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.

Reflections upon Roy Bhaskar’s ‘Critical Realism’

Reflections upon Roy
Bhaskar’s ‘Critical Realism’

Wal Suchting

‘ … quamquam ridentem dicere verum quid vetat?’


(‘ … but what is to stop anyone with a smile on his face
from telling the truth?’)

2 ‘Empiricism’ and Its Critique

(Horace, Satires, I, i, 25)

1 Introduction
1.1 Over the last fifteen years or so Roy Bhaskar has
published a considerable body of work. Though it has been
praised by some, and has even been influential here and there, it
has not yet been the subject of a comprehensive critical scrutiny,
at least in print. I This could not be undertaken in any great detail
within the fairly brief compass of a paper like the present; but I
shall attempt at least to sketch the bare outlines of such a critique.

1.2 In order to have a reasonably circumscribed presentation
to discuss I shall concentrate on just one text, Reclaiming Reality. A Critical Introduction to Contemporary Philosophy (Verso,
1989), which is Bhaskar’s latest work (at the time of writing) and
deals with all his main and characteristic themes and positions.

(All further page references, unless otherwise attributed, will be
to this book.)
1.3 It is general useful, and indeed important, to distinguish
between criticism within the terms of a general conceptual
framework or ‘problematic’, to use contemporary jargon (‘immanent’ or ‘internal’ criticism) and criticism of the framework
itself (‘ extrinsic’ criticism’). 2 The latter sort is most convincing
when it proceeds from a developed alternative framework. However, constraints of space forbid the presentation of the latter
here, so criticism of Bhaskar’s doctrines will be largely of the
former sort. 3
1.4 In earlier writings Bhaskar has called his general philosophy of science ‘transcendental realism’ and his special philosophy of the human sciences ‘critical naturalism’. He is now
inclined to telescope the two and call the whole position ‘critical
realism’ (vii, 190). Naturally, I have followed his preference
here in the title of this paper, and in one of two places elsewhere,
but will keep to the older terminology in referring to the constituent parts.

This is a condensation by at least two-thirds of another on the
same subject (unpublished). Hence many of the formulations
and arguments are unavoidably somewhat elliptical, though I
hope clear enough to be understood and evaluated. The text is
made up of numbered paragraphs, both for ease of reading and
to facilitate critical discussion.

Radical Philosophy 61, Summer 1992

2.1 Though in the presentations of transcendental realism
(TR henceforth) ‘positivism’ seems often to be used interchangeably with’ empiricism’, the latter appears to be meant as
the generally more inclusive term; anyway, I shall use it as such.

2.2 Empiricism is characterised within TR in two ways. In
terms of its ontology, ‘the world … consists essentially of
atomistic states of affairs which are constantly conjoined’. In
terms of its epistemology, such states of affairs ‘are known by
asocial, atomistic individuals who passively sense (or apprehend)’ them (8). Statements about constant conjunctions are,
when true, laws.

2.3 Bhaskar brings two sets of theoretical (as distinct from
political, ideological, etc.) charges against em~iricism, corresponding to the above twofold distinction.

2.31 The charges against the ontology of empiricism are in
turn of two main sorts.

2.311 Empiricism cannot provide a sufficient condition for a
statement’s being a law. This is shown by its incapacity to handle
successfully ‘the Humean problem of induction’, which concerns that ‘warrant’ or ‘guarantee’ we have for ‘supposing that
the course of nature will not change’ (38). This is, Bhaskar says,
citing C. D. Broad, ‘the scandal of philosophy’ (30), and ‘any
theory of science as rational depends upon a resolution’ of it
(39). Empiricism interprets this basically ontological problem as
equivalent to the problem of what ‘warrant’ we have for ‘supposing the regularities in our experience will continue’ (38) or for
‘supposing some general proposition, statement or theory is
true’, which is an epistemological problem (38, 39). In fact, this
is a special case of the ‘epistemic fallacy’ characteristic of
empiricism, which is the thesis that ‘ontological questions can
always be reparsed in epistemological form: that is, that statements about being can always be analysed in terms of statements
about our knowledge’ of being (13). But empiricism cannot
solve even this reformulated problem. So it cannot account for
the necessity characteristic of laws.

2.312 Empiricism cannot provide a necessary condition for
a statement’s being a law. This is shown by its incapacity to
handle successfully the question of the ‘transfactuality’ of laws,
or the problem of ‘transduction’ (181), which is that of the
applicability oflaws outside the domain of actual experience (for
example, the centre of the sun). For the constancies that lawstatements record occur only within systems that are ‘closed’ to
the influence of disturbing factors by experimental manipula-


tion, and yet we apply laws to problems pertaining to situations where experiment does
not or even cannot occur. So empiricism cannot account for the universality of laws.

2.32 The charges against the epistemology of empiricism are also of two sorts. As
we have seen, empiricism holds that knowledge is ultimately (a) of a direct, perceptual
sort, which (b) occurs in the context of socially atomistic individuals. Each has an
unacceptable consequence. (a’) It follows from (a) that empiricism cannot account for
the fact that the development of knowledge exhibits radical discontinuities, since
directly perceptual knowledge would be simply cumulative, (b’) It follows from (b) that
empiricism cannot account for the fact that knowledge (in particular scientific knowledge) is a result of social endeavour.

3 . Transcendental Realism and Empiricism
3.1 TR has both a genuine ontology (unlike the experiential pseudo-ontology of
empiricism) and an epistemology.

3.11 The ontology of TR consists of: (a) events, (b) experiences of events, and (c)
what are referred to differently in different places, but on p. 90 as ‘structures, generative
mechanisms or agents’ – what I shall call compendiously,. using a traditional term,
‘powers’. (c) beget, in appropriate ‘triggering’ circumstances (a) and (a) beget, when in
the appropriate relation to sentient beings, (b). Thus the ontology of TR differs crucially
from that of empiricism with regard to (c), understood as irreducible items in the
‘furniture of the world’. Laws record the ‘tendencies’ of (c) to produce (a) in certain
situations. As irreducible to (a) and hence independent of (b) and indeed of consciousness in general, (c) are what Bhaskar calls the ‘intransitive objects’ of knowledge
(called thus, presumably, because they endure through ‘transitory’ attempts to understand them).

3.12 According to the epistemology of TR, know ledge arises not by virtue of some
direct relation between know er and known but via certain conceptual means called
‘transitive objects’ , which are also socially evolved and applied. Experiment is the main
means for discovering the existence and nature of powers through its ability to create

3.2 With the apparatus outlined in 3.1, TR is, it is argued, able to avoid both the sets
of charges against empiricism outlined in 2.31.

