by Theodor W. Adorno, 1957
translated by Glyn Adey and David Frisby, 1976
The modes of procedure assembled under the name of sociology as an academic discipline are united in an extremely abstract sense, namely, in that all of them in some way deal with society. But neither their object nor their method is uniform. Some apply to societal totality and its laws of movement, others, in pointed opposition, apply to individual social phenomena which one relates to a concept of society at the cost of ostracization for being speculative. Accordingly, the methods vary. In the former case, insight into the societal context is supposed to follow from structural basic conditions, such as the exchange relationship. In the latter, such an endeavour, even though it may in no way desire to justify the factual from the standpoint of an autocratic mind, is dismissed as philosophical residue in the development of science, and is to give way to the mere establishment of what is the case. Historically divergent models underlie both conceptions. The theory of society originated in philosophy whilst, at the same time, it attempts to reformulate the questions posed by the latter by defining society as the substratum which traditional philosophy called eternal essences or spirit. Just as philosophy mistrusted the deceit of appearances and sought after interpretation, so the more smoothly the façade of society presents itself, the more profoundly does theory mistrust it. Theory seeks to give a name to what secretly holds the machinery together. The ardent desire for thought, to which the senselessness of what merely exists was once unbearable, has become secularized in the desire for disenchantment. It seeks to raise the stone under which the monster lies brooding. In such knowledge alone meaning has been preserved for us. Sociological research into facts opposes such a desire. Disenchantment of the kind that Max Weber accepted, is merely a special case of sorcery for such research, and reflection upon that which governs secretly and would have to be changed, is viewed as a mere waste of time on the way towards the alteration of the manifest. This is especially the case since what nowadays generally bears the name empirical social science has taken, more or less avowedly since Comte's positivism, the natural sciences as its model. The two tendencies refuse to be reduced to a common denominator. Theoretical reflections upon society as a whole cannot be completely realized by empirical findings; they seek to evade the latter just as spirits evade para‑psychological experimental arrangements. Each particular view of society as a whole necessarily transcends its scattered facts. The first condition for construction of the totality is a concept of the object [Sache], around which the disparate data are organized. From the living experience, and not from one already established according to the societally installed control mechanisms, from the memory of what has been conceived in the past, from the unswerving consequence of one’s own reflection, this construction must always bring the concept to bear on the material and reshape it in contact with the latter. But if theory is not to fall prey to the dogmatism over whose discovery scepticism—now elevated to a prohibition on thought—is always ready to rejoice, then theory may not rest here. It must transform the concepts which it brings, as it were, from outside into those which the object has of itself, into what the object, left to itself, seeks to be, and confront it with what it is. It must dissolve the rigidity of the temporally and spatially fixed object into a field of tension of the possible and the real: each one, in order to exist, is dependent upon the other. In other words, theory is indisputably critical. But, for this reason, hypotheses derived from it—forecasts of what can be regularly expected—are not completely sufficient for it. What can merely be expected is itself a piece of societal activity, and is incommensurable with the goal of criticism. The cheap satisfaction that things actually come about in the manner which the theory of society had suspected, ought not to delude the theory, that, as soon as it appears as a hypothesis, it alters its inner composition. The isolated observation through which it is verified belongs, in turn, to the context of delusion which it desires to penetrate. The concretization and certainty gained must be paid for with a loss in penetrating force; as far as the principle is concerned it will be reduced to the phenomenon against which it is tested. But if, conversely, one wishes to proceed in accordance with general scientific custom from individual investigations to the totality of society then one gains, at best, classificatory higher concepts, but not those which express the life of society itself. The category 'a society based on the division of labour in general' is higher and more general than 'capitalistic society'—but it is not more substantial. Rather, it is less substantial and tells us less about the life of the people and what threatens them. This does not mean, however, that a logically lower category such as 'urbanism' would say more. Neither upwards nor downwards do sociological levels of abstraction correspond simply to the societal knowledge value. For this reason, one can expect so little from their systematic standardization by means of a model such as Parsons' 'functional' model. But still less can be expected from the promises repeatedly made, and postponed since sociological prehistory, of a synthesis of the theoretical and the empirical, which falsely equate theory with formal unity and refuse to admit that a theory of society, purged of the substantive contents, displaces all its emphases. It should be remembered how indifferent recourse to the 'group' is as opposed to recourse to industrial society. Societal theory formation, based on the model of classificatory systems, substitutes the thinnest conceptual residue for what gives society its law. The empirical and the theoretical cannot be registered on a continuum. Compared with the presumption of insight into the essence of modern society, empirical contributions are like drops in the ocean. But according to the empirical rules of the game, empirical proofs for central structural laws remain, in any case, contestable. It is not a matter of smoothing out such divergences and harmonizing them. Only a harmonistic view of society could induce one to such an attempt. Instead, the tensions must be brought to a head in a fruitful manner.
