chapter  3
Bhaskar’s transcendentalism and related philosophies
Pages 13

I mentioned in Chapter 2 that Bhaskar ([1975] 1978: 13-15) even stratifies reality according to the degree we are able to observe or not observe reality. To be precise, he distinguishes between three domains of reality: the empirical (consisting of empirically observable events), the actual (containing not only observable events but also non-observable ones), and the real (harbouring the two preceding domains plus the generative mechanism of natural laws and their causes). This last domain may seem surprising to academics trained in empirical methodology, but it appears to be a major reason why Bhaskar and his particular brand of Critical Realism (CR) have gained considerable popularity during the last quarter of a century – particular in the UK and among social scientists. Roy Bhaskar is a British philosopher of Indian descent who tried – through his ‘transcendental realism’ – to overcome the limits of empiricism and venture into domains that lie beyond the observable – i.e. to entities that are inaccessible to our senses. His realist theory of science (Bhaskar [1975] 1978) is oriented towards the natural sciences, while his subsequent work, The possibility of naturalism (Bhaskar 1979), deals with the social sciences. The latter presents a critical ‘naturalism’ that assumes the social sciences can, at least to some degree, be treated similar to the natural sciences. It is claimed to be a realistic but flexible approach positioned between positivism and the poststructural trend. Bhaskar considers his brand of realism to be ‘socially situated’ but not ‘socially determined’; it even aims at motivating social change. As to Bhaskar’s categories of reality, the question arises: ‘why go beyond the rational and empirical categories?’. Is it ever possible to gain knowledge by trying to transcend those categories? But to probe the limits of the empirical is by no means a new quest. The mere fact that Bhaskar calls his realism ‘transcendental’ (emphasizing his intent to go beyond the purely phenomenal realm), indicates that he is taking up the legacy that Immanuel Kant left us over 120 years ago. However, Kant applied the expression ‘transcendental’ to idealism rather than realism. Nevertheless, there is no doubt that Bhaskar goes back to Kant who also pondered about that which transcends the realm of sense experience – namely the unfathomable thing-in-itself (das Ding an sich, in Kant’s terminology). However,

for Kant, as an idealist, the thing-in-itself was inaccessible. While Bhaskar, a critical realist, seems to have higher ambitions. And this boldness seems to have sparked the imagination of many young scholars of our time. However, during the long period dominated by empirical methodology, there were other attempts to go beyond Kant’s brand of transcendental idealism and his belief that the thing-in-itself cannot be known. Among the most famous attempts was that of Schopenhauer who considered himself the ‘true’ heir of Kant. And since I find some parallels between Schopenhauer’s ‘solution’ to the thing-in-itself, and Bhaskar’s notion, I shall here concentrate on such a comparison – even if this may appear somewhat outlandish. But the quest for some closure about ultimate reality is by no means trivial – indeed, it is indispensable in any ontological exploration. Schopenhauer’s ([1819] 1936) magnum opus is called Die Welt als Wille und Vorstellung (usually translated as The world as will and idea). But ‘Vorstellung’ means in German not only ‘idea’, it also means ‘representation’. Thus, Schopenhauer’s work may be interpreted as a dual analysis: on the one side, of reality itself (the universal Will to exist – and not only in a biological sense); on the other side, as the conceptual representation of the bits and pieces of reality. Such a duality also seems to appear in Bhaskar’s work. Indeed, Cruickshank (2004) wrote a critique on Bhaskar’s work entitled ‘A tale of two ontologies’. Cruickshank’s (2004: 567) paper pivots on the idea that Bhaskar and his followers seem to operate with two different conceptions of ontology. Ontology is the conceptual representation of reality; and this has to be strictly distinguished from reality itself. Bhaskar might regard this as committing the ‘epistemic fallacy’ (i.e. taking ontic entities for epistemic ones – that means confusing real entities with knowledge about them). But this is here not the case. If it were, it would mean to confuse the realm of knowledge, truth, confirmation, corroboration, etc. with the realm of existence, reality and its