The following text has been automatically reproduced by an Optical Character Recognition (OCR) algorithm. It may not have been checked over by human eyes. For matters of precision please consult the original pdf.

In Defence of Epistemology

R2~

If I believed in the imminent collapse of capitalism

from forces internal to the economy (falling rate of
profit etc), then my evaluation of movements would
be affected by this. But Hindess and Hirst, while
pointing this out, insist that my calculation of
consequences; is itself a function of my political
position (evafuation). So we are in a messy circle.

Contrary to them, it ,might be thought vital for
political activity that predictions be minimally
contaminated by wishes (‘pessimism of the
intellect’: Gramsci). Here again Hindess and
Hirst’s contempt for empirical reason lands them
in gross subjectivism – a politics of wish-fulfilment
“R1

Hindess and Hirst are right to stress the distinction
between the ORIGIN and the political tendency of
ideas and ideologies, and right to highlight the
double talk that has tried to conceal this distinction
in the name of preserving dogmatic versions of
‘class analysis’. But they offer a pragmatist
reductionism in place of a class-origin reductionism. Thus, for example, in denying that
‘bourgeois ideology’ is ‘bourgeois’ in virtue of its

origins they rure oUt of account the possibility that
some aspects of the bour.geoisie’s class outlook
might be progressive even true. Or rather they..

insist on their own doubletalk which would denY’its
bourgeois statU8; in virtue of its supposed serviceability to socialist revolution.

Generally spealdng, a pragmatist reduction fails
to avoid the epistemic issue. For the question
always arises: will this line have these (desirable
or ,undesirable) effects? (See my reply to Peter
Binns RP3; and see Andrew Collier’s ‘Truth and
Practice’ RP5). Moreover, a politics, which not
only downgrades questions of truth (Machiavellianism) but systematically seeks to extrude such
issues from its frame of reference, must, I
suggest, be a politics of contempt – a practical
anti-humanism. The extrusion can never be
achieved. Always it wiil be a matter of hiding uncomfortable truths or promulgating useful fictions.

It is handy, however, to think that, outside one’s
system, no justification need be sought for one’s
beliefs and one’s practices.

IN DBFBNCE or BPIS’I’BMOLOGY·
Andrew Collier
My aim in this paper is to criticise a postAlthusserian tendency which urges us to ditch the
whole project of epistemology; I shall also say
something about the conditions for an epistemology
which will not lay itself open to the objections
raised against epistemology by this tendency insofar
as those objections are valid. And I shall make
some brief comments about some of the outstanding problems for an epistemology which is to cope
with the human sciences – problems which are not
made to disappear, but merely evaded, by the
rejection of epistemology.

My task is therefore a polemical one – the defence
of what I regard as already established pOSitions of
materialist epistemology, against new versions of
idealis m, albeit shamefaced (or as they say in the
trade, ‘de-negated’) versions. (1)

1) Epistemology after Althusser
‘The identiffcation ofwhich ‘I spoke in introducing
(the Hegelian Marxists ‘) work – of the problem of
the unity of theory and practice and the problem
qf the relation between science and its object is
an invalid and illicit conflation of questions of
quite a different order. The first problem is the
fundamental problem of Marxist politics: how to
give ideas a material force. • .. The second
problem is an epistemolOgical one: how to guaran-I
i I admit to fee’ling that iUs somewhat Shameful to be re-iterating this
position at a time when we need to break new groundtn scientific epistemology, and when philosophers such as Roy Bhaskar are doing so. But it
is a shameful necessity, for while idea).1sm enjoys a revival unparallelled
since Edwardian times – and precisely among self-styled radicals … really
new knowledge will be prevented from having the political effects that it
deserves.

What is most disconcerting is that modern idealists are not only unaware
that their sopbtsms have long since been refuted; they are even unaware
that th8y are idealists. Who can’doubt that, if Bishop Serkeley had been
alive today, he would have re-titled h1s major work ‘Towards a
Mater1alis~ _Theory of Perception r?

tee that a theory does in fact provide a knowledge
of the reality it claims to explain. ‘

(Alex Callinicos, Althusser’s Marxism, pp22-23)
Part of the value of Althusser’s work in epistemology is that he has cleared up this ‘historical’

confusion. Theoretical production is itself a practice with its own criteria of success, not a mere
effect of other – economic, political or ideological practices. The political question is then, not the
relation of theory-in-general to practice-in-generaI”
but rather the problem: how to secure the transformation of the ‘political’ practice of class struggle
from an economistic and reformist one (which it
will spontaneously tend to be) into one which raises
the issue of state power, through the intervention
of Marxist theory into that practice. In this context,
for Althusser as for Lenin, the unity of theory and
practice is not a theoretical given but a practical
task. The Marxist theory which must be united with
the class struggle is not itself a mere epiphemomenon of that struggle; its relation to the struggle is
that it yields knowledge of the society that generates it; and because it does so it enables the
workers’ movement to fight clearsightedly, without
the blinkers of bourgeois ideology.

Yet, surprisingly enough, Callinicos goes on to
criticise Althusser for not realising that he has
provided the basis for abolishing epistemology
-altogether. if he ‘has done so~ what comes of his
anti-historicist work which1:ook place within
epistemology ?

There seem to be three points of departure for the
anti-epistemologists in Althusser’s thought. Firstly!

there is his contention – through all phases of his
work – that the criteria of validity of a theoretical
practice are internal to it. This is said to rule out
any general criterion; and epistemology is said to
)e precisely the pursuit of general criteria. Here

I thi~ a correct point and an incorrect one have
been confused. It-Is perfectly true that the procedures of validation are quite different in different
sciences; one doesn’t prove propositions of chemist·
ry, evolutionary biology and linguistics in the same
way. But in each case what the science aims at is
knowledge of its object; the differences stem from
the differences in the objects; what the sciences
share is that they are all attempts to get at the
truth about their, respective ~bjects. In their timehonoured formula, they aim to produce a theory
which corresponds to its object. And the nature of
their procedures will be determined by this project.

This is important because there are theories whose
procedures recognise no such constraints, and we
need to distinguish between these pseudo-sciences
(astrology, palmistry, etc) and genuine sciences.

Secondly, there is Althusser’s definition of
philosophy as ‘class struggle in theory’. With this
definition Althusser rightly repudiates his earlier
view that philosophy can be a sort of masterscience, ‘the theory of theoretical practice’. But
this is not a rejection of epistemology; it emphatically doesn’t mean that philosophy should abandon
its commitment to objectivity. On the contrary, if
it is partisan, it is the partisan of ‘the materialist
thesis of objectivity’. To defend the sciences is to
engage in class’ str~ggle in theory, because ‘true
ideas always serve the people; false ideas always
serve the ene mies of the, people’ (Althusser, Lenin
and Philosophy, p24). Only the exploiting classes
can gain from ideological error and ideological
obstacles on the path of the sciences. Hence philosophy still has the epistemological function of
“‘drawing a dividing line” inside the theoretical
dom,ain between ideas declared to be true and ideas
declared to be false, between the scientific and the
ideological’ (Lenin and Philosophy, p61).

Thirdly, there is Althusser’s Essays in SelfCriticism, where he does become embarrassed
about his earlier epistemological terminology; but
if he turns his back on epistemology here, it is to
embrace historicism as a long-lost brother. So
long lost perhaps thathe fails to recognise him, but
that shouldn’t stop’,us from doing so. For in a cryptic footnote on p124, ‘he tells us that a materialist
interpretation of epistemology ‘could lead us to
study the material, social, political, ideological
and philosophical conditions of the theoretical
“modes of production” and “production processes”
of already existing know ledge: but this would
properly fall within the domain of Historical
Materialism!’ But Historical Materialism can
only study science as ideology, and the results of
such study will be strictly irrelevant to its truth.

For considered as one social practice among
others, science can indeed legitimately be studied
as ideology, as located in the superstructure where else? The ideological roots of the discoveries of Newton or Darwin or Marx or Freud can· be
laid bare without invalidating those dis coveries.

The science/ideology distinction is an epistemo-logical one, not a social one (2).

2 Althusser does place science outside the superstructure. But it is not
necessary to do So in order to avoid historiCism, as I hope my next paragraph shows. The ‘scientific community’, its institutions and practices,
its relation with various state and economic apparatuses’ etc are quite
obviously part of the social formation in question, and as such part of the
object of historical materialism. I am sure that Althusser did not intend
his denial of science’s place in the superstructure to deny this, but it can
only give that impression. When he finally asserts it, it is to sell the
pass to historicism.

It is extraordinary how strong is the prejudice that if one practice can be
known by virtue of another, the former loses its autonomy and is expla1ned away by the other. I can only assume that there is some primaryprocess thinking here – perhaps an infantile identification of knowledge
with eating.

It is impossible to think the relation of science
to other social practices by means of the concept
of relative autonomy. Considered as one social
practice among others, with miscellaneous economic, political and ideological relations to other
social practices, science has no special autonomous status. Considered as the appropriation of
the real in objective knowledge, however, it is
radically and categorically autonomous. For as
such it is defined by its norm of correspondence
to the r~al, and insofar as it achieves this norm,
it succeeds, whatever its genealogy. Insofar as
we want to know about the reality of which a
science speaks, the social, psychological and
linguistic accidents concerning the practice of that
science are simply of no interest, any more than
the colour of the scientists’ hair. It is therefore
as true and as misleading to say that a science is
also an ideology as to say that a human being is a
pile of chemicals or that Beethoven’s ninth symphony is a succession of noises. And indeed as false,
remembering that the same human individual will
be at different times composed of different molecules, and that many successions of noises have
been performances of Beethoven’s Ninth. Scientific
know ledge doesn’t change its nature when the
writings embodying it are translated from German
into English, or when it is applied in the service
of the proletariat rather than the bourgeoiSie, or
when the experiments that test it are performed in
Peking instead of New York. The reduction of
science to ideology is not a misidentification of a
social institution, it is a category-mistake.

