JÜRGEN HABERMAS
A Reply to a Pamphlet1
Hans Albert has criticized an essay on the analytical
theory of science and dialectics in which I took up a controversy which developed
between Karl R. Popper and Theodor W. Adorno at the
Before entering into the discussion, I must make certain comments in order to establish agreement concerning the basis of our dispute. My criticism is not aimed at research practices in the exact empirical sciences, nor against those in behavioural-scientific sociology, in so far as the latter exists. It is another question whether this can exist beyond the confines of small-group research with a social-psychological orientation. My critique is exclusively directed at the positivistic interpretation of such research processes. For the false consciousness of a correct practice affects the latter. I do not dispute that the analytical theory of science has stimulated actual research and has helped to elucidate methodological judgments. At the same time, however, the positivistic self-understanding has restrictive effects; it silences
1 Cf. Hans Albert, 'The Myth of Total Reason'. Page references in the text refer to this essay.
2 The Positivist Dispute, pp. 131ff.; in addition, Albert refers to several passages in my essay 'Dogmatism, Reason, and Decision' in Jürgen Habermas, Theory and Practice (London/Boston, 1974), pp. 253ff. He does not take the book as a whole into account.
[198/199] any binding reflection beyond the boundaries of the empirical-analytical (and formal) sciences. I reject this masked normative function of a false consciousness. According to positivistic prohibitive norms, whole problem areas would have to be excluded from discussion and relinquished to irrational attitudes, although, in my opinion, they are perfectly open to critical elucidation. Moreover, if those problems connected with the selection of standards and the influence of arguments on attitudes were inaccessible to critical discussion and had to be abandoned to arbitrary decisions, then the methodology of the empirical sciences themselves would be no less irrational. Since our chances of reaching agreement on contentious problems in a rational manner, are in fact, quite limited, I consider reservations of principle, which prevent us from exhausting these chances, to be dangerous. In order to guarantee the dimension of comprehensive rationality and to penetrate the illusion of positivistic barriers, I shall adopt what is really an old-fashioned course. I shall trust in the power of self-reflection. When we reflect on what happens in research processes, we realize that we move the whole time within a spectrum of rational discussion, which is broader than positivism regards as permissible.
Albert isolates my arguments from the context of an immanent critique
of Popper's view. Consequently, they become confused—I myself scarcely
recognize them. What is more, Albert creates the impression that, with their
help, I intend to introduce something approaching a new 'method' alongside the
already well-established methods of social-scientific research. I have nothing
of the sort in mind. I selected Popper's theory for discussion because he himself
had already confirmed, in some measure, my doubts about positivism. Influenced
by Russell and the early Wittgenstein, it was above all the
The charge of misunderstanding, however, has already been levelled at me by Albert. In his opinion, I am mistaken on the following points:
Furthermore, Albert asserts that the pragmatist interpretation of the empirical-analytical sciences is erroneous. In the last analysis, he considers that the opposition between dogmatically fixed and rationally substantiated positions is a falsely posed alternative which has been made redundant by Popper's critical rationalism itself. I shall discuss these two objections in the context of those four 'misunderstandings', which I wish to resolve in that order. The reader may then decide on whose side they lay.
I do not like encumbering a sociological journal with details of the theory of science, but we cannot carry on a discussion as long as we stand above matters instead of in their midst.
1 CRITIQUE OF EMPIRICISM
The first misunderstanding relates to the methodological role of experience in the empirical-analytical sciences. Albert rightly points out that experiences of diverse origin can intervene in theories regardless of whether they spring from the potential of everyday experience, from historically transmitted myths, or from spontaneous impressions. They merely have to fulfill the condition that they can be translated into testable hypotheses. For the test itself, on the other hand, only a specific mode of experience is permitted, namely, sense experience, which is organized by experimental or analogous procedures. We also speak of systematic observation. I, for my part, have in no way questioned this influx of unordered experiences into the stream of imaginative [200/201] leaps out of which hypotheses are created, nor would I fail to recognize the merits of test situations which organize sense experiences through replicable tests. But if one does not wish to enthrone philosophical innocence at any price, the question must be permitted whether, through such a definition of the preconditions for testing, the possible meaning of the empirical validity of statements has not been established in advance. And if this is the case, one might ask what meaning of validity is thereby prejudiced. The experiential basis of the exact sciences is not independent of the standards which these sciences themselves attribute to experience. Apparently, the test procedure, which Albert suggests is the only legitimate one, is merely one amongst several. Moral feelings, privations and frustrations, crises in the individual's life history, changes in attitude in the course of reflection—all these mediate different experiences. Through corresponding standards they can be raised to the level of a validating instance. The transference situation existing between doctor and patient, which is utilized by the psychoanalyst, provides an example of this. I do not wish to compare the advantages and disadvantages of the various test procedures; instead, I simply wish to elucidate my questions. Albert is unable to discuss them because he calmly identifies tests with the possible testing of theories against experience in general. What I regard as a problem he continues to accept without discussion.
This question interests me in connection with Popper's objections to the empiricist presuppositions of more recent positivism. Popper challenges the thesis of the manifest self-givenness of the existent in sense experience. The idea of an immediately attested reality, and of a manifest truth, has not withstood critical epistemological reflection. Since Kant's proof of the categorial elements of our perception, the claim of sense experience to be the final court of evidence has been dismissed. Hegel's critique of sense certainty, Peirce's analysis of perception incorporated in systems of action, Husserl's explication of pre-predicative experience, and Adorno's attack on First Philosophy have all, from their various points of departure, proved that there is no such thing as immediate knowledge. The search for the primary experience of a manifest immediacy is in vain. Even the simplest perception is not only preformed categorially by physiological apparatus—it is just as determined by previous experience, through what has been handed down, and what has been learned [201/202] as by what is anticipated, through the horizon of expectations, and even of dreams and fears. Popper formulates this insight in the statement that observations always imply interpretations in the light of experiences made and knowledge acquired. More simply, empirical data are interpretations within the framework of previous theories; as a result, they themselves share the latter's hypothetical character.3
Popper draws radical conclusions from this state of affairs. He reduces all knowledge to the level of assertions, of conjectures, with whose help we hypothetically complete an insufficient experience and interpret our uncertainties concerning a concealed reality. Such assertions and conjectures are differentiated merely in the extent to which they may be tested. Even tested conjectures which are constantly subjected to rigorous tests do not attain the status of proven statements; they remain suppositions, admittedly of a kind that hitherto have withstood all attempts to eliminate them—in short, they are well-tried hypotheses.