3.21 With regard to the charges in 2.311 TR’s responses are the following.

3.211 The problem of induction in a serious ontological one, the solution to which
is that, since (according to TR) nature is ultimately a complex of powers, existing
independently of any form of awareness, and by their very nature invariant in their
operation (how could a power characterised as, say, ‘that which – in appropriate
circumstances – enables the scratching of glass’ ever do anything but enable the
scratching of glass, given those circumstances?) (40), we have a guarantee of the
uniformity of nature. Or, to put the matter in terms of laws, these record the tendencies
of powers to realise themselves in appropriate circumstances, and these are necessary
truths about powers. So, once a power!law always a power!law, as it were.

3.212 As regards the problem of transduction, since laws are about the tendencies
of powers, which are perfectly objective, and not, except very indirectly and derivatively, about the events they produce when manipulated (in particular, experimentally),
there is no problem about the application of laws in situations which exclude manipulation by human beings.

3.22 The responses to the epistemological problems are these.

3.221 Since knowledge is not a matter of direct confrontation between knower and
what is known (or to be known), but occurs via transitive objects, it is not only
explicable, but positively to be expected, that the development of knowledge, in
particular scientific knowledge, should exhibit discontinuities, as one set of transitive
objects is replaced by others in the course of endeavours to grasp the character of
intransitive objects.

3.222 The generation of knowledge is inherently a social affair, as many people
must cooperate in the creation and use of transitive objects, experimental set-ups, and
so on.

3.3 TR consists not only of an ontology and an epistemology, but also of a ‘metaphilosophy’. Various things are said about this, in different ways. I shall try to
systematise the main points by reference to the Aristotelian schema of the ‘four causes’ .

3.31 As to its ‘formal cause’ (given by answering the question: ‘What is it?’),
philosophy is ultimately a ‘philosophical ontology’, which consists in ‘some general
account of the nature of the world, to the effect that it is structured and differentiated’

(150). As such it is irreducible to a ‘scientific ontology’, which specifies ‘the structures

Radical Philosophy 61, Summer 1992

which, according to the science of the day, it [se the world- WC]
contains and the particular way in which they are differentiated’

(150). TR as a ‘metaphysical realism’ consists of ‘an elaboration
of what the world must be prior to any scientific investigation of
it’ (12). Or, perhaps better, it is an account of what the world
must be like if any scientific investigation of it is to be possible:

philosophy is a determination of ‘the necessary conditions of
conceptualised activities’ (14).

3.32 This already foreshadows philosophy’s ‘material
cause’ (given by answering the question: ‘From what is it
made?’), since, if its content is as just indicated in 3.31, then it
must take ‘as its premises scientific activities as conceptualised
in experience (or in a theoretical redescription of it)’ (14), it must
be ‘the analyst of intelligible activities’ (22).

3.33 What is said in 3.331 also foreshadows philosophy’s
‘efficient cause’ (given by answering the question: ‘How is it
made?’). If philosophy is a body of propositions not identical
with any scientific one, it must have a distinctive method (14),
which, since that of science is a posteriori, must be a priori (14),
consisting in the establishment of the conditions for the possibility of ‘conceptualised’, ‘intelligible’ activities; in other words,
philosophy’s method is ‘transcendental in Kant’s sense’ (14).

3.34 Philosophy’s ‘final cause’ (given by answering the
question: ‘For the sake of what is it made?’) is, in effect, the
subject of a number of different formulations. There is space for
citing at most two passages. One occurs in the first paragraph of
the last and latest chapter in the book, and can also be found
almost word for word in the first paragraph of the preface. The
author says that he is concerned with the form of the realism
required ‘to aid and empower the sciences, and especially the
human sciences, insofar as these illuminate and inform projects
of human emancipation’ (180). This may be supplemented by
some words from the beginning of the first essay: ‘ … philosophy
… is the discipline that has traditionally underwritten both what
constitutes science or knowledge and which political practices
are deemed legitimate’ (1). Putting these formulations together it
may be said compendiously that the point of philosophy is to
increase the power of the sciences, particularly the human sciences, as they bear on the political programme of liberation, by
‘underwrit[ing]’ the former and ‘legitimat[ing]’ the latter. The
two terms thus picked out may be taken to come to the same
thing, for to underwrite is to provide a guarantee for something,
a warrant, to make or show that something is legitimate, that is,
in accordance with or has the character of a law (legis), or a right
(juris), to show that something is not only de facto but also de
jure. So the ultimate aim of philosophy is to guarantee or justify,
and to do this for, in the first instance, knowledge. To answer the
question as to how this is to be done we may call on other
passages. For example, any theory of knowledge ‘must logically
presuppose a theory of what the world is like for knowledge … to
be possible’ (13), and such a theory is a (philosophical) ontology
(49). So the point of philosophy, so far as knowledge is concerned anyway, is to guarantee or justify the latter by reference
to t,he general nature of the world.

At the same time, Bhaskar stresses that realism, qua philosophy, ‘is not, nor does it license, either a set of substantive
analyses or a set of practical policies. Rather, it provides a set of
perspectives on society and nature and on how to understand
them. It is not a substitute for, but rather helps to guide, empirically controlled investigations into the structures generating …

phenomena’ (3). To cite other formulations, philosophy is, in the
words of Locke to which Bhaskar subscribes, ’employed as an
under-labourer in clearing the ground a little, and removing
some of the rubbish that lies in the way to knowledge’; it is ‘an
analyst and potential critic of conceptual systems and the forms
of social life in which they are embedded’ (2), so that ‘it can
Radical Philosophy 61, Summer 1992

sustain a critical orientation … to the existing practice of a
science … can criticise [its] lack of scientificity’ (18) and even
act as ‘occasional midwife to the sciences’ (24).

4 On the Ontology of Transcendental Realism
4.1 We have seen that the key idea in the ontology of TR is
that of what I have called ‘power’. This being so, it is not a little
surprising that next to nothing is to be found in the book being
examined (nor anywhere else in Bhaskar’s writings for that
matter) by way of elucidation of the notion. Certainly some such
idea is pervasive in quotidian, and even in informal parts of
scientific language (for example, ‘lethal’, ‘perishable’, ‘magnetic’, ‘produce’, ‘suffer’). But it is present in a quite ‘spontaneous’ form, and its being there does not, by itself, guarantee that it
has an irreducible meaning or even that it refers at all (compare,
for example, ‘average man’ or intentional language applied to
inanimate or abstract objects). So a clear understanding of it
cannot simply be assumed.