Nowadays, in the train of disappointment with both cultural-scientific [Geisteswissenschaftlich] and formal sociology, there is a predominant tendency to give primacy to empirical sociology. Its immediate practical utilizability, and its affinity to every type of administration, undoubtedly play a role here. But the reaction against either arbitrary or empty assertions made about society from above is legitimate. Nevertheless, empirical procedures do not merit simple priority. It is not merely the case that there exist other procedures besides these. Disciplines and modes of thought are not justified by their mere existence but rather their limit is prescribed for them by the object [Sache]. Paradoxically, the empirical methods, whose power of attraction lies in their claim to objectivity, favour the subjective—and this is explained by their origins in market research. At most, this preference abstracts from statistical data of the census type—such as sex, age, marital status, income, education and the like, and also opinions and attitudes—the behavioural modes of human subjects. So far, at any rate, only within this compass has what is specific to them asserted itself. As inventories of so‑called objective states of affairs they could only be distinguished with some difficulty from pre‑scientific information for administration purposes. In general, the objectivity of empirical social research is an objectivity of the methods, not of what is investigated. From surveys of varying numbers of individuals, statements are derived by means of statistical processing which are generalizable and independent of individual fluctuations in accordance with the laws of the theory of probability. But even if their validity be objective, in most cases the mean values remain objective statements about human subjects, and, in fact, they remain statements about how human subjects see themselves and reality. The empirical methods—questionnaire, interview and whatever combination and supplementation of these is possible—have ignored societal objectivity, the embodiment of all the conditions, institutions and forces within which human beings act, or at most, they have taken them into account as accidentals. At fault here are not only those interested in commissioning research who consciously or unconsciously prevent the elucidation of such conditions and who in America are careful to make sure—even when distributing research projects on mass communications for instance—that only reactions within the dominant 'commercial system' are recorded and that the structure and implications of the system itself are not analysed. Moreover, even the empirical means are objectively fashioned to this end. This involves the largely pre‑ranked questioning of many individuals and its statistical evaluation which, in advance, tend to recognize widely‑held—and, as such, preformed—views as justification for judgment on the object itself. In these views, objectivities may also be reflected but certainly not entirely, and often in a distorted form. In any case, as the most cursory glance at the manner in which working people function in their jobs will demonstrate, the weight of subjective opinions, attitudes and modes of behaviour is secondary compared with such objectivities. No matter how positivistic the modes of procedure, they are implicitly based upon the notion—derived from the ground rules of democratic elections and all‑too unhesitatingly generalized—that the embodiment of the contents of man's consciousness or unconsciousness which form a statistical universe possesses an immediate key role for the societal process. Despite their objectification, in fact on account of it, the methods do not penetrate the objectification of the object, or in particular, the constraint of economic objectivity. For them, all opinions possess virtually the same validity, and they capture such elementary differences as that of the weight of opinions in proportion to societal power purely through additional refinements such as the selection of key groups. The primary becomes the secondary. Such shifts within the method are not, however, indifferent to what is investigated. For all the aversion of empirical sociology to the philosophical anthropologies which became fashionable in the same period, it shares with them a standpoint; namely, the belief that already in the here and now it is man as such who is central, instead of determining socialized human beings in advance as a moment of societal totality—in fact, predominantly as the object of the latter. The reified nature [Dinghaftigkeit] of the method, its inherent tendency to nail down the facts of the case, is transferred to its objects, that, to the subjective facts which have been ascertained, as if they were things in themselves and not hypostatized entities. The method is likely both to fetishize its object and, in turn, to degenerate into a fetish. Not for nothing—and quite rightly as far as the logic of scientific procedures under discussion is concerned—in discussions of empirical social research do questions of method outweigh substantive questions. As a criterion, the dignity of the objects to be examined is frequently replaced