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.

Lukács and the Marxist Criticism of Sociology

Lukacs aacl tbe MaJlxist Ca-iticislll
of Sociology
This paper is situated in the context of three interrelated arguments. The first and central issue is
epistemological, concerning the grounds upon which
one theory of ‘point of view’ claims to be superior
to others, to represent ‘the truth’, to be ‘scientific’,
to produce ‘knowledge’. The last decade has seen
the emergence, within academic sociology on the
one hand and Marxism on the other, of two radically
opposed positions. The former can best be called
‘agnosticism’: the implicit relativism of much
sociology is taken to its extreme on the grounds that
its traditional dilemma is insoluble. If it is true,
then it is false, since it would then be an absolute
truth; if it is false, then it is true, since relativism
itself would be relative. All that can be done is to
accept it. The result is that no claims at all can be
made about the superiority of one point of view over
another; this seems to me to be a position implicitly
shared by ethnomethodology (where it appears as
‘indifference’ to ‘practical sociological reasoning’)
and by much allegedly ‘phenomenological’ sociology
and by some forms of symbolic interactionism.

Although all these approaches might offer their own
prescriptions for sociological activity, they make
no attempt to repair the relativist dilemma, either
ignoring it or more or less explicitly accepting it. 1
Juxtaposed to this position, we find within Marxism,
and particularly in the work of Althusser, the claim
to be a rigorous science, with developed criteria
of scientificity that clearly distinguish it from other
ways of seeing the world and establish its
superiority as a science.

My starting point is a dissatisfaction with both
positions. There is, I believe, another alternative,
Marxist and ‘dialectical’ in the proper sense of the
word, the basis of which can be found in the work
of the early Lukacs, especially in History and Class
ConSCiousness, and which is rarely elaborated in a
convincing and systematic way. This leads to the
second argument, taking place within Marxism and
concerning the nature of that work, which has been
attacked precisely for its ‘unscientific’ and even
‘anti-scientific’ status. I will take Gareth StedmanJones’ (1971) criticism as representative of this
attack; it displays, I believe, a fundamental misreading of Lukacs, against which I want to offer a
reading based on the work of·Merleau-Ponty (1974a)
developed, unfortunately often chaotically, by the
group of writers around the journal Telos. 2
The third argument concerns the practical co~se­
quences of Althusser’s Marxism. Over the last few
years, a number of articles and books have
appeared criticising sociology from an Althusserian
point of view, 3 all displaying to a greater or lesser
extent certain undesirable tendencies: a tendency to
misread what is being criticised (as Stedman-Jones
misreads Lukacs) a-nd thus not to criticise at all;4
and a tendency to reject in toto the ‘criticised position. This is undesirable on a theoreti~al level
because – as I will try to show – Marxism’s super1 Perhaps the clearest account of the agnostic argument can be found In
PhllUl’s (1974); the a!,’Ilostlc cr1t1clsm of Mannhelm, with which I deal
later, is combined with similar criticisms of Mannheim and Mills.

2 Sac in particular Aralo (1972a, b); Feenberg (1971,1972); Piccone (1972)
3 See in particular Hindess (1972, 1973a, 1973b) and Hlrst (1973)
• IIIndess (1973<1, b) are particularly clear E'xamples of such misreading.


iority lies not in its rejection of ‘bourgeois knowledge’ or ‘ideology’, but in encompassing and going
beyond it; on a more immediate level, it is undesirable for its consequences in theoretical
debate, ‘ideological struggle’. 5
My procedure will be as follows: to begin with,
I will outline the agnostic position through a discussion and criticism of Mannheim’s sociology of
knowledge. Mannheim, like other traditional
theorists in the area, was aware of, and attempted
to avoid the dangers of agnostiCism, and the nature
of his failures provides much of the strength behind.

the modern agnostic position. Further, Mannheim
provides a useful comparison to Lukacs, with whom
he might at first be thought to share a number of
positions: he claims for the sociology of knowledge
a superiority similar to that which Lukacs claims
for Marxism, and each bases his claim on the
nature of a specific social group which, because of
its position in society, is able to produce a more
adequate knowledge of that society than other social
groups. In Mannheim’s case, of course, it is the
intellectuals, in Lukacs’the proletariat.

I will then present an interpretation of Lukacs
that attempts to establish that, on the one hand, he
avoids Mannheim’s failures and robs agnosticism
of its main justifications, and that on the other hand,
his epistemology both foreshadows. and goes beyond
that underlying Stedman-Jones’ argument and modern ‘scientific’ Marxism. Finally, I will attempt
to outline the procedures involved in a Marxist
criticiFlm of sociology based on Lukacs’ work.

These are by no means minor tasks and in the
space of a paper I can make no claim to go beyond
schematic arguments; in particular, I offer no
systematic critique of Althusser, relying rather on
a number of specific implicit and explicit criticisms
although I will try to indicate the direction of a
more systematic criticism. Unfortunately, the position for which I will be arguing suffers precisely
from a surfeit of over-general schematisation. My
justification for much of what is to come is less
its intrinsic merit, but rather that, if understood
properly, it enables the definition of some very
specific and important tasks to which Marxist
theory must address itself.

r”Om Mannheim
10 Agaosticism
The criticism which moves from Mannheim to
agnosticism is really quite simple. According to
Mannheim, the sociology of knowledge replaces
relativism with ‘relationism’; it appears when
different points of view have invalidated each other
by revealing each other’s social roots and consequent ideological characteristics. It offers the only
way out of the impasse by replacing the notiOn of
absolute or eternal truth by one of a permanently
developing truth. Thus:

‘Relationism, as we use it, states that every
assertion can only be relationally formulated.

5 Later in this connection, see Ranclere (1974); I will return to tlds point

It becomes relativism only when it is linked

with the older, static ideal of eternal, unperspe ctivistic truths independent of Jhe subjective experience of the observer, and when
it is judged by the alien idea of absolute
truth’ (Mannheim 1972 p270)
Having asserted the ‘perspectivistic’ nature of
truth, he still attempts to distinguish between at
least more or less adequate knowledge, if not between truth and falSity. The criticism is straightforward: such a distinction is impossible given
Mannheim’s starting point. For MannheiIl,l., more
adequate lmowledge is achieved by synthesising
different points of view into an ever-widening
totality of lmowledge, and he suggests two ways in
which this might be done. In the first place:

‘What has been correctly but differentially
perceived by the two perspectives must be
understood in the light of differences in the
structure of these varied modes of perception.

An effort must be made to find a formula for
translating the results of one into those of the
othe r and to discover a common denominator
for these varying perspectivistic insights.

Once such a common denominator has been
found, it is possible to separate the necessary
differences from the arbitrarily conceived and
mistaken elements which here too should be
considered as errors’ (P270)
Alternatively, we might achieve a synthesis
through recognising the situational determinants
in opposing or differing pOints of view, and, as a
consequence, neutralising them.

Each alternative implies some ‘neutral’ area
(in the second case, apparently immediately visible) between or within different points of view an implicit ‘real world’ against which ‘situational’

determinants can be identified, or a neutral
language which enables the same result. This
implication remains despite his attempt to situate
‘superior lmowledge’ in the perspective of intellectuals as a group: they are simply able to see more
of the real world than others, or have easier
access to t~ neutral language.

In this way, ‘truth’ and ‘error’ creep back, and
having gone this far, it is no surprise that
Mannheim reaches the point of advocating ‘a direct
‘” HE~ bo coR-RECI IPEt!’:> c.~E

~oN? 1:>0 1″‘”tte), b~oP fROf’-I -rItE
SKIE’b? NO. ~ ifte’f INNtlTE


‘SoQ/h… f~IC£ ….

observation of the facts’ (P256) in order to choose
between different points of view. This is in direct
contradiction to the earlier assertion of the perspectivistic nature of all truth which, moreover,
Mannheim has taken to its logical conclusion in
asserting that in an important way, ‘facts’ are
constituted by conceptual structures (P91). 6 Other
formulations – such as the argument that the superior point of view is the more fruitful or comprehensive of the alternatives – are even less satisfactory:

the earlier formulations would imply that the terms
can have no absolute meaning apart from the J»erspectives in which they occur, and they will of
course have different meaI1ings in different perspectives. Thus it appears that we cannot escape
relativism” or agnosticism.

The Lukacs of History and Class Consciousness is
self-evidently concerned with the same problems
as Mannheim, but they are answered in a less obvious way. It is not an easy work and it is not made
any easier by the intrinsic difficulties of Lukacs’

dialectical form of argument; at times his arguments are confused, and a great number of assumptions remain implicit. Superficially, he appears to
reproduce Mannheim’ s central contradiction, insisting that all lmowledge is rooted in its social
context, but then arguing that it is possible to
choose between different points of view, implying
the existence of trans-contextual criteria. StedmanJones (1971 p47) puts forward what is perhaps the
most common interpretation when he sums up
Lukacs’ ‘startling but elegant’ reply to relativism

‘. .. all truth is relative to the standpoint of
individual classes; the proletariat is by its’

essence a universal class; its subjectivity is
universal; but a universal subjectivity can only
be objective’

I want to argue that this formulation is perhaps
the least elegant and that Lukacs’ position is in fact
considerably more complex, often implicit in his
practice rather than explicit in his argument. He
suggests that there are a number of inter-related
features which enable Marxism to avoid relativism
and ensure its superiority over other forms of
knowledge, one of which is a by no means simple
relation between Marxist theory and the proletariat.

