Andy Blunden November 2009

Critical Realism and Reality

This article addresses Roy Bhaskar’s philosophy with reference to the so-called epistemic fallacy, Bhaskar’s main weapon against postmodern scepticism. As Bhaskar’s writing is notoriously impenetrable, I rely a great deal on Andrew Collier’s “Critical Realism. An Introduction to Roy Bhaskar’s Philosophy” (Verso 1994).

According to Bhaskar, the ‘epistemic fallacy’ is answering a question about ‘ontology’ (by which he means whether some particular thing exists) with an answer about our knowledge of it, i.e., ‘epistemology’, as if whether something exists were the same question as whether we know it exists. Apparently, the epistemic fallacy has ‘dominated philosophy for three centuries’.

Let’s presume we are not a brazen deconstructionist who claims in the context of everyday life (in a performative contradiction) that ‘nothing exists outside the text’. Otherwise, if the existence of something is insistently problematized, we would probably respond by asking the questioner just what they meant by ‘exist’; or if the question is posed in terms of problematizing the reality of some thing, what is meant by ‘reality’. So far as I can see Bhaskar’s works are extremely thin on definitions of existence or reality, though full of definitions of various related isms. When Collier sets out to give us a definition of the meaning of ‘real’, he says

“The word ‘real’, in many contexts, draws its content from its contrast with ‘apparent’. A theory is realist in a stronger sense than others if it makes the following claims for knowledge: ... 1. Objectivity ... 2. Fallibility ... 3. Transphenomenality ... 4. Counterphenomenality ...”

Note how the definition of ‘real’ morphed into a definition of ‘realism’, an epistemological current, and ended up a definition of ‘real’ in epistemological terms. So if we ask Bhaskar an ontological question:, “What do you mean by ‘real’?,” we get a purely epistemological answer – the ‘epistemic fallacy’. Later, in tackling the question of whether a belief exists, he commits a brazen ‘epistemic fallacy’ by citing appeal to evidence, and then follows up with an apology. Likewise in dealing with the question of the existence of a poltergeist.

The premises of the sciences

Before going any further I shall spell out where I am coming from, using the notion of the ‘substances’ or ‘premises’ of a philosophy. These are the things which are deemed to exist at the most fundamental level of a philosophy. Substances characterize not just a particular science or problem, but the character of the entire intellectual project or world view in question. Natural science, by definition, is the science of Nature, and Nature means the world existing outside of and independently of human activity. The substances of natural science are matter and motion according to natural laws which are knowable. ‘Matter’ means all that exists outside of our consciousness. So the whole idea of the project of natural science is founded on the premise that its subject matter exists outside of and independently of human activity. That is what Nature is. What is not so, cannot be the subject of natural science. But a project is not completely determined by its substances. In fact, natural science shares its substances with everyday life insofar as people are concerned with events, processes and phenomena beyond the horizon of their immediate sphere of activity (public discourse), but it is self-evident that natural science differs in many other respects from the everyday life.

But at a certain point in the development of the natural sciences, specifically around the turn of the 20th century, mainly as a result of the ever-increasing precision of measuring practices flowing from the development of industrial activity , certain well known contradictions arose in physics. It turned out that it was not possible to fully describe the subject matter of quantum physics independently of the activity of taking measurements, because the interactions with the measuring apparatus formed an essential part of the behavior of the matter in question. It was shown, for example, that the idea of an electron having a trajectory was not coherent. The speed and position of an electron could not in principle be stated independently of the interaction used to measure it. Now, after an initial period of confusion, natural science reconciled itself to the fact that certain concepts (for example, the trajectory of a particle) could not be reified, and descriptions of certain objects could exclude reference to human activity only by letting go of the idea of there being such ‘objects’ as such. Having overcome this hurdle, natural science was able to continue with premises which did not include human activity, and reifying its concepts, and avoiding claims that would lead to incoherence. It was never in question that matter existed independently of activity, but there were restrictions on statements about material entities which could be valid and coherent without reference to activity, just as there are certain claims in mathematics which are incoherent, and natural science could avoid reference to activity only by respecting these restrictions. Reification of concepts is valid, but only up to a very definite point, beyond which reification is simply not coherent.

