RE: [OPE] intermission: value of knowledge

From: Paul Cockshott <>
Date: Mon Nov 30 2009 - 05:20:34 EST


Take Paul C's example of the manufacturing of paper, a process that reduces
the entropy of the raw material (wood pulp). Note, first, that this process
involves concrete labor - it is the process of changing one use-value into
another. Second, that although entropy is being reduced, value is being
These two points suggest to me that the process of increasing (or, in
this case, reducing) information in the sense used here belongs to the
creation of use-value, not value.

You are right that entropy is reduced and value is added, but I am trying to answer a more fundamental problem in the opening chapters of the book :
1. why is it that labour is possible and How is it that labour can reshape parts of the real world.
2. why is it that labour creates value.
3. why is it that human labour, unlike the labour of horses, has proven indispensible.

In the process of answering these questions one has to ask some fundamental questions about the relationship between work in the physical sense developed by Watt, Labour in the sense used by Smith and the classical economists, entropy as discussed by Maxwell and Boltzmann, and information as described by Shannon and Chaitin.

If we take a typical mass production process like the manufacture of cars or newspapers, there are two phases involved. The first phase involves producing a low entropy material : sheet steel or paper, from high entropy material : rough cast ingots or paper pulp. The laws of thermodynamics imply that this reduction in entropy can only be achieved at a cost in energy. Although the amount of labour required to produce a sheet of steel or a sheet of paper declines as manufacturing technology improves : ( consider hammering by hand as opposed to a steel rolling mill ), there remains an energy cost. The labour for this phase comes to be dominated by the labour of energy production and that of machine supervision rather than direct human expenditure of energy as in the pre-industrial period.

The second phase of manufacture involves giving shape to this raw material : using a hydraulic press in the case of car body, or a printing press in the case of a newspaper or book.

Now the shape of the die for the car, or the sequence of letters on the solid type used to produce the newspaper, will themselves contain information, and it is this information that gives the product its usevalue. A newspaper with no type on it, would obviously be useless.
Less obviously, a newspaper that contained the same text every day, whilst not useless, would be less usefull.
In both of these cases, Shannons information theory indicates that these would contain strictly less information than the actual newspaper whose edition changes each day.
The maximum information would actually be conveyed by a newspaper with an apparently random sequence of letters and digits, but, since our normal language perception depends on there being substantial redundancy in text, this would again be less useful to us. In specialised applications however, such as the transmission of newspapers over the internet, the transmission of enciphered messages etc, this maximum entropy density representation actually turns out to be very useful.
But, third, note also that paper-making is not part of the 'knowledge
economy' as most business people, consumers and economists understand it. Of
course, in a sense, all economies are 'knowledge economies'. But this is not
the sense that interests us here. We are interested in economic activities
that do not produce material things - the writing of the novel as opposed to
the making of the book, etc.

Now the key point about copying technologies like printing or pressing by die, is that the labour expended on the pattern can be amortised over many copies. My contention is that what we call the modern information economy is a generalisation of a principle that has been apparent from the very first phase of industrial capitalism, one which Babbage called attention to in his book ' Economy of Machinery and Manufactures '.
The writing of a novel does produce a material thing : the draft that is sent to the printer. The making of a pattern by a master pattern maker in a casting shop is an analogous process, it is only the prejudice of the intellectual classes that sees it as something distinct.
There is still some justice to Smiths remark:

Adam Smith
The difference between the most dissimilar characters, between a philosopher and a common street porter, for example, seems to arise not so much from nature as from habit, custom, and education. When they came into the world, and for the first six or eight years of their existence, they were perhaps very much alike, and neither their parents nor playfellows could perceive any remarkable difference. About that age, or soon after, they come to be employed in very different occupations. The difference of talents comes then to be taken notice of, and widens by degrees, till at last the vanity of the philosopher is willing to acknowledge scarce any resemblance.

It appears that the limitations of the 'quantitative' notion of information
are also an issue for philosophers. The 'Information Theory' entry of my
Routledge Concise Encyclopedia of Philosophy has this to say:

"The information studied by Shannon is sharply distinct from information in
the sense of knowledge or of propositional content. It is also distinct from
most uses of the term in the popular press ('information retrieval',
'information processing', 'information highway', and so on). While Shannon's
work has strongly influenced academic psychology and philosophy, its
reception in these disciplines has been largely impressionistic. A major
problem for contemporary philosophy is to relate the statistical conceptions
of information theory to information in the semantic sense of knowledge and
content" (the author of this entry is Kenneth M. Sayre).

More details about all this are at

ope mailing list

The University of Glasgow, charity number SC004401
ope mailing list
Received on Mon Nov 30 05:24:42 2009

This archive was generated by hypermail 2.1.8 : Wed Dec 02 2009 - 00:00:03 EST