Empiricism, like the traditional critique of knowledge in general, attempts to justify the validity of exact knowledge by recourse to the sources of knowledge. Yet the sources of knowledge—pure thought, established tradition and sense experience— all lack authority. None of them can lay claim to immediate evidence and primary validity and consequently to the power of legitimation. The sources of knowledge are always contaminated; the way to their origins is barred to us. Hence the question of the source of knowledge must be replaced by the question of its validity. The demand for the verification of scientific statements is authoritarian because it makes the validity of statements dependent upon the false authority of the senses. Instead of inquiring after the legitimating origin of knowledge, we have to inquire after the method by means of which definitively false assertions can be discovered and apprehended amidst the mass of assertions which are, in principle, uncertain.4
3 Karl
R. Popper, Conjectures and Refutations (
4 Conjectures, pp. 3ff, 24ff.
Popper carries this critique so far that it unintentionally makes even his own suggested solution problematical. Popper strips the origins of knowledge, enlisted in empiricist studies, of their false authority. He rightly discredits every form of primary knowledge. But even mistakes can only be falsified on the basis of criteria of validity. For their justification we must adduce [202/203] arguments; but where, then, are we to look for these if not in the very dimension—not of the origin but of the formation of knowledge —which has been ruled out? Otherwise, the standards of falsification remain arbitrary. Popper wants to sublate the origins of theories, namely, observations, thought and tradition alike, in favour of the method of testing which is to be the only way of measuring empirical validity. Unfortunately, however, this method, in its turn, can only be grounded by recourse to at least one of the sources of knowledge, to tradition, in fact to the tradition which Popper calls the critical tradition. It becomes clear that tradition is the independent variable upon which, in the final instance, thought and observations are just as dependent as the testing procedures, which are composed of thought and observations. Popper is too unhesitating in his trust in the autonomy of the experience which is organized in the testing procedure. He thinks that he can dismiss the question of standards in this procedure because, for all his criticism, he shares, in the last analysis, a deep-seated positivistic prejudice. He assumes the epistemological independence of facts from the theories which should descriptively grasp these facts and the relations between them. Accordingly, tests examine theories against 'independent' facts. This thesis is the pivot of the residual positivistic problematic in Popper. Albert does not indicate that I might have succeeded in even making him aware of what is at issue here.
On the one hand, Popper rightly counters empiricism with the objection that we can only apprehend and determine facts in the light of theories.5 Moreover, he occasionally describes even facts as the common product of language and reality.6 On the other hand, he assumes for protocol statements, which are dependent on a methodically secured organization of our experiences, a simple relationship of correspondence to the 'facts'. Popper's adherence to the correspondence theory of truth does not seem to me to be consistent. This theory presupposes that 'facts' exist in themselves, without taking into account that the meaning of the empirical validity of factual statements (and, indirectly, the meaning of theories in the empirical sciences too) is determined in advance by the definition of the testing conditions. It would instead be more meaningful to attempt a basic analysis of the connection between the theories of the empirical sciences and the [203/204] so-called facts. For in this way, we would apprehend the framework of a prior interpretation of experience. At this level of reflection, it would seem obvious to apply the term 'facts' only to the class of what can be experienced, a class which has been antecedently organized to test scientific theories. Then one would conceive of the facts as that which they are: namely, produced. One would thus recognize the concept of 'facts' in positivism as a fetish which merely grants to the mediated the illusion of immediacy. Popper does not complete the retreat into the transcendental dimension but the consistency of his own critique leads in this direction. Popper's presentation of the basis-problem shows as much.
5 Conjectures, p. 41, note 8.
6 Conjectures, p. 214.
2 THE PRAGMATIC INTERPRETATION OF EMPIRICAL-ANALYTICAL RESEARCH
The second misunderstanding, of which Albert accuses me, refers to the so-called 'basis-problem'. Popper gives the name 'basic statements' to those singular existential statements which lend themselves to the refutation of a law-like hypothesis expressed in the form of negative existential statements. Normally, these formulate the result of systematic observations. They mark the point of contact at which theories strike the basis of experience. Basic statements cannot, of course, rest upon experience without contact with it, for none of the universal expressions which occur in them could be verified, not even with the aid of a large number of observations. The acceptance or rejection of basic statements rests, in the last instance, on a decision; but the decisions are not made in an arbitrary fashion. Rather, they are made in accordance with rules. Such rules are only laid down institutionally, not logically. They encourage us to direct decisions of this sort towards an implicitly pre-understood goal, but they do not define it. We behave in this way in the course of everyday communication and also in the interpretation of texts. We have no choice when we move in a circle and yet do not wish to forego explication. The basis problem reminds us that even applying formal theories to reality entangles us in a circle. I have learned of this circle from Popper; I did not invent it myself, as Albert seems to suppose. Even in Albert's own formulation (p. 181) it is not difficult to discover it. [204/205]
Popper explains it in a comparison of the research process with the process of trial by jury.7 A system of laws, regardless of whether we are dealing with a system of legal norms or empirical-scientific hypotheses, cannot be applied unless agreement has previously been attained concerning the facts of the case to which the system should be applied. Through some kind of decision, the jurors agree which representation of a factual occurrence they intend to approve. This corresponds to accepting a basic statement. The decision is more complicated, however, since the system of laws and the facts of the case are not given independently of one another. On the contrary, the facts of the case are even sought under categories of the system of laws. The comparison of the research process with the process of trial by jury is intended to make us aware of this circle which is inevitable when general rules are applied. 'The analogy between this procedure and that by which we decide basic statements is clear. It throws light, for example, upon their relativity, and the way in which they depend upon questions raised by the theory. In the case of the trial by jury, it would be clearly impossible to apply the "theory" unless there is first a verdict arrived at by decision; yet the verdict has to be found in a procedure that conforms to, and thus applies, part of the general legal code. The case is analogous to that of basic statements. Their acceptance is part of the application of a theoretical system; and it is only this application which makes any further applications of the theoretical system possible.'8
7 Karl R. Popper, The Logic of Scientific
Discovery (
8 Logic, pp. 110ff.
What does this circle, resulting when theories are applied to reality, signify? I think that the area of the empirical is established in advance by means of theoretical assumptions concerning a certain structure, in combination with a certain type of testing conditions. Such things as experimentally established facts, upon which empirical scientific theories could founder, are only constituted in a prior context of the interpretation of possible experience. This context is produced in an interplay of argumentative discourse and experimental action. The combination is organized with a view to controlling predictions. An implicit pre-understanding of the rules of the game guides the discussion of the investigators when they are deciding whether to accept basic statements. For the circle within which they inevitably move [205/206] when they apply theories to what has been observed refers them to a dimension in which rational discussion is only possible with the assistance of hermeneutics.