4.2 What is said is not of much help. As we have seen,
powers are said to be ‘structures, generative mechanisms or
agents’. The second and third characterisations are just intuitive
synonyms, whilst the first poses a puzzle of its own. For, strictly,
‘structure’ signifies a set of relations between a set of elements,
and powers are presumably properties of elements, which are
not themselves relations. The traditional view just assumed is
indicated by talk of ‘real essences’ captured in ‘real definitions’

(e.g. 85, 190). But, whilst powers as thus traditionally conceived
have been taken to be related to their effects by ‘logical necessity’, Bhaskar says that it is a question of ‘natural necessity’ (12,
17, 52, 154, 173). But as to what this is supposed to be – ‘the rest
is silence’.4 So it seems that all that is said of the basic concept in
the ontology of TR is that a power is a power is a power.

4.3 It is not difficult to think of ‘dialectical’ arguments (in
the ‘Socratic’ sense) against the idea of power. .For instance,
there is a traditional distinction between ‘active’ and ‘passive’

powers (a match has a power to bum, but it can actually bum
only something with a power to be burntV Something which is
said to have a power must then surely have the power to have that
power, and so on ad infinitum. This is not a vicious regress. But
it is, to say the least, somewhat curious that the existence of an
infinite sequence of properties should be a presupposition of the
existence of one property, so that the truth of a proposition about
the former follows from that of one about the latter (a sort of
Ontological Argument for the Existence of Powers). Of course,
the sequence might be cut short at some point by claiming that
there must be an ultimate power. But this is an at least equally
curious conclusion, and not only smacks of arbitrariness, but
suggests the question as to why powers should be assumed at all.

For if at some point we must say that something just is the case
why not do so to start with and simply say that there is a regular
concomitance between events? (The analogy here is with the
Argument from/to Design for the existence of God.) Note that
this is not meant as a defence of a ‘regularity’ theory or ‘dispositions’, laws or whatever (like Bhaskar, but largely for different
reasons, I regard such an account as untenable6). Yet these sorts
of arguments tend, as Hume said of Berkeley’ s, to ‘admit of no
answer and produce no conviction’, to be, in Kant’ s words this
time, ‘mock combats [Spielgefechte]’.7 Instead, let us turn to a
consideration of how the notion of power is used within TR,
however that notion is understood; for the problem of determining what Bhaskar actually means by it, if anything, is probably
best given up as a bad job (in advance of future elucidations). In
this new line of questioning we shall have to refer also to the
epistemological doctrines of TR already presented.



4.4 According to TR, introducing the notion of power
pennits the solution of the problem of induction, understood
properly as an ontological problem, as distinct from the empiricist understanding of it as epistemological, concerning the future
course of experiences or the truth-value of general statements.

Prescinding from the point that these two fonnulations are in fact
quite distinct, and concentrating on the second, it is clear enough
that TR is in no better boat than empiricism. For even if the
supposition of the existence of powers were to give a ‘warrant’,
‘guarantee’, or whatnot, to beliefs about what is not yet observed
(it would not, but that is another story), since, for TR, knowledge
about the world is always via transitive objects, TR no less than
empiricism is faced with the problem of assessing evidence for
the truth-value of statements (of various degrees of generality).

So ‘the problem of induction’ , once posed, may be expelled from
the empiricist epistemological field by the ontological pitchfork
of TR, but it keeps coming back in its own epistemological
backyard. Of the use of ‘powers’ here Wittgenstein might have
asked: ‘isn’t the engine idling?’. We are simply offered what
Hegel used to call ‘assurances [Versicherungen],. 8
4.5 If the preceding argumentative point has been taken,
then it is not necessary to labour the point that TR can no more
solve the ‘problem of transduction’ than that of induction, for the
ascription of powers to non-‘ closed’ systems must be made on
the basis of evidence derived from work on ‘closed’ systems,
which is also the evidential basis for alternative interpretations
of scientific knowledge of these, including empiricist ones. So
‘transduction’ does not introduce any basically new factor into
the debate.

5 On the Epistemology of Transcendental
5.1 Readers of Bhaskar’ s book who also know something of
Althusser may well be reminded, on first meeting with the
fonner’s distinction between ‘intransitive objects’ and ‘transitive objects’ of the latter’s distinction between ‘real objects’ and
‘theoretical objects’/’objects of theory’. Indeed, such a reader
might be forgiven for thinking that the two tenninologies mark
essentially the same conceptual distinction. But Bhaskar forthrightly rejects such an identification (e.g. 188) and criticises
Althusser at a number of places. Since I do not have the space to
demonstrate this point, I simply register my opinion that the
discussions are utterly confused. I shall restrict myself to taking
up what is probably his main point – namely that Althusser’s
epistemology is basically idealist – not because Bhaskar’s argument on this point is of any special value, but because it is not
much worse than an immense number of similar ones, and so
may contribute to a more general discussion. It will also be an
invaluable background for a ‘depth’ analysis of TR and its

5.2 Bhaskar does not cite Althusser in this regard, but
various passages from the latter are part of the stock-in-trade of
Althusser-criticism – they are guaranteed to disarm philosophers
– at least the ‘realists’: ‘The production … of knowledge, and
hence that of its object … takes place entirely in knowledge … in
thought.’ ‘Theoretical practice is … its own criterion .. , the
sciences … once truly constituted and developed … have no need
for verification from external practices.’ ‘ … this radical inwardness of the criterion of practice for scientific practice.’9 Since
commentators on Althusser seem generally to have had insuperable difficulties in understanding such passages, I shall first
discuss them in tenns of a simple, everyday example, even if this
does not occur anywhere in Althusser’s own writings.