I will deal in turn with these features: Marxism’s
grasp of the totality, its ability to transcend the
dualisms of bourgeois thought, and finally its
relationship to the proletariat implied in the notion
of praxis.

(a) The totalising movement of Marxism

E”1W:fiI cA-N You
M~~ l-IKe-iitM?



Lukacs describes the notion of ‘totality’ as the most
fundamental distinguishing feature of Marxism;
whether or not he is correct is less important for
present purposes than what he means by ‘totality’,
and what is involved in Marxism’s grasp of the
totality. There is (or was) no shortage of bourgeois
theories of ‘society as a whole’, so, self-evidently,
there is more to it than that. It becomes apparent
later in the book that ‘totality’ in the frame\Ork of
Lukacs’ thought is related to a more familiar
Marxist claim concerning its method of producing
lmowledge. This production involves a totalising
movement, whtc;n, in relation to praxiS, he calls an
‘aspiration towards totality’ (P198), the first step

Although he Is always careful to avoid a thorough”,oinll: idealism.


of which is the penetration by thought of immediate
‘external’ appearance – the penetration of reifieation and the revelation of the determim tions behind
what bourgeois thought takes to be the given ‘facts’.

To take an oversimplified example, bourgeois
,economics takes as given the ‘facts’ of price and
attempts to explain their fluctuations in terms of
supply and demand (equally ‘given’ as ‘facts’).

Marxism, on the other hand, is able to grasp the
way in which these ‘facts’ came to be in the first
place – to grasp the origins of commodity fetishism:

the way in which human relationships manifest
themselves as relationships between things (price,
precisely, expreSSing a relationship between

This movement beyond the immediate leads to the
revelation of hidden, ‘internal’ relationships relationships unavailable to immediate perception.

To continue the example (and the oversimplification)::

as it reveals the nature of the wages system and the
origin of prices in labour, Marxism at the same
time reveals the foundations of social classes and,
beyond that, through a grasp of the way in which the
mode of production reproduces itself, it reveals
those relationships that are usually subsumed by
the terms ‘base’ and ‘superstructure’. In this way,
each immediately evident ‘fact’ is understood not in
terms of its independent existence, or in terms of
an external causal relationship; rather its existence
is understood as the product-of a number of relationships – a structure of relationships; in Lukacs’

terms, its immediacy is mediated, and these relationships are in turn mediated by others. The
‘totalising movement’ of Marxist thought is precisely the steady exploration of these increasingly
wider, more complex and more fundamental relations, and the discovery of these structural forms
is the production of knowledge – there is clearly a
variety of realist epistemology underlying Lukacs’

understanding of the Marxist analysis of social
formations, one which sees access to the underlying
structures of the social world as a task of theoretical construction.

Thus what appears to bourgeois thought as ‘real’

– eg price – is shown to be abstract – ie isolated
from the totality of relationships that determine its
existence – and what appears to be abstract – eg
social classes (since we cannot point to an existing
object which is a class) – is shown to be real – ie a
concrete determim tion of external appearances.

This seems to follow closely Marx’s own methodological statement in the introduction to the
Grundrisse. frequently quoted in part by Lukacs,
but worth quoting in full:

In other words, we start and finish with ‘reality’,

the ‘concrete’ (or perhaps better, we start with
‘reality’ and finish with the ‘concrete’), reconstruct·
ing it, in the process of theoriSing, through the
relationships into which we place the immediately
appearing phenomena. There is a significant difference here between Mannheim and Lukacs: apart
from the one qualified assertion noted earlier,
Mannheim nowhere investigates what is involved in
the ‘theoretical construction of reality’ – thought
remains a reflection of reality from a particular
angle. For Lukacs, however, the reconstruction of
‘facts’ in theory is fundamental:

‘. .. integration into the totality . .. does not
merely effect our judgment of individual phenomena decisively. But also as a result, the objective structure, the actual content of the
individual phenomenon – as individual phenomenon – is changed fundamentally. ,7
Whereas for Mannheim, differences between
theories are differences between points of view on
the same ‘facts’, for Lukacs the differences lie
in the extent and method of the reconstruction of
the ‘facts’ and there is thus no road by which he
can return to advocating a direct examination of the
facts. In this case at least he escapes Mannheim’ s
contradiction and, .moreover, escapes without the
implicit acceptance of agnosticism – the view that
the ‘real world’ (if there is one) remains unknowable. The totalising movement of Marxism, the
ability to understand and penetrate reification and
grasp the way in which the facts we are reconstruct·
ing are produced as facts in the totality of their
determilnations, reveals that these facts are the
product of human activity and that social classes
are the subject of that activity. Thts places the
‘facts’ we study in the ongoing movement of

‘. .• the function of these unmediated concepts
that have been derived from the fetishistic forms
of objectivity is to make the phenomena of capitalist society appear as supra-historical essences.

The knowledge of the real, objective nature of a
phenomenon, the knowledge of its historical
character and the knowledge of its actual function
in the totality of society form, therefore, a
single, undivided act of cognition. ‘

(p14; see also pp142-9)
It is the mediation of history that enables us to

grasp the concrete interpenetrations of the dualisms
which have typified bourgeois thought.

For r.;ukacs, then, what Mannheim calls the
‘activist’ element of thought is an awareness that
, . .. it seems to be correct to begin with the
real and the concrete, with the real precondition, thought is a reconstruction of reality and since that
reconstruction leads us to history, we realise that
thus to begin, in economiCS, with e. g. the poputhought is not the only activity involved in the relation, which is the fowldation and the subject of
construction of reality, that it is only one moment
the entire social act of production. However, on
what Marx called ‘sensuous human activity’. It
closer examination, this proves false. The population is an abstraction if I leave out, for example, is this activity – and not thought alone – which
gives us access to the ‘real’ (in the sense of
the classes of which it is composed. These
‘objective’) world.

classes in turn are an empty phrase if I am not
The totalising movement of Marxism thus leads
familiar with the elements on which they rest.

us directly to the transcendence of the dualisms of
E. g. wage labour, capital etc. . .. The concrete
bouirgeois thought and to the concept of praxis.

is concrete because it is the concentration of
many determinations, hence unity of the diverse.

(b) The Transcendence of dualisms
It appears in the process of thinking, therefore,
as a process of concentration, as a result not as
The most significant dualism that Marxism transa point of departure, even though it is the point
cends, according to Lukacs, is that of subject and
of departure in reality and hence also the point
object, although, as we shall see, a number of
of departure for observation and conception. ‘

other dualisms are left behind – in particular those
(Marx 1973 pp100-101)


of truth/falsity, relative/absolute, thought/reality.

It is the subject/object dualism and the attempt to

overcome it which provided the impetus behind the
development of bourgeois philosophy and although
the solution remained unavailable, the problem was
pursued to the point where it became possible for
Marxism to realise the transcendence.

There is an implicit and, in the last analysis,
superficial point of contact here between Lukacs’

pOSition and that of Althusser. The latter also sees
bourgeois philosophy in terms of its entrapment in
the problematic of subject/object, the ’empiricist
problematic’, of which traditional empiricism and
traditional rationalism provide two variants.

Lukacs is primarily concerned with German idealist
philosophy, and in particular with the development
of the contradictions of Kantian rationalism, but in
the course of his argument he too reveals the close
relationship between the two variants, the way in
which an internal dynamic can lead to the transformation of one into the other. For Kant, the
starting point is not that our knowledge must conform to objects but that objects must conform to
our knowledge: he returns to the familiar principle
that we can only know what we create; the object
‘in-itself’ remains unknowable and the world that
we know is the ‘product’ of rational mind. But there
is a ‘necessary correlation’ between the principles
of rationality and irrationality according to Lukacs,
and when rationalism claims to be the universal
principle, the correlation becomes crucial and
erodes the whole system. In Kant’s case, the
irrationality, the course of the erosion, is the unknowable ‘thing-in-itself’; there is a radical separation between content and form, the overcoming of
which would require the development of a concept
of praxis. Failing that, all that can be produced
from Kant’ s starting point is a structure of highly
systematised formal laws which, because they are
laws, exclude the creative subject (cf. ‘Reification
and the Consciousne~s of the Proletariat’ Part 11
passim). Knowledge thus becomes a matter of the
passive contemplation of something outside the
knower. This analysis is the basis of Lukacs’

criticism of the way in which the natural sciences
conceive of their activity and of Engels’ use of the
natural sciences to refute Kant’s conception of the
‘thing-in-itself’. His reference to Comte (P154)
indicates that he sees positivism as (me possible
attempted solution to the dilemma of traditional