The demand for controlled observation as the basis for decisions concerning the empirical plausibility of law-like hypotheses presupposes a pre-understanding of definite rules. It is certainly not sufficient to know the specific aim of an investigation and the relevance of an observation for certain assumptions. Instead, the meaning of the research process as a whole must be understood before I can know to what the empirical validity of basic statements is related, just as the judge must always have grasped the meaning of judicature as such. The quaestio facti must be determined with reference to a quaestio juris understood in its immanent claims. In legal proceedings, this question is prominent in everyone's mind. The whole affair revolves around the question of an offence against general prohibitive norms, positively set down and sanctioned by the state. But, what does the quaestio juris look like in the research process, and how is the empirical validity of basic statements measured in this case ? The form of propositional systems and the type of testing conditions which are used to measure their validity, suggest the pragmatist interpretation: namely, that empirical-scientific theories reveal reality under the guiding interest in the possible informative safeguarding and extension of feedback-regulated action.
Popper himself provides clues to this interpretation in his own work. Empirical-scientific theories are significant in that they permit the derivation of universal propositions concerning the covariance of empirical quantities. We develop such law-like hypotheses in anticipation of law-like regularity itself, without being able to justify empirically such anticipation. This methodical anticipation of possible empirical uniformity, however, corresponds to the elementary requirements of behavioural stability. Feedback-regulated actions can only be secured for a long period of time if they are guided by information as to empirical uniformities. In addition, this information must be capable of translation into expectations of behavioural regularity under given conditions. The pragmatist interpretation refers the logical general to general behavioural expectations. Viewed pragmatically, the disjunction between universal propositions, on the one hand, and the, in principle, finite set of observations and the corresponding singular existential statements, on the other hand, can be explained [206/207] by the structure of feedback-regulated action, which always allows itself to be guided by anticipations of behavioural regularity.9
This interpretation, according to which the empirical-analytical sciences allow themselves to be guided by a technical cognitive interest, enjoys the advantage of taking account of Popper's critique of empiricism, without sharing a weakness of his falsification theory. For how is our uncertainty—in principle— about the truth of scientific information to accord with its generally varied and quite permanent technical utilization? Certainly, by the time that knowledge of empirical uniformities is incorporated into technical productive forces and becomes the basis of a scientific civilization, the evidence of everyday experience and of a permanent regulated feedback is overwhelming; logical misgivings are unable to assert themselves against the plebiscite of functioning technical systems, a plebiscite which is renewed daily. However great the weight of Popper's objections to verificationism, his own alternative thus seems less plausible, since it is only an alternative under the positivistic presupposition
9 In this context. Popper's comment that all universal terms can be regarded as dispositional terms, is of interest. (Logic, pp. 94f.; appendix X, pp. 423ff.; and Conjectures, pp. 118f.). On the level of individual universal terms, the problem of universal statements is repeated. For the dispositional concepts implied in such terms can, in their turn, only be explicated by means of assumptions about a law-like behaviour of objects. This is shown in doubtful cases when we imagine possible tests which would be sufficient to elucidate the significance of the universal terms used. In all this, recourse to the testing conditions is hardly fortuitous. For it is only the relation of the theoretical elements to the experiment which closes the functional circle of feedback-regulated action, within which such things as empirical regularities first exist. The hypothetical surplus beyond each specific content of an immediately perceived entity which, in the logical form of law-like statements and in the universal expression of observational terms, comes into its own, does not relate to a regular behaviour of things 'in themselves' but instead to a behaviour of things in so far as this forms a part of the horizon of expectations of actions requiring orientation. Thus, hypothetically, the degree of generality of descriptive content in perceptual judgments far exceeds the particularity of what is perceived in each case because, under the selective pressure towards the stabilization of the results of actions, we always gather experiences and articulate meanings—'for what a thing means is simply what habits it involves' (Peirce).
A further clue for a pragmatist interpretation is given by Popper in connection with a sociology of tradition ('Towards a Rational Theory of Tradition', in: Conjectures, pp. 120ff.). He compares the analogous roles which traditions and theories acquire in social systems. Both inform us about reactions which we can regularly expect, and in accordance with which we can confidently orientate our behaviour. Likewise, they bring order into a chaotic environment in which, without the capacity for prognosticating answers or events, we would not be able to form suitable behavioural habits.
[207/208] of a correspondence between statements and actual states of affairs. The moment we abandon this supposition, and take technique in its widest sense seriously as a socially institutionalized regulatory system which, in accordance with its methodical meaning, is designed to be technically utilizable, one can conceive of another form of verification. The latter is exempt from Popper's objections and concurs, in fact, with our pre-scientific experiences. All the assumptions, then, are empirically true which can guide feedback-regulated action without having been previously rendered problematic through errors experimentally striven for.10
Albert imagines that by referring to Popper's criticism of instrumentalism he is released from any argument of his own against my interpretation, which he does not even reproduce. But I do not really need to answer his criticism since it is directed against theses which I do not expound. In the first instance, Popper concentrates on the thesis that theories are instruments.11 Here he can easily counter that rules of technical application must be tried out, whilst scientific information must be tested. The logical relationships in the case of suitability tests for instruments and the testing of theories are not symmetrical—instruments cannot be refuted. The pragmatic interpretation which I wish to give to empirical-analytical sciences, does not include this form of instrumentalism. It is not the theories themselves which are instruments but rather that their information is technically utilizable. Even from a pragmatic viewpoint the failures, whereby law-like hypotheses founder under experimental conditions, possess the character of refutations. The hypotheses refer to empirical regularities; they determine the horizon of expectation of feedback-regulated action, and consequently can be falsified by disappointed expectations of success. Yet the law-like hypotheses,
10 According to this view. Popper's reservation regarding incontestably valid knowledge can be quite compatible with the pragmatic corroboration of knowledge. Popper admits experimental tests exclusively as an instance of falsification, whilst in the pragmatic view, they are controlled experiments which refute assumptions but can also confirm them. However, corroboration through the results of an action can only be globally allocated, and not in a strictly correlative manner, since, with a given theory, neither in their scope nor in respect of their area of application can we definitely ascertain the factually working elements of knowledge. Definitively, we only know that parts of a theory which is controlled through the results of an action—and that means that it is a prognostically tested theory—are proved correct in the sphere of application of the test situation.
