Quick notes on: Archer, M., Bhaskar, R., Collier, A., Lawson, T. & Norrie, A. (eds.) (1998) Critical Realism: Essential readings. London: Routledge.

Dave Harris

General Introduction ( Bhaskar)

Bhaskar kicked it all off.  The term critical like transcendental suggest affinities with Kant, but realism shows the differences.  The argument went through a definite sequence as indicated in the essays below, with a later emphasis on the dialectic and then the totalizing and transformative aspects of the philosophy.

Positivism had come under attack especially in its views that natural laws simply reproduced facts in thought.  It featured a 'monistic theory of scientific development and a deductivist theory of scientific structure' (x).  Popper and his associates argued for falsifiability and the role of revolutionary breakthrough, Kuhn and others indicated the real social processes involved, Wittgenstein and others argued that facts and science were not atomistic and were theory dependent.  However, the issue of the reality of being, compared to the relativity of knowledge, remained.  This arose with Kuhn and Feyerabend implying that meaning might not be shared at all between one theory and the next, that they might be incommensurable, that there was no rational choice, and perhaps, no 'theory independent world'.

It is different if theories merely conflict, because this presupposes that they are alternative accounts of the same world, and that there is therefore rational theory choice.  This would then reconcile ontological realism, epistemological relativism, and rationality, and this is what critical realism claims.

Structure was to be deduced, but conventional ways of understanding it were insufficient.  Bhaskar was to argue that positivism in particular could not explain the necessity nor the universality, and not the transfactuality of laws [the logical problem with induction?].  He argued for an ontology that was not reduced to epistemology, and that also did not collapse 'the real, the actual and the empirical' (xi), which implied that it was stratified, emergent and differentiated, that it had ontological depth resulting in 'intransitivity, transfactuality and stratification'. 

Popper and Hempel argued for deductivism as explanation, seeing empirical regularities as deducible from universal laws [this is not my reading of Popper I must say].  However, this still confuses empirical regularities with the law like ones.  A genuine explanation would involve 'the introduction of new concepts not already contained in the explanandum', generative mechanisms and models.  However, these might operate at deeper levels of reality - they could come to be seen as real, assuming that we could actually get to examine them, perhaps through causes.  We therefore have a necessary 'vertical or theoretical realism'.  Science could be seen as moving from the manifest phenomena through modeling and experimentation to the identification of 'generative causes' which then have to be studied in turn.  This is a move towards a deeper knowledge 'of natural necessity a posteriori' (xii).

However, a horizontal realism, or transfactual one, is also necessary if we want to explain the universality of these generative mechanisms, in the form of scientific laws describing transfactuals rather than empirical generalities: this is what scientific work does as well and it explains and diagnoses.  The trick is to identify universals which are necessary, and not just empirical.  We have to see laws at work as 'powers, or more precisely tendencies, of underlying generative mechanisms which may be'(xii) working horizontally, appearing as actualized or unactualized possibilities [compossibilities presumably?], or vertically producing 'an ongoing irreducibly empirical open-ended process of scientific development' [irreducibly empirical in the sense of having to be referred ultimately to empirical experiment or observation?  Not theological or speculative?].  We study structures as well as events, open systems as well as closed [actualized, here described as differentiated].

Up until now, philosophers have committed 'the "epistemic fallacy"', with their views of the empirical world.  Science is social, but the mechanisms it identifies are independent of their discovery '(existential intransitivity)', and we must separate the transitive from the intransitive if we are not to reify science.  Also, the laws of nature are independent of scientific systems and the extent to which they are closed (transfactuality): not understanding this leads to 'the fallacy of actualism'[the objective myth for Deleuze].  Open systems are ubiquitous and must be grasped through various scientific procedures including experimentation.  Laws are only universal within their range, but are not actual.  'Constant conjunctions are produced, not found'.  Theoretical explanations explain these laws 'in terms of the structures which account for or perhaps merely ground them': practical explanations explore transfactuality.  Both nature and science must be stratified, and this stratification is detectable in any actual science.  Once we recognise this, we can solve problems like the logical difficulties of induction, which can now be seen to work if we accept actualization [there is increasingly a notion of the necessity of reason here as well, so that if water boils when it's heated, 'then it must do so'].  Natural mechanisms multiply [multiplicities!] and this produces 'the real plurality of sciences which study them': these multiple mechanisms are not reducible to each other even though they might be linked in explanations, so that we have 'emergence', 'so that the course of nature is different than it would have been if the more basic stratum alone operated'(xiii), and this has implications for the social sciences.

Human sciences are traditionally 'dominated by dichotomies and dualisms, and critical naturalism attempts to transcend them.  One main split was between positivism and hermeneutic, and the transcendence involves 'a qualified critical naturalism'.  There is also a split between individualism and collectivism, to be resolved by seeing society 'relationally and as emergent'.  The split between voluntarism and reification could be transcended by a 'transformational model of social activity'; the fact value split would be transcended by the theory of explanatory critique; the dichotomy between reasons and causes was replaced by 'the critical realist conception of causality' (xiv); the mind body split was to be overcome by seeing mind as emergent from matter.

The first to these questions was addressed in The Possibility of Naturalism, contrasting uncritical naturalism with anti naturalism as above, allowing for those who wanted to synthesise positivist and hermeneutical principles, including Weber and Habermas, and dualists like Gadamer or Winch who denied any relevance to positivism, and the more hybrid forms like post structuralism.  Critics share the same view of science as positivists, but another possibility arose that there might be a qualified, critical, non reductionist naturalism, involving transcendental realism and allowing the social to become specific and emergent.  Winch can be critiqued for pursuing a positivist ontology.  So can strong calls for deterministic accounts since conjunctions are not necessary or sufficient for transcendental realism, and human beings still need to discover intelligible connections.  The real is not just what can be grasped by positivist science.  The difficulty positivists have in explaining social phenomena shows this, while hermeneuticists are also wrong in claiming distinctiveness for social phenomena - social and natural phenomena both display particular conjunctions, and these can be studied empirically in both domains.  Social science in particular can begin with the pre-interpretations of social life, but these do not exhaust the possibilities nor do they prevent corrigible, perhaps even causal explanation.

Reasons may not be the same as causes, but they do not offer strong explanations - they are not logically independent of the behaviour they explain, and they are not 'causally efficacious in producing one rather than another sequence of bodily movements' (xv), so it is hard to see any grounds for preferring one set of reasons as explanations to another.  However, critical naturalism offers the possibility of an independent analysis of social objects, neither invoking subjective meaning and intention or social facts.  Instead, society preexists individual agency and is thus '(transcendentally and causally) necessary' (xvi), but societies only exist and persist via human agency - society is both a condition and outcome of human agency, even though for actual individuals, the social world exists first.  Although this looks like Giddens on structuration, this is the difference [for Archer] - structural constraint and possibility always affects agency, so even  unintended consequences are inexorable.

The accounts of actors themselves are corrigible and limited by these constraints and unacknowledged conditions and outcomes.  They must always form the starting point for inquiry, since social action is transformational, 'recursive and non teleological'.  Social action must always be relational not individualistic or collectivist.  When we do sociological abstraction, we should not just offer empirical abstractions but ground these in the ontological depth of nature and society, attempting to grasp generative mechanisms and causal structures.  Marx can be rescued here, but all the founding fathers combined some aspects of such realism.

Human agency is limited by the nature of ontology, since social structures that result from action are dependent on concepts, activities and 'greater space - time specificity' (xvii).  Social systems should be seen as intrinsically open, 'the most important epistemological limit', but there is also a a 'relational limit' affecting the notion of causal interdependency between social science and its subject matter [I am not at all sure what this means is, unless it is the argument that there are no causals entirely external to it]. The epistemological limit means that there can be no simple decisive tests and we must use instead explanatory criteria as the basis for rational choice, but at least this offers a way forward from critical realist conceptions.  When we offer theoretical explanation, we are describing significant features, tracing them to possible causes, eliminating alternatives, and identifying the generative mechanism at work, which then requires further explanation.  We proceed by resolving complex events and components, theoretically describe these components, and then try and trace these components back to possible antecedents, and eliminate alternatives.  In this way, social sciences can be sciences exactly as in the natural ones.  If social sciences tend more towards the philosophical, this just reflects that these forms of arguments are simply 'retroductive' options complementing those of natural science.  [Which also lets in arts of course].

Conceiving of the subject matter of social science as social objects together with beliefs about them, leads to an explanatory critique which is different from the natural sciences, a form of 'ethical naturalism' bridging facts and values.  It starts with a critique of Hume that moving from factual to evaluative statements is 'logically inadmissible' (xviii).  This is not to deny that factual and evaluative statements can causally influence each other.  It is clear that much social scientific discourse is value impregnated, and so is social reality, so that is is impossible to describe social situations without pursuing value implications.  However it might still be possible to oppose the values implied, however functional or likely to lead to flourishing social practices are claimed to be. 

We can make progress by refusing to 'detotalize or extrude (e.g. by hypostatisation) social beliefs from the societies in which they are found'.  Such beliefs might be contradictory or even false, and factual social science can show this - this will then lead to an evaluation that says they should be rejected.  If we accept that beliefs can be causally explained, and provide a true account of causes, we might immediately proceed to evaluation.  This is explanatory critique, and 'we may be able to discover values, where beliefs prove to be incompatible with their own true explanation'[presumably if we can show that beliefs are ideological and not really in someone's interest?].  Of course this assumes the truth is itself a good, but this is implied by any factual discourse.  It is true that truth may not be the only social good and that there may be other considerations, but 'other things being equal truth is good and falsity is ill' (xix).  Showing that something should be done to remove false beliefs is not the same as saying what exactly should be done, leaving room for some 'insider, shared, tacit, "movement based" knowledge' as well as theory.  Of course we are assuming that all other things are equal, but this is exactly what happens in scientific discourse - the same clause helps us move from fact to fact and from fact to value, and other fundamental parallel between social and natural science.

A dialectical phase of critical realism ensued, aimed at enriching critical realism, and developing a general theory of dialectic, incorporating hegelian dialectic.  A totalizing critique of western philosophy was on the cards.  It is crucial that 'determinate absence' be included, even though Hegel did not do so.  Dialectics then becomes 'the absenting of constraints' on more mundane absences.  Dialectical critical realism itself was a dialectic system, with a first moment of non identity [between levels of reality as in transcendental realism]; a second 'dialectical edge'[some sort of valorization of absence and negativity]; a third level of totality or 'holistic causality'; the fourth dimension of 'transformative practice, the unity of theory and practice in practice' (xix--xx).  Although hegelian dialectic refers to identity, negativity and totality, Hegel could not really describe negativity, and all the totalities were closed [sounds like Deleuze again].  The moral good is implicit in every action or remark that leads to truths, so that 'moral realism'becomes the same as ethical naturalism, and explanatory critique produces a 'radical emancipatory axiology'(xx).  Again this is not the same as actually existing morality.

There is a clear link with a reassessment of Marx as a scientific realist, at least in Capital - essential relations are different from and sometimes out of phase with, or possibly even opposed to the phenomenal forms, but this was never properly theorised in the form of scientific realism.  He also failed to generate a thorough enough critique of empiricism or objectivity, at least in its intransitive dimension [labour produces objects in the transitive dimension].  Nor did he consider the geo- historic dimension [ontology, structural causality] within the epistemically relative, and the critique of political economy became a substitute for a proper research programme of materialism.  This explains the hegelian residues in Marxist thought, and some of the ambivalences and contradictory tendencies, especially the contradiction between idealism [the role of labour] and crude materialism.

However, we can detect trends of a more promising position in his criticisms of Hegel, the inversion of subjects and predicates, and Hegel's '"ontological monovalence"', 'a purely positive account of being' without absences [which explains everything] there was also the principle of identity, where being is equated with thought, and logical mysticism.  Again, there was no systematic attempt to regard nature as intransitive or to explore the geohistorical processes producing social forms [not just a realisation of thought as in monovalence].  There is however a rational kernel.

We can reconstruct marxian dialectic in more scientific terms, seeing the contradictions in thought and crises in socioeconomic life as the result of contradictory essential relations which generate them; as historical, rooted in these relationships and circumstances; as critical because it shows the historical conditions of liberty and the limits of theories; and systematic, trying to trace the features of capitalism back to 'certain existentially constitutive features of its mode of production'(xxi).  The most important feature was the contradiction between use value and exchange value, and between concrete and abstract labour.  These oppositions are seen as real and as inclusive [in the sense that they presuppose each other], and as related to 'a mystifying form of appearance'.  There are no logical contradictions here, because they can be 'consistently described'.  They are not scientifically invalid if we except the notion of 'a non empiricist, stratified ontology in which thought is included within reality'. 

In Hegel, the rational kernel is 'an epistemological learning process' whereby inconsistencies are removed by referring them to a more general totality.  This is done either by bringing out what is implicit in some notion, or correcting some inadequacy in it.  In this way, any absence or incompleteness is seen as 'an inconsistency which is remedied by resort to a greater totality'.  This is the same as the epistemology of the logic of scientific discovery.  However, the mystical shell is its 'ontological monovalence', with no notion of determinate absence or uncancelled contradiction, no open totality or transformation through praxis.

This explains the emphasis on absences, as arguments, change, or aspirations to freedom.  Here mistakes and constraints'have to be positively identified and eliminated, but absence is ontologically prior to positive being' [multiplicities precede actualities?].  Absences refers to processes as well as products. Analytical reason is a part of the overall process of dialectical reason which goes on to critique the fixity of the subject and the traditional notion of subjects and predicates, leading to a form of identity thinking - again which misses out absence [almost exactly the same as Deleuze emphasizing difference which is not thoroughly explained by repetition?]

To spell out the moments of the system:

The first stage requires non identity relations and the rejection of epistemic and other fallacies, of identity theory and of  'actualism' [objectivism], and alterity is the key concept.  This leads to scientific intransitivity, and 'referential detachment' (xxii) [dividing the referent from that to which it refers]. Reality and a necessary ontology is required, focusing on the 'transcendentally necessary stratification and differentiation of the world', leading to a notion of cause , powerful generative mechanisms and transfactuality, even 'natural necessity'. There is also 'alethic truth', referring to things rather than propositions, and this provides the basis for the dialectic between things, and depends on ontological stratification.  Discovering this truth overcomes problems of induction which operates only with the actualised and the destratified.

The second stage emphasises the category of absence, aimed at ontological monovalence, and both Plato's and Kant's attempts to domesticate the negative.  It implies genuine negativity, contradiction and critique.  It sees causality, space and time as unified in producing 'tensed "rhythmic" spatialising process'[phase transitions for Delanda], connecting the past with the present, and emphasizing constitutive processes.  Contradictions can be both internal and external, logical and dialectical: the latter involve internally related oppositions which are capable of change, process, oppositions including reversals.

The third stage is based on the category of totality, to include thought and its qualities such as 'reflexivity, emergence, transcendence, constellationality, holistic causality, concrete universality and singularity, internal relationality and intra activity, but also detotalization, alienation, split…  illicit fusion and fissure' (xxiii).  The dialectic is between centre and periphery, form and content, figure and ground, generation and alienation, 'retotalization in a unity - in - diversity'.

The fourth stage involves transformative praxis or agency, which is implicit in the other three moments when we look at human social relations, especially if we see agency as '(intentional) causality, which is absenting'.  The justification for agency is found in emergent powers and in the different dimensions of social being - 'material transactions of nature, interpersonal relations, social structures and the stratification of the personality', and these are 'dialectically interdependent'.  There is a moral openness to the human species, where ideological and material struggles are also ruled by 'absolute reason', embedded in theory guided practices [which looks like the desire for emancipation and freedom].

Perhaps we should present these arguments by starting with the concept of absence say as in desire. [so -- a pedagogy!]  This will allow immediately lead to referential detachment, existentially intransitivity, and therefore ontology.  Then we can proceed to classification and causality, alethic truth, and transfactuality.  The importance of negation appears in the form of all the categories in the second stage, and space time rhythms.  Contradictions lead to an understanding of emergence and from there to all the categories in the third stage.  When considering the fourth stage, we can derive from the notion of activity dependence of social structures a transformational model of social activity and the factors that limit it [we can even remind ourselves that facts are not values here, it seems].  We experience absences in constraints which we encounter when we act [maybe, xxiv].  We then see all similar constraints as absences, and consider how we might consider them 'to incorporate flourishing and potentials for development, and the negative generalisation of constraint to include details and remediable absences generally.  All this will produce 'totalizing depth praxis', which is 'already implicit in the most elemental desire'. [comnpare this pedagogial structure with the one in Huckle].

Science is a social product, but this should draw attention to the work that is required to produce it.  Science is also about acquiring knowledge of things which are not socially produced and which do not depend on human activity when they work -things attracted by gravity, or sound travelling.  We can use terms like transitive and intransitive to refer to these objects of knowledge.  Here transitive objects are those that have been 'fashioned into items of knowledge'(16) by scientists, facts and theories paradigms and models, techniques of enquiry and so on.  Something like Darwin's theory of natural selection would not have been possible without them [reminds me of Althusser's points that the constituents of Marxism were also bourgeois political economy and early socialism], but it also included intransitive objects -'the mechanism of natural selection' (17).

It is easy to imagine a world without science [an assertion that leads him to reject some epistemologies below].  Chemical and physical processes will continue, regardless of our knowledge.  We can know about them, but they are intransitive 'science - independent'.  However, we cannot imagine science without transitive objects, including various models of natural processes - pumps or idealised households, atoms is little billiard balls.  If science requires intransitive objects, this leads to philosophy, and ontology especially in the form of the'transcendental question "what must the world be like for science to be possible?"'(18).  It is not just an abstract ontology, because science has already gained knowledge of intransitive objects: if we investigates intelligibility we will come to necessary concepts about reality.  Any adequate  philosophical account of science must take into consideration both its social character and the independence of its objects: the social character must be seen as a productive force, so that knowledge is not just spontaneously achieved; the independence of objects must lead to an essential realism 'the independent existence and activity of causal structures and things'.

Earlier traditions can now be examined and criticised. Much modern philosophy is not located neatly in these traditions:

Classical empiricism as in Hume.  Atomistic events are seen as given facts, and their conjunctions are to be studied.  Knowledge and the world become isomorphic, so science is almost an automatic response to the facts and their conjunction.  Realism is to do with the perception or with universals, exclusively so with empiricism.  Empiricism can explain neither the social character of science nor the independence of its objects, but turns to 'solipsism [individualism anyway] and phenomenalism' (20)

Transcendental idealism as in Kant.  Science becomes a matter of deriving certain models and ideals of nature, and these are 'artificial constructs'associated with human activity in general.  Natural necessity appears as constant conjunctions of events.  The natural world is a construction of the human mind, or the scientific community in modern versions.  Objectivity becomes a matter of intersubjectivity, which addresses the social nature of science against empiricism, but it has problems with the intransitives, which either do not exist at all, or if they do, are unknowable.  Cognitive activity imposes an order on nature.

Transcendental realism.  Knowledge tries to gain understanding of the structures and mechanisms that produce phenomena, and is produced by social activity in science.  The structures and mechanisms are real, neither surface phenomena nor human constructs, not events [but transfactuals] , and intransitive.  Causal laws do not depend on constant conjunctions of events, but themselves arise from real structures and their differentiation.  The order of nature exist independently of human thought, although thought expresses it, and this is integral to science, and based on its intelligibility, not on some prior 'dogmatic metaphysical belief' (21)

Empiricism and transcendental idealism share an ontology, believing in an empirical world ['empirical realism'(21)].  However, this depends on experience defining the world, and on epistemology dominating or replacing ontology.  They also assume that our understanding the world is an essential property of it.  Instead, we should see connections with our experience as 'an accidental property of some things, albeit one which can, in special circumstances, be of great significance for science' (21).  However, the real and the empirical are not coextensive: otherwise, we would have to accept idealist positions that statements about the world really expressed propositions 'about the way men understand it', with the structures of reality somehow becoming a function of human needs.  This does not sufficiently grasp the rationality of science and the way in which it is 'continually extended and corrected'(22).

Ontological questions and the philosophy of science cannot be dismissed, and an ontology is always assumed, including views that empirical events somehow reflect what the world is actually like, that conjunctions between the world and sense experience are necessary.  This would explain why the world is such that science can understand it, but the issue of necessity is different for transcendental realists - its structures can be investigated by substantive science.  Those structures are themselves such that the categories of transcendental realism emerge, such as the difference between structures and events, or open and closed systems.  These categories in turn are 'presupposed' by experimental activity.  Ontology should always bear in mind the account of science, but it is not that science clarifies or produces being, rather the reverse - it is 'the order of the world that, under certain determinate conditions, makes possible the cluster of activities we call "science"' (23).  However, science is not the only way to understand the nature of the world, because the structure of the world is not determined by science - to argue that propositions of science constitute ontology is the '"epistemic fallacy"'.

Similar arguments are found if we examine the category of experience.  It is inadequate as an account of science, although it occurs.  However, coming to make it intelligible presupposes intransitive structures which cannot be accessed by experience alone.  We see this with experimental activity, where human beings act as causal agents as well as perceivers.  The usual notion of perception implies an intransitive object, for example, which must be independent of experiences, depending on some distinct objects which can be experienced differently  The need for a scientific training and for changes in scientific theories shows this as well.  In practice, processes and things can be seen as the 'primary objects of perception' anyway, which are then used to reconstruct events and states [dynamic processes rather than static events - argued below].  Events could easily take place in the world without being experienced, rephrased as 'actualities unperceived' (24): some could be unperceivable [dark matter?].  Events which are not experienced remain possible, however, and meaningful statements [predictions?] can be made about them - existing knowledge about the world is not the same as a philosophical conception of it.  The history of science reveals these changing relations between objects and experience.

Turning to the empirical activity itself, this presupposes intransitivity and the structure of nature of objects investigated.  We can see this with causal laws, usually seen as a constant conjunction of events, but this conjunction is actually produced by experimental conditions [it appears in experimental conditions] implying that the experimenters themselves are part of the causal law.  An ontology is really being implied, especially when we argue that causal laws operate outside of experimental situations as well.  But this in turn suggests that constant empirical conjunctions are not necessary or sufficient for causal laws, which operate in a much more open way, without necessary regular sequences or constant conjunctions.  It is experiments that establish closed systems [except for astronomy, he argues].  Many philosophers and scientists have acknowledged that active interference is necessary if experiments are to work, but they have not seen that this implies an ontological distinction between empirical regularity and the causal operations outside.  This is normally taken as intuitively acceptable, as when experiments often fail, but without leading to the rejection of causal laws, which are beyond our powers to affect.  'Thus the intelligibility of experimental activity presupposes the categorical independence of the causal laws discovered from the patterns of events produced' (26).  We can now also see a rational explanation for unpredictable phenomena occurring -- they demonstrate open systems.

Bhaskar goes so far as to say that without human interventions, very few constant conjunctions would occur, so we would never get from experience to causal laws as empiricists assume.  Causal laws would continue to operate nevertheless.  This human intervention is often disguised in accounts of science - there is 'a concealed anthropocentricity'.  Events take place beyond experience, and causal laws beyond empirical conjunctions or sequences, operating in open systems.  Experience is only important in science if the perceiver is 'theoretically informed' and if the system can be closed - then experimental scientists 'can access underlying causal structures'.  This takes 'enormous effort - in experimental design and scientific training'.

If there are causal laws outside what can be grasped, for example by experiment, there must be an intransitive and structured world beyond human beings, a non empirical world, one which does not correspond neatly to the empirical [nor to idealist categories].  [Here, 'structured' means not just empirically patterned events].

It follows that philosophical ontology is possible, it proposes that causal links are distinct from events and experiences, yet nevertheless it is consistent with the practice of science.  Ontology should not be seen as describing 'a world apart from that investigated by science', however (27).  There is no 'mysterious underlying physical realm' as in Leibniz, and metaphysics should remain 'a conceptual science' focused on what makes science possible.  The old metaphysics commits the epistemic fallacy, where the intention is to produce statements about knowledge which will then produce statements about being, that ontology is really epistemology.  This displaces the notion of an independent world which can be investigated by science.  There are no 'transcendent entities'(28) [so a major disagreement with Deleuze here].  A general commitment like this means we cannot trace specific epistemological questions about, say, science and experience.  Causal laws can not be reduced to statements about knowledge of them, since they would operate even if they were not known.

