Habermas J The Analytical Theory of Science and Dialectic -- a Postscript to the Popper/Adorno Controversy

In discussing totality,  Adorno uses categories which are indeed Hegelian in origin, and he rehearses arguments found in Dialectic of Enlightenment. His concept of totality is not referring to some organic whole, nor is it a 'class' [referring to the work of Nagel -- unknown to me I fear]  . Adorno's conception, like Habermas's, is a constitutive one. This conception has been attacked as ideological or mythological  [and discussion in Dialectic of Enlightenment is cited about the interconnections between myth and critique]. The positivist conception of the whole is different, a functional connection of empirical regularities. There are important differences between the concept  'system', and the concept  'totality', but this is difficult to discuss. If we choose to argue using only formal logic  'totality' dissolves, and if we use dialectic  'system' is transcended. In order to make progress in the Dispute we need to approach the issues  'from the outside' [surely Habermas is not suggesting some neutral observation language?]. Discussion might be organised around four distinctions:

1. The system can be seen as consisting of the formal connections of variables, existing independently of experience, correctable in laws or generalisations, and subjected to deductive hypotheses. How might this conception be applied to the concrete, though? It might serve as a framework, but there is no discussion about applying this system to the objective world. Any coincidence with the objective seems to be  'fortuitous', external to the theory itself  [this is usually glossed, via processes such as operationalism or coding etc]. Dialectical theory by contrast insists on  'appropriateness' [adequacy?]. Some fortuitous coincidence is acceptable if we're only interested in technical control, but there is a cognitive interest beyond the technical in dialectical theory.

Scientism extends frameworks used for technical control to other areas, and thus distorts the object. In social sciences this distortion of the object is especially ironic since it constrains the very subjectivity and liberating potential that positivism is designed to further.

Only the dialectic can offer a proper connection between objects and theory. There is no empiricist 'immediacy' (134), but a process of dialectical exploration, as in the hermeneutic circle. Thus  'categories gain their determinacy gradually' (134), relational at first, and then expressing substance and function. The concepts of dialectic are therefore  'reflexive... they retain a moment of the objective content which... they subject to analysis' (134).

2. The relation between theories and objects is transformed, and so is the relation between theory and experience. Analytical philosophers tolerate only one type of experience, gained from controlled observation of physical behaviours in laboratory conditions as they test intersubjectively valid judgments. All experience must be checkable in principle by this procedure. But this is a blind, restricted, abstract methodology which cannot possibly reunite full experience with theory: it is a restriction of experience. Experience is permitted only in some forms, although other forms still persist as pre-science: functionalist notions are untestable too at their core .

Dialectical theory can transcend positivism here as well. Positivism does and must relate to some kind of experience, but dialectic wants a greater amount of experience to count, even if it is untestable. Indeed, experience is better than empirical tests for some of the central theorems [such as the constitutive role of a totality]. Dialectical notions prove themselves in more than just instrumental ways, 'in the course of explication' (136). Here, even the categories used must be related to experience rather than grounded in formal logic which is then 'applied'. It is not just the restricted experience of falsificationism that matters, but the equally 'authentic' experience of authors and poets. Although Adorno does not make this specific point in his Introduction, he does so in practice, when preferring authentic versus commercially produced art, or when citing the Odyssey [in D of E].

Critical theory aims to relate to a context of German philosophy and culture which has been debased and has to be rescued or transcended. Critical rationalism relates to a British context of pragmatic social science which needs clarification and improvement away from common-sense. Perhaps these contextual matters lie behind the differences in approach? But are they irreconcilable? Critical rationalism aims to drag common-sense up to the level of philosophy, but critical theory then aims to transcend philosophy.

[Of course, critical theory has also attacked common-sense in its material on the culture industry. Further, Habermas wants to go beyond hermeneutic to depth hermeneutic. However he transcends this material by immediately invoking totality. This could be seen as being just as dismissive of experience as common-sense as positivism is! The intent behind clarification seems to matter as well. The critical rationalists choose examples where clarification is liberating, cases which would be agreed by critical theory as well, no doubt -- totalitarianism and myth, phoney rationalism as in fascism. Critical theory can also point to examples where clarification is not liberating, where it attempts instead to pin down issues in order to control them, as in various vulgar operationalisms of experience in order to manipulate people. This sort of intent is important and goes beyond mere philosophical discussion of abstract terms. There's a danger of hypostatisation in both positions -- critical rationalism celebrating clarity, while critical theory celebrates ambiguity and dialectic. In this Dispute, both sides back away from abstract commitments like this -- Adorno does allow a role for the principle of non-contradiction, while Albert does allow role for critical theory in exposing  'crude positivism'. However, both sides also stereotype the other side, and express indignation if it happens to them! Perhaps both approaches need more immanent critique deployed on themselves -- both need to be examined and judged by their own standards of clarity, or rigour, or explanatory power?]

