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.

Scientific Explanation and Human Emancipation

Scientific Explanation
and Human Emancipation
Roy Bhaskar

1. Introduction
What connections, if any, exist between
explanations in the human sciences and the
project of human emancipation? I want to
addr~ss this issue in the light of the
transcendental realist reconstruction of
science (2) and the critical naturalism
which that reconstruction enables (3).

My main target will be positivism, and
the doctrine of the value-neutrality of
social science.

But I will also be attacking a rationalistic intellectualism, which
sees social theory as (actually or potentially) immediately efficacious in practice.

In opposition to positivism, and its
historicist/hermeneuticist displacements, I
want to argue that the human sciences are
intrinsically critical and self-critical;
that accounts of social objects are not only
value-impregnated, but value-impregnating;
and that the possibility of a scientific
critique of lay (and proto-scientific) ideas,
grounded in explanatory practices based on
respect for the authenticity and epistemic
significance of those ideas, affords to the
human sciences an essential emancipatory
impulse, in virtue of which, subject to the
operation of various ceteris paribus
clauses, we pass securely from statements
of fact to value.

However, in opposition to the idealist
(theoreticist) notion of the unmediated
efficacy of social science, I want to insist that it always occurs in the context
of a situation co-determined by noncognitive features too.

Social theory
appears, then, as conditioned critique: as
subject, in its genesis and effect, to nontheoretical, as well as theoretical, determinations (whose critical understanding is
itself part of the task of theory).

This is
of course an implication of historical

To conceive critique as conditioned by factors outside itself is not to
impugn its normative power, merely to be
realistic about its practical impact.

On the view advocated here, knowledge,
through necessary, is insufficient, for

For to be free is (i) to know,
(ii) to possess the opportunity and (iii) to
be disposed to act in (or towards) one’s
real interests.

Freedom can thus be no
more the simple recognition of, than it is

escape from necessity. Hegel (and Engels)
and Sartre (and perhaps Marx, at least in
his more chiliastic proclamations) are
equally wrong – on the condition that circumstances or wants contain any non-cognitive components.

It is salutary to remember
that there is a logical gap between ‘knowing’

and ‘doing’, which can only be bridged by
‘wanting in suitable circumstances’.

It is
the argument of this paper that the special
qualitative kind of becoming free or liberation, which is emancipation, and which consists in the transformation, in ‘selfemancipation’ by the agent(s) concerned,
from an unwanted to a wanted source of determination, is both causally presaged and
logically entailed by explanatory theory,
but that it can only be effected”iti practice.

2. Explanatory Schemata and Transcendental Realism
To explain something is to resolve some
agent’s perplexity about it: it is to render
the unintelligible intelligible – by the
elucidation, extension, modification or
replacement of that agent’s existing conceptual field (4).

In particular, scientific explanations do not resolve problems by
subsuming some particular problem under a
more general one, but by locating such
(normally already generalised) problems in
the context of a new cognitive setting; it
is (new) concepts, not (universal) quantifiers which accomplish explanatory problemresolution in science. But the empirical
adequacy of any such resolution must be
tested by devising or finding conditions
under which the referent of the (conceptual)
object posited in the explanans operates
free from extraneous influences. Now the
enduring and transfactually active nature
of such referents is a condition of the
intelligibility of this experimental/
exploratory activity; and so the philosophy
of science must draw ontological distinctions between structures and events (the
domains of the real and the actual) and open
systems and closed, indexing the stratification and differentiation of reality (5).

Typically, then, to explain an event,
regularity etc. is to bring it under a new

scheme of concepts, designating the structures, generative mechanisms or agents producing it.

But, in line with their undifferentiated ontology, the dominant traditions in the philosophy of science have not
clearly distinguished theoretical from
practical (concrete, ‘historical’ or applied)
explanations, neither of which are either
deductive or inductive in form.

explanations are iteratively analogical retroductive: i.e. antecedently available
cognitive resources are used to construct
plausible models of the mechanisms producing
identified patterns of phenomena, which are
then empirically checked out, and, if
deemed adequate, in turn explained – in a
continuingly unfolding dialectic of taxonomic and explanatory knowledge (6).

Practical explanations involve the RRRE
schema: i.e. resolution of complexes
(‘conjunctures’ or ‘compounds’), redescription of their components, retrodiction to
possible antecedents of these components and
elimination of alternative possible causes

Thus if theory assumes the form of
the abduction of the abstract from the concrete, applied work characteristically
depends upon the reverse movement, leading
to the recovery from the abstract of the
concrete, now reconstructed as the product
of a multiplicity of abstractly apprehended

Knowledge of structures and
of their contingent modes of articulation
in time thus appear as distinct moments of
scientific activity.

Between abstract
sciences and the reconstructed concepts of
concrete objects, lie the concrete sciences
(like biography) which study the ensemble of
significant truths about a given thing and
the intermediate sciences (like ecology)
which study the confluence of two or more
orders of determination.

Of course in as
much as these types of explanation succeed
in identifying real, but hitherto unrecognised, conditions and patterns of determination they immediately augmen~ our knowledge,
and hence (on the definition enlisted above)
ceteris paribus our freedom.

On the metaphysics implied by the new
analysis of science, ontology is vindicated
as a study of the presuppositions of scientific practice, and the error of its reduction to epistemology is isolated.

the world, as we actually know it (i.e.

under the descriptions currently available
to science), is now revealed as characterised by situations of dual and multiple
control and by the phenomenon of emergence.

But transcendental realism does not
license the simple-minded transapplication
of results derived from reflection on the
conditions of the natural sciences to the
social sphere.

Rather, it is only in virtue
of an independent analysis, that we are in a
position to see that there is a paramorphic
relationship between the natural and the
human sciences, such that there are knowable
structures at work in the human domain
partially analogous, but irreducible, to
those identified in nature.

Thus the material causality of social forms appears as a
condition of intentional agency, and the
efficient causality of beliefs as a condition of discursive thought.

But a realist
interpretation of non-physical (sui generis

sociological, psychological) explanations
of human phenomena is only justified if it
can be shown that there are properties
instantiated in the human world inexplicable
in terms of different sets of conditions of
purely natural laws.

In concrete terms, the
emergence of society is manifest in the
causal irreducibility of social forms in the
genesis of human action (or being), and the
emergence of mind in the causal irreducibility of beliefs in the explanation of those
changes in the states of the physical world
which are the result of intentional agency.

(Of course the relations are two-way.

the human effects of natural causes are
normally mediated as cultural products, and
the social effects of human actions in
The resulting critical naturalism has
nothing in common with either positivism or
scientism, because clear differences transpire between positivism and science, on the
one hand, and the human and natural sciences,
on the other.

Nor is it ‘objectivist’ in
either method or result: for it is predicated on the analYSis of (existing conceptualisations of) historical practices, and it
situates these analyses within the framework
of the same historical processes which
social science describes and philosophy

But positivism’s anti-scientific hermeneutical foil is shown to be
equally untenable – for the very features
it picks upon (such as verstehen) themselves require for their intelligibility
crucial aspects of the categorial framework
of natural science (existential intransitivity, causality etc.).

Nor do neo-Kantian
syntheses of dual criteria or multiple
interests fare any better.

This is not
only because the components of the attempted
syntheses are faulty (e.g. in being based on
a positivistic misconception of natural
science), but because the very project of
rendering ontological mediations as epistemological divisions is fundamentally mistaken.

Thus conceptuality is a condition of generality in the historical domain; and there too
an emancipatory conatus is initiated as an
effect of explanatory power, in circumstance~
where it cannot be a universal or constitutive condition for it.

