applicability* of the theorem from the assumptions.
I was trying to do something model-building makes quite
difficult, which is to establish the prior abstractions the
discussion presumes. Since my aim is to establish them, the
last thing I want is to rule them out of order. As far as I
am concerned, *all* simplifications can go in the pot. V=0,
V = constant, maximum profit rate, constant real wage, s/C
or s/(C+V), you name it, you got it. I want to discuss
them, not silence them. For the record, I disagreed with
Andrew's response to Allin and Patrick on the maximum rate
of profit, which I think is a useful construct for the same
reason I think V=0 is useful for discussing certain ideas,
namely what is the exact domain of applicability of Okishio.
Okishio's theorem is profound and beautiful.
I think I have been wrong not to make this clearer, but
perhaps I can rectify that here. I sought to establish two
cardinal points:
(~E) Equilibrium is not real: a system whose abstractions
are derived from stationarity cannot correspond to any
possible reality.
(D) Reality is dynamic: any real market necessarily
possesses dynamic properties, in the strict sense that
the state of the system at any time depends, and only
depends, on the state at previous times.
Okishio follows from (E). It does not follow that Okishio
tells us nothing or is useless. It only follows that it
cannot be applied as a theorem about reality or a refutation
of Marx.
But the literature - including Okishio's own assessment of
his own theorem says the theorem applies to reality and
interprets Marx accurately. In my book this makes it a false
theorem. I take any theorem to include its proposed domain
of applicability. The theorem 2 + 2 = 1 is valid for the
field of numbers modulo 3 but not for counting things.
I failed to state more clearly the domain of applicability
of simultaneous theory. Let me say now: I think nearly all
fixed-point (simultaneous) theorems are valid theorems about
ideality.
I also think Marx was very concerned with ideality. I think
that is why his major economic works all contained the word
'Critique' in the title (Zur Kritik, Grundrisse der Kritik,
Das Kapital: Kritik der politischen Oekonomie, usw).
I *don't* think theorems about ideality are useless. I just
don't think they are theorems about reality. Okishio tells
us the capitalists can never get the rate of profit right;
i.e. rational expectations cannot happen. I think that's a
damn important result, but not to be confused with a correct
prediction of profits. Sraffa produced a tremendous (sorry
Riccardo) critique of neoclassical price theory. Things went
wrong when his supporters took this for a derivation of
actual prices.
Fixed point or simultaneous theorems inform us in my view
about the internal structure of capitalist *thinking*, which
for various reasons I think is necessarily simultaneist.
This is hardly surprising: we can *think* a dynamic process
in almost any order we choose because our thought is free to
roam through all of time. The actual process, however,
happens in a definite sequence. It is quite natural that
logic tends to jettison historical movement; but the result
tells us more about our thought process than about the world
out there. However I do not think it is a futile enterprise
to study thought as such. Indeed, for social science it is
indispensible because the 'substance' of social relations is
objectivised thought.
I think Marx perceived this. I find his work superior to all
positivism because of the constant dialectic between reality
and ideality; between a temporal, historical actuality and
its static, eternalised perception by capital.
Naturally and perhaps unfortunately, it is quite easy for
vulgar economics to take Marx's discussions of ideal forms
of consciousness out of context (such as the concept of an
actually equalised profit rate and the vulgar concept of the
'cost' of the 'factor' capital to which it leads) and read
them as if Marx asserted that these really happen.
I'm sure few will dispute that Marx perceived the ideal is
determined by the real. But it is not adequately recognised
that this necessarily implies that the ideal is less rich
than the real. Otherwise we could equally well explain the
real on the basis of the ideal, as the ideal on the basis of
the real.
Underlying the idealisation of an equal profit rate is the
reality of a general profit rate. The real consists of
countless actual rates of profit, but the ideal is only one
hypothetical rate which the capitalists baptise the 'cost of
capital'.
This subtle issue should not be reduced to a crude dualism
between non-ideal reality and a pure ideal 'model' because
the real *contains* the ideal. Ideas are held by material
beings. The Ideal equal rate of profit exists, not just as
the Platonic average of the actual rates of profit but in the
heads of Real capitalists.