By reification I mean claiming that some finite thing, captured in a concept, exists outside of and independently of activity.

Now oddly enough, the human sciences share with ‘unrestricted’ quantum physics, the fact that activity is included in their premises. As Marx made explicit in “The German Ideology,” the substances or premises of the human sciences are real individuals, their activity, and material conditions, both those they inherit from the past and those they create themselves. These are the premises of the human sciences, taking it for granted that the human sciences work ‘in alliance with’ natural sciences insofar as natural processes intervene in social life. Now if you are participating in the project of the human sciences, you must perforce make these your premises. That is the nature of the project. Every day we move from interactions which are part of everyday life, to participation in specialized institutions and forms of practice, and in doing so we adjust the premises of our activity. Activity appropriate to natural science, everyday life, public discourse or artistic production are not invalidated by this, they are simply distinct from the project of the human sciences. It is a different way of appropriating the world, bringing different kinds of insight. The human sciences are sciences nonetheless, since they are held to the same epistemological standards as natural science, and exist within the same kind of institutional arrangements as the natural sciences. Someone participating in the human sciences does not cease to believe in the existence of Nature (that is, a material world prior to, outside of and independently of human activity), but reference to it is not internal to the practice of the human sciences. Contrariwise, to import activity into ‘ordinary’ natural science (i.e., natural science other than microphysics and cosmology) and to disallow the reification of the concepts of ‘ordinary’ natural science, is to violate a characteristic feature of natural science, a kind of category error.

At this point it should be clear that the type of claims to reality which are valid in one project (everyday life, natural science, ...) are incoherent in another project (artistic production, human science, public discourse, ...). In this context it should be noted that Bhaskar accepts that nothing in his philosophy is controversial within the practice of ordinary natural science. His philosophy is grounded in ordinary natural science, but is targeted at other projects, where, as we have said, different ontological foundations apply.

One more clarification. What is a concept? A concept is the coincidence of three moments: a word, symbol or other artefact by means of which it is represented and accumulates connotations and associations; individual human beings who know it and use it in their actions with each other; and particular systems of activity through which the concept is given meaning. Concepts arise in the life of institutions of some kind and may be meaningful only within such systems of practice, but as their activity merges with the general life of the community, the meaning of the concept becomes objectified, fixed and it enters the language of everyday life. But this needs to be clear: a concept refers to, or more precisely is indigenous to a system of activity. But as a concept it enters people’s lives and becomes ‘naturalised’, it is reified: that is, it is regarded as if it existed independently of human activity (is natural rather than constructed). Reification is not necessarily wrong. Within the context of everyday life or within relevant institutions it would be pedantic and wrong to deny reification. But this always has its limits. Times change, institutions fall, errors are exposed, practices change. Reification is always contingent.

Lenin said quite correctly in Materialism and Empirio-Criticism: “the only ‘property’ of matter connected with philosophical materialism is the property of being an objective reality.”

So while philosophical materialists affirm that Nature – a material world prior to, outside of and independent of human activity and human consciousness – exists, we are not philosophically committed to admit the existence of any particular thing or kind of thing: that is always a question of epistemology and particular forms of action, which entail the subject-object relation, an entirely different question from the ontological question which affirms at its outset, the distinction between consciousness and matter, between material things and our ideas of them.

These are my premises; different according to the nature of the project I am involved in. Let us now move on to look at how Bhaskar deals with this.

Bhaskar’s Answers to the Epistemic Fallacy

Let’s leave aside the various mad men and first year students of hermeneutics, who provide easy targets. Bhaskar himself claims that ‘generative mechanisms’ can be real, and we could clarify matters by teasing out the meaning of this claim. By ‘generative mechanism’ he means some process held to account for an observable event, such as maternal instinct, economic stimulus, universal grammar, Oedipus complex, HIV, etc. Bhaskar is at pains to distinguish between the theory (sociobiology, economics, etc.) in which the concept of the mechanism figures (which he calls the transitive object), and the objectively existing mechanism to which it refers (which he calls the intransitive object). How should Bhaskar answer the question: “Does the maternal instinct exist (even in respect to people who don’t believe in it)"? His answer is broadly and correctly of the form of “Carry out an experiment to test the hypothesis.” But an hypothesis is part of the theory, not the ‘intransitive object’. In other words, he cannot answer in any particular case in philosophical terms (and he explicitly agrees on this), but he can test the validity of the theory by experiment or other investigations, i.e., in practical activity.

