‘At the crossroad of Magic and Positivism’. Roots of an Evidential Paradigm through Benjamin and Adorno

A Hamlet without the prince

As David Frisby stated, the famous ‘Positivist Dispute’ [PD] (Adorno et al., 1976) was actually not a positivist dispute at all, in that ‘positivism provided no reason for dispute’ (Frisby, 1976: xxiv). Putting it in more literary terms, Anthony Giddens described the dispute held in 1961 in Tübingen as ‘a Hamlet without the Prince’ (Giddens, 1974: 18). In fact, not one of the main participants in the controversy (Adorno, Popper, Habermas, Albert) claimed to be a positivist in the true, historical sense of the word, that is, adherents to the philosophy of the founding fathers of sociology, such as Saint Simon, August Comte and Hebert Spencer.

On the contrary, all the debaters had claimed allegiance to either critical rationalism or critical theory. Indeed, in his introduction, Adorno argues against a naïve positivism, which was hardly at issue among any of the disputants, even though it may remain operative in much of social science practice. Adherence to the central role of theoretical criticism is one reason that led Dahrendorf, with reference to the Popper-Adorno debate, to suggest that a superficial glance at the controversy might incline one to the view that ‘it could indeed have appeared, astonishingly enough, as Popper and Adorno were in agreement’ (Dahrendorf, 1976: 123). This is clearly far from being the case, since as the controversy proceeded, even the notion of a debate between competing standpoints became problematic. If the positivist dispute is not in the strict sense a Methodenstreit (like the earlier Schmoller–Menger Methodenstreit in the 1870s and 1880s in Germany), the central issue in the controversy is to be found in the role and political aim of social science, in general, as a form of practical reason. ¹ The controversy between Critical Rationalism and Critical Theory was between those who supported the idea of western society as an ‘open society’ (Popper) and those espousing a conception of it as ‘unidimensional society’ (like Marcuse would later name it: Marcuse and Popper, 1976 [1972]). After more than half a century, politics, science and philosophy do not seem to be so interrelated as they used to be in the 1960s. The relation between theory and political praxis is less immediate, and hence, the methodological question has gained more importance. In other words, although the question knowledge for what? is still relevant (see, for example, the recent debate on ‘public sociology’), the question of how to know – the concrete modus operandi of social research – becomes equally important.

The methodological question has been largely neglected in the dispute. In this sense, ‘the Prince’ missing in the Hamletian dispute regards not just ‘positivism’, but ‘sociology’. Among the various contributions, very few indications can be found on ‘how to do sociology’ from the perspective of critical rationalism and critical theory. Adorno himself was aware of this fact when he wrote that Popper and he ‘both failed to achieve complete mediation of their theoretical interests with sociology as such’ (Adorno, 1976a [1969]: 2). Practical indications can only be discerned ‘between the lines’ or indirectly. ²

In the following pages, we will attempt to show that Adorno implicitly made reference to an epistemological model (a paradigm, if you prefer) in the social sciences. This paradigm has failed to attract sufficient attention, though it is very much operative in spite of never having become explicit theory. Adorno evokes it in the PD repetitively as ‘physiognomy’, affirming icastically, ‘knowledge of society which does not commence with the physiognomic view is poverty-stricken’ (Adorno, 1976a [1969]: 33). He moreover cites Walter Benjamin, Georg Simmel, Sigmund Freud and Karl Kraus as practitioners of this approach. Generally speaking, physiognomy is that para-scientific discipline whose aim is to identify the psychological and moral character of people by inferring it from their physical appearance. The main criticism of physiognomy is that it is a deterministic and astorical form of knowledge, one which denies the importance of cultural and social factors. In this sense, physiognomy can be a source of racism and stereotyping. Nevertheless, many other authors (Goethe, Dilthey, Cassirer) have, in various ways, included as part of a complex and anti-reductionist theory of culture that focuses on aspects ignored by other approaches (Dodd, 2008; Moynahn, 1996). Recently, Honneth (2005) observed that ‘the heading’ under which Adorno formulates his sociological programme in his writings is ‘a physiognomy of the capitalist form of life’ (Honneth, 2005: 56). ³ Moreover, this philosophical–hermeneutical project cannot be understood unless it is set in relation to the philosophy of Walter Benjamin and the various stages of the Benjamin–Adorno debate over Benjamin’s opus magnum: The Arcades Project. The central focus of this work was an attempt to develop a materialistic physiognomy of urban culture. Even if Adorno would criticise Benjamin sharply, he eventually re-functionalised his ‘empirical metaphysical’ method, applying it in his sociological research and using it as an aesthetic correction to Max Weber’s conception of ideal type. Before re-visiting Benjamin and Adorno’s physiognomy, we will see – with the help of the Italian historian Carlo Ginzburg’s (1989) work – that the roots of physiognomy are actually much older, and that in the late nineteenth century (more precisely in the decade 1870–1880) an ‘evidential paradigm’, based specifically on semiotics, began to assert itself in diverse disciplines of the human sciences, such as art history, medicine and psychoanalysis.