But that itsn’t the end of this shocking affair,
for if Historical Materialism were granted the
right to pass judgement on the sciences by virtue
of the fact that they fall within its object as social
practices, the other sciences could claim equal
credentials for the role of master-science. Psychology, because the thinking of scientists, like any
other mental process, is subject to its laws;
linguistics, because scientific discourse is produced in accordance with the rules of language;
logic, whose laws it must obey if it is to be
coherent thought; evolutionary biology, for the
capacity for scientific discovery is an adaptive
species -specific trait of human beings; and – to
reduce the whole thing to its absurdity – physics,
because SCientists, laboratories, textbooks etc
are after all compoS!ed·pf atoms.

So the requisite critique of Althusser’s epistemology can’t be that which he himself provides; he
exorcises one devil – ‘theoreticis m’ – that seven
devils may enter. Rather, the critique should be

9

directe~ at what Timpanaro calls his ‘Platonism’,
and should reinstate the notion of putting questions
to nature (3) as the characteristic of scientific
enquiry. Because this notion involves the idea that
the results of scientific validation-procedures
must be causally dependent on the nature of the real
object, it may appear to fall under Althusser’s fire
against genetic accounts of knowledge. Althusser’s
anti-geneticis m is quite correct; it is a mistake to
try to determine the nature of something by reference to its past history (cf. ‘degenerate workers’

states ‘). Experiments though are not necessarily
the means by which scientific theories originate;
but they are the means by which they are tested,
and as such belong to the ‘synchronic structure’ of
a science. In the final section of this essay I shall
say something about the problem of applying such
a notion in the human sciences. First I shall consider an alternative line of ‘immanent critique’ of
Althusser – one which moves further still from the
taint of the empirical, and compounds Platonism
with Kantianis m.

2:i

.Road to Kaliningrad

‘Lenin criticizes Kant’s subjectivism in the name
of a materialist thesis which is a thesis conjointly
of (material) existence and of (scientific)
objectivity. ‘

(Althusser 7 Lenin and Philosophy, pl14)
‘There is no question here of whether objects of
discourse exist independently of the discourses
whi.ch specify them. Objects of discourse do not
exist at all in that sense: they are constituted in
and through the discourses which refer to them.

The distinction/ correlation ‘structure of epistemology depends on the conception of objects existing
independently of knowledge yet in forms appropriate to knowledge itse If. To deny that conception
is to reject epistemology and the field of problems
defined within it.

(Cutler, Hindess, Hirst and Hussain, Marx’s
‘Capital’ and Capitalism Today, pp216-17)
Marx’s ‘Capital’ and Capitalism Today is a book by
four Marxists which rejects almost every main
Marxist idea: the theory of value, exploitation (in
the classical Marxist sense), laws of tendency, the
possibility of correspondence or non-correspondence
between forces and relations of production (the
central dialectical discovery of Marx) , modes of
production, structural causality, the determinance
in the last instance of the economic structure,
politics as class struggle, the distinction between
reform and revolution, and (though in a rather
tongue-in-cheek sense) the possibility of a classless society.

As far as I can tell, the only survivors of this
epistemological holocaust are: classes, defined by
the relation of possession or separation between the
economic agents and the means of production; and
the methodological principle of rejecting explana3 The idea of putting questions to nature doesn’t presuppose that nature has
her own language in which to answer, independent of our theories. Much
of the anti-epistemological case rests on the unargued assumption that it
does. For instance:

‘If testing is a rational procedure then there must be an a-theoretical
mode of observation governed by-a pre-established harmony between
language and the real. To maintain, as Popper does, both the rationality of testi~ and the thesis that observation is an interpretation in the
light of a theory is to collapse into a manifest and absurd contradiction.’

(Hindess, Philosophy and Methodology in the Social Sciences, pl.S6)
thiS passage, I take tt, is meant to be a complete argument, not Just a
step in one. n: is supposed to be ~ incoherent both to recognise
that there are no theory-independent observation-statements, and at the
same time to test the theory by reference to observations (experiments).

This would only be true if the theory determined, not only the nature of the
experiment and the way the result is interpreted, but the result itself.

10

tions of historical processes in terms of the will
and consciousness of agents. These survivors are
certainly crucial parts of Marxist theory, but
hardly peculiar to it, and not conjointly sufficient
to distinguish Marxist practices – either theoretical
or political – from many others.

Such a drastic revision ought not to be ruled out
a priori. It would, I believe, be more accurate to
call it an abandonment of Marxis m rather than a
revision of it, but it would be dogmatic to refuse
to consider this possibility. If Marxism will not
stand up to scrutiny, or does not give us the best
available account of the workings of society, it
should be rejected. Such rejection would not necessarily lead to the abandonment of socialist politics,
any more than the replacement of Newtonian science
by Einsteinian undermined technology – on the contrary, it made new practices possible within it.

It can’t be ruled out in prinCiple that Marxism will
be replaced by a superior scientific theory on which
an improved strategy for the workers’ movement
could be’ based.

But it seems’ to me that the common element of
their critique of these diverse concepts is the idea
that classical Marxism is essentialist; this essentialism is seen as flowing from a rationalist epistemology; and this epistemology is rejected, not in
favour of Humean empiricism (the only alternative
epistemology of which they seem able to conceive)
.

but ~n favour of no episte~ology at all.

I will argue that-classical Marxiam is-not essentialist,
rationalist or indeed empiricist, but represents a
genuine breakthrough to a materialist epistemology,
and that the rejection of epistemology leads these
authors to a rather extreme form of idealism.

First let us look at their criticis m of some
specific Marxist concepts:

(a) The labour theory of value:

The point of this theory is not, I take it, primarily
to explain the prices of commodities, but to give an
account of the necessity for a proportionate distribution of social labour , and the various mechanisms
specific to various modes of production, by which
equilibrium in this distribution can be restored. I
shall therefore proceed straight to their criticis ms
of this aspect of the theory.

It is just the concept of equilibrium that they dislilte most, both in the work of RudoIf Hilferding and
in that of 1. 1. Rubin. There is something of an
attitude of ‘heads I win, taUs you lose’ in their
treatment of the respective arguments of these two
Marxist political economists; also a sort of conceptual witch-hunting technique of guilt-by-association.

The concept ‘equilibrium’, it seems, has suckled
two devils: economism and teleology_
‘In the concept of “equilibrium” we see the
economistic hope of an end to capitalist relations
of production. “Equilibrium”, the state of
capitalism’s vitality, is also the threat of its
death.

Should equilibrium conditions be threatened,
systematic non-reproduction is pOSSible, a
systematic non-reproduction which undermines
the relations of production. This .fantasy is the
dark side of a functionalism – the death that
awaits the organism if its vital mechanisms are
.inhibited. ‘ (p71)
I would hate fo see this style of argument generalised. Consider the following:

Fred: Man must eat before he can think.

Joe: You are a Heideggerian! I shall prove it!

He who speaks of the necessity of eating,
admits the possibility of starving to death.

Bence your whole philosophy is based on
anxiety in the face of death. Ergo: You are
a Heideggerian. Q. E. D. Take him away and
cut his concepts off!

But people who talk of equilibrium generally believe
that it can indeed be disturbed (what’s happening to
the economy now, comrades ?), but also that a new
equilibrium can often be established. They are in
no way committed to the idea o£.a ‘terminal crisis’.

And even if they were, they would not be corn mitted
to the idea that such a crisis would lead to the
supersession of capitalism without political
“struggle, as is insinuated later in the same passage
(cf. Norman Geras’s excellent exposition of
Luxemburg’s theory of the collapse of capitalism
and the historical alternatives of socialis m or
barbaris m, in his book The Legacy of Rosa
cLuxemburg).


But the idea that there are mechanisms for the reestablishment “of equilibrium is treated as equally
flawed – this time because it is ‘teleologicalfunctionalist’. In their discussion of Rubin, they
tell us:

‘This variant of the Marxist theory posits a
functionalism (a certain composition is necessary
in.an economy, this composition regulates
production) and then makes ~ working economy
exhibit this functionality (it must have a composition structure since prodUction must be regulated
– these goods would’ not be bought if they were not
use-values). ‘ (p87)
‘What these gentlemen lack is dialectics.’ The only
concept of a totality of which they seem capable of
conceiving, is of one which excludes dysfunctional
aspects ,which~ in short, cannot generate contradictions. Either teleology is in its heaven and all’s
well with the economy, or there can be no selfregUlating mechanisms, and every crisis is terminal.

Yet the Marxist dialectic is nothing if not a theory
of structures which reproduce themselves without
teleology, and generate dysfunctions which can be
contained but not eliminated within the structure (4).

My accusation is not the dogmatic one of infidelity
to Marx , but that they fail even to consider a central
Marxian theoretical innovation as a possibility. It
seems to be necessary to remind these epigoni of
Theoretical Practice that there is such a thing as a
‘de-centred structure-in-dominance’.

(b) Hunting” the trans-historical subject:

This treatment o”f equilibrium-theory as a theory of
economic systems as expressive totalities (for that
is what is going on) is an instance of a more general
tendency to read Marx in essentialist terms, and
reject his theories about society on the ground that
their alleged essentialist premises must be rejected.