If the experiment is positive and the hypothesis is confirmed, but someone insists: “Yes, you confirmed the theory, but does maternal instinct exist? Or is it just part of how you make sense of the world (just as your experiment was)?,” then Bhaskar can only point to the evidence, direct or indirect. That the other’s in-principle grounds for disbelief is some social theory doesn’t matter, they have to be shown how they can know that it is real, they need to be given evidence. The only answer he can give is an epistemological one of some kind, depending on the evidence available for the reality under question. In principle he can only answer such a question in particular, and he can and does answer it only epistemologically and practically. His core claim seems to be that generative mechanisms may exist, in general. One example is sufficient to settle this question, but if the other continues to deny such a thing could exist, then they need to be asked what they mean by ‘exist’, and on what evidence or criteria they would affirm or deny existence.

The main target of the ‘epistemic fallacy’ is the claim that such things as “instinct” or “complex,” which are meaningful (like any concept) only within the framework of some institution or community of practice, do not exist independently of that institution. Before leaping to mutual denunciations, simple clarification is obligatory here. All agree that the concept exists only within the relevant community of practice and its theories. Concepts belong to systems of human activity. Bhaskar claims that a concept refers to something which may exist in the world beyond human activity, though knowing whether it does exist, still depends on practice and evidence. A concept and its real referent is seen to be like the map of country and the country itself: we have the map, but do all the places on the map really exist?

The idealists deny, in principle, the existence of a ‘real’ (i.e., material) referent. That’s their problem. My response would be to say that there is or is not (within the given institution or social formation) a material basis for the concept (i.e., a specific system of activity) in the objective world, within the limits of what I know. I could not go further than to say that there is/is not a basis for it. Ultimately, it is always a judgment about a system of activity: knowledge can never be greater than the system of practices which support it. But whether the generative mechanism of a natural process can exist, i.e., whether the reification of the concept is valid, depends. Reification is valid so long as the conditions and systems of practice in which the concept figures allow it. To reify a concept means that the relevant system of actions has become fixed and ‘naturalized’.

It is just as inappropriate to impose the premises of the human sciences, in which activity figures as a substance, and in which different criteria for existence and reality apply, on either everyday life or natural science, as it is inappropriate to impose (as Bhaskar does) the substances and reality criteria of ordinary natural science on the human sciences. The point is that in human societies, belief in the reality of some concept is integral to the constitution of that society. Economic stimulus mechanisms may work in economic societies but not in other societies. Reality questions in this domain cannot be answered without reference to human activity and the concepts constituted in that activity. Now, although it would be pedantic in the extreme to equivocate on reality questions in reference to natural generative mechanisms, in certain contexts, it is appropriate. All societies have theories about the natural world which function as part of regulative mechanisms within their social life. In this context, it is more appropriate to answer a question like “Does chi exist?” with answers like “It does (does not) have a basis in traditional Chinese medicine.” That is, within the specific practice where this concept is reified, it acts as an organising principle, so long as the practice as a whole has a legitimate place in human life. And this is the point. Without the wish or the opportunity to take a reified concept out of its social context in order to test its reality in a different social context, when we know, likely as not, that the efficacy of the theory of which it is a part depends on the social context in the first place, it makes no sense to simply say that the reified concept (phlogiston, star sign, salvation, economic stimulus) exists. It has a basis in certain practices and the only material bases for judging it are its place in those practices and the practices themselves, insofar as they interact with other forms of life. To elide the distinction between practices and the theories they realize, and answer a question about the reification of a concept in terms of other forms of knowledge but detached from the forms of social practice to which such terms are indigenous, as if they were universal, opens the way to an ‘epistemic fallacy’. This is what Bhaskar does: he detaches a concept from the project to which it is indigenous and problematises it as if it were universal. Economic stimuli only work (if at all) within the theory and practice of economics and cannot exist outside of that. Strictly internal interpretation of texts makes sense only within hermeneutics. Unconditional reification only works within ordinary natural science.