Unlike Althusser, Lukacs does not step outside of
(suppress, conceal) the dualism but attempts to
transcend it through a grasp of the concrete unity of
subject and ob~ect in the movement of history,
reached in the totalising movement of Marxist
thought (cf. p14 and pp142-9). Through history we
c~n grasp the way in which the subject produces itself as object, the way in which humanity creates
itself. Marxism further reveals that it is because of
the internal expansionist dynamic of capitalism that·
a proper grasp of the totality becomes pOSSible, and
that for the first time the possibility exists that
history may be made consciously, that the proletariat, because of its pOSition in the production process, is the first class in history capable of
becoming conscious of itself as the subject and
object of history at the same time. Conversely,
because of its position in the production process,
the bourgeOisie is unable to achieve a theoretical
transcendance of the subject/object dualism

The precise nature of this unity of subject and
object requires further exploration, for it is here
that we come up against the central difficulty of
dialectical thought that Lukacs himself warns us of:

‘It is of the essence of dialectical method that
concepts which are false in their abstract onesidedness are later transcended. The process
of transcendance makes it inevitable that we
should operate with these one-sided, abstract
and false concepts. These concepts acquire
their true meaning less by definition than by
their function as aspects that are then transcended in the totality. Moreover, it is eTen
more difficult to estabUsh fixed meanings for
concepts in Marx’is improved version of the
dialectic than in the Hegelian original. For if
concepts are only the intellectual forms of
historical realities, then these forms – onesided, abstract and false as they are, belong
to the true unity as genuine aspects of it. ‘ .(p. xlvi)8
In this context, there is the possibility of two
non-dialectical misreadings of a dialectical text.

The first restores in toto the dualisms transcended
in the dialectic. Gareth Stedman-Jones takes this
tendency to its extreme: where Lukacs is attempting to transform, in a new synthesis, previously
antiilbmic tendencies of bourgeois thought into a
dialectical science of society, Stedman-Jones sees
only an anti-scientific Romantic coming into confused contact with Marxist science (and in the process, he reduces Marxist science to the positivism
that Lukacs is attacking). Where Lukacs tries to
develop a political practice and a theory of organisation from his philosophical pOSition, as he. does in
the later essays in History and Class Conscious~ and also – arguably – in his book on Len~n,
Stedman-Jones can see only an unbridgeable gap. 9
The second misreading involves the opposite
tendency: to read the claim to have transcended the
dualism as an assertion of the Simple, immediate
unity of opposites. In the case of the subject/object
dualism, Lukacs can be read as asserting an a
priori unity rather than a unity grasped in the concrete, in history. Matters are not helped by the
fact that this is the way in which the older Lukacs
came to read himself (p. xxxlii). Stedman-Jones
again takes this misreading to its extreme, taking,
without understanding, Lukacs’ argument that a
change in the self-consciousness of the proletariat
involves a change in the world that is knpWll through
that self-consciousness and transforming it into
the crassest idealism. According to Stedman-Jones
Lukacs’ argument is that:

‘. .. once the proletariat fulfils its vocation as
the identical subject-object of history by
acquiring an adequate consciousness of capitalist society, it abolishes capitalism in a final
interiorisation of it. The exact analogy of this
procedu’ie with the movement of Hegel’ s· Spirit
needs no emphasis. All that it omits is the
brute, material ‘struggle for power … ‘

(Stedman-Jones 1971 :p53)
‘1 Paul Piccone (1972) suggests a TatheT odd Teading of·Lukacs’ notion of
totalisation, implying that it involves a simple compound of immediate
appeaTances. It is difficult to see how this stands up to the quotation.

The ‘TeconstTuction’ of course takes place in thought; the ‘external’ world
remains the same as it comes to be known and is ‘transformed’ Insofar as
that knowledge is a moment of praxiS in the world. Further dimensions of
the Telationship of ‘sameness’ and ‘tTansformation’ will be explored later.

8 The last sentence hints at an important point on which Lukacs can be criticised: a regular confusion of the l’elationship between theoretiCiU and
histoTical development.

9 On the second point, see in parttculaT Feenberg (1972), Piconne (1972).

See also the latte.r’s ~lemic·a~t.S~edmanrJ?ne8 (ppU9:20).


Of course, Lukacs holds no such bizarre position:

‘For a class to be ripe for hegemony means that
its interests and consciousness enable it to organise the whole of society in accordance with
those interests. The crucial question of every
class struggle is this: which class possesses
this capacity and this consciousness at the
decisive moment? This does not preclude the
use of force. It does not mean that the class
interests destined to prevail and thus to uphold
the interests of society as a whole can be
guaranteed an automatic victory. On the contrary,
such a transfer of power can often only be brought
about by the most ruthless use of force (as e. g.

the primitive accumulation of capital). But it often
turns out that questions of class consciousness
prove to be decisive in just those situations where
force is unavoidable and where classes are locked
in a life-and-death struggle. ‘

(Lukacs 1971 pp52-3)

produced by praxis and underlying immediate appeal’

ances. Such an analysis is the task of Marxist
science and this will be elaborated later; however
its foundations and its necessity are revealed preCisely by a philosophy of history (or ‘praxis), by an
anthropology which produces a dialectical conception
of the subject/object relationship. Such a conception
seems beyond the comprehension of Althusserian
thought and the absence accounts for its apparent
inability to attempt any sort of founding enterprise
in relation to Marxist scientific knowledge; instead,
we are left with a number of magically bridged
gaps, the epistemological break, the chasm between
the real object and the object of knowledge. It is
this dialectical conceptualisation which, I would
suggest, represents the real break with what
Althusser calls the empiricist problematic, not
leading directly to Marxist science but providing
the foundation for it, providing an initial revelation
of the ‘area’ t~at will constitute its object. 11
It is now possible to Identify a second way in
which Lukacs avoids Mannheim’s contradictions and
agnosticism. The latter both remain caught within
the same fundamental dualism. Mannheim attempts
two false solutions which from within the dualistic
problematic appear mutual exclusive. It is possible
to reinterpret his position in terms taken from
Lukacs’ critique of Kant: in the first place he
(Mannheim) holds that objects must conform to our
knowledge of them, and he talks about the way in
which our ‘categories’ can construct facts: he then
returns to the contemplative stance that is the
reverse side of the same pos ition – knowledge is a
‘reflection’ of the object, albeit from different
points of view, and since both knowledge and object
must conform to a system of objective laws
(implied but not explicitly recognised by Mannheim)
the situational elements may be eliminated and
‘truth’ discovered. The transformation the development of which is traced by Lukacs remains ‘behind’

Mannheim’s work as a static contradiction. Agnosticism, on the other hand, gives priority to and confines itself to the first half of the contradiction. 12
It should be evident by now that on the basis of
Lukacs’ work, it should be possible to construct a
coherent synthesis of both positions; it should become evident that such a synthesis transcends each
pOSition insofar as it is able to grasp thought as one
moment of a wider praxis.

But the problem still remains of exactly what the
transcendance of the dualism means. There can be
no doubt that he talks of the identical subject/object,
but it is app~rent that it is meant in its peculiar
Hegelian sense. Lucien Goldmann (1971) talks about
the partial identity of subject and object, and although he makes no attempt to conceptualise the
relationship, he offers a clue. Despite his use of
the term ‘identical’ and his later self-criticism,
Lukacs does not offer an a priori unity but insists
that it be discovered in history; in other words it
is a mediated unity, the primary mediation being
history itself. The proletariat becomes conscious
of itself as both subject and object through the
mediation of the developing social and economic
structures of capitalist society, through the mediation of its own developing relationships with other
classes. Its cti,scovery of itself as an object produced in the past, and as the object of present
production processes, is the precondition of its
action as subject to produce itself as a new, future
object through a transformation of those production
processes and their accompanying social structures.

The relationship of subject and object is in this
sense one of identity and separation at the same
time 10 – a formulation which Lukacs approaches
on several occasions (cf. esp. p142).

The fact that the subject/object identity is
mediated – by economic and social structures and
(c) Praxis
other classes – provides the ‘space’ for the ‘brute
material struggle for power’. It further makes
necessary a comprehensive grasp of the mediating
It was argued earlier that the ability to penetrate
structures – there is no question of an immediate
reification led theory to become conscious of itself
reading of iessences’ from history:

as one moment of ‘sensuous human practice’. It
is this practice as a whole, uniting various levels
‘. .. the categories of dialectics must be applied
of ‘thought’ and ‘action’ that Lukacs seems to
to man as the measure of all things in a manner
embrace in his concept of praxis; praxis is what
that also includes simultaneously a complete
‘abstract thought’ on the one hand and the
description of the economic structure of
the world of structures and appearances,
bourgeois society and a correct knowledge of
into a developing totality that transthe present. For otherwise, any description
11 This is an appropriate point to say something about the nature of Lukacs’

will inevitably succumb to the dilemma of
conception of totality – which depends in turn on the concept of mediation.

