From: Ian Wright (ian_paul_wright@HOTMAIL.COM)
Date: Mon Nov 17 2003 - 16:56:41 EST
Paul wrote: >Such constraints based theories are also commonplace in >physics. Both quantum and classical mechanics have the >same structure at the level of abstraction you define. > >One has to be careful when one suggests that cause is >something other than constraints plus boundary conditions. Jerry wrote: >This implies that Marxian theories where there are >sets of simultaneous equations are also not causal >theories. Do you agree with that inference? > >Put in the context of some of the many discussions >that we on OPE-L have had over the years, does this >mean that your conception of value is temporalist (rather >than 'simultaneist')? Ajit wrote: >A theory of value is essentially a timeless >problematic. A dynamic theory of value is bunk. A >problem of asertaining the reasons for changes in the >price of a commodity over a period of time is not a >theory of value but rather an entirely different >problem. Let me try to answer all these issues at once. There are so many interesting and deep issues here that I won't be able to say anything completely satisfying or comprehensive. But I'll have a go. Pure constraint-based theories break down when the constraints are contradictory: they don't have a lot to say in such situations. For example, consider tiling a plane with a set of tessellating shapes. The contraints are the set of shapes. For some sets of shapes, the constraints are satisfiable, and lead to a solution which tiles the plane. For other sets, the constraints are unsatisfiable, and the plane cannot be tiled. So we can develop a constraint-based theory that can identify sets of shapes that are impossible to tile. But consider that we are not engaged in logical possibilities for tiling the plane, but engaged in modelling actual causal occurrences that occur in historical time, for example, modelling a child trying to tile the plane given a set of shapes that happen to embody contradictory constraints. Now the child has a strong desire to tile the plane, but does not know our constraint-based theory, and therefore does not know its desire is unrealisable. So what happens? The pure constraint-based theory has nothing to say about what happens dynamically, only that a complete tiling of the plane will never occur. The constraint-based theory does not model the mechanism by which the constraints are satisfied, which in this case is the child's strivings, so this kind of theory in principle cannot tell us what will happen, cannot theorise the unfolding dynamics. (Clearly, the child will try lots of different combinations, may repeatedly try different solutions in a cyclical manner, come nearer to tiling a large part of the plane only to fail in some corner, retreat and repeat and so on). Now it seems to me that the status of Sraffa's theory is analogous to the situation just described. It is a constraint-based theory of the economy, which identifies logically possible (and logically impossible) economic configurations. But it has nothing to say about dynamics, because it does not model the mechanisms that try to satisfy the constraints (it does not model the child, it only models the problem). In my view this is a severe limitation of the theory, as I'm interested in modelling what actually happens over time. But a further limitation of this theory, or at least a limitation of how it has been used, is that it does not admit of contradictory constraints, for example Marx's aggregate equalities. I don't want to argue whether Marx's equalities do or do not hold. Instead I want to point out that some interpreters of Sraffa's theory deny the existence of contradictory constraints in reality, for no other reason, it seems to me at least, than sets of overdetermined equations cannot be solved (or they think they cannot be solved). Now if there are economic mechanisms that are in contradiction (or opposition) to each other, and strive to attain mutually incompatible configurations of the economy, then a constraint-based theory will be unable to theorise the resultant dynamics, and instead will simply (and correctly) state that the constraints cannot be satisfied simultaneously (just like the tiles and the child). But unfortunately some interpereters of this theory go further and commit a logical blunder: that because the constraints cannot be satisfied simultaneously then the underlying mechanisms that generate those incompatible constraints do not exist, or cannot exist together. For example, denying the law of value is an active mechanism, or denying that there is a tendency for profits rates to equalise. The fallacy consists in reducing transfactually active mechanisms in open systems to the effects they generate in closed systems. For example, reducing the mechanism of the law of value to the definition of value as embodied labour, or reducing the mechanism of the equalisation of profits to the assumption of equal profit rates. This is a real blunder, particularly as it seems to me that the interesting dynamic questions can only arise in theories that admit of contradictory tendencies in the economy that interact over time. Paul: I take your point. It depends on what we include as constraints, and what we include as solution methods. The matrix inversion solution methods that solve Sraffian economies clearly do not correspond to the dynamics of an evolving economy, dynamics that could be viewed as an unfolding solution to the same set of constraints. The difference is between a logico-deductive model and a causal model, between modelling the problem and modelling how the child solves the problem. Jerry: In general I think that models that employ simultaneous equations are not modelling dynamics, and are therefore unlikely to get the causality right. But I really don't want to deny the utility of static, simultaneous modelling, just the kinds of conclusions that may validly be drawn from them. I do think that the concept of value can only be fully understood in a dynamic context, so I suppose that makes me a temporalist. Ajit: I think you are compartmentalising things that are in causal interaction with each other. For example, I have a computer model of a simple commodity economy. The causal interaction between labour values and prices results in an efficient allocation of social labour time, at which point prices are proportional to labour values. Irrespective of whether this is a good or bad, relevant or irrelevant model for understanding real economies, it is nonetheless a real dynamical system, which can be an object of scientific study. It just so happens that I try to understand it by employing ordinary differential equations, not simultaneous equations. This leads me to explain what happens dynamically in terms of a theorem that states that labour values are global attractors for prices. Does your philosophy rule out the existence of the object I have constructed? Does it rule out my explanation for its behaviour over time? Or does it have nothing to say about it? Presumably the causal interactions that are instantiated when the model is run are a figment of my imagination? I would really like to know if I've been labouring under such a delusion! -Ian. _________________________________________________________________ The new MSN 8: advanced junk mail protection and 2 months FREE* http://join.msn.com/?page=features/junkmail
This archive was generated by hypermail 2.1.5 : Thu Nov 20 2003 - 00:00:00 EST