Empirical realism shows the epistemic fallacy best.  It is found in Kant's view that the categories of thought have meaning only when applied to objects of experience in the world of sense.  It is found in logical positivism which argues the propositions are meaningless if they cannot be verified or falsified empirically.  Verificationism offers a particular form in confusing the meaning of propositions of reality with the grounds for holding that proposition, 'which may or may not be empirical'(28).  Instead, we must distinguish between natural and logical necessity and between 'natural and epistemic possibility'.  We need no longer assume that nature is ordered and structured in the same way as knowledge.  We can allow that experience is important in epistemology, without presupposing its significance ontologically.  Seeing science as a continuing process of discovery means that epistemic certainty is not relevant.  The tendency to take the conditions of knowledge as an implicit concept of the world lies underneath the problem of induction, which only becomes important if we take as central 'the necessity for an epistemically certain base'(29).  There is an idealist version of the epistemic fallacy as well.  In transcendental realism, it is not necessary for the existence of the world that science occurs, but it is necessary the other way around, that the world exists of a particular type so that science can occur, so scientific knowledge is not essential, nor a defining characteristic of the world.  Instead, 'on a cosmic scale, it is an historical accident', but a lucky one.

It might be argued that to do any kind of ontology, we need to do more general epistemology, about knowledge in general.  But this assumes that philosophical clarification of knowledge will lead to science.  We should start instead by working on 'what must be the case for science to be possible'[so science is taken as some crucial or defining part of knowledge].  It is true that scientific research sometimes has the same logical character as other activities, including 'criminal detection', but this means it does not just confine itself to what we can know to exist.  Variables have other values outside what simply exists, nor is it a simple process to identify and measure things as they exist - in this sense, 'knowledge follows existence, in logic and in time'.

Those positions that think they have denied any possibility of ontology are simply working with an implicit ontology 'and an implicit realism'.  For empiricists, experience produces an ontology and a realism based on its characteristics.  The assumptions it works with shows that it aims at 'incorrigible foundations of knowledge', and a methodology that is consistent with epistemology, but which is 'of no relevance to science', or is at least out of phase with science.  Hume's ontology, built on the value of impressions as opposed to essences, developed into a notion of the empirical world which has been largely accepted since.  The central issue, whether knowledge is exhausted by a knowledge of atomistic events, has been addressed by philosophers asserting the need for some theoretical component as well: but they have not criticised the central argument that impressions lead to knowledge.  Knowledge is still reduced to knowledge of atomistic events, and this is then identified as the principle component of knowledge of the world.  This fortunate 'isomorphic correspondence' (30) between knowledge and reality has been rarely criticised in subsequent philosophy, leading to 'a continuing "ontological tension"'(31) between subsequent philosophiesof science and this inherited ontology.  This has resulted in insoluble problems, such as how experiences might be generalised, or whether the necessity of events comes from the world or from the need to uphold philosophical rules.

For transcendental realism, adequate methodology 'must be consistent with the realist practice of science', not with the assumptions of empirical realism.  Hume actually developed a realist method which contradicted his epistemology allowing understanding to correct the senses for example, and similar hints are found in Kant.  An implicit ontology resulted from not really resolving this problem, entering a vacuum left by a proper discussion of ontology.  Facts are really social products, not just the particulars of the world, and any conjunctions are 'doubly social products', since these facts are then grasped as conjunctions.  This does have the benefit of permitting normal science [derided as conservative here] , and merely descriptive or instrumentalist accounts of theory, leaving no room for proper ontological criticism.  Transcendental realism does permit philosophical criticism [because reality is not grasped fully by science].  Ironically, 'to be a fallibilist about knowledge, it is necessary to be a realist about things.  Conversely, to be a sceptic about things is to be a dogmatist about knowledge' (32).

The conditions of science were also responsible for this attention, because there was a great deal of scientific consolidation going on when Hume was writing, and philosophy seem to be limited just to justifying scientific knowledge not showing how it was produced [still the default position today?].  Abandoning ontology left science uncriticised.  Restoring it makes it possible to criticise some approaches ['say social psychology'] as not doing science, even though it might be possible.  When this certainty collapsed, the reaction that set in only split the creations of the mind, and argued for '"poetic intuition"' (33).  This shows the anthropocentric nature of philosophy, turning questions about the nature of the world into questions about the nature of human beings, where the conception of science became a 'flat surface' dominated by human needs, not a 'multi dimensional structure independent of man'[so we have multiple dimensions even if not a distinct transcendental reality?].  The human mind or the scientific community came to be seen as determinate, and philosophy seemed to have no practical relevance for the critical activities of science.

How do causal laws have a real basis independent of events?  We must move away from the anthropocentric, and address the issue of necessity.  Empiricism and idealism agree that there must be some generative mechanisms at work in necessity, but idealists see it as imposed by human beings as 'an irreducible figment of the imagination'.  Transcendental realists see these mechanisms as real, emerging from the on going activity of science, and this is necessary to account for 'the rationality of theory construction' (34), which depends on external factors to produce change.  These generative mechanisms generate events, but persist even when they do not do so, and even when some possibilities are unrealised because of 'intervening mechanisms or countervailing causes'.  Experimenters try and eliminate interference to study the mechanism in its active phase, but this is possible only if the system can be closed, in the artificial situations of the laboratory - but they exist outside the laboratory and can still be used to explain phenomena.  They must do so if they are to be known and if they are to be seen as necessary, outside any rules that humans might generate.  Real connections that produce empirical ones can be seen understood as necessary.

It is mechanisms which we should study not events, actual states and happenings even if they are not actually manifest or empirically identified.  We get to know about them not just with experimental situations but by blending 'intellectual, practico- technical and perceptual skills' (35).  What results is neither an artificial construct nor an essence.  [Bhaskar argues that the notion of a world without men can be justified and made intelligible, 35, pointing again to the social activity of science as presupposing it - such activity is a useful counter to the passive images of the world held by philosophers which are often anthropocentric].

Baskhar argues it is valid to infer that generative mechanisms exist from his analysis of experimental activity and its transcendental elements.  Again, it is the intelligibility of scientific experiment that justifies this assumption of mechanism, including the unpredictable elements that we can explain in terms of an open system.  The claim that we have closed and limited mechanisms in experimentation is important because it assumes such open and unpredictable systems.  It is different with concepts of causal laws, which are too independent of patterns of events [maybe, 36].  Causal laws involve causal agents - the generative mechanisms.  We have explained everything in terms of 'things', some of them endowed with causal powers - 'it cannot be wrong to reify things'!' (37). Generative mechanisms simply express 'a way of acting of a thing'.  (38).

It is better to think of causal laws not as powers but as tendencies, because this would then explain activities that are not necessarily being realized or made manifest.  Again this is presupposed by the discussion of the intelligibility of experimental activity. It follows that we should make law like statements to include conditionals, specifying possibilities beyond what is made manifest, making statements that are 'normic, rather than subjunctive', specifying what is happening even if it might not be manifest, rather than saying what would happen.  These involve transfactuals -'things are really going on irrespective of the actual outcome' (37).  Invoking a causal always means invoking a normic statement.  Laws are not just empirical statements, nor just about events, instead they describe the ways of acting of transfactual things.

Most things in the world are complex objects with 'an ensemble of tendencies, liabilities and powers'[compare with Deleuzian haecceities]. If we can understand this ensemble, we can explain phenomena, and explanations need to be referred back to this 'essential nature of things' (38).  Prediction in science is only a derivative, and rarely possible - if it takes place, it's useful only for explaining the essential nature of things.

So the objects of scientific investigation are structured and intransitive things, producing mechanisms and causal structures independently of human beings.  This is not just an activity for metaphysics, because 'it is the end to which all the empirical efforts of science are directed'.  Ontology is not about the mysterious realm beyond the present one, but 'as providing a set of conditionally necessary truths about the ordinary world as investigated by science'.  Philosophy can suggest that real things exist, but not account for what they actually do to produce reality - that requires science.  Philosophy analyses the concepts of science.

So experimental scientists trigger the mechanism to ensure that it is active, and then try to study it without interference.  Both of these involve changing the course of nature, however.  Experiment does help us establish relationships between antecedents and consequents, in an experimentally closed system, but achieving things like variations of antecedent requires 'substantial theoretical labour in itself' (38).  The practical dimensions of experiment are crucial.  This alone suggests that there is some ontological distinction between experimental sequences and causal laws.  This explains the importance of specific experience in science.  The analysis also agrees that human beings are causal agents, but that experiments are a significant aspect of science.  Theory is also important of course because it clarifies the questions to be settled by experiments and also interprets the reply.  The development of scientific instruments has been crucial in testing and extending theory [balances, clocks and telescopes for example].

However, thought is also as important as experiment, but this raises the rationalist case that sufficiently strong axioms could help us deduce all the laws of nature.  Some sciences have proceeded even though experiment is impossible [Bhaskar seems to be working up to arguing for surrogates for experiment].

Mechanisms, or events and experiences operates in 'three overlapping domains of reality, viz. the domains of the real, the actual and the empirical' (41) [these are the multiple dimensions of reality?  None of them are virtual or transcendental?].  Empiricism collapses these three into one.  This prevents creative thinking about what experiments reveal in terms of the harmony or phase between the different levels [note that the experiences operate in all three domains, presumably meaning special scientific experiences or special philosophical experiences].  Science as a social activity unites these levels of reality, and experiences and the facts they ground are social products.  [Amending the notion of anthropocentricity] empirical realism is also 'an implicit sociology'. Empiricism neglects the hard work required to do science, to transform things so they can be added to scientific knowledge, especially in experimental activity.

We can do transcendental deduction with sociology as well to get at the notion of the sort of society that makes science possible. There must be a social character to science as well and this has  also been ignored..  We need a sociological dimension to explain intelligible scientific change, to revitalise the transitive dimension.  The concepts 'empirical' and 'sense experience' belong to this transitive dimension, even though experiences can be epistemically crucial to science.  Only transcendental realism explains this adequately, since experience, organised in the form of an experiment, gives access to 'the enduring and active structures, normally hidden'(43){ that is, if we accept the ontology, science would be doing this]. Scientific experience is as real as the structures it accesses, and scientific objects are not mere artefacts nor necessarily something illusory.  The relationship between the levels of the world is 'natural generation, not an interpretation of man', or relation between two kinds of real objects.  Levels of analysis do not 'correspond', but are linked causally [the example is the relationship between electrons and dining room tables], and every day objects can be explained in terms of the very small or the very large, but this is no different from explaining chemical reactions in terms of atomic structure.

Transcendental realist laws are independent of human beings, although our knowledge of them is not.  This position allows for the fact that there may be unknowable laws, or those that produce instances that are not perceivable, or are not manifest at present  to human beings.  However, social activity constantly extends science, and there can be no a priori limits.  Problems in the physical exploration of materials does not show the failure of science, but the failure instead to extend inquiry into matters that cannot be sensed [such as atoms].  In the case of Locke [seen here as an empiricist], there was no attempt to see the science of the day as only a particular state of science, and an error of basing a philosophy upon current science [almost a theory of ideology] [and various obstacles to progress in science are discussed].  Philosophers still tend to be 'prisoners of the scientific thought of the past' (44).

Anthropocentric and epistemic fallacies in philosophy have meant that false conceptions can thrive, usually based on seeing things only in relation to human possibilities and limits [what Deleuze says about Kant or Husserl] .  We need instead to fully discuss the transitive dimension and its social aspects, and then the intransitive dimension based on an adequate philosophical ontology, independent of humans and their activities.  Transcendental realism 'will be vindicated by its capacity to render intelligible the underanalysed phenomenon of science' (45) [naive --ignores the  circling between the philosophy and the definitions of science , an abductive one at best] .  Any adequate philosophy of science should regard knowledge as socially produced and thus as transitive ['having a material cause of its own kind', apprently the original definition of the transitive in Aristotle,  not just a reflection of the natural world or of the human mind].  It should also address objects of knowledge as independent and intransitive, not depending on human activity or perception.  Only this sort of philosophy can explain how [proper] science operates, investigating 'the real mechanisms that generate the actual phenomena of the world, including as a special case our perceptions of them'

Part II

Introduction.  Realism in the social sciences (Archer) 189 -205

The first attempt to develop social sciences saw direct parallels between natural and social reality, as in Comte.  The science of society was to imitate natural science which was then empiricist, leading to the study of isolated events and the conjunction in the form of correlations.  This reduced human beings to mere material.  Critical realism is not advocating this kind of scientific sociology with a unity of method based in positivism, because both natural and social reality are reduced.  Instead, there is the same 'quest for non observable generative mechanisms whose powers may exist and exercised will be exercised unrealized' (190), because they work in open systems.

Social systems can never be closed either externally or internally [since human beings are themselves creative and reflexive].  The subjects can do thought experiments just as well as the investigators.  Social reality therefore remains as a real puzzle - it depends on intentional humans but they never conform to it, it relies on understandings which are never full, it depends on activity but never just follows activity, and it is emergent.  This is the debate between structure and agency, and it has taken different forms in social theory.

Initially, social reality was produced by individuals (Mill), which led to methodological individualism, and this was opposed by social determinists (Comte) [Mill wrote an essay on Comte which is a masterpiece of polite academic refutation].  The main problem is that individuals always operate within a social context, that there can be no individual economic act without assuming an economic system, and so on.  Collectivism initially displayed 'ontological timidity'(191), for fear of reification and with an empiricist orientation: social facts became remainders, explaining what was left when individualistic explanations ran out.  Instead of doing ontology, lots of early sociologists deployed the notion of social structure as an heuristic [with Lockwood as an example].  Gellner was an exception in suggesting that structural properties were real.  The empiricist turn failed to uncover social structures, since they are not observable.  Social structures do not just emerge from combinations of individuals, but must be seen as causal, again not just in producing empirical regularities.

When empiricism finally declined, sociology still avoided ontology and turned instead to epistemology in the form of 'the born again idealism of the "discursive"' (193), committing the epistemic fallacy.  Story telling replaced explanation, as in the development of postmodernism.  Foucault is an example, operating 'by persuasion without any context of justification, yet...  immune to critique' [see DeCerteau's critique] .  When challenged, relativism becomes the defence.  Yet real economic and social structures produce the conditions for post modernists to operate in their playful way.  Structures as discourses do not help to emancipate human beings from these structures, and there has been a shift towards a non political development [flourishing].  There has also been an anti humanism.  The result has been a mere 'campus language game' where anything goes, instead of seeing humanity as 'a natural kind (meaning species beings who are more than their biology but less than their socialization)'.

The decline of positivism did reduce the role of human experience in knowledge, and this also avoided the charge that the non observable features of society were reified rather than real.  Transcendental realism emerged and was opposed by the mere 'celebration of local incommensurable language games' (194).  Transcendental realism permits corrigible explanations, without determinism, although there can still be substantive debate about what the domain of the real actually consists of.  Other approaches implied an ontology anyway, so there seems to be a necessary logical relationship between ontology, explanatory methodology, and practical social theories.  Instrumentalism would deny this sequence, focusing on methodology, as displaying what works: mutual inconsistency among methods ensues.  Postmodernism denies the ontological stage, but has no explanatory methodology as such, merely 'aesthetic appreciation' with no commitment to practical social theories.

Critical realism assumes 'intransitivity, transfactuality, and stratification' (195) as the basis of its ontology.  There is an ontology of structures instead of events, intransitive entities.  Explanation now requires us to research a state of matter in itself regardless of how we view it, to prevent the collapse into epistemology, and sees this as limiting what can be theorised.  However society is 'morphogenetic', with a capacity to change shape or form, with no necessary tendency to social equilibrium.  If we are looking for intransitive mechanisms here, we must look at neither individuals nor groups, but at social life and its relations, which produces a relational conception at the level of methodology: relations are irreducible to individuals or to determinist social forces.  Turning to transfactuality, the implication is that any current form of society is historically contingent, but there is no pure contingency, and not everything is flux [in other words there are grand narratives, and even little narratives assume a certain durability].  The relative endurance of social relations means that we can abandon any 'baggage of preconceptions' such as organicism or mechanism, or any attempt to reduce social life to language or any other part.  Instead we have to provide 'an analytical history of [a society's] emergence, of why it is so and not otherwise'(196).  It is not enough to rely on analogies with other mechanisms because this produces 'inadequate retrodictions' and untested assumptions.

This instance on the stratification of reality means that surface sense data alone will not do.  There are the three domains of reality instead and they should not be collapsed.  [In examples, experiencing a cherry tree in England depends on an economic system to import it, or experiencing educational discrimination follows from a particular institutionalisation of a definition of achievement].  What is required is a vertical explanation not just a horizontal association between events, but one that goes into generative mechanisms and relationships, 'vertical causality which simultaneously entails temporality'.  Horizontal causality is technically endless, and is usually packaged by an implied vertical dimension, as in Collier's examples of ideological practices, where specific ideologies presuppose a general ideological stratum [the example is presenting Thatcherite economics as a form of household shopping, which itself already carried an ideological charge].  This is a variant of the argument that the social forms the necessary conditions for individual acts, that they preexist [it also seems to be a basis for distinction between Bhaskar and Giddens].

Outhwaite argues that critical realism supports a number of different approaches, including those based on action and structure, but this assumes that it acts as a kind of philosophical underlabourer with no direct transfer into scientific ontology.  However, realism assumes at least that there are intransitive objects in social life, and advocates need to show that either structure or agency are candidates [and Archer doubts whether they are].  It could be argued that the intransitive objects are constituted differently, for example that they might be enduring sets of interpretations or structural properties, and Bhaskar seems to support this view, at least in giving a major role to beliefs and meanings which build on those that went before.  However, this assumes some conscious action in the creation of social facts, but this equates interpreted experience with efficacious facts, omitting all the other influences which have not been grasped in consciousness or interpretation [the examples here are economic processes like inflation and their impact on personal spending power].  There is also a problem with unintended consequences which would only be operative once they had been interpreted or mediated [the example here to counter this view is that pensioners still have to manage their budgets even though they do not understand the index linking of incomes].  Structures provide different distributions of material and cultural resources, and this can go on behind our backs.

Archer sees less potential in social action approaches or hermeneutic.  Not all interpretations do lead to understanding structures.  Structures can be understood by detecting their 'causal efficacy' (199) which might appear in consciousness or not.  Critical realism addresses these limits and constraints on hermeneutic understanding, and it is this that provides emancipatory potential and also 'epistemic fallibility'.  There is no point interviewing people to get at these structures.  Critical realism is more than a philosophical underlabourer inducing discussions only about ontology, but should lead to 'practical social theorising', moving beyond a purely 'mentalistic' stance.  Endless philosophical agnosticism will not permit sociological progress.

However, all social things are dependent on activity, and so structural properties can be seen as dependent in the same way.  This is what Benton says, and this limits naturalism.  For Benton, the account of individuals as having power only in relations is inadequate to explain social structures.  Apparently, Manicas says something similar, arguing that social structures are only incarnate in practices, and this leads him to see similarities between Bhaskar and Giddens -since the agent and the structures generated coexist, 'social realism collapses into structuration theory' (200).  Other contributions in the collection argue against this view, seeing social characteristics as different categorically from individual characteristics, in the process of emergence: Bhaskar sees emergence as taking place through various mediating systems, positions occupied by individuals and the practices associated with those positions [sounds a bit like role theory].  It is not just individual practices alone, which would involve an inevitable individualistic basis for social life and lead to Giddens.  If the positions and products of individual activity are different from mere practices, differences with structuration theory remain.  Porpora argues this in terms of saying that one does social relationships are established, they are 'analytically prior' to subsequent rules following behaviour, that positions predate practices, and produce constraints and limits from interests, resources and powers.  Practice is therefore not seamless but goes through stages. [Could be Berger and Luckmann]

Thompson has also criticised structuration for ignoring the positions and organisations as an intermediary, so that patterns of action are not just voluntarist but 'positionally conditioned'.  Archer's own position turns on asking about the origin and power relations of positions, roles and institutions which predate current agents and limit them.  These effects can be detected through the impact they have on agents' actual practices, and on how agency itself is considered.  It is not just a temporal difference, but rather a definite temporal relation involved: phenomena are '"tensed"' (202).  The temporal dimension helps us distinguish structural conditioning and its effects on the interaction, which then provide structural elaboration, as in the 'transformational model of social action' in Bhaskar, and in his dialectic, or Archer's morphogenesis.  If we use a temporal model, we can explain social structuring over time as a relation between structure and agency, producing actual research [and politics?], rather than just gaining sensitivity to complexity, as in Giddens on the duality of structure.

So critical realism does imply a choice of theories, one which rejects reductionist accounts or conflation approaches [also called '"elisionism"', 203].  Conflation approaches assume that structure and agency can never be separated, and that every aspect of structure is activity-dependent in the present tense, while structure is only ever effective through individual agency.  Realists reject these views, and this is an ontological difference but also a methodological one.  Once we see structure and agency and separable, we can examine the interplay between them, and thus develop practical social theories based on actual structural constraints, and isolate causals rather than assuming that there must always be joint responsibility.  Emergent properties which produce preexistence also produce relative autonomy and causal influence, and we can analyse these partly to develop 'strategic uses of our freedoms for social transformation'.

Chapter 8.  Societies (Bhaskar) 206 -57

The first stage is to argue that societies are irreducible to people, and then to examine their connection with people.  Social forms are necessary for any intentional act, and this preexistence makes them autonomous.  They are causal in their effects and it is this that 'establishes their reality' (206).  What human activity does is to reproduce or transform the social forms, and it is this that makes social science possible.  Social science moves from the manifest phenomena of social life, the experience of the social agents, to the essential relations 'that necessitate them' (207).
Transformational models of social activity entail an inherent relational conception of the subject matter of social science, where a society can be seen as the sum of relations within which individuals and groups operate.  Human individuals may or may not be aware of these.  This also explains the emancipatory potential of social science.

In natural science, the objects that are thought about may be real in their own right - intransitive.  Is society intransitive in this sense?  It is common to assume that society is made up of persons and their actions, and only exist through human experience.  However, such reductionism could go further than individuals, to the neurophysiological level.  Social atomism and methodological individualism are the results.  Instead, it is necessary to argue that 'societies are complex real objects irreducible to simpler ones such as people' (208).  [Note that Popper is a methodological individualist here.  Several others are reviewed as well, 208].  However, all the apparently characteristic properties of humans 'presuppose a social context for their employment', 'irreducibly social predicates'[cashing a cheque is the example].  Nor are facts about individuals any easier to observe or understand than when trying to describe social facts.  Attempts to modify methodological individualism through the use of devices such as ideal types, or statistical generalisations, actually undercuts the epistemological claims.  There is often an implicit assumption that the social simply refers to group behaviour, but this leaves out relations between groups.  The underlying assumptions of the position arise through the connection with political liberalism.

Sociology should not just concern itself with large scale or group behaviour, but with relations between individuals and groups, the relations between these relations, and the relations with nature.  Relations do not always involve collective or mass behaviour [we can specify those that relate to individuals such as between husband and wife].  Arguing that human beings are rational explains only how action occurs, and is not a sufficient explanation of the action, more 'an a priori presupposition'(210).  Rational accounts often contain 'a normative theory of efficient action' rather than offering an explanatory theory for actual empirical episodes. At the same time, it is true that social effects appear as changes in people, and material affects as the result of actions of people, but this needs to be understood adequately without reductionism.

There is also a problem with collectivist conceptions, as in, say Durkheim, which emphasize the group [in the full social sense].  This means that relationships arise from collective phenomena instead of individual relationships.  Durkheim combined this conception with a positivist methodology [to understand the collective as empirical patterns].  Weber embraced individualism with a neo kantian methodology, breaking with utilitarianism only by specifying other forms of action.  Again an empiricist ontology constrains possibilities, thus he and Durkheim were prevented from developing an adequate social realism.  Marx did better with a realist ontology and a relational sociology.

Thus we have individuals producing society with Weber, and societies producing individuals with Durkheim, voluntarism with the first and reification with the second.  Berger and Luckmann (and others)  offered a third model where first individuals create society, and then society produces the individuals, 'in a continuous dialectic' (213), which seems to do justice to both voluntarism and reification, although it splits social facts entirely from natural ones.  Social objectivations can produce alienation under certain conditions, with the former inevitable.  However, this model combines the errors of voluntarism and mechanism.  Bhaskar insists that people and society are'radically different kinds of thing' (214).