Mediation implies the existence of elements which are united in their sepempiricism and utopianism, of voluntarism and
aration, just as the subject/object unity is mediated. Far from reducing
each aspect of the totality to an essence, mediation maintains at the same
fatalism … ‘ (pp190-91)
Thus the initial theoretical penetration of reification
reveals that history is the product of a human praxis
and at the same time it reveals the power of, and
the ne~d fo_r, ~ _preCise analysis of those structures
10 Arato (1972a) presents a sophisticated discussion of the problem, although
largely unsympathetic to the standpoint taken here; CoUctti (1973) offers
a more sophisticated version of Stedman~Jones’ Interpretation. Revai
(1971), in an early revtew, despite – or because of? – “Js Hegelian
position offers a formulation similar to that offered here.


time the unity of the totality and the relative Independence of its levels:

‘ •.• the category of totality does not reduce its various elements to an
undifferentiated uniformity, to identity. The apparent independence and
autonomy which they possess in the capitalist system production is an
illusion only insofar as they are involved in a dynamic dialectical relationship with one another and can be thought of as the dynamic dialectical
aspects of an equally dynamic and dialectical whole’ (Lukacs 1971:12-13;
my emphasis)
12 In his criticism of relativism, Lukacs demonstrates how the irrationality
of the unknowable ‘thing-In-itself’ is a central assumption, and how it
(relativism) rests on an implicit absolutism – an absolutism of ‘man’;
thus the necessity for the categories of dialectics to be applied to ‘man
as the measure of all things’. See esp. ppl86ff.

cends both. It is here th3.t we find the relationship
between Marxism and the proletariat, the nature· of
which Lukacs tends to gloss; to this extent, there is
a justification for some of Stedman-Jones’ criticisms (cf. esp. Stedman-Jones 1971 p48), but there
is already a reply implicit in History and Class
Consciousness, particularly in the final essay,
‘Towards a Methodology of the Problem of
Organisation’, and it is made explicit in a striking
way by Merleau-Ponty.

Lukacs moves to and fro between the most
abstract levels of philosophy and an analysis of the
class position and experience of the proletariat.

The implication is not that the proletariat can
immediately and spontaneously achieve a fully
developed theoretical awareness of its position,
otherwise there would be no point in Lukacs turning, in the later essays, to the problem of the
nature of the revolutionary party. The relationship,
as always, is mediated. The proletariat is a class
~hich in its immediate experience of acting in the
world, achieves a practical consciousness of itself
as subject and object at the same time; its very
existence opens up the possibility of a theoretical
transcendance (carried out by theorists) of the subject/object dualism and – since the existence of the
proletariat presupposes the existence of capitalism
as a potential world system – the possibility of a
theoretical grasp of the totality of social relationships as a coherent whole. The existence of the
proletariat reveals the way forward for thought; in
turn, thought reveals the way forward for the proletariat – i. e. raises to a theoretical consciousness
its potentiality as the revolutionary subject of history, and articulates the means by which that potentiality may be realised and the problems involved in
that realisation. In this sense, Marxism as a
theoretical system is fully conscious of its social
roots and its existence as one moment of a wider
praxiS, and it gives a concrete a;-ticulation to this
consciousness precisely in the revolutionary party,
where theorist and proletariat come together’, the
former to theorise the experience of the latter, the
latter to base its practice on the knowledge
produced by the former. 13
If thought is one moment of a developing totality,
its ability to grasp the concrete conditioned by the
concrete, and its relationship to the concrete articulated in a praxis that changes the concrete, then
there is a constant movement of thought from an
adequate to an inadequate and from an inadequate to
an adequate grasp of the concrete; in other words,
thought moves from ‘true’ to ‘false’ apd from ‘false’

to ‘true’, from ‘absolute’ to ‘relative’ and from
‘relative’ to ‘absolute’. Knowledge is ‘true’ insofar
as it claims to offer only an incomplete grasp of its
object, ‘false’ insofar as it claims to offer a total
knowledge; ‘absolute’ insofar as it has reached the
limits placed upon knowledge by historical development, ‘relative’ insofar as that development will
continue and the limits placed upon knowledge
change; ‘true’ insofar as it was once ‘absolute’,
‘false’ insofar as it is now ‘relative’; ‘absolute’

insofar as it is ‘true’, ‘relative’ insofar as it is
only partially ‘true’.

If Marxism is the possibility of the full selfcpnsciousness of the proletariat, then it must be
able to take account of its own relativity, its own
developing falseness, and again it is through the
revolutionary party that this is achieved: through

its relationship to the class’s experience of the
developing social totality, it is able to theorise that
experience and development. It is here that we can
understand the meaning of one of Lukacs’ most
.obscure statements:

‘Let us assume for the sake of argument that
recent research has disproved once and for all
everyone of Marx’s individual theses. Even if
this were to be proved, every serious ‘orthodox’

Marxist would still be able to accept all such
modern findings without reservation, and hence
dismiss all of Marx’ s th~ses in toto – without
having to renounce his orthodoxy for a single
moment. .• Orthodoxy refers exclusively to
method. ‘ (p1)
This is immediately dismissed as nonsense by

‘In fact, such a credo would simply be an intellectual suicide for Marxism: what scientific
method in history has been able to survive the
systematic disproof of everyone of its findings?

What possible charter could there be for it?’

(Stedman-Jones OPe cit. p47)
AllOwing for Lukacs’ deliberate exaggeration, the
charter is this: as the totality of social relationships
develops, so our knowledge of it will move from
adequacy to inadequacy; through its relationship
to the practices conditioned by, comprising and acting upon that totality, (i. e. through praxis) Marxism
is able to revise itself in a consistent rather than
an arbitrary or piecemeal way. Thus the superiority of lVlarxism does not lie in its individual·statements about the world, but in its ability to revise
. those statements as soon as they become inappropriate without creating destructive internal contradictions, without fragmenting its insights; uniting
each through its totalising movement which is not
achieved once and for all but is rather a ceaseless
movement of totalisation and retotalisation
(cf. esp. p24).

Thus there is a complex of reciprocal developing
relationships between ‘thought’ and ‘reality’, the
motor of which, identified at its most general level,
is ‘praxis’. It can be seen now that Lukacs avoids
agnosticism by breaking with the latter’s ‘problematic’, directing us to a complex of new and different
questions. For agnosticism, there are two poles:

‘thought’ and ‘reality’, and problems can only concern the correspondence or lack of correspondence
between them, the problem of ‘truth’ and ‘falsity’.

From Lukacs’ pOSition, truth and falsity ~re intertwined and the questions that enable a grasp of that
intertwining concern the conceptual structure of
knowledge, the extent to which it remains caught in
or goes beyond immediate appearance, its awareness of its own multidimensional relationships with
what it grasps as the ‘concrete’, its articulation
with other moments of praxis.

“~ SYMP/’I1lff:Tic,


1oE ~l1l a naive spontaneist and a Leninist conception of the revolutionary party – not, as Pleco.le suggests, a straightforward capitulation
to the Stal1nist ‘seeds’ present in Lenin’s thought.



The Marxist Criticism
Non-Marxist Knowledge
It might still be possible to interpret the exposition

of Lukacs to this point as implying that Marxism is
superior to bourgeois thought because it meets
certain criteria that the latter fails to meet. If this
were the case, he could still be accused of reproducing Mannheim’ s contradiction, of putting
forward absolute criteria by means of which we can
assess knowledge in an argument explicitly asserting that knowledge is always rooted in a changing
social context and therefore itself always changing.

However, none of the features of Marxism so far
discussed can be regarded as ‘objective’ measures
against which we can lay Marxism on the one hand,
bourgeois thought on the other and find the latter
wanting. It must be emphasised that they are all
features internal to Marxism, not properties
achieved once and for all but potentialities the
fulfilment of which is an ongoing (and never ending)
process. It has already been argued that the movement of totalisation must be conSistently renewed;
and the renewal of this movement involves a steady
deepening of our grasp of the relationship between
the dualisms transcended by Marxism; for example,
one need only look at the work of Merleau-Ponty
and Sartre to realise that Lukacs’ grasp of the
subject/object relation remains at a very simple
level. 14 In the same way, the fully self-conscious
relationship .between theory and practice in the
revolutionary party is not an immediate achievement, but a permanent task, an ongoing struggle
that may be diverted into innumerable forms and
combinations of dogmatism and blind activism.

This can be established by the most casual glance
at the history of the revolutionary movement.

If these criteria are potentialities within one
form of thought, then how can they provide the
basis for establishing the superiority of that form
over others? Since they must be defined by the
form that contains them, what relevance can they
have for other forms? Lukacs offers, but does not
elaborate (except implicitly in his own critical
positions) an answer. Marxism does not reject
bourgeois. knowledge but accepts it and goes beyond
it, including it as a moment within Marxism:

‘If, then, the standpoint of the proletariat is
opposed to that of the bourgeoisie, it is nonetheless true that proletarian thought does not
require a tabula rasa, a new start to the task
of comprehending reality and one without any
preconceptions. In this, it is unlike the thought
of the bourgeoisie with regard to the mediaeval
forms of feudalism – at least in its basic tendencies. Just because its practical goal is the
fundamental transformation of the whole of
society, it conceives of bourgeois society together with its intellectual and artistic productions as the point of departure for its own
method. ‘ (P163)

And again:

‘ … the “falseness”, the “one-sidedness” of
the bourgeois view of history must be seen as
a necessary factor in the systematic acquisi• tion of knowledge about society. ‘ (p164h5, _ __
14 Sartre’s wOl’k (Cb!!. Sarlre 1960) develops a number of notions that remain
inlpllcit. ill lli.,h;I·Y anti Class Consciousness – in particular, the concept
of ItJtalisatlon.