by the objectivity of the findings which are to be ascertained by means of a method. In the empirical scientific process, the selection of the research objects and the starting point of the investigation are guided, if not by practical administrative considerations and not so much by the essential nature of what is investigated, but rather by the available methods which, at most, must be developed further. This explains the undoubted irrelevance of so many empirical studies. The procedure of operational or instrumental definition generally current in empirical techniques—which will define a category such as 'conservatism' by means of certain numerical values of the answers to questions within the investigation itself—sanctions the primacy of the method over the object, and ultimately sanctions the arbitrariness of the scientific enterprise itself. The pretence is made to examine an object by means of an instrument of research, which through its own formulation, decides what the object is; in other words, we are faced with a simple circle. The gesture of scientific honesty, which refuses to work with concepts that are not clear and unambiguous, becomes the excuse for superimposing the self-satisfied research enterprise over what is investigated. With the arrogance of the uninstructed, the objections of the great philosophical tradition to the practice of definition are forgotten. What this tradition rejected as scholastic residue is dragged along in an unreflected manner by individual disciplines in the name of scientific exactitude. But as soon as there is any extrapolation from the instrumentally defined concepts even to the conventionally common concepts—and this is almost inevitable—research is guilty of the impurity which it intended to eradicate with its definitions.
It is in the nature of society itself that the natural scientific model cannot be happily and unreservedly transferred to it. But although the ideology suggests otherwise, and this is rationalized by the reactionary opposition to new techniques in Germany, this is not because the dignity of man, for the gradual abolition of which mankind is avidly working, would be excluded from methods which regard him as a part of nature. Instead, it is more true to say that mankind commits a flagrant sin in so far as man's claim to domination represses the remembrance of his natural being and thus perpetuates blind natural spontaneity (Naturwüchsigkeit) than when human beings are reminded of their natural instincts (Naturhaftigkeit). ‘Sociology is not a cultural science (Geisteswissenschaft).’  Insofar as the obduracy of society continually reduces human beings to objects and transforms their condition into 'second nature', methods which find it guilty of doing just this are not sacrilegious. The lack of freedom in the methods serves freedom by attesting wordlessly to the predominant lack of freedom. The enraged, indignant protests and the subtler defensive gestures provoked by Kinsey's investigations are the most powerful argument for Kinsey. Wherever human beings are, in fact, reduced under the pressure of conditions to the 'amphibious' mode of reaction,  as they are in their capacity as compulsive consumers of the mass media and other regimented joys, opinion research, which infuriates lixiviated humanism, is better suited to them than, for instance, an 'interpretative' sociology. For, the substratum of understanding, namely human behaviour, which is in itself unified and meaningful, has already been replaced in the human subjects themselves by mere reaction. A social science which is both atomistic, and ascends through classification from the atoms to generalities, is the Medusan mirror to a society which is both atomized and organized according to abstract classificatory concepts, namely those of administration. But in order to become true, this adaequatio rei atque cogitationis requires self‑reflection. Its legitimation is solely critical. In that moment in which one hypostatizes that state which research methods both grasp and express as the immanent reason of science, instead of making it the object of one's thought, one contributes intentionally or otherwise to its perpetuation. Then, empirical social research wrongly takes the epiphenomenon—what the world has made of us—for the object itself. In its application, there exists a presupposition which should not be deduced from the demands of the method but rather the state of society, that is, historically. The hypostatized method postulates the reified consciousness of the people tested. If a questionnaire inquires into musical taste and, in so doing, offers a choice between the categories 'classical' and 'popular', then it rightly believes that it has ascertained that the audience in question listens in accordance with these categories. Similarly, one automatically recognizes, without reflection, when one turns on the radio, whether one has found a popular music programme, or what is considered serious music, or the background music to a religious act. But as long as the societal conditions for such forms of reaction are not met, the correct finding is also misleading. It suggests that the division of musical experience