Nevertheless, it would be a grave error to treat natural science as simply a community of practice with its own indigenous beliefs and concepts (like some religious cult or moral panic); science is a project subject to sociological and historical development, but its content is essentially universal, because it is characterized by reality criteria, and institutional forms and practices in which epistemological criteria are embedded. Science does not ‘believe in’ anything. It is nonetheless a distinct project whose reality criteria and substances cannot validly imposed on other projects.

Even Bhaskar’s most skilled acolytes cannot explain his philosophy without committing the ‘epistemic fallacy’ and nowhere in his works or those of his acolytes can I find an ontological definition or explanation of ‘real’ or ‘exists’. So the charge of ‘epistemic fallacy’ is empty. I know the difference between the thing itself and a concept of the thing, but there is nothing I can say about the thing other than what I can say about the activities which constitute it as an idea. A generative mechanism may or may not have a basis in the objective world. Within a certain domain there is no harm in reifying the concept and saying that it exists. But the fact is that there are limits to the validity of such reification.

Take a simple idea. The hardness of a thing cannot be determined without some test, whether scientific or everyday, but we know that subject to certain qualifications, such as temperature, hardness can normally be taken as a property of an object without reference to the actions by means of which it is measured. But what about capacitance? This is a concept defined within electrical theory and only rarely relevant to everyday life. If I were to introduces capacitance into everyday life I would be obliged to outline what this concept means in terms of how it is measured. And what about calling something a “phallic symbol” – a property solely of the size and shape of the object? The simplest empirical observation is always an action mediated by this or that artefact and/or interpreted on the basis of theories about the nature of the world. There is no such thing as direct or immediate observation. So we are always reifying. Whether the reification is valid may or may not be in question, depending on the context, but we are all born realists.

Bhaskar’s Concept of Thought

In a philosophy which aims to foreground ontology but cannot define ‘reality’ or ‘existence’ it is a serious problem that Bhaskar’s only commitment to a concept of thought is a qualified endorsement of Chomsky’s linguistics and Freud’s psychoanalysis, both based on the hypothesis of innate psychic structures within the individual human organism. He explicitly hedges his philosophical bets on psychology, allowing that consciousness could be anything from a Cartesian substance to an epiphenomena of brain tissue, but he seems to consistently regard consciousness and its forms a function or epiphenomenon of physiology, albeit physiology which interacts with social processes.

He claims that psychology and sociology are distinct levels or strata of nature, and that these strata are reflected in the structure of the sciences, rather than the structure of the sciences having constructed a stratification of nature, reflecting the stratification of social practices. In other words, he reifies the social/psychological dualism found in modern academia. Given that he sees thought as a function of physiology, shaped by experience, no wonder that he has such a problem with reification. He claims to transcend the differences between structuralism and methodological individualism by defining the unit of analysis as a ‘social relation’. But without any definite body of science resting on this unit of analysis, distinct from existing science, this is nothing but a gesture.

How can you insist on a realism, without clarifying the ontological status of consciousness? If consciousness as a property of individual physiology is supposed to settle the matter, then this is an ontology that would have passed muster 100 years ago, but it is not viable in these times. A philosophy claiming to address itself to matters of ontology and epistemology must have a foundation in some empirical theory of psychology, that is to say, a practically meaningful concept of what consciousness is. Otherwise, arguments about whether something is just a thought-form or exists outside of consciousness are empty. If thoughts exist between the ears, then the entire universe is outside consciousness (everything barring a few hundred grams of nerve tissue).