Thus the ‘laws of tendency’ of capitalis m and the
succession of modes of production are seen as
instances of an essential tendency of mankind to
progress, to develop its productive forces; this is
traced to a ‘trans-historical subject’ and myths of
the self production of man etc. Cutler ~ quite
rightly feel that there is no longer any need to argue
against these myths, and that doctrines which stand
or fall with the m – must fall. But they fail to show
that Marx does start from these premisses.

:Capitalis m develops the forces of production becau~e
of a dynamic internal to its structure, not because It
is one avatar of an inherently self-developing Eternal
l.lIan. There are in each case specific reasons why
particular modes of production generate the forces
wl1ich make pgseible transition to particular other
;4 There are, of course, homoeostatic mecha.n1Sms in nature. I wonder if
these authors would regard them as teleological, or pretend they don’t
exist,.

modes of production – e. g. feudalism to capitalism
or capitalis m to socialis m. It was only when, in
The German Ideology, they decisively – not to say
contemptuously – turned their backs on selfproducing subjects etc, that Marx and Engels were
able to formulate the materialist conception of
history as a research program me.

The a priori determination of Cutler et al to find
essentialism in Marx perhaps stems from a tendency to push the epistemological break further and
further forward until it postdates the deaths of
Marx, Engels and probably Lenin, and perhaps
shifts out of history altogether, into e’s ch~tologylo
Of course, they don’t deny that there are place~ in
Marx’s writings where he says things incompatIble
with essentialism and the consequent unilinear conception of progress in history, and it is this side of
his thought that they wish to develop. But in interpreting his central concepts, they impute
essentialism on the slenderest evidence.

Perhaps the most serious pol~tical consequence of
this approach is the rejection of the analysis of
politics in terms of class struggle, for if such an
analysis is misconceived it is difficult to see how
Marxis m is an advance over utopian socialis m.

They seem to be claiming that political movements
could represent classes only as expressions of the
will of those classes, not as complex effects of
economic class struggle. It is a repetition of a
mistake which has arisen before in connection with
the Marxist theory of the state. Marx and Engels
had (in the Communist Manifesto) said of certain
states – namely the co~titutional states based on
limited suffrage which existed in the most advanced
countries at the time of writing – that ‘The executive of the modern state is but a committee for
managing the common,affairs of the whole bourgeoisie’~ Some have taken this as a general theory of
states as purpose-built instruments in the collective hands of the ruling class, but it is not so.

Marx’s analysis of France under the second empire,
or of Germany under the Prussian monarchy, show
the possibility of forms of state which, while they
are effects of the class structure and the struggle
it generates, and while they necessarily serv~ to
perpetuate the economic power of the possessmg
classes, are not related to the ruling class in this
way.

The relation of parties to class struggle is
similar. The British Labour Party, for instance,
does not represent the will of any class. It tries to.

resolve economic class struggle in a particular
manner determined by the effects of various economic cl~ss forces upon it (through its need to retaih
its trade union base, appease petty bourgeois
‘public opinion’ far electoral reasons, co-operate
with monopoly capital in the management of the
economy, etc). In doing so, it serves the needs of
~he bourgeoisie in preserving capitalism.

It is not then, always possible to map parties
onto the ciasses the·y represent, but this doesn’t
mean we have to see them as a plurality of pressure
groups with no closer connection to the class
struggle than the anti-vivisection society. The .

assumption that it does stems from a wrong poslng
of the problem. It is posed in terms of the unity of
Has

~

Radical

subsQription to
Philo8op~

expired?

UNBW NOW!

11

classe~ at the economic, political and ideological
levels. But such unity does not necessarily exist Marxism has never claimed that it did. It is
normally only in revolutionary situations that
economic classes close their ranks and act as
collective historical agents. But economic struggle
goes on all the time.

Classical Marxism (5) certainly defines classes in
economic terms (Pace Poulantzas), but it was never
asserted that classes so defined would necessarily
Cohere at the political and ideological levels.

Rather, the organisation of the working class as a
political unity was to be the result of a protracted
ideological struggle, by which the economic
struggle which is given as a spontaneous effect of
capitalist class-relations would be transformed
into a struggle with revolutionary political
objectives.

But for Cutler et al this original discrepancy
between economic and political class organisation
is a theoretical problem rather than a practical
one. Either it is a matter of a discrepancy between
objective interest and subjective consciousness of
that interest, or the failure to develop a political
class consciousness is itself an effect of the structure, determined in the last instance by the economy
itself. In the first case, individual subjectivity is
given an explanatory role, and no account is given
of why other corn munal interest groups should not
take precedence over class loyalties. In the second
case, we are committed to a structural causality
which, we are assured, is essentialist.

Why the recognition of the discrepancy between
class interest and political class consciousness
should lead t~ ‘subjectivism’ we are not told. Could
it not be explained by an objective but non-economistic theory of ideology? Perhaps, granted such
a theory, we are back on the ground of Althusserian
structural causality; but then what is essentialist
about such a position? We are told it is functionalist – ideology is produced because the reproduction
of the economic structure requires it. This may be
a valid criticism of Althusser1s paper on Ideology
and the State (in Lenin and Philosophy) of wh ich I
have argued elsewhere that it depicts the production of ideology as ‘a conspiracy without a subject’.

But such functionalism is not a necessary consequence of any scientific account of ideology. It is
possible torecognise that capitalism produces
dysfunctional ideological formations (e. g. Marxism)
and that the production of ideas in bourgeois society
is as chaotic and unplanned as is material production, without Jeaying anything to ~ublecttve arbitrariness. Certainly ideology does, overall, serve to
make possible the reproduction of capitalist classrelations -insofar as it doesn’t, there is ideological crisis. But the evolution of ideologies is
Darwinian, not Lamarckian.

But all these attributions of some form of subjectivism to Marx or other Marxists (whether the
‘subjects’ involved are de-historicised individual
subjects or a trans-historical Absolute Subject),
are ‘must have’ arguments: Marx must have
assumed. this that or the other thing (which he
would not have admitted to assuming). In such
cases it is useful to ask whether there is some
other explanation of his position – and there is.

Cutler .eLal don’t really consider the possibility
of an objective, materialist theory of subjectivity,
which neither ignores the fact of human subjectivity
5 This intentionally vague term iii intended to include at least the mature
Marx, Engeis, Lenin, Luxemburg and Trotsky; and to exclude historicist
Marxism, whether of the Stal1n1st variety (Soviet Ma~m in the 30s and
408 waa profoundly historicist), or the western Hegehan variety (Korsch,
Lukacs, Gramsci, the Frankfurt School, Sartre).

12

nor treats it as an irreducible datum. Yet classical
Marxism has always assumed the possibility of
such a theory, and although there is much work to
be done in this field, there is some reason to
believe that Marxism, with a little help from its
psychoanalytical friends, will be able to achieve
this.

But this neglect is an instance of an approach on
the part of these authors which makes all that is
original in Marxism invisible to them – their
interpretation of any structure, any interconnected
whole, as an expressive totality, of the sort that
Althusser identifies in Regel and the Hegelian
Marxists.

(c) IDmressive totalities and determination in the
last instance:

Pre – Marxian thought – with the possible exception
of Spinoza – conceives of the ultimate structure of
reality, whether physical or social, in one of two
ways: either as the mechanistic interaction of
atoms, themselves unstructured; or as forming a
whole or wholes in which the parts had no autonomy,
their nature following from the nature of the whole.

The former was the view of 17th and 18th century
empiricis m and materialis m, the latter of Leibniz
and Hegel, with Kant representing a mixed pOSition.

In order to conceive of social structures in a
materialist way, it was necessary to reject both
atomis m and holis m, both of which depended on the
idea of irreducible subjects (individual or Absolute),
and neither of which could allow for the possibility
of really existing conflicts, internal to and generated by the structure, yet threatening to destroy it.

For Marx, a social formation is a structure which
is not the mere expression of its ‘idea’; which has
no centre, no purpose, no directing agency; which
determines the social nature of its elements, and
the contradictions between them. Its elements
double as material beings subject to the laws of
their own material nature, and as terms of social
relations in which they occupy definite roles
generated by the stru~ture. There are various
interlocking sub-structures whose nature cannot
be read off from the overall structure.

Marx’s achievement in theoriSing this type of
structure is marked by Althusser’s concept of a
‘de-centred structure-in-dominance’, as contrasted
with the ‘expressive totality’ of the idealists.

Althusser points out that the metaphor of ‘inverting’

Hegel is inadequate. A society is not a mere expreSSion of its economic structure, as according to
Hegel it is an expression of its ‘Idea’. Over-literal
interpretation of the inversion of Hegel leads at
once to mechanistic-economistic distortions of
historical materialis m, and the repetition of
idealist philosophical positions.

All these points, I had thought, were bridges
burnt behind us (6).

But no. We are now assured that the only way to
conceive of the effectivity of a structure on its
elements is in terms of an expressive totality
which constitutes them; and that the economic
structure cannot be determinant in the last instance
6 It might be thought by ‘humanistic’ Marxists that there is an alternat1ve
to Hegelian and atomistic social explanations. other than the one that I
defend here – namely, that the explanatory structures of capitalist society
are ‘reified praxis In that case I would want to know: How does this
praxis get reified? Is it self-reifying? In that case we are back with
Hegel; and the objectivity, ‘coefficient of adversity’ and sheer cussedness
of capitalist reality are not given their due. Is it the activity of the
capitalists that reifies the praxis of the workers? In that case, we would
be back with an atomisttc account of some individuals oppressing others;
unless this is a way of saying that the structure of capitalist society
(which constitutes the capitalists as capitalists) has this effect. But in
that case we are back with the notion of an objective structure, independent of the will and consciousness of individuals; which is the concept I
am defending.