15 Thc’se quotations ha’e sC’U-evident implicatlons for Stedman-Jones’

argum(‘nt that for Luk.”lcs, capitalism is never progressive. It should be
clear by /lOW th:lt eve:rything that Luk.”lcS says about Marxism requires the
existence of capitalist society and Its advances over pre-capitalist


These points indicate a central danger in the way
in which I have approached the problem: if the
superiority of Marxism is contained within certain
of its potentialities, the establishment of that
superiority must be in the realisation of those
potentialities – in the ‘practical’ criticism of a
specific study or theoretical system” not in the
formal exposition of the potentialities. It is this
that gives the discussion of Lukacs so far its
rather glib, self-satisfied air. It has remained a
purely formal exposition and as a consequence is
likely to appear a purely verbal solution, particularly since the all too obvious difficulties of a
‘practical’ realisation have not been considered
(and in the last analysis, these are the difficulties
of the revolution itself). As I indicated at the begin.

ning, the justification for approaching the issues
in this way is that it enables a preliminary definition of the problem that is an alternative to the
Althusserian definition in terms of ‘scientificity’

alone. The significance of the problem should not
be underestimated: it is the problem of what
exactly we are doing and should do when we engage
in criticism or debate with non-Marxist points of
view, the problem of theorising an everyday practice vital to our political action. Even at the
present level of generality, the discussion has not
been pursued as far as it could be, and there are
still apparent contradictions in the argument: what
are we to make of the statement that Marxism must
accept bourgeois knowledge as its starting point in
relation to the earlier comments about the reconstruction involved in theorising? How can knowledge
be accepted and remain the same, and be reconstructed at the same time? The examination of this
problem will reveal a major theoretical gap in
Marxism; a revelation which is perhaps the most
Significant result of the present project. I will continue to confine myself to the discussion of the
Marxist criticism of sociology, since it was developments within sociology that occasioned this paper
in the first place, and as I indicated earlier, it is
the area in which much Marxist criticism has
recently appeared.

Perhaps the best starting point is the dismissal
of a fallacy common to both Mannheim and agnosticism: that to situate a work in its social context,
to reveal its social roots, is by and in itself a
criticism of that work, a proof of its lack of ‘objectivity’. Tbe consequence of this assumption – which
in turn implies the whole framework of relativism
and agnosticism – is a concern to reveal only that a
work does have social roots, at the expense of any
grasp of the complexity of its relationship to those
roots. A Marxist form of the same fallacy involves
the dismissal of non-Marxist thought as ‘bourgeois’

or ‘academic’, almost invariably in the form of a
dogmatic assertion since to argue the position
through to its logical conclusion would lead to an
unacceptable relativism.

Althusserian critiCism, on the other hand, rejects
the fallacy, but at the cost of not Situating the work
in any satisfactory way at all. Instead, criticism
is confined to the conceptual level with the result
that eventually the very reification that Marxist
theory begins by penetrating is re-introduced at
its very centre. W~th, perhaps, the exception of
the conjuncture which makes an epistemological
break pOSSible, 16 knowledge and its development
16 And when it comes to concrete conjunctures, such as that which saw the
appearance or .Marxist Science, Althusser’s specificatior.s .lrc very loose,
amounting to htlle more than a gloss on Lenin’s generalisations.

(see Althusser: 1969,1974)

become contingent, unintelligible in the sense that
Kantian rationalism ends as unintelligible from the
perspective of its starting point: knowledge develops according to, and its correctness is guaranteed
by, formal laws which ‘just happen to be the case,17
Criticism becomes a matter of reducing a work to
its ‘problematic’, the formal rules of discourse that
make it possible; a work is read only to discover
these rules behind it. 18
To overcome this unintelligibility, two steps
must be taken. The first is taken by Lukacs himself: the rules of discourse, the problematic, must
be related to the immediate experience (practical
consciousness) of a particular class. However,
once a problematic has been articulated, there is
no necessary relation between the class position or
origins of the individual theorist and the fact that
the work occupies a position within that particular
problematic; the first step is a very general situating which points the way to the second step, that of
situating the work in the more immediate social
context within which it is produced.’ If we remain at
the level of the underlying problematic and its general social origins, the specificity and complexity
of a position within that problematic are lost. A
work is only truly intelligible when we can grasp through the project oflt~e writer and the conditions
in which it is pursued – why it should occupy
this position within this discourse, why it is produced at this time by this individual, why it should
have these pecularities, features not directly necessitated by the underlying problematic. Such an
exercise, not in itself criticism, is a necessary
preliminary. What is unintelligible, what cannot be
understood, cannot be coherently criticised. The
next step is the conceptual criticism itself, which
is a matter of identifying the partial and one-sided
nature of what is being critiCised, its ‘unconscious’

relativity and its internal contradictions, in turn
rooted in its reliance on unreconciled dualisms.

There are a number of different ways in which
knowledge may present a partial and one-sided
view of its object, each rooted in its failure to go
beyond the immediate. Lukacs does not deal in a
detailed or systematic way with the identification
and transcendance of ‘partial’ knowledge, but an
examination of his arguments in relation to the
fragmentation and false separation of disciplines
in the social studies, combined with the earlier
exposition, makes it possible to follow up more
rigorously some of its implications. The unification
of the fragmented disciplines is not simply a matter
of adding together, rather they must be ‘trans'”
formed inwardly’ by an ‘inwardly synthesising
philosophical method’ (Pl09). It might be inferred
from such an argument that Marxism is apbilosophical method and nothing else, and that knowledge produced by science is worthless until it has
undergone a philosophical transformation. Lukacs’

critique of scientific knowledge tends to support
this interpretation: the subordination of philosophy
to science is seen as a product of reification: to
accept uncritically the results of science is to
accept the reified world. Yet there is a lack of
clarity in his argument, almost a contradiction. that
17 Cf. esp. Glucksmann (1972)
18 Hindl’ss illustrates the danger (lUndess 1973a, 1973b). In the former,
HlH::serl’s ‘Crisis’ is reduced to the problematic of ‘transcendental
empiricism’ when its Significance for the development of phC’nomenology
is precisely its apparent partial abandonment of transcendentalism.

In the latter ethnomethodology is condemned as leading, through its notion
of ‘background assumptions’ to an infinite regress: in fact the starting
pOint of ethnomethodology is the rec6gnition of such an infinite regress as
inevitable. It then seeks to identify the ‘glossing’ operations which are
employed to avoid it – a project which Hindess Simply faHs to recognise.

19 These conditions arc, of course, exceedingly complex; a network
comprising all levels of the social fOl·mation.

leaves room for a different conception of science.

We have already seen that he recognises that the
reified world is open to analYSiS, indeed that such
an analysis is necessary and that it is founded upon
his philosophical method. At the same time, he
states that his argument is based on the analysis
that Marx presents in Capital. 20 If the latter is
Marxist SCience, a ‘truly scientific study’ (P8), then.

it appears to have its own specific role ‘within’

Marxist philosophy, the relationship of dependence /
independence between them re~aining to be properly
concept~alised. The followjng paragraph is quoted
in full because it shows the possibility of my interpretation and the ambiguity of Lukacs’ position:

‘Marxism, however, simultaneously raises and
reduces all specialisations to the level of aspects
in a dialectical process: this is not to deny that
the process of abstraction and hence the isolation of the elements and concepts in the special
disciplines and whole areas of study is of the
very essence of science. But what is decisive
is whether this process of isolation is a means
towards unq.erstanding the whole and whether it
is integrated within the context it presupposes
and requires, or whether the abstract knowledge
of an isolated fragment retains its “autonomy”
and becomes an end in itself. In the last analysis,
Marxism does not acknowledge the existence of
independent sciences of law, economics or
history etc: there is nothing but a Single, unified
– dialectical and historical – science of the
evolution of society as a totality. ‘ (P28)
The implication is that the philosophical method
is the ‘highest’ moment of totalisation and that its
object is precisely the knowledge produced by
Marxist science; there are also implicit and explicit
distinctions between Marxist and non-Marxist
science, and this in turn implies that a preliminary
transformation of the knowledge produced by the
bourgeois disciplines is necessary before the philosophical transformation may be operated. 21 These
different levels of ‘science’, the philosophy / science
relationship, and the different levels of transformation are frequently blurred by Lukacs. For our
present purposes, sociology can be taken as
‘bourgeois science’, and the lower level transformation is the problem. I will try to work out an
initial, schematic clarification of what is involved,
firstly in relation to empirical sociology, then to
theoretical sociology.

