Ladies and gentlemen,
In the previous session I showed you that a thinker, a sociologist, who can be considered a positivist in the sense that he rejects any autonomy of the concept from the facts encompassed by the concept, is nonetheless forced in spite of himself – and precisely because of the material, the facts, the data in which he immerses himself – to incorporate theoretical elements into his work far beyond what his own epistemology would lead one to expect. In this context I reminded you both that he constructs a large, continuous social tendency and that there are dialectical elements in his work, and the example I cited to illustrate this was the reversal of charismatic authority into traditionalist authority and the transformation of rational and democratic administration, bureaucracy, into a blind, opaque, irrational rule. My reason for addressing these things was not that I intend to give you a lecture on Max Weber. My concern is for you to see things in their correct proportions from the start, by which I mean that you should understand such evidence – though it is always relevant to the content too – as contributions to the general theoretical problems, or the problem of the theory of society we are dealing with here; I simply wish, at least to a certain extent, to elaborate on these problems for you with a constant eye on Max Weber’s work.
The conclusion to be drawn from what we covered in the previous session is that even such essentially anti-theoretical or atheoretical thought and research as that of Weber, at least in terms of his intentions – I will remind you of just one principle, that of the ‘ideal type’ – necessarily either turns into theorizing or cannot dispense with theory. And the very simple conclusion I would draw from this is that one should not accept this shift into theory as a form of necessary evil that cannot be avoided but, rather, that one becomes aware of and theoretically formulates this connection between even extremely material-based work and theory. I just told you in this context that, with Weber, this move towards a theory formation resulted precisely from the pressure, the weight of the facts he had to deal with, and this provides an indication I would like to make use of immediately for the things we must fundamentally consider in these lectures, namely an indication that the form of theory whose concept I wish to elaborate for you here is not a concept of free-floating, a priori theory. Rather, the kind of theorizing I mean really depends on the facts in question, so – to exaggerate somewhat – one does not come closer to theory in general by moving away from the facts, by isolating some particular characteristics of the facts, forming theories from them and neglecting the rest; with the notion of theory I have in mind, it is precisely the immersion in the concretions that allows us to move beyond the merely factual. One must devote oneself to the individual materials so thoroughly as to reveal more than simply the blind, conceptless material.
Ladies and gentlemen, I think you will not expect me already to carry out this programme, which is really an epistemological programme, in the elements of a theory of society that I hope to present to you; but I do at least wish to indicate the direction of these considerations and to draw your attention to a difference that should be noted at this point compared to the usual and widespread concept of theory. For as long as prevailing scientific thought, including the field of sociology, engages in forming theories, it is simply obeying the logic of classification – that is, theory consists merely in the ordering of materials, the establishment of logical classes and the possible conclusions to be drawn therefrom. In other words, then, to the extent that theory is in evidence, it lies at a distance from the factual material, which leads to that somewhat unfruitful dualism between a blind material on the one hand and relatively empty concepts on the other, which means that theory, instead of leaping out of the material itself, really becomes – and I will show you this in detail – a shell for the facts encompassed by it. So the opposition to the prevailing formation of theories in the social sciences in which I find myself is not – as some of you will perhaps think – that of a boundless speculation which overrides the actual materials. For, if there is a concept of theory to be found that lies beyond that unfruitful dualism, it can only be one that disregards the usual rigid antithesis of fact and concept.1 I will mention only in passing that this marks the very nerve centre of dialectic; those of you whose interests are primarily philosophical will have noticed that anyway, and for the others I would not offer any more than that in this present context. The fact that even so-called positivist research, if carried out seriously, cannot avoid forming theories has naturally been observed by the positivists; it is not my discovery, ladies and gentlemen, and one can even say, generally speaking, that the willingness to concede this in social science research becomes all the greater the more those people truly surrender to their research, the more they saturate themselves with facts. I would almost say that those who are generally most hostile to theory are the epistemologists, the people who are, in a sense, the most theoretical and abstract, and who do not allow the experience of their objects to lead them towards theory. An example of such an extreme, to name only one, is the American sociologist Lundberg, who probably exhibits the most severe anti-theoretical stance; you should familiarize yourself with his philosophy to get a true idea of this extreme, for naturally one can only assess these matters if one has first-hand knowledge of them, as opposed to simply listening to the pre-chewed opinion of an academic teacher, which obviously has a tendency to harden into a form of authoritarian judgement, whereas I mean the opposite. So I will not only be grateful to you but actually ask you to examine closely and