On the misuse of the word ‘reificatlon I I see note (11) below.

I.

withil the social structure, unless the superstructure is a mere epiphenomenon of it.

At this point one might be forgiven for thinking
that we are going to be treated to a materialist
version of methodological individualism: come back
Hobbes, all forgiven.

Wrong again! We are told that this essentialism
springs from a rationalist epistemology, and atomism from an empiricist epistemology. -These are
seen as the only alternative types of epistemology,
and both wrong. It remains to be seen what alternative mode of seeing society springs from – no
epistemology at all.

The argument is set out most clearly on
pp213-14:

‘It should be noted that the different epistemological conceptions of the relation between
discourse and its objects entail different
conceptions of the relation between objects
themselves.’ For einpiricism,
‘Relations between objects •.. can only be
conceived as given in experience itself. The
classical empiricist conception of relations
between objects is therefore in terms of a
mechanical, external causality representing
nothing more than the existence of regular and
recurrent correlations between observed
phenomena. ‘

‘In the r4tionalist epistemology, on the other
hand, where the world is conceived as a
rational order, concepts give the essence of
the real ana relations between concepts therefore represent the essential form of the relations between objects. The classical rationalist
conception of relations between objects is
therefore in terms of an expressed causality,
an internal relation between an essence and
the phenomenal forms of its appearance. These
relations between objects may be established
through purely theoretical argument. ‘

I shall not stop to dwell on the fact that there is a

great deal of work in the field of scientific epistemology that is neither rationalist nor empiricist in
these senses. Philosophy has not stood still since
Leibniz and Hume. It is not only Marxists, either,
who have transcended this alternative.

It is difficult to fit even Engels’s statement that
materialism means presenting the facts in their
real and not an imagined connection, into this
schema. The reference made in the same passage
as the last quotes, to the thesis of determination in
the last instance as ‘regarded by classical Marxism as beyond any merely empirical refutation’ -is
astounding; if anyone can produce a single passage
written by a ‘classical Marxist’ to substantiate this,
I would be prepared to hang a portrait of Sir Karl
Popper on my lavatory wall. But perhaps the best
text to look at is Marx’s 1857 introduction, and ask
of it: Is classical Marxism rationalist and essentialist?; and: Are the only alternatives to rationalism, empiricism or no epistemology at all? Short
as this text is, it contains enough pointers to keep
us clear of the main errors in the field of epistemology – I shall quote enough of it to make this
evident, I hope.

. One word of warning. It is thanks to Althusser
and his followers that this text owes its justifiably
high reputation in modern Matxist epl,sJemology.

But that should not lead us to read the whole’ of
Althusser’s epistemology into this text – he
deserves some credit for his originality, but also
for leading us up some original blind alleys.

mff

a reministjoumal

mlf is a new journal which aims to devei?p feminist theory a~d
practice through work in the areas of mantlsm a~d psy~hoanalYSls,
The first issue will include articles on ‘Snual Liberation and the
Family’. ‘Natural and Social Division of Labour – Engels and the
Domestic Labour Debate’, ‘Woman as Sign’, translation of Michele
Montrelay’s ‘Inquiry into Femininity’, ‘Representation and Sexuality’,
‘Bourgeois Economics and Women’s Oppression’,
The journal will appear twice a year and is self· financed so please
support mlfby subscribing.

Subscription for two issues: £2 or £3 Europe, £4 or $7 USA and
elsewhere; single copies: £1.20, £1. 70 Europe and £2 or $4 USA and
elsewhere; payable to: TIllf 69 Randolph Avenue London W9 lDW
England

‘DAS

P 22352 F

ARGUMENT

107

Gewerkschafts·Diskussion (I)

”lY·
–.. –

20. Jahrgang

“:..

~-‘,

Editorial:

Die Argument-Konzeption
Eberhard Schmidt
aerat die Sozialpartnerschaft in die Krlse?

Walther Miiner..Jentach
Die Neue Linke und die aewerkachaften
Peilpekti..en arbeltsorlentlerter Reformpolltlk
K. H. Tiaden
Terfltorlalstruktur und antlkapltallatlacha Reformpolltik
Stallnlamua-Forachung
aerd Meyer
Induatrlalialerung, Arbelterklal”, Stalinherrschatt
In def UdSSR (11)
Akt,u*11e Analv.en
Wolfllang Abendroth
Bahro, ,,Alternative”
, Ralner Kawa
Terrorlamu,: $trategien der Verunalcherung
Kongre6berlchte
Hochachulpolitlacher KongreB de, BdWI
WSI-Forum
Kulturtagung de, IMSF
Deut,che G..en,chaft rOr Verhaltenstherapie
Ba’prechungen
Aathatlk/ Uteraturgeschlchta, aewerklchaften,
Theorla und auchlc:hte, StIat,recht
Zaitschrlttenachau

••

I

Januar/Fabruar 1978

Einzelpreis: 9.80 DM
Einzelpreis fUr Studenten: 8,– DM

Argument-Verlag
Postfach 210730
D-7500 Karlsruhe

13

their inIler connection. Only after this-work has
been done can the real movement be appropriately
presented. If this is done successfully, if the life
of the subject-matter is now reflected back in the
ideas; then it may appear as if we have before us
an a priori construction.’

(Marx, Preface to 2nd edition of Capital vol. I,
pl02)
This ‘appearance’ is common to all the sciences
(see Bhaskar on the ‘Leibnizian dimension’ of
science, in A Realist Theory of Science). The
ontological assumption is not that real relations
are structured like conceptual relations, but that
relations between concepts can be made to map
real relations.

At the same time it is clear that Marx doesn’t
~hare the connation of epistemological and ontological questions which characterises empiricism,
at least from Berkeley on .. This ‘epistemic fallacy’

as Bhaskar calls it, consists essentially in the idea
that truth must have a subject, that fo~p~o be true,
there must be an answer to the questiori~Who is to
!3ay that.·£? Bhaskar argues against this that it is
necessary to distinguish what he calls the transitive
object of science from its intransitive object. The
intransitive object is the reality which science
seeks to know, which pre -exists science and is
unaffected by the knowledge which seience- requires
.of it; which science never fully appropriates,
but acquires ever deeper knowledge of. The transitive object is the real as known by the science of a
particular time~It is never assumed to be final
and definitive appropriation, but it is aimed at
deepening the knowledge of the intransitive object.

Progress in science is essentially progress towards
an ever closer approximation to objective truth.

Essentially, the intransitive object is Marx’s
‘real subject’ (see (it) above)” or Althusser’s ‘real
object I. Lenin marks the same qtstinction by the
terms relative and absolute truth (Materialis m and
E mpirio – Criticis m).

In this way the implicit assumption of an original
complicity between subject and object, which
characterised both rationalism and empiricism,
is overcome. Yet it is not assumed that the intransitive object is unknowable, only that progress
in the knowledge of it is a never-ending process of
apprOximation. If science doesn’t yield knowledge
of the real, it loses its pOint.

The authors of Marx’s ‘Capital’ and Capitalism
Today, however, have a different alternative to
rationalism and empiricism, and an older one.

‘Far from providing an external measure for
discourse, the entities referred to in discourse
are constituted solely in anq through the forms
of discourse in which they are specified.

Objects of discourse cannot be specified extradiscursively …. ‘ (p216)
One might as well say: We can’t talk about things
without using .words; therefore we can’t talk about
things at all, only about words. Or again: Our
knowledge conSists of ideas; therefore we can only
know ideas. Not a bad effort this, as the speculations of a youthful future bishop. But hardly clever
for four learned post-Althusserians.

Granted, there are two differences from Berkeley:

tdeas are conceived as linguistic realities, not as
perceptual ones – this gives the whole thing a 2Othcentury look. And it is not denied that there are
things-in-themselves, only they would have-1:o be
unknowable. But such ideas also have been put
forward before, .and with a subtlety absent here –

by ~mmanuel Kant. In the words of Cutler et al
‘Now we have argued that the epistemological
project is not a necessary one and that the relations between discourse and its objects does not
need to be conceived in terms of both a distinction
and a- correlation between a realm of discourse
and an independently existing realm of objects.

But in-the absence of such an epistemological
conception it is no longer possible to conceive
of objects existing outside of discourse (and
represented in its baSic concepts) as the measure
of validity of discourse. On the contrary, in the
absence of such speciiiable yet extra-discursive
objects the elements specified iri discourse must
be conceived solely in and through the forms of
discourse in which they are constituted. What is
specified in theoretical discourse cannot be specified extra-discursively: it can be conceived only
thro~ that discourse or a related, critical, or
complementary one.’ (ibid, pp228-29)
Compare the following:

‘Whatt!len, is to be understood when we speak of
an object corresponding to, and consequently also
distinct from, our knowledge? It is easily seen
that this object must be thought only as something’

in general = x, since outside our knowledge we
could have nothing which we could set over against
this knowledge as corresponding to it. ‘

(Kant, Critique of Pure Reason, p134)
The difference is that Kant’s thing-in-itself had the
function of accounting for objectivity in appearance,
for an element of passivity on the part of the knower
in relation to the known, for an ’empirical realism’

within a ‘transcendental idealism’. In short, Kant
had a ‘grain of materialism’, which these authors
have cast away.