Society exists prior to individuals, just as the natural world does - it 'is always already made'. Human agents make it exist, 'Society ... exists only in virtue of their activity',  but they do not create it, rather 'they reproduce or transform it'.  Human agents do not objectivate, but rather work on given objects within preexisting social forms, of communication or action, for example.  Human action becomes transformational, working with materials that exist already, just as natural scientists do - as in Aristotle, a material as well as an efficient cause is necessary, and we can regard social activity as producing and transforming material causes [in other words transitive objects].  So society is different from human agency, it is both its condition or material cause [intransitive] and its outcomes: it is dualistic. [objects and products coexisting?] Agency both produces and reproduces, so it is dualistic as well [again, as mixtures of both?].

Human action is intentional, which implies a capacity to monitor and control performance, and to monitor the monitoring and offer a commentary.  Societies do not have this capacity.  Mostly, this human capacity results in the unconscious reproduction of social structures, however, as when we marry and in the same process reproduce the nuclear family as an inexorable unintended consequence.  Sometimes human agency leads to social change, but not always.  If human intentions lie in the realm of psychology, and social forms in sociology, it might be possible [but it is not particularly necessary] to think of a linking science which explains 'how people reproduce any particular society' (216).  [Otherwise, we should can be content with different social sciences explaining different parts of the overall process?] Conscious human action can be described in terms of the agent's reason for engaging in it or in terms of its social function or role: it follows that human choice often 'becomes functional necessity'.  In this way, 'necessity in social life operates in the last instance via the intentional activity of agents'.

None of this accords with common intuitions, even though we might well be aware 'that the reason why garbage is collected is [not] necessarily the garbage collector's reason for collecting it (though it depends upon the latter)'.  We are familiar with the idea that structures, unlike the rules of grammar, limit speech but do not determine it.  Society therefore can be seen as 'an ensemble of structures, practices and conventions which individuals reproduce or transform, but which would not exist unless they did so'.  Similarly, 'human action always expresses and utilises some or other social form' (217). Certain skills, competencies and habits are required for reproduction and this is socialization: the skilled reproduction and transformation of society is an achievement not a mechanical consequence.  This achievement itself requires certain 'conditions, resources and media', as material causes, and this can never be just reduced to the individual.  Breaking with alienation does not mean a breaking with these conditions and so on, but rather self consciously transforming them, 'for the development and spontaneous exercise of their natural (species) powers'

This account offers a better notion of social change and history: Weber can only explain change as a matter of contrast, [eg alternations of types of action?] Durkheim as something external, Berger and Luckmann as a matter of 'recreation with genuine novelty.' In this account, social events can themselves initiate or constitute ruptures and mutations or transformations [both as external and as a result of innovative individuals?].

What are generative mechanisms in social science?  We can develop analogies.  Human work is an analogue of the natural events that change objects.  Social structures are an analogue of generative mechanisms, even though they exist only through the activities they produce and cannot be identified independently of them: 'they must be social products themselves' (218).  So human beings must also reproduce the structures that govern subsequent human activity.  This makes social structures only relatively enduring and only relatively autonomous.  They link together to produce society as 'an articulated ensemble of such relatively independent and enduring generative structures…  a complex totality'.  Change can arise both from these components and their interrelations.  Since human beings can theorise about their own activities, they are also objects of transformation, and are both relatively enduring and relatively autonomous.  Social structures are social products and therefore cannot be explained 'by reference to non social parameters', even though these might act as constraints.  All this makes societies 'sui generis real'. However, there is a difference between social structures and natural structures which do exist independently of human activity, and also of human conceptions, and natural structures are not only relatively enduring.  These are 'real differences in the possible objects of knowledge' (219) for social and natural science.  In essence, societies as an ensemble of tendencies and powers, but these have to be exercised by intentional activity 'and are not necessarily space - time invariant'. [At this stage, Bhaskar declines to explain these different objects in terms of a more general theory, as Deleuze might, and says the differences are irrelevant for the purposes of epistemology-- which he has degraded to a second place issue earlier!]

Emergence characterises both the natural and the human worlds [and we can understand intentional action as producing this emergence in the social world?].  We can also suggest that the social is a precondition for certain actions, and this enables us to employ causal criteria, and thus to assert [!]  that society is real.  Durkheim argued like this when asserting that social facts appear as external and constraining, although we must reject any determinism.  It is also the case that social structures enable as well as coerce.  Nevertheless, within these limits, Durkheim was right to use causal mechanisms in social science.  The same arguments can be used with relational conceptions of social facts.  [Then a strange bit about how this can include non empirical objects as well, 220, and this is somehow guarantees the autonomy of social theory -- not linked to the empirical alone?]

We have to remember that the relational conception of sociology accepts that social forms exist but maintains that in being social, they depend on the relationships between people, and those with nature.  Here, we need 'mediating concepts' to explain both the social and the individual dimensions, the '"slots"'which people must inhabit to reproduce the social structure, which also refers to human agency and human action [the old concept of role? Looks a bit functionalist here, assuming congruency etc].  The terms we need are positions and the practices they entail, the 'position-practice system', which must be considered relationally.  Sociology therefore takes as its main interest the initial conditions of any particular event which include or presuppose reference to some other social relation.  The background relations can be analysed as differentiating, producing and reproducing, changing, or shifting, to produce 'particular social forms and structures' (221).  It is these relations that will endure, not individuals and not social groups.  Relations take the form of various kinds of interactions [personal and positional].

A particular interest lies in the distribution of the structural conditions of action, the differential productive resources they attract, and how persons and groups are allocated to particular functions and roles in the division of labour.  This also raise the possibility of conflicts, and of 'interest motivated transformations in social structure'.  It is not just a matter of exchange, and nor just a tension between societies and individuals as in an awful lot of orthodox sociology.  Marx added to such a conception a notion of historical materialism that privileges material production, but this can only be justified to the extent that it generates research programmes and theories 'progressively richer in explanatory power' (222) [very conventional stuff on theory choice --extended below], and there are problems articulating the two components.

The issue of internal relations is also problematic.  Rigorous causality as in Hume sees all relations as external, leading their opponents in idealism [including Bergson] to insist on the universality of internal relations, and this dispute arises within Marxism as well.  Instead we need to recognise both - 'it is in principle an open question'.  [Logical discussions of internal and external relations ensue, 222.  The gist of it seems to be whether the terms are symmetrical as in bourgeois proletariat or asymmetrical in essence.  Bhaskar wants to suggest that we are interested in 'real definitions...  Produced a posteriori in the irreducibly empirical processes of science' (223), and we should ignore the internal relationships at the logical level.  What relations really show is 'ontological depth or stratification', and this is 'consistent with relational internality' [but explains it better?].  We have to remember that 'surface structure is necessary for deep just as langue is a condition of parole and intentionality of system'].

Most social phenomena and most natural events 'are conjuncturally determined', with multiple causes, but whether these multiple causes constitute a totality as such, with its own internal relations, 'remains open' [Deleuze would make that step].  It might be possible to take an apparently contingent relationship, such as an industrial accident, as indicating a totality with ['deeper'] relations between economic activity and state structure, for example.  This is always possible and it gives sociology a 'chameleon like and "configurational" quality [with a reference to Elias].  Although totalization involves thought, it is also real - our understanding might suggest a contingency, but there can be a reality where aspects are conjoined [seems to be a support for Deleuze then, but he also likes accidental conjunctions etc].  'Social science does not create the totalities it reveals, although it may itself be an aspect of them'. 

Marxism has always claimed that social life is a totality, drawing upon a theory of history to explain the modes of articulation of the moments of that totality, so we can use historical materials to point to a totality.  We can see the specific social sciences [such as linguistics, economics, history, sociology] in the same way, explaining how 'generative complexes'produce particular types of social activity.  It would be wrong to hypostatise these types as independent of other social activities.  Such specialised activities can be seen to 'logically presuppose a totalizing one' (224), but they themselves do not offer a general theory.  Thus we can reconsider sociology as considering particular kinds of relations, especially relations of production [in a general sense].  If those relations themselves are internally related, sociology must 'either presuppose or usurp the place of just such a totalizing and historical science of society as Marxism has claimed to be'.  [An additional possibility is to accept pragmatic limits by denying internal relations. You could make the same points, but substituting Deleuze for Marx - Bhaskar is presupposing or usurping Deleuzian virtual reality, or is content to accept unanalysed limits on his ontology to address particular problems in social realism, such as a preference for repetition rather than difference].

Returning to the ontological limits outlined above,  social structures are clearly dependent on activity, concepts and space - time, and therefore social sciences have a different subject matter from natural science.  We can add some problems.  Since society cannot be directly perceived, it is necessarily theoretical, but this is the same for many natural scientific objects as well - we know them through their effects.  However, societies do not actually exist independently of their effects, which produces a strange ontology [this is where he leaves off ontology for epistemology as he says explicitly].  However, the real problem is that social objects manifest themselves in open systems which do not permit empirical regularities and which cannot be experimentally closed.  This is what makes simply borrowing methods from the natural sciences 'totally inapplicable' (225) [this is directed at classic theories of causality and laws, and even Popper's theories of rationality and criteria of falsification].  Given the absence of any possibility of a decisive empirical test, other criteria for rational development and choice 'must be explanatory and non predictive', and this includes the capacity to generate research programmes, and to explain novel events 'without strain', perhaps before they are even realized.  This does not weaken the ontological claims, but simply affects the form of laws: these are still understood to be tendencies just as in natural sciences, and we should not assume that laws are not operative in open systems [they would have to be laws that permitted transfactuality] .  Generally, although we might not be very confident about our analyses, if they are independently validated [how? Anyhow?]  we are warranted 'in applying it transfactually as a natural scientific one' (225 -6).  Usually, we have to choose between theories, which implies a relative preference [and a strange piece which seems to say that we can use a number of different grounds to justify this preference].

Given that the subject matter of social sciences is structured by relations of various kinds, there is a constraint on what might be said by theory.  There might always be a necessary historical dimension relating to totalities, but this is only the same sort of difficulty that we find when discussing open systems.  There might be problems with meaningful measurement, however.  One is that irreversible social processes mean that we must consider qualitative not just quantitative change [if these processes refer to events that are necessary?].  The greater problem is that 'meanings cannot be measured only understood'(226), stated in human language.  This means that language must be precise just as measures must be accurate in natural sciences, and the capacity to generate precise language is 'the a posteriori arbiter of theory'.  However, neither accurate language nor measurement necessarily leads to the rejection of theory.

Experimental activity in natural science also enables practical manipulations, and laboratory manipulation can be crucial in the process of scientific discovery.  However, social sciences have an analogue, and an advantage, in the internal relation they have with their subject matter.  The objects of social science can be affected by social science, or even constituted by it, in the case of the sociology of social science.  Conversely, social science itself can be affected by social developments in the rest of society.  In this sense, the objects of knowledge are not independent of the process of production of knowledge, unlike natural sciences.  However there is a difference between causal interdependency and 'existential intransitivity' (227), which is taken to be a difference between contingency and an 'a priori [logical?] condition of any investigation'[in other words, existential intransitivity is to be asserted or produced by abduction?].  Whether social or natural, once an object exists, 'however it has been produced', it becomes a possible object of scientific investigation, and its properties are independent of the process of investigation.  This notion of existence is 'univocal: "being" means the same in the human as the natural world, even though the modes of being may radically differ'. [Looks like
arguing for 'reality as we define it'?]  Human sciences thus have intransitive objects, although they themselves can both be an aspect of and cause processes of investigation.  Positivism ignores this interdependency between explanation and object; hermeneutics denies the resulting intransitivity.  Both limit scientific critique 'upon which the project of human self emancipation depends'[so we are now making political arguments to support this particular conception, in addition to the ones about explanatory power and generating research programmes. It would be hard to deny the productivity of either approach in terms of research programmes or explanatory ambition].

All societies generates a theory of themselves, and we can consider this as protoscientific knowledge.  Transformational activity requires this set of ideas to be replaced by social scientific theory which in turn requires cognitive resources.  Social change itself affects this process [provides both different sorts of protoscientific knowledge and different sorts of social theory?  The example suggests both, Marxism emerging as a result of the 1840s crises and the subsequent effects of Stalinism and fascism, the Cold War and the consumer boom] Social crisis can make generative structures more visible and this is 'a partial analogue to the role played by experimentation in natural science'(227).

It already been argued that 'deductively justified predictability' (historicism) (228) is impossible because of open social systems.  Transformations can also bring about qualitative change which is not easy to anticipate.  This is why social science must be necessarily incomplete, for ontological not epistemological reasons.  Possibilities and effects of social development can often be understood only afterwards, implying a particularly tight link between the development of the object of knowledge and the development of knowledge itself.  [Apparently, this leads to a difference with Lakatos about how to judge research programmes - presumably, Bhaskar is saying that social change itself affects the productivity of research programmes].

If we have suggested a generative structure, we can test this empirically, 'albeit exclusively in terms of its explanatory power'.  Constructing theories in the first place can be difficult if it is not possible to specify an appropriate 'explanatorily significant' object of inquiry, especially if we have ruled out studying merely events as in empiricism.  Without adequate specification, a number of different theories might be able to compete over definitions and boundaries, and the distinction between theory and its applications might become merely arbitrary.  In these circumstances, falsification is not possible, because 'every theory, if interpreted empirically, is false' (229).  Social science is fortunate here in being able to accept nominal definitions of social activities as transitive objects, based on agents' descriptions, or theoretical redescriptions of them.  If we are to transform protoscientific knowledge into theory, we can begin with 'a real definition of the form of social life' that has already been identified: this will prevent reliance on the arbitrary, and it is an essential step in any causal explanation.

Real, non arbitrary, definitions in social science are different, because we can see the mechanisms at work only in terms of their effects, unlike in every other kind of scientific knowledge.  However, another analogy can help here, when considering the relation of philosophy to science.  Philosophical objects do not exist in their own right, but are those already defined by science.  In the case of science and every day social definitions, the objects presuppose the second order discourse, however.  The goals of philosophy and social science are different - the former seeks to establish the a priori conditions of knowledge, and the latter to find 'mechanisms and relations at work'; the former offers formal conclusions, the latter historical ones subject to empirical test.

Generalising, we can now suggest that transcendental arguments [in social science] are simply a species of retroductive ones more generally [counteractualised arguments for Deleuze?], except that they attempt to explain the conceptualised activities of agents, and can isolate only necessary not sufficient conditions for these activities, given an open system with multiple causes.  This is not to collapse the two sorts of argument, especially philosophy and science, because philosophy uses 'syncatgorematic'(230) terms, rather than terms with a precise reference.  Also, social science discourse operates with two different levels of reality, '(social structures, and their effects)'whereas philosophy accepts that the only level of reality is that defined by science [not true with Deleuze of course].  In both cases, explanation in the second order discourse is not adequate without 'supplementary considerations'[reasons for theory choice?].  In social science, though supplementary consideration should include 'the provision of independent empirical grounds' for the existence of the generative mechanisms.  Philosophy operates at a more autonomous level: sciences develop an immediate reference to the world and also require possible 'a posteriori grounds for the analysis'.

We have now deduced the possibility of social science from the necessary social forms which preexist intentional action, in a 'formal use of the transcendental procedure'.  We can use this definition of an adequate social science to evaluate existing sociological theories.  Marx in Capital comes closest to a transcendental argument, attempting to show what must be the case if the phenomenal forms of capitalist life exist; specifying how economic phenomena in capitalism can be understood in a more 'pure' way; specifying categories that are required for any concrete investigation.  However, it is necessary to argue that all social phenomena 'must be grasped as productions'.

Social activities 'as conceptualised in experience'are to be the middle term,'minor premise'  for transcendental argument, and these are necessarily dependent on space and time.  Here hermeneutics has done much to clarify such experience, although it is too committed to empirical realism.  As a result, it fails to explain the conditions for the phenomena and their intransitive reality, and to recognise that phenomena can actually be false or inadequate.  Transcendental arguments particularly acknowledge the second point, if structural complexes produce experiences which may not be immediately represented in intuition.  We have already established that agents may be unaware of the structures behind their productions, as in grammar and speech.  Transcendental analysis makes possible a 'second order critique of consciousness', as in Marx's work on commodity fetishism, or where historically specific practices become naturalised, or idealised as something eternal.  However, Marx also offered a first order critique of consciousness, where the phenomena themselves are false or inapplicable - in his critique of the wage form as an expression of the value of labour, which he saw as imaginary and irrational, akin to 'confusing machines with their use'(232).

We can see that social theory's relation to every day knowledge involves analysis of the 'real but non empirical and not necessarily adequately conceptualised conditions' of every day knowledge, and implies a critique as both conceptual criticism and change.  We can call every day conceptions 'ideology' only if we can argue that they are necessary, that they need to be explained as well as criticised, showing why false beliefs are held in the first place.  There is no parallel in natural science.  Social criticism and advocacy of change then becomes possible.  In addition, theory fuses into practice 'as facts about values, mediated by theories about facts, are transformed into values about facts'.  Value neutrality collapses, 'the last shibboleth in the philosophy of the social sciences', as values can be shown to be false [based on false facts].

If social activity is both historical and interdependent, the social world must be open.  If social activity is to be socially explained, social science is part of its own subject matter.  Transformational arguments when applied to beliefs and cognitive material clearly imply a 'commitment to a principle of epistemic relativity' and again this makes moral and political argument limited and open.  Nevertheless, this ontological status makes it the possible object of knowledge which is 'still scientific' (233), even if not like natural science.  Any law-like statements will express only 'historically restricted tendencies operating at a single level of the social structure only'.  Some of these tendencies may never actually be manifested even though they are important in understanding different forms of social life [the example is the tendency for rates of profit to fall in capitalism].  Society is 'a complex and causally efficacious whole --  a totality, which is being continually transformed in practice'.  We cannot read it off from a given world nor reconstruct it only from subjective experiences, but it is at least 'on a par with the objects of study in the natural sciences'. 

[I must say I would have been relieved if this very long chapter had actually ended here, but there is another 20 pages or so!]

[A section follows on a critique of the fact value distinction, 233 - 43, but we have grasped the basics in the stuff about criticising ideology above.  Perhaps the most interesting bit is the discussion on relativism.  Bhaskar starts by considering two objections - that relativism is self refuting, and that it denies the possibility of translations between cultures.  The first argument confuses epistemic relativity, the view that all beliefs are socially produced, with 'judgmental relativism', that says that all beliefs must therefore be equally valid and can only be distinguished by some kind of irrational decision.  As we have seen, his transcendental argument leads to epistemic relativity, but he argues that this does not lead to judgmental relativism, since there must be rules for preferring one belief to another.  I must say I find these rules rather vague so far - explanatory power, research productivity, and some sort of reference to independent empirical validity, combined with some goal of achieving emancipation, sometimes expressed in a strange form that says that rational discourse is both cognitively and politically essential,better expressed in Archer.  Translation clearly is possible, but this does not necessarily suppose absolute standards - we can also interact with animals.]

Appendix on Marxist conceptions of ideology

Marx uses ideology in two ways, as part of the superstructure to be explained in terms of the base, and as part of the analysis of the base itself as in commodity fetishism.  This is because Marx actually had two distinct research programmes, one theorising and critiquing the capitalist mode of production in Capital and another referring to historical materialism, as in the sketch in the 1859 Preface.  These were never properly integrated, hence the absence of any 'theory of capitalist society'(244) [original emphasis].  His followers tried to solve this division.

An enduring problem is how to explain both relative autonomy and determination in the last instance for the super structures.  There has been discussion of two errors, an idealist separation of superstructure, and an economistic reduction of it.  The problems can be seen when trying to locate science: it becomes entirely autonomous for Althusser, but the mere expression of reification for Lukacs.  Implications arise for analysing ideology.  In Capital we have a theory of 'false or superficial economic ideas', but we cannot generalise the analysis to consider all ideas.  The general work on historical materialism suggests a framework for the development of these ideas, however but there is a risk of including all ideas into some 'undifferentiated bloc'.  Also, activity clearly contains an element or component of ideas, some conception of what agents think they are doing, even if this might be mistaken.  Materialist arguments have not been theorised adequately, but were simply asserted rhetorically, for example when criticising Hegel.

Instead, we should see different ideologies as associated with different practices, including scientific practices.  Ideologies might indeed relate to each other, but this has to be investigated.  Different practices might have different degrees of autonomy from the base, and in some cases, might become relatively autonomous with 'bases of their own' [the examples are 'physics, technology, literature, warfare'].  Additional problems arise from contrasting ideology to science, and seeing it as essential to social reproduction, and as false consciousness.  However, the only fully worked out example is the critique of political economy.

Bhasker wants to reserve the term ideological to cases where we can develop three characteristics of a theory - critical, explanatory, and categorial.  In the first case, we will want to explain most of the phenomena already explained by ideology, and then explain an additional significant set as well.  In the second case, we need to explain the reproduction of ideology and specify its limits and how it might be transformed, especially by connecting it to to a level of structural relations operating in a critical theory, but absent in ideology [there is another quality, being able to 'explain, or at least situate, itself within itself' (246) -- account for its own emergence instead of seeing it as natural or an individual  discovery etc?].  In the third case, we can reject a set of beliefs as ideological if they fail to meet the criteria of scientificity or of 'domain adequacy' [the nature of its subject matter].

If we can demonstrate these characteristics, we can argue that theory is cognitively superior to ideology, especially in its possession of an 'ontological depth or totality'.  Critical realism clearly achieves the first, but it can also demarcate social science from natural science, explain how social theories arise from connections with their subject matter, and how reflexivity can therefore be generated.  The claim is that it can explain the appropriate conditions for ideology. 

Marxism met some of these criteria but not all [for example it failed to specify what it meant by science].  We can see that Capital is a triple critique - of bourgeois political economy, of the economic conceptions of everyday life which underpin political economy, of a mode of production that makes these conditions necessary.  Marx goes on to argue that bourgeois economics cannot penetrate to the essential realities but must deal with phenomenal forms which mask real relations - reality itself is deceptive and produces deceptive appearances.  For example fetishism naturalizes value concealing historically specific class relationships, and the wage form confuses powers with their exercise, again concealing reality, especially in terms of the production of surplus value.  So these misunderstandings are 'readily explicable'(247), and there is no crude reductionism.  Mepham describes this as a circular process where real relations generate phenomenal forms which are reflected in ideological categories, which underpin ordinary practices in the economy.

Ordinary persons see their practices as connected directly with phenomenal forms. Marx operates differently, moving 'retroductively' from phenomenal forms to real relations, which helps him criticise ideology and informal practices.  The analysis explains the conditions for phenomenal forms as well. This removes any additional obvious value judgements, except those preferring rigorous explanations and claiming superiority over bourgeois theory.  Marx sees that ideology is both false and necessary.  Theory's subversive challenge is based on its cognitive power.  The gap between what an object is and what it appears to be arises from pointing to structures and relations which are not apparent to experience or bourgeois ideology.  It is misleading to argue that this gap is a necessary contradiction, however.  Social life is riddled with necessary category mistakes, but these and their interrelation reproduce the structures that generate them.  There is no split between a political and a scientific Marx.  Thought is part of social reality, and there are logical contradictions in both.  Further, thought cannot exhaust social reality, and does not constitute it on its own, so there must be misrepresentations of reality, and some of those are going to be 'necessary for what they misrepresent' (249).  We are describing causal relations not logical ones alone.  We have contradictions, but not just contradictions in thought.  Misrepresentations also relate to each other, 'are necessary for each other', and are not just contingently or externally related as if they were separate forces.  [All this is used to ground a rejection of Colletti and Hegelian readings of Marx as operating with contradictions and idealist dialectic].

Chapter 10.  Realism and Social Science (Outhwaite) 282 -312

Social realism insists on a distinction between transitive and intransitive objects of science; the stratification of reality into the real, the actual, and empirical; causal relations as tendencies produced by  generative mechanisms; the rejection of empiricism and conventionalism, turning on the notion of the real definitions which are 'statements about the basic nature of some entity or structure'(282) [the example is a statement about the chemical composition of water as opposed to what humans think of it.  But why should a chemical statement particularly depict reality, as opposed to, say, the thermodynamic definition of systems and states?].  Finally, there must be explanatory mechanisms which have to be demonstrated to exist.  There is a difference between philosophical and scientific ontologies, however, and only the latter actually tells us what the structures and mechanisms 'actually are'(283). 

We might start by noting accounts of social reality that definitely rule out realism of this kind, ones that insists there are no intransitive objects of social science, and that generative mechanisms do not explain social objects.  Intransitivity certainly needs to be modified in social science because agents' conceptions make up part of the reality themselves, as when distinguishing between, say, a real and a simulated quarrel.  However, events such as wars or political developments of the past might be intransitive in the sense that they actually happened regardless of anything that's written about to day [now enter Baudrillard and the controversy over the Gulf War].  The radical conventionalist view would argue that social situations never exist independently of the way they are interpreted, and that such interpretations 'are essentially arbitrary'. 