study the very works whose positions stand in extreme opposition to the position I am developing – all the more because in these lectures, if I am to fulfil my own theoretical programme, I cannot do much more than point you towards such works. But if we take those positivist sociologists who are not as radical as Lundberg, and who are well disposed towards theory, one will generally find that they are paying lip service rather than truly recognizing the weight of theory. They do say that one cannot dispense with theory: they observe that, for example, an empirical study that is not based on some undetermined, theoretically articulated expectation, on a certain anticipation of facts yet to be confirmed, generally has few prospects of finding anything fruitful or productive; thus, like it or not, they feel forced to speak of theory and grant it some validity in its own right, for example by imposing certain – how to put it? – certain conditions on it, the way one might impose them on a disreputable social or professional group, namely that such a theory must regularly report to the fact-finding police, as it were, and point out at every conceivable opportunity that it is only a theory, for heaven’s sake, not something already confirmed or even confirmable by the facts. In the language of positivism, this is known as scientific cleanliness. I do not entirely understand this demand, because I think that the difference between something that is genuinely an idea and a blind fact should really be obvious even without such assertions. What we are dealing with is actually the aspect I have already touched on in the name of the autonomy of theoretical elements. At this point, the social-scientific argumentation of positivism is, I feel, at best highly inconsistent. For, on the one hand, the necessity of forming theories as opposed to pure fact-finding is conceded, but, on the other hand, the development of theoretical reflections is really only tolerated as a formulation of hypotheses that can then be fulfilled by finding the facts. To put it dialectically, then – and even the positivists sometimes find themselves in a dialectic – something is supposedly necessary, because one cannot do without it, but simultaneously superfluous, because, according to this view, once the hypothesis is fulfilled, once it proves true or is refuted, it can be discarded; so, in such a view of things, the autonomy of theory as an expression of something in society that is not limited to the ascertainment of individual facts is reversed again.
Ladies and gentlemen, I am not a natural scientist, nor do I presume to act as if I were. It does seem to me, however, as if the actual terminus ad quem in the natural sciences, that is, the ultimate goal of scientific work, is precisely the formation of theories, the summarizing explanation of the individual observable facts, and that, in general, scientific experiments are only carried out from the perspective of forming theories; no experiment, no ascertainment of any pure, existent fact takes place for its own sake. In positivist, statistically oriented social science, which prides itself so greatly on operating more mathematico, on following mathematical, scientific rules, the exact opposite is the case: here it is really mere fact-finding that takes precedence, and, from this perspective, the construction of theories itself seems to be a mere useless ingredient that is essentially superfluous, or at least a somewhat disreputable matter. Now, ladies and gentlemen, the fact that things happen in this way is obviously no coincidence but, rather, points to the heterogeneous character of the subject areas in so-called nature and so-called society. That is to say, it expresses the fact that natural science generally works as far as possible with elements that have no properties of their own, that are not already preformed, not already objectively qualified, and that, the more it can reduce phenomena to such propertyless smallest units or elements, which can in turn be integrated into a seamless theoretical summary, the greater is its sense of triumph. In the social domain, of course, which is defined qualitatively to an incredible degree, where qualitative differences are virtually the decisive aspect, this is possible only to a very limited extent – to the extent, for example, that strictly statistical surveys whose elements are fundamentally equated with one another at random can even be carried out. And it has transpired that this is indeed possible in this specific domain, simply due to how reliable the predictions of opinion polls are, for reasons that we can perhaps discuss a little more later on. But naturally it cannot be said that in the domain of hierarchically structured society, where the behaviour of these elements is not a last object of observation but, rather, something eminently predetermined by the overall context, there can be a reduction like the one that is possible in the natural sciences; consequently, it is equally unfeasible to bring together the elements thus subsumed and classified similarly with a uniform classificatory theory, which would ultimately mean functional equations, the way one can in natural science. That is why it so readily occurs – precisely where the social sciences follow the model of natural science – that they really produce only something like a parody of the natural-scientific method, namely ascertaining and presenting facts, which in natural science leads to theoretical conclusions, to formulas, to some more or less final conceptual definitions, whereas the statement and arrangement (if at all) of ascertained facts so often remains the same. If you pick up the Journal of Sociology,2 for example, and take a general look at the short abstracts that tend to follow each report on whatever empirical investigation, you will find that there are really only classifications of facts, perhaps only in the form of charts, without any of the theoretical conclusions being drawn that are characteristic and central for the genuinely scientific method.