Although Bhaskar is fond of quoting “Theses on Feuerbach,” I never see him quoting 8: “All mysteries which lead theory to mysticism find their rational solution in human practice and in the comprehension of this practice.” Bhaskar never follows this lead, but a close study of the category of ‘activity’ does lead to a solution of the problems Bhaskar is failing to clarify. The main problem with the various social theories resting on discourse theory is that discourse is detached form the activities within which text gains meaning. Once discourse is restored to its place within forms of practice constituted by and realizing discourse amongst really existing individuals, the idealism of discourse theory is negated. Practice is both objective and subjective.

Recall Lenin’s advice that it is only the existence of matter as a whole, independently of human consciousness, to which philosophical materialists are committed. Everything else, all other concepts arise through the elaboration of human practice. Bhaskar’s main rhetorical technique is to make statements which are incontrovertible if made in reference to objective reality as a whole, but he makes them in relation to specific things. (Matter exists, but do strings?) Philosophical materialism has no commitment to the existence of any particular kind of thing. Only nave realism believes that the existence of particular things is self-evident. Nothing can be said about matter by philosophy other than that it exists independently of human activity. Any particular part, capacity or aspect of that material world, any ‘generative mechanism’, cannot be shown to exist without resort to practical investigation and the invoking of epistemological criteria.

At first sight, Bhaskar’s claim to investigate the transcendental conditions of possibility of experiment looks interesting. He calls this enquiry into conditions of possibility ‘underlabouring’. Whenever we do something, we begin with a presumption about the conditions for our intentions to succeed, but if something goes wrong we reflect on the conditions of possibility for the event. Scientific experiments have a rationale which is tied up with the continuity and intelligibility of Nature. This is an intriguing and profound question which goes much further than everyday experience: how is it that Nature is intelligible? Why are natural laws constant? Is randomness in nature a coherent concept? But I can’t see such questions being resolved on the basis of a philosophical ontology which lacks concepts of reality, existence and consciousness.

I think Bhaskar is uninteresting and misleading, and given that he fails to define or clarify the key concepts of his philosophy, his philosophy cannot be taken seriously.

Vygotsky on the Mixing up of Epistemology and Ontology

Bhaskar is not the only figure to have raised the sceptre of mixing up ontology and epistemology; Lev Vygotsky also raised this problem in 13 of “The Historical Meaning of the Crisis in Psychology: A Methodological Investigation” written in 1927, but he is talking about quite a different issue.

The context is what Vygotsky describes as the ‘crown and centre’ of his study of the world’s psychological literature which laid the foundations for the construction of his own psychology over the following years. The issue centres around the problem of if and how it is possible to observe consciousness. ‘Subjective psychologists’ claimed to be able to observe consciousness by introspection, whilst Behaviourists rightly rejected introspection as unscientific but wrongly declared that because the scientific observation of consciousness was impossible, the concept of consciousness must be excluded entirely as an object of science and there could be no Psych-ology, as such.

According to the various approaches to this problem, Vygotsky drew a sharp line between: on the one hand Phenomenology, the study of thought-forms in which being and appearance coincide, developed consistently by Husserl, and on the other hand, materialistic Psychology, which begins from Feuerbach, for which “even concerning thinking one must distinguish the thinking of thinking and the thinking as such” (quoting Feuerbach).

In developing this latter perspective he warned us not to “mix up the epistemological problem with the ontological one,” where, Vygotsky says “Epistemological consciousness as part of the antinomy ‘subject-object’ is confused with empirical, psychological consciousness.” The epistemological problem of the mind concerns that subject-object relation and with how we may gain knowledge of the object, here consciousness. The study of consciousness presents all the same problems as are presented in the study of any phenomenon. Remember, there is no such thing as immediate or un-mediated observation; all scientific work involves reconstruction of the object from objective facts (phenomena). Psychology is no different here. We gather evidence and with continued investigation we get a better and better picture of consciousness and over time the subject-object problem is successively resolved. Mind and body are the objects of our study and mental objects are objective in the same sense as any of the body’s functions. But the relation between consciousness and matter is an ontological problem. Matter is everything outside of and independent of consciousness. What is reflected in our consciousness will be different according to whether I am asleep or awake, drunk or sober. Consciousness is the immediate, the subject matter of Phenomenology, appearances, not something real, but nonetheless, the means by which we come to know the real.