structures, etc. The former deals with propositions and their truth-value; the later refers to entities such as things, relations, processes, natural laws, and other properties. Although there are close ties between ontology and epistemology, we not only have to make a distinction between reality and its ontological representation, but also between the latter and representation in the epistemological sense (cf. Mattessich 1970, 1992). Unfortunately, the term ‘representation’ is often exclusively reserved for the epistemic. But ‘representation’ does not only refer to talking about correspondence, truth, confirmation; it is equally employed in talking about existing entities. How else could we ever talk about reality if not by means of our conceptual and linguistic representations? In defence of Bhaskar, it has to be admitted (as the quotation from Cruikshank reveals) that by making the distinction between the ‘transitive domain’ (i.e. our fallible theoretical interpretations of reality), on one side, and the ‘intransitive domain’ (i.e. reality beyond our knowledge), on the other side, the confusion between the two domains is mitigated. Yet, this does not eliminate the potential category mistake pointed out above. In this connection it ought to be mentioned that Bhaskar introduces a novel conceptual apparatus that may be

bewildering to scientists and philosophical novices. Although he initially uses the term ‘transcendental’ in characterizing his type of realism, he then, surprisingly, borrows unrelated terms from grammar. Instead of speaking of ‘transcendental’ (for the empirically unknowable) he speaks of ‘intransitive’ categories; and instead of speaking of ‘non-transcendental’ (for the empirically knowable) he speaks of ‘transitive’ categories. Hence, disciples of CR regard the intransitive dimension as the ontological one, and the transitive dimension as the epistemological dimension (cf. Groff 2000: 413) – though it might be more accurate to regard the intransitive dimension as that of the Real instead of the ontological. Remember our assertion above that ‘ontology is the conceptual representation of reality’, not reality itself. In a later publication (Bhaskar and Laclau 2002), Bhaskar seems to have modified his formulation such that it is knowledge which has two dimensions: a transitive, artificial one, used as conceptual reference to the world; and an intransitive dimension referring to the referent (i.e. reality). Yet, let us continue comparing Bhaskar’s and Schopenhauer’s search for the thing-in-itself. Indeed, there exists another parallel between them, because Schopenhauer might also be accused of a category mistake. He too endows both of his domains – the representative one (his Vorstellung) as well as actual reality (his Wille) – with the term ‘Welt’ (‘world’ or reality). For him, ‘the world is given twice’. Hence, in Schopenhauer’s case we have the reverse situation from Bhaskar’s. From the latter’s writings one may infer that it is ‘ontology’ that has two meanings, while Schopenhauer claims that it is reality (‘die Welt’) that appears to us twice. However, the most interesting aspect of the comparison between Schopenhauer’s theory and that of Bhaskar lies in the following question. If Kant could not find a way to comprehend the thing-in-itself, what are the solutions of Schopenhauer, on one hand, and of Bhaskar, on the other? In the case of Schopenhauer, the solution to comprehend reality in its very essence is the universal, ceaselessly active Wille that manifests itself in every human being no less than in other biological creatures as well as in all physical phenomena. But then another question arises. If the thing-in-itself manifests itself in such a variety of ways would it not be more correct to call it ‘force’ or ‘energy’ instead of ‘Will’ (cf. Copleston 1965: 37)? Hence, one might interpret Schopenhauer’s solution to the essence of reality and the thing-in-itself, as the universally penetrating and restless power of energy that seems to be the ultimate cause of all change and everything else – save for the ‘unmoved mover’, postulated by Aristotle.1 And that seems to be as far as Schopenhauer could get in what might be one of the most original ontological explorations ever undertaken. But what is Bhaskar’s solution? If one can speak of a ‘solution’, it is surprisingly similar. One of Bhaskar’s major concerns is the problem of causality. He rejects Hume’s proposition that is considered to go hardly beyond equating causality with constant correlation (see our note 9 to this chapter). Bhaskar, in contrast, wants to transcend the empirical; he searches for the generative mechanism of causal forces in general that he designates as ‘causal powers’. If Bhaskar speaks of ‘a cause’ he does not mean the mere