into ‘classical’ and 'popular' is final and even natural. But the societally relevant question only arises with this division, with its perpetuation as something self‑evident, and necessarily implies the question whether the perception of music under the a priori sectors most acutely affects the spontaneous experience of the perceived. Only the insight into the genesis of the existing forms of reaction and their relationship to the meaning of that experienced would permit one to decipher the phenomenon registered. The predominant empiricist habit, however, would reject any discussion of the objective meaning of the particular work of art, and would discuss such meaning as a mere subjective projection by the listeners and relegate the structure to the mere 'stimulus' of a psychological experimental arrangement. In this manner, it would, from the outset, exclude the possibility of discussing the relationship between the masses and the products forced upon them by the culture industry. Ultimately, the products themselves would be defined through the reactions of the masses whose relation to the products was under discussion. But it is all the more urgent today to proceed beyond the isolated study since, with the hold of the media on the population growing stronger, the pre‑formation of their consciousness also increases so that there is scarcely a gap left which might permit an awareness of this very pre‑formation. Even such a positivistic sociologist as Durkheim, who in his rejection of Verstehen was in agreement with social research, had good reason for associating the statistical laws, to which he also adhered, with the 'contrainte sociale'  and even for recognizing in the latter the criterion of society's general law‑like nature. Contemporary social research denies this connection and thereby also sacrifices the connection between its generalizations and concrete, societal determinations of structure. But if such perspectives are pushed aside and considered to be the task of special investigations which must be carried out at some point, then scientific mirroring indeed remains a mere duplication, the reified apperception of the hypostatized, thereby distorting the object through duplication itself. It enchants that which is mediated into something immediate. As a corrective, it is not then sufficient simply to distinguish descriptively between the 'collective realm' and the 'individual realm', as Durkheim intended, but rather the relationship between the two realms must be mediated and must itself be grounded theoretically. The opposition between quantitative and qualitative analysis is not absolute. It is not the last word in the matter. It is well known that whoever quantifies must always first abstract from qualitative differences in the elements, and everything that is societally individual contains the general determinations for which the quantitative generalizations are valid. The proper categories of the latter are always qualitative. A method which does not do justice to this fact and rejects qualitative analysis as incompatible with the essence of the collective realm distorts what it should investigate. Society is one. Even where the major societal forces have not yet made their influence felt, the 'undeveloped' spheres are functionally inter‑related with those spheres which have advanced towards rationality and uniform socialization (Vergesellschaftung). Sociology, which disregards this and remains content with such weak and inadequate concepts as induction and deduction,  supports what exists in the over‑zealous attempt to say what exists. Such sociology becomes ideology in the strict sense—a necessary illusion. It is illusion since the diversity of methods does not encompass the unity of the object and conceals it behind so‑called factors into which the object is broken up for the sake of convenience; it is necessary since the object, society, fears nothing more than to be called by name, and therefore it automatically encourages and tolerates only such knowledge of itself that slides off its back without any impact. The conceptual dichotomy of induction and deduction is the scientistic substitute for dialectics. But just as a binding theory of society must have fully immersed itself in its material, so the fact to be processed must itself throw light on the societal totality by virtue of the process which apprehends it. If, however, the method has already rendered it a factum brutum, then no light can subsequently penetrate it. In the rigid opposition and complementation of formal sociology and the blind establishment of facts, the relationship between the general and the particular disappears. But society draws its life from this relationship, which therefore provides sociology with its only humanly worthy object. If one subsequently adds together what has been separated, then the material relationship is stood upon its head by the gradation of the method. The eagerness to quantify immediately even the qualitative findings is not fortuitous. Science wishes to rid the world of the tension between the general and the particular by means of its consistent system, but the world gains its unity from inconsistency.