Consider proceedings in a court of law. Let us assume, for

purposes of discussion, that the question before the court is the
detennination of the guilt or otherwise of someone charged with
murder. In tenns appropriate to the present context we may
describe the situation as one of the production of a certain result,
a verdict, from certain ‘raw materials’, like the testimony of
witnesses, material exhibits, forensic evidence, and so on, using
certain ‘means of production’ such as cross-examination and reenactments of events. Now this process takes place entirely ‘in
knowledge’, ‘in thought’, it is characterised by ‘radical inwardness’, in the sense that it occurs entirely within the context of a
set of intra-legal nonns and procedures. This is the case at least
from the time of laying of the charge, which is intelligible only
within a set of legal concepts; indeed, the laying of the charge
may well be the result of a previous series of proceedings. The
‘raw materials’ are in general at least in large part before the
court in the fonn of representations of ordinary life. The ‘working up’ of such materials proceeds according to the legally
defined rules of court-functioning (rules of evidence, principles
prescribing the conduct of judges, of deliberations by the jury,
and so on). A verdict is properly so called if and only if the
relevant procedures have been correctly followed in reaching it.

The producers of production of the verdict neither requires nor
allows any intrusion of ‘external practices’, such as newspaper
stories, results of extra-judicial inquiries, and the like – though of
course, material of such provenance can be introduced once it
has been appropriated by legal procedures. Even the results of
purely scientific procedure have to confonn to legal nonns to
count as ‘forensic evidence’. But there is obviously nothing here
to give aid and comfort to idealism. The raw materials have their
origins in ordinary perception, in familiar material processes, a
verdict may be overturned on appeal or retrial involving ‘inputs’

from ‘external practices’, the legal nonns and procedures themselves may be the subject-matter of socio-historical explanation
and be changed as a result of broader social changes.

All this is pretty obviously applicable, appropriately modified, to experiment, which Althusser himself ca11s ‘the criterion
of the theory of the “experimental” sciences .. , the fonn of their
theoretical practice’ .10 Everything that enters into an experimental situation qua experimental does so only within the tenns of a
conceptual-procedural framework. One of the raw materials
enters, say, not as a heap of powder of a certain colour, texture,
etc., but as a sample of such and such a chemical compound of a
certain degree of purity as established by certain standardising
procedures, and so on. The instrumentation is constructed in
tenns of certain theoretical principles and has to be used in
certain ways, and no others. The readings of the instruments will
generally have to be corrected, using various empirical principles, statistical techniques and the like. In this sense, the experiment takes place entirely in knowledge, in thought, the theoretical practice of a science that employs experiment has no need of
verification from practices external to the one it has itself constituted in its experiments, the criterion of practice is, in this sense,
‘radically inward’. It should go without saying that none of this
is inconsistent with (indeed it depends upon) the science’s being
ultimately about something other than itself considered as a body
of statements, rules of inference, and such like; that is, about real
objects which supply the causal input and ultimately detennines
the outcome of the experiment. But it is crucial not to confuse the
domain of knowledge and the domain of things: what the latter is
like is what ultimately counts, but it can count only through, and
by means of the fonner, as nothing here ‘speaks for itself’. To
say that the real object is only cognitively accessible through
some set of representations that are related to that real object in
and through interactions in a practice is obviously not to say that
the real objects are existentially dependent on the cognitive order
or that the latter is some sort of veil behind which the natural
Radical Philosophy 61, Summer 1992





order is hidden. That we cannot eat an apple without biting
pieces off it, chewing it, swallowing it, without our digestive
tract working on it, and so on, hardly means that we never really
eat apples, but only experience internal states of ourselves, or

5.3 Now the main theoretical thrust of Althusser’s position
here was against the idea that the knowledge-situation is one in
which a relation is set up between two tenns, the know er (‘subject’) and what is to be known (‘object’) such that the fonner
reflects (mirrors, represents) something about the object. ll One
of his basic objections to this is its ultimate idealism: a representation can only be of something of a similar nature to itself, and
so the object of the representation must be conceived of as
having the character of knowledge itself, lodged, as it were, in
the object in general, like Nothung in the tree-stump, to be freed
by the knowing Siegfried. There are, of course any number of
variants of this fundamental picture. In fonns of ‘direct realism’

there is an unmediated relation between subject and object. In
some of these positions traditionally known as ’empiricism’, the
relation is mediated by sense-perceptions which are cognitively
transparent, as it were. From this, it is but a very short stop to the
picture of perceptions that are ‘theory-laden’, hence turbid with
assumptions of human provenance. And the piquant figure of
this situation is that proponents of the latter generally think that
they are rebelling against old-fashioned empiricism/positivism,
when their range of positions is really just another variant of the
same basic framework. So when Bhaskar presents the idea of
‘transitive objects’ as an alternative to empiricism, this cannot be
taken at face value, and indeed nothing he says suggests that it is
anything but a new brand of what used to be (and no doubt still is,
in some backwaters) a trendy version of the old epistemological
problematic. Apart from the idealism, the incoherence of the
latter was already brilliantly revealed in the first few paragraphs
of the introduction to Hegel’s Phenomenology 0/ Spirit, where,
in brief, it is pointed out that the accuracy of the representation
could only be checked by gaining access to the object independent of representations, that is, by knowing about it without
knowing. Althusser’s conception rejects the framework completely, for knowing is always a matter of comparing the results
of appropriating the object in various cognitive-instrumental
ways. So the relation-to-the-object is already included in the
result of inquiry and there is no further question of that but only
of the comparative adequacies to similar problems of different
results of appropriation. As Goethe once said in a related regard,
there is an almost irresistible urge to ask about what lies ‘behind’

the result; 12 but this is simply a result of being in the grip of a
certain picture of knowing.

But we are already involved in ‘metaphilosophical’ questions, and to these we now turn explicitly.

6 On the Metaphilosophy of Transcendental
6.1 We have already seen (para. 3.33) that Bhaskar says that
philosophy’s (that is, his) method is ‘transcendental in Kant’s
sense’. Now this is, at best, crucially ambiguous. It could refer to
(a) the most general, fonnal structure of Kant’s transcendental
method; that is, roughly, an argument from an assumed ‘A’, to
the necessary and sufficient conditions ‘B’, postulated for’ A’ to
be the case. Or it could refer to (b) the specific use Kant made of
it, in the first place in the paradigmatic first Critique; namely, to
detennine the necessary and sufficient conditions for the possibility of there being a knowable world at all, that there should be
‘experience’. Now Bhaskar never attempts anything like (b), so
we must assume that his reference to Kant in effect concerns (a).