conjunction of actual events, but a potential mechanism, a power pertaining to an entity. I find it striking that such related notions as universal will or energy, on one side, and generative mechanism of causal power, on the other side, constitute the centrepiece in each of these two, in many respects utterly different, philosophies. Even if no one else seems to have bothered connecting these two apparently dissimilar ontological perspectives, there can be little doubt of the astonishing affinity that exists between them. There is a further aspect to comparing Schopenhauer’s and Bhaskar’s search for the thing-in-itself (‘will’ or ‘power’ or ‘generative mechanism of causality’). As to the ‘solution’ to this problem, Schopenhauer seems to have a trick up his sleeve. The universal force or ‘will’ to strive and exist is not only ‘out there’ but is also inherent in every human being. This is claimed to afford us a more ‘direct’ personal access to that all-pervading power, the thing-in-itself – you might say, through a more intimate ‘back door’. But Bhaskar could use a similar argument. He could claim that ‘causal powers’ (or the natural laws acting in them) are not only ‘out there’ but are constantly experienced by each of us personally, and thus are experienced ‘intimately’. Yet, Bhaskar’s solution refers less to this ‘personal experience’ than to the universal essence of ‘causal powers’ – though neither in the Humean sense, as a mere conjunction of events, nor in the Kantian way, as a necessity given to us a priori. If causal laws (i.e. ‘natural laws’) are ‘tendencies’ (or ‘propensities’) then they are properties of things, and that conforms to Bunge’s definition of such laws (cf. Bunge 1977: 77-80). What Bhaskar seems to convey by ‘tendencies’ (in the penultimate sentences of the above quote) is this: natural laws maintain their tendencies whether experimentally invoked or not. Pertinent experiments may fail, but if so, it may be because some of the antecedent conditions (beyond the law itself ) may not have been fulfilled – or there may have been some other intervening events. Hence natural laws must not be seen in the confines of a closed experimental system but from the viewpoint of an open system. In other words, natural laws transcend the mere experiential and experimental, their tendencies and potential powers exist even if not invoked (cf. Section 1.2, ‘fourth’ item – where I compare the logical explanation of a cause with the inference of Hempel’s covering laws). Indeed Bhaskar provides a model of science different from the traditional one; a model where causal laws are not regarded as ‘universally true statements but as references to the natures and powers of things’ (Outhwaite 2000: 11). But we got side-tracked and should return to the ‘thing-in-itself ’. What are the ‘solutions’ of modern science to this problem? For me, the most plausible interpretation has been advanced by Donald T. Campbell (1974a) under the name of ‘hypothetical realism’ (see also Lorenz [NP 1973] 1977: 8) – which I take to be a special interpretation of critical realism (for a more precise definition of ‘hypothetical realism’, see Hancil 2010: 1). The pivotal assumption is that our cognitive capacity has evolved genetically through constant interaction with the external world. Even if our knowledge is

hypothetical and open to correction, our tools for knowing reality are not limited to reason and sensory experience (as well as their technical extensions). They also consist of the inherited endowments of instinct and emotional wisdom, collectively accumulated during billions of years of biological (and ‘lately’ cultural) evolution and interaction with the environment – i.e. with reality. Indeed, for me cosmic evolution is equivalent to the impact of an immense number of chance events upon reality. This combination of personal experience and innate capacity may resemble Kant’s synthetic a priori – yet it is different. Kant tried to resolve or reconcile the conflict between the rationalism of Descartes, Spinoza, and Leibniz, on one side, and the empiricism of Locke and Hume, on the other side, by recognizing the contribution made by reason as well as sense experience in the creation of knowledge. Kant postulated, in addition to an ‘analytic a priori (e.g. the propositions of logic and mathematics), also a ‘synthetic a priori’ (i.e. a kind of ‘factual knowledge given to us in advance’). Yet, the latter notion is no