11 'Three Views Concerning Knowledge', in Conjectures, pp. 111ff.
[208/209] in their methodical sense, refer to experiences which are constituted exclusively in the functional sphere of such action. Technical recommendations for a rationalized choice of means under given ends cannot be derived from scientific theories merely at a later stage, and as if by chance. But these theories themselves are not therefore technical implements. This is possibly true in a figurative sense. Technical utilization of knowledge is, of course, in no way intended in the research process: actually, in many cases, it is even excluded. Nevertheless, with the structure of propositions (restricted prognoses concerning observable behaviour), and with the type of testing conditions (imitation of the control of the results of action which is built naturally into systems of societal labour), a methodical decision has been taken in advance concerning the technical utility of empirical scientific information. Similarly, the realm of possible experience is prejudiced, namely, that realm to which hypotheses refer and upon which they can founder.
The descriptive value of scientific information cannot be disputed, but it is not to be understood in such a way that theories represent facts and relations between facts. The descriptive content is only valid with reference to prognoses for feedback- regulated actions in predictable situations. All the answers which the empirical sciences can supply are relative to the methodical significance of the questions they raise and nothing more. No matter how trivial this restriction may be, it contradicts the illusion of pure theory which has been preserved in positivistic self understanding.12
12 Another of Popper's objections is directed against operationalism, according to which basic concepts can be defined through modes of procedure (Conjectures, p. 62; Logic, pp. 440f.). Rightly, Popper asserts that the attempt to trace dispositional concepts back to measurement operations, in its turn presupposes a theory of measurement, for no operation could be described without universal terms. This circle, in which universal terms point to empirically regular behaviour, whilst the regularity of behaviour can only be established through measuring operations, which in turn presuppose general categories, seems to me, however, to require interpretation. The operationalist point of departure rightly insists that the semantic content of empirical scientific information is only valid within a frame of reference which has been transcendentally posited by the structure of feedback-regulated action, and furthermore the semantic content cannot be projected onto reality 'in itself. It is incorrect to assume that such content could be simply reduced to criteria of observable behaviour. The circle in which this attempt is ensnared shows instead that the systems of action, of which the research process forms a part, are mediated through language, but, at the same time, language is not subsumed in categories of behaviour.
[209/210]
3 CRITICAL JUSTIFICATION AND DEDUCTIVE PROOF
The third misunderstanding to which, according to Albert, I
have succumbed, refers to the relationship between methodological and empirical
statements. He finds me guilty of a particularly crass positivism since I do
not forego empirical arguments in methodological contexts and thereby
confound in an unacceptable manner the logic of inquiry with the sociology of
knowledge. After Moore and Husserl, from different standpoints, had effected
the strict division between logical and psychological studies, and, in so doing,
had reinstated the old Kantian insight, even the positivists broke with their
naturalism. Influenced in the meantime by the advances which had been made in
formal logic, Wittgenstein and the
Reflection on what Popper himself does, makes us aware of the peculiar form of meta-theoretical discussions in so far as they advance beyond linguistic analysis. On the one hand. Popper pursues the immanent critique of given theories and, in so doing, employs the systematic comparison of logically compelling deductions. On the other hand, he develops alternative solutions; he makes suggestions of his own and attempts to support them with arguments. In this case, he cannot confine himself to the verification of deductive connections. Rather, his interpretation is aimed at critically altering old attitudes, of making new standards of judgment plausible and new normative points of view acceptable. This takes place in the hermeneutical form of argumentation which evades the rigid monologues of deductive systems of propositions. It sets standards for critical discussion as such. This is revealed in every choice between possible techniques of inquiry, between several theoretical starting points, between various definitions of basic predicates; it is revealed in decisions as to the linguistic framework within which I express a given problem and form its hypothetical solutions. A choice of standards is constantly repeated, as is the attempt to support this choice through suitable arguments. Morton White has shown that, even at the highest level, meta-theoretical discussions remain bound to this form of argumentation. Even the distinction between categorial and non-categorial being, between analytic and synthetic statements, between descriptive and emotive contents, between logical rules and law-like regularities, between controlled observation and moral experience—even these fundamental distinctions, upon which exact empirical science rests, are in no way exempt from discussion. They presuppose criteria which do not result from reality itself, that is, criticizable standards which, in their turn, [211/212] cannot be strictly substantiated by arguments but can be supported or weakened by them.13
13 Morton White, Toward
White makes the attempt which Popper neglects; the attempt to examine the logical relations of this non-deductive form of argumentation. He demonstrates that methodological decisions are quasi-moral decisions and, as a result, can only be justified in discussion of the kind familiar from the old topics and rhetorics. For neither the conventionalistic nor the naturalistic interpretation does justice to the choice of methodological rules.
Critical argumentation differs from deductive argumentation in progressing beyond the dimension of the logical connection of statements and includes a moment which transcends language— attitudes or outlooks. A logical relationship of implication between outlooks and statements is impossible: attitudes cannot be deduced from statements nor, vice versa, statements from attitudes. Nevertheless, agreement upon a mode of procedure and the acceptance of a rule can be supported or weakened by arguments; at any rate, it can be rationally considered and judged. This is the task of critique with reference to both practical and meta-theoretical decisions. Since the supporting or weakening arguments do not stand in a strictly logical relation to the statements which express the application of standards, but instead only in a relation of rational motivation, meta-theoretical discussions can also include empirical propositions. However, the relation between arguments and attitudes does not, in this way, itself become an empirical one. It can be taken as such, as for instance in a Festinger experiment on change of attitude, but then the argumentation would be reduced to the level of observable language behaviour, and the moment of rational validity, which forms part of every motivation, would be suppressed.