(The critical cutting
edge that Habermas’ work retains despite
this is achieved only by the effective
noumenalisation of discourse as a counterfactual counterpoint to the realm of historical agency (8).)

3. Social Structure and Human Agency
On the transformational model of social
activity (TMSA), entailed by the new critical naturalism, the ontological structure of
human activity or praxis is conceived, after
Aristotle, as consisting in the transformation by efficient (intentional) agency of
pre-given material (natural and social)

A criterion for differentiating the
social from the purely natural material
causes is given by their property that,
though necessarily pre-given to any particular agent, and a condition for every intentional act, they exist and persist only in
virtue of human agency.

On this model, then,

social structure and human agency are seen
as existentially interdependent but essentially distinct (9). Society is both everpresent condition and continually reproduced
outcome of human agency: this is the duality
of structure (10). And human agency is both
work (generically conceived), i.e. (normally
conscious) production and reproduction of
the conditions of production, including
society: this is the duality of praxis.

Thus agents reproduce, non-teleologically
and recursively, in their substantive motivated productions, the unmotivated conditions
necessary for – as means of – those productions; and society is both the medium and
result of this activity. From this model
flow a series of limits on naturalism, which
may be summarised as the activity-, conceptand space-time-dependence of social forms,
in virtue of which (as I have attempted to
argue elsewhere (11» a sui generis social
science is possible. Of course the holistic,
hermeneutical and historical character of
social objects necessitate differences in
the structure of social scientific explanations so that, paradigmatically, social
complexes must be understood as partially
con’ceptually articulated totali ties in continual transformation.

Similarly, the
impossibility of artificially producing, and
the unavailability of spontaneously occurring, closed systems requires reliance on
purely explanatory (non-predictive) criteria
of confirmation and falsification, and more
generally theory-development and -assessment.

However, in relation to the specificity of
social objects, (non-scientistic) scientific
knowledge of them is possible.

The TMSA allows us to pinpoint a double
spt of paired mistakes: the ontological
errors of reification and voluntarism, and
the epistemological ones of (social) determinism and (methodological) individualism.

(Both may be combined to produce various
pseudo-dialectical hybrids.) And it allows
us to isolate the closely affiliated weaknesses of the substantive traditions of
structuralism and functionalism, on the one
hand, and action-oriented and interpretative
sociologies, on the other. For its part,
the TMSA respects a methodological distinction between the social sciences, which
abstract from human agency, studying the
structure of reproduced outcomes; and the
social psychological sciences, which
abstract from reproduced outcomes, studying
the rules governing the mobilisation of
resources by agents in their everyday interaction with one another and nature.

If the
object of the former is social structure,
that of the latter is social interaction.

They may be linked by the study of society
as such, identified as the system of relations between the positions and practices
agents reproduce and transform, the subject
matter of the social science of sociology.

The TMSA can allow that the form of psychology, the study of mental processes, may b~
species-general, but its content will always
be historically specific.

The transformational model and the
structure/praxis connection are represented
as in Diagrams 1 and 2 below
On the TMSA unintended consequences, unacknowledged conditions and tacit skills



i i





,’ :,







Diaer8111 1: The Transformational Model of Social Activity



reproduo tiorV


3 production

rib: 1, l’



Dia6ram 2: Structure and Praxi.s
unaoknowledged conditions;
= unacknowledged motivation; 4 = taalt sk11la



(cf. 1, 2 and 4 in Diagram 2) limit the
actor’s understanding of his social world,
while unacknowledged (unconscious) motivation (cf.3) limits his understanding of himself.

Corresponding to each of these limits,
knowledge has a distinct emancipatory role
– at 2 and 3 via the conditions and at 1 and
4 via the effects and form of praxis.

Now the continuity, depth and reflexivity
of human agency suggest the model of it
represented in Diagram 3, based on a model
proposed by Anthony Giddens (12).

ratiooination R


oonditions of




-….. unintended
I consequenoes










Diagram 3: The Stratification of Aotion

Discursivity presupposes a distinction
between real and possible, including ratiocinated, reasons, grounded in the causal
efficacy of the former.

Ratiocination, Rm ,
is a property of the reflexive monitoring
of conduct. Where Rm f Rr there is the
possibility of rationalisation. Real
reasons are the wants that prompt motivation
and ceteris paribus issue in action (13).

As such they may be regarded as efficacious
beliefs, which may be conscious or unconscious, trained on objects of desire.

as such they consist in a cognitive-conative
vector or perhaps better ensemble (see
Diagram 4).

beliefs (knowledge)


(dispositions to act)


Diagram 4: Beliefs. Desires and Actions

The error of ‘theoreticism’ (see Section 1)
is now clear: it involves the attempted
elimination of the conative component, no
doubt as rooted in our ‘inner nature’ as
the cognitive component, from the generative
matrix of action.

In social phenomenology unintended consequences may take the well-known forms of
counterfinality and suboptimality (e.g. in
a prisoner’s dilemma) (14). The conditions
figuring on the left-hand side of Diagram 3
include the rules and resources agents
command in such games.

Like all such conditions they may be unmotivated and unacknowledged.

These are features that a general
social phenomenology, whether rational or
empirical, cannot itself, without vicious
circularity, explain.

For such rules and
resources are at once historical deposits,
and so always subject to a potentially unrecognised possibility of supercession.

The games of the life-world (lebenswelt) are
always initiated, conditioned and closed
outside the life-world itself.

4. On the Critique of Interpretative Fundamentalism
The TMSA shows what may escape (and so be
misconstrued by) consciousness in our conscious activity.

But are there perhaps
elements in our experience or aspects of our
consciousness of which we must be certain,
and which (perhaps in virtue of this) are
not subject to the possibility of historical

The history of post-Cartesian philosophy
is largely the history of the attempt to
establish just such an Archimedean point for
knowledge, free from the possibility of
error and impervious to every form of doubt.

Thus in a recent empiricist avatar, scientific knowledge was conceived as incorrigibly
grounded in (or even exhausted by) sensedata or operations.

Of course we now know
that there are no foundations of knowledge,
that there is no uniquely privileged level,
moment or type of operation, that there are
no brute data; that the facts already contain a certain ‘sedimented’ reading of the
world (that natural facts are social institutions), and that the relationship between
theories and facts is between the contents
of two interdependent kinds of conceptual
schemes, one of which is taken as referring
to objects apprehended in experience.

short we now know that the facts are theorydependent and changeable; and science itself
appears, as one might anticipate on the
TMSA, as a historical process of levels and
connections, a weighted network, without
foundations, developing in time.

This view
does not dispute the epistemic value of

However, it interprets thjs
not as the absolute privilege of a content,
but as dependent upon the ontological and
social contexts within which the significant
experience occurs (15).

Now in as much as there has been a
‘coupure’ in the recent philosophy of the
human sciences, it lies in recognition of
the significance of the condition that man
is a self-interpreting and self-motivating

animal, whose language and beliefs are in
some manner necessary for and productive of
his life; so that human reality faces the
scientific neophyte as already pre-interpreted, as (as it were) linguistically and
cognitively ‘done’, prior to any scientific
investigation of it. These pre-interpretations are not externally related and contingently conjoined to what happens in the
human sphere, but internally related to and
constitutive of it (16).

It was natural,
then, in the wake of this understanding, to
suppose that these interpretations (or
beliefs) would constitute the base or foundations of social knowledge; to regard them
as consisting, so to speak, in brute interpretations (or beliefs), whether such dataanalogues were conceived positivistically
as immediately available to the investigator
or dialogically as dependent upon work
within his own culture. Thus one had a
transposition of the familiar thematics of
classical philosophy in a hermeneutical key
– more plausible than in the original,
perhaps, because nature is not self-interpreting, but little different in logical
form or epistemological effect. For both
the reductionist thesis that social knowledge is exhausted by, and the milder position that it is rooted in (and so must be
consistent with) self-interpretations lead
inexorably to a displaced hermeneuticised
scientism and a consequent ‘disavowal of
reflection’ (17).