Conventionalism looks plausible because it argues that society is not observable; it is theoretical; and any assertion about society is as good as any other.  The first is uncontroversial because once we talk about the collection of people in some geographical site, we are talking about some licenced theoretical moves and not others.  The second alludes to some problems with theoretical terms and the way in which they are legitimated.  The challenge these days comes from the notion of individualism, especially methodological individualism: ontology turns on individual persons and their actions and social structures are 'merely summary, metaphorical redescriptions of these'.  Individuals do seem unproblematic because there are individual and distinct human bodies.  The usual counter this to argue for the importance of social relations as a precondition for individual actions, as necessary conditions for significant human actions, even if there is some controversy about how to characterise them.  The problem is not that social effects are difficult to measure, rather that they are difficult to explain. Nevertheless, some social patterns are 'as real and general as one could reasonably expect' (285), like the link between parental social position and educational achievement of children.  Here we see that social sciences are more real compared to common sense than are the natural sciences.

So far, despite the difficulties, it is possible to ask 'questions of a realist kind about social structures'.  However there is a 'radical interpretivist position' that says that structures of society are nothing but interpretations.  There is nothing interesting to be explained in realist terms and generative mechanisms.  Actors' reasons are different from causes, and this is easy to sustain if we take cause to refer to empiricist conjunctions of independent events.  However, there are certainly bodily states like fatigue which can interact with mental states.  It is also possible to argue that 'reason - explanations are the analogues of mechanism - explanations' and can therefore support a realist interpretation of explanatory models.  We can define reality as something that makes action rationally compelling.

In general, everything turns on whether there are or are not intransitive objects in social science, whether social sciences display naturalism, in Bhaskar's  terms.  If so, there would be a unity of method between natural and social sciences, despite significant differences in method dependent on real differences in subject matter.  Bhaskar argues that societies do possess properties that might become objects of knowledge, that they are irreducible to people, that social forms are a precondition for any intentional act, and that this establishes the autonomy of social forces as objects of scientific investigation, since they exhibit causal powers.  This conception permits transformational social activity, since social dimensions are both a condition of activity and the outcome of human agency [then a Bhaskar example that looks a bit like Bourdieu on the coincidence of individual motives and reproductive requirements - people marry for personal reasons, but the unintended consequence is the reproduction of the nuclear family].  Social relations take the form of structurally and relationally 'defined positions' (287), and the relations can either be internally meaningful [as when buyers are linked to sellers] or contingent and external [shoppers related to traffic wardens].  The connections with other social theorists are acknowledged [not specified here, but Giddens seems especially important for Archer].

It follows that not all actions are tightly structured - for example work relations vary in the amount of autonomy offered to the worker, but the real issue turns on the distinction between voluntary action and structure of constraint.  Here, even an apparently creative activity like writing a book can be seen as 'the more or less automatic product of the set of theoretical and ideological structures', a 'hyperstructuralist conception'.  Agency is still required to make the structures work, however.  In the opposite direction, social structures can be radically questioned - Harré argues, for example, that experience is always required even of social facts: in other words, when experienced, they involve theory and interpretation.  Global patterns might exist outside this experience, but if so, they can never be understood, they exist in a 'noumenal realm of latent structures' (288).  Certainly, although most important issues for a social psychology focus on structures as they are perceived by actors, where, say, the structural relations of roles are interpreted as a set of imperatives and constraints.  This still leaves a possibility for understanding at the nonexperiential level, however.  Socialization techniques involve 'the inculcation of structural conceptions'.  Nevertheless, for Harré, anything beyond empirically oriented social psychology must remain speculative, and offer a debatable rationale.

Bhaskar says naturalism would be threatened if social structures only existed through the actions which they are supposed to govern, and the agents' conceptions of what they're doing.  Social structures might also be only 'relatively enduring'(288-9).  The last one is not particularly significant, since many natural structures are also are only relatively enduring.  The other objections can be dealt with too if we acknowledge that structures can sometimes apparently generate only possible actions, some of them contradictory, appearing as negatives such as deterrents or other effects of a power structure [so what looks like exception to a structure is still an effect of the structure].  [There also seems to be another repetition of the argument, that objective possibilities structure individual relations, as when a gift relation presupposes some 'structure of gift exchange'].

The issue of independent existence can be defended if we remember that human agents do not always have a correct perception of the situation nor a correct conception of the activity - they can be misled.  They also need not be conscious 'of their implication in structures' such as that of the capitalist economy, which nevertheless govern their actions' (289).  Ideologies typically offer 'a mixture of conscious and unconscious beliefs' for example.

Bhaskar argues both for causal independence between social structures and representations of them and 'existential intransitivity'(290)  [of external objects].  This will argue against, say hermeneutic, and positivism.  The issue really depends on the relation between social sciences and common sense social knowledge.  Social sciences must remain closer to common sense, because we all have 'intuitions about the structure of almost all the social processes we may care to think about', and this gives us access to our subject matter.  Intuitions about physical phenomena are much more limited by comparison.  The 19th century belief was that social sciences will eventually produce the same kind of explanatory power as physical sciences, but most of the important findings have turned on not contradicting common expectations, but rather conceiving of them in new ways.  At least the objects of social science are already partly defined, and they must be taken into account because they are the perceptions of agents which have definite social effects, especially in some social situations.  In this way, 'concept-dependence and activity-dependence of social structures' (291) is an asset.

We can then return to the basic logical project, asking what a given society must be like in order to explain common sense behaviour.  [Apparently Bhaskar says this is like a philosophical investigations of the transcendental presuppositions of activity] A good example for Outhwaite would be Marx in Capital, investigating the mechanisms of the capitalist mode of production then critiquing its representations. Marx is close to transcendental realism, but so are Durkheim and Weber, operating in a neoKantian framework [so one conception of the transcendental is as good as any other?].

Beyond that, we should see sociological approaches as adequate to their 'conception of the object of enquiry'.  Ethnomethodology will not help understand capitalism, but nor will Marxism understand how to terminate telephone conversations.  Historical analysis may or may not be relevant.  Everything depends on how the observable social phenomenon is to be described.  We might then go on to questions of social ontology to ask whether this object is structured by deeper causes, or just a product of the interpretation of human beings, and social realists can contribute - but the disputes really focus upon the nature of human societies, not the nature of social scientific theories [I think this is saying that realist philosophy must quietly privilege particular kinds of social theory.  I think it also must privilege certain kinds of scientific theories as well].

In this way, it becomes possible to construe interpretivist and structuralist approaches in realist terms, with both focusing on the fundamental structures and generative mechanisms.  'Realism does not uniquely license either of these approaches' (292) [a big argument with Archer here], but it does focus attention on issues of social ontology rather than ignoring them as in positivism or conventionalism.  Positivism is rejected by a focus on generative mechanisms understood through a complex network of theoretical and observational statements, reminding us that it is difficult to close systems and therefore to predict.  This affects social science in particular, so rational theory choice here, for realists, involves having to produce a mechanism to explain the phenomenon; to provide good reasons to believe in its existence; and to rule out alternatives.

This is still not terribly helpful because there are maybe several alternative accounts with different mechanisms, and objects may be 'complex and over determined' (293), but some basic principles remain.  First theoretical abstraction is valuable, with no particular privilege given to observational statements only: we might aim at entities that explain phenomena, even if they are not particularly observable.  Second, if reality is stratified, that might provide us with a wider context for explanation, especially attempting to link up micro with more meso and macro accounts.  Thirdly, social realism can provide us with certain 'a priori considerations' to evaluate social theories.  Beyond that, it is difficult to go, and it is quite possible that two or more theories might do equally well on these criteria.  We then might opt for the more simplified one, but this would drive us to conventionalism which offers the most simple explanations.  It might be better instead to go for theories which are testable, which may or may not favour the simpler theory.  We might also an attempt to maximize 'the chances of further intertheoretical debate'.  Again most of the conventional approaches do not do this, but block off theoretical criticism from the rival camp.  Realism offers a better framework for the rational discussion of ontology. Maximum discussion can also be defended on political grounds, using Habermas on the value of serious discussion of political matters.

Realism has reawakened the importance of theoretical argument, against dominance of empirical research, but it now needs to engage in more detail 'with the conceptions of interpretation which have been worked out within the frameworks of hermeneutics and critical theory' (294) [the rest of his book apparently attempts this].

Part III The Theory of Explanatory Critiques

Chapter 15 Introduction: Explanatory critiques (Bhaskar and Collier) 385-95

The dogmatic insistence that we cannot move from an 'is' to an 'ought' can be challenged, if we assume that ethics is necessarily linked to '"human flourishing"' (385) [a phrase apparently associated with Foot and Anscombe].  There are definitions of good which contain clear factual criteria as in good friend.  There is also the case that factual statements can criticise things [in value terms, I assume] including all other factual statements, and this led Edgley to argue that other social theories and ideas can be seen as ideological and contradictory, which leads to a critique of social structures.  This is what is meant by 'explanatory critiques' in Bhaskar's work.

It is certainly true that many people, including social scientists, act as if factual or explanatory theories do have practical implications, as when Capital turns into socialist advocacy.  Critics like Marx, Freud and Nietzsche explain ideas and doubt the truth of those ideas, but the issue of truth itself is not being dealt with by examining causal origins.  Instead, the project is to show why a certain kinds of false belief prevails.  This is a methodology involving 'suspicion', and again it is not unknown in common sense.

Addressing critical potential is not a category error, as in those that believe in a value free science.  It has become common to states values openly at the outset and then mingle facts and values, but Marx and Freud and the others tried to keep value judgements 'out of their premisses'(387).  Critics say that they smuggled values back in.  But their procedures if accepted as 'logically kosher' lead to another possibility - 'that an objective study can actually discover values that it did not assume beforehand'.

Bhaskar's explanatory theories do this.  They studied both ideas and what ideas are about, structures, as in class structures, for example, and ideas about class structures.  The argument is that class structures can cause 'directly or indirectly' ideas about themselves, and that these ideas can have effects such as reproducing class systems.  In general, any society that generates false beliefs about itself shows that 'something [is] wrong with that society which (other things being equal) ought to be put right'.

It could be argued that this is a value judgment, but it is not 'an optional value judgment; it follows from the nature of the case'.  It is an objective fact that people have beliefs, and they believe them to be true.  Indeed, pursuing the truth 'is a condition of any discourse at all', and a commitment to truth inconsistency is not a 'concealed (value) premise' (388).  Pursuing falsity is harmful.  This can also be questioned in that it is sometimes better if people have false beliefs, as in white lies, or as in denying dangerous knowledge that might be put to social harm [as in the Nazis].  Some scientific research needs to be closed down, perhaps because results might be misused.  However, this shows that true knowledge is neither good nor false in itself, but 'only good or evil "other things being equal"'.  Thus facts are true whatever else is true, but when they imply that something should be done 'it almost never does so whatever else is the case'.

Thus it might be acceptable to argue that even a class structure is better than any other alternative.  Plato argued this for the Republic and the need to accept the 'noble lie close single quote about the origin of classes.  However, this possibility except that facts may logically lead to values, in that false belief is 'a disvalue', even though it might be outweighed by other values.  It is acceptable to assume that when the truth is hidden, people are trying to get away with wrongdoing.  'A free and equal society would not need lies'(389).

Explanatory critiques are emancipatory, exposing false beliefs and their political implications, whether or not these are conscious.  By attacking the fact/value division, we can develop 'arguments from the conditions of human flourishing (one might add: not only human flourishing)'.  We can extend realism into the realm of values and morality, finding an intransitive dimension there too, a 'moral realism which is naturalistic' that looks for real values in the real world.  Some recent developments in critical realism include the suggestion that universal emancipation may be implicit in every free action, for example, or that 'all ills may be thought of as absences'.  These ideas are developed in the dialectical stage of critical realism.  The principle is that 'any rational value judgement must have factual grounding', and that trying to isolate values from facts in explanation is 'radically incomplete'.

Chapter 16. Reason as Dialectic (Edgley) 395--408

The economic and social crisis in advanced industrial states is reflected in an intellectual crisis: scientific knowledge becomes an economic resource, a factor of production, but there is 'radical uncertainty' about its status (395).  In the colonial context, the crisis is apparent in, say, the contest the claims of European medical science.  Universities are obviously involved: anthropologists have become aware of Eurocentrism; psychologists worry about the boundaries between sanity and madness; many English speaking philosophers assume they are inhabiting some ivory tower of timeless forms, but are no longer fully insulated from dissent from outside.  Epistemology in the philosophy of science, and the emergence of a philosophy of social science show the problems, where the main issue is the distinction between science and ideology.

Changing conceptions of science already included Popper on 'the less stringent requirement of falsificationism'(396) [less stringent than philosophies of induction or verification]; Kuhn suggested that was to stringent; Feyerabend insisted that anything goes when it comes to methodology.  In social sciences, Chomsky saw scientificity as offering behaviourist constraint; Popper on the unity of the natural sciences has been opposed by claims that the social sciences have their own special logic and methodology, as in Winch.  Overall, there seem to be 'equally unacceptable alternatives'-an empiricism, or a relativism that makes radical criticism impossible and 'seems to be self refuting'.

Marxism has a distinctive place.  It is not empiricist.  It does not offer an alien rationality, although it can appear as one when it becomes a European cultural export.  Marxism is often criticised by analytic philosophers as unscientific, or as pseudo science in Popper.  The problems have also been identified within Marxism, especially the notion that it offers scientific socialism.  It is clearly both the scientific theory and revolutionary practice.  Critics have often operated with conventional conceptions of science and reason in this debate, based on a shared idea about the distinctions between 'fact and value, theory and practice, description and prescription, science and morality' (397).  This shared idea emanated from analytic philosophy, but also with struggles to understand the impact of science, seen as something intrusive and threatening.  One common reaction was to separate the 'indicative' statements of science from the 'imperative' statements of policy or practice [the actual example is from Poincare, 398]. 

Science can lead to a technology, but this technology itself involves values that 'cannot be settled scientifically', and this has been applied to Marxism, hence the debate about whether Marxism is an ethical or a scientific approach.  The latter was particularly appropriated by the Third International and Stalinism to mean implying laws of inevitable social change.  Both ethics and socialist science diminished Marxism 'as a programme of revolutionary action'.  This might indicate the classic ways in which capitalism reproduces itself, but the whole debate can be reopened - for example, the concept of science in Stalinist Marxism 'is itself ideological' (399) and its deeper contradictory nature can be reawakened, as Marx himself intended.  As a result, 'the question is whether Marxism embodies a different conception that supersedes its rivals'.

Marx offered a dialectic conception, part of the hegelian inheritance, used to oppose metaphysical especially mechanistic conceptions of science.  The opposites that are united could be seen as facts and values theory and practice and so on.  The key to it lies in contradiction, but not as a category of logic alone.  For Hegel, there were contradictions 'within the essence of things' (400) and the opposition and resolution of opposition is what produces social change.  However, Marxist dialectic is materialist, and locates contradictions in material reality: in this, it breaks with analytic logic.  As a result, it is common to rephrase Marxism as involving collisions rather than contradictions as such.

Popper's criticism sees the dialectic can be acceptable as an empirical theory about the development of thought, but there are problems considered from the point of view of logic.  Firstly, 'there are no contradictions in reality'.  But this assumes that reality is the immediate material reality, not the wider reality where thoughts becomes a part of reality and thus can introduce contradictions.  Rigorous logic might suggest that contradictions are meaningless or logically impossible.  Secondly, logical relations abide in 'what Popper now refers to as the Third World'.  Thirdly logical relations exist outside of time and historical change.  These arguments are characteristic of analytic philosophy which have extended into the logic and methodology of science.

However, we can draw on Wittgenstein and the Tractatus, where truth concerns the relation of a proposition to its subject matter or its object.  Here, scientific theories 'are descriptive, explanatory and predictive'(401), and these are ways in which scientific theories relate to their object.  Here, the theory can only be 'self contradictory' if reality cannot be truthfully described by it.  Reality itself is not contradictory.  The implication is that there can be no evaluative or practical outcomes.  What of technology? Technology must be involved if we are to define science as explanatory or predictive - we are given implications about what to do.  This is how scientific knowledge becomes power, giving us mastery over nature.  This is found in Marxism too with its hopes that human beings can be liberated from nature and necessity.  The possibilities might at last be turning into reality in advanced industrial societies.

Technology is therefore crucial, but it has been relatively neglected by analytic philosophers.  Instead, many contributions have been directed to ethics [examples 403].  Pure science has been distinguished from applied.  However, it is possible that interrogating technologies can imply value judgements in science itself.  Technology 'represent a crucial breach, from within science itself so to speak, of the supposed logical barrier between fact and value'.  We have to notes that technology is not just the use of knowledge for some practical purpose, a means to an end, but that power, control and domination are inherent [implied in experiment, we might argue].  It is power and control over particular objects just as in science, and the subject with the knowledge also has power over the object.  We know this from the potentially oppressive nature of human sciences, where people are particularly subject to power, for example in producing technologies to control behaviour.

This connection between technology and science on the one hand and practice on the other still might not be what Marx meant by science as dialectical: it looks 'essentially non dialectical', with the play of power operative only at the level of practical implications.  Criticism seems to have no place.  However, this is implicit as well, if we 'enrich' the basic model of the connections.  We do this by it considering relating one theory to another, and here it is perfectly possible for these theories to contradict each other or for one to be criticism of the other.  This is not a matter of evaluating technology nor of morals, since values are not always moral values - such a distinction breaks with the notion of reason as value free [where, usually, values are seen as moral ones].

Even Popper sees criticism is crucial to science, although he fails to develop the idea very far.  For example in his discussion of dialectic, he argues that it is a critical attitude which produces the antithesis, not an inner relation between thesis and antithesis.  It is ultimately our decision not to accept contradictions, which implies that contradiction is only a category of logic and is not there for evaluative purposes or critical itself.  Logical analysis is different from criticism for Popper.  However, to identify a logical contradiction is 'at least by implication, to criticise it' (405), and when we criticise a theory we are implying its possible acceptance by some actual or possible subject.  Popper's notion of a Third World of knowledge operates without a theorising subject.  Edgley wants to argue that logical criticism has always involved the notion of acceptance by actual or possible subjects, that logic is always connected with the more general notion of reason, and must do, unless we are to accept a merely contingent relation.

People, at least, can contradict themselves, hold contradictory views.  This 'implies or presupposes that in this sense there can be contradictions in reality'(406) because we are making statements about real people, and if we criticise them for being contradictory, we are making a statement that is 'at once empirical, logical and evaluative, i.e. critical'.  It is believing contradictory things that is 'logically impermissible'.  Thus scientific evaluation is not just a matter of logical implication, but a series of more complex value judgements, and contradictions can 'have a real existence in some subjects' thoughts and attitudes'.  Criticism should be understood as the activity of opposing, and this is crucial to the development of science, which is more than just the logical structure of theories, but rather 'the activity of arguing'.  Science does not just understand the world, but tries to change 'that part of it which consists of misunderstanding'.

It can be argued that science always criticises some other theory, and never the relation between theory and its object.  The relation is always descriptive explanatory and predicted - thus disputes between cosmological theories criticise each other, but they do not criticise the object [I am not at all sure this is so these days, where there seems to be some dispute about actually what the object is. Ironically, these disputes never seem to challenge the 'true purpose of science' itself].  However, even here, there is a social target for the criticism of theories, including various institutions.  Nevertheless, we can settle with the argument that Marxist dialectic is characteristically about social science - natural science might possess some crucial differences about the extent to which it is dialectic.  To press the argument for a dialectical natural science would mean accepting that 'the reality investigated by natural science has an underlying core ("essence") that differs radically from...  its phenomenal appearance' [which Deleuze does of course]; that this core is essentially in conflict; that the natural sciences develop theories as a result of determinate contradictions between them, so that new ones 'both negate and preserve the old'(407).  [ i.e. accept Hegel?]

In social sciences, the dialectic would address logical contradiction at the level of the object and in the relation between theory and objects, since it focuses on people in society, and 'people are peculiar as objects of science in being also subjects with their own theories...  about their activities'.  There is therefore 'a much closer logical relation' between these theories and social practices and institutions.  Contradiction can not only arise between thoughts, but also 'between actions and practices'.  It is in this sense that Marx said that ideas about practices and institutions reflect particular societies, since societies are reproduced and produced partly through the self understanding of their members.  These ideas appear in surface features of social structure 'and thus form an ideology that obscures the underlying realities of that structure'.  If we expose ideologies to scientific critique we can see that their attempts to appear consistent turn out to be self contradictory because appearances contradict this 'deeper nature'.  In this way, the critique of ideological social theories criticises the society in whose structure these ideas become realized - thus Marx's critique of bourgeois economy is a critique both of the categories of political economy and of capitalism itself. [So critique must involve a surface/depth metaphor]

What of Marxist materialism?  For Edgley, it asserts that the economy is basic to any structure in producing material goods and satisfying material needs.  Its links with ideas is discussed in the section on the fetishism of commodities.

Overall, Marxist social science practically opposes the 'basic self contradictions of capitalist societies'(408).  It is not a moral or ethical critique and nor does it advocate moral practices alone - the emphasis is on 'society's structural irrationalities' not the conduct of individuals, for example.  Nor does it think that ideas alone can produce social change, through the process of 'reasoning with and exhorting people'.  [So material crisis is needed as well?].  The Marxist notion of reason does not separate science from morality.

What remains of a role for philosophy?  Must it be only analytic and descriptive, describing the structure of other people's language and knowledge [- the handmaiden view].  But this whole argument has been philosophical, comparing science and the need for it to be critical with social science.  Philosophy is therefore part of the 'same general project of social criticism', aiming at fundamental and basic categories, including that of reality.

Chapter 20.  Addressing the Cultural System (Archer), 503 -43

Against conflationism, we need the dualistic approach of distinguishing between system and social integration.  Such integration is not to be achieved by pursuing some analogy with some structural analysis.

We begin by talking about what is distinct about the cultural system as opposed to the socio cultural level.  We can think first in terms of Popper's 'third world knowledge', a corpus of everything known, expressed in a common language at least in principle, so that they are intelligible.  This involves there being only one system at any particular time, denying cultural relativism.  The system has its own logical principles which are not the same as the judgements of social actors.

Popper distinguished subjective mental experiences on the one hand from objective ideas on the other, and we can see objectivity in the way in which some ideas influence others causally, through thought processes.  The products of these thought processes therefore 'stand in logical relationships' (505), and they are objective.  The actual logical relations may involve consistency or contradiction.  There is no 'causal harmony' with a socio cultural system, however, no universal integration, but an interface between causal relations and objective logical ones, and this is contingent.  The socio cultural system itself has different groups and individuals in causal relationships, leading, overall to a process of elaboration of the cultural system as a new ends with new logical relations to the others are introduced.  In other words we have 'a morphogenetic cycle', involving cultural conditioning [from the socio cultural] leading to cultural interaction leading to cultural elaboration.  It is a cycle because the end products can constitute new inputs.  We need two sorts of analysis though, causal for stage one, logical for stage two [I thought these were causal too]  --  a dualism.

This chapter focuses upon the cultural system and its logical characteristics.  It is necessary to establish that these are independent of any reference to the socio cultural level.  If we get this right, we can explain cultural stability and change.  It is necessary to describe them with a certain formality to avoid parochial contents.

Taking contradiction first, it is argued that this is an ontological relation not dependent on people's awareness; it's also claimed that these are known [epistemology] only by referring them to 'invariant logical principles'; methodologically we should establish ways to research the problems (506).  Taking the ontological claim first, the cultural system has an objective existence with autonomous relations among its components, having been produced by particular socio cultural interactions and having emerged.  It features constraints and possibilities, and can introduce new problems in the relations between the emergent entities themselves, or between them and different environments or particular human actors.  None of this necessarily involves people noticing contradictions, and the same goes for compatibilities [although putting theories into practice does involve individual capacities if not understandings]. Opposing views see contradictions as located solely in the socio cultural level, as when Winch argues that logical relations depend on social relations, a form of conflation and depending on meaning equating with the use: there is no way outside linguistic conventions.  However, social relations vary a lot more than logical ones do, and the use theory of meaning has been criticised, and can even be reversed, so that the usage of concepts depends on 'exploiting their lack of meaning, double meaning or ambiguous meanings'(508).  Thus meanings can be separated from use and have to be given 'ontological status'(508).