Sociology’s entrapment in the immediate is most
obvious in what is conventionally called ’empirical’

sociology: sociology concerned with revealing,
describing and relating ‘facts’; it lies in the deliberate isolation of ‘facts’ as dependent and independent variables, the dividing up of the empirically
‘given’, the separation, for example, of ‘social
interaction’ from ‘social structure’ and the separation of different aspects of human activity from
each other; the sub-disciplines of sociology – the
sociology of the family, of deviance, of religion,
industrial sociology etc – are excellent examples of
the fragmentation of a discipline according to
’empirically given’ (leaving open, for the moment,
what is meant by ’empirically given’). It lies also
in the relationships that are sought between the
‘facts’ in, for example. statistical correlation. or
20 Cf. esp. pM. This relationship of mutual foundation between phUosopliy and
science is much cleal’er in Sartre (1960) although, like Lukacs, he is
primarily concerned with the philosophical totalisation. It is a concf’pUon
that owes a debt to Begel, and the specification of the differences between
the Begellan and Marxist verstons would be an intriguing and far-reaching

21 The distinction implicit in Lukacs between the natural and social Sciences
indicates that a different type of transformation will be required for each.


vaguely understood ‘interpretive procedures’, the
latter either assumed (eg Goffman) or sought
empirically (eg ethnomethodology). These are
‘formal’, ’empty’, ‘abstract’ relationships in the
sense that there is little or no attempt at rigorous
conceptualisation – they are taken for granted or
presented in a ‘story telling’ form. The fragmented
and partial nature of such work is self-evident; the
task for Marxism is to integrate these partial
descriptions into a totalising thought in which the
‘meaning’ of the starting point is inevitably transformed.

Now this transformation is not a straightforward
transformation of the ‘object of knowledge’, in
which we end with a different ‘object’, abandoning
the one we started with. The latter remains because
it is ‘given’ not only in thought but also in the world,
it is an aspect of the way in which the world initially
presents itself to us. If this were not the case, if it
were given only in thought (ideology), or if it were
given by the world and passively perceived, we
would fall back into the dualistic irrationalism
already described. The transformation is not of the
‘given’ in either of these senses but of the mediations grasped as determining it. The most abstract,
general mediation that is transformed is the specific
form of the subject/object dualism that lies behind
the production of the partial knowledge with which
we start. The knowledge produced by positivist
sociology, for example, rests on the separation of
the object and a contemplative subject; the participation of the subject in the creation of the object is
ignored and part of the work of Marxist criticism is
to reveal this contribution. Perhaps the simplest
example concerns the results of questionnaire
surveys: significant here are the ways in which the
choice and organisation of questions limits possible
answers, the effects of the construction of the
setting in which the questions are asked, the less
general theoretical presuppositions behind the
questions and the interpretive procedures involved
in making sense of the answers. 22 Many of the
tools needed for such criticism have already been
developed by the various interpretive ‘alternatives’,
and by Althusserian Marxism itself, but with the
aim of rejecting rather than re-interpreting the
product of positivist procedures. 23 In the case of
the various forms of interpretive sociology, it is
again a matter of drawing out the various interpretive procedures and their theoretical presuppositions!

and also of relating the observed interaction to the
social and material structures that produce its
setting and condition its progress: it is a matter
of reintroducing the world of objects – and not only
those of immediate experience.

This is, of course, a very general outline of what
could be the first steps of the critical movement;
it is intended to give a very general idea indeed of
what is involved in the totalising process. However,
it enables two important points to be made. The
first is that it is clear that the ‘originally given’

object (questionnaire results, the observed interaction) remains at the same time as it is transformed, reconstructed. As the mediating structures
underlying its production are explored, it takes on
a new meaning: it is no longer, for example, the
ultimate point of reference against which any theory
22 For exarnp!’?, qllcsti0llinr; resp0l.denl” sinr.:ly rather than In groups is
likely to predetermine types of anSWl’rs; Sartre’s concept of serialisation
providE’s a theon’tieal basis for re-interpretatlon In this case.

Wesi.ergaard (l!f;O) dUTlollstrates an irnplicilfault in the interpretive
procedures of the Affluent Worker study.

23 It is quite posbible that such a reinterpretation will tell IJS as much about
positivist sociology and its techniques as it does about the objed of those
te(“hnlques. Sartre (1963a) argues that the results ~ a rigorous analytical
study are the most easily integrated into Marxism.


must be measured, as it is for most forms of
empirical sociology; nor does it vanish completely
into the reconceptualisation, as appears to happen
with Althusserian criticism. Rather, it comes to
exist in the tension of a part-whole relationship, it
is no longer an ‘individual’ to be related to other
‘individuals’ but it becomes the specific point of a
general mediating structure: the specific in which
thought seeks the totality of determinations (whilst
the totality is that in which thought seeks to discover the specific). 24
Secondly, there is a noticeable gap in the above
formulations: they lack a conceptualisation of the
perceptual, imaginative and logical steps by which
the ‘real’ object is appropriated by thought, by
which mediations are grasped and transformed,
and a structure of concepts produced. The gap is
precisely that between the real object and the object of knowledge, or between science and ideology
and ideology and the world, and what is required to
fill it is a phenomenology of concept formation.

That is the gap in Marxist theory to which I
referred earlier. I will return to consider it again
later in the discussion, but its absence haunts this
paper, a second vital factor restricting it to the
description rather than the conceptualisation of the
processes and results of Marxist criticism.

It is an absence that will be encountered again in
discussing the criticism of theoretical sociology,
although a rather different set of problems is involved here. Sociology’s theoretical insights are
attempts to grasp the mediations, the determining
structures behind ‘facts’. From the argument so
far, these can be expected to be incomplete, to
tend towards the positing of empty or unintelligible
relationships. Criticism in this respect is a matter
of displaying the partial nature of theoretical
categories, again not a rejection bui”a deepening
and ~t the same time a limiting of their meaning.

The partial nature of a theoretical category or
concept lies not in the empirical isolation of some
thing immediately given, but in its ‘internal’

incompleteness (and therefore its abstractness).

The ‘ideological’ function of such a concept lies in
its implicit or explicit claim to provide a full grasp
of its object. 25 Again, criticism and incorporation
into a Marxist framework are the same activity,
carried out by Lukacs himself on the theoretical
categories of Max Weber. Andrew Arato (1972a)
has described what happens to Weber’s categories
in Lukacs’ exposition,. and it is worth follOwing his
discussion through. As Arato points out, Weber
was concerned with demonstrating that historical
materialism was only one system of ‘ideal types’

among many from which he (Weber) had chosen
another equally useful system; this makes Lukacs’

achievement all the more significant.

The first category adopted and transformed by
Lukacs is that of ‘formal rationality’, for Weber
the ‘reason’ of science and industry, resting on
quantifiability and calculability and excluding all
‘values’. A rational economy requires the organisation of all areas of life – law, politics etc according to the principles of formal rationality.

According to Arato, Lukacs ‘fuses the Marxist
category of abstract labour with Weber’s category
of formal rationalityt (P35). enabling him to go
24 This r(‘!el”!-; back to Lukacs’ c(,ncept of totality discussed III nIl. The
nl:1l1oll.-;hip of lenuion between pari and whole (a relationship of d<'(lclldelll'l'/indl'jlCIlc<'IlCe) iJll(llil'R in turn that the totality is both 'slrul'lured'

i!lli! ‘expressive’; thel'(, is no absorption of evel’ylhinl~ into some ‘ess<'nce'.

25 There is of (“ourse an inlcJ”:lctioll between the conceptual criticism ,lIld the
situalill{: of a work in it:-; social context. In Ihe same way that the latter is
J(‘ces,~al’y for the full inldJil’ii>ility of a work, it ,lso provides the full
intelligibility of Its mlel’ifi£. fOflllS of one-sidedness,

beyond Marx in arguing that
, . .. free labour in itself is not enough to allow
the complete self-realisation of capitalist production, or even the total rationalisation of a
single factory. The culmination of capitalist
rationality is only possible when the “fate of
the worker becomes the fate of society as a
whole”, when the “internal organisation of the
factory” becomes the microcosm of “the whole
structure of capitalist society”. ‘ (P35) 26
If Weber’s category enables an extension, or at
least a clarification, of Marx, then what has
happened to the category itself? At one level it
remains the same: it ‘refers’ to an analytic logic
necessary to capitalist production and reproduc~d
in every area of social life; at the same time it is
modified, it becomes a concept ‘required’ in the
totalising framework orMarxism, and the conceptual structure of which it becomes a part limits its
meaning and fills it.out. Limits, in the sense that
it ceases to be the (eventually unintelligible) dominant feature of industrial society as ‘such, but becomes ~ feature of capitalist society, a ‘dependent’

rather than a central, organising concept – dependent upon the analysis of the development of capitalist relations of production, the analysis of which it
follows in Lukacs’ exposition. It is ‘filled out’ in
the sense that it takes its place in a structure of
concepts appropriate to capitalist society which
specify in more detail its significance; whereas
previously in Weber’s work it was specified
principally in relation to the principles of organisation of non-capitalist society – as for instance in
Weber’s analytic classification of types of authority.