In either variant the
doctrine of the incorrigible, because ontologically constitutive, foundations of
social knowledge secretes, like its positivist prototype, as an inevitable corollary:

the doctrine of the neutrality o~ social

Of course Hegel demonstrated long ago
(18) that the fundamentalist programme is
both radically incomplete and viciously
circular: in that it not only cannot establish its own legitimacy, but must (implicitly or explicitly) presuppose some unvalidated ‘knowledge’.

And it is clear that, in
these respects, any Viconian facimus must
share the same limitations as the Cartesian

For just as Descartes must assume
some content to initiate his axiomatics; so,
for Vico, God or man must already possess
some matter for their constructions, that is
to make their worlds, and what any agent
does not make (what it must take to make) it
possesses no privileged understanding of
(just as what an ego cannot demonstrate it
must remain uncertain about).

It should be
noted that on the transformational model we
do not make the conditions or consequences,
skills or motives of our intentional making
(cf. Diagram 2 above); so that our beliefs
about, or interpretations of, our actions
cannot be constitutive in the requisite

In considering the social-incorrigibilist
position in slightly more detail, it is convenient to distinguish two sub-arguments for
it: one Viconian, the other hermeneutical in

The more strictly Viconian argument contends that one and the same knowledge is
used to generate as to explain behaviour;
so, as it were, superimposing a transcend-


ental unity of agency on that of consciousness.

But agency may consist in the exercise of tacit skills (19). Moreover, the
consciousness involved, and knowledge
exploited, in action may be practical and
so cannot immediately ground, even if it is
held to be the ultimate empirical touchstone of, a supposedly discursive, theoretical science. Thus we need not be able to
say how we do what we know very well how to
do (or vice versa), even when, as Chomsky
has made abundantly clear, the first-order
skills are themselves verbal, discursive

Secondly, while it is surely the case
that communication (and interaction generally) would be impossible unless we were
normally able to identify agents’ immediate
reasons for acting, it does not follow
(a) that we must be always able to do so,
or (more fundamentally) (b) that we must be
able to identify the underlying reasons for
(or causes of) those reasons.

For example,
we may know that a person is washing his
hands or polishing an icon, but not why he
is doing so.

And so the possibility arises
of the systematic misdescription of reasons
in rationalisation or ideological-mystification, i.e. in the self-misunderstanding of
agents or forms of life.

The hermeneutical argument for social
foundations maintains that it is interpretations that uniquely and completely differentiate the social world from mere assemblages
of physical happenings, so that it is only
and sufficiently by reference to them that
its sui generis character can be sustained.

Elsewhere I have attempted to show that the
social world is not exhausted by its conceptual aspects, and that such aspects are
in any event not necessarily immediately
available to consciousness (20).

Thus although the immediate intentions of agents
and meanings of acts cannot normally be misdescribed for mutual understanding or functioning language-games to be possible, both
intentions and meanings may be opaque to
agents (a) occasionally, at the level of
everyday interaction and (b) systematically,
at the level of the underlying explanations
and descriptions of the reasons motivating
their behaviour in such interaction. Particularly significant here is the possibility
of a’contingent generalisation of Gedel’s
theorem in the direction of what I shall
call ‘meta-critique’. This consists in a
critique of a language on the grounds of its
incapacity to adequately express ideas or
institutions which are customarily described
by means of them.

Such a critique aims to
pinpoint precisely what cannot be said in a
particular language about what is said or
done by means of it.

In general, then, the generative role of
agents’ skills and wants, and of agents’

(and social) beliefs and meanings must be
recognised without lapsing into an interpretative fundamentalism by conferring discursive and/or incorrigible status upon them.

But how are beliefs and meanings in particular to be identified in the face of the
corrigility of statements of them? Now
agents’ accounts are more than just evidence;
they are an internally related aspect of
what they are about. Thus any resolution of
this problem must be two-way: the social

investigator must avoid both the extremes of
arrogant dismissal of and of fawning assent
to first-person accounts (21).

But agreement between agent and investigator hardly
seems either a necessary or sufficient criterion for an adequate interpretation.

Rather, it would seem that the adequacy of
any interpretation, or more generally of any
set of self-understanding, can only be shown
in relation to the point of the interpretation (or understanding) in the always more
or less contingently circumscribed context
of the agents’ self-formation, that is total
developing life-activity (22).

If judgments about belief cannot be separated from judgments about activity, judgments of meaning, again presupposing a
dialogical fusion of horizons (23), cannot
be separated from judgments of explanatory
adequacy (presupposing a degree of causal
interaction). Thus the so-called ‘problem
of the indeterminacy of translation’ is
resolved in practice by selecting that translation which is explanatorily most adequate
(whether or not it is the most ‘charitable’)
in the context of what is already known
about the organisation of the particular
society in question (and societies in general).

The most adequate explanation will
save the maximum of significant phenomena in
the subject matter at stake, showing in that
subject matter precisely the degree and type
of ‘irrationality’ that does so.

5. Facts and Values: Hume’s Law and Helices
I now want to show that the human sciences
are necessarily non-neutral; th-atthey are
intrinsically critical (both of beliefs, and
of the objects of beliefs), self-critical
and value-impregnating; and in particular
that they both causally motivate and logically entail value-judgments ceteris paribus

I will not be concerned to argue
against the scientistic misconception that
factual judgments are value-free, partly
because this connection has been, if not
always adequately theorised, widely recognised (inside as well as outside the analytic tradition (24)), but mainly because I
want to address myself more to an aspiration
than what is characteristically misconstrued
as a ‘difficulty’: the hope that the human
sciences might yet come to be in a position
to cast some light on what we ought to do
and say, feel and think.

—-In fact of course one is dealing with a
fact-value helix here (see Diagram 5).

Diagram 5: Fact(value Helix

And it is clear that the scientistic denial
of the value-impregnation of factual discourse, involving the reification of propositional contents, shares with the positivist
denial of its converse, as a common condition of their plausibility, a naive extensionalist theory of meaning (whether in
physicalist, sensationalist or Platonist
guise). Moreover it shares with the theoreticist (rationalist) conception of the unmediated efficacy of theoretical discourse
a neglect of the conative and affective
bases of action, involving a voluntarism of
theoretical praxis. The converse ‘practicalist’ error – of anti-intellectualist irrationalism – ignores of course the cognitive
bases of action.

These four errors can be
represented as in Table 1 below.

positiYism (and displacements)

theoretioism (idealism) ~ P ~ T
Table 1

F standsibr facts and theories
P stands for practioe

Theoreticism, as defined above, leads naturally to the denial that practice (to the
extent that it is not merely a redescription
of ‘theory’) possesses any efficacy in the
generation of theory.

Once the value-implications of theory,
and the rational assessability of wants (in
virtue of their grounding in beliefs), are
accepted, then Diagram 4 can be modified as
in Diagram 6.


beliefs (knowleoge)



Diagram 6



Of course there is a feedback between values
and actions, mediated by practices, including scientific (knowledge-producing) ones,
so that they should be understood as connected by a loop as in Diagram 3.

There is an important asymmetry between
the F ~ V and T ~ P relationships, on the
one hand, and the V ~ F and P ~ T relationships, on the other.

Factual and theoretical considerations not only predispose and
motivate but, in favourable circumstances
(and subject to the operation of CP clauses),
logically entail value and practical judgments.