And no wonder, for what they denounce as
‘epistemology’ – the thesis that a discourse and its
objects are distinct, yet discourse should seek to
correspond to its object – this thesis is precisely
mate’rialist epistemology (7). The complicity of
subject and object present in Des cartes , Hume and
others consists not in their ‘idea that thought
should correspond to its object, but tn their failure
to give adequate recognition to the distinctness of
thought and object – 1. e. not in their assertion of
the second, but in their denial of the first postulate
of materialist epistemology. In Des cartes , and
rationalism generally, this takes the form of the
imposition by choice of method of substantive ontological conclusions – e. g. the distinctness of mind
and body. In Berkeley, HUme and later empiriCism,
it takes the form of phenomenalism or some other
form of neutral monis m.

Yet the question remains: if we are to take the
materialist epistemological option, are we not
assuming that reality is, at least, knowable by us?

Why should reality be knowable? Someone once
said to me, when I was arguing that the distinctions
made by science reflect real distinctions: ‘You
seem to be assuming that there is Joe Reality some·
where out there, making sure that things turn out
alright. ‘

Indeed it is bound to look as if there is such a
Being; because an unknowable universe – though
there is nothing logically ~ontradictory about such
I an idea – could not have given rise to beings who
7 I sometimes get the impreSSion, from the manner and frequency with
which they repeat thiS formula for epiStemology as !!2Yl distinctness I!!S!

correspondence of the two realms, that they think that distinctness in
itself precludes correspondence, and that the epistemological project
therefore only has to be expressed this way in order to be refuted. This
would be on a level with the idea that ISne can’t have any knowledge of
other people’s mental states, as one can’t introspect them.

15

can ~chieve knowledge, and who depend for their
survival on such knowledge. In an unknowable
universe, there would be no knowing subjects to
know that it was unknowable.

It is a useful exercise for an epistemologist to
reflect on the manner in which Darwin’s theory
dealt a death-blow to the idea of teleology in
nature. There appears to be an ‘original complicity
between species and their environment – because
where the two are ill-suited, the species die out.

Nature will always produce the appearance of
design if it produces the appearance of anything,
because where it doesn’t produce that appearance,
there will be nothing for it to appear to. There is
no necessity about the knowability of nature, any
more than about the existence of Des cartes; but
whenever Descartes thought he existed, he existed
sure enough. And anyone who asks: Is the universe
knowable? lives in a knowable universe alright.

If we return to Kant armed with these Darwinian
insights, we can discard the transcendental idealist
shell, and extract a very valuable kernel.

‘Experience is itself a ‘species of knowledge which
involves understanding and understanding has
rules which I must presuppose as being in me
prior to objects being given to me, and therefore
as being a priori. They find expreSSion in ~
priori concepts to which all objects of experience
neces’sarily conform, and with which they must
agree. ‘

(Critique of Pure Reason, pp22-23)
Of course, we know that Kant thought that the
understanding actively imposed its categories on
the manifold of intuition, so that the ‘must’ at the
end of the passage means: ‘We have ways of making’

objects conform to our knowledge!’ But we could
make creative use of its ambiguity, and say:

In order for empirical knowledge to be possible,
reality must be ‘structured in a certain way; and
knowledge is possible” as the achievements of the
sciences have shown; so reality must have that
structure, objects must conform to our knowledge
(in the sense that we say: ‘It must have rained in
the night’, when we find the kitchen floor flooded
in the morning).

In this way, we could really have access to
‘synthetic ~priori’ knowledge – not !Lp-riori in the
absolute sense, but as the limiting case of the
relative sense referred to by Kant on page 43 of
the Critique of Pure Reason. As knowledge, that
is, of the conditions for the possibility of knowledge – conditions which are only contingently
realised, but the reality of which follows necessarily
from the fact that there is such a thing in science.

Radical
Science Journal
No. 6/7 now available £1.75 176 pages
Contents: EDITORIAL-Science and the labour Process; LES lEVIDOW-A Marxist Critique of the IQ Debate; MICHAR BAR NETTTechnology and the Labour Process; BOB YOUNG-Getting Started on
Lysenkoism; RSJ SUBGROUP-Marxian Psychoanalysis and Feminism;
LES LEVIDOW-Grunwick as Technology and Class Struggle; MICHAR .

YOUNG-Althusser’s Marxism and English Social Science; GREG
McLENNAN-WiII the Real Louis Althusser Stand Up? Short Reviews:

Technology of Political Control; Sabotage/Industrial Conflict; Television/Technology & Cultural Form; Sociology of Knowledge_ Notes:

Energies, Priorities & Activities; Marxist Readings on Science, Technology, Ideology, Labour Process, Psychoanalysis.

No. 5 still availabl. £1.00 136 pal’s
contents: PATRICK PARRINDER-The Black Wave: Science and Social ConSCiousness in Modern Science Fiction; DAV/D TRIESMAN-The Institute of Psychiatry
Sackings; BOB YOUNG-Science is Social Relations_ Letter: LOUP VERLETLa course ou la Vie.

Subscription: £3.00 (individual), £9 (institutional) for 3 issues postpaid.

Individual copies: please add 15p each for postage, etc.

Bulk orders: one-third reduction on 10 or more copies.

Radical Science Journal, 9 Poland St, London W1

16

Some of these conditions may indeed by Kantian
ones: spatio-temporalordering, causality,
countability, a principle of conservation and
continuity and so on.

As Kant says of the reproducibility of representations in imagination (which is a condition of the
possibility of knowledge):

‘This law of reproduction presupposes that
appearances- are themselves actually subject to
such a rule, and that in the manifold of these
representations a co-existence or sequence takes
place in conformity with certain rules. Otherwise
our empirical imagination would never ,find opportunity for exercise appropriate to its powers’, and
so would remain concealed within t1ie mind as a
dead and to us unknown faculty.- If cinnebar were
sometime.&’l”ed, sometimes black, sometimes
li,ght, sometimes heavy, if a man changed sometimes into this and sometimes into that animal
form, if the country on the longest day were
sometimes covered with fruit, sometimes with
ice and snow, my empirical imagination would
never find opportunity when representing red
colour to bring to mind heavy cinnebar. ‘

(Critique of Pure Reason, p132)
If our faculties of knowledge depend for their
possibility on the structure of the world outside
them in this way, why should we not assume that
the world really has that structure and hence makes
them possible, rather than that they have the additional, magic ability to force the world into a knowable form which it doesn’t have in itself?

Such a materialist inversion of Kant would, I
suggest, be far more fruitful for epistemology than
all the somersaults and other circus tricks that
Regel has been compelled to perform posthumously
by his Marxist ringmasters (8).

But to return to Kant’s post-Althusserian successors: we must now ask where their proposed final
solution to the- eptstemological problem gets them.

Does it rid them of the errors of rationalism?

Does it enable them to ‘think the relations between
objects in a way that is neither essentialist nor
mechanical? The answer to both questions, I fear,
is No.

For what procedures are left to a theory in order
to justify itself? Only its internal consistency, and
its openness to criticism by another theory. But
how do we decide between two internally consistent
but conflicting theories? By-their explanatory
power? But what do they explain? Not ‘the facts’,
that’s for sure. A number of logically irrelevant
criteria present themselves: political, aesthetic,
humanistic. A consistent rejection of epistemology
couldn’t stop these being appealed to, any more
than it could stop complete arbitrariness. The
motto would be Feyerabend ‘s: anything goes! Our
present authors want to let the sheep out of the fold
and Feyerabend wants to let the wolves in, but the
effect is the same-.

No doubt these authors would differ from
Feyerabend in demanding at least high intellectual
standards, logical rigour etc. This brings them
closer to traditional rationalis m, but they are
certainly at no- advantage with respect to it.

Everything happens- in thought; theory is-not only
8 To avoid a possible misinterpretation: I have no sympathy for historicist
readings of Kant, which have found some favour among Marx:l.sts (e.g.

Lucien Goldman). According to such views, we do indeed shape reality
in the process of knowing it, but in socially determined and historically
relative ways. The ‘inversion’ which I favour is the opposite of this: a
realist insistence on the objectivity of knowledge, a denial that reality
is our product; and consequently a re-interpretation of the status of the
so-called ‘synthetic a priori’, which remains as the set of the conditions
ofthe possibility of experience (knowledge), but contingently and
objectively realised conditions.

comp)sed of concepts, it is- atlOut concepts; and if
they can claim one up on rationalism for not
imagining that reality itself is constructed
deductively, by the same token they -are in a
worse case when it comes to the practical value
of theory, for their theory has !lQ. relation with
extra-theoretical reality.

Theories are useful when they tell us about the
world; then they can help us to act on it more
effectively. It is all very well to clarify the relation
between the concept of -the -possessing class and the
concept of the state apparatus, but the concept of a
boss can’t sack you and the concept of the police
can’t bust you.

They argue (pp219-20) that rationalism misled
Marxists into regarding the superstructure as
derivable in essence from the base, while in empiricism the relations of base and superstructure are
held to lie ‘beyond the range of theoretical determination ‘. According to their own view, all that can
be said about the relation of base to- superstructure
is that each economic structure requires as a
necessary condition of maintaining its existence, a
superstructure from a definite range of possibilities. This is surely to reduce the relations of base
and superstructure to logical relations of the concept of base to the concept of superstructure (if
this interpretation is not right, I fail to see the
connection between this conclusion and their view
of epistemology). But in this case they get stuck on
both horns of the dilemma: their theory deals only
in concepts, so what can it dO,but ‘derive essences’;
and as for the reality to which the theory doesn’t
refer, how can its elements be related except
accidentally? Did the struggle of the rising bourgeoisie against feudal superstructural institutions
just happen to promote the development of nascent
capitalism, and so further the interests of the
bourgeoisie? Surely, real causal relations can be
discerned here, without any recourse to essentialism.