3. The issues relating theory and experience also apply to the relation between theory and history. It should be possible to provide law-like knowledge for history as well as the natural science, for the positivists  [develop 'analytical/empirical' knowledge in Habermas's terms]. Thus history and science must aim at gaining the same form of knowledge. Empirical testing is supposed permit prediction through general laws and limiting conditions (expressed as causes). For history, the goal is to explain particulars, and it tends to use  'trivial laws, mostly psychological or sociological rules derived from experience' in order to infer some hypothetical cause for particular events  (137). Thus it is supposed to have the same logical form as a natural science but limiting conditions can only be hypothetical causes, for Popper. For him there can be no general historical laws, and no specific laws of history -- sociological rules are the focus of attention.

The dialectic, by contrast, relates the individual to the totality, and so its laws are broader. Objective context is the issue and how this plays a part in historical development. Dialectic does not deal with abstractions, so it is not anthropological, and there are no historical constants, but rather particular stages of unique development, which are also irreversible (138). Historical knowledge is revealed in knowledge of the object rather than just from analytic approaches. Dialectic does not operate with isolated cause and effect chains, but with entire general/particular links. It proceeds via hermeneutic analysis. The meaning so derived is constitutive rather than heuristic (139) -- the categories arise from the actual consciousness of actors in the  'objective spirit of the life world' (139).

Meaning is articulated through identification and critique. Thus dialectic does not just study subjective meanings, but needs an objective theory to  'account for that moment of reification which objectifying procedures exclusively have in mind' (139)  [in other words, to explain the reified reality that positivism just abstracts from?]. Thus objectivity is important, for not just as a series of laws relating objects, and so subjectivity, but not just as naively self-evident. When dialectic focuses on subjective meaning it is only in order to assess it  'behind the backs of subjects and institutions' (139) to reveal the historical totality. It is equally interested in the subjectively meaningless but important elements of constraint. In this way, specialist approaches used in social science such as Weber's verstehen are transcended.

In this way, theory transforms the concepts from outside, operating in that field of tension described by Adorno. In this way, the separation between theory and history is overcome. That is, history now has a theory rather than focusing on isolated specifics. Hermeneutics can also have a history. History is connected to the future too, in suggesting different possibilities for the same totality. Thus dialectic has a practical intent [in demonstrating all these possibilities] rather than aspiring to value neutrality for the social sciences (140).

4. The relations between theory and practice can be explored in a similar way. Retrospective connections between the two are available by examining history. For positivists, the intention is to generate laws and therefore predictions. But these must be restricted by the instrumentality of positivism -- for example in trying to develop more rational administration. There is an assumption that particular fields will be isolated and relatively static, and more complex connections escape attention. However, total analysis is unavoidable even for effective social planning  (141). This is admitted by Popper, but only as an  'heuristic'. But this means that the model of society can only be chosen arbitrarily -- there is in positivism no equipment to choose one option rather than another (142).

By contrast, critical theory needs to proceed properly, in the interests of emancipation, to ground projects in possibilities and real histories of the totality  [for example, to see what is reified and what is not?]. Thus the dialectic offers the more comprehensive rationality: a full account is needed for the dialectic, rather than the partial self-imposed bans of positivism.

Value freedom

There are differences between facts and decisions, says Popper, which is based upon his argument that laws are independent from subjects' influence and, more generally, unmediated by social forces. Thus we have two kinds of test -- an empirical one for the sciences, and a decisionistic one for social science. There is also an is/ought separation, with science belonging to the empirical side, and thus being unable to deal with values. However, knowledge is confined to the former side, while the latter offers only evaluation. Philosophy is also restricted -- to the latter side. In Popper's work, it concerns itself with  'objective value ethics', involving some purely ideal level of being above ordinary experience, and not based in history. This produces a dilemma: although philosophers try to be rational by pursuing 'intellectual honesty', and  'existential commitment', this means that their decisions are seen as a 'reduction' of norms  [that is, they are committed to value judgments stressing intellectual honesty and commitment which are assumed to be universal?]. Even this scheme is sometimes further positivised, and values become axioms with deducibles. But all depends on the decision to be rational. This decision is either a personal one, or it is anthropological, political, or even the result of a return of myth (146) [Popper rules out all but the first?].