On the other hand, value and practical considerations, while they may (and in
general will) predispose and sometimes moti-

vate, do not (non-trivially) entail factual
and theoretical judgments (25).

It is just
this asymmetry which makes the helices in
Diagram 5 (and in its theory/practice analogue) potentially relationa1ones: that is,
progressive, i.e. developing, spirals,
rather than viciously self-confirming, and
so self-destroying, more or less rapidly
vanishing, circles.

My core argument is very simple.

turns on the condition that the subject
matter of the human sciences includes both
social objects (including beliefs) and
beliefs about those objects. Philosophers
have characteristically overlooked, or concealed, the internal relations connecting
these aspects: empiricists by objectivising
beliefs, idealists by bracketing away objects. Now these relations, which mayor
may not be intra-discursive (depending upon
whether the first-order object is itself a
belief), are both causal and cognitive – in
the ontological or intransitive dimension we
are concerned with relations of generation;
in the epistemological or transitive dimension of critique.

But it is the causal
relation of generation that grounds the
epistemological programme of critique.

Now I am going to contend that if we
possess (i) adequate grounds for supposing
that a belief P (about some object 0) is
false and (ii) adequate grounds for supposing that S (co-) explains P, then we may,
and must, pass immediately to (iii) a negative evaluation of S(CP) and (iv) a positive
evaluation of action rationally directed at
the removal of S(CP).

To elaborate; in as
much as we can explain, that is show the
(perhaps contingent) necessity for some
determinate false consciousness, or perhaps
just some determinate consciousness under
the determinable ‘false’, then the inferences to a negative evaluation of its
source(s) and a positive evaluation of
action oriented towards their dissolution
are ceteris paribus mandatory.

It should be stressed straight away that
such action can only be rationally justified
CP to the extent that there are grounds for
supposing the source to be dissoluble; and
that the TMSA does not in itself license the
supposition of a society without some false
consciousness. The notion of false consciousness here involves simply in the first
instance the notion of disjuncture, mismatch
or lack of correspondence between belief and
object. But, as I shall presently show,
this general pattern of argument may be
readily extended to accommodate both the
cases of more interestingly specific forms
of false consciousness and that of other
types of inadequate consciousness (and,
indeed, more generally, defective being).

In principle this pattern of inference
applies equally to beliefs about natural, as
well as social, objects, on~condition
(and to the extent) that the relevant source
of false consciousness S, is itself a social

But in this case S cannot be the
same as, or internally related to, 0, and
neither S nor P can be causal conditions for
the genesis or persistence of 0, as in the
cases of psychological rationalisation and
ideological mystification, where S, P and 0
are typically causally interrelated. Only


in the case of beliefs about social objects
can the illusory (or more generally defective) character of consciousness be a condition of what it is about. However, given
that beliefs about nature are social objects
all the modalities of false consciousness
may clearly apply to our beliefs about our
beliefs about nature: that is, to our understanding of – as distinct from in – science.

I shalr-call (i) the criticar-and (ii)
the explanatory condition. Of course even
if the critical condition alone is satisfied
then we also pass immediately to a negative ‘

evaluation of P(CP), and of action based on
or informed by P(CP).

But I want to distinguish this kind of ‘criticism’ which
although it formally violates and so’refutes
‘Hume’s Law’ (26), remains silent on the
causes of error, from an explanatory critique.

Criticism, in Marx’s words, ‘knows
how to judge and condemn the present, but
not how to comprehend it’ (27).

The essence
of Marx’s objection to criticism may, I
think, be stated thus: it employs value (and
particularly, although contingently, moral)
terms in the absence of any kind of causal

At its best, i.e. if displayed
in- naturalistic (i.e. non-intuitionist or
-emotivist) form, it can furnish grounds for
belief and action which, if true, a fortiori
increase our freedom.

But criticism says
nothing about, although it may of course
(intentionally or unintentionally) causally
affect, the (causal) conditions of actions,
the springs (so to speak) of belief and behaviour, the sources of determination.

so criticism cannot contribute to the discursive elucidation of the topic of the
transformation of the sources of an agent’s
determination from unwanted to wanted ones:

i.e. of emancipation. Only a discourse in
which the explanatory, as well as the critical, condition is satisfied can be intrinsically emancipatory.

As the concept of a ‘critique’ is betterknown I shall not discuss it here.

structures of the various types of ‘depthexplanation’, which may be undertaken at
several different levels (including the analytical, phenomenological and ideological),
is considerably more complicated than that
depicted in the bare form of an explanatory
critique, but the transition from fact to
value is effected in essentially the same
way. The possibility of an explanatory
critique constitutes the kernel of the emancipatory potential of the human sciences.

But to illustrate the possibilitIes here
fully, I want to develop the argument on a
series of levels, which may be regarded as so
many ratchets of reason.

Level I: Technical Rationality
Patently, the human sciences may be used,
like any other sciences, to achieve (more or
less consciously formulated, and justified)
ends, which may of course be adjudged
equally good or bad.

In particular, explanatory theories may be used, in conjunction
with statements of particular initial conditions, to generate technical imperatives
akin to ‘put anti-freeze in the radiator
(if you want to avoid it bursting in winter)

If such imperatives ever appear to
depart from the ends-means schema, it is
only because they already presuppose a context of human purposes in the domain of
their intended applications.

Level 11: Contextually-Situated Instrumental Rationality
The human sciences, even at the level of
instrumental rationality, are not symmetrically beneficial to the parties involved in
relations of domination etc.

For, in the
first place, explanatory knowledge increases
the range of real (non-utopian) human possibilities, which may mean of course decreasing the range of assumed or fancied ones.

But CP this will tilt the ‘balance of – in
a broad sense – political argument’ against
the status quo.

This is quite consistent
with the existence of only a simple external
connection between knowledge and politics.

Secondly, even on an instrumental interpretation, explanatory knowledge appears as
a necessary condition for rational selfemancipation (whether from the-oppression of
individuals, groups, classes, organisations,
systems of relations, structures of interaction etc. or from the oppression of conscious or unconscious systems of ideas, in
which the agent is entrapped).

Hence the
dominated, exploited, oppressed, repressed,
etc. have an interest in knowledge (in the
straightforward sense that it facilitates
the achievement of their wants).

And the
dominating, in as much as their interests
are antagonistic to those they dominate,
possess an interest in the ignorance of the
dominated (and perhaps even in their own
ignorance of the nature, or even the fact,
of their dominance).

Thus the human
sciences, and at a remove philosophy, cannot
be regarded as equally ‘a potential instrument of domination’ as of ‘the expansion of
the rational autonomy of action’ (28). The
human sciences are not neutral in their

6. Instrumental vs. Critical Rationality

Level Ill: Intra-Discursive (Non-Explanatory) Critical

At the first two levels, no attempt is made
to question the logical heterogeneity (and
impenetrability) of facts and values.

Despite this, the human sciences may still
have emancipatory implications (contingently,
so to speak) in virtue of (i) their use as
sheer technique and (ii) their effects in
the context of the existence of relations of
domination, exploitation and oppression.

The point has been made, particularly effectively by Roy Edgley (29), that any science
involves intra-discursive criticism, i.e.

criticism of other actually or possibly
believed (and therefore potentially efficacious) theories, hypotheses, etc. Acceptance of some theory T entails, ceteris
paribus, a series of negative evaluations:

on theories etc. incompatible with it, on





beliefs such theories underpin, on actions
they sustain or inform. Granted that ‘X is
false’ does not just mean ‘Don’t believe
(act on) X’ it certainly CP entails it.

is only if one denied any ontological connection between beliefs and action, or theory
and practice, that one might have grounds
for supposing that a change in theoretical
does not entail a change in practical judgments (CP).