longer tenable. Kant believed, for example, that space, time, causal necessity, etc. are given a priori to our mind; but this does not stand up in the face of general relativity theory. However, the advances of biology and neuroscience lead to a notion that resembles Kant’s ‘synthetic a priori’ – though it must not be identified with it. First, even if our ‘emotional wisdom’ (including other inherited qualities that help us to perceive structural analogies of external reality) is given to each of us a priori, it is not given synthetically, but evolutionary. And second, if you look at it not from a personal but a collective-evolutionary point of view – namely through the inherited experience and environmental interaction of our ancestors with the environment (reaching back to the earliest beginnings of life) – then this knowledge is acquired not a priori but a posteriori (i.e. not in advance but ex post through the evolutionary experience).2 Although we may never know the very essence of the thing-in-itself, we cannot deny the ability to comprehend crucial structural and dynamic feature of reality through a combination of various cognitive abilities. Hence, reason, experience, and emotional wisdom, all these together replace Descartes’ ‘guarantor’, assuring us that we are not utterly deceived about reality. Apart from such thoughts, Lorenz possesses a similar trump-card as does Schopenhauer (and possibly Bhaskar). Remember, for Schopenhauer it was the universal Will that constitutes ultimate reality, but that Will also resides in us. Lorenz emphasizes the brain/mind system that enables us to ‘mirror’ reality, but that mirror itself is an integral part of the very reality we try to grasp. And it is the constant and intimate interaction between reflector and the reflected (tested over millennia) that is the guarantor for reasonable approximations to reality and truth. If looking at such facts from a pragmatic or commonsense point of view, we might ask: if structural and dynamic analogies to reality were not accessible to us (at least in an approximate way), how could we have ever attained the insights for creating all those scientific-technological ‘marvels’ – from spaceships and hydrogen bombs to computers and nanotechnologies? Not only that; if evolution had not endowed creatures to grasp their environment in some way, then not

even a primitive jellyfish could orient itself in the depth of the ocean. Perhaps the most striking illustration in favour of a realist stance is the amazing camouflage behaviour of the octopus (and many other animals). It has been documented that the ability of the octopus to assume the pattern of its environment is a function of its brain. But how could any animal reflect its environment (i.e. its immediate reality) so perfectly, if it had not the ability to perceive and represent structural and dynamic similarities of this reality? Of course, there is the big challenge from quantum theory which demonstrates that the reality of atomic and subatomic particles is observer-determined (see our Chapter 11). This is hardly deniable but, as Pagels (1983: 161) points out, ‘we only have to worry about the observer-determined reality of quantum sized objects’ because ‘the macroworld can store information while the microworld cannot.’ However, we still have to connect the above analysis with Bhaskar’s distinction of three domains of reality (the empirical, the actual, and the real). Hence, the first of these domains is a subset of the two preceding ones, and the second domain as a subset of the last one. Yet, the reader might ask: why do we need, in-between the empirical and the real, an actual domain? What is the difference between the actual and the real? Does Bhaskar mean the actual domain includes also entities not yet observed but potential observable and explainable in the future? Examples that come to mind are dark matter (about which we know very little) and dark energy (not to be confused with dark matter) which is a more recent notion about which we know even less (see also Chapter 11). Or does Bhaskar mean something else? It seems that the boundaries of this second category (the actual) are vague and might not warrant to be considered as a separate category. Since the entities of the actual, as well as those of the real, are unobservable, why make a distinction between the two? How do we know that those of the latter are unfathomable while those of the former are not? These are questions that inevitable arise when contemplating Bhaskar’s CR. Nevertheless, it is through this stratification that Bhaskar tries to transcend the