This inconsistency is the reason why the object of sociology—society and its phenomena—does not posses the type of homogeneity which so‑called classical natural science was able to count upon. In sociology one cannot progress to the same degree from partial assertions about societal states of affairs to their general, even if restricted, validity, as one was accustomed to infer the characteristics of lead in general from the observation of the characteristics of one piece of lead. The generality of social-scientific laws is not at all that of a conceptual sphere into which the individual parts can be wholly incorporated, but rather always and essentially relates to the relationship of the general to the particular in its historical concretion. In negative terms, this attests to the lack of homogeneity of the state of society—the 'anarchy' of all history up till now—whilst, in positive terms, it attests to the moment of spontaneity which cannot be apprehended by the law of large numbers. Anyone who contrasts the human world with the relative regularity and constancy of the objects in the mathematical natural sciences, or at least in the 'macro‑realm', does not transfigure this world. The antagonistic character of society is central and this is conjured away by mere generalization. Homogeneity, rather than its absence, requires clarification insofar as it subjects human behaviour to the law of large numbers. The applicability of this law contradicts the principium individuationis namely that, despite everything, it cannot be overlooked that human beings are not merely members of a species. Their modes of behaviour are mediated through their intellect. The latter certainly contains a moment of the general which can very easily recur in the statistical generality. Yet it is also specified by means of the interests of particular individuals which diverge in bourgeois society and, even given uniformity, tend to be opposed to one another, not to mention the irrationality in individuals, reproduced under the societal constraints. It is only the unity of the principle of an individualistic society which unites the dispersed interests of the individuals in the formula of their 'opinion'. The currently widespread talk about the social atom certainly does justice to the powerlessness of the individual confronted with the totality, yet it remains merely metaphorical when compared with the natural scientific concept of the atom. Even in front of the television screen, the similarity of the smallest social units, that is the similarity of individuals, cannot be seriously asserted with the strictness possible in the case of physical‑chemical matter. Yet empirical social research proceeds as if it took the idea of the social atom at its face value. That it is to some extent successful, is a critical reflection upon society. The general law‑like nature of society, which disqualifies statistical elements, testifies that the general and the particular are not reconciled, that precisely in individualistic society the individual is blindly subjected to the general and is himself disqualified. Talk about society's 'character mask' once recorded this state of affairs, but contemporary empiricism has forgotten it. The communal social reaction is essentially that of social pressure. It is only on this account that empirical research, with its conception of the collective realm, is able to brush individuation aside in such a high‑handed manner, since the latter has remained ideological up to the present, and since human beings are not yet human beings. In a liberated society, statistics would become, in a positive manner, what today it can only be in negative terms: an administrative science, but really a science for the administration of objects—namely, consumer‑goods—and not of people. Yet despite its awkward basis in the social structure, empirical social research should retain its capacity for self‑criticism to the extent that the generalizations which it achieves should not immediately be attributed to reality, to the standardized world, but instead they should always be attributed to the method as well. For even through the generality of the question put to individuals or their restricted selection—the cafeteria—the method prepares in advance what is to be ascertained—the opinions to be investigated in such a manner that it becomes an atom.
Insight into the heterogeneity of sociology as a scientific construct, that is, insight into the heterogeneity of the categorial, and not merely graded and easily bridgeable, divergence between disciplines such as social theory, the analysis of objective social conditions and institutions, and subjectively orientated social research in the narrower sense, does not imply that one should simply accept the sterile division between the disciplines. The formal demand for the unity of a science is certainly not to be respected when the science itself bears the marks of an arbitrary division of labour and cannot set itself up as if it could discern without difficulty the much‑favoured totalities, whose social existence is, in any case, questionable. But the critical amalgamation of divergent sociological methods is required for concrete reasons, for the cognitive goal. In view of the specific nexus of social theory formation and specific social interests, a corrective of the type offered by the research methods is salutary no matter how entangled with particular interests the latter may be by virtue of their 'administrative' structure. Numerous stalwart assertions of social theories—and here we shall only mention for the purpose of illustration, Max Scheler's assertion about the typical lower‑class forms of consciousness —can be tested and refuted with the aid of strict investigations. On the other hand, social research is dependent upon confrontation with theory and with knowledge of objective social structures, otherwise it would degenerate into irrelevancy or willingly comply with apologetic slogans such as those of the family, which occasionally gain popularity. Isolated social research becomes untrue as soon as it wishes to extirpate totality as a mere crypto‑metaphysical prejudice, since totality cannot, in principle, be apprehended by its methods. Science then pledges itself to the mere phenomena. If one taboos