But denying such a connection
makes practical idscourse practically

Again, this point is consistent
with a contingent relationship between a
science and its subject matter; and it
applies, quite indifferently, at the level
of intra-discursive critical rationality,
to all sciences alike. All the sciences,
then, irrespective of subject matter, are
intrinsically critical, and so evaluative.

Level IV: Explanatory Critical Rationality


All the sciences make judgments of truth or
falsity on beliefs about their object

But the human sciences, in virtue
of ,the distinguishing feature of their
object-domain that it includes beliefs
about inter alia social objects, also make
(or at least entail) judgments of truth or
falsity on (aspects of) that domain.

such belief/object correspondence, or lack
of it, appears immediately as a legitimate
object of social scientific explanation.

However, in as much as the natural sciences
are also concerned in their own substantive
critical discourse not just to isolate and
criticise, but to comprehend and causally
explain, illusory or inadequate beliefs
about the natural world, then they too,
assuming the second-order standpoint of the
intermediate science (in the terminology of
Section 2) of the natural-sociology (or
-psychology) of belief – in which natural
science is seen as a resultant of natural
and cultural determinants (30) – may come to
explain false consciousness of nature at
least partially in terms of human causes
(e.g. faulty instruments, inadequate funds,
superstition, the power of the church, state
or corporations).

In virtue of their explanatory charter, and in as much as they are in
a position to give well-grounded explanations of false consciousness, then, the
human sciences must, and the natural science~
may (mediately, via the natural-sociology of
belief), arrive at value judgments on the
causes, as well as the contents, of consciousness.

To recapitulate the central argument,
then, if we have a consistent set of
theories T which (i) shows some belief P
to be false, and (ii) explains why that, or
perhaps some such false (illusory, inadequate, misleading), belief is believed; then
the inferences to (iii) a negative evaluation of the object S (e.g. system of social
relations) accounting for the falsity of the
belief (i.e. mismatch in reality between the
belief P and what it is about 0) and (iv) a
positive evaluation of action rationally
directed at removing (disconnecting of transforming) that object, i.e. the source(s) of
false consciousness, appear mandatory CP.

This could be represented, informally, in

the inference scheme below as:

I . S . 1 (i) T P. ( i i) T exp I (P)
V(S -+ I~P» -+ (iv) V~ -S


(i i i) (31)

and we certainly seem to have derived value
conclusions (CP) from purely factual premisses.

Now for some possible objections.


It might be objected that ‘P is false’

is not value-neutral.

But if it is not
value-neutral, then the value-judgment ‘P
is false’ can be derived from premisses
concerning the lack of correspondence, or
mismatch, of objects and beliefs (in the
object domain).

Moreover as, assuming that
such judgments are intrinsic to any factual
discourse, we are nevertheless able to infer
from them, together with explanatory premisses, conclusions of a type which are not
intrinsic to every factual discourse (viz.

those specified in (iii) and (iv», we do
have a transition here that goes against
the grain of Hume’s Law, however precisely
that is supposed to be here interpreted or

On the other hand, if ‘P is false’

is value-neutral, then the inferences to
‘P ought not be believed (CP)’ and ‘Don’t
believe (act upon) P (CP)’ certainly seem


The suggestion that science itself presupposes, or embodies commitment to, certain
values, such as objectivity, openness, integrity, honesty, veracity, consistency,
coherence, comprehensibility, explanatory
power, etc. should certainly be welcomed suggesting, as it does, that the class of
the ‘value-neutral’ is as empt~ as that of
Austin’s original ‘constatives’ (32).

it does nothing either to rescue Hume’s Law,
or to deny the validity of inference-types
(iii) and (iv), which turn on the special
feature of the sciences of beliefs that
commitment to truth and explanatory power
entail the search for theories which will
possess value-implications that cannot be
regarded as conditions of, or as already
implicit as anticipations in the organisation of, scientific-activity-in-general.


It might be maintained that, although
inference-type (iii) is valid, (iv) is
faulty, so that no commitment to any sort of
action is entailed by the critical explanatory theory.

But this is not so.

For one
can reason straight away to action directed
at removing the sources of false consciousness, providing of course one has good
ground for supposing that it would do so,
that no ill (or sufficiently overriding ill)
effects would be forthcoming that that there
is no better course of action which would
achieve the same end.

Of course the inference scheme does not itself, conceived as a
philosophical reconstruction, determine what
such practical (‘critical-revolutionary’)
action is: that is the task of substantive

Of course ‘remove (annul, defuse,
disconnect, dissolve, transform) sources of
false consciousness’ does not specify what
the sources are, any more than ‘lying
wrong’ says which statements are lies.

Behind this objection, however, lie two
considerations of some moment.

First, the
kind of theory underpinning (iv) may be



different from that informing (iii). Diagnosis is not therapy.

We may know that
something is causing a problem without knowing how to get rid of or change it.

Secondly, an explanatory critique of this type
does not in general specify how we are to
act after the source of mystification (false
consciousness) is removed.

It focuses on
action which ‘frees’ us to act, by eliminating or disconnecting a source of mystification acting as an unwanted source of (co-)
determination, replacing that source with
another wanted (or perhaps just less unwanted) one, so achieving (absolute or relative) liberation from one stream of constraints or compulsions inherited from, as
the causalities (and casualties) of the

But it does not tell us what to do,
if and when (and to the extent that) we are

Thus emancipated action may, and
perhaps must, have a different logical form
from emancipatory action.

called ‘cognitive ills’. Their manifest
includes the explanation of the ‘practical
ills’ of ill-health, misery, repression,
etc.; and in between such ills and the cognitive ones, what might be called the
communicative ills of deception (including
self-deception), distortion, etc.

This indicates two further lines of consideration. First I.S.1 can be straightforwardly generalised to deal with the
explanation of such non-cognitive ills, with
a corresponding deduction of value-judgments,
as in I.S.3 below, where I-H stands for


T exp I-H.








However, as will be immediately obvious,
this deduction, despite its evident social
and epistemic power, is now no longer from
purely factual premisses, or from what is
immediately or self-evidently constitutive
of purely factual discourse.

And so it
cannot be used to achieve a formal refutation of Hume’s Law.

It is precisely on this
rock that most previous attempts at its refutation, including Searle’s notorious
attempted derivation of an ‘ought’ from the
rather tenuous institution of ‘promising’

(33), have broken. But further reflection
shows another possibility here: namely that
there are non-cognitive conditions, such as
a degree of good health and the absence of
marked asymmetries in political, economic
and the other modalities of power, for discourse (including factual discourse) -ingeneral to be possible.

If this is correct
then a formal derivation of an ‘ought’ can
proceed as in I.S.4 below:




T exp (I-H








Is there a sense in which I.S.1 and 2 are
epistemically prior to their non-cognitive
generalisations? Yes, in as much as empirically-controlled retroduction to explanatory structures always occurs in the context
of, and typically (in science) assumes the
form of, criticism of beliefs (consciousness) – scientific, proto-scientific, lay
and practical.

The human sciences, then, must make
judgments of truth and falsity, in virtue of
their explanatory charter.

And these, in
the context of explanatory theories, entail
value-judgments of type (iii) and (iv).

Mutatis mutandis similar considerations
apply to judgments of rationality, consistency, coherence, etc.

Thus I.S.1 can be
generalised in the cognitive direction represented in I.S.2 below, where C(P) stands
for the contradictory character of some
determinate set of beliefs.