positivistic-empirical approach, and attain a conception (though no comprehension) of the generative causal mechanisms or its ‘tendencies’ (as a possible explanation of the thing-in-itself ). This might offer a possibly more complete picture of an ‘outer’ and ‘inner’ reality as well as its relation to scientific knowledge (as preliminary or fallible as the latter might be). But some might ask: even if the thing-in-itself is thus clarified as the force of a universal Will (in Schopenhauer’s case), or as an evolutionary endowment (in the case of Lorenz), or as the causal powers or tendencies (in the case of Bhaskar), did we attain a final solution? In the Middle Ages and during Renaissance as well as the Enlightenment, the answer to the question of ultimate cause was God, and nowadays, in our modern scientific area, the answer is Nature. And since the latter is best represented by its laws and their causality, we may have come as close to an answer as we might ever get. And yet, there is one further aspect to be considered. It is the fact that some philosophers distinguish scientific laws (in my terminology ‘scientific law

statements’, describing the contingent scientific view of such fundamental regularities) from laws of nature (‘prescribing’ the actual causal forces of reality as a necessity).3 Others (e.g. Scriven 1961; Cartwright 1983; Swartz 2009) distinguish, in addition, between two theories of the second category – i.e. of ‘laws of nature’. I shall not enter into the details of the philosophical controversy between those two competing camps but shall refer the reader to Swartz (2009). However, this issue does concern several points of some concern to ontology. Hence, I shall sketch it in rough outlines. In reading Swartz’ quote, one might get the impression that it is not at all a juxtaposition of two views of laws of nature but, on one side, the description or representation of laws versus, on the other side, the laws themselves (similar to the distinction between ontology vs. reality). But despite the deceptive formulation, this is not the case. We are dealing with two fundamentally different interpretations of far reaching consequence for the beliefs of those who hold one view versus the other view. The Regularists or Humeans and Neo-Humeans – despite the twist that Hume himself was a Necessitarian, because his scepticism was purely epistemological, not ontological (cf. Beauchamp and Rosenberg 1981) – deny the physical necessity of any such laws. In the view of Regularists (e.g. Swartz 2009) natural laws do not govern nature but merely ‘describe’ or, as it might be less confusing to say, ‘present us’ with those regularities. But Regularists are on the verge of obliterating the distinction between lawful and accidental regularities and assert that explanations can be given without nomologically necessary laws. Necessitarians (e.g. Ellis 2001; Bird 2005; Carroll 2012), in contrast, insist on the physical or universal necessity of such laws. But Regularists complain that ‘Necessitarians – unwittingly perhaps – turn the semantic [correspondence] theory of truth on its head. Instead of having propositions taking their truth from the way the world is, they argue that certain propositions – namely the laws of nature – impose truth on the world’ (cf. Swartz 2009: item d). They also claim that the Necessitarian view is unempirical and cannot be reconciled with the existence of statistical laws of nature. Regularists also claim that their view is able to ‘dissolve’ the problem of Free Will. As to myself, leaning in direction of the Necessitarian view, I find the Regularists’ claims tenuous and refer the reader for detailed arguments to Swartz (2009) who is more favourably inclined towards the Regularist position. One might now ask whether Bhaskar’s CR conforms to the Regularity theory or to that of the Necessitarians. In Bhaskar ([1975] 1978, 1979) he hardly offers any direct answers to this question (the subject index of neither of these two earlier books refers to Regularists or Necessitarians), but since he deems laws of nature as dispositional properties, there can be little doubt that his view conforms to that of the dispositional Necessitarians (like Ellis 2001; Bird 2005 and others). There still is a further item to be considered when discussing CR: namely, its preferred mode of inference. Empiricists rely almost exclusively on induction and deduction. But the disciples of CR believe that for the social sciences, such lesser known modes of inference as abduction and retroduction are more appropriate (for a discussion of these modes, see Wuisman 2005: 367).