Popper does not consider that a rationalization of attitude is out of the question. This form of argumentation is the only possible one for tentatively justifying decisions. Yet since it is never conclusive, he considers it to be unscientific in comparison with logical deduction. He prefers the certainty of descriptive knowledge, a certainty guaranteed by the deductive combination of theories and the empirical constraint of facts. Yet even the interplay of statements and experiences of this particular type presupposes standards which require justification. Popper evades [212/213] this objection by insisting on the irrationality of the decision which precedes the application of his critical method. According to him, the rationalistic attitude consists in the willingness to decide upon the acceptance of theories on the basis of experiences and arguments. It cannot, however, be grounded either through arguments or through experiences. Certainly it cannot be justified in the sense of a deductive proof but it can in the form of a supporting argumentation. Popper himself, in fact, makes use of it at some length. He explains every critical attitude in terms of certain philosophical traditions. He analyses the empirical presuppositions and consequences of scientific criticism. He examines its functions within the given structure of public political life. In fact, his methodology as a whole is a critical justification of criticism itself. It may be that this non-deductive justification is unsatisfactory for a logical absolutism. However, no other form of justification is known to scientific criticism which goes beyond an immanent critique and tests methodological decisions.
Popper terms the critical attitude a belief in reason. Therefore he claims, the problem of rationalism does not consist of the choice between knowledge and faith but rather in a choice between two sorts of faith. But, he adds paradoxically, the new problem is which faith is the right one and which the wrong one.14 He does not totally reject non-deductive justification, but he believes that he can avoid its problematical combination of logical and empirical relations if he foregoes a justification of criticism—as if the 'Black Peter' were not already present in the criticism itself.
Albert saddles me with the onus of proof for the problem of foundation [Begründungsproblem]. He seems to assume that all problems are resolved for him with the abstention of rationalism from the problem of self-foundation [Selbstbegründung]. Apparently, he draws upon William W. Bartley, who attempted to demonstrate conclusively the possibility of such an abstention.15 However, it seems to me that this attempt was not successful.
14 Karl R. Popper, The Open Society and its Enemies (
15 The
Retreat to Commitment (New York, 1962), exp. Chs. 3 and 4; also,
‘Rationality Versus the Theory of Rationality’ in: M. Bunge (ed.)
The Critical Approach to Science and Philosophy (
Bartley starts out from the assumption that, for logical reasons, a deductive self-foundation of rationalism is out of the question. Instead, he discusses the possibility of a critical philosophy [213/214] (Kritizismus) which does in fact accept every proposition which can be rationally grounded, but not exclusively such propositions. He holds no views which cannot be accountable to criticism but neither does he demand that all views, including the critical attitude itself, be rationally grounded. Is this view still tenable, however, when logically the conditions of critical testing are themselves exposed to criticism? Bartley neither questions the standards according to which experience is organized in test situations, nor does he pose sufficiently radically the question of the sphere of validity of deductive justifications. For by means of a stipulation, he exempts from criticism all those standards which we must presuppose in order to criticize. He introduces a so-called revisibility criterion: “namely, whatever is presupposed by the argument-revisibility situation is not itself revisable within that situation”.16 We cannot accept this criterion. It is introduced in order to secure the argumentation, but it would stifle argumentation in the very dimension in which the latter's peculiar achievement is revealed; namely, in the subsequent revision of previously applied standards. Something approaching critical justification consists precisely in the fact that it produces a non-deductive connection between selected standards and empirically secured propositions. Consequently, it also supports or weakens attitudes by means of arguments which, for their part, are first found within the sphere of these attitudes. As soon as it progresses beyond the verification of deductive systems, argumentation takes a reflexive course. It employs standards which it can only reflect upon in their applications. Argumentation differs from mere deduction by always subjecting the principles, according to which it proceeds, to discussion. To this extent, criticism cannot be restricted in advance to conditions which form the framework of possible criticism. What can pass as criticism always has to be determined on the basis of criteria which are only found, elucidated and possibly revised again in the process of criticism. This is the dimension of comprehensive rationality which, although incapable of a final grounding [Letztbegründung], develops in a circle of reflexive self-justification.
16 ibid., p. 173.
Bartley's unconditional rationalism makes too many reservations. He does not recognize criticism as the sole and ultimate horizon within which the validity of theories about reality is [214/215] science and ethics. For, on the one hand, theoretical knowledge determined. As a makeshift, we can conceive of criticism—which cannot be defined because the standards of rationality can only be explained within criticism itself—as a process which, in a domination-free discussion, includes a progressive resolution of disagreement. Such a discussion is guided by the idea of a general and unconstrained consensus amongst those who participate in it. Here, 'agreement' should not reduce the idea of truth to observable behaviour. Rather, the categories, with whose help agreement can be achieved in each case, are themselves dependent upon the process which we interpret as a process for achieving consensus. The idea of agreement does not therefore exclude the distinction between true and false consensus; but this truth cannot be so defined that future revision is ruled out.17 Albert objects that I presuppose as a fact something resembling a rational discussion in methodological contexts (p. 193). I presuppose it as a fact since we always find ourselves in a communication which is intended to lead to agreement [Verständigung]. At the same time, however, this empirical fact possesses a distinctive feature: namely, a transcendental precondition. Only in discussion can agreement on the standards be reached on the basis of which we differentiate facts from mere spectres. The discarded link between formal and empirical statements attempts to do justice to a context in which methodological questions can no longer be meaningfully separated from questions of communication.
17 See D. Pole, Conditions of Rational Inquiry (
4 THE SEPARATION OF STANDARDS AND FACTS
The fourth misunderstanding with which Albert charges me relates to the dualism of facts and decisions. This can be elucidated on the basis of the difference between natural laws and cultural norms. Assumptions about empirical regularities can decisively founder on the facts, whilst the choice of standards can at most be supported critically by additional arguments. One can easily differentiate a realm of scientifically reliable information from that realm of practical knowledge which we only secure through a hermeneutic form of argument. I want to question this optimistic distinction, which is traditionally termed the separation between [215/216] which has been proven against facts is constituted within a normative framework which is capable of a critical but not of a deductive-empirical justification. On the other hand, the critical discussion of standards certainly involves empirical considerations —that is, recourse to so-called facts. A critique which creates a rational connection between attitudes and arguments forms the comprehensive dimension of science itself. Even theoretical knowledge can be no more certain than can critical knowledge. Yet again, the 'misunderstanding' appears to result from Albert having in no way apprehended my intention. I do not deny the distinction between facts and standards. I merely ask whether the positivistic distinction, which permits a dualism of facts and decisions, and correspondingly a dualism of propositions and proposals—that is, a dualism of descriptive and normative knowledge—is an appropriate one.