T exp C(P)



C(P) )


But the human sciences are of course not
only concerned to explain what might be

7. Depth Rationality
Level V: Depth-Explanatory Critical Rationality
The most thoroughly explored applications of
I.S.1 and 2 involve the phenomena of psychological rationalisation and ideological

These phenomena are characterised by two distinctive features.

a doubling of necessity between misrepresentation (P) and source (S); so that the, or
some such, misrepresentation is not only
causally necessitated by, but causally
necessary for, the persistence or modulation,
reproduction or limited (non-essential)
transformation of its source.

Secondly, an
internal relationship between source (S) and
object (0); so that the misrepresented object

is either the same as, or at least causally
dependent upon, the source of the misrepresentation.

Thus, in a simple depth-psychological
model, an agent N may misdescribe his real
(i.e. the causally efficacious) reason, s,
for some action, 1/J , by p.

If P is itself a
contingently necessary releasing condition
for 1/J and s itself generates, in context, p
then we have:








To explain this we now posit a structure S
such that 1/J is (perhaps contingently) necessary for its persistence or modulation, as




Given s


p. sp






p the deductions proceed as in

I. S .1.

This paradigm may be easily extended to
include ‘outer’ as well as ‘inner’ causes,
including the self-mystification of forms of
social life, or systems of social relations,
in. ideologies. Thus the contradictions
which mystify Colletti (34) turn simply on
the necessary co-existence in social reality
of an object and a (categorially) false
presentation of it, where it is the inner
(or essential) structure of the object which
generates the categorially false presentation (or appearance).

Schema (7) is isomorphic with (5):

(7) E -+ A.




and (8) is isomorphic with (6) :

(8) R









R’ ,

where E = essence, A = appearance, P =
practices, and R, R’ the modulated reproduction of the capitalist mode of production.

Are there any general conditions on the
internal structure (E) of a self-reproducing
system (T) which generates and contains within itself (i.e. T) a functionally necessary
misrepresentation (A) of itself? It seems
plausible to suppose that E must possess at
least sufficient internal differentiation to
justify attributing to it a ‘spaltung’ or
split; and that if T is to be capable of
endogenous (essential) transformation,
rather than merely modulated reproduction,
the split must constitute, or be constituted
by, antagonistic (opposed) tendencies.

apart from the Colletti-style contradiction
built into the notion of the system’s misrepresentation of itself, it seems a priori
unlikely that what the human sciences may
empirically discover about the various
structural sources of false consciousness
will justify the application of a single,
unified category of ‘contradiction’ to those

Instead one might conjecture a
galaxy of concepts of contradiction, clustered around the core notion of the axiological indeterminacy generated by the logical archetype (together with the evaluative
connotations this secretes). The specific
concepts of contradiction would then achieve
their individuation in the constraints they
impose upon such indeterminacy and in their

thematisation of its form.

Perhaps the most famous depth-explanation,
Marx’s Capital, has the structure of a
triple critique: of theories, of the practical consciousness such theories reflect or
rationalise, and of the conditions explaining such consciousness.

But in Marx, and
the Marxisn tradition generally, the criticised (discursive and practical) consciousness is regarded not just as false but as
‘ideological’ – where ‘ideology’ is counterposed to ‘science’.

In addition to the
critical and explanatory conditions, one
thus finds a further set of categorial conditions.

Here beliefs are typically criticised for their unscientificity simpliciter,
or for their inadequacy in sustaining the
(irreducible) specificity of the subject
matter of their domains.

Thus in reification, fetishism, hypostatisation, voluntaristic conventionalism, organicism, etc.

social life is presented, in one way or
another, in an a-social mode – a condition
rooted, for Marx, in the alienation and
atomisation characteristic of capitalism as
a specific form of class society.

example, on Marx’s analysis, the wage-form
collapses a power (labour-power) to its
exercise (labour), the domain of the real to
the actual, while the value-form fetishistically represents social relations in the
guise of natural qualities.

The critique of
these gross categorial errors could be represented as:


T exp I (P) . T exp -S c (P)

-+ -V(S -+ -S . I (P) ) -+ VC/)




T exp I (P) . T exp -S (P)
-+ -V(S -+ -S . I (P) ) -+ VC/)


where -S and -S stand for the unscientific
and des08ialisin~ character of the forms in

What are we to make of Engels’ celebrated
rebuke to Lefargue: ‘Marx rejected the
“political, social and economic ideal” you
attributed to him.

A man of science has no
ideals, he elaborates scientific results, and
if he is also politically committed, he
struggles for them to be put into practice.

But if he has ideals, he cannot be a man of
science, since he would then be biased from
the start’ (11 August 1884)? While interests
both predispose and motivate analyses (and
their acceptance/rejection) in the human
sciences, so that Engels’ scientistic repudiation of the V -+ F connection is disingenuous; it remains the case that no value judgments other than those already bound up in
the assessment of the cognitive power of
Marx’s theory are necessary for the derivation of a negative evaluation of the capitalist mode of production (CP) and a positive
evaluation of action rationally oriented
towards its transformation (CP) – so that the
political commitment that Engels attributed
to Marx as, so to speak, a contingent extra,
can (on the assumption that Marx’s depthexplanation is correct) be logically grounded
in his scientific practice alone.

Of course
the theories now required to confirm, extend,
develop or refute Marx’s own analyses can
only be consequent upon engagement in invest-


igations of comparable scope and penetration.

Level VI: Depth Rationality
tha t clear paradigms exist in the
human sciences of I.S.1-4, most notably in
the traditions inaugurated by Marx and Freud
but also in some of the work of the theorist s of the 1 i r(~-world of social interaction,
is there a spnse in which the application of
these inference schemes, and hence of the
type of explanatory critique they presuppose,
is transcendentally necessary?

Now assume two interlocutors X and Y.

Suppose X believes himself to possess a
rational argumentative procedure RA, a reasoned argument Ar and a conclusion Q; but that
Y does not or cannot (perhaps ‘in spite of
himself’) accept or act upon RA, Ar or Q.

(The ‘reverse conditions may apply symmetrically to X, but we can ignore this complication here.) What is to be done when rational
argument fails? Clearly there are three
general kinds of possibility here:

(i) Y continues to mistakenly believe (and
act upon )-Q;
(ii) some non-discursive process (e.g.

force, medication) induces in Y a belief in
Q; or
(iii) X and Y jointly initiate an inquiry
into the conditions blocking or compelling
yts beliefs.

Adoption of solution (i), i.e. stoic
acceptance of irrationality, error, etc. is
a counsel of despair.

Moreover it cannot be
generalised to the first person case of
doubt (or more generally, choice) without
vicious axiological regress.

Solution (ii)
can be ruled out on the grounds that drugs,
force, etc. can only simulate the acceptance
of Ar or RA’

Further it is not emancipatory,
in that it does not replace an unwanted with
Gi Vl’Yl


a wanted source of determination, but merely
counteracts the effects of one unwanted
source of determination with another.

has the corollary that in as much as the
original source of determination is not
defused, it may continue to exercise a
latent power.

The alternative (iii) of a depth-investigation (D-I) is possible where reason fails
but has not yet exhausted its resources; and
it is practicable where yts beliefs are generated or underpinned by unreflected (unacknowledged) processes, and Y seeks to
understand, in order to undermine or abrogate, these processes.

A depth-investigation may be defined generally as any cooperative inquiry, which includes the agent,
into the structure of some presumed set of
mechanisms, constituting for that agent an
unwanted source of determination (which,
whether cognitive or not, will always possess some cognitive manifestation), with a
view to initiating, preserving or restoring
that agent’s ability to act and think

Four points must be immediately made about
this definition. First, what is rational
cannot be stipulated a priori, but must itself be discovered, in relation to antecedent
notions of rationality (its nominal essences,
so to speak), in the context of the explanatory critique such a depth-investigation presupposes.