Popper's  'open society'is the  'residual determination of thought' (147). But this involves either an  'unarticulated metaphysics', or a view that rationalism is an act of faith. It certainly cannot be based on any insight into experience. Its role is really the result of prioritising unreflected technical interests: the concept is rooted in pragmatism, which says that it is possible to direct our fate by developing rational social techniques  (148). This notion is at the heart of rational administration, and depends on an assumption that because we can be rational over nature, we can also be rational about our own fate.

The dualism we began with is the source of the problem. It is true that social norms are not just natural -- but knowledge is never totally value free.

Basic statements

In Popper, basic statements are supposed to validate the role of other statements, rather than experiences in themselves. Protocols are to tap experience. Falsification demonstrates the problems of inductivism, but also the power of any one existential statement. However, basic statements also involve universal terms which are themselves hypothetical and unverifiable, especially by some 'sense certainty'. This has long been recognised by Hegel and Husserl  [the latter with his discussion of the pre-predicative role of perception]. Peirce recognised the role of previous experience, and the anticipation that it will be continued  (151). Popper tells us that we cannot decide in advance, but that each basic statement is to be tested.

However, this process of falsification by basic statements produces a provisional consensus only. This leaves us with decisionism again [in deciding whether to proceed]  (151). The process works like the action of the jury, says Popper, but this reveals a circularity: in law, the facts first of all have to be established before the law can arbitrate, and the law itself pre-decides the relevance of the facts  (152). Only a hermeneutic analysis can help to clarify the pre-understandings and social functions behind these processes.

The actual organisation of the scientific 'jury' usually reveals very little dispute about basic statements or universals. Scientists tend to operate with unreflected norms and to be rather pragmatic about the role of experience. They may have doubts about single issues, but rarely doubt the whole apparatus  [this was to be argued by Kuhn in his defence of  'normal science']. Falsification is probably unwelcome as problematising general assumptions. Scientists find solutions in the end from their experience, grounded in work and the intersubjective agreement about their work, as well as a long-term interest in the domination of nature. Procedures seem less important than these basic constituting processes. Underlying interests like technical domination affect the whole processes and rules of science. [Odd this -- these underlying interests would also not be apparent in the everyday procedures that are so important in beating off Popper?] An interest in work is that force that binds the scientific community together. Again, only hermeneutic analysis can reveal this, scientific methods are incapable of clarifying the activities of scientists.

The interest in work is a constant, although actual norms rules or techniques can differ between say a court room and the science lab. This constancy provides the underlying basis for consensus, despite the differing cultural variables (155) [hence the apparent worldwide success of science]. This interest is so powerful that it is often forgotten, so universal that it is taken for granted (155), hence the  'illusion of pure theory', found in philosophers from Plato onwards. Only in modern bourgeois societies does the practical intent of science dominate, and the needs of new kinds of manufacture feed back into science. The spread of this society explains the monopoly of natural science.

This forgetting of the important social connections with science is assisted by an interest instead in the more specific connections within science, like the theory- practice relation  (157). In this way a 'theoretical' attitude can develop which is apparently devoid of interests -- or value-free.

In this sort of analysis, we can see that the forces that constitute modern science cannot be explained by science alone. What is required is a proper connection with the life process, a discussion of the relation of objects and subjects. Value freedom, and technical banishments prevent these problems being clarified. Thus, science is normative, and it reflects one normative interest in particular.

On Weber

Weber's analysis [of the relation between means and ends or goals in rationalization] indicates some of the problems. However, Weber's own view of science allows important value judgments to operate only in the choice of goals -- analysing the situation and choosing the best means to proceed apparently involves purely logical  'if - then' relations.

However, some means/ends relations are not like this. For example, some goals become the means for the attainment of goals -- so values must be attached and then detached? Even in Weber's scheme, technical calculations are still connected to the life process, because these are always problematic when applied to social life. The conditions affecting choice are particularly complex. Moreover, means and ends often cluster in practice [the choice is not calculated afresh each time]. Practical action aims to make all this unproblematic rather than open to calculation (160).

Again, we need to understand these social processes. The finding that Weber's calculations do fit social life is not because they are simply right and that there are no discrepancies between technical rationality and practical activity, but rather because the practical has already been smuggled into the technical. The enormous possible combinations of means and ends has already been reduced, with some excluded as irrelevant (161). Only after the  'great mass of all conceivable constellations... [is]... eliminated... can value-free investigation commence' (161).

All apparently formalised theories are in danger of operating in the same way. The ageing not realise the importance of the technical/cognitive interest which pre-selects aspects to investigate  'universally'. Proper investigative procedures would recognise this and control it. Once more the dialectic is needed. Although tight methodological control can be valuable in practical terms, it makes for poor and  limited theory .

contents page