In the appendix to The Open Society, 18 Popper develops the asymmetrical relation between standards and facts: “. . . through the decision to accept a proposal (at least tentatively) we create the corresponding standard (at least tentatively); yet through the decision to accept a proposition we do not create the corresponding fact”.19 I should like to examine this relation in more detail. We can discuss proposals and statements. Yet the discussion entailed no more produces the standards than it does the facts. Rather, in the first case, it draws upon arguments in order to justify or contest the act of accepting standards. Such arguments can include empirical considerations. But these, for their part, are not under discussion. In the second case, the reverse occurs. Here it is not the choice of standards which is under discussion, but their application to a state of affairs. The discussion draws upon arguments in order to justify or contest the act of incorporating a basic statement with reference to a given hypothesis. These arguments include methodological considerations. Their principles, however, are not under discussion in this case. The critique of an empirical-scientific hypothesis and the critical discussion of the choice of a standard is not symmetrical. Yet this is not because the logical structure of the discussion is different in the two cases—it is the same.
18 4th Edition,
19 ibid., p. 384.
Popper terminates this reflection by reference to the correspondence theory of truth. Ultimately, the dualism of facts and [216/217] standards is based upon the assumption that, independent of our discussions, there exists something resembling facts and relations between facts to which propositions can correspond. Popper denies that the facts themselves are only constituted in combination with the standards of systematic observation or controlled experience. In so far as we intend true propositions, we always know that their truth is measured against a correspondence of propositions and facts. In the following manner. Popper meets the obvious objection that this concept of truth necessarily implies that the criterion or the standard or the definition has been introduced which itself must be exposed to critical discussion: 'It is decisive to realize that knowing what truth means, or under what conditions a statement is called true, is not the same as, and must be clearly distinguished from, possessing a means of deciding—a criterion for deciding—whether a given statement is true or false.'20 We must forego a criterion, a definable standard of truth; we cannot define truth—but, nevertheless, we 'understand' in each individual case what we intend when we test the truth of a proposition: ‘I believe that it is the demand for a criterion of truth which has made so many people feel that the question "what is truth" is unanswerable. But the absence of a criterion of truth does not render the notion of truth non-significant any more than the absence of a criterion of health renders the notion of health non-significant. A sick man may seek health even though he has no criterion for it.’21
20 The Open Society, vol. 2., op. cit., p. 371.
21 ibid., p. 373.
In this passage, Popper makes use of the hermeneutic insight that we understand the meaning of statements from the context even before we can define individual terms and apply a general standard. Anyone familiar with the business of hermeneutics would certainly not conclude that we intend the meaning of such terms and statements without any standard at all. Rather, the pre-understanding which guides interpretation prior to any definition —even Popper's interpretation of truth—always includes standards implicitly. The justification of these prior standards is not really excluded; instead, it is the abstention from a definition which permits a continuous self-correction of the diffuse pre-understanding in the progress of the explication of the texts in hand. The interpretation throws the light of a growing understanding from the text back onto the standards through which it was initially [217/218] made accessible. With the adaptation of the originally employed standards, the hermeneutic course of exegesis also provides their own justification. The standards and the descriptions which they permit when applied to the text still stand in a dialectical relationship. It is just the same with the standard of a truth based on correspondence. It is only the definition of standards and the establishment of criteria which tears apart the standards and the descriptions which make them possible. It is only they which create a deductive connection which excludes a retrospective correction of the standards through the object measured. Only at this point does the critical discussion of standards free itself from their usage. Yet standards are also used implicitly before a critical justification on the meta-theoretical level is differentiated from the object level of applied standards.
For this reason, Popper does not manage to evade the dialectical connection between descriptive, postulatory and critical statements by reference to the correspondence concept of truth. Even this concept of truth, which allows such strict differentiation between standards and facts, is, no matter how implicitly we orientate ourselves by it, still in its turn a standard which requires critical justification. A critical discussion, regardless of whether it concerns the acceptance of proposals or propositions, includes a threefold usage of language: the descriptive, in order to describe states of affairs; the postulatory, in order to establish rules of procedure; and the critical, in order to justify such decisions. Logically, these forms of speech mutually presuppose each other. The descriptive usage is in no way limited to a certain class of 'facts'. The postulatory usage covers the establishment of norms, standards, criteria and definitions of all types, no matter whether practical, logical or methodological rules are involved. The critical usage employs arguments for considering, evaluating, judging and justifying the choice of standards; it includes, therefore, language-transcendent outlooks and attitudes in the discussion. No proposition concerning reality is capable of a rational test without the explication of a connection between arguments and attitudes. Descriptions are not independent of standards which are used, and standards rest upon attitudes which are in need of justification through supporting arguments, but which are, at the same time, incapable of deduction from assertions. If attitudes are altered under the influence of arguments, then such a motivation apparently combines a logically [218/219] incomplete constraint with an empirical one. The only constraint of this sort originates in the power of reflection, which breaks the power of the unpenetrated by rendering it conscious. Emancipatory insight translates logical constraint into empirical constraint. This is achieved by critique; the latter overcomes the dualism of facts and standards, and in this way, first produces the continuum of a rational discussion which otherwise would degenerate immediately into arbitrary decisions and deductions.