Radical Philosophy 61, Summer 1992

What then are we to take as ‘A’? Let us say science as such. But
the history of ideas records many different conceptions of what
constitutes ‘science’.13 But can one of these be non-arbitrarily
selected a priori as specially privileged? Of course, it is possible
to attempt simply to detennine the most general principles which
have governed various activities historically called ‘scientific’. 14
But such results would not have any metaphysical-nonnative
force of the sort Bhaskar desires. Conclusions: a ‘transcendental
deduction’ of this sort is either dogmatic/circular, or merely
historical in import. Suppose then we take ‘A’ to be, say, experimental science. The choice of this as the subject-matter of
deduction will still require independent justification of a non’transcendental’ sort, but perhaps some metaphysical ‘foundation’ can be ‘deduced’ for the experimental aspect. To make
things simpler and more favourable to the idea of a ‘transcendental deduction’ , let us suppose, for the sake of argument, that it is
possible to infer sufficient ‘conditions for the possibility’ of
experiment. But what about necessary ones? Suppose someone
who holds that a sufficient condition (at least) for the possibility
of experimental science is the existence of natural powers (detail
is not necessary here) is confronted by an Occasionalist who
holds that the possibility of experiment really depends on the
existence of a bon Dieu who always matches experimental
procedure and experimental result, so that the first always appears to be, but really is not the causal effect of the other.

Bhaskar says in a passage that is relevant here: ‘I do not claim
that my analyses are certain or unique (though they are the only
plausible analyses I know of)’ (15). The part of the sentence
before the parenthesis is no other than astonishing when read in
the light of the frequently repeated claim that ‘transcendental
deduction’ is an a priori procedure: how could the latter not be
both certain and unique? But let that pass. The part of the
sentence within the parenthesis suggests the question: ‘plausible
to whom?’ , since presumably something is not just plausible per
se. In this case, the claim of plausibility is presumably not meant
just as an autobiographical remark, but as an imp1icit appeal to
an assumed consensus of all and, more precisely, a consensus of
the scientifically inclined. But this is just to presuppose the
adequacy of the transcendental, ‘warrant’ for which is supposedly being demonstrated. In fact one is reminded of what
Descartes wrote in the letter of dedication of his Meditations
(‘To those most learned and distinguished men, the Dean and
Doctors of the sacred Faculty of Theology at Paris’) where he
says that, as Christians, ‘we must believe in the existence of God
because it is a doctrine of Holy Scripture, and conversely, that
we must believe Holy Scripture because it comes from God’,
though, he adds, ‘this argument cannot be put to unbelievers
because they would judge it to be circular’.

6.2 The conclusion to be drawn from the above is that
Bhaskar’s idea of a ‘foundation’ for knowledge (inter alia)
which warrants, justifies, legitimates, guarantees the latter, in
particular a ‘philosophical ontology’, is just one more avatar of a
traditional aim of philosophy (indeed it is partly constitutive of
the latter as a distinctive discipline) – whether the foundation be
thought of as the Will of God, sense-data, transcendental apperception, Wesenschau, or whatever. Like all the others, it is hit off
by the devastating image that Wittgenstein used in discussing
one particular area: ‘The mathematical problems of what is
called foundations are no more the foundation of mathematics
for us than the painted rock is the support of a painted tower.’ 15
Looked at most charitably, the idea is probably the result of a
confusion between the basic concepts and principles with which
or according to which a practice proceeds, on the one hand, and,
on the other, concepts and principles/rom which (as it were) it
proceeds. It is as if one thought of a dictionary of a natural
language as the ‘foundation’ of its vocabulary, rather than as the

codification of usage at a certain time from a certain point of
view; or its syntax as having a similar relation to actual patterns
of formation of phrases and sentences from words.

6.3 It may be added that the traditional picture of knowing
alluded to above (para. 5.3) leads irresistibly to the project of
specifying ‘foundations’. For if particular items of knowledge
involve an alignment between the thing and the idea of the thing
in the particular case, there must surely be general truths, logically prior to, and independent of any particular item of knowledge about the conditions under which that alignment obtains;
and these are precisely the ‘foundations’ provided by one philosophy or another. But the whole idea is a mere chimera conjured up by a wrong way of looking at knowledge.

6.4 I have suggested above that Bhaskar’ s doctrines as so far
examined share with empiricism the traditional ‘problematic’ of
epistemology. Their continuity with traditional philosophy has
further emerged in the discussion of his foundationalism. Now
the surest index of a problematic is the sort of questions it
licenses, either by explicitly posing them, or by affirming what
can only be interpreted as answers to certain implicitly presupposed questions. That TR and empiricism, though allegedly
opponents, really belong in principle to the same problematic is
revealed by the fact that both take certain questions to be genuine, and indeed serious ones. Here the most obvious cases are the
questions of induction and ‘transduction’. That empiricism cannot solve them is not taken by TR as a sign of something’s being
amiss with the problematic that generated them, but as a challenge to offer adequate answers. That it cannot do so, as we have
seen, is only further evidence, if such is needed, that the common
framework is flawed in principle, and that the ‘problems’ it
generates offer challenges not for their solution but for their


7 Critical Naturalism: Preliminary Remarks
‘Critical naturalism’ (‘CN’ henceforth) is a philosophy of
social theory of science which presents it as generated by the
application of the fundamental ideas of TR to the domain of
society’ or, put otherwise, it is the social-philosophical aspect of
‘critical realism’. It is set out in the book being examined in the
form of a set of substantive positions – by the via affirmativa as
it were, but also in part by the via negativa of the way in which
TR, as exemplified in the natural sciences, has to be qualified in
the social domain. These qualifications concern ontological,
epistemological, ‘relational’ and ‘critical’ features. I shall say
something about the first three only, since the fourth is merely
asserted, leaving no real room for argument. Still, I shall have to
be even more expositively and critically abstemious here than
before, though the consequences of this are somewhat mitigated
by the fact that there is at least a small amount of critical
literature on this area. 16

8 On the Ontology of Critical Naturalism
8.1 The crucial idea in the ontology of TR is that of what I
have called ‘power’. As so far understood, it has at least two
‘ontological’ characteristics: (a) it is a property of a situation,
which property is marked out by its tendency to produce certain
specific sorts of effects, and (b) it exists independently of any
forms of human awareness.

8.2 The ‘analogue’ (78) of natural powers in the field of
society is ‘social structures’ (78). What is distinctive about CN is

that it is a ‘relational conception’ (7). This distinguishes social
powers from natural powers as characterised in para 8.1(a)
above. (For further on this point see, e.g., 3, 4, 93). What do the
relations hold between? This is not crystal clear, but it would
seem to be between persons, for it is said that material objects’

being social in character (as well as simply material) ‘consists
only in the relationships between persons or between such relationships and nature that such objects causally presuppose or
entail’ (81 – emphasis added). Though the formulation is hardly
pellucid, it appears to mean that any relationship between interpersonal relationships and nature stems from the causal character of objects qua objects, so that the interpersonal relationships
are the fundamental ones as regards the constitution of specifically social subject-matter. Let us register these results as: (CNla) Social powers are relations, and (CN-Ib) the relations are
primarily interpersonal. The second point distinguishes social
powers from natural powers as characterised in para 8(b) above.