Secondly, although the concept of
a depth-investigation has been introduced as
an ideographic practically-oriented application of some or other determinate explanatory-critique, the theory at the heart of the
critique itself depends crucially for its
own development and empirical confirmation
on such investigations (whether on living or
reconstructed, e.g. historical, materials).

It follows from this that the links between
theory and practice, and between pure and
applied research, though not abrogating
their distinctions, are bound to be tighter
than in the natural sciences.

corresponding to the different types of
inference scheme outlined above, there will
be different forms of depth-investigation.

These must not, however, by hypostatised.

For of course the explanation of cognitive
ills will in general involve reference to
practical and communicative ills, and vice
versa. Finally the desire for emancipation
which motivates the depth-investigation can
neither be posited a priori (for although it
is a necessary truth that people act on
their wants, it is not a necessary truth
that they act on their interests), nor predicted in historicist fashion on the basis of
some particular theory of individual development of history.

But as a socially-produced
social object, the desire for emancipation
will of course be a crucial topic for metainvestigations. And such investigations will
need to be continually reflexively incorporated into the substantive theory of the
practice etc. from or for which emancipation
is sought.

The structure of a simplified D-I may be
elucidated as follows:

(1) Y is not capable of ~; scientific
realism suggests there is a mechanism M
preventing this.

(2) General theory T investigates the

structure of blocking/compelling mechanisms,
under the control of empirical data and

(3) The application of T to Y investigates
the agent Y, as well as X. For it is Y’s
interpretations, actions and determinations
that are at issue.

Subjectivity in the
human sciences is not an obstacle; it is (an
essential part of) the datum.

But ontological authorship does not automatically
carry over into epistemological authority.

Now the Y-dependence of the D-I means that Y
must have a motive or interest in disengaging
M, or in a range of aspects that M prevents.

And that co-investigator X must not have an
interest in the distortion of M-descriptions.

Concretely, this raises the questions of the
costs of emancipation for Y and of the conditions under which emancipation may be a
second-best solution for Y; and for X it presupposes both the willingness to learn (in
the general spirit of Marx’s ‘Third Thesis
on Feuerbach’) and the continuing development of X’s own self-understanding. At a
deeper level, the success of the detailed investigation of the modus operandi of M in or
for X must depend upon an internal differentiation within the experience of X, so that
the empiricist/utilitarian notion of emancipation as a process of the alteration of the
circumstances of atomistic individuals must
be rejected.

Moreover it should be reiterated that cognitive emancipation will in general depend upon non-cognitive (and extradiscursive) conditions; and that cognitive
emancipation is necessary, but insufficient,
for full emancipation (as shown by the example of the slave who knows very well he is
a slave but still remains a slave, i.e.

unfree) .

In fact dissonance, not liberation, may
be the immediate result of enlightenment.

And such dissonance may lead either to ‘revolutionary-critical’ activity or to despair.

Moreover constraints upon cognitive emancipation itself are imposed by the pre-formation
of thought-contents (in psycho-analysis),
the projects of others (in social phenomenology) and the non-discursive aspects of
social reality (in historical materialism).

Hence emancipation cannot be conceived
either as an internal relationship within
thought (the idealist error) or as an external relationship of ‘educators’, ‘therapists’

or ‘intellectuals’ to the ‘educated’, ‘sick’

or ‘oppressed’ (the empiricist error).

Now I want to propose that the possibility of a depth-investigation is a transcendental condition for any science of man and
hence (at a remove) for any science at all;
and that in particular to inquire into the
nature of the real grounds for beliefs is
the same thing as to inquire into the possibility of rationalisation, self-deception,
deception of others, counterfinality and
systemic mystification; and that to inquire
into the conditions of possibility of these
cognitive-communicative malaises immediately
raises the question of the conditions of the
possibility of practical ones – from illhealth to brutal oppression. The issue of
the causes of belief and action, presupposing a distinction between real and possible
(including assumed or fancied) grounds, can
only be taken up the depth human sciences.

But a moment is reflection shows that this
distinction, and hence the possibility of a
depth-investigation at the analytic, phenomenological and historical levels, is a condition of every rational praxis or authentic
act of self-understanding at all.

It is
necessitated by the existential intransitivity and enabled by the causal interdependency of the phenomena of sociality. Thus in
the human sciences the problem of error
(oppression, etc.) must make way for the
problem of the causes of error (oppression,
etc.), as part of the programme, paramorphic
(but non-identical) to that of Kepler,
Galileo and Newton, of the investigation of
the underlying structures producing the
manifest phenomena of social life.

The object of the depth-investigation is

Emancipation may be conceived
either as the process of the changing of one
mode of determination D, into another D2, or
as the act of switching from Dl to D2, both
Dl and D2 perduring but Dl in an inactivated

Now if the emancipation is to be
of the human species, then the powers of
emancipated man must already exist (although
perhaps only as powers to acquire or develop
powers) in an unactualised state.

The key
questions for substantive theory then become:

what are the conditions for the actualisation of the powers?: are they stimulating
(of the socialist tradition); or releasing
(of the anarchic/liberal traditions)?; do
they lie in social organisation or individual attitudes etc? (35)

8. Conclusion
Can anything be said about the conditions of
the possibility of emancipatory practices in
general? I think that, for emancipation to
be possible, four general types of condition
must be satisfied.

First, reasons must be causes, or discourse is ontologically redundant (and scientifically inexplicable).

But the potentially
emancipatory discourse, given the TMSA and
the general conception of an open world, can
only co-determine action in an already prestructured, practical and collective context.

Second, values must be immanent (as latent
or partially manifested tendencies) in the
practices in which we engage, or normative
discourse is utopian or idle.

I think that
Marx, in conceiving socialism as anticipated
in the revolutionary practice of the proletariat, grasped this.

And it is on this
feature that Habermas’ deduction of speechconstitutive universals also turns (36).

But if there is a sense in which the ideal
community, founded on principles of truth,
freedom and justice, is already present as
an anticipation in every speech inter-action,
might one not be tempted to argue that equality, liberty and fraternity are present in
every transaction or material exchange; or
that respect and mutual recognition are contained in the most casual reciprocated
glance? (37).

It is an error to suppose
that ethics must have a linguistic foundation; just as it is an error to suppose that
it can be autonomous from science or history.

Third, critique must be internal to (and

conditioned by) its objects; or it will lack
both epistemic grounding and causal power.

But it follows from this that it is part of
the very process it describes, and so subject
to the same possibilities, of unreflected
determination and historical supercession,
it situates. Hence continuing self-reflexive auto-critique is the sine qua non of any
critical explanatory theory.

Finally, for emancipation to be possible,
knowable emergent laws must operate (38).

Such laws, which will of course be consistent with physical laws, will be set in the
context of explanatory theories elucidating
the structures of cognitive and non-cognitive oppression and the possibility of their
transformation by women and men. Emancipation depends upon the untruth of reductionist materialism and spiritualistic idealism
alike. On reductionism – if the physical
process level is Lp, and the level at which
emancipation is sought is Le’ then either Lp
completely determines Le and no qualitative
change is possible; or qualitative change is
possible, and the laws of Lp are violated.

On idealism – either emancipation is entirely intrinsic to thought, in which case it is
unconditioned and irrationality is inexplicable; or if it is conditioned, it cannot be
intrinsic to thought.

Emancipation depends
upon explanation depends upon emergence.

Given the phenomenon of emergence, an emancipatory politics or therapy depends upon a
realist science.

But, if and only if emergence is real, the development of both are
up to us.

The-Possibility of emancipation is not of
course the reason why an emergent powers
theory, if it is, is true.