As soon as we discuss a problem at all with the aim of reaching a consensus rationally and without constraint, we find ourselves in a dimension of comprehensive rationality which embraces as its moments language and action, statements and attitudes. Critique is always the transition from one moment to another. It is, if I may put it like this, an empirical fact in a transcendental role of which we become aware in the execution of criticism. It can also, of course, be repressed and disguised from that moment on in which, with the definition of the initially implicitly applied standards, a language-immanent realm of logical relations is freed from living reflection. This repression is expressed in Popper's critique of Hegel: “To transcend this dualism of facts and standards is the decisive aim of Hegel's philosophy of identity—the identity of the ideal and the real, of right and might. All standards are historical: they are historical facts, stages in the development of reason, which is the same as the development of the ideal and of the real. There is nothing but fact; and some of the social or historical facts are, at the same time, standards.”22 Nothing was further removed from Hegel's mind than this metaphysical positivism which Popper counters with the insight of logical positivism that statements and states of affairs belong to different spheres. Hegel in no way reduced the logical and empirical, the criteria of validation and factual relations, the normative and the descriptive to the level of historical facts. However, he did not neglect the experience of critical consciousness, namely, that reflection also holds together the well separated moments. The critique moves from the argument to the attitude, and from the attitude to the argument, and acquires, in this movement, the comprehensive rationality which, in the natural hermeneutics of everyday language, is still, as it were, naturally at work. In the sciences, however, this rationality must be re-established between
22 The Open Society, vol. 2, op. cit., pp. 394-5.
[219/220] the now-separated moments of formalized language, and objectivized experience by means of critical discussion. Only because this criticism relates chosen standards non-deductively to empirical states of affairs, and can measure one moment against another, is the statement correct—which according to Popper's own presuppositions would be untenable, namely: '. . . that we can learn; by our mistakes and by criticism; and that we can learn in the realm of standards just as well as in the realm of facts.'23
5 TWO STRATEGIES AND A DISCUSSION
Albert seizes upon a series of questions, polemicizes and lets them drop again. I cannot discover any principle behind this sequence. I have attempted to clear up four fundamental misunderstandings in order to create a basis of agreement upon which further problems—for instance, the role of historical reflection, the postulate of value freedom, or the position of the critique of ideology—could be discussed without linguistic confusion. Now, I believe, my intention will no longer be open to misunderstanding. In opposition to positivism, I should like to justify the view that the research process, which is carried out by human subjects, belongs to the objective context which itself constitutes the object of cognition, by virtue of cognitive acts.
The dimension in which this combination of the research process with the social life-process is formed belongs neither to the sphere of facts nor to that of theories. It stands apart from this dualism, which only has meaning for empirical scientific theories. Rather, in the comprehensive communicative context of scientific criticism one moment links itself to another. I would say, in old-fashioned language, that the transcendental preconditions of possible knowledge are here created under empirical conditions. As a result, neither the sociology of knowledge nor a pure methodology are sufficiently appropriate at this level of reflection. Their combination, which used to be called the critique of ideology, is more appropriate. I do not like to use this expression, for I do not wish the present discussion to cover all randomly situated interests. I am concerned with knowledge-guiding interests which, in each case, form the basis for a whole research
23 ibid., p. 386.
[220/221] system. In contrast to positivistic self-understanding, I should like to point out the connection of the empirical-analytical sciences with a technical cognitive interest. But this has nothing to do with 'denunciation', as Albert insinuates. It has quite escaped Albert's notice that it is far from my intention to criticize empirical-analytical research itself. He imagines that I wished to play off the methods of understanding against those of explanation. On the contrary, I regard as abortive, even reactionary, the attempts which characterized the old methodological dispute, namely, attempts to set up barriers from the outset in order to protect whole sectors from the clutches of a certain type of research. It would be a bad dialectician who immunized himself in this way.
Naturally, reflection upon cognitive interests is not without consequences. It makes us aware of attitudes upon which fundamental decisions concerning the methodological framework of whole research systems are dependent. Only in this way, do we learn to know what we are doing; only in this way, do we know what, when we do something, we can learn. We become aware, for instance, of the fact that empirical-analytical research produces technically utilizable knowledge, but not knowledge which makes possible a hermeneutical elucidation of the self-understanding of acting subjects. So far, sociology has primarily—and by no means in an unproblematic manner—assisted the self-reflection of social groups in given historical situations. It cannot escape this today, not even where it has professed its intention to provide mere information on empirical regularities of social behaviour. I agree with Albert that in our discipline we ought to devote all efforts to acquiring more and better information of this kind. I do not agree with him that we could, should or even must restrict ourselves to this. I shall not examine here the reason why in this country sociology has taken over the role of a historically orientated theory of society, whilst other social sciences were free from this burden and have therefore made faster progress within the limits of an exact empirical science. But what would it be like if a successful, positivistic, scientific strategy were able to reject this task completely and banish it to the vestibules of scientific discussion? For in the hands of the positivists, the critique of ideology serves this purpose. It concerns itself with cleansing the practical consciousness of social groups of those theories which cannot be reduced to technically utilizable knowledge, and yet [221/222] defends their theoretical claims. How would it be then if this purge were feasible and were successfully carried out?
Under the conditions of reproduction of an industrial society, individuals who only possessed technically utilizable knowledge, and who were no longer in a position to expect a rational enlightenment of themselves nor of the aims behind their action, would lose their identity. Since the power of myth cannot be broken positivistically, their demythologized world would be full of demons. I fully accept the risk of this language. It belongs to a sphere of experience which is in no way reserved for a clairvoyant elite. However, I do have to admit that the power of imagination is only formed in contact with traditions which one initially acquires and does not immediately immerse oneself in. The possibility of rational agreement even in this dimension can be verified by reading a recently published book by Klaus Heinrich.24
A sociology which restricted itself in its critical intention to empirical-analytical research would only be in a position to examine the self-preservation and self-destruction of social systems in the sphere of pragmatically successful adjustment processes, and would have to deny other dimensions. Within sociology as a strict behavioural science, questions relating to the self-understanding of social groups cannot be formulated. Yet they are not meaningless on that count, nor are they beyond binding discussion. They arise objectively from the fact that the reproduction of social life not only poses technically soluble questions; instead, it includes more than the processes of adaptation along the lines of the purposive-rational use of means. Socialized individuals are only sustained through group identity, which contrasts with animal societies which must be constantly built up, destroyed and formed anew. They can only secure their existence through processes of adaptation to their natural environment, and through re-adaptation to the system of social labour in so far as they mediate their metabolism with nature by means of an extremely precarious equilibrium of individuals amongst themselves. The material conditions for survival are most closely bound up with the most sublime conditions; organic equilibrium is bound up with the distorted balance between separation and unification. Only in this balance, through communication with others, is the identity of each ego established.