We cannot place our own thoughts outside of reality because
we cannot prevent these ideas and thoughts from influencing
this reality, and because these ideas and thoughts are
objectively conditioned not just by their correspondence to
reality but because we, the thinkers, are real and actual
material beings with a past and a place in society which
conditions our thoughts.
This is why we cannot accurately conceive of reality, above
all social reality, as a kind of pure Kantian Being-in-self
independent of thought. I think the failure to comprehend
this is the fundamental error of the model-building
approach. If we forget that the thinkers themselves are part
of the material world, the result is necessarily partial.
When we think about the world we must include our own selves
and thought in this thinking. The price is paradoxes and
contradictions. Dialectical thinking consists in confronting
these instead of trying to dispense with them. They arise
from the nature of consciousness itself.
Hegel sums this up in the profound idea that philosophy
consists of thought thinking itself. But this seems to me to
lead to the peculiar idea that the world is the realisation
of its concept. This requires that the determinations of
thought be more complex than the determinations of the
world. I think he should have put it exactly the other way
around; reality consists of matter becoming conscious of
itself. A truly materialist science would identify which
properties of matter make this possible. Its starting point
would be the fact that thought is the finite product of an
infinite reality.
That is why I sought, in formulating my argument, to re-open
the old discussion about reality and ideality, objectivity
and subjectivity, and show it has a mathematical expression
in the debate between successivism and simultaneism. I wanted
to show how it could come about that the capitalists' ideas
were insufficiently internally rich in determinations to be
self-determined; so that on the one hand these ideas could
be explained on the basis of an objective reality richer in
determinations than capitalist thought (because dynamic),
and on the other that these ideas for this reason were
inadequate as a theory of reality, because of the loss of
determinations.
A very powerful result that follows from this is that we
cannot explain the real world from the behaviour of agents
operating on the basis of these ideas. Hence and finally, a
subjective account of economics is a logical impossibility.
I think we need this kind of general theorem to confront
subjective economics in its original, Austrian form.
At the time I thought that there was an important agreement
with Jerry on this general area but I now see that I was
mistaken, and that we are really talking about different
things, with a coincidental (though useful) agreement that
the assumption of equilibrium is unreal.
I think Jerry is interested in something different from us.
His position seems to be, with nuances, a general criticism
of nearly all deployment of simplifying assumptions in
economic analysis, related to what seems to be the idea that
the goal of economic theory is the construction of models.
I respect this outlook but I don't share it. My objective is
not to build models, nor have I any generalised objection to
assumptions, nor even any inviolable objection to any
particular assumption, even price stationarity or profit
rate equalisation in the correct context. My concern is not
whether assumptions are wrong in principle but (i) what
conclusions do they impose on their thinkers and (ii) when
we can and cannot use them.
My objections to simultaneism do not centre on the realism
of even very fundamental simplifications such as lack of
price stationarity and profit rate equalisation. My
objection is that it derives its categories - value, price,
profit, money and causality - from these assumptions, and
does not admit that it has done so. They are not really
simplifications at all, but disguised axioms. Therefore they
can never be dropped; and any conclusions derived from them
can only ever apply to systems for which these axioms are
valid. But this excludes all real systems.
Reduced to its bare minimum, I object to the claim that the
domain of applicability of such systems includes the real
world. And, hence, that they are an adequate representation
of either value theory in general or Marx's value theory in
particular.
Superficially, this chain of argument sounds quite like some
of the points Jerry has been making but I think it is really
very different. Jerry [3345] writes
"quite frankly, I am amazed at the resistance that you
have towards dropping this assumption [V=0]".
Terry Pratchett calls this 'two minds backing up a one-way
street.' I think we are talking at complete cross-purposes.
There is no resistance to dropping this assumption. We drop
it all the time. It just makes it easier to talk to other
guys. The issue is not how to make theories; it is how to
have discussions.
Jerry comes across as objecting to this *discussion*. Some
discussions have to be stopped, but I don't think this is
one of them.
Please Mister moderator, can we have our discussion back?
Alan