At the epistemological level, all contradictions always involve a lack of internal logical consistency, so that nothing can be both P and not P: the various claims made for particular relationships between items is not relevant, such as the certainty of a community of believers.  However, there may well be local variations, if we accept that intelligibility and logic itself varies, as with 'extremely "alien" beliefs'.  Generally this would show the effect of context or modes of social life even on logic, for example differentiating science and religion.  Claiming a universal logic is ethnocentric.  However, it is argued that even the most alien thought depends on logical relations, and apart from anything else, we assume this when we translate: we often proceed by trying to find the local word for 'no', for example, which implies shared principles.

Methodologically, identifying contradiction can be a problem.  If the cultural system is firmly embedded in a socio cultural context, the two cannot be studied separately, and that the items in the cultural system become contaminated.  These are serious objections requiring sustained argument.

[Ontology] Relativism has been revived as a strong program, proposing to examine the entire cultural domain from the perspective of the sociology of knowledge [the advocates appear to be Bloor and Barnes].  One argument is that causal analysis [in this special sense that one item in the cultural system causes another] would not separate out true and false beliefs, and that truth and falsity is provided by the social character and causation of knowledge.  This includes relational properties like contradiction.  Beliefs are relative and incommensurable, and truth and falsity only a local idiom.  Relativists even accept that their own proposals are relative.  Incommensurability cannot produce contradiction, there can be no transcontextual criteria, no translations.

In defence of the notion of contradiction, even Mannheim argued that mathematics and logic were universal, against Bloor and Barnes, who suggest that even mathematics and logic showed social variation, for example in the form of abstraction from physical reality that they pursued.  These can only be social conventions.  However, there is a shortage of alternative mathematics and logics, and most of the example supplied illustrate at best a certain variability, along with 'brute regularities' (512).  Again, the amount of mathematical variation is much more limited than sociocultural variability. [A discussion of Greek number theory ensues, turning on whether this shows alternative logics or just different conventional definitions displaying the same deductive logic.  The regularities include identical solutions, even though particular forms of mathematical thinking varied].  The same problem arises with demonstrating alternative logics: here the argument is that common usage very often does violence to formal logic, for example in declaring that something can be both X and not X.  Archer's response is to say that X can be not X if we are introducing, possibly implicitly, dimensions of time and place, and she denies that common sense utterances are fully propositional - thus the contextual bits can be seen as missing elements which if replaced would conform to logical rules.  Common sense often offers shortened responses in an assumed dialogue, and we need to restore them to the full version before we can decide if the law of contradiction is holding or not [detailed discussion 514-5].  The social basis for these conventions is asserted only against a supposed biological one.  Even when they discuss peculiar academic extensions to formal logic, for example like intuition or three valued logic [the Trinity is one example], these can all be seen as a simply parasitic on standard logic, and these examples are not unintelligible [lots more examples 516 F, include contradictory identifications as a form of irony, the use of metaphors and similes in non propositional forms like poetry.  The Trinity is discussed, and Christian doctrine is seen as an attempt to avoid contradiction, by advocating 'a semi propositional belief' in a mystery beyond human beings.  This has shaded over into the public understandability of the doctrine: the mystery and its solution only in faith really upholds the law of contradiction which would apply otherwise, and we can all see that this is necessary to defend the authority of the church].  There may well be some substantive differences across cultures, but again we would have to analyse them before we could say that they did not share the formal properties of logic. 

The bigger issue is translatability. [Several densely argued pages 518 -523 ensue on the issue of translatability.  There are serious problems with the strong relativist view here that says translatability is radically impossible because of the strong influences of cultural context.  There is agreement that pragmatic forms of translation can take place, where people understand each other well enough to decide simple issues like whether or not a cow is in a cornfield, but full translatability is not possible.  However if this were true, we would not be able to communicate with other people at all, but could only talk to them having been fully socialised into their culture.  However, having done that, we would also not be able to talk to anyone outside their culture, and we would even have problems relating our own internal languages and thoughts.  Other minor problems include knowing where to draw the line between linguistic differences and incommensurable cultural ones -is a minority linguistic community the same as an incommensurable one?  As an external perspective on this debate, I am reminded about some of the problems in modern ethnography in finding a linguistic community which has not been exposed to commercial or colonial conventions in the first place, which makes the issue a rather abstract one, ending in whether we could or not talk to aliens.]

There is still a problem about whether contradictions are universal or radically context dependent.  Archer argues that we can only answer this by separating out the cultural system from the socio cultural level.  The problem arises because it might be possible to argue that if we know the complete cultural context, there will be no contradictions [which is rather close to what she said before about spelling out particular everyday statements, and it also reminds me of the old slogan that to understand all is to forgive all].  We do have to know something about context to recognise a contradiction, say in a particular religious belief, but if we acknowledge this, there is a danger that radical contextuality will result: there seem to be no rules about how much context to consider.  The implication again is that Archer's explanations of cultural social change as a result of contradiction and inconsistency which is subsequently corrected would have a serious problem.

Sometimes, explaining cultural consistency involves social criticism.  This arises with 'incoherent "bobility" type concepts'[an excellent book I found on Google, edited by Clifford explains that this arises when privileged members of society claim that they are able to practice particular virtues, because the same name is used to describe privileged members and practitioners of particular virtues.  This arises in the context of describing the same Gellner article that Archer is considering here].  In other words, we can use social explanations to account for apparent cultural inconsistencies in beliefs [as in ideology] - so we argue that the Zande are forced to believe in witchcraft even though they are skeptical, because the beliefs support the power structure in society, even though this is not known to the Zande themselves.  In this way, everything can be rendered consistent as before.

The problem is solved if we distinguish between the logical/cultural and the causal/socio cultural, and Gellner fails to do this adequately [perhaps because he does not wish to grant any independence to textual analysis at the cultural level].  Archer prefers to discuss contradictions in beliefs as a matter of connection strictly with other ideas, and here, logical contradiction is clearly possible [the example relates to Berber society, where various blessed people are supposed to be both rich and display great generosity and indifference to wealth, 527].  However, we do not need to explain this away by reference to the sociocultural level immediately, even though this level explains how the blessed people actually cope with  these contradictions: what the beliefs mean in practice is not the same as what they mean at the cultural level.  In practice, the blessed ones offer lavish feasts but also solicit donations from pilgrims, but Gellner deploying this practice as a solution to contradiction loses detail about how social practices actually operate.  Strategic action is not the same as logical relation, and it is wrong to include them both in 'context'.  Both offer meanings to participants.  It is true that cultural inconsistencies need to be made tolerable in daily life, but we need to keep separate how these two levels are managed and how the two levels interact - for example, if daily life fails to deliver enough donations, will this feed back to the notion of the selection of the blessed by god?

It is a mistake to argue that the socio cultural level mixed with the cultural explains what people really mean [in total, as it were, facing both culture and daily life].  This cannot be solved except by arbitrarily defining context, since there are no rules.  If we separate the two levels, however, and deployed different methodologies, we can make progress.  Combining the two levels tries to do too much at once, and we need to proceed more slowly and analytically, methodologically isolating the cultural system first.  This involves seeing the cultural system itself as separate, so that the logical relation between ideas only refers to other ideas in the cultural system: here there can be contradictions [I think she is arguing that we can subsequently turn to the socio cultural to look for strategic solutions.  There are some debates about sayings and meanings that are far too esoteric to follow.  At one stage, Archer's position  seems to depend on arguing that cultural categories actually have '"core features", for example, "focal colours", "natural prototypes", "best examples" or "basic level objects"'(530), and that these anchor the variety of meanings that are found in cultures]. There is a temporal dimension again, but this does not deny change going on through usage itself, simply asserting that it must take time.  We have to reconstruct this temporal sequence to understand what looks like the 'instantaneous registration' of the effects (530) -morphogenesis at the micro level.  Any differences at the socio cultural level could also produce a new set of resemblances in languages, provided they were not too different.

Contradictions are only found where ideas relate to other ideas in a relation of logical dependence of some kind.  The simple example involves an apparent contradiction that water boils at 100° C with an empirical finding that it does not do so at the top of the mountain, and we solve this by adding conditions to the first proposition, to allow for standard temperature and pressure.  Note that we have first of all turned ideas or beliefs into propositions, and only propositions can display contradiction or consistency, they cannot be ambiguous.  This might reduce the content of cultural systems, including those beliefs which do not come in the form of propositions.  However, Archer proposes that we can render these other beliefs as propositions, drawing upon the cultural context to do so.  This is an improvement at least on cultural apologists who can find 'intelligibility from the oddest sayings' (532) especially if this is done by sampling cultures or texts and then abstractly spinning off meanings.  This is as erroneous as extracting individual scientific principles of the past, like the concept of phlogiston, and putting them in a modern context. We have to extend the same considerations to what subjects themselves think about this socio cultural environment, allowing that this could arise from manipulation by other people.  However, this would explain some of the origins of the proposition, and still leave logical relations between them as 'emergent properties and thus irreducible' (533) to the social or the biological.

So we focus on propositions and look at the relations between them.  These relations might well include items that are not propositions themselves even though they are meaningful [for some, meaningful only to an observer].  In practice, it seems that ordinary people hold factual beliefs separately from representation or ones which are not propositional, and it is a mistake to treat them at the same level, as if all utterances are equal and all contribute to some homogenised world view.  Empirical investigation is required to disentangle the representational from the factual, and to keep distinct those that originate in the socio cultural system.  At least we will take what people say or write seriously, although we know that public sayings and private meanings might be different, but even here what people say tells us about interconnections with the cultural system and what is logically entailed by it.  We can argue that two items are contradictory if no other cultural item can reduce them, whether or not the population involved is aware of the inconsistency - these are objective contradictions even if they might not actually trouble anybody, as with the Azande who are aware of incoherence: the socio cultural system is not always involved in reconciling contradictory propositions.

Indeed, strategic coping or whatever does not resolve a logical contradiction, and logical relations are independent of causal ones.  In practice, 'the SC level crucially affects the CS level' (535), but over time, permitting an analytical separation.  This also explains the way in which 'the items populating the CS realm have escaped their creators and have logical relationships among one another which are totally independent' (536) [so culture follows the same kind of development as science].  Managing any inconsistencies might have effects, as when a new theory replaces an old one, again this might operate at a logical level or at the level of relations with ideas, without any connection with the socio cultural.

Overall, the socio cultural level is not just the context of the cultural system.  The effects of the socio cultural system are important, but this is distinct from any contradictory relations in the cultural system itself - these have to be lived with and dealt with, and how people do this is is an  important issue in its own right.

READER BEWARE! WE ARE NOW ENTERING DEEP WATERS (UNLESS YOU ARE A PHILSOPHER)

Part IV

Chapter 22.  Introduction.  Dialectic and dialectical critical realism.  (Bhaskar and Norrie) 561 -74

Baskhar developed a realist account relating to the sciences, but then went on to pursue the notion of dialectic through Marx and Hegel.  This helped him place critical realism in the tradition of western philosophy, without revising any of the earlier concepts [typical philosophical project].  This would 'provide a philosophical basis for marxian social theory'(561), and there are ethical implications which might help develop a suitable praxis.

The work proceeds to add to the Hegelian dialectic by suggesting that non identity should be the starting point, and adding the notion of transformative agency.  This will enable further commentary on the terms of critical realism, introduce additional terms such as 'tense and process' and also add a dimension of emancipatory practice.

The notion of dialectic is to be extended as a logic of argument, a dynamic of conflict, a mechanism of change, and 'the axiology of freedom'.  To enable all this to be reconciled, the concept of absence becomes important, and the dialectic is seen as ways of overcoming [negating] absences or ills/constraints [to give an example of the horrors to come, the actual phrase here is that dialectic offers 'the absenting of constraints on absenting absences or ills'(562). This is done to preserve a general function for absence? Same as Deleuze noting different differences? Twats!] The emphasis on the absent helps counter the earlier emphasis in classical philosophy of the positive, which at the most general level, assumes some positive actual notion of reality, 'ontological monovalence'[so the absolute does the same job as an insistence on proper difference does for Deleuze?].

Overall, we have a dialectical process with four moments: the first moment (1M) leads to a second 'edge' (2E), then to a third level (3L) and finally to a fourth dimension (4D). 1M  involves the critical realist concepts that we know already  - 'structure, differentiation, change, alterity, transfactual efficacy, emergence, and systemic openness'.  These are developed at 2E via dialectical concepts 'such as negativity, negation, becoming, a contradiction, process', development and decline, mediation and reciprocity'.  These are presupposed by the concepts at the first level, which are limited to analysing science - 'the marriage of critical realism and dialectics opens up the prospect of an understanding of the way sciences must be if they are to be adequate to "cosmology...  human geohistory...  personal biography, laborious or routinized work but also joyful or idle play' (563).  The second level implicitly invokes the third level, since particular absences or omissions at 2 requires an explanation of some greater totality, or a follow-up theory that explains it.  The dialectic leads to this notion of the totality and to 'constellationality' [borrowed from Adorno, it seems , and referring to clusters of matter or events associated in  various ways?].  Further internal forms appear at this third level: 'collectivity, relationality, reflexivity, concrete universality, subjectivity and objectivity, autonomy-within-duality and hiatus'.  Once we recognise these internal forms within constellations, we might be able to develop emancipatory knowledge or agency, in the form of 'new philosophical accounts of reason, rationality and...(practical wisdom)', and this will lead to the fourth dimension where the theory and practice are unified in practice.

The actual book [Dialectic: The Pulse of Freedom] explains these themes, with the centrality of absence, emergence and contradiction as reconsidered through a discussion of Hegel and Marx [see below].  Absence becomes a central component of positivity, argument 'absents' mistakes in thought, while social change absents various blocks.  Dialectic therefore 'becomes the "great <loosener>, permitting empirical "open texture"...  structural fluidity and interconnectedness"' (564).  Emergence shows the interplay of absence and positivity, and at this level produces various kinds of 'new beings', which could not have been induced or deduced, 'matter as creative or autopoietic'.  These new entities have causal powers and can demonstrate qualities of 'meshwork' of material being [same phrase as in DeLanda] .  We also step outside the normal kind of space time to discover new 'processual rhythmics'.  This explains the emphasis on geohistory.  Overall, we can also see these extra characteristics present in the here and now [which will loosen then 'absent' any empiricist misunderstandings].

So, absences or real negations exists together with positivity as a 'bipolar dual',and this idea is applied to various philosophical positions including those of Hegel, Marx, Kant and some dead Greeks. Contradictions become 'connections between entities...such that they are in principle distinct [ie really exist] but inseparable [opposed]'.  Implications are raised for analytic philosophy and formal logic.  Hegel is rebuked for thinking that everything can be reduced to logical contradictions, which assumes that being is logical: for Baskhar, contradictions arise from different levels of reality and causal contexts, and appear in actual geohistory, various 'structural and causal contexts', (565) and rhythmics.  In this way, dialectical reason goes beyond analytical reason, and logical contradictions are treated as a sign or a site of real contradictions.

Having moved from 1M to 2E, a further clarification of possible 'dialectical motifs' can be explored.  'Mediation', for example suggests an intermediary which manages the bipolar nature of being and absence, and this might include 'ideas of internal connection and particularity within totality'.  'Duality' combines both 'existential' [really existing] interdependence and essential distinctions, for example between the specific and the general, agency and structure, freedom and the conditioned.  This is developed into notions of 'hiatus-in-the-duality', where autonomy is asserted against reification or other kinds of collapse or dislocation, and 'perspectival shifts', where we can emphasize one pole rather than the other for explanatory purposes [for example agency and structure for sociology].  'Constellationality' argues that things are necessarily connected, as with the dialectical unity between dialectical and analytical reason, where the first builds on the latter to over reach it, and the latter [assumes or presupposes the former?].  Without dialectic, analytical logic apparently can deliver lack or absence, and this sometimes can be used politically as a defence or compromise - the  TINA syndrome [younger and non UK readers might not have remembered the British Prime Minister Margaret Thatcher and her constant refrain that 'there is no alternative', TINA, to her appalling policies].  Here, political manipulation takes place to manage the contradictions, equivocations, '" extreme plasticity...and rational indeterminacy"'of particular formations or ensembles.

We can penetrate these manipulative formations by understanding them as part of a deeper, more complex context or structure, as part of our '"drive to totality"'.  'Totality' is the final concept to be considered at 2E and it helps us move on to 3L, just as 'absence' suggested 2E in its turn.  The totality consists of embedded ensembles reduced by their geohistories, interacting so as to produce change, in an 'open and potentially disjointed process' as a context.  It goes beyond our normal conceptions of 'identity, or causality, space and time'[but is still not a virtual level as it would be with Deleuze?  We are still close to DeLanda's conception of different sorts of phases producing different states of matter, rooted in vectors of forces rotating around attractors?].  In particular, we have to go beyond 'identity thought', and sees things as necessarily related to other things, a concept of a '"entity relationism"': again this will mean moving beyond analytic thinking.

We can pause at this point and think about the implications for the more familiar notions of critical realism.  Baskhar now sees these as a preliminary statement of the full dialectical position.  However we now need to add to the notion of human nature as a 'four-planar social being, or the "social cube"' (566).  Considering social life as a totality will lead us to grasp the four dialectically interdependent planes: 'material transactions with nature, interpersonal action, social relations, and intrasubjectivity'.  These elements are multiply determined and mediated, sometimes contradictory.  When we grasp the totality, we can extend ['dialecticise'] our earlier ideas about the transformative potential of social agency and the system of positions and practices.

We also need a new conceptualisation of power as 'generalised master - slave - type relationships', with a clear reference to Hegel on the master slave dialectic, but also to Marx on the idea of 'wage slavery, ideology and the fetishism of commodities', and, by extension, to all socially structured power relations including those of gender and race [power here is defined as being able to get your way against the wishes or interests of others].  So we have substantive arguments about inequalities, as a background to dialectics or practices.  We can also see the effects of historical power on western philosophy [and this is argued apparently by seeing analytical philosophy as somehow expressing 'an ontology of stasis'.  This connects it with the other areas such as ontological monovalence].  Baskhar uses this critique of philosophy to argue, that it is somehow systematically connected to current power relations, including 'money, instrumental reasoning, reification, alienation'.

Critical realism acts as an 'underlabourer' for science and social science, explaining what the world must be like for them to be possible, but it has always had an ethical dimension as well, hinted at earlier in some of the work on how critique is emancipatory, and how we can bridge the gap between facts and values [by ideology critique]. This is also developed [I might take some notes on the specific chapter unless I have hanged myself by then].  Here, we are at the level of 4D, which will lead us to a moral theory about 'universal human flourishing', connecting the flourishing of an individual, to the relations established by the flourishing of all.  We can see that emancipation is a possibility, immanent to the structure of reality, but currently constrained by power relations at the global level.

The ethics is grounded in Bhaskar's ontology, and the notion of 'alethic' truth [had to look it up again -- 'truth in the world' as opposed to epistemic truth in the mind of the knower].  There seems to be a test suggested here for such a truth, the connection with theory and 'the absence of heterology (unreconciled otherness)'[so we know we are close when we have explained all the apparent contradictions and absences].  It is this sort of truth that drives the 'dialectic of scientific discovery', as a 'non arbitrary principle'(568).  It is but a step then to argue that the true equals the moral good,  which equals freedom and ' universal human emancipation'.  The same ontology produces grounds both for the cognitive and the normative aspects of human being as it operates on its four planes.

The alethic is central to a dialectic with its attempt to show that 'any ill' can be conceived as an absence or a constraint [and the dialectic attempts to identify these and overcome them], and this can be seen to apply on the moral level as well, where a human freedom and the pursuit of the good life can also be seen as dependent on negating constraints and ills.  Again, the moral is not tied tightly to the factual, but embraces the transfactual, unlike other systems that are shackled to ideal speech situations or some original statement of human society.

The dialectic also heads for 'universalisability' by its very mechanism aimed at negating constraints.  In this way, aiming at the truth necessarily implies human emancipation, since power mechanisms are constraints [I simplify quite a lot here].  The desire to overcome constraints implies a move towards knowledge of all  four planes of human existence.  [The argument seems to be that if we intend to negate one constraint, this somehow commits us to negating all of them in similar circumstances, and thus eventually to negating {'absenting' is the annoying term used throughout} all constraints as such, including those introduced by power relations.]

So we have an argument based on 'moral realism and ethical naturalism'(569).  Again morality is seen as something objectively real, alethic, and opposing this to actual moral systems leads to critique.  This in turn implies that we can study ethical systems and moral properties using the social sciences, and we can do this critically, showing existing systems to be limited or false.  This will also help us to evaluate any actions designed to negate them.  As before, falsity implies a critique, including critical analysis of any actions based on falsity.  Critique should also lead to practice ['immediately', the argument is].  There is also implied 'the conception of emancipatory axiology', and emancipation can also be linked to the failure to satisfy other '"axiological needs, necessities and interests beside truths...  Basic health, education and ergonic efficiency"'[in other words, we take these as fundamental axioms of human existence.  'Ergonic' means to do with work?].

So a universalising move is inherent in the argument.  Baskhar argues that any true judgment has to be oriented towards concrete individuals, as well as being universally applicable to any other individuals.  This is not taken as an imperative, but rather an argument that Kant called 'assertoric' (570) [seems to be some sort of reasonable generalisation.  It takes place when the analyst stands in the place of the concrete subject and expresses solidarity and commitment which takes the form of placing concrete subjects in a theory of human nature needs or interests].

However, it seems to be the moral dialectic, aimed at freedom that drives the search for truth, still based on an ontology of absence and a desire to absent it, a feeling of constraint or unfulfilled needs, 'elemental desire'.  All these involve a certain ability to detach from the experience, recognise the difference between the desirer and the desire, and this in turn implies the whole 'intransitive world' of causality, ontological stratification and alethic truth'[this is Bhaskar's own example about how thinking about the absences attached to desire is a way into dialectical thinking, just as good a way in as speculating about the scientific experiment].  This implies that desire is not abstract, but something related to to specific things which it needs and lacks.  At the most general level, this will involve understanding the lack of freedom, of all kinds, from the freedom to do otherwise, to more 'negative and positive' forms of liberty, moving from unwanted and oppressive states of affairs.  If we follow the dialectic, we will see that these are universalisable.

We therefore have a critical morality directed at lived experience and constraint, suggesting a process to develop inner urges emanating from desire, encountering blocks, especially in the form of power relations.  We can see this drive to freedom as an 'ontic pulse', both real and moral, and Bhaskar thinks it is ultimately '"irrepressible"'.

By way of criticism, Collier [who has a chapter in this collection, although, irritatingly, it has a page missing] notes that the whole philosophical heritage is being challenged here.  At the core is the doctrine of absence, which is not grounded on human being and which offers causal and perceptual criteria for recognizing it.  But at the same time, Collier sees Baskhar as entirely political in his argument.  Collier attacks the idea of universalisability of morality, which is rendered as like a problem relating theory to practice.  Why should such contradictions demand autonomy?  It is also the case that we can universalise particular acts in ways which produce different moral maxims.  This leads to the argument that there are some facts we do not wish to universalise, including 'the exaltation of the fascist thug'(571).  Collier thinks that each case should be considered before we assert a general universalism.

Baskhar could respond by saying that of course we must see dialectical universalisability as related to the concrete moment, including its implications on the four planes of social being, and its power relations.  We need to establish this sort of social scientific knowledge to explain agents in their real contexts, and this will help us to judge universal statements and statements about what they ought to do - thus we can explain the reactions of the fascist thug [as ideology] and would not want to universalise them.  Baskhar would also argue that the best and most adequate description of a particular social act, an explanation grounded in the most adequate theory, would provide a suitable basis to make universalised statements 'assertorically'(572).

Norrie, a criminal lawyer, has developed Bhaskar's work, originally through the notion of explanatory critique and its links with praxis [its function as 'praxiology'].  He saw the evolution of legal forms as entailing concealed contradictions, and thus expressing power relations.  In this way he, he was able to draw upon Baskhar and critical realism [and one of his chapters is included].  Norrie's approach can be seen as 'dialectical phenomenology', focusing on 'personalism' as an ideology.  Norrie develops the moral basis of legal ideas of justice and their contradictions, including an notion of individual responsibility which repress the social responsibilities, and argues that we need a dialectical approach to understand the location of individuals in social relations.  Like Baskhar, he uses immanent critique to expose liberal theories of justice and what they suppress, and then traces limits to social history.  We have a dialectical emphasis on contradiction instead of legal analytical reason.  Particular laws can be seen as TINAs.  There are hints of the four planes of social being.  There is also an argument that political practice is required if criminal justice is to develop.  Another strategy is to point to other systems of law to show absences within ours, and to trace legal reform in terms of a pulse to freedom.