Even so, empirical sociology is naturally far too cunning and self-reflecting to avoid an awareness of this element that can be termed atheoretical in a deeper sense. The answer it usually provides is that sociology is such a young science, and, because of its youth, it is not yet able to construct larger theories in the manner possible in the exact natural sciences. I must say that, to me, this explanation, which one encounters time and again – at congresses too, one constantly hears people talking about the youthfulness of sociology – such explanations are simple nonsense, somewhat reminiscent of a lady of advanced age (I am thinking of an example from Frank Wedekind)3 who declares that she was extremely attractive in her youth. And this supposed youth is a spurious claim; if one starts with Comte,4 sociology is a good 150 years old, and I would actually say, if one traces it back to Saint-Simon,5 it is roughly 200 years old. Now, I think it is a somewhat crude analogy, but an acceptable one, if one imagines what happened in the roughly 200 years between the discoveries of Copernicus and the development of Newtonian physics – that would be around 200 years, perhaps a little more, I don’t recall the exact dates – what happened in the natural sciences, then whatever has ensued in the way of social theory in the meantime, at least in the framework of the established, institutionalized social sciences, can really be considered quite meagre. And, despite its advanced age, sociology does not show the slightest inclination to approach the ideal of a genuine theory of society, one that would illuminate its workings. On the contrary: if one looks at the history of sociology, it is more fitting to speak of it moving away from the formation of theories than towards it. If I can remind you of what I said last time, for example the fact that, for Durkheim, Comte’s statements about the development of society as a whole already seemed like metaphysics, then I am sure you will understand that; and Lundberg would undoubtedly consider the principle of the collective spirit’s objectivity, which Durkheim espoused, no less of a metaphysical prejudice than Comte’s idea was in Durkheim’s eyes. Hence the direction of development within this established science is not at all towards the theory that people promise, but rather the opposite. So the reason for this problem, as I have already hinted, is not the mere temporal factor that the time is not yet right, but rather something categorial. The anti-theoretical trait lies in the restriction of sociology to pure empiricism, and sociological empiricism – as I learned in my disputes with Mr Silbermann6 – is at its most sensitive when one reminds it that the purpose of sociology is not fact-finding, not the mere collection of opaque data, but rather a concept of theory that reveals something substantive – I am deliberately putting it so vaguely – about society.
I remember – and I am telling you this because I would like especially to try and explain the relationship between theory and empirical research as closely as possible, with reference to the approach of empirical sociology – the dispute with a young and, as it happens, very talented assistant on a project at the Institute for Social Research whom I had given certain – well, they were definitely hypotheses for a particular study – and already at that point – in this respect he was highly characteristic of the mindset that is really displayed by sociological empiricism as such – he considered this formulation of hypotheses something questionable and said, ‘Well, if one approaches the facts with something like a hypothesis, then this hypothesis is really always on the verge of becoming a form of prejudice, and then one can no longer devote oneself purely to the facts at all.’ Now, I do not actually mean to dispute the claim of this young sociologist, who is probably not so young any more, that this is possible. Those of you who have read the piece ‘Opinion Delusion Society’ from my Interventions7 will find a fairly detailed elaboration of the argument that, because of its inherent weight, the anticipatory, unfulfilled, not factually posited opinion – for metapsychological, objective reasons, I would say – has a tendency to establish itself, to harden, to become objectified and resist correction, and certainly this much is true about that attitude: it is a necessary part of social knowledge, especially theoretical social knowledge, to remain open, to examine itself and improve itself.