It is rather
that if human beings, and social forms in
general, are emergent from but conditioned
by nature, then there is at least the possibility that the human sciences, provided
they ‘do not anticipate the new world dogmatically, but rather seek to find the new
world through criticism of the old’ (39),
could still be of some benefit to the
greater majority of mankind.


Cf. esp. N.R. Hanson, Patterns of Discovery, Cambridge
University Press, Cambridge 1958; R. Harr~, The Principles of
Scientific Thinking, Macmillan, London 1970; and M.B. Hesse,
The Structure of Scientific Inference, Macmillan, London 1974.

7 See The Possibility of Naturalism, p165.

8 See e.g. J. Habermas, Theory and Practice, Heinemann, London
1974, pp16ff.

9 Cf. A. Collier, ‘Materialism and Explanation’, Issues in Marxist Philosophy Vol.Il, ed. J. Mepham and D.H. Ruben, Harvester
Press, Brighton 1979, p37, and the unpublished essay by M.

West lake referred to there.

10 Cf. A. Giddens, New Rules of Sociological Method, Hutchinson,
London 1976, p121 & passim.

11 See The Possibility of Naturalism.

12 A. Giddens, Central Problems in Social Theory, Macmillan,
London 1979, p56.

13 See The Possibility of Naturalism, pp121-23.

14 See e.g. J. Elster, Logic and Society, Wiley, Chichester 1978,
Ch.5, and E. Ullman-Margalit, The Emergence of Norms, Oxford
University Press, Oxford 1977.

15 See A Realist Theory of Science, Ch.1.

16 Cf. e.g. C. Taylor, ‘Interpretation and the Sciences of Man’,
Review of Metaphysics 25(3), 1971 (reprinted in Critical
Sociology, ed. P. Connerton, Harmondsworth 1976).

17 J. Habermas, Knowledge and Human Interests, Heinemann, London
1972, p.vii.

18 G.W.F. Hegel, The Phenomenology of Mind, AlIen & Unwin, London

19 Cf. M. Polanyi, Personal Knowledge, Routledge and Kegan Paul,
London 1958, Ch.4.

20 See The Possibility of Naturalism, esp. Ch.4, Sections 3 and 4.

21 Cf. R. Bernstein, The Reconstruction of Social and Political
Theory, p203.

22 Cf. Knowledge and Human Interests, passim.

23. Cf. H.-G. Gadamer, Truth and Method, Sheed and Ward, London

24 Cf. C. Taylor, ‘Neutrality in Political Science’, Philosophy,
Politics and Society 3rd Series, ed.s P. Laslett and W.

Runciman, Blackwell, Oxford 1969 (reprinted in The Philosophy
of Social Explanation, ed. A. Ryan, Oxford University Press,
Oxford 1973).

25 If ‘ought’ implies ‘can’, the non-trivial implication of a
power is a presupposition, not an entailment, of the oughtstatement; which depends upon a theory (i.e. ‘factual knowledge’) of the agent and his circumstances.

26 See R. Hare, Freedom and Reason, Oxford University Press,
Oxford 1963, p108.

The title has been (in my opinion, unconvincingly) disputed – see e.g. A. Maclntyre, ‘Hume on “is” and
“ought”, Philosophical Review 1959, reprinted in A. Maclntyre,
Against the Self-Images of the Age, Duckworth, London 1971.

(The vexed passage in Hume is A Treatise of Human Nature, ed.

L.A. Selby-Bigge, Oxford University Press, Oxford 1965, Ill,
i.1, pp468-70.)
27 K. Marx, Capital Vol.I, Lawrence and Wishart, London 1965, p505
28 A. Giddens, New Rules of Sociological Method, p159.

29 See esp. R. Edgley, ‘Reason as Dialectic Radical Philosophy 15,
1976, and ‘Marx’s Revolutionary Science’, Issues in Marxist
Philosophy Vo1. III , eds. J. Mepham and D.H•. Ruben, Harvester
Press, Brighton 1979.

30 Cf. e.g. D. Bloor, Knowledge and Social Imagery, Routledge and
Kegan Paul, London 1962, Ch.2.

31 An explanatory critique in the natural sciences could be
represented as follows:









I would like to take this opportunity to acknowledge some

First, I have benefitted greatly from discussions with
Roy Edgley on this and related topics.

Secondly, I owe much
to the stimulus of the pioneering work of JUrgen Habermas in
this field, even where (as will be obvious} I come to rather
different conclusions. Thirdly, I am indebted to the writings
of Alasdair Maclntyre, who did perhaps more than anyone else in
the ‘analytical’ tradition to open up the possibility of a
historical treatment of moral and practical philosophies.

Finally, this present paper developed out of another ‘Emergence, Explanation and Emancipation’ presented at a conference
organised by Paul Secord under the auspices of the University
of Houston (and forthcoming in conce)tual Issues in the Human
Sciences, ed. P. Secord, Oxford 1981.

I am extremely grateful to the participants at that conference, and at the seminars
where I have read drafts of this paper, for their criticisms,
questions and comments; and in particular to William Outhwaite
for sharpening my thinking on the nature of an explanatory
critique of consciousness in natural science.

See my A Realist Theory of Science 1st ed. Leeds 1975, 2nd ed.

Harvester Press, Brighton and Humanities Press, New Jersey
1978, and my ‘Realism in the Natural Sciences’, Logic, Methodology and Philosophy of Science VI, ed. L.J. Cohen et aI,
North Holland, Amsterdam 1980.

See my The Possibility of Naturalism, Harvester Press, Brighton
and Humanities Press, New Jersey 1979.

Cf. e.g. M. Scriven, ‘Truisms as the Grounds for Historical
Explanation’, Theories of History, ed. P. Gardiner, Free Press,
New York 1959; and P. Achinstein, ‘Explanation’, American
Philosophical Quarterly Studies in the Philosophy of Science,
ed. N. Rescher, Blackwell, Oxford 1969.

See A Realist Theory of Science.

(i) T>P.

(ii) T exp I(PN} ~ (iii) -V(SS ~ I(P N }}
(iv) V9lss
See J. Austin, ‘Performative-Constative’, La Philosophie
Analytique 1962, trans. by G.J. Warnock in Philosophy and
Ordinary Language, ed. C. Caton, University of Illinois Press,
Urbana 1963 (see esp. p30).

J. Searle, ‘How to derive “ought” from “is”‘, Philosophical
Review 1964, and Speech Acts, Cambridge University Press,
Cambridge 1969, Ch.8.

See L. Colletti, ‘Marxism and the Dialectic’, New Left Review
93 (1975).

Is the present direction of argument necessarily incompatible
with a substantive utilitarian ethics? Yes and no.

The utilitarian tradition has generally been willing to concede that a
world with more possibilities is CP better than one with less,
so that supposing that a happy or healthy man could make himself miserable or ill (not that he would – in virtue of his
state – of course normally want to) but not vice versa,
Bentham and Mill would be bound, on this kind of ground alone,
to approve an emancipated as better than a non-emancipated

But could they approve the kind of measures substantive
depth theories indicate as necessary for the transition to such
states? I doubt it.

In Kantian terms, one could say that
although they might will the end, it is highly unlikly that
they could will the means in all but the most improbable

See J. Habermas, ‘Towards a Theory of Communicative Competence’, Inquiry 13 (1970).

Cf. R. Harr~, Social Being, Blackwell, Oxford 1979.

See my ‘Emergence, Explanation and Emancipation’, Conceptual
Issues in the Human Sciences, ed. P. Secord, Oxford 1981.

See Writings of the Young Marx on Philosophy and Society,
ed. Easton and Guddat, Anchor Books, New York 1967, p212.



Buy the newest RP in printDownload the PDF