the question of being as an illusion, as something which cannot be realized with the aid of the method, then the essential connections—what actually matters in society—are protected a priori from knowledge. It is futile to ask whether these essential connections are 'real', or merely conceptual structures. The person who attributes the conceptual to social reality need not fear the accusation of being idealistic. What is implied here is not merely the constitutive conceptuality of the knowing subject but also a conceptuality which holds sway in reality (Sache) itself. Even in the theory of the conceptual mediation of all being, Hegel envisaged something decisive in real terms. The law which determines how the fatality of mankind unfolds itself is the law of exchange. Yet, in turn, this does not represent a simple immediacy 'but is conceptual. The act of exchange implies the reduction of the products to be exchanged to their equivalents, to something abstract, but by no means—as traditional discussion would maintain—to something material. This mediating conceptuality is, however, not a general formulation of average expectations, nor is it an abbreviating addition on the part of a science which creates order. Instead, society obeys this conceptuality tel quel, and it provides the objectively valid model for all essential social events. This conceptuality is independent both of the consciousness of the human beings subjected to it and of the consciousness of the scientists. Confronted with physical reality and all the hard data, one might call this conceptual entity illusion, since the exchange of equivalents proceeds both justly and unjustly. It is not an illusion to which organizing science sublimates reality but rather it is immanent to reality. Moreover, talk about the unreality of social laws is only justified critically, namely with regard to the commodity's fetish character. Exchange value, merely a mental configuration when compared with use value, dominates human needs and replaces them; illusion dominates reality. To this extent, society is myth and its elucidation is still as necessary as ever. At the same time, however, this illusion is what is most real, it is the formula used to bewitch the world. The critique of this illusion has nothing to do with the positivistic scientific critique according to which one cannot regard the objective nature of exchange as valid. This validity is unremittingly corroborated by reality itself. But if sociological empiricism claims that the law is not something that exists in real terms, then it involuntarily denotes something of the social illusion in the object—an illusion which sociological empiricism wrongly attributes to the method. It is then precisely the alleged anti‑idealism of the scientific mentality which benefits the continued existence of ideology. The latter is supposed to be inaccessible to science since it is not, of course, a fact. Yet nothing is more powerful than the conceptual mediation which conjures up before human beings the being‑for-another (das Füranderesseiende) as an in‑itself, and prevents them from becoming conscious of the conditions under which they live. As soon as sociology opposes recognition of what is known as its 'fact' and remains content simply to register and order it—in so doing, mistaking the rules distilled for the law which governs the facts and in accordance with which they develop—then it has already succumbed to justification, even if it does not suspect that it has done so. In the social sciences, one cannot therefore proceed from the part to the whole as one can in the natural sciences, since it is something conceptual, totally different in its logical extension and in the unity of features of any individual elements, which constitutes the whole. Nevertheless, because of its mediated conceptual nature, this whole has nothing in common with 'totalities' and forms, which necessarily must always be conceptualized as being immediate. Society has more in common with the system than with the organism. An empirical research devoid of theory which gets by with mere hypotheses is blind to society as a system, its authentic object, since its object does not coincide with the sum of all the parts. It does not subsume the parts nor is it made up, like a geographical map, of their juxtaposition of 'country and people'. No social atlas in the literal and figurative sense represents society. Insofar as society is more than the immediate life of its members and the related subjective and objective facts, research which exhausts itself in the investigation of such immediacy misses its mark. For all the hypostatization of the method, even by virtue of such hypostatization as the idolization of what can be simply observed, it produces an illusion of being alive, an illusion of neighbourliness, as it were, from countenance to countenance. A dissolution of such an illusion would not be the last of the tasks for social knowledge if it had not already been dissolved. Today, however, it is repressed. In this respect, the transfiguring metaphysics of existence and the rigid description of what is the case are equally guilty. Moreover, to a considerable extent, the practice of empirical sociology does not even comply with its own admission that hypotheses are necessary. Whilst the necessity of the latter is reluctantly conceded, each hypothesis is met with suspicion since it might become a 'bias' and lead to an infringement of impartial research.  