I have been arguing that classical Marxist epistemology is not, in intention, rationalist; that its
ontology is not, in intention, essentialist; and that
its notion of structural determination_ is not, in
intention, holistic or teleological.

Anti-Marxists who level these charges at Marx
generally treat it as axiomatic that Marxism’ is at
least essentialist and holist. They direct their
energies to showing these positions to be incorrect.

Popper’s anti-Marxist writing is like the Maginot
Line: a magnificent defence against Marxist theory,
constructed along a frontier across which it was
never the intention of Marxists to attack.

The authors under consideration however are in a
different case. They know that the structures of
which Marx and Althusser speak are not supposed
to be essentialist ‘expressive totalities’, but they
think that in fact they can’t be anything else.

So we must ask ourselves: how is non-essentialist
structural determination possible? And, before
that: what are the distinguishing features of a
Marxian (materialist) structure, as against a
Hegelian totality or a Humean bundle?

It is ‘classical MarxiS m’ that I am concerned to
defend here, for although the terms ‘structural
causality’, ‘de-centred structure in dominance’,
etc entered Marxist parlance with Althusser’s work,
the concepts were already present; and Althusser~
ians have sometimes got into real problems in
applying these concepts to transitional periods.

In the first place, the structure has the property
of maintaining itself in being while its elements
change: hence it is not reducible to the sum of its
elements; and at the same time, its self-reproduction is non-teleological – it doesn’t depend on there
anywhere existfng a purpose of perpetuating the
structure: hence it is unlike the holistic ‘society as
subject’ which ‘creates itself’, which Marx and
Engels ridiculed in THe German Ideology.

Secondly, the structure assigns definite powers
and limitations to its elements, which therefore
can’t be understood atomistically; yet it doesn’t
constitute those elements, which also obey laws
other than those of determinat~on by the structure.

Thirdly, a structure of this sort depends for its
self-perpetuation on nature external to it. Thus, a
society must find its raw material for production in
its environment, and must satisfy the biological
needs of its population, sufficiently to secure the
physical survival and reproduction of its ‘supports’.

Every society exists under the constraint of geological and biological laws which it didn’t constitute.

Fourthly, the structure may have irremediably
imperfect mechanisms for securing its own reproduction; these mechanisms are not for the most
part systematically or ‘deliberately’ produced; and
they are not only accidentally imperfect, i. e. they
can’t in all cases be perfected within the structure,
so that that .structure could achieve ‘immortality’.

In particular, class societies produce class antagonisms irreconcilable within these societies,
which can (and to some extent always do) disrupt
the society in question, and which may lead to its
overthrow and supersession.

Social structures have no ‘soul’, no single principle or directing agency; but neither are they
reducible to the interplay of many ‘souls’, the ideas
and purposes of many individuals. ‘Holism’ and
‘atomism’ share the assumption that social explanation must be in terms of will or wills, conscious
agency and purposes. In rejecting teleology, and
every idealist or voluntarist form of explanation,
a materialist theory of society, so far from falling
midway between holism and atomism, clears itself
of both at one stroke, and founds a theory of
structural determination.

Marx demonstrates at length in Capital how
capitalist firms, whose ‘purpose’ is to produce
surplus -value (and this ‘purpose’ is itself determined by the structure), not only- produce surplusvalue but in doing so reproduce the material and
structural conditions of existence of capitalis m riew means of production, new workers, the same
relation of separation between them.

The question how political and ideological formations are produced is more complex, but the solution
is no more ‘teleological’. The state and ideology
$;.re not for the most part purpose-built to secure
the reproduction of capitalist relations. The policeman-defends ‘law and order’ and in doing so defends
the privileges of the bourgeoisie. The parson
preaches spiritual values and in doing so offers
sufferers an alternative solution to the relief of
their suffering; he is not a conscious opium-pusher.

In order to achieve a materialist analysis of the
superstructure, it is necessary to separate the

17

3) Requirements of an Epistemology
for the Human Sciences

question of the production of ideological and
political institutions, from the question of the role
of such institutions in the reproduction of class
relations. It is their failure to separate these
questions, rather than any rationalist epistemology,
which forces many Marxists into teleological explanations (e. g. the ‘functionalism’ of Althusser’s
essay on ideology and the state).

But if these questions are separate, how do~s it
come about that the superstructure does in fact
promote the reproduction of economic class relations? Once again, the answer is not to be found
in consciousness, teleology, providence, or Joe
Reality; it is simply that if a hypothetical mode of
production had no mechanism that secured this
effect, that mode of production could not be instantiated. So that it is no accident that the only modes
of production that exist are ones with mechanisms
which do in fact do the job of creating the political
and ideological conditions for their reproduction.

While we are on the subject of teleology, it is
perhaps worth pointing out that biological organisms – so beloved of holistic/teleological theorists
as analogies (‘society as an organism ‘) – in fact
are themselves materialist structures, governed
not by a single constitutive principle, but by the
complex, structured causal interaction of their
elements with each other and with the outside
world ..

One who denies the possibility of materialist
structures invalidates not only materialist theories
of society, but also materialist theories of biology.

The break with atomistic behaviorism and essentialistic psychologies of consciousness achieved by
psychoanalytic theory would likewise be called into
question. If Spinoza and Marx are cast into the pit,
it is with Darwin and Freud clutching their heels.

And indeed, when Hindess criticises Popper’s use
of Darwinian homologies in Philosophy and Methodology in the Social Sciences, he does accuse Popper
of ‘teleology’ in a context which implies that Darwin
was also guilty. The greatest victories over teleological explanation in the history of science are
thereby tarred with the brush of teleology just
because the phenomena they explain look teleological superficially. In this way, these authors
situate themselves in pre-Darwinian as well as
pre-Marxian theoretical space (9).

This has taken us away from epistemology, but
on the evidence of other texts by some of these four
authors, it would seem that the rejection of essentialism may have motivated the rejection of epistemology. It is necessary to show that Marx is not
guilty of essentialism or teleology in order to .

plead guilty to epistemology with a good conSCIence.

9 The question ‘How are non-teleological self-reproducing structures
possible?’ is best answered by referr!ng to actual structures of this kind
– to answer it in general terms is to risk falling into the sort of a priori
dialectics which was the chief error of the older forms of dialectical
materialism (from Engels to Mao). If the detailed account of the reproduction of capitalist social relations in Marx ‘s ~ does not convince
these authors. I can only suppose an U£iQ!:i rejection of dialectics on
their part, and recommend that they take another look at modern biology.

However. there do exist some general accounts of this kind of structure. ,
produced independently of Althusser’s account. For instance in Buldtartn s
much maligned Histori-cal Materialism; and, in exposition of Spinoza. in
Bans Jonas’s paper ‘Spinoza and the Theory of Organism’, in Spinoza’ a
Collection of Critical Essays, ed. Marjorie Grene.

18

Doubtless it will be asked: Doesn’t an epistemology
which stresses the role of ‘questions put to natu-re’

(i. e. procedures so set up that the result depends
on the structure of external reality) constitute just
one more form of dogmatism? For if it is a definite enough criterion to exclude certain theoretical
practices, must it not be claiming privileged
access to reality for one form of enquiry?

This is easily answered. All that such an epistemology demands is that a theoretical practice be an
enquiry into reality, i. e. aims to measure its
propOSitions against reality. Discourses which
make no effort to conform themselves to reality
can hardly object if a discourse which does attempt
to do so claims superiority precisely in this respect: its correspondence to the real. Needless to
say, epistemology doesn’t rule out discourses
which make no claim to yield objective knowledge poetry, for instance. It simply insists that discourses that do make such claims try to fulfil them.

Experiment is simply the attempt to make the truth
of one’s conclusions no accident.

But it will be alleged that I am not arguing from
the nature of science, but from the nature of truth;
and that there is pre-scientific knowledge as well
as a scientific one.

Naturally, it is true that we stumble on ‘facts’ in
our everyday non-scientific practices, and that the
element of truth in our everyday conception of the
world is also ‘no accident’ in that without some
degree of knowledge, no practice would be possible.

But – even without bringing in the fact that a Marxist theory of ideology shows us that the error in this
conception is no accident either – the testing of
these conceptions within non-scientific practice is
necessarily haphazard, being as it is no more than
a by-product of our pursuit of our practical needs.

The whole value of science as a separate specialised activity (10) is that it lets the ~hings (ll)them10 ‘Specialised’ in the sense that it is distinct from other activities, not
necessarily in the sense of being the activity of a socially distinct body of
specialists. Naturally, in a bourgeois society, the social division of labour
is superimposed”n the technical division, and the ‘scientific community’

as a rule forms part of the bourgeoiSie or petty bourgeoisie. This does not
of course mean that the contents- of their discoveries are bourgeOiS, any
more than one can taste the surplus value in one’s beer.

Having said that, though, I certainly don’t imagine that communist society
could do without ‘experts’. The simple fact that human knowledge accumulates necessitates the permanent increase of specialisation. The antispec1alism lobby stems from a petty bourgeois conception of knowledge as
the private property of individuals.