This is in turn generalised to look at whole systems of legality, seeing it as partly autonomous and partly as a result of 'existential violence'.  There is a plea to develop a totality including the legal system and the social structure.  Overall, the law both emancipates the legal subject and suppresses alternative forms of social organization - it is structurated, and the notions of duality connect with this argument.

[Friends I am not at all sure how to manage this next bit.  There are two chapters by Bhaskar himself, and they are massive, very densely argued, full of references to philosophers I have never read, and spattered with Greek terminology.  You would just have to read the detail for yourself.  It does look ingenious, but it is beyond me.  Instead, I can offer you a brief summary and gloss of the chapter on critical realism and dialectic (chapter 23) and then some notes of the bits I understand.  Then the same for the chapter on dialectical critical realism and ethics (chapter 24).]

Chapter 23.  Critical realism and dialectic (Bhaskar) 575-640.

[My gloss.  Bhaskar is positing the existence of a complex totality and he is using this to rebuke the procedures of normal science and philosophy, which have overemphasized only one part of this totality or only one process within it.  The totality itself has stratified levels and particular regions where some processes are at work or have been completed, while at other levels and regions, new possibilities are emerging, and in some levels and regions, this process has barely begun.  These are the characteristics of the open system which can never be closed, partly because human beings themselves are emerging as agents.  It is these characteristics that explain many of the features already discussed such as counterfactuality, the differences between the real, the actual, and the empirical and so on.  The actual clusters of real things that emerge can be produced by the operation of causal laws, or organic developments, or they can even be purely contingent.

The processes themselves can be quite complex, involving the negation of absences at the most general level.  Absences can be material or logical, so we can move beyond Hegel with his emphasis on logical contradiction, and beyond Marx who wants to privilege particular kinds of economic and political contradictions.  What we get is an ingenious attempt to incorporate all sorts of major philosophers and to see them as offering incomplete accounts of this totality. For example if we use the concept absence rather than negation, we can specify that there might be all kinds of absences or just logical ones.  The same goes for different sorts of emergence. The whole argument outlines a myriad of possibilities, in thought alone, swept along by the dizzy implications that occur -- classic philosophical delirium. The problem almost becomes one of having to limit the possibiities to proceed at all, except as a series of  critiques of other positions as necessarily insiffuciently complex. Bhaskar's justification for his own approach is really weak in the end I think -- a good explanation preserves ontological depth and seems to offer a good enough account of something.

Reading through this stuff, I was put in mind of the equally 'delirious' Deleuze, who also has a complex totality producing the actual and the real, although of course one of its levels is a virtual one.  I'm still not sure if this is the case for Bhaskar.  Deleuze also talks about different sorts of conjunctions which join things together, although it seems to me that he emphasises the purely contingent as in the haecceities, and leaves little room for any sort of causality unless it is one of those mysterious processes which he talks about in terms of 'dark precursors' and  resonances.  In general, the concept of absence seems to work like the concept of difference for Deleuze having a major critical role in denying the adequacy of purely positive accounts, including positivism, but also the view that reality is benevolent and as it were working with human beings to reveal itself.

Now the notes. The actual sections in this collection seem to be extracted from a number of chapters in the book...]

Section five Prima facie objections to critical realism.  [sections 1--4 are in an earlier chapter] Critical realism needs to extend away from just explaining science.  Nevertheless, the analysis of science is useful producing a way of developing a 'metacritical analysis of philosophies' (575).  By thinking about science, we can also make more concrete some of the general propositions of philosophy.  The same goes with ethics and its transformation into politics.

We can now see that critical realism develops by reflecting on philosophy, including Heidegger [mercifully in another chapter], involving 'reflection on the presuppositions of the pathology of every day life' (576).  Science is not to be particularly privileged and there are other dimensions of human activity.  Nor do the objects of science comprise reality, only a particular slant on it.  Finally, there are aesthetic and moral realities which are equally important.  Overall though, we should not just confine our enquiries to 'that which lies within the bounds of human cognitive competence' at the moment, since 'reality is a potentially infinite totality, of which we know something but not how much'[that is an open system which can never be closed off by say Hegel or Lenin].

Section six. On the sources and general character of the Hegelian dialectic.  Hegel takes dialectic to be both a logical process of reason, and as offering a dynamic, a process involving the practice of negation.  The term was originally derived from the Greeks, referring to conversational discussion, or a method of philosophical splitting [lots more detail to follow].  Eventually, though, two processes emerged: 'historical determination by rationalist epistemology, structural domination by empiricist ontology' (577).  Kant attempted to settle the split by turning to a transcendental subjectivity to satisfy rationalist demands, leaving empiricist criteria applying to being, even if these were then unknowable.  This tradition left a series of dichotomies to be addressed by Hegel, who tried to restore the notion of process, the way in which reasoned precedes by generating itself, differentiating itself, and then particularising itself.  The process could work in an ascending or descending form [from universal to particular and the reverse].  The ascending form points towards god and Spirit.  Together, they offer a pattern of unity, loss and return.

Hegel attempted to overcome the earlier dichotomies by referring to an expressive unity, 'a monism'(578), but one which recognised diversity and the importance of subjectivity, at least before transcending them in the final identity of being and thought [more details 579].  In effect, this is a kind of '"constellational identity"', however.  Hegel wants to let the idealist terms envelop or contain the materialist ones, while preserving their distinctiveness as a necessary stage.  This helps an accommodation with the present, because we can see it as something pointing to reason.  Thus reconciliation is always present as a possibility even in the middle of the most apparently separate events or identities.  Hegel demonstrates this in a number of these works (580).

Turning to the motor of the process, we need to discuss the relation between actual things as something other than what appears as a relation of opposites, or terms which cancel each other.  This is where we get transformation from, a never ending immanent process of reconciliation, not mutual canceling, despite what seems to be the case with empirical science [and formal logic].  Indeed, empirical opposites are necessary for overall progress, pointing to the need for something more comprehensive.  Opposites can not be accepted using analytic logic, hence the need for dialectical logic.  Once we use dialectic, we can see that apparently eternal opposites are really in a relation of 'reciprocal interdependence'(581).  Material opposites are then united with dialectical reason in a 'constellational identity'.

We can see the progress of this 'determinate negation' in our experience of past processes.  That experience often takes the form of demonstrating 'a theory/practice inconsistency', to be resolved by speculative reason, and then subsequently challenged in the next phase of the cycle.  If this is happened in the past in experience, we can no longer easily identify contradictions without some sort of transformation of our understanding - this is the 'Alpha transform' [we have met it before as a transformation of everyday understanding into theory.  Bhaskar uses the Greek letter for Alpha, the pesud.  I'm going to use the Latin A], We then need another 'transform' [I'm going to call it a B transform] to resolve these difficulties in theory.  This is not often understood as a part of Hegelian dialectic, although Kant nearly got there.  Bhaskar admits that these transforms 'are my own gloss on Hegel'.  [Then a strange bit about why Hegel thinks that Reason is superior to ordinary reason - because it can work both transforms.  It implies that ordinary reason itself is a transformation of common sense, which we are now going to call the P transform -- I thought that was covered by the A transform].  Hegel insists that we have to translate prereflective experience into ordinary reason and then Reason, before returning to it - the V transform.

In this way, we pursue the truth through a process of Reason, and we must pursue this process to avoid the errors of incompleteness.  The process itself generates contradictions first before incorporating them, but nothing is ever lost in this overall transcendence.  Hegelian dialectic brings out what is implicit, and then fixes problems with ordinary reason.  Bhaskar wants to describe both of these processes as 'real negation'.  Only the A transform represents any kind of social or political progress, which might be located in being itself.  Dialectic goes beyond reflective or analytic thought, however to show how concepts and forms of life are systematically connected, not just different, and how existing states can be seen as about to develop in some sense, in a process of becoming.

Section seven.  On the immanent critique and limitations of the Hegelian dialectic.  We can already see some links with Bhaskar's own position, but we need to distinguish local from global dialectics.  We can take the idea of a determinate negation which preserves the earlier characteristics as representing a principle of 'stratification, structuration or superstructure formation', [later] and accounts for emergence.[Very difficult to understand here, referring to terms used in earlier chapters]. The process can be seen as a negation or absence becoming apparent to the understanding -a real negation -then that understanding being transformed in order to be resolved at the dialectical level - the transformative negation.  Bhaskar wants to say that these are still real negations [presumably using the same argument as when he insists that anything transitive is also real?] The transformative negation 'absents the absence' at the lower level [I still think this is terribly clumsy terminology], appearing as 'the negation of the negation'.  However, the initial concept is retained or incorporated, its inherent tensions or inconsistencies identified, especially in the form of a tension between theory and practice.  Subsequent progress proceeds to the next stages in an endless process of bracketing and retention.  This expands the conceptual field.  Only absences are lost.

The end is a totality which is 'constellationally closed, completed' (584), but which still reveals a sequence of stages which have led up to it.  At the final stage, where spirit becomes fully self conscious and absolute, the process ends.  The future that ensues follows a different process, which does not require dialectic reason.  In practice, it is the this notion of the totality which really informs more of the local practices of the dialectic, rather than emergent problems, making the analysis retrospective [only the end can we see the deficiencies in the earlier stages].  Thus Hegel is not really separating out properly immanent tensions, and we should really distinguish between 'genuinely autosubversive negations' and retrospective ones, with transcendental tensions.  In the Hegelian totality all these contradictions are internal, neglecting external constraints and contradictions, and this has limited marxian social theory too.

Seeing a fixed end in a totality also means Hegel's theory overemphasises linear processes, at least at the global level.  In practice, at the local level, there is 'an incessant variety of perspectival  switches', reflecting his more local concerns to deal with a range of phenomena.  Processes other than linear ones are clearly possible as in 'recursively unfolding matrices, gestalten, or any of a variety of topological modes' (585).  Linearity is introduced by the narrative form of the argument, but we should separate this from the epistemological issues.  The actual processes seem to take three possible forms: they are fully self generative and autonomous with no external source; they offer a process of transformative critique, demonstrated best in the development of Hegel's own thought in reaction to a number of other philosophers; there is a dialectic in reality itself which is only described by philosophical dialectic, as in the Phenomenology.

There can be good and bad radical negations [applied, curiously to the immanent and retrospective negations described above], and good and bad totalities, depending on whether totalities are open or closed, constellationally or otherwise.  Hegel wanted to close his totality to avoid infinite regress, but an open totality might not simply repeat same sorts of features.  The thought of an open totality is easier to manage  anyway [!].  The thought of an open totality is more comprehensive, and can contain a closed totality as one possible case.

The dialectic precedes mostly through the A transform and then the B transform.  Bhaskar wants to see these as particular cases of  transcendental argument that can be either retroductive/desending, or explanatory/descending.  They do not belong exclusively to idealism.  The A transform in particular is the most basic form of any kind of critique, immanent critique.  Hegel wants to see these processes confined only to thought, but there are implications for practice.  In particular, 'if reality is out of kilter with the notion of it, it is reality which should be adjusted, and not its truth' (586).  Practice will unite theory and practice [hence the 4D stage].  If we hypostatise thought, we separate ourselves from the practical elements of reality.  Thus Hegel is incomplete.  Reconciling everything into thought restores positivity, so 'Hegelian dialectic is un-hegelianly-dialectical' (587).

Hegelian processes are also special ones.  The various transforms imply possible 'indeterminate and underdeterminate negation', so linear radical negation 'is clearly untypical', and more and more determinations are required in actual examples.  There is also the case that 'real transformative negations in geohistory' rarely preserve earlier stages, given that most actual beings are finite anyway.  If we are to preserve all the qualities of the earlier elements, we are surely also preserving error or partiality, and this must be cancelled if proper sublimation is to work.  In any actual cases, 'it is clear that something has to be lost, even if it is only time'.  It is impossible to predict whether sublimation will would not occur, and other possibilities include different forms of reasonable resolution, including, for example, 'a distinction between essential significant or valuable characteristics and those which are not', or an agreement to exercise mutual recognition.

It follows that a proper general theory of dialectic will have to include different possibilities, including 'non-resolutary results, and non-reasonable resolutions, non-radical-preservative-determinate-negational reasons, and non-reconciliatory radical preservative determinate negations'(588).

Dialectic

Section one absence.

Real negations are more common [or important] than transformative ones or radical ones.  We can call the real negation the absent where  it is either something inexplicit or incomplete.  Hegel did not have the final stage of practice to remedy the absences in the real: it is more than a matter of just overcoming incompleteness.  Using the more general terms we can argue that absence leads to inconsistency which itself leads to greater completeness in an endless process.

Real negation involves an absence in some 'more or less determinate entity, thing, power, events, aspect  or relation, etc.' (589).  It can be something as simple as tool missing from a workbench, something which in this first case which is clearly intransitive, whether or not it has been identified by an observer, but there are clearly 'less determinate kinds' of absence, especially if we are allow the region [the workbench in the above example] in question to be indefinite or open.  The missing entity may not be observable.  The absence might be 'real but not actual' (590) -the absent thing might be simply not normally confined in a region, or it might be distant at the time.  Absences can be related to the past and the outside, and also connected to the presence or absence of other things.  It might even be difficult to say whether something is present or absent, especially 'at the boundary of the space - time region'.  These general qualities mean that real negation or absence is 'the more basic category than transformative negation (change)'.  Absences can also be seen as 'process - in - product'[I think the example mean something that was once important but is now being lost in time], and 'product - in - process', something emergent or divergent.

We can be misled if we think that negation involves an action to absent something, because absence also covers what can be seen as '"non being"'.  This also exists, and might even have 'ontological priority over being so that 'in short, negativity wins'.  Stressing negativity helps us realise the importance of contingency when it comes to examining entities that actually exist, unlike ontological monovalence which stresses positivity.  We need this to argue for the emancipatory qualities of the dialectic [to be explained later]. 

Certainly, reference alone does not presuppose existence, especially positive factual existence.  There can be intransitive ['intransitive objects of specific epistemic enquiries'] 'ontics' (591), and Bhaskar calls the positive ones 'onts' and the negative ones 'de-onts'.  [A lot of incomprehensible philosophical discussion ensues in terms of how propositions might refer to or imply the existence of 'onts', 591.  The essence of it seems to be that we normally think that references have to detach themselves from a referential act to claim an ontological dimension, and this implies 'existential intransitivity'- the example is the objectivity of scientific references.  A theory of truth is implied.  Bhaskar wants to operate with a more general and wider notion, again, to help him refer to non existent things, and also things that are outside science, such as 'laws, powers and tendencies…  totalities, relations and aspects'.  There is some nothing to which we can refer, including 'fictional discourse'].  When we attempt real negation, we might be presupposing some factual version of negativity itself.  This is also found in other discourses including fictional ones [apparently, there are three modes in which we might say something about negatives, 592]. 

In any event, real negation 'is vital to dialectic', in its full sense, allowing for 'de-onts'.  [Seems to me the argument is letting the end determine what we must find at the start]. We have to grasp absenting processes if we want to explain change, and if absences are seen as constraints, we need to consider them if we are interested in 'the logic of freedom'.  Argument itself depends on the ability 'to absent mistakes'.  Absences can be transfactual, actual, static, internally related, an inaction.  Dialectic proceeds by isolating the absence, often in the form of a 'theory/practice inconsistency or irrelevance', and understanding it in terms of dialectical universalisability.  This is how we have moved from 1M to 4D above [argued in detail 593, with an incomprehensible diagram].  In particular, normal philosophy has lacked 'the concepts of structure and heterology, concretion, relationality  and totality, agentive agency and, above all, absence itself'.

Real negation is also important to counter '"fixism"'[Deleuze's objectivist illusion?  Seems to be applied particularly to the notion of the subject here, especially the subject as a source of rigid and arbitrary definitions or predicates.  It looks like this rigid subject, separated from the predicates it offers, is going to be a crucial part of a formal logic which rejects the notion of contradiction].

One extreme case of absence 'is never anywhere existence', and something which refers to non existence is normally excluded from science, for example by Popper [because statements cannot be falsified].  Absent things have played a part in the development of sciences, such as phlogiston or the aether, however, but in most normal science, existential questions take on a particular force in specific contexts, where there are particular criteria [as in whether or not the Higgs boson existed?].  However, generally, when we falsify, we are clearly identifying and fixing mistakes, something 'absentive'(594), and assuming that error is necessarily ['paradigmatically'] dependent on absence and that we need to absent absence. 

Indeed, this sort of dialectic process is found in 'every learning process'.  It is implied in human agency in the possibility that states of affairs can be created, sometimes by acting on what is given.  It follows that epistemic change, and changes in beliefs, must also be possible and necessary.  This is where we get to dialectic as 'the great "loosener"', stressing 'structural fluidity and interconnectedness' in [heroes here are Marx and Bakhtin].  Reflexivity implies acting which implies absenting.  If we try exclude absences, we are engaged in 'performative contradiction', itself a contradiction between theory and practice.

However, it still might be possible to argue that the positive is more important, some 'matrix of factual discourse', possibly crucial in identifying absences.  However, the very identification of something positive involves an absenting [in the sense of a selection of material neglecting some aspects of the context], so here we are deducing the category of absence transcendentally.  Also, we need the notion of negatives to define what might be positive.  If we do not accept that a mixture of the positive and negative exists, we are left only with an 'eternal, all pervasive token monism' (595), and this will exclude becoming.  We can now see causal determination and change as a matter of transformative negation - 'all causes are in space-time and effects are negations'.  So we have used transcendental argument again to reject monovalency and monism.

We still need to tidy up the differences between change and difference [sic].  The two do not collapse into each other.  Change implies something continuing 'in a tensed process'.  Difference involves 'two or more non identical tokens' which cannot be reduced to a single origin.  If there are two non identicals, we are allowing for the possibility that there is an infinite number.  Again this is 'transcendentally necessary' for any referential detachment, and it clearly implies intransitivity.  Sometimes, the two categories are merged, but we are not permitted to think about fundamentals, unique beginnings or fixed foundations.

[Still on the argument for the negative].  A world composed only of the positive would be a world without voids or absences, with a continuous smooth connection between objects [as in Cartesian mechanism], where movement would be particularly difficult to explain, let alone action at a distance.  We could, as a thought experiment, think of an entirely negative world, 'a total void, literally nothing', and this possibility is not rigorously excluded by a fully positive one [classic philosophical argument, all about why there is something rather than nothing]. There are other arguments as well.  The arguments against fundamental or unique origins similarly argue that it will be difficult to posit the beginning that also explains actual forms of positives and negatives. We do need positive being to understand negative being, and it is possible to argue that the positives are more common or more important than the negatives - but ontological monovalence is still wrong and dogmatic [argued very densely 598].  It is more interesting to allow for negative being, not least because it explains different perspectives 'where something is absent in one perspective and present and the other' and raises a number of interesting complex possibilities in terms of how the negative and positive are related - 'all sorts of topologies become possible: hidden depths, tangled loops, inverted hierarchies, or mediatized, virtual and hyperrealities; holes within wholes...  binds and blocks, intra as well as inter action, juxtaposed, elongated, congealed, overlapping, intersecting, condensed spatio-temporalities; intertwined, dislocated and punctured processes' (598).  We now have the power of absences in everyday life as well, of persons or of opportunities.  All these are 'transcendentally and dialectically necessary for any intelligible being' (599).

Section two.  Emergence.  Hegel pointed to contradictions that led to the expansion of conceptual fields.  So far emergence has only been discussed at the 1M level, and it is necessary to bring out its specific ontological implications.  New entities or concepts are generated by preexisting material 'from which they could have been neither induced nor deduced'.  This is a spectacular quantum transformation, 'matter as creative'.  If we understand it in material terms, it is a parallel to the processes described in Hegel, and will produce 'types of superstructure'.  Marx has only examined one kind of super structural emergence, where a level is superimposed or intraposed on or within an existing one.  But there are other formal possibilities, especially if we add complexity [including divergence and detachment, or disembedding].  The concept is useful [in Bhaskar's hands] to grasp the notion of agency together with causality, the existence of real novelty without a transcendent cause.

However, in the real world, emergence does not follow an Hegelian linear path.  Complexity and multiplicity in the open system is what produces it.  Teleology is rare, and full awareness of the processes is unusual and arises from lower levels.  Emergence can produce a new type of structure or a microscopic 'structuratum'[a highly obscure footnote, 600, explains this as something that condenses a multiplicity of different structures and their ways of acting], and is usually still 'conditioned and controlled' by the level on which it is superimposed.  There is also the obverse process,'disemergence', where a level collapses back into the one below.  Emergent layers can sometimes contain problems which limit their attempts to preserve themselves.  Complexity can result as in 'a "laminated" system', and these can sometimes be decentred, asymmetric 'and contextually variable'[the example is the notion of the self in Joyce!], together making up 'an internal pluriverse...  Populated by a plurality of narratives', some of which might be discordant.

The emergence is driven by a dialectic between real and actual levels, and can lead to genuine otherness ['real determinate other - being'], with its own specificity which must be grasped before we impose any scientific or historical explanation on it [the examples include how chemical phenomena first had to be explained in their own terms before they could be reduced to physics, and how vulgar theories of change, including Marxist ones were premature, forced to supply some myth of origins and linear direction or teleology].  The implications extend to the notion of human agency which itself has emergent causal powers, traceable ultimately to the biological matter which forms agents, but allowing for them to react back on that matter in an open system.  [The argument for this emergence seems to be that human beings are capable of destroying natural mechanisms even against their own best interests].  It is not just that human agents are dualist [both biological and social], because agency is impossible without our particular biology, even though it cannot be reduced to biology - and it is impossible for materialist philosophers to argue that while maintaining their own autonomy from biology.

We can treat human reasons as causes, and we do so in scientific experiments, which are caused to happen by scientists, in order to gain 'epistemic  access to the real, transfactually efficacious, but normally empirically counter factual causal structures of the world' (601).  This helps us to use the same transcendental argument for the deduction of emergence, at least for humans.  It helps us reject simple 'reductionist materialism' [where the actual is confused with the real, and sequences are understood to be simple blocks of space and time].  [In the middle of the dense discussion we find an argument about causality which raises problems with determinism.  I first came across in Deleuze.  The problem is that if an event is caused by an earlier event, it was somehow bound to happen in the past so it has no real future dimension.  Bhaskar sees this as confusing real processes, with  epistemological arguments, and proposes an open system, where things are genuinely 'other - beings' (602). Causal processes are seen as special cases of transformative negation occurring in 'rhythmic' processes, 'a tensed process in space - time'.  Agency can be seen as 'intentional causality', engaging in 'efficacious absenting'].

In pluriverses, emergence results from 'duel, multiple, complex and open' systems of control.  In a sociological example, higher order agencies can set the boundaries for lower order laws, as when capitalist systems explain particular selections of physical processes in engineering.  This is how we should explain marxian superstructures - the relations of production set the boundaries for the forces of production and also for the relationship between the political and the economic.  If we apply the second notion of superstructure here, where super structures become intrastructures, it is also possible to see that the economic also sets a framework for the political, which can be sometimes relatively autonomous, sometimes even 'supervenient'[I still prefer Poulantzas'model of the social formation as a cnd badge where the economic is an outer ring as well as a segment].  It is in the second sense, that postmodernism has arisen from globalised capitalism, for example.  We can even apply the two models 'concurrently'. [What happened to the even more complex possibilities?]

Emergence also suggest the idea of 'emergent spatio-temporalities' (603), possibly related to a new system of material things, or a preexisting one.  New rhythmics can also emerge, extending to the transfactual, agentive, or negative, or their opposites.  Rhythmics can be non substantial, with the latter seen as 'a sui generis causal power of space - time itself'.  We have to reconsider space - time as well: it could be a reference grid, or measure, a set of mutual exclusion relations, something potentially emergent, perhaps with causal powers of its own, or something entropic. 

Bhaskar goes on apparently, to link together space, time and causality around the notion of an irreducible tensed dimension to space and time. By considering normal emergent social things like people, institutions or traditions as being either constituted by it, or containing their own 'geohistory' which itself offers possibilities for future development.  They should also be understood as consisting of a set of irreducible relations, various 'connections and interdependencies with other social (and natural) things', argued at 3L.  Humans live for the future,  but they also live literally in the past, and if we think about this, we end with all sorts of possibilities of overlap, intersection, condensation, elongation, or divergent and convergent 'rhythmics (causal processes)' (604).  We can see all sorts of different 'sedimented rhythmics' in the regions of modern cities, combining elements of the past with the present and the future.