The physiognomic or ‘Morellian method’

Although Carlo Ginzburg (1989) quotes Adorno only once, and even then only marginally, there are several affinities between the former’s ‘evidential paradigm’ and the physiognomy proposed by Adorno in his discussion with Popper.

Ginzburg showed that towards the end of the nineteenth century, the humanities saw the emergence of an epistemological model that was based on physiognomy applied as a complex semiotic practice. The art historian Giovanni Morelli, who first published under the pseudonym Ivan Lermolieff, provoked controversy and lively discussion by creating what is remembered still today as the ‘Morellian method’. Morelli’s conviction was that attributing a painting to its true author should not depend on analysing its most conspicuous characteristics, which are the easiest to imitate (Perugino’s eyes turned towards the heavens, Leonardo’s smile, and so on). Instead, one should examine the most trivial details that would have been least influenced by the mannerisms of the artist’s school. For this reason, Morelli’s books are sprinkled with illustration of earlobes, fingernails, and the shapes of fingers and toes. Using this method Morelli identified and faithfully catalogued the shapes of the ears in figures painted by Botticelli, Cosmé Tura, and others – features that were present in the originals, but not in copies. He ended up proposing new attributions for many works hanging in Europe’s major museums, with some sensational new identifications: Giorgione, Titian, among others. Despite his success, his method was heavily criticised as arrogant, mechanical and crudely positivistic.

What Morelli actually practised was a detailed type of physiognomy of figurative art, one which was so successful as to attract even Sigmund Freud’s attention. ⁴ Morelli’s method shared with psychoanalysis an analogous process of interpretation based on discarded information, on marginal data, which were instead considered in some way significant. For Morelli, as well as Freud, marginal facts were revealing because they constituted the instances when the subject’s control (for Morelli, the artist’s) would relax and purely individual touches were disclosed, those ‘which escaped without his being aware of it’ (Morelli, quoted by Ginzburg, 1989: 101). Psychoanalysis supports Morelli on this central point: our inadvertent little gestures reveal our character far more authentically than any formal posture that we may carefully prepare. Morelli’s method can be compared not only to psychoanalysis, but also to the process ascribed, almost contemporaneously, to Sherlock Holmes by his creator, Arthur Conan Doyle. The art connoisseur resembles the detective who discovers the perpetrator of a crime (or the artist behind a painting) on the basis of evidence that is imperceptible to most people. In each of the three cases (Morelli, Holmes and Freud), marginal traces enable understanding a deeper, otherwise unattainable reality: traces – more precisely, symptoms (in Freud’s case), clues (for Sherlock Holmes) and pictorial marks (Morelli). This threefold analogy is not coincidental: Freud was a physician; Morelli had a medical degree and Conan Doyle had practised medicine before turning to literature. In each of the three cases, the epistemological model is that of medical semiotics: that discipline which permits the diagnosis of diseases inaccessible to direct observation based on superficial symptoms, irrelevant to the eyes of the uninitiated. Towards the end of the nineteenth century, this presumptive paradigm, based specifically on semiotics, began to assert itself in the human sciences. Its roots, however, were much older ⁵ and can be traced to ‘the oldest act in the intellectual history of the human race: the hunter squatting on the ground, studying the tracks of his quarry’ (Ginzburg, 1989: 105).

The group of disciplines that can be called ‘evidential’ and ‘conjectural’ (medicine included) are difficult to relate to the scientific model derived from the Galilean paradigm. In fact, they are all highly qualitative disciplines, in which the object is the study of individual cases (humans, situations and documents), precisely because they are individual. Galilean science is endowed with totally different characteristics: mathematics and the empirical method implied quantification and the repetition of phenomena, respectively, while the individualising perspective by definition excluded the latter and admitted the former only as a mere instrument. Modern physics cannot call itself ‘Galilean’, but its epistemological and symbolic significance for science in general cannot be denied.

Popper himself warns in the PD against the myth of ‘false naturalism’ or ‘scientism’, fighting (and winning) a battle that actually was at the very heart of Adorno’s thought and intended to be the (supposed) centre of the dispute. Popper states that the misguided and erroneous methodological approach of naturalism or scientism has often been quite unconsciously and uncritically accepted and absorbed by the social sciences, in general, and sociology, in particular. This form of scientistic common sense ‘urges that it is high time that the social sciences learn from the natural sciences what scientific method is’ (Popper, 1976 [1969]: 90). This method expects the social scientist also to ‘begin with observations and measurements’; ‘proceed, next, by induction to generalisations and to the formation of theories’. In diametrical opposition to such method, we have ‘deductivism’ – the prince of all possible scientific methods: ‘in the sciences we work with theories, that is to say, with deductive systems’ (Popper, 1976 [1969]: 99). The deductive system is an attempt to solve a scientific problem, which is a problem of explanation. Popper believes that a theory formulated through a deductive system can be criticised rationally through its consequences. He concedes some role to observation, as long as its function is to reveal problems: observations that ‘show us that something is not quite in order with our knowledge, with our expectations, with our theories’ (Popper, 1976 [1969]: 89).