Ladies and gentlemen, I am well aware that this is easier to formulate in a general way, to set up as a norm, than actually to execute; I have no illusions about this, but I would at least say that a decent scholar, by which I mean a scholar – or a researcher, one should almost say – who takes the truth seriously will consciously and emphatically incorporate the possibility of such self-reflection as one of the most important aspects of their work. But I do think that, for there to be anything resembling examination or reflection, there must first be something to examine or which can be subjected to such reflection. So, if one begins simply with examining theoretical views before there are actually any to examine, it is quite certain that no theory will be formulated at all. Let me give you an example of this too from practical social science work. When I had to take over as head of the ‘Community Study’8 in Darmstadt – that would have been twelve years ago, or maybe more – I asked one of the American directors what he and his group actually intended to find out with this study, what they actually wanted to know, what the point of the study actually was, and I asked him [in English, trans.], ‘What do you want to know or what do you want to find out?’ And the respective American colleague, a very friendly and affable man, replied, ‘We want to know just everything’, to which I responded, ‘If you want to know everything there is to know about this city, Darmstadt, then you won’t know anything, you’ll suffocate in such a conceptless material that nothing will be visible at all.’ When I later began to look at the material, I actually encountered things such as a thorough statistical report on the weather in Darmstadt for every day over the duration of the study. I will not even deny the possibility of finding some form of correlation between some socially relevant trends, for example the number of children produced, and the weather; but, for that, one first needs to have the idea of doing that in the first place, which a strict empiricist would probably forbid as an impermissible pre-emption of the future, so that would not be possible either. In other words, then, and I would already like to draw a conclusion from this today – I have to proceed fairly quickly in these lectures, as we will miss such an inordinate number of sessions this semester,9 so we might travel across country a little at times, which I ask you to forgive, ladies and gentlemen – namely that, in sociology, theory is not simply something that results from the facts automatically as long as one follows the rules, but that there is a sort of leap between theory formation, between an insight into socially relevant factors and the gathering of facts, that there is not a straightforward and reliable continuum between the fact and the theory, but that, to a certain extent, these two areas, as closely as they are obviously and necessarily connected, cannot be imagined in such a way that the path leads from the one to the other without the addition of a further element. And I would say that, if one is to develop a genuine authority for sociology and social contexts, one must, on the one hand, always be aware that one cannot simply move from the concept of society to the facts – as George put it, simply make butter out of the Milky Way10 – but also that, on the other hand, it is equally impossible to arrive at a theory by proceeding from a mere arrangement and gathering of facts and merely placing these in a vaguely logically stringent context. So I think that the crux of the question of fruitful sociological work lies in how clearly one is aware of this necessary and inevitable qualitative leap and, to continue this metaphor, whether one succeeds in leaping or makes this leap unconsciously, and thus incorrectly, in such a way that it fails. In part, the reason for this is simply a matter of scientific history, namely that what we group together under the name of sociology or social science today, and to a considerable extent also political science, is not an internally consistent area of knowledge with a consistent categorial structure of the same kind as in most of the traditional, the so-called classical sciences, but simply a more or less unconnected and unconsidered agglomeration of very different things – such as philosophical reflection on society and its purpose, institutional and historical analysis of social facts, and finally what are more strictly considered ‘research techniques’ on the model of so-called market research. And I do not think that the future of sociology and sociological theory formation lies in attempting to knock this agglomerate into shape by adapting it to some consistent categories and making a unity of it, a unity that corresponds neither to its category nor to the matter itself; rather, to the extent that there is any unity, it can be determined only by determining precisely these qualitative differences and by first marking the different areas off from one another and then thinking through the extremely complex and mediated relationships that exist between these sub-disciplines and sub-complexes, which are in themselves entirely dissimilar things.
I would like to say especially to those of you who are in the first semester or the first few semesters of the sociology course, and who naturally begin by thinking, ‘So, we’d like to gain an overview of the scientific field as a whole’ in the same way a student of jurisprudence can expect (I am deliberately using this example), as jurisprudence, aside from the natural sciences, is the epitome of such a completely consistently structured, dogmatic theory – I would like to tell you from the start that you will not find any such thing in sociology. Rather, to the extent that there is a formation of theories, it assumes precisely a reflection on these discontinuities, and you should not believe that you can presuppose any implicit unity that would show you the way, as it were, through a field that is not only immense but also full of cracks. In telling you this, I am pointing out once again that the interpretation of social facts is closely connected to the facts themselves and their divergence, and necessarily changes with them. But I think it is necessary at this point to move on with our subject and to ask why the facts within society do not simply lend themselves to theory, and to theory formation, of the same type as one finds in the natural sciences. So, in other words, I think that reflection on the relationship of theory and fact in sociology, in the social sciences, requires a reflection on the role played by the so-called facts within society itself; and this role of the facts, the status of the facts within the complexion of a total society, is really the crux of the entire problem of sociological theory formation – simply because, for a number of reasons, these facts are not merely identical to society itself, because society is by no means limited to its own facts in the same way that jurisprudence rests on the findings which can be subsumed under the law, let alone the way in which facts correspond to the laws of natural science. In the first session after the Whitsun holiday, then, we will have to start by considering the status in society of the facts themselves, of the factual, and the consequent implications for our knowledge, which will simultaneously bring us closer in content to an idea of a theory of society.