Thus what I referred to earlier as its ‘unconscious relativism’ is revealed – it loses its
‘absolute’ implications and becomes the grasp of
an historically limited phenomenon.

There is a neat irony in the fact that Lukacs
appropriates another Weberian category – that of
‘objective possibility’ – precisely to point to the
transcendance of the type of industrial society in
which ‘formal rationality’ is dominant. Arato
shows how Lukacs transforms ‘objective possibility’ from a category referring to the past structured by the more or less arbitrary interests of
the individual historian to a category referring to
a present structured by the interests of a social

‘Weber examines the possibilities of the past
for the sake of the cognitive interests of the
present, while Lukacs seeks to interact with
the possibilities of the present for the sake of
future praxis. ‘ (P63)
Again, the category always remains the same, a
grasp of the possible outcomes which lay within a
particular historical situation, in turn enabling a
grasp of the ‘subjective’ contribution to the eventuai
outcome. It is here that what can best be called the
‘fluidity’ of concepts becomes apparent: it is the
case that the meaning of a concept depends upon its
relations to other concepts within the same conceptual system, yet it is also true that the same concept
is not unrecognisable when it is situated within
another system. Like the ‘given’, it is transformed
and remains the same at the same time. Again for
the conceptualisation of this process, we are
referred to the absent phenomenology of concept
26 I have reservations about Aratl”s formulations in this passage. It leaves
the impression of an ‘expre:;sivc’ ecollomic determinism: the economic
rosc (‘xpresslng itself idcutirally at dUfercnt levels of the SOCial formation. Thr; notion of ‘expression’ might offer an apprOXimation to the nature
of ehange oC forms of thought (with which Lukacs Is primarily concerned)

formation, but it is possible to make one more
point about the fluidity of concepts, which in turn
leads to a further. point about the process of Marxist criticism. This involves returning to the ‘transcendance of dualisms’: a theoretical understanding
of the dualisms of bourgeois thought is essential for
the coherent comprehension and rigorous criticism
of non-Marxist thought.

Any theory, whether articulated or ‘practiced’ in
empirical research, presupposes a conception of
the subject/object relationship; organisation by
means of the possible variations of this relationship represents the most general level at which it
is possible to distinguish between different theories
or ‘approaches’. As long as the variation remains
dualistic, ‘static’ or ‘dynamic’ contradictions
appear within the thought that presupposes it.

Lukacs traces the most general dynamic form of
contradiction set in motion by the most highly
articulated presence of the dualism in pre-Marxist
philosophy – ie the working out of the contradiction
in rationalist philosophy. The basis of the fluidity·
of concepts lies precisely in this: that one dualistic
variation of the s.ubject/object relationship, pushed
to its limit, will transform itself into its opposite
and this in turn will involve the transformation, not
the destruction, of the conceptual articulation of the
dualism. In the working out of the transformation of
Kantian philosophy, there is no point at which a
previous set of categories is abandoned for a new
and opposite one, yet the transformation is radical.

A dialectical conception should be able to reveal
and theorise the transformation as part of its own
conceptual structure, become self-conscious, of its
own necessary internal fluidity.

In its ‘static’ form the contradiction appears as a
number of implicit or explicit contradictions in the
same work. In Ideology and Utopia the contradiction
is glaring; in less directly epistemological writings
it has more manifold results and is less obvious.

In any case, however, it involves the coexistence
in the same work of two (or possibly more) opposing variations of the same dualism ‘behind’ the
explicit argument of the work. Here, an understand·
ing of the dualisms enables the identification of
contradictions within the work being studied, and
provides the unifying principle of those contradictions. Crudely, the consequent process of ‘incorporation’ is a matter of elaborating the contradictory
positions into a coherent whole, in the way that
Lukacs constructs a dialectical position out of the
poles of traditional rationalism.

Before going on to look briefly at the problem of
a phenomenology of concept formation, some
further concluding observations are possible. To
begin with, the transcendence of relativism can be
underlined. Relativism implicitly or expliCitly conceptualises different systems of thought as self-.

contained unities, between which there can be no
mediation; otherwise there would be a transcontextual unity which would enable judgments between
different systems. The above discussion has shown,
I hope, that such mediation is possible and that it
does not re-introduce either a notion of a ‘neutral’

or ‘objective’ reality independent of conceptualisation against which ‘theory’ can be tested, or reference to some eventually empty ~:rrnscontextual
notion of ‘comprehensiveness’.

Rather, the
but the material articulation of a form will dUfer at different levels of the
SOCial formation. This does not affect the main drilt of the argument: that
the development of abstract labour involves other changes before the
domination of the capitalist mode of production is achieved.

27 A full critiCism of Mannht’im would have to show how, on occasions, he
seems to be struggling tov.oartls a dIalectical conception without ever
making it.

mediation is built into the nature of conceptual
thought itself; dialectical reasoning is a matter of
pushing each system to the point at which a transformation takes place, yet maintaining both systems as essential moments of a wider totality. Of
course, this is not always as simple a task as it
might appear from the above formulations: the
disc;ussion has been primarily concerned with
highly articulated systems and many of the complex,
ities involved in dealing with less systematic
theories, and with theories of the natural sciences
(where the problems are likely to be very different)
have been lost.

Secondly, the criticisms of Althusserian Marxism
that have been running through the argument have
been based on its failure to achieve a self-conscious
transcendence of the dualisms of bourgeois thought,
positing instead a break which amounts to a false
suppression of the problem. This results in the
presentation of Marxist science as an eventually
unintelligible phenomenon governed by formal laws
– a return to a form of classical rati.onalism.

Glucksmann (1972) sees Kant’s rationalism in the
a priori identity of structures that Althusser posits
between the different levels of production, and the
latter’s discussion of the production of knowledge
modelled on economic production is perhaps the
core of ~he resulting reification of thought processes. 8 A more systematic critique than that
offered here would search out the contradictions and
gaps consequent upon the suppression of the dualisms and point towards their resolution.

Another result of the suppression is a misreading
of non-Marxist and much Marxist ,work, Lukacs
being an ideal example, and this has certain more
immediate implications when it is seen in connection
with the sort of criticism operated by Althusserian
Marxism. Despite the generality of the diSCUSSion,
it should be apparent that the process of Marxist
criticism involves the demonstration that it can
understand non-Marxist thought more completely
than that thought can understand itself .. It can grasp
the complex of relationships that relate that thought
to its social context, draw out its conceptual presuppositions and push them to the point at which
they become self-destructive and reveal hidden
contradictions and limitations that remain ‘unconscious’ from within the criticised system.

However, the earlier discussion of praxis revealed
that conceptual criticism of this type cannot be isolated and the continued ability of Marxism to
appropriate and unify the scattered and partial insights of non-Marxist thought depends in the last
analysis ‘on its ability to articulate itself as a
coherent, concrete social and political practice.

It is only through its practical relationship to a
developing totality of social relationships that
Marxist theory can recognise the partial truths
presented by non-Marxist thought.

NOW, the gap between epistemology and political
organisation is not inconsiderable: it would be a
nonsense to ascribe the absence of an effective
political practice to a rigid epistemology, and as
big a nonsense to ascribe a rigid epistemology to
the absence of an effective political practice. But
without implying, that one is a principle cause of the
other, it is possible to see how a mutual reinforcement can take place. A Marxism that presents itself
in terms of a radical choice between an unfounded
science and an ideology to be rejected will have an
inbuilt tendency to different organisational forms,
and a different political effectivity, to a Marxism
that attempts to transform an ideology by working
28 Paradoxically, 1t is allD the closest he comes to the type of phenomenology I am advocating here.


within it. And a politically.isolated and ineffective
Marxism or a Marxism’rigidified in fixed institutional forms will tend towards a different conception
or its own knowledge producing activities than a
Marxism engaged in a living interaction with the
class whose experience it is theorising.

Towa.-ds a Ma.-xist Phenomenology 01 Concept fOl’mation
Tl)e need for a Marxist phenomenology can be
indicated in a number of ways. 29 That for which I
am arguing here is based eventually on the theoretical description of perceptual processes. Its foundations have, I believe, already been laid, although in
a fragmented way, in the work of Husserl, Sartre
and Merleau-Ponty, 30 but a coherent structuring
and elaboration of the insights of these very different writers is still needed. Beyond pointing to this
task, it is only possible in the present context to
outline its scope. and significance.