When it comes to conceptual emergence, when we perform the various transforms in epistemology, we often draw upon 'past or exterior cognitive resources'.  But we can also perform a 'perspectival switch' without any additional inputs, a new gestalt.  For example, we can see the emergent totality as an organism, and even there, we can see it as reacting to the 'affordances' of its environment rather than insisting on some internal teleology [this new notion of evolutionary biology is used by DeLanda to critique the old deterministic notions relying on some blueprint or some teleology].  This argument reveals a perspectival switch which has uncoupled itself from normal explanations [which might be what he means, 604-5].  Although we switch perspectives to gain better explanations, this is not an entirely subjective process, because it is still governed by the reality principle [misappropriated from Freud]. 

Dialectic is particularly good at taking a looser view of totalities and how they might be linked or segmented, and does not attempt to restrict complexity through some notion of expressivism or teleology.  Instead, it stresses alterity, with all sorts of spaces and forces at work, including negative external and contingent ones, as well as the normal positive determinations and relationships.  The precise patterns are 'up to science to fathom'(605) [abandoning the dialectic for something simpler to do this?].

Section three.  Contradiction 1: Hegel and Marx.

Not all contradictions are direct and logical as in Hegel.  In human affairs, it often seems to imply a dilemma where we have to choose one end at the expense of achieving another.  Subjects experience a block of one line of action, and become aware of another one which might be opposed to it.  Both Hegel and Marx focused on internal contradictions, but there can be external constraints on action as well, and these are pervasive [such as the laws of nature].  It would be wrong to insist that everything faces contradiction: instead, everything depends on 'grounded ensembles and totalities' (606).  However, there must be some sort of complicit relation between the contradictory things and subsequent change in its essence, unless we want to argue for unchanging eternal things.  Instead, finitude offers the basic form of the contradiction - it is our fate to perish.  Other common contradictions are between 'mind and body, fact and fantasy, desire and desired, power and need, eros and thanatos, master and slave, self determination and subjugation'.

Formal logical contradiction arises from 'axiological indeterminacy', and this may be common in social life too.  It might result in something determinate and intransitive if agents obey axiological imperatives, showing the 'constellational' link between reason and cause.  Sometimes logical contradictions can themselves become part of problems of the life world, and we can explain them as intransitive objects, normal parts of social reality.

Dialectical contradictions depend in turn on dialectical connections which are linked aspects of the totality which are distinct but inseparable ['synchronically or conjuncturally internally related', (607) where one existentially presupposes the other]. The internal connections can take a variety of forms from absolute to conjunctural.  For real dialectical contradictions, elements have to be opposed so that one aspect negates the other or the whole - they may be more or less radical.  In Hegel, they are assumed to be linear, and Marx tended to assume this too, partly from adopting a particular narrative form in his writings.  Sometimes logical contradictions can also be dialectical ones if they are grounded in a particular mistake [with its particular real ground], but normally, they are two species of internal contradiction. Dialectical contradictions can involve various degrees of antagonism, at different levels of explicitness, and operating over different sorts of cases.  They can be both contingent and necessary, external and internal.  Not all conflicts and all struggles can be the result of a dialectical contradiction, and sometimes there are contradictions at a deeper level.

We take power to be a transformative capacity [power 1], drawing upon structures of domination or control [power 2].  Baskhar wants to see the latter as 'generalised master - slave (- type) relations'(608), which preserves the possibility of reversibility.  We can assume a universal structural asymmetry [in order to explain the direction of change?] and this may produce dialectical contradictions over time, associated with particular hierarchies or complexes of power or totalities, and this in turn might produce subsequent secondary or tertiary contradictions.

All these relations have to acquire meaning in the social world, including logical contradictions.  Materialist contradictions as described above are stressed by Marx, and the way to resolve contradictions is to transformatively negate the ground of the contradiction itself -transformative praxis of the totality [an awful phrase instead appears on 609 - 'transformed transformative totalising transformist praxis'], and the result will be to sublate the contradictions in socialism.  Marx does not emphasize logical contradictions even though his work might contain them.  The unity of opposites in his work is grounded, with a concrete conception of capitalism.  There might well be an epistemological dialectic too, borrowed from Hegel, involving immanent critique of political economy, the identification of insoluble contradictions and anomalies, a research programme to solve the problems and so on.  There is another dimension to ideology as well though, not just a lack of understanding, and this appears in the description of commodity fetishism.  This permits Marxism to move beyond the immanent, and to discuss new kinds of causal and relational dialectic in the context of power 2 relations.

[Differences between the Hegelian and the Marxist dialectic are spelled out in more detail, 610-11 - Hegel operates with logical contradictions in philosophy, but also has a kind of concrete social history, but this does not depend on material contradictions.  Marx breaks with actualism and empiricism by developing a structure with ontological depth, and this increases the possibilities of identifying contradictory and opposing tensions, although the whole thing is still grounded in explaining the contradictions of the modern world.  Greater complexity will also permit a role for transformative praxis.  Bhaskar goes on to say that there are other ways to read Hegel, especially on history - which does lead to him presenting some analyses of real societies.  We also get a return to Bhaskar's earlier language which sees Marxism as operating both A and B transforms as defined earlier.  Marx borrows some of Hegel's epistemological dialectics as well, so there is considerable overlap.  However, 'Hegel's resolution is in theory.  Marx's is in practice' (613), and the latter depends on much more detailed analysis of actual societies - both also referred to 'practice' as some general human activity.] 

Emphasis is given to the motives affecting both in terms of resolving discrepancies between theory and practice, which Baskhar argues has always been 'of special interest', because it uniquely prompts attempts to resolve it, and offers a powerful way to begin to criticise power 2 by practice as well as theory, or by noticing discrepancies between the theory and practice of the powerful.  Also, working with theory helps us adequately identify its Achilles heel, its weakness in the middle of its apparent strength.  It helps illustrate an alienation of the subject, and it raises questions about universalisability.

All theory has a practical nature, and all practice has 'quasipropositionality''(614), not least because it relies on certain conceptual resources and beliefs.  Taking a transcendental perspective, we can see that theory and practice occupy a linked duality [and, that logical contradictions and inconsistencies in theory can be seen as a theory/practice inconsistencies, but I am not clear why.  It is something to do with the way in which theory offers axioms for life, so that problems produce practical difficulties in knowing how to act].  The sense of lack of fit between theory and practice produces 'the important dialectical figure of the non - identity, alterity or hiatus-in-the-duality'.  [I think the issue about universalism arises because people can start to see that their particular theories might be located in some particular social arrangement]. Generally [and somehow as an implication from all this] we can develop a broader notion of dialectical connections, which do not have to be contradictions, nor to produce 'axiological inconsistencies', which can be described without invoking contradiction, and which do not infringe the rules of formal logic, except in some cases where logical contradiction is involved.  This covers all theory/practice inconsistencies, and gets over the objection from formal logic that contradictions are untenable.

[Back to theory and practice] theories often invoke practical reasoning about how things work, and practices often involve theoretical reasoning about the nature of the world, although there can be different emphases given to the different components: nevertheless, they occupy a duality at least when it comes to reasoning, even though they are distinct [baffling diagrams on 616].

Back to Marxism and the theme of the unity of opposites, which implies some internal relationality between antagonists in an overall structure of domination.  If this informed materialist dialectics in Engels, there is still the notion of immanent critique as central to negation in Lukacs.  Both claim to be offering concrete analysis.  Marx himself in the later writing used the notion of contradiction to refer to logical inconsistencies, other external  forms of non dialectical opposition [as in a clash of supply and demand], contradictions in structure found in particular social forms, and specific geohistorical contradictions where social forms develop [as in the tension between the relations and forces of production].  Generally, this is a list of forces which arise from social forms themselves as ground, which produces contradictory tendencies.  The rhythmics arise from structural tensions or conflicts, between wage labour and capital, or between use and exchange value, implying 'a meta ontological dialectic' with the last two types generating the first two.  This explains the apparently fundamental original separation between producers and their production, alienation, but operating with a real dialectic.  It might be possible to extend this in a more open direction, however. 

Note that we are describing consistently some dialectical contradictions, and these are not particularly mystifying.  We can consider Marx's work as part of a more general critical realism, even though he did not particularly come to develop it.  Those who want to suggest something different are mistaken [especially Colletti, who was one of the main people arguing that formal logic must reject contradiction, 619].

Section seven.  Dialectical motifs: Tina formations, mediation, concrete universality etc..  We can start with heterology which takes three forms: where something is not true of or applicable to itself; not the same as itself; not true for and/or to itself.  The first sense is the most usual one and is found in Hegelian dialectic where 'A is necessarily also not-A' (619), and this is what drives the dialectic forwards - something is true of but not present in some basic concept or form.  Eventually, at the end of the dialectic, we get to see that everything actually is true of or for itself, once 'mediated by absolute spirit' (620).

Mediation is the same as determination for Hegel.  Ultimately, it is the whole that provides the mediating terms.  Mediation also refers to something that is an intermediary, as when Marx argues that labour mediates humanity and nature, with various secondary and tertiary forms of mediation developing such as private property.  Here, mediation implies a medium, something that offers some distance or stretching between terms, or something that permits communication, or, in postmodernism, 'virtualization or hyperrealization'.  This will lead us to the notion of the 'concrete universal' [don't see how] which manifests or individualizes itself in one 'particular differentiation', the 'concrete singular', a particular moment which mediates another one.  Any temporal moment or spatial determination can be said to be mediating others.  Sometimes processes are mediators [the examples are the social world between structure and agency, and at the most general level 'the  tensed tri-unity of causality, space and time'].  If A leads to C by means of B, B is the mediator of the relation - thus social relations mediate agency, and 'rhythmics mediate causality'.

Alienation means 'being something other than', (621) being estranged from one's self or from what is essential to one's nature.  This implies that there is something intrinsic to one's self which need not necessarily be inside physically but can be a characteristic.  Alienation means losing autonomy.  There is also alterity, where language produces some notion of 'existential intransitivity'.  It is a mistake to attribute this to some fundamental origin or to let linguistic referents collapse into non-alterity.  We need to recognize our alterity as irreducible.  In politics, it means we cannot talk of some 'undifferentiated expressive unity' embodied in the state.

There is also constellational identity and constellational unity, relating to processes of containment and connection respectively.  We can use this to describe how being and thought occupy a constellation, without reducing them to each other or to a simple unity - thought is within being, but also emerges from it.  Similarly, dialectical and analytical reason can be seen as constellational.  This helps us avoid collapsing identity and difference, the problem with Hegel, so that being becomes only the 'demi-actual existent'[expanded into another ungraspable critique of Hegel, 621].  It seems to be that opposites are identical in constellations, which stops them being put into some other sort of relation, like a hierarchy.

Duality is slightly different, referring to 'existential interdependence' and some 'essential distinction' (622) [again examples of things like theory and practices or structure and agency].  The 'hiatus-in-the duality' refers to some dislocation, again guarding against 'reificatory collapse'.  There are also perspectival switches which may be grounded in intransitive features, or epistemic interests.  They can be transcendental if the eventual form is transcendentally significant [examples here include the dyads of tacit and explicit knowledge in Polanyi -- presumably the argument that the tacit grounds the explicit].  Dialectical perspectival switches can follow the result of a successful dialectical argument or can take the form of a reflection.  A 'degree of perspectival fluidity and multifacetedness' is essential for any productive concrete or totalizing inquiry, and Hegel claims to have developed this at its best.

Sometimes, a dialectical necessity to solve some theory/practice inconsistency becomes 'an axiological (or practical) necessity'as well.  [Baskhar argues that this is possible and inherent in and  for all dialectical necessities].  This involves the concept of a TINA -such necessities are seen to have no alternative.  In these cases, an axiological imperative forces us to adopt some sort of defence mechanisms, some sort of supporting or reinforcing connections ['duals, complements and the like', 623], or some 'cloak' of compromise.  Together, these describe the TINA syndrome.

Anything that assumes an identity between subject and objects can take an anthropocentric form, for example, but that must then appear as an 'anthropomorphic correspondence theory', or as a compromise with one [this could be a long winded way of saying that special interests have to turn into interest that represent the whole of humanity?].  Actualism is also typically found in TINAs, if the actual is seen as a some invariant combination of subjectively defined particulars [appeals to be 'realistic'and accept what is?].  It is also common to have some ceteris paribus clause to defend particular combinations or conjunctions, and this can manage theory practice inconsistencies in particular, sometimes by developing some metaphysics.  In these ways, axiological necessities can be seen as 'autosubversive, radically negating, internally split, axiologically inconsistent'.  TINA compromises and formations are always internally contradictory, however, and need to be 'transcendentally or otherwise refuted', to expose their duplicity and equivocation, or their pliability or indeterminacy.  Such compromises have produced characteristic paradoxes and effects, including schizoid skepticism, introjected or projected duplicates or fragments [as in Hegel's 'Unhappy Consciousness'] or the 'scotomatic'[God knows, also defined as stoicism, possibly meaning a deliberate focusing down of perception].  Philosophically, empiricism and idealism can be seen as TINAs. Once we have identified them, TINA clauses appear as inconsistencies or contradictions, or splits and 'detotalizations' (624).  It can imply that something else is needed as well as a particular argument [and Derrida on the supplement fits here somewhere]. 

TINA formations are usually repressed, linking to issues of ideology and power.  This has been discussed in philosophy in terms of what happens when an axiological necessity, such as the need for food or sex [and Bhaskar wants to include autonomy] is blocked.  Freud is clearly relevant here, but other critical philosophers ask the question too.  Hegel's answer was that a particular form can be both false and necessary, but Bhaskar suspects that this is really a TINA, and indeed all his dialectical stages can be seen as TINAs, clarified only at the last stage.  Nevertheless, this leaves him chronically likely to be seen as offering a compromise with the prevailing order of things 'rationally transfigured' into absolute ideas.  (625).

TINA compromises can end in 'axiological indeterminacy', or the abandoning of any attempt to produce progress in thinking or practice.  We can proceed only by reawakening the idea of some incompleteness.  [Again Hegel is seen as compromising on this, 625].  We must be careful that this is not infinitely replicating the status quo, nor limiting the transformative capacities of praxis [for example leaving structures intact].  Any attempt to announce 'the constellational closure of the history'[Hegel rather than Fukuyama here] risks reawakening 'tensed geohistorical processes'[in the form of Marxism for Hegel -autonomism for Fukuyama?].  Challenging closure itself depends on holding to the concepts of ontological depth and open practical totalities.

Ideology can be seen as originating where power intersects with discursive and various kinds of human relations as in the social cube.  It takes a particular form when it embodies  'categorial error'[which is why we need philosophers, of course].  This can sometimes take the form of weak analogies, such as war as a game, or [positivist forms] such as the capitalist notions of the value form.  These activities are essential to TINA formations, but are particularly open to explanatory critique [as above].  However, since they reproduce social relations and interests, explanatory critique alone is not enough.  We need as well transformed, transformative, transformist, totalizing agency.  This will take the form of politics of various kinds - life politics [turning on health, career, focused on concrete singulars, with an ethics of 'derived virtue']; politics of movement [in the sense of movements like feminists or various other collectivities]; representative politics [preserving the freedoms and rights of particular communities]; participatory-emancipatory politics [aimed at the fundamental structural change in the interests of 'universal human flourishing'(626).  Each one requires effective working together, maintaining solidarity and trust, and monitoring transformations from struggles over power 1.  The end will be  a 'eudaimonistic pluriverse', and the minimisation of [bad] heterology, so each could be true to of and for themselves and each other.

We must now discuss totality and concrete universality, which might be seen as polar opposites, if totalities seeks to exclude heterology and establish unity.  There are different sorts of totalities.  Hegelian dialectic offers a concrete totality, with contradictory elements and producing constant transcendence, or 'preservative superstructuration'(627).  Kantian dialectic is different [discussed 627], where human intelligence is limited because it can never know 'the infinite totalities of reason'.  Both are accused of making excessive claims beyond the simple one, that the world is contradictory and that we can never know all of it.  Both attempted to ground this idea in something unconditioned [or absolute].  Critical realism is of this more limited kind, operating with a grounding based on 'the less conditioned', and avoiding empiricism.  They also did not sufficiently examine the social context of their day, tracing dialectical limits  to  'geohistorical specificity' and its effects on existing concepts.  [More of this on 628].  They would have discussed concrete powers of agents, and practical totalities as something realizable, 'informed by the supreme ethical virtue of wisdom' (628), and this would have meant they were discussing the majority or entirety of human beings, which is the aim of critical realism.

Science attempts to maximize its explanatory power by aiming at totality.  However, it has neglected the issue of the internal relations of its subject matter, 'intra-action'[because it has headed too readily for the empirical?].  This can involve a matter of the constitution of something where one element or aspect of it is essential and intrinsic although not necessarily a physical part of another element; where a particular element is present within, not essential, but somehow implied by the other elements; where one element acts causally on another element internally related to it.  The last one describes dialectical connection, noting that this does not necessarily involve contradiction, and allowing for the relations to be complex, not always reciprocal, sometimes transfactual, sometimes related to agents, and so on. 

Can we go on to transcendentally deduce the concept of totality from these interrelations?  In social life, we can see that some important objects, like books, existentially presuppose lots of others and certain traditions which have made it possible.  The text itself can be seen as a totality internally related, so can language, the monetary system, or a musical tune.  Given that all of these totalities are essential or transcendentally necessary for knowledge of them, we have performed a transcendental deduction of a totality - they must exist for social life to be possible.

What about the natural world?  We can see dialectical contradictions which presuppose internal relations, which take the form of change, again argued that without some understanding of these science would be impossible, for example scientific taxonomy requires that certain things have properties which are essential to them and which help explanations to develop.  These definitions are essential if the open system of the world is not to become arbitrary -we could not operate for very long with a classification based on superficial resemblance, which would constantly produce paradox.

We have to break with the ordinary notions of identity, causality, space and time, themselves based on some implicit analogy with a classical mechanistic world view.  We have to see things instead as existentially constituted by relations with others, so that the ordinary notion of identity is an abstraction not only of properties, but from processes of formation and activity, for example in identifying causality.  Specific regions of the totality are best seen as generated by the totality.  Otherwise, we are left with arbitrary definitions of identity or individuation, and explanations of behaviour, aporias for philosophy and real problems for science.  We need to restore and develop a notion of 'entity relationism' (630).

We research totalities by starting from individual elements.  We then discover more and more elements or the whole thing ['at the intensive margin'], contained or condensed or implicated in or causing the original elements, in various modes [through their presence or absence], with their accompanying paradigms and syntagms..  Working more extensively, we can connect the initial element with other elements, for example through various 'constitutive, permeative and causal relations' (631).  We must not confuse what is intrinsic and what is merely internal [the example is discussing permeation either in Cartesian or Newtonian terms].  We can pursue reflections and then consider the implications for the 'saturated result'.  Reflections of different types may be compounded to produce totalities of their own.  Here we have simple totalities which are consistent, sub totalities which have discontinuities and spaces and boundaries, and partial totalities which may or may not have relations, including contingent ones between its elements and sub totalities - the latter describes the social world for example.  We can identify all sorts of materialist processes going on here, including stratification, emergent rhythmics, transformative agency. 

Together, considering the complexity of these possibilities could lead to infinite activity of theorising, and it is easy to see why people like Hegel decided to introduce a particular form of closure for example with unilinear presentations.  Real partial totalities are themselves limited, however and the purposes of science limit inquiry as well [here, closure seems to take the form of some post pragmatic decision that the definition has been achieved].  Openness remains as a potential, however especially given flexible systems and the possibility of 'multiple perspectival switches'.

We can think about holistic causality [why not?  I've got the rest of my life] as a way of getting to de-alienation.  Causality already presupposes transfactualism, various kinds of mediation and multiple determinations, including milder forms like 'conditioning, limiting, shaping, influencing' (632), and lots of other processes like stimulating, enabling, coalescing.  There is also self generation.  However, let us consider all these as forms of determination and define holistic causality as a matter of the coherence of a complex, so that the totality itself causally determines the elements, and the elements themselves causally codetermine each other and thus codetermine the whole.  The second case describes an emergent totality.  Given that we are mostly describing partial totalities, we must resist any attempt to make them  expressive or centred.  Indeed, one particular element within a totality may constitute the ground of that totality, having an asymmetric weight. Elements can also show different forms of attachment and detachment, such as relative autonomy.  Totalities may be grounded in some deeper structure, acting as a mere mediator of a deeper holistic causality.  In any of these cases, the totality becomes structured.  Its elements might be linked by dialectical contradictions and antagonisms, or be mutually reinforcing or supporting [Bhaskar raises doubts about whether this second one is in fact a TINA connection].  Totalities are therefore open processes, even though they are constituted by geohistorical formations.  They can exhibit continuous configuration or change in the form of transformative negations in rhythmics.  They may include 'intentional absenting or agency'.

To take an example, a generative separation can create an alterity, and produce an absence or transformative negation in a rhythmic, appearing as the exercise of causal powers.  If it is an agent that is alienated from something intrinsic and absented, intentional causal agency can produce a subsequent 'happy totalisation' (633), which may turn out to be a stage in a dialectic of alienation and dealienation, where alienation becomes an essential stage of achieving totality and self identity. [Doesn't seem to be much more than saying we can overcome alienation if we intervene --to overcome alienation by putting ourselves back in some totality by recognizing the processes involved!].

Now let's turn again to the connected notion of the concrete universal.  We can define concrete first in terms of its contrasting notion of abstract, then ascribe a more positive meaning in the sense of being complete or adequate for the purposes at hand.  Actualities and experiences may be concrete in these senses.  But we should not equate the concrete with the actual or the empirical.  If we look at Capital for example, 'capital in-concretion' is an intransitive object produced by theory, and it is transfactual [we're getting close to the idea of the diagram or the machine in Deleuzian thought]: manifestations of it are determined by other economic modes or by particular mechanisms, and Marx did not adequately describe these, but they might include particular states of labour power, ethnicity or the unconscious, for example. Manifestations also appear following other moments.  Marx actually worked towards some 'pan-concrete totality' as the ultimate intransitive object, and this is not just the result of many determinations found in particular conjunctures, and nor is it some Hegelian multiplicity [it might be a Deleuzian multiplicity, though, with both actual and virtual bits].

Critical realism insists on the separability, multiple determination, and spatial-temporalisation of the concrete universal.  It is 'a multiple quadruplicity' (634) [four dimensions at least?], Specific, structurally sedimented, located in space and time, containing moments of universality and singularity.  In open systems, there is 'a multiplicity of quadruplicities'.

In Kant and Hegel, the issue was whether concrete universals were real or imaginary, and if the universal components in them were transcendent, that is transcending space and time [this is an implication of Hegel's view that the real is also ideal, and the infinite embodied].  Critical realism insists instead on the three characteristics above.  Separability arises from the fact that we can control determinants and causals, say in an experiment, in order to remove mediations and arrive at 'the instantiation of a universal law in a singular', where the singular is assumed to be replicable.  Hegel closes the notion of totality, and therefore change, at the final stage of his dialectic [I can't see the implications in this very dense argument, 634.  I think Baskhar is simply saying that we don't want to close it off like this but maintain all sorts of different processes of change emanating from different structural levels and modes, and that this is necessary for explanation in social science].  Hegelian idealism also forbids 'referential detachment'[since spirit and matter are ultimately fused?], and this again reduces the autonomy and the interesting properties of concrete singulars that lie outside the reference.  Finally, transfactuality in structures produces 'unactualized possibilities', and these are not exhibited in singularities - and this is a possibility not grasped by Hegel.

There is a problem with all this complexity and multiplicity,  however.  Mediations may occur in a number of modes or rhythmics.  This is useful if we wish to criticize particular 'modes of illicit abstraction...destratification, deprocessualization, demediation and desingularization'(635).  There are other simplifications as well, because elements might be connected to totalities or require explanations [show the results of processes] at the different stages of the dialectic [1M, 2E and all that].  This produces further forms of illicit abstraction -- 'destratification (again), denegativization,  detotalization and deagentification'. Making concrete universals subject to human interaction adds further complications which will require different forms of investigation.  All these add multiple determinations and possible structures constraining them [further baffling diagrams on 636].