24 Versuch uber die Schwierigkeit,
Nein zu sagen (
[222/223]
A failing identity for people attempting to assert themselves, and an unsuccessful communication between those talking to one another, are both self-destructive—this ultimately has physical effects. In the individual sphere, these are familiar in the form of psychosomatic disturbances; the dismembered life-histories reflect the dismembered reality of institutions. We are acquainted with the painful processes of constantly renewing our identity from Hegel's Phenomenology of Mind as well as from Freud's psychoanalysis. The problem of an identity which can only be produced through identifications, and that means solely through externalizing identity, is at the same time the problem of communication, which makes possible the happy balance between silent, isolated existence and silent estrangement, between the sacrificing of individuality and the isolation of abstract individual persons. Everyone repeats such experiences of impending loss of identity and the silting up of verbal communication in the crises of his life-history. Yet they are no more real than the collective experiences in the history of the species, which the total societal subjects have made for themselves in their confrontation with nature. Questions concerning this realm of experience, because they cannot be answered by technically utilizable information, are not capable of explanation by empirical-analytical research. Nevertheless, since its beginnings in the eighteenth century, sociology tried to discuss these very questions. In so doing, it cannot do without historically-orientated interpretations; nor, apparently, can it evade a form of communication under the spell of which alone these problems pose themselves. I refer to the dialectical network of a communicative context in which individuals develop their fragile identity between the dangers of reification and formlessness. This is the empirical core of the logical form of identity. In the evolution of consciousness, the problem of identity presents itself as a problem of survival and; at the same time, of reflection. From here dialectical philosophy once developed.
In the shirt-sleeved world picture of many a positivist, dialectics plays the part of a bogeyman. For others, who occasionally become aware of the fact that they lapse into dialectical trains of thought, dialectics only expresses the fact that we think and are able to think when, according to the traditional rules of logical inference, we really should not be able to do so. In dialectics, thought does not become entangled because it scorns the rules of [223/224] formal logic but because it clings to them in a particularly stubborn manner—even on the level of self-reflection, rather than breaking off reflection at this point. The self-reflection of the strict empirical sciences, in my opinion, strikes a cautionary note as far as positivistic expectations are concerned. It includes the realization that our theories do not simply describe reality. On the other hand, it does not permit itself to be discouraged by definitions from explaining such connections which, according to the demarcations upon which—for good reason—empirical scientific analysis is based, should not exist.
Given these points of departure, a discussion between positivists and those who are not ashamed of dialectical trains of thought has its moments of treachery. Nevertheless, since both parties are convinced of the unity of human reason, as well as of the possibility of a consensus achieved in a rational manner and, in addition, do not intentionally deny the comprehensive rationality of an unreserved criticism, it is possible for them to carry on a discussion. In so doing, however, both parties pursue a different strategy.
Albert accuses me of a quite unscientific strategy. He calls it immunization and masking. If one considers that I subject to discussion the very conditions of validation themselves—upon whose exclusiveness Albert insists—then neither description seems particularly meaningful to me. I should prefer to talk of a flanking strategy. You have to make it clear to the positivist that you have already taken up a position behind his back. I have no idea whether this is a sympathetic manner in which to proceed but, at any rate, it was dictated to me by the course of the discussion. Albert's objections rest on presuppositions which, in their turn, I have questioned. Albert's strategy,25 on the other hand, I could characterize, with a certain symmetry to his accusation of obscurantism, as one of pretending to be stupid. One refuses to understand what the other person is saying. This strategy, intended to force the opponent to accept one's own language, is several centuries old and has been extraordinarily successful since the days of Bacon. The advances of the exact sciences rest, to a large extent, upon translating traditional questions into a new language. They find no answer to questions which they themselves have not
25 I do not wish to include here Albert's slip which creeps in on pp. 188f. I do not imagine that Albert makes a commonplace anti-communism a part of his strategy.
[224/225] formulated. Yet, on the other hand, this very same strategy becomes a formidable hindrance if one wishes to discuss the status of such research as a whole. The systematic pretence of inability to understand dries up a discussion since any discussion must always move within the compass of a pre-understanding which is mutually taken for granted. In this way, one promotes an ethnocentricity of scientific subcultures which destroys the candour of scientific criticism.
The accusation of unintelligibility belongs to this context. In so far as this touches me as an empirical subject, I take it repentantly to heart. But in so far as it is aimed at a structure of thought and expression, it requires explanation. Understanding is a two-sided relationship. Whilst carrying out my required reading of ingenious positivistic studies, I have had the painful experience of not, or not immediately, understanding a great deal. I attributed the difficulty to my defective learning processes and not to the unintelligibility of the texts. I would not venture to exclude altogether the impression that the same thing could happen the other way round with someone who quotes Hegel at second hand.
I speak here of tradition with regard to learning processes which it makes possible and not in the anticipation of authorities to which a descent could be traced back. Perhaps it is precisely for this reason that Popper's works belong to the series of great philosophical theories, because he still maintains learned acquaintance with traditions which many a member of his retinue hardly knows even by name.
SOURCE: Habermas, Jürgen. "A Positivistically Bisected Rationalism," in The Positivist Dispute in German Sociology, translated by Glyn Adey and David Frisby (London: Heinemann, 1976), pp. 198-225. Boldface added by RD.
The Analytical Theory of Science and Dialectics by Jürgen Habermas
The Myth of Total Reason by Hans Albert
On Habermas’s Theory and Practice by R. Dumain
“Historical
Remarks on the Question of Organization”
by Jürgen Habermas
“The
partisanship of the critique of ideology in favor of technological rationality”
(Excerpt)
by Jürgen Habermas
Jürgen Habermas on centralized automation of social control
On
the self-reflection of rationalistic “faith”
by Jürgen Habermas
Addendum on Heine, Habermas, et al by R. Dumain
Commentary
on Matthew Piscioneri, Habermas: The Myth of Reason
by R. Dumain
Metacritique Critiqued by R. Dumain
“Komunika
Etiko kaj Esperantismo”
[Communication Ethics & Esperantism]
de Helmut Welger (in Esperanto)
Positivism vs Life Philosophy (Lebensphilosophie) Study Guide
Reflexivity & Situatedness Study Guide
The Philosophy of Theory and Practice: Selected Bibliography
Home Page | Site
Map | What's New | Coming Attractions | Book
News
Bibliography | Mini-Bibliographies | Study
Guides | Special Sections
My Writings | Other Authors' Texts | Philosophical
Quotations
Blogs | Images
& Sounds | External Links
CONTACT Ralph Dumain
Uploaded 18 February 2009
Site ©1999-2021 Ralph Dumain