To begin with, it would itself be an epistemology,
a guide to and delineation of the sCientificity of an
analysis qf the structures of society, the foundation
and ‘guarantor’ of Marxist science. It is the
specific tool by means of which a philosophy of”
history (or praxis or consciousness) founds, points
to the necessity of, and defines the means of producing, the precise and careful analysis of social

In the second place, it is an essential base for
the reflexivity of Marxist thought, and this is of
prime importance for political practice. The rooting of our concepts, at all levels, in the immediate
experience of the class is necessary to comprehend
and clarify the way in which agitational and propaganda work is to be carried out; such wo~k is precisely the attempt to root a Marxist analysis of
society in the class itself, and involv~s ‘rendering
that analysis comprehensible in terms of class experience. Yet its epistemological role also enables
a critique of class experience, to distinguish
between that experience which is an inessential
surface effect of the development of the mode of
production and that which indicates an important
change in that development and comprises a fundamental part of it. The material structures of society do not change in the same way as the experience
of those structures, and in this sense the analysis
of each must necessarily be rooted in that of the

Finally, and most importantly, a phenomenology
of concept formation is only part of a phenomenology
of consciousness as such, and thus a phenomenology
of class consciousness. Within orthodox phenomenology and in its already existing variants (and
especially in Sartre) there exist in embryonic form
the necessary tools to grasp consciousness in its
widest sense: to bring within the scope of Marxist
comprehension imaginative, emotional as well as
‘ideological’ and conceptual processes; and eventually to develop a coherent politics of interpersonal
relationships. The significance of this last possibility cannot be overstated: it involves the development of a ‘science of lived experience’, a science
different to that of social formations, with a different internal logic and with its own (self-founded)
criteria of sCientificity, but one complementary to
and articulated with structural analysis. The
importance of a phenomenology along these lines
must be self -evident; as must the difficulties of
developing it in a coherent and systematic way.

211 Piccone (op. elt.), for example, uses Lukacs to point to the need for
phenomenology of needs (which by itseU, it seems to me, is likely to
remain an idealist project). See also Paci (1972).

30 See in particular Husscrl (1970); Merleau-Ponty (1962, 1974a, b);
Sartre (1949,1957,1960, 1963a, b).


AltJlUsser, L. 1969 For Marx, London, Penguin; 1974 Elements
d’autocritig~, Paris, HacheUe
Arato, A. 1972a ‘Lukacs’ Theory of Reification’, Telos 11, pp25-66;
1972b Notes on ‘History and Class Consciousness’, Philosophical Forum
Vol. JII, pp386-400
Colletti, L. 1973 Marxism and Hegel London, New Left Books
Fecnberg, A. 1971 ‘Heificationand the Antimonies of Socialist Thought’


1972 l&nin. London, Ncw Left Books
Mamilieim, K. 19721cE’ol(lgy :l~ London, Routledge & Kegan Paul
Marx, K. 1968 ‘Theses on Feuerbach’ in Selected Works London, Lawrence
and Wishart; 1973 Grundrisse London, Pcnguin
Merleau-Ponly, M. 1962 The Phenomenology of Perception, London,
Routledge & Kegan Paul; 1974a Adyentures of thc Dialectic, London,
Heinemann; 1974b The PI”OSe of the World, London, Heinemann
Paci, E. 1972 The Function of the Sciences and the Meaning of Man,
Evanston, Northwestern Univcrsity Press
Phillips, D. 1974 ‘Epistemology and the sociology of knowledge’, Theory and

Glucksmann, A. 1972 ‘A Ventriloquist Structuralism’, New Left Review
72, pp68-92
Goldrnann, L. 1971 Reflections of ‘History and Class Consciousness’ in
Mes7.’Uos (ed. ) AspE’cts of Histo!’y and Class Consciousness London, Merlin
Hindess, B. 1972 ‘The “Phenomenological” Sociology of AUred Schutz’,
E£Q.!]pmy and Society Vol.l ppl-27; 1973a ‘Transcendentalism and History’,
E.£Quomy and SO(“i(‘ty Vol. 2 pp309-42; 1973b The Use of Official Statistics
in SOciology, London, Macmillan
Hirst, P. 1972, ‘Marx, Law and Crime’, Economy and SocjPty Vol.l pp28-56.

Husserl, E. 1970 TIle Crisis of European Science and Transcendental
Plwn(lmenolor;y, EV;J.nston, Northwestern University Press
Lukacs, G. 1971 IIistory and Class Consciousness London, Merlin;

Piccone, P. 1972 ‘Dialcctic and Materialism in Lukacs’, ~ 11, ppl05-33
Ranciere, J. 1972 ‘Althusser and Ideology’, Radical Philosophy 7
Revai, J. 1972 Review of Georg Lukacs’ ‘History and Class ConSCiousness’,
Theoretical Practice No. 1
Sartre, J”:P. 1949 The Psychology of the Imagination London, Rider;
1957 Being and NOUJinilllg§ji, London, Methuen; 1960 Critique de la Raison
dialedigue Paris, Gallimard; 1963a The Problem of Method, London,
Methuen; 1963b Sketch for a Theory of the Emotions, London, Methuen
Stedman-Jones, G. 1971 ‘The Marxism of ,the Early Lukacs, New Left
Review 70 pp27-64
Westergaard, J. 1970 ‘The Rediscovery oE the Cash Nexus, Socialist
Register 1970



Kate Soper
Agnes Heller, The Theory of Need in Marx,
translated from the German, Introduction by Ken
Coates and Stephen Bodington, Alison & Busby,
London, 1976, 135pp, hardback £5.25, pb £2.95
As Hemingway, I seem to remember, somewhere
said of Pernod, so it is with this book: it takes you
up as much as it brings you down. The analogy,
however, is perhaps too frivolous for a work whose
scholarly sobriety borders on dryness; moreover,
it suggests an ease of absorption that might mislead
readers who are unaccustomed to that strange brew
of half-developed concepts, potent good sense and
flights of fancy that can be concocted from Marx’s
works and labelled (somewhat euphemistically) ‘a
theory of needs’. For it does not seem to me that
Heller has managed to offer us anything much more
readily digestible than Marx himself on this subject,
even though her project is largely one of exegesis
and synthesis – a~d I speak as one who has spent
some time in the attempt to ascertain the meaning
and coherence of Marx’s various remarks on the
subject of needs. On the other hand, it may be true
that I have approached Heller’s book with too many
preconceptions and expectations about what a work
on the theory of needs should achieve, and that
others less steeped in this aspect of Marxism will
find a good deal to interest and inspire them in this
book, if only because it sketches out an area for
consideration that is scarcely ever discussed in any
detailed way, and because it is the product of a good
deal of reflection on that area. All the same, I suspect that many readers will wish that Heller had
provided more opportunity to share in this process
of reflection. As it is, she tends merely to chart
its results, and these are often presented in an
over-condensed and disjointed form.

In all fairness, it should also be said that she has
not been well served either by her translator or by
her editor in this English edition. There is a nervous recourse to literal rendering in the translation
which betrays a failure to have construed Heller’s
precise meaning (and in several instances I have
still not managed to decipher this). Even where the
meaning is clear, it is frequently couched in rather
bizarre expressions, and the reader is confronted
with an array of undefined concepts (eg ‘community
structure’, ‘society of associated producers’, the
‘antinomies’ of capitalism, its ‘formation’, and so
on). In the case of these and other terms, some

explanation for their choice either within the text
or in a glossary would have been welcome. So too
would have been more indications (if only in the
form of section headings and bridge passages) of
the overall direction and design of the wou. ,As it
is, we are offered the pieces of a jigsaw – which is
tantalizing because we are not sure if we have all
the pieces, and wearisome because so much of the
work of assembly is left to a reader who has little
idea of the final picture to be constructed.

There are two further general features of this
book which some may find disappointing. In the first
place, there is scarcely a reference to other work
bearing on the question of needs, by which I mean
either to work outside historical materialism in
anthropology or psychology or biology, all of which
are pertinent studies, or to attempts by other Marxists to confront the vexed question of needs. Admittedly in the latter case there are a few directly
relevant works, and it may be that Heller has not
had much opportunity to assess them 1 – here I have
in mind such writers as Seve and Timpanaro, and
the debate on Marx and Fretld. Yet she also never
mentions- either Sartre or Marcuse nor any of the
economic studies that bear on the issues she
raises (Mandel, Betteiheim, Rubin) and there is
scarcely a reference to any work by Lenin or
Trotsky or Stalin. In other words, there is no
attempt to place her contribution in the context of
developments in Marxist study either in the East or
the West, though her debt to Lukacs is obvious.

There is an advantage to this in the sense that her
book is refreshingly unparasitical; it also means
that it avoids any facile classification in terms of
allegiances within current Marxology (it does not,
for example, adopt either a straightforward humanist or anti-humanist stance and cannot be located
easily in terms of such disjunctures. ) Its disadvantage is that it is restricted to Marx’s work
alone, and thus to a large extent remains a piece
of academic Marxology – an exegesis of texts which
themselves are regarded as self-sufficient ends:

getting at Marx’s meaning, rather than assessing
its worth or relevance to contemporary events,
stin seems the dominating concern. Since it scarcely
ever ventures beyond Marx’s own dicta either for
its substance or its exe~plification, the book re-1 Though her book was ori(tinally published in German, Heller is herseU
Hungarian and associated with a group of Hungarian philosophers of
Lukacian inspiration ‘Who have rel’ently been subject to a certain amount
of persecution in Hun~ary.


Buy the newest RP in printDownload the PDF