Nevertheless, the notion of concrete singularity is 'the key to the realm of freedom' (636).  [I think the argument is that we resolve complexity by specific theories aiming at maximising explanatory power, which must nevertheless follow particular principles of depth and completeness, translated as stratification and totality, 637].

We can conclude [thank Christ] by discussing levels.  We've already seen ontological stratification, in the transcendental movements between 1M and 4D, and how this solves lots of philosophical problems [further explicated on 637].  We can now specify a model for 'applied scientific explanation' [presumably, as well as philosophy].  We start with some concrete phenomenon, found in a particular conjuncture, and see it as composed of interlocking partial totalities: the aim is to 'resolve it into its components'.  Then we have to redescribe the components theoretically, so we can bring theoretical science to bear.  Then, we retrodict to antecedent causes, allowing for the affects of context and geo history.  Then we work through the large number of multiple causes and mediations 'until one has identified a full enough set... for concrete applied explanation to have been said to have been provided'[said by who?].  Then we have to regress to redescribe the initial phenomenon in terms of its causes.  Taken together, this provides a model of scientific explanation which includes correction.  However, we must be prepared for much more complexity and mess, with equally plausible explanations found at different levels.  It would be a mistake to privilege theory in these efforts, and this shows in various disputes about Marxism, where we're talking about intransitive objects which are changing, requiring some geohistorical context.  We are also attempting to find the  'fundamental processual dynamic' (638), but it would be wrong to misapply this level of analysis to more surface elements of change -- to do so would be 'actualism' [which I have taken throughout to mean privileging the actual, a sin rather like empiricism]. [Add together the initials of the key words and you get an RRREI(C) model - -which comes up later]

[Changing the terminology a bit] pure models can help us isolate areas and regions which require more local investigations.  But this should not lead to disconnection and compartmentalisation.  Dialectics can not be confined by disciplinary boundaries, with their own definitions of phenomena, but must totalize, focusing on 'the intrinsically connected'.  We need 'level specific concepts'and they are a way toward investigating both the concrete universal and the totality.  It is a notion of incompleteness or absence that then drives the dialectic.

Chapter 24.  Dialectical critical realism and ethics. Bhaskar (641--88)

[My gloss.  The first thing to notice is the massive contradiction between universal emancipatory intentions, and the appalling dense difficult self referential and elitist way in which this section is written.  The whole thing depends upon philosophy and social science connecting with lay interests in pursuing their desires, but the last way to do it would be by writing like this.

The argument itself has really been foreshadowed in some of the earlier staff.  We know that Bhaskar has already argued for a four-fold ontology, and this section begins by summarising the characteristics, and arguing that some ontology is essential, and that to deny one is to bring about some implicit reliance upon one, which is probably right.  Then there is a discussion on truth and its dimensions , again which is quite interesting, and can be understood as a tweak of Habermas and his material on valid arguments.  En route, Popper and other correspondents theorists are briefly dismissed, which is quite interesting for those who want to see Popper and Bhaskar as somehow in the same camp.

The last bit I could bear to read [I left one section out altogether - sorry] was on the possible emancipatory tendencies of social science.  It is all good solid stuff, rather familiar about the role of social science being to examine the constraints on people's lives, using the dialectic and its central emphasis on absence.  Morality is grafted on by seeing any ill as an absence, and then pursuing the argument that we can move from explanation to moral implications, in the rather formal guise of a critique of the naturalistic fallacy.  After that, we end with a lyrical description of a possible utopia where all constraint seems to have been ended, and human beings are fully autonomous, although there are some weasily bits about the constraints that might remain, and much is just assumed, such that we would all agree to a distribution of resources globally, and that the only reason we don't at the moment is because we have false ideas about reality. 

In general, there isn't too much about practical politics here after all -radicals just somehow connect to lay people by explaining to them how constraints arise and how this might fit any particularly current crises that they might be going through, or that their economies are going through.  This is the classic supposed role for the radical philosopher, helping people understand their lives by offering them incomprehensible dialectical arguments.  There is no real discussion of what he has called TINA arguments after all, and why they are actually quite plausible to lay people -because they do not have the resources, including cultural capital, or access to radical philosophy to be able to thoroughly interrogate the terms of the argument.  Picking up on one of Collier's points as well, there is a certain naivety in the view that people like fascists would be prepared to reconsider their views after a thorough discussion with a radical philosopher, nor any real critical analysis of any fascist or other universals such as those involving race or nation.  It is too general.  The baby's primal scream is as much a valid protest against constraint, as that of a Marxist militant - but we're good at explaining baby screams in ways that don't exactly liberate them from parental control, and, similarly, we infantalise Marxist militants.  Babies are actually fairly easily mollified by giving them some nice consumer item, and it would be rather strange to insist that they should still be screaming because they are merely well fed rather than fully autonomous].

Section one ontology.  Ontology means both everything that has being, and something that relates to a particular set of arguments as opposed to epistemological or ethical ones.  We can distinguish between philosophical and scientific ontologies where  science studies 'specific ontics' for a particular purpose [so who studies de-ontics?].  We can specify different levels of abstraction or enquiry, domains.  We can also identify characteristic mechanisms such as contradiction or emergence.  We can develop an ontology that is conditional, based on the world investigated by science, for example, pursuing transcendental argument, or we can develop more philosophical grounds going beyond deductions from science.

Any theory of knowledge 'presupposes an ontology' (642).  It might be empirical realism, for example.  Attempting to operate without an ontology means in practice the deployment of an implicit one.  The central focus on ontology avoids various fallacies - the epistemic fallacy, where our knowledge of being guides our investigations; the anthropic fallacy, where reason is seen as some attribute of human being; the ontic fallacy, operating with reified facts and some 'compulsive determination of knowledge of being by being'; the linguistic fallacy, where the intransitive collapses into discourses about it.  [exponents of these fallacies include Rorty, Heidegger, Hume, and post modernists respectively].

Critical realism is to be theoretically deepened and 'topically enrich[ed]'(643), especially by emphasis on negativity and totality, with further expansion of the four moments. 1M stresses non identity, 2E negativity, 3L totality, 4D transformative agency.  Topical enrichment involves developing a suitable moral realism.

• (1M) We can see the importance of non identity in the earlier arguments distinguishing transitive and intransitive dimensions, or the real and the actual, which leads to transfactuality and constellations.  Non identity also involves 'the critique of centrism', especially in the anthropic fallacy, where being and knowledge arises from the realms of subjectivity and actuality: paradoxically, this leads to the reification of facts, which restricts a proper notion of agency.  More generally, the identity between subject and object has led to all sorts of philosophical problems, and it must be replaced with a proper discussion of alterity, including referential detachment, and, at its most general, absence.
• (2E) develops the notion of absence in the most general sense, so that the effects of causes can be seen as absentings - 'changes are absentings' (644).  Any ills can be seen as absences or constraints, suggesting emancipatory praxis to overcome them.  Also implied is a critique of ontological monovalence, which reasserts the existence of negatives of various kinds, including 'non-existents', and leads to arguments for the priority of the negative, and eventually an analysis of power 2.  Similarly, geohistoricity is reasserted as involving real connections of space, time and causality, in various rhythmics. The connection between absences and ills lead to the notion of dialectic as freedom,  as a matter of absenting constraints.
• (3L) centres on the concept of totality against 'ontological extensionalism' which disconnects, say, thought from being producing alienation, and repression.  Alienation can be considered as alienation from product and then from the four planes of social life discussed above.  Instead, 'totality depends on internal relationality' (645), including intra-activity of various kinds, including permeation and connectedness.  We can identify both concrete universality and concrete singularity, and even 'concrete utopianism'[some nice new kind of totality?].  We have identified partial totalities resulting from 'complex and dislocated open process', which can take the form for example of 'global commodification'.  We can discuss reflexivity again as a fully grounded world phenomenon in a specific geohistorical context [based on some understanding of the complexities of totality, and allowing individuals to have causal powers]. Totality also implies universalisability.  There are however limits introduced by reality itself, and by alethia based on it.
  
• (4D) introduces transformative agency.  It can be ignored in three ways in any argument that does not talk about 'embodied intentional causal agency', by reducing agency to bodies or spirit or both.  Such arguments can also reify facts or fetishise processes.  The way to overcome this is first to argue that reasons may be causes and vice versa, in a 'dialectics of reversal' (646), or one that links consciousness and self consciousness.  Appeals to material interest can take the agent through instrumental reason to a much deeper explanatory critical rationality.  It is possible to detect a certain 'if highly contingent' direction in geohistory leading to the free flourishing of all, and this is useful apart from anything else in challenging various arguments for the closure of the future.  It is praxis which totalises, transforms, and brings about transitions which will 'unite the interests of the human race', and this capacity is to be understood as 'a species - specific ineliminable fact'.

Turning to lateral extensions, we can expand the notion of tenses to rethink the reality of the past as caused and determinate, the present as indefinitely extendable and indeterminate, or a 'moment of becoming' (647), and of the future as a 'mode of possibility of becoming'.  These conceptions prevent a belief that only the present exists, or that the future might be blocked.  This has a role in morality too [because it will guide actions in a good direction?].  Moralities are not just those that actually exist at the moment, where 'ought' is currently seen as separated from 'is'.  Moral realism can be grounded in 'ethical naturalism', involving a transition from fact to value, and thus of theory to practice.  It is possible to develop alethic ethical knowledge based on 'conceptions of human nature', based on the four plane model, and a commitment to openness.  Arguments like those that split ought and is are attempting to demoralise, but are actually 'reflecting the morality of an actually existing already moralised world'.  We need to counter this with the notion of universalisability.  No one can deny a certain epistemic relativity in this field, but we can still reconcile this with 'judgmental rationality' [that is develop some reasons for choosing one form of morality rather than another?].  Ultimately, we ground arguments 'in the truth of things not propositions', and we can escape subjectivity once we develop referential detachment of the terms that explain things. This will lead us to discuss truth, in various ways - faithful to the communicative needs of the social cube, adequate to the intransitive, both expressive and referential, and alethic.  The alethic truth 'is freedom, depending on the absenting of constraints on absenting ills' (648).

We also need to discuss ontology fully embracing the complex, including the complexity of what seems to be simple, identifying what is embedded, intermingled or constellational, for example.  Social beings similarly are constituted by the presence of others and they possess four kinds of tendency -natural absences, empty selves, hidden depths...  The void at the heart of being'.  What we have done is to make a general argument for ontology, for 'existential intransitivity' and referential detachment.  Realism appears in any discourse which claims to be about something other of itself, in praxis, and in desire for something outside oneself.  It is easy to demonstrate the necessity of referential detachment for any one who claims to speak about something. [I am not convinced this explains expert or considered referential attachment as well as lay forms]. Such detachment leads to 'the first taste of alethic truth' (649), As soon as we start investigating the properties or ground of things we are doing science [again, doubtful -- what if we investigate the ground via astrology?].  Scientific terms like differentiation and cause involve absences, and this leads us to the dialectic.

Problems arose particularly after the collapse of positivism, challenged by relativity theory and quantum mechanics in science.  Instead of attempting a new ontology, though, developments occurred in the transitive.  First there was a 'sociological conventionalism', in Bachelard and Kuhn, requiring scientists to accept authoritative statements about reality while quietly carrying on, a 'stoic indifference to reality'.  Then we had a 'post structuralist collapse to skepticism', where nothing existed outside the text, a  clear acceptance of theory/practice inconsistencies, that was to lead to the 'unhappy consciousness of the pragmatist like Rorty' who acknowledged reality but did not wish to talk about it.  This in turn led to Feyerabend and 'explicit dadaist contradiction' where there are no limits on what we can say about the real world.  Behind the struggles was one for 'symbolic capital, money and power', which Bhaskar wishes to describe as mimicking dialectical struggle.  This tendency is 'irrealism' (650).

Section two the dialectic of truth.  To say something is true is to assent to it, but this also involves a claim about the world, and how things are, a basic correspondence theory.  Pragmatism makes such claims.  When challenged, the argument turns to the ground of truth, invoking coherence theories.  So we have the 4 dimensions of assenting to something, going on to describe it, producing evidence, and basing some action upon it.  This can be seen as a basic structure for all judgments.  Universalisability is implied.  Telling the truth might satisfy an active axiological need to find our way in the world. However, difficulties have arisen with nearly every theory about truth, and associated matters including perception, causality and agency.  There are also quite variable contexts affecting content, and problems with developing universalisability [the example is the disputes about what traces on a monitor tell us about particle physics].

The most important theories of truth 'have been the correspondence, coherence, pragmatic, redundancy, performative, consensus and Hegelian theories of truth (and Marxist ones) [discussion ensues].  Correspondence theories, including Popper's, have difficulties in identifying an Archimedian point to compare competing items, run the risk of reification, accept only some kinds of knowledge [not immediate knowledge says Bhaskar] , and operate with a basic metaphor of matching, since transitive theories cannot be exactly like what they are about  (651).  Coherence theories refer to criteria rather than meanings, and sometimes assumed that there is a correspondence between coherent texts and true accounts of the world.  Pragmatism refers to 'warranted assertability', but not all warranted statements are necessarily true.  For Nietzsche, the will to power is the crucial drive, and truth becomes merely '"a mobile army of metaphors"' (652) [Bhaskar sees this view as both necessary and impossible, with the tendency to self erasure - not sure why].  [The other kinds are also criticised on 652].

In Marx, truth normally means correspondence with reality, with metaphors such as reflection, but the criterion for evaluating the claims of truth depends on human practice.  Reflection means both that the immediate form and its essence can be reflected upon, requiring explanation, and then some test of scientific adequacy developed - so political economy can be criticized as a mere reflection of social relations, but it leads to a discussion of an adequate representation of the surface forms, how reflections are produced.  Both description and explanation are important though, and the real object itself is seen as constraining both.  Other metaphors are found in later Marxist work, such as images and copies, which imply an explanation in the description.  Later writers still saw truth as a matter of the practical expression of the collective subject [and this includes Gramsci], and it is here that the inadequacy lies for Bhaskar [I think because social production is much more complex].  There is also a tendency to neglect the independent existence of the objects of truth.

We have to address the four dimensions of truth as above -requiring action from assent, as adequate or assertable intransitive terms, as adequately referential, and as alethic, referring to the truth of things not just propositions.  This is the '"truth tetrapolity"'(653).  We can see how it works by thinking of a simple dialectic.  First scientists are subjectively certain about the reasons for something, then manage to convince they colleagues that that something is factual.  Scientists work on the reason for that something or its objective truth.  So this is a track from subjective certainty to 'intersubjective facthood' to alethic truth.  The last stage involves accepting the existence of that something after the necessary referential detachment, and exploring, say, necessity or groundedness.  They do not need to continually assert the truth of the object.  We have moved from propositions as well.  Alethic truths are still expressed in language and can be corrected.  Different propositions might need to be judged [and Bhaskar says we should judge these by taking into account the material and cultural superstructure of society].  Nevertheless, the last stage does mark a distinct change in the direction of scientific inquiry.  Progress still depends on uncovering some notion of totality and pursuing a suitably emancipatory form of practice [I am glossing this very dense section].  Bhaskar here is relying on his own earlier arguments showing that dialectical critical reason somehow overcomes and completes other forms of scientific inquiry based on different notions of necessity, like Hume's or Kant's.  [A baffling diagram, on 655, shows how these are connected.  Popper and falsification is a stage that proceeds retroduction and transformation] The test of alethia is that it is grounded in dialectical reason and theory and leads to 'the absence of [bad] heterology'- it is non arbitrary.

The process itself might well be social as in a research programme, or result from natural processes of development.  There are no agreed ways to assess consistency, or rather, these are internal and intrinsic to the process.  However, universalisability is important as a test for consistency and as a criterion for truth.  We can also use a postulated end state, even though it might not be realisable - and thus introduce a notion of progress towards an [agreed?]  desired end.  Theory/practice consistency must be practical, again refer to some end state with which it is consistent, be grounded in an explanatory theory of both the current state and the transition to the desired end state.  Universalisability involves a notion of the transfactual, an attempt to explain all the moments of the concrete universal [as actualizations of something universal?] It must be agent specific [?], and transformative, oriented to achieving the desired end state.  We know we are on the right lines to develop universal judgments in both theoretical and practical reasoning if we can demonstrate

'expressive veracity: "if I had to act in these circumstances, this is what I would act on"'; fiduciariness: "in exactly your circumstances, this is the best thing to do"'; descriptive: "in exactly the same circumstances, the same result would ensue"'; evidential: "in exactly the same circumstances, the reasons would be the same" (656). 

[These statements are then rendered in more philosophical terms, so that the last two become the principle of sufficient reason.  They do imply each other.  Any advice must only be seen as'assertoric'as above].

Theory and practice is best seen as a duality because theoretical reason 'in turn implies a commitment to act on it' and requires only a perspectival switch to become practical.  It still requires that there is solidarity between the agents involved, and that agents are sufficiently empowered and capable of sufficiently deep praxis [maybe this is what is meant by agent specific?].  In other words, it must lead to 'axiological commitment'.  Theory that does not lead to practice in this way introduces an inconsistency between the two terms.  [something dense and incomprehensible ensues about the relation of all this to the 'four planes' theory of social being, 657].  The overall process shows how particular aspects of the dialectic are aligned, however.

We also need to discuss meaning and reference, beginning with the familiar semiotic triangle [signifier, signified, and referent, rendered as force or the communicative dimension, conceptual distanciation and the use of metaphors, and detachment respectively.  Bhaskar also wants to translate the signified as the transitive dimension and the referent as the intransitive].  Various errors are identified in different ways to connect these, to demote the signified, or ignore the referent.  The role of playful communication is seen as important in model building, including developing metaphors for the signified.  Paradigmatic and syntagmatic dimensions connect semiotic triangles with others, affected by geohistories and intra-activity.  The most important thing to stress is referential detachment and the subsequent development of a generalised concept of the referent, which is couched in such terms as to permit the treatment of any aspect of reality, once we have detached it from human acts.  [I think the argument then says that the philosophical debates about the status of the referent, like whether it is real or imaginary, become less relevant].  The basic model can be applied to both natural and social sciences, although making ethical statements requires retrodictive [moral or political] commentary upon the law like statements produced by theory [I think].

Section seven.  Social science, explanatory critique, emancipatory axiology.  Much of this follows from arguing that the social sciences should be seen as '"dialectical critical naturalism"', which will overcome the famous dichotomy between hermeneutics and positivism.  It depends on social scientist being able to expose the influence on the subject matter of their own concepts, and that the activities of social science and the way it deals with the unconscious motivation or unacknowledged conditions can provide a guide for life for lay people.  Human nature is seen as having four planes as above, with non human and social dimensions, and covering 'power, normative and discursive' (659) practices.

Extending the argument, moral realism argues that morality is a real property, with both transitive and intransitive dimensions [roughly, relating to existing social and interintersubjective understandings, and that morality embodied already in the world].  Critical realism does not just describe but criticises, as a result of this gap, and describes, or redescribes and explains morality as something that guides practice.  It enables us to argue that moralities embedded in societies can be false, arising from some false idea of necessity.  Ethical naturalism means that moral propositions can be known and even tested by social science.  Thus explanatory critique leads to emancipatory axiology, through the dialectic, naturally.

We begin by arguing that any ill can be seen as an absence, and any absence a constraint, including constraints rooted in social structures.  These constraints produce moral falsehoods, and if they include [deliberate?] category mistakes, we can consider them to be ideologies.  They can all be seen as unwanted, unnecessary and removable.  Freedom then means autonomy and the ability to self determine.  The [natural?] disposition to reason may be linked in a constellation with reason itself, despite the differences, and human beings can sometimes desire to be free but lack the sufficient power or knowledge to do so.  Freedom also requires being able to access skills and resources [treated as a mere minor qualification, rasied and then dismissed in a sentence].  Knowledge can be tacit as well as propositional [politically correct but doubtful -- can tacit knowledge develop the dialectic?]

Moral reason is itself dialectical, involving transfactuality, concreteness and the ability to generate action.  It aims at a universal flourishing of human beings, implies both negative critique and a positive theory of how to remedy it, diagnosis, explanation and an appropriate action.  It might exploit the earlier model of applied social scientific explanation, which itself depends on the more general form of pure scientific explanation, and these can be combined in 'the meta-ethical virtue of phronesis or practical wisdom' (661). Depth praxis involves all the other abilities of transforming and transitioning as before, and it can be separated from instrumental reason, explanatory, critical and other kinds of rationality.

The analysis is clearly saturated with values, and it also implies values [so we can solve the naturalistic fallacy and slide from is to ought as before].  It studies beliefs and they can be false, and their falsity explained, if falsity lies in real things.  This also is to criticise any action based on false beliefs.  We can criticise purely technical reason and instrumental rationality if it does not attempt to explain its roots in power 2.  We have seen from the model of truth above that to urge anyone to act implies a statement of truthfulness, and how explanation can become an emancipatory axiology. We can also link form and content in the same way as fact and value, since substances are implied. 

Social trust and solidarity is grounded on satisfying all four elements of truth, and any missing elements can help us think out alternatives to the existing state of affairs, or even suggest a transition.  Centrism again is to be critiqued, and figures are to be seen as emerging from grounds.  If any ill is unwanted or unneeded, there is a universal interest in absenting it.  This can be generalised to argue that there is a universal interest in absenting all constraints, including master/slave relations and heading towards universal human autonomy.  Thus autonomy is implicit in every judgment, including those of truth and those involving deeds.  The goal of universal autonomy is even 'implicit in an infant's primal scream' (664) [seen as a protest against some absence].

This provides a general or formal argument for autonomy, but we can move on to substance as well if we see the possibilities of making shrewd judgment as 'naturalistically grounded'.  We can start to critique concrete singularities as a naturalistic impulse, requiring no additional justification [such as the ideal speech act].  We can check any particular criteria of thought or action by asking whether they are 'sincerely universally universalisable', and a negative answer leads to immanent critique.

Social science connects with lay people if its redescriptions can extend into a counter hegemonic struggle over matters such as the dominant moral ideology - 'personalism', which ignores social processes and mediations, and insists that values are not naturalistically grounded.  We can demonstrate that we can act as 'rational causal/absenting agents' (665).  We can also indicate a necessary dialectic between desire, wants, interests, and emancipation, and introduce the objective dimensions including power relations.  All this can be brought into an extended 'moral imagination'.  We do have to enlighten and empower ourselves first.  Political intervention depends on emancipatory agents being able to develop on the latent tendencies and processes.

We can be reassured that human beings are open in moral terms [more baffling diagrams on 666 show how ethical and political issues are linked.  Politics is subdivided as above -- life politics, movement politics and the rest.  Another diagram on 667 shows how universal human nature is embedded in a system of mediations and rhythmics to produce the concrete singularity of the individual]. 

[Bhaskar's utopia is spelled out a bit 667-8].  We end all master -slave relations and alienation, we extend rights and freedoms, with lots of local autonomy and participatory democracy as long as this does not interfere [!] with the 'rational regulation' of resources.  We operate at the global level as well, for example with a 'Council of the Peoples', trust, solidarity and care abounds, minimal standards of living are maintained 'without the compulsion to work, in the sense of selling one's labour power', tacit knowledge would be empowered, there would be cooperatives, the slogan would be 'to each according to their essential needs and innovative enterprise' [an already hijacked term, alas] (668), the rights of other species would be taken into account, education would be 'creative flourishing' and so on.

Do any constraints remain?  Any inequities, or pathologies, or contradictions about how to proceed?  We have to operate with 'a principle of balance' when considering totalities [disappointingly, specified in a later section].  The discoveries of actual social science might also help us realise any unrealised possibilities.  Open totalities mean that everything is transitional and that the status quo is being constantly challenged.

We can condense it all into the dialectic of desire mediated by the reality principle, the 'axiological necessities'.  At least this helps us remember that there is 'a conatus to knowledge' [I looked it up not long ago, and it refers to the intentions for action that follow, the other end from the affective in the classic sense], and this is an immanent tendency 'to get the dialectic of freedom going', so that once we see how to overcome one constraint, we will be driven to attempt to overcome all of them.  We also have to remember that rational autonomy depends on social relations and the way in which they relate to human nature and its four planes - this is a notion of the reality principle that drives the dialectic of freedom.  We could see the natural science and technology as already satisfying material desires, and we should be demanding that they also lead to a drastic reduction in labour time, new information-based technologies with an empowered workforce and a developed global production of commodities aimed at transformation which is negational.  Genuinely transformative depth praxis will be spread by the global movements, so that geohistorical processes would finally catch up with dialectical rationality.

And now, at last, enough....

back to social theory page