This last position is emphasised in two further formulations
which may be listed thus. (CN-2a) ‘ … social structures, unlike
natural structures, do not exist independently of the activities
they govern’ (79). (CN-2b) ‘ … social structures, unlike natural
structures, do not exist independently of the agents’ conceptions
of what they are doing in their activity’ (79). Further, we may list
as (CN-3): ‘social structures, unlike natural structures, may be
only relatively enduring (so that the tendencies they ground may
not be universal in the sense of space-time invariant)’ (79).

The following are also distinguishable theses, though some
of them clearly are, and some might be made out to be, corollaries of the preceding, or of one of the others. (CN-4): CN is an
anti-individualist social theory (70-73). (CN-5) CN is an anticollectivist social theory (73). (Collective phenomena are reducible to ‘expressions of enduring relationships’.) Though (CN-2),
human activities are always subject to the structuring and hence
constraining effects of certain social structures/relations. Therefore (CN-6): CN is opposed to all forms of voluntarism (176f.).

Nevertheless, since (CN-2_, (CN-7): CN is opposed to all forms
ofreification/determinism (93). (CN-4 may be paired with CN-6
and CN-5 with CN-7). The result of all this is termed ‘the
transformational model of social activity’ (e.g. 77), that is, social
change occurs by virtue of the agency of human beings’ acting to
restructure restructures/relations in which they find themselves
at some time.

8.3 I shall now make some brief critical comments on these
positions, starting with (CN-I). As regards (CN-Ia), it is completely opaque how a power could be said to consist in relations
(as distinct from, say, revealed in/identified by them). For example, the power of gravity (‘to speak with the vulgar’, as Berkeley
says, even if we ‘think with the learned’) varies with change in a
relation (distance), but the varying effects are due to the power
and not to the spatial relations. Thus (CN-Ia) must be supplemented with something else, here (CN-Ib). But whence do these
person-terms derive their powers? Are they inherent? In this case
we would be in individualist country, contrary to (CN-4). Are
they relational? In this case we would be back to where we
started. Do they derive from some supra-individual totality? But
this would be contrary to (CN-5).

8.4 (CN-2a) is unclear. If it means that social structures do
not exist unless they are governing relevant activities in fact,
then it is plainly false: for example, much of the life of almost
any society just ‘free-wheels’ along the road of habit, custom,
and so on (what Marx called ‘second nature’). Of course, it is
true that, say, an institution is generally identified in terms of at
least its potential functions of governing certain activities. But
this is trivially true, and, moreover, even natural powers are
generally identified in terms of what they effect (‘solvent’) or
how they are affected (‘soluble’).

Radical Philosophy 61, Summer 1992



8.5 If (CN-2a) means that the agent of an action must have a veridical conception
of that action in order to be an agent, then it is obviously false. If it means only that for
someone to be said to ‘act’ (‘behave’) rather than just be in some state of motion
(stepping on to the road rather than being jostled on it) the person must have some
conception of what s/he is doing (intention, goal, motive, or whatever) then it is true,
though trivially so.

8.6 (CN-3) is false on two counts anchored severally in the two parts of the sentence
quoted. Firstly, natural structures are not necessarily non-relatively enduring. Consider
many geological structures. Even ‘ultimate’ elements of the natural world (sub-atomic
items) often last only a very short time and are often the quite temporary products of
human manipUlation. Secondly, what is said within the parentheses confuses: (i) the
question of the universality of the relation asserted by a law; and (ii) the question of the
universality of the spatio-temporal distribution of instances in the antecedent of the
law. The two are logically independent (and indeed a ‘law’ may be true without having
any instances at all).

8.7 Finally, with regard to the positions bearing on social change, namely, (CN-6)
and (CN-7) – and the associated (CN-4) and (CN-5) – CN in fact leaves social change
a mystery. For the alleged ultimate items in the ontology of CN, namely, relations,
presumably cannot change just by themselves: change musts occur by virtue of the
actions of their tenns, namely, people. But we are told that people can act only within
the constraints of the relations within which they find themselves, which suggests
detenninism, though we are also told that relations can be changed by people, which
suggests that they are not, after all, entirely constrained by these relations, which
suggests voluntarism, which is, however, ruled out. In fact we are back with the preMarxist conundrum presented by Marx, and solved in principle in the third of the
Theses on Feuerbach, namely, how people who are detennined by their ‘circumstances’ can themselves change the latter. 17

9 On the ‘Relational’ Specificity of Social Subject-Matter in
Critical Naturalism
9.1 There seem to be at least two different positions involved here, and each offers
difficulties of interpretation, an analysis of which will have to be omitted here in favour
of my assuming in each case the one that I take is meant. These positions may be put in
the author’s words as follows. [1] ‘ … social theory and social reality are causally
interdependent. This is … to say that social theory is practically conditioned by, and
potentially has practical consequences in society.’ (5); ‘ … the objects of social scientific
knowledge, although … independently real … are causally interdependent with the
knowledge of which they are the objects.’ [2] ‘ … the social sciences are part of their own
field of inquiry … so that they are internal with respect to their subject matter in a way
in which natural science is not’ (84).

9.2 [1] is true. But what has it to do with the ‘limits of naturalism’ in social theory?

On the one hand, any theory is conditioned by reality in the sense of being dependent
upon it for its subject-matter and ultimate source of evaluation. Also, any theory, those
of the natural sciences included, works in and through ideologies. So there are no
significant differences between social science and natural science here. On the other
hand, any theory is any science can, and many have, had consequences in society. So
what is all the fuss about?

9.3 [2] is true in the sense that, for example, a theory in social science may attempt
to explain the origin and development of some, or even all theories in social science in
a way in which a theory in physics, say, would not attempt to explain the origin and
development of theories in physics. But it would be somewhat eccentric to describe this
situation as a ‘limit’ of any sort on social science.