This view is based upon a 'residual theory of truth', upon the notion that truth is what remains after the allegedly mere subjective addition, a sort of cost price, has been deducted. Since Georg Simmel and Freud, psychology has realized that the conclusiveness of the experience of objects, if the latter in turn are essentially subjectively mediated like society, is increased and not decreased by the degree of subjective participation of the knowing subject. But this insight has not yet been incorporated into the social sciences. As soon as individual common sense is suspended in favour of the responsible behaviour of the scientist, people seek salvation in procedures which are as free from hypotheses as possible. Empirical social research ought to dismiss completely the superstition that research must begin like a tabula rasa, where the data that are assembled in an unconditioned manner are prepared. In so doing, it ought to recall epistemological controversies which are indeed fought out long ago, but are forgotten all too willingly by short-winded consciousness in its reference to the urgent requirements of the research process. Scepticism with regard to its own ascetic ideals befits a sceptical science. The readily‑quoted statement that a scientist needs 10% inspiration and 90% perspiration is secondary and leads to a prohibition on thought. For a long time, the abstinent work of the scholar has mainly consisted in renouncing for poor pay those thoughts which he did not have in any case. Nowadays, since the better paid executive has succeeded the scholar, lack of intellect is not only celebrated as a virtue on the part of the modest well‑adapted person who is incorporated into the team, but, in addition, it is institutionalized through the establishment of levels of research which hardly recognize the spontaneity of individuals as anything other than as indices of friction. But the antithesis of grandiose inspiration and solid research work is, as such, of secondary importance. Thoughts do not come flying along but rather they crystallize in protracted subterranean processes, even if they emerge suddenly. The abruptness of what research technicians condescendingly call intuition marks the penetration of living experience through the hardened crust of the communis opinio. It is the long drawn‑out breath of opposition to the latter, and by no means the privilege of highly gifted moments, which permits regimented thought that contact with being which is often inexorably sabotaged by the distended apparatus that intervenes. Conversely, scientific assiduity is always both the operation and exertion of the concept, the opposite of the mechanical, doggedly unconscious procedure with which it is equated. Science should be the recognition of the truth and untruth of what the phenomenon under study seeks to be. There is no knowledge which is not, at the same time, critical by virtue of its inherent distinction between true and false. Only a sociology which set the petrified antitheses of its organization in motion would come to its senses.
The categorial difference between the discipline is confirmed by the fact that what should be fundamental, namely the combination of empirical investigations with theoretically central questions, has—despite isolated attempts—not yet been achieved. The most modest demand and yet, in terms of immanent critique, the most plausible demand for empirical social research in accordance with its own rules of 'objectivity', would be to confront all its statements directed at the subjective consciousness and unconsciousness of human beings and groups of human beings with the objective factors of their existence. What seems merely accidental or mere 'background study' to the domain of social research provides the precondition for the possibility of social research ever reaching the essential. Inevitably, in these given factors, it will first emphasize what is connected with the subjective opinions, feelings and behaviour of those studied, although these connections, in particular, are so wide‑ranging that such a confrontation ought not really to content itself with the knowledge of individual institutions but instead should have recourse to the structure of society. The categorial difficulty is not removed by means of a comparison between certain opinions and certain conditions. But even with this weighty reservation, the results of opinion research acquire a different value as soon as they can be measured against the real nature of what opinions are concerned with. The differences which thereby emerge between social objectivity and the consciousness of the subjectivity, no matter in what form this consciousness may be generally distributed, mark a place at which empirical social research reaches knowledge of society—the knowledge of ideologies, of their genesis and of their function. Such knowledge would be the actual goal, although not of course the only goal, of empirical social research. Taken in isolation, however, the latter does not have the weight of social knowledge. The laws of the market, in whose system it remains in an unreflected manner, remain a façade Even if a survey provided the statistically overwhelming evidence that workers no longer consider themselves to be workers and deny that there still exists such a thing as the proletariat, the non‑existence of the proletariat would in no way have been proved. But rather, such subjective findings would have to be compared with objective findings, such as the position of those questioned in the production process, their control or lack of control over the means of production, their societal power or powerlessness. The empirical findings concerning the human subjects would certainly still retain their significance. One would not simply have to ask within the content of the theory of ideology how such modes of consciousness come about, but also whether something essential has changed in social objectivity through their very existence. In the latter, the nature and self‑consciousness of human beings, no matter how this is produced and reproduced, can only be neglected by erroneous dogma. Even the existence of such consciousness, whether as an element of the affirmation of what exists or as a potential for something different, is a moment in societal totality. Not only theory but also its absence becomes a material force when it seizes the masses. Empirical social research is not only a corrective in that it prevents blindly superimposed constructions, but also in the relationship between appearance and essence. If the task of a theory of society is to relativize critically the cognitive value of appearance, then conversely it is the task of empirical research to protect the concept of essential laws from mythologization. Appearance is always also an appearance of essence and not mere illusion. Its changes are not indifferent to essence. If no one in fact knows any more that he is a worker then this affects the inner composition of the concept of the worker even if its objective definition—through separation from the means of production—is still fulfilled.