Vygotsky explains this with a metaphor about reflection. The image in a mirror is nothing real; it is a ‘virtual reality’, the result of the intersection of two material processes, the optical properties of the mirror with its related optical processes, and the material processes producing the objects reflected in the mirror. It is just as wrong to identify the image in the mirror (the virtual object) with the object of which it is an image, as it is wrong to identify the image with the optical processes alone, which cannot create the object which is reflected. The image is nothing real, but an appearance, a ghost, which arises through the intersection of two material processes, and can be fully understood through the scientific study of these two material processes and their intersection. And you can’t get a reflection of a reflected image, because the reflection (virtual object) is not a real thing; it has no optical properties or any other material properties.

Likewise, our consciousness is the result of the intersection of human behaviour and human physiology, and can be understood through the study of the intersection of these two material processes. “My joy and my introspectional comprehension of this joy are different things”; states of consciousness cannot be ‘directly observed’, either by the study of physiology or of behavior or by introspection. But consciousness can be scientifically reconstructed through the study of human behavior and physiology.

Further however, human behaviour and physiology, it turns out, can only be understood by means of the concept of consciousness. Consciousness as a concept cannot be excluded from the science of human behaviour. Consciousness mediates between behaviour and physiology, just as behaviour mediates between consciousness and physiology and physiology mediates between consciousness and behaviour. But consciousness is no real thing; it is a phenomenon. Reality is what lies outside of consciousness. That is the meaning of reality after all.

“In the end, the question can be reduced ... to the differentiation of the ontological and the epistemological problem. In epistemology appearance exists, and to assert that it is being is false. In ontology appearance does not exist at all. Either mental phenomena exist, and then they are material and objective, or they do not exist, and then they do not exist and cannot be studied. No science can be confined to the subjective, to appearance, to phantoms, to what does not exist. ... The unreal must be explained as the non-coincidence, generally as the relation of two real things; the subjective as the corollary of two objective processes.”

Vygotsky noted that “Marxist psychologists” of his time either equated appearance with being, by taking introspection to be ‘objective’ observation of consciousness (Phenomenology), or substituted either physiology or behaviour for consciousness (Physiological or Social Behaviourism), while excluding consciousness from investigation on ontological grounds.

The difficulty of resolving these problems has led to innumerable confusions. The correct conviction that dichotomies set up barriers to the solution of epistemological problems is now conventional wisdom. Descartes’ original insight was that consciousness must be a substance (i.e., foundational concept) distinct from matter. The importance of this ontological insight is usually overlooked amidst denunciations of his dualism. But those who neglected his insight wound up with homunculi, or today, the identification of consciousness with its physiological basis. But a thought is not a neuron or even a system of neuronal reactions. An image is not a mirror or ‘mirror-state’. Descartes’ mistake was to set off epistemologically from this ontological dualism. Only classical German philosophy, much later, began to make headway on this problem of the subject and object.

Hermeneutists and decontructionists may claim that only texts, discourses or forms of consciousness exist, and if anything at all, only featureless matter exists outside of mind. On the other side are those who say, not that consciousness does not exist, or that consciousness is an appearance, but that consciousness is a form of matter or a ‘material process’. In the language of his times, Vygotsky calls this position ‘mechanical materialism’. This tendency has come back into popularity because the deconstructionists have convinced us all that dichotomies and dualisms are bad. So, if consciousness exists, and matter exists, consciousness must be a form of matter. But this is wrong. It is mixing up the epistemological problem of the subject-object relation where dualisms and dichotomies are anathema, with the ontological problem where the distinction between matter (which is real) and consciousness (which is not real, but is an appearance) is absolute.

Andy Blunden,
21 November 2009

PS. I believe the “maternal instinct” to be a myth, which nonetheless functions like all myths as a regulatory concept within a society.


Vygtosky on Ontology and Epistemology | Critical Realism site | Wikipedia on Bhaskar | Lenin on Ontology