10 On the Epistemology of Critical Naturalism
10.1 Para. 8.1 listed two ‘ontological’ characteristics of ‘natural’ powers which
Bhaskar alleges are not characteristics of social powers (paras. 8, 9 above). Now there
is a third characteristic of social powers, of an epistemological type: (c) such powers
can be isolated, identified and studied in closed systems by means of experiment. But,
according to Bhaskar: ‘because social systems are intrinsically open and cannot be
artificially closed, our criteria for the empirical testing of social theories cannot be
predictive and so must be exclusively explanatory’ (5). This invites a large number of
objections of which I shall list just the following on the grounds of their being both
Radical Philosophy 61, Summer 1992


decisive and susceptible of being put briefly.

10.2 In fact predictions about social systems are made, and
indeed regularly. See, for example, actuarial tables for insurance
purposes or stock market forecasts. So being a statement about
social subject-matter does not entail that it cannot be the basis of
a rational prediction.

10.3 Nor does the converse entailment hold, for many
clearly natural scientific theories have been or still are non- or
minimally predictive. See the histories of, at the large end of the
scale, cosmogony/cosmology, at the smaller end, geology, and at
the smaller still, the theory of evolution which, at least as first
presented by Darwin, depended for its immense scientific persuasiveness almost entirely on the wonderful coherence it lent to
otherwise isolated facts, and the simplicity of the general principle underlying such explanations. So, even were it true that
social theory is essentially non-predictive, this would not mark it
off from natural science.

10.4 No argument whatever is given for the claim that social
systems are ‘intrinsically’ open, and so ‘cannot’ be closed.

Furthermore, the history of the sciences shows that primarily
explanation-oriented sciences often become predictive through
the development of new techniques (computer modelling procedures, carbon-dating, and so on), despite the pronouncements of
philosophy ex cathedra about what is ‘intrinsically’ this or that
and hence ‘cannot’ be done.

10.5 A final consideration is in a sense the most crucial, for
it goes to the very heart of Bhaskar’ s whole programme. What is
to be said about a doctrine that on the one hand claims to be
devoted to ‘projects of human emancipation’ (para. 3.34 above)
and on the other denies that social theory can be predictive, that
is, aimed at the future, rather than explanatory, that is, aimed at
has been or is the case? Is there anyone who needs to have
spelled out the premise or two that would permit the deduction of
a formal contradiction here?

11 Concluding Remarks on the ‘Limits of

11.1 As we have seen, CN is the result of an attempt to
determine, by a priori philosophical means, proceeding from the
general principles of TR, ‘the’ ‘foundations’ of social science.

The procedure largely consists in trying to determine the ‘limits’

of the applicability of ‘the’ methodology of natural science to the
domain of social phenomena. What is presented is to a great
extent obscure. Where it is not obscure, it is at least very deeply
puzzling (e.g. ‘powers’ as relations), trivially true, or substantive
but false. What are regarded as ‘limits’ of ‘naturalism’ are
generally just uninteresting statements about differences between natural and social subject-matter (e.g. that the existence of
the former does not depend on human beings whilst the existence
of the latter does). Where there is what would genuinely be a
‘limit’ had a case been made out (the matter of the place of
explanation and prediction) there is no argument at all; and if the
limit had been established then it would have contradicted the
whole point of the enterprise of CN.

11.2 Of more general significance than the particular flaws
is the question of the fundamental defect of the whole procedure
of arriving at ‘the’ methodology of ‘social science’. This is,
briefly, to start with certain assumptions about social subjectmatter, established prior to specific scientific theories, and to
infer from them, allegedly a priori, what the methodology of
those particular theories must be. But this precludes from the
start (if the procedure is followed consistently) scientific criticism of those initial assumptions, so that the whole proceeding is
in principle dogmatic and methodologically obscurantist. Prop30

erly considered, neither ‘subject-matter’ nor ‘methodology’ is
absolutely prior with respect to the other: at most, one may be
contextually prior. What happens in the actual constitution of a
science is that a roughly delimited subject-matter is first handled
with whatever means are available. The resulting theories, to the
extent that they prove satisfactory, suggest improved methods,
which in turn permit the construction of better theories, and so
on. (Of course, this is simplified. There may, for example, be
imports of methodologies from other domains. But I am concerned here only with the central point.) There is no a priori
reason for thinking that, even at one time, this leapfrogging of
‘substanctive’ ??? and ‘methodological’ factors will yield any
single methodology over a whole domain; indeed no such reason
to think that there must be anyone item appropriately called
‘social science’ – except at a very high, unenlightening and
maybe misleading level of generality – rather than a perhaps
interlocking body of ‘social sciences’. In fact, exactly the same
holds for ‘natural science’, rather than ‘natural sciences’. Finally, the very contrast, natural versus social sciences, is not one
that can simply be taken for granted. For it may well be that
‘natural sciences’ and ‘social sciences’ interlock by virtue of
more or less complex similarities and differences, and that even
the way this is changes over time.

12 General Conclusion
None of this is allowed for in a foundationalist philosophy of the
sort which is exemplified by innumerable doctrines in the history
of philosophy. Bhaskar’s is just a recent example, which has
probably attracted a little attention by virtue of its ‘realism’ as
contrasted with the debilitating positivism which for a longtime
held the field of philosophy in science, and for its ‘Marxist’

pretensions. But in the end all such foundationalists are, as
Spengler said of certain modem artists, ‘acrobats who bustle
about with hundred-kilo weights made of cardbcrard’ .18


Chalmers (1988) is oriented more towards general metaphysical
issues; Albury, Payne, Suchting (1981) and Benton (1981) more
towards issues of social theory. Bhaskar (1989) responds to the
first and third. I was able to read this only after I had effectively
finished the present paper. But I have nothing to change as a
result of having read it, as there is nothing there but more of the

Cf. Marx’s distinction between ‘vulgar criticism’ on the one
hand and ‘true’ or ‘genuine’ criticism on the other in his 1843
critique of parts of Hegel’s Philosophie des Rechts: MarxEngels, Collected Works 3 (London: Lawrence and Wishart,
1975),91 (= Werke [Berlin: Dietz, 1956ff) 1: 296).


For more detailed discussions of certain issues see mainly
Suchting (1983), (1986), (1991).


For a criticism of one account, which is at least worked out
sufficiently to come to grips with, see Nerlich and Suchting
(1967) and Suchting (1969). For a demolition of necessitarian
approaches in general see now Van Fraassen (1989).


See Aristotle’s distinction between dynamis tou poiein and
dynamis tou paschein in Met. IX, 1-6 and Locke’s between
‘power … to make, or … to receive any change’ (Essay, 11, xxi.2).