Empirical social research cannot evade the fact that all the given factors investigated, the subjective no less than the objective relations, are mediated through society. The given, the facts which, in accordance with its methods, it encounters as something final, are not themselves final but rather are conditioned. Consequently, empirical social research cannot confuse the roots of its knowledge—the givenness of facts which is the concern of its method—with the real basis, a being in‑itself of facts, their immediacy as such, their fundamental character. It can protect itself against such a confusion in that it is able to dissolve the immediacy of the data through refinement of the method. This accounts for the significance of motivational analyses although they remain under the spell of subjective reaction. They can indeed seldom rest upon direct questions; and correlations indicate functional connections but do not elucidate causal dependencies. Consequently, the development of indirect methods is, in principle, the opportunity for empirical social research to reach beyond the mere observation and preparation of superficial facts. The cognitive problem of its self‑critical development remains, namely that the facts ascertained do not faithfully reflect the underlying societal conditions but rather they simultaneously constitute the veil by means of which these conditions, of necessity, disguise themselves. For the findings of what is called—not without good reason—'opinion research' Hegel's formulation in his Philosophy of Right concerning public opinion is generally valid: it deserves to be respected and despised in equal measure.  It must be respected since even ideologies, necessary false consciousness, are a part of social reality with which anyone who wishes to recognize the latter must be acquainted. But it must be despised since its claim to truth must be criticized. Empirical social research itself becomes ideology as soon as it posits public opinion as being absolute. This is the fault of an unreflectedly nominalistic concept of truth which wrongly equates the 'volonté de tous' with truth in general, since a different truth cannot be ascertained. This tendency is particularly marked in American empirical social research. But it should not be dogmatically confronted with the mere assertion of a 'volonté générale' as a truth in‑itself, for instance in the form of postulated 'values'. Such a procedure would be loaded with the same arbitrariness as the installation of popular opinion as objectively valid. Historically, since Robespierre, the establishment of the 'volonté générale' by decree has possibly caused even more harm than the concept‑free assumption of a 'volonté de tous'. The only way out of the fateful alternative was provided by immanent analysis; the analysis of the consistency or inconsistency of opinion in itself and of its relationship to reality (Sache) not however the abstract antithesis of the objectively valid and of opinion. Opinion should not be rejected with Platonic arrogance, but rather its untruth is to be derived from the truth: from the supporting societal relationship and ultimately from the latter's own untruth. On the other hand, however, average opinion does not represent an approximate value of truth, but instead the socially average illusion. In the latter, there participate what unreflective social research imagines to be its ens realissimum: those questioned, the human subjects. Their own nature, their being as subjects, depends upon the objectivity, upon the mechanisms which they obey, and which constitute their concept. This can only be determined, however, if one perceives in the facts themselves the tendency which reaches out beyond them. That is the function of philosophy in empirical social research. If it is not realized or suppressed, if merely the facts are reproduced then such a reproduction is at the same time a corruption of facts into ideology.
1 Cf. Kant, Critique of Pure Reason, trans. N. Kemp Smith (London/New York, 1933), pp. 586f.; Hegel's Science of Logic, trans. A. V. Miller (London/New York, 1969), pp. 795ff.; and numerous passages in Nietzsche.
2 'Sociology and Empirical Social Research' in Aspects of Sociology (London/Boston 1973), p. 124 (amended translation).
3 M. Horkheimer and T. W. Adorno, Dialectic of Enlightenment (New York, 1972 / London, 1973), p. 36.
4 Cf. Emile Durkheim, Les Règles de la méthode sociologique (Paris, 1950), pp. 6ff.
5 Cf. Erich Reigrotzki, Soziale Verflechtungen in der Bundesrepublik (Tübingen, 1956), p. 4.
6 Cf. 'Ideologie und Handeln' in Max Horkheimer and Theodor W. Adorno, Sociologica II, Reden und Vortrage, Frankfurter Beitrage zur Soziologie, vol. 10, 2nd ed. (Frankfurt, 1967), pp. 41f.
7 Cf. René König, 'Beobachtung und Experiment in der Sozialforschung', in Praktische Sozialforschung (Cologne, 1956), II, p. 27.
8 Cf. Hegel's Philosophy of Right, trans. T, M. Knox (Oxford/New York, 1952), §318, p. 205.