11 It has been suggested to me that referring to the objects of science as
‘things’ is revealing, indicating some ‘reiflcation’ underlying my thought.

The eXpression I USe was suggested by the phrases ‘res ipsa loquitur’

and ‘to the things themselves’ (the latter a translation of Husserl’s
‘Zu den Sachen ‘).

But as this is a common objection to the ‘sc1entistic’ Marxism which it
is my concern to defend, it is worthy of a few comments. The O. E. D.

gives as the primary sense of ‘thing’: ‘Whatever is or may be an object
of thought’, and derives it from Old High German’ding’ meaning ‘public assembly’. Equivalent words in other European languages also indicate a
primary sense of ‘object of discourse or activity’ rather than ‘bit of hardware’. which is what people have in mind in connection with ‘reification’.

In common usage, ‘thing’ is the most general ontological-term in the
English language, ranging over events (‘a funny thing happened to me .•. ‘),
people t’you poor old thing ‘), statements, actions (‘what a thing to say,
do’), relationships (‘baby, we’ve got a good thing going’), intellectual
problems (‘The time has come, the wairus-said, to speak of many things
.•• why the sea is boiling hot and Whether pigs have wings’) – the list
could be extended indefinitely.

Two ‘things’ (conclusions) follow from this. Firstly. the term ‘reification’

will be quite contentless unless we restrict its meaning to something like
‘the fallacy of misplaced concreteness’. Secondly, there is likely to be a
hidden motive behind the objection to ‘re1fying’ discourse (in the wider,
contentless sense). I would suggest an epistemophobia which finds it
threatening to be the object of knowledge (it is noteworthy that the word
‘objectifying’ is used of knowledge with the same pejorative sense). I
would seriously suggest that the whole anti-scientific-ideology which sees
objective lmowledge as a threat in this way can be eXplained without
residue in Freudian terms as an effect of a reaction-formation against
scopophilia. Cf. Nietzsche: ‘SCience offends the modesty of all genuine
women. They feel as if one were trying to look under their skin ~ or
worse! under their clothes and finery.’ (Beyond Good and Evil p81).

Needless to say, this doesn’t just-apply to women, let alone to ‘genuine’

women, whatever they might be.

selves~ speak instead of interpreting them.in
terms of our practical needs; that it so oraers our
practice (experiment) that its result depends on the
structure of external reality, not on us. Science
thus enables us to overcome the anthropocentricity
which necessarily qualifies our everyday knowledge
based on practical experience. This in turn makes
possible the development of radically new practices
which could never have been arrived at by mere
practical experience. The im mense progress
achieved by the human race in the last three
pundred years was made possible by this mutation
which allowed science to emerge and liberate itself
from immediate practical concerns. Of course this
mutation had something to do with the rise of the
bourgeoisie, but then the bourgeoisie of that era
was carrying out an unprecedented development of
human liberation. Reactionaries in radical clothes
who wish to renege on the scientific achievements
of the bourgeoisie should be recognised as partisans of the other historical alternative, of
barbaris m, not socialis m.

For the rest, all that needs to be said about prescientific knowledge is that it is possible because
everyday practice includes something analogous to
experiment, though limited by its immediate practical purpose; to use the Freudian term, it includes
reality-testing. Nevertheless, pre-scientific
beliefs must give way to scientific ones when they
clash.

Finally we come to the question of the applicabil. ity or otherwise of the approach I have been defending to the human sciences. After all, Popper used
a version of experimentalism to condemn Marxism
and psychoanalysis. And with the exception of
certain forms of psychological research (of rather
doubtful value), experiment in the laboratory sense
has little use in the human sciences. How does
historical materialis m, for instance, put its
questions to nature (or to society)?

As a preliminary to answering this, it is useful
to distinguish (using a passage from Husser! ‘s
Logical Investigations, vo!. I, pp230-31) between
abstract or theoretical sciences, which ‘are
nomological, insofar as their unifying principle,
as well as their essential aim of research, is a
law’, and concrete sciences, in which ‘one
connects all the truths whose content relates to
one and the same individual object, or to one and
the same empirical genus’. Husser! goes on to say
that ‘the abstract or nomological sciences are the
genuine, basic sciences, from whose theoretical
stock the concrete sciences must derive all that
theoretical element by which they are made
sciences.’ No doubt this group (abstract sciences)
includes physics, chemistry etc. As instances of
concrete sCiences, Husserl cites ‘geography,
history O. e. presumably, historiography – A. C
astronomy, natural history, anatomy etc.’

Now many of the characteristics of the concrete
sciences are commonly ascribed to the social or
human sciences. For instance, experiment as it
exists in the abstract sciences is out of place in the
concrete sciences listed above. They proceed by
observation, description, and explanation in terms
of abstract-scientific concepts, whereas the abstract sciences proceed by abstraction, deduction
and experiment. Furthermore, prediction in the
concrete sciences is always probabilistic, and the
falsification of predictions in them has no immediate theoretical consequences. The abstract science$
on the other hand make no predictions in the sense
of forecasts, only conditional predictions. The

J,

foundation.of Popper’s shadow boxing with Marxism
was his belief that Marxism was a concrete science
and nothing else. This belief is false, but not merely
perverse; the relation between historical materialism
as an abstract science and the concrete analysis of
the concrete situation which Lenin called the heart
of Marxism, is one which desperately needs clarifying. Embarrassment about the unclarified nature
of this relation is no doubt one of the unconscious
motives for the various forms of epistemological
Luddis m which are so prevalent on the left. For
abstract sciences in general are tested independent1y of their use in various concrete sciences, by
means of experiment. But in the case of Marxism,
it is difficult to point to tests which are not in
themselves applications. Where are the experiments? Here let us refer to Marx: ‘In the analysis
of economic forms neither microscopes nor chemical reagents are of assistance. The power of
abstraction must replace both.’ (Capital vo!. I, p90).

However, in the natural sciences too, such abstraction has to be made before experiment becomes
possible. The ‘experimental conditions’ of the
laboratory are essentially reproductions in reality
of abstractions that have already taken place in
thought. The essence of ‘laboratory conditions’ is
that irrelevant variables that would affect a process under natural conditions can be artificially
eliminated. Any abstract science – natural or
social – must abstract from aspects of the real of
which it IS not at that moment treating; but only in
the natural sciences is it possible to actualize this
abstraction and hence measure the real forces, in
terms of the concepts of the science in question.

For this reason, although the human sciences
formulate concepts of forces, relations etc which
are in principle quantifiable (libido, the rate of
exploitation and so on), they are not able to assign
accurate numerical values to specific instances of
these. Both historical materialism and psychoanalysis assume an ontology of quantifiable forces,
without being able to quantify them (except as
‘more’ or ‘less’, ‘high’ or ‘low’, ‘ris ing’ or
‘falling’) .

This is not just a ‘difference in method I between
natural and human sciences – it is not even a difference in method. It is a disadvantage under which the
human sciences labour, and which prevents them
from achieving the same preciSion as the natural
sciences. The persistent disagreements within
Marxist or Freudian theory are only partly explained by the proneness of these sciences to the
intervention of ideology (by virtue of their social
implications). Their unavoidable imprecision is
also at work here. And it is no use seeking humanistic consolations for this impreciSion (‘people are
complicated, can’t be mathematically quantified,
resist scientific analysis’ etc etc etc). These theoretical impreciSions lead to miscalculations in the
practices based on them, which can have consequences far more tragic than any physicist~s or
pharmacologist’s mistake has yet been (Stalin,
Hitler).

Nevertheless, these imprecise sciences are
better than no sciences at all. If they were not, we
should forget about Marxism and revert to an
empiriCist practice of politics, in which there is
no theory but ‘learning by experience’ – a sort of
left Oakeshottism. Some Marxists would not be
averse to this option, I think. It is implicit in
Mandel’s statement that it is by a ‘historico-genetic
method that we will succeed, rather than by an abstract attempt to work out concepts ~hat risk being
19

cha~enged by the next historical experiences. It· is
be, not merely the class composition or mode of
, election (though those are of course important), but
only the balance -sheet of history and revolutionary
that the workers’ assembly would rest on a pyramid
practice that will teach us to think more correctly.’

of democratic organs of power (economiC, military
(New Left Review 100, p102). Concepts which don’t
and adminis trative), whereas the bourgeois parliarisk being challenged by future historical experiment rests on a hierarchic apparatus. The crucial
ences don’t give us any guidance about how to predivide between Lenin and the believer in the PRS
pare for those experiences, and hence are useless
for practice, while leaving theory to those with
then consists in their view of the nature of the
hindSight. If we need theory at all, we can’t wait
organs of state power in bourgeois democracies.

for the owl of Minerva.

Are they simply instruments which can be trans’The proof of the pudding is in the eating’ – very
ferred from brougeois to proletarian hands ,. or are
well, but if you have reason to suspect that the
they essentially institutions of bourgeois power
pudding is poisoned, you will be well advised to
which must be smashed? Once the question has
subject it to scientific tests before eating it. The
been put in this form (a Marxist form, determined
practical advantage of theoretical science over
by the higher-level – but still empirical – hypopractical experience is precisely this separation
thesis that all states are forms of class power) it
it makes possible between testing a theory and
can be answered by investigating the daily running
applying it.

of bourgeois states. All sorts of empirical facts Marxist science is not accurate enough to give us
about Watergate, the refusal of army chiefs in
a fail-safe strategy for revolution, but it has got
Ulster to obey their parliamentary masters, the
enough content to exclude certain apparent possiapproval given by the organs of the British
bilities, and so takes us further than a mere
bourgeoisie (The Times, The Economist) to the
empiricist politics of wait-and-see. For example,
Chilean atrOCities, the co-operation of police chiefs
Marxist theory (as developed by Lenin) excludes
with the fascists at Lewisham and Manchester, a
the ‘possibility’ of a parliamentary road to socialjudge giving his blessing to an advocate of genocide
ism (PRS). Let us make the question about the
– these facts acquire a theoretical importance far
epistemological status of Marxism more concrete
beyond that granted them by the liberal who sees
by looking at this example.

in them only ‘abuses’.