I earlier on criticised the regularity theory in my paper ‘Regularity and Law’, Boston Studies in the Philosophy o/Science, Vol.

XIV (Dordrecht: Reidel, 1974). These criticisms were made
from within a problematic I now reject, but were damaging to the
account being criticised. Pretty much definitive criticisms of the
regularity theory, also from within the problematic of laws, will
be found in D. M. Armstrong, What is a Law 0/ Nature?



Radical Philosophy 61, Summer 1992





(Cambridge: Cambridge University Press, 1983) and M. Tooley,
Causation. A RealistApproach (Oxford: Clarendon Press, 1987).

A rejection of the problematic of laws will be found in Suchting
(1991) and, from a different point of view, Van Fraassen (1989).

Hume, An Enquiry Concerning Human Understanding, Sect.

XXI, Part I, note to penultimate paragraph; Kant, Critique of
Pure Reason, Bxv.

Hegel, for example, Phenomenology of Spirit, trans. Miller, p.


Althusser and Balibar (1970), pp. 41, 59, 60.

Althusser and Balibar (1970), p. 59; also Althusser (1969), p.

184 note 2).

This characterisation of epistemology in the mainline received
sense can also be found in, for example, Karl Mannheim’ s
excellent early study (1922/1953). For the various choices
within this scheme, as well as the aporiai (to use the author’s
own tenn) in the straightforward realist epistemology that he
defends (in the context of the belief, shared by TR, that ‘epistemology’ is founded upon ‘ontology’) see the comprehensive
exposition in Hartmann (1925).


Goethe to Eckennann, 18 February 1829 – GesprNche (lnsel
ed.) I, p. 298.


See, for example, Bihme (1980).


These are what Collingwood (1940) calls ‘absolute presuppositions’.


Remarks on the Foundations of Mathematics (Oxford: Basil
Blackwell, 1956), p. 1713 (para. 13).

See note 1 above.

The root of the trouble here (and indeed in many places elsewhere) is the taking of relations as fundamental, which locates
the account within the same problematic as individualism (different choices within the conception of society seen as individuals-in-relations), instead of practices. See my (1983), Chs. 12
and 17.

Oswald Spengler, Der Untergang des Abendlandes (Munich:

Beck, 1980), p. 378. (The English version – The Decline of the
West, I, p. 294 – is an inadequate paraphrase.)



Non-incidental References

Spinoza and Leibniz
is the central theme of Volume 6 (1990) of

An International & Interdisciplinary Series
On Spinoza’s relation to Leibniz, esp.: Truth, method and correspondence
(Garrett) / Reason and Experience (Biasutti) / Mathematical truth (Yakira) /
Language and cognition (Dascal) / Comparati’Oe lines (pombo) / Ehistoi’1f!

(Boss) / Ethics and politics (Salas Ortueta) / GesprĀ«h au/ tier Grenzlinie (de
Further: Articles on Physique pantheiste et determinisme: Spinoza et Huygens
(Vuillemin) / Some early responses to Spinoza’s philosophy in Jewish thought
(Levy) / – New Documents of Proto-Spinoza Franziscus van den Enden,
Stenos statements on Spinoza and spinozism (Klever) / Van den Enden

Reports & Information on congresses & activities in France and the Nethderlands.

Reviews of 17 Spinoza publications from 1980-89

390 pp. ISBN 3-88479-552-X. ISSN 0179-3897. Format 21 x 14 cm. Price:

Subscr. DM 48,- / US $ 24,80. Single copy DM 63,- / US $ 32,50.

Already published (and still available):

Vol. 1 (1985): Spinoza’s philosophy of society. ISBN 3-925-247-009.

Vol. 2 (1986): Spinoza’s epistemology. ISBN 3-925-247-01-7.

Vol. 3 {1987}: Spinoza and Hobbes. ISBN 3-925-247-02-5.

Vol. 4 (1988): Spinoza’s early writings. ISBN 3-88479-461-2.

Vol. 5 (1989): Spinoza and Literature. ISBN 3-88479-546-5.


Next issues: Vol. 7 (1991): The ethics in the “Ethics”. Vol. 8 (1992): Spinoza’s Psychology and Social Psychology. Vol. 9 (1993): Spinoza in the ethical
and political diScourse of modernity.


Verlag Konigshausen & Neumann, P.O. Box 6007
D-8700 Wiirzburg (FRG)

Radical Philosophy 61, Summer 1992

Albury, R, G. Payne, and W. Suchting (1981)
Review of Roy Bhaskar, The Possibility of Naturalism, Economy and Society 10, pp. 367-79.

Althusser, L. (1969) For Marx, London: Allen
Lane The Penguin Press.

Althusser L. and E. Balibar (1970) Reading Capital, London: NLB.

Benton, T. (1981) ‘Realism and Social Science’,
Radical Philosophy 27.

Bhaskar, R (1989) ‘Postscript’ to The Possibility
of Naturalism, 2nd ed., Sussex: Harvester Press.

Bihme, G. (1980) Alternativen der Wissenschaft,
Frankfurt/M: Suhrkamp.

Chalmers, A. (1988) ‘Is Bhaskar’s Realism Realistic?’ Radical Philosophy 49.

Collingwood, R G. (1940) An Essay onMetaphysics, Oxford: Clarendon Press.

Hartmann, N. (1925) Grundzuge einer M etaphysik
der Erkenntnis, 2nd ed., Berlin/Leipzig: W. de

Mannheim, K. (1922/1953) ‘Structural Analysis
of Epistemology’ in: Essays on Sociology and
Social Psychology, London: Routledge and Kegan

Nerlich, G. and W. A. Suchting (1967) ‘Popper on
Law and Natural Necessity’, Brit. J. Philos. Sci.

18: 223-235.

Suchting, W. A. (1969) ‘Popper’s Revised Definition of Natural Necessity’, Brit. J. Philos. Sci. 20:


Suchting, W. A. (1983) Marx: An Introduction,
Brighton: Wheatsheaf.

Suchting, W. A. (1986) Marx and Philosophy,
London: Macmillan.

Suchting, W. A. (1991) ‘On Some Unsettled Questions Touching the Character of Marxism, Especially as Philosophy’, Graduate Faculty PhilosophyJournal (New School for Social Science, New
York), 14: 139-207.

Van Fraassen, B. C. (1989) Laws and Symmetry,
Oxford: Clarendon Press.


Download the PDFBuy the latest issue