In the first place, we must avoid the temptation of
The structure of the enistemology governing our
revolutionary theory then is as follows: our·theoretlooking for direct empirical verification of this
doctrine – e. g. in Chile. Someone can always give
ical postulates are tested, not indeed under laboraplausible reasons for thinking that Chile was a
tory conditions, but not merely by applying them in
revolutionary practice either. They generate quesspeCial case. And Lenin had already long since
formulated the essentials of the case against the
tions which can be asked of the various realities
possibility of a PRS. But on the other hand we are
thrown up by history, such that the answers that
not dealing with something that follows in rationalhistory gives (12) are not just facts, but evidence
istic fashion from a priori definitions. This could
for the theoretical postulates, which can then be
used to guide strategy in new circumstances too.

hardly. be so when, as is well known, Marx and
May we never have to prove the poisonous pudding
Engels, in historical conditions rather different
of the PRS in the manner of the Chilean comrades
from today’s, did believe a PRS to be possible in
– by eating it ..

a few countries (Britain, USA, Holland – see
Engels’s introduction to Class Struggle in France.

12 In case there are any readers who stil1Imagine that Marxists personify
‘history’ and spell it with a capital ‘H’, I should point out that it is a figure
and Marx’s speech on the Hague Congress of the
of speech similar to ‘experience teaches us •.. ‘ or ‘time will tell’.

first International – both postdating the Paris
Commune).

The impossibility of the PRS follows from the
nature of the state apparatus in advanced bourgeois
societies. Not primarily the nature of parliament
Althusser, Louis, For Marx, NLB; Politics and History, NLB; Lenin and
itself, but of army, bureaucracy, police etc. When
Philosophy, NLB; ESSays in Self-CritiCism, NLB.

Marx and Engels envisaged a PRS, they nevertheAlthusser, Louis and Balibar, Etienne, Reading ‘Capital’, NLB
Bhaskar, Ray, A Realist Theory of Science, Alma Books
less recognised that all that an electoral victory for
Bukharin, Nicolai, Historical Materialism, Ann Arbor
a revolutionary workers’ party would do is show
Callinicos, Alex, Althusser’s Marxism, Pluto
Cutler, Hirst, Hindess and Hussain, Marx’s ‘capital’ and CapItalism
that class consciousness had reached the point of
IQggy, RKP
ripeness for worker’s power – it would not in itself
Engels, Friedrich, Preface to Marx’s Class Struggle 1n France
Geras,
Norman, The Legacy of Rosa Luxemburg, NLB
constitute that power. A ‘pro-slavery rebellion’ on
Grene, . lIvIarjorie (ed.) Spinoza’ a collection of critical essays, Doubleday
the part of the bourgeoisie could be expected. By
Anchor
Hindess, Barry, Philosophy and Methodology in the Social Sciences, Harvester’

the time of the Russian revolution it was obvious
HIndess, Barry and HIrst, Paul, Pre·Capitalist Modes of Production, RKP;
to Lenin that the parliamentary democracies had
Mode of Production and Social Formation, Macmillan
Husserl. Edmund, Logi-cal Investigations, .RKP
developed sufficiently large, speCialised, powerful
Kant, Immanuel, Critique of Pure Reason, Macmillan
and hierarchical state apparatuses to remove the
Lenin, Vladimir, What is to be Done?, Progress; Materialism and
Emp1rio-Criticism, Progress
struggle for power from the parliamentary site
Mandel, Ernest, ‘Revolutionary Strategy in Europe’, New Left Review 100
altogether (which doesn’t mean that the workers’

Marx, Karl, Grundrisse (for 1857 Introduction), Penguin; ~ Vol.I,
Penguin; ~ Vol. n, Lawrence and wtshart; The Flrst International
movement can’t use parliament, as the bourgeoisie
and After (for The Civil War in France and ‘Speech on the Hague
does, for ideological purposes). The confrontation
Congress ‘), Penguin
Marx, Karl and Engels, Friedrich, The Communist Manifesto in Ilw
between classes in future revolutionary situations
Revolutions of 1848, PenguIn; The German Ideology, Lawrence and Wishart
therefore would take the form, not primarily of
Nietzsche, Friedrieh, Beyond Good and Evil, Penguin
confrontations between parties in parliament, but
between the whole hierarchy of bourgeois state and
economic apparatuses on the one hand, and a net;…

(I would like to thank Martin Barker and Tony
work of democratic workers’ institutions – workers’

Skillen for their comments on an earlier draft of
councils, militias etc – on the other. The essential
this. essay – the former’s comments being mainly
difference between a parliament and the supreme
critical, the latter’s mainly favourable. )
representative assembly in a workers’ state would

BIBUOGRAPHY

’20

4) SiImmary of Issues, and a note on, ‘de ..negations’ of Idealism
.Perhaps it will be useful to conclude by sum marising the anti-epistemological theses of Cutler et aI,
and the antitheses presented in this essay. (I am not
a Hegelian, and don’t believe in the possibility of
‘synthesis’ It is a question of Either / Or. )
Antithesis:

Thesis:

epistemology need only demand that knowledge be
that all epistemology is dogmatic in that it assumes
produced by enquiring into the real, by ‘putting
that some particular method of enquiry has privilquestions to nature’. This requirement is not
eged access to the real.

imposed by any dogma, but by the fact that science
aims at truth, i. e. at ‘matching’ (13) the real.

——————-=~~~~==~~
Antithesis:

Thesis:

on the contrary, the whole purpose of putting questhat any such correspondence assumes a pretions~o
nature is to establish a harmony between
established harmony between discourse and reality.

discourse and reality, which can’t therefore be
pre -established.

Antithesis:

Thesis:

it does not, for the ‘replies’ of nature are results
‘putting questions to nature’ presupposes some
of theory-determined procedures of enquiry, and
theory-independent observation-statements
will therefore originally be couched in terms of the
against which to test theories.

theory, even if they lead to its revision or abandonment. But the results themselves are not determined by the theory, but by the structure of reality.

————————~~—–Antithesis:

Thesis:

these possibilities are not logically exhaustive, and
all epistemology must be rationalist or empiricist
Marx’s methodological sta,tements fall into neither
Rationalism reduces relations between objects
category. Modern scientific epistemology has
to conceptual relations, and empiricism to conworked out an alternative to both more thoroughly
tingent conjunctions. .

– the best example being Bhaskar’s A Realist
Theory of Science.

f

Finally, it is worth looking at one or two of the
passages in Mode of Production and Social
Formation in which Hindess and Hirst deny that.

they are idealists.

We must inSist, confident that we shall be misread, that our rejection of the- epistemological
category of “concrete” is not a denial of the significance or reality of (material) objects. That
denial is a position within epistemology which
substitutes other (spiritual) objects as apprpriate to know ledge. We do not deny the existence of
social relations – that would render our project
absurd. What we reject is the category of
“concrete” as object-of-knowledge. It is the
relation of “appropriation” or of “correspondence”
of knowledge to its objects which we challenge. ‘

(pp6-7)
(a) IdeaIi’sm is taken”here in the sense of idealist
ontology, where it is idealist epistemology which is
at issue. The two are’logically independent.

(b) Saying that their position is not idealist because
it is not epistemology is as fatuous as A. J. Ayer’s
claim that he is not an atheist since on his view it
makes no sense to talk of God, and therefore makes
no sense to deny his existence.

(c) The non-existence of social relations would
render their project no more absurd than would the
non-correspondence of knowledge and its objects.

‘This does not commit us to denying that tables
exist so said Berkeley – A. C. or cause us
intellectual discomfort when we refrain from
walking out of the top windows of high buildings. ‘

(pS)

have lost no sleep worrying about the phYSical
safety of Hindess and Hirst. If idealists acted on
their sophisms, a Darwinian process would have
considerably simplified the task of materialist
philosophers. Unfortunately the dislocation between
theory and reality which idealism excuses does
have practical effects at more complex levels. The
elimination of the bigger part of Marxist theory in
Marx’s ‘Capital’ and Capitalism Today could have
very serious political effects if it convinced the
left.

‘ThIS is not to deny “reality” exists, is ordered,
or to assert that it is infinite and unknowable all of these are sceptical or critic~l positions
within epistemologfcal dis course. ‘(p8)
‘All epistemologies share the conception of an
independently existing realm of objects that may
none the less be correlated with their representations of appropriations in determinate forms of
discourse. To deny epistemology is to deny that
correlation. It is not to deny forms of existence
outside of discourse but it is to deny that
existence takes the form of objects representable
in discourse. ,t (p21)
If it is denied that reality is representable in discourse, it is asserted that reality is unknowable,
despite protestations to the contrary.

The gentlemen must make up their minds. If they
think that knowledge has some relation to reality,
they must tell us what it is. If not, they must say
what the point of knowledge is. If they can’t, they
should shut up.

1

13 I owe thls” expression to Rip Bulkeley.

2i

Buy the newest RP in printDownload the PDF