What would Morelli’s method look like in Popper’s view? Since physiognomy is an art of deciphering the unknown from the known, it has something in common with Popper’s ‘observation which creates a problem’ (Popper, 1976 [1969]: 89). On the other hand, deduction plays only a small role in the conjectural paradigm. It is not only the conjectural disciplines (art history, psychoanalysis, medicine) that apply a method that is difficult to codify and reproduce (Ginzburg speaks of ‘flexible rigour’). Physiognomic scrutiny can guarantee only a weak logical inference: not a necessary, but a probable, connection between facts. If Giorgione used to paint fingers or nails in a certain way, it is highly probable that a painting attributed to Giorgione that does not exhibit such fingers or nails is not by Giorgione, but a copy. More than on deduction, a conjectural paradigm is based on abduction (Peirce): a form of logical inference that goes from the observation of surprising circumstances to a hypothesis that could possibly explain the reliable date. Ab-ducere in Latin means ‘to lead away from’, which is the exact opposite of the linear thinking of deductive logic: it indicates ‘displacement’, moving from the conventional reality to the unconventional one. In abductive, as opposed to deductive, reasoning, the premises do not guarantee the conclusion. Despite all this, abduction plays a central role in the logic of scientific discovery. ⁶

Peirce (1929) also used the common term ‘guessing’ to describe abduction. Ginzburg dares, rather more perilously, to call it ‘intuition’. He specifies that there are at least two forms of intuition: one ‘mystical’ (the super-sensible intuition of various nineteenth- and twentieth-century forms of irrationalism), as well as other forms of ‘discernment and wisdom’ that are instruments of conjectural knowledge. ⁷ This ‘low intuition’ is based on the senses (though it skirts them) and can be found throughout the entire world, without limits to geography, history, ethnicity, sex or class. Thus, it is far removed from higher forms of knowledge, which are the privileged possessions of an elite few: ‘it is the property of the Bengalese, their knowledge having been expropriated by Sir William Herschel; of hunters; of sailors; of women. It binds the human animal closely to other animal species’ (Ginzburg, 1989: 125).

Now we shall see how a type of conjectural paradigm – based on a physiognomy of social and historical reality – was theorised and put into practice in Walter Benjamin’s Arcades Project and Adorno’s constellations method. Adorno’s contribution to the PD cannot be understood if we fail to consider the intense – and conflictual – dialogue he had with the thinker who probably influenced him most: Walter Benjamin.

Benjamin’s attempt at an urban physiognomy

In the Introduction to the PD, Adorno clearly states the origins of his interest in physiognomy. It was during his discussions with Benjamin during the long task of writing the Arcades Project (1924–1940). ⁸ Since Adorno’s physiognomy in this work is largely due to Benjamin’s influence, it is important to revisit the ‘old controversy’ and see how later Adorno himself, after criticising Benjamin, uses physiognomy in his own sociological investigations, such as the study On Jazz (Adorno, 2002 [1936]) and the collective study on The Authoritarian Personality (Adorno et al., 1950).

The epistemological roots of Benjamin’s physiognomy are to be found not directly in the materials of the Arcades Project, but in the dense and esoteric ‘Epistemo-Critical Prologue’ (Erkenntniskritische Vorrede), which is the foreword to his study on ‘German Baroque Drama’ (Trauerspiel) published in 1922. As Benjamin reveals at the beginning of the chapter, a fundamental property of any philosophy that does not wish to be merely ‘propaedeutic to knowledge’ is precisely the problem of the ‘representation’ (Darstellung) of truth (Benjamin, 1998: 27). From this, it follows that the form of the Trauerspiel is not a mere concept of a genre, but an ‘idea’ that must be represented by means of a ‘constellation’ of concepts and empirical materials. The researcher’s role – to draw connections between the phenomenal elements – is not unlike that of the astrologer, who perceives figures in the sky: ‘Ideas are to objects as constellations are to the stars’ (Benjamin, 1998: 34). Truth (the idea) literally springs from the assemblage of materials, even the most ‘extreme and imperfect ones’, as an ‘illumination’, as a form of ‘unintentional truth’. As mystical as this conception may appear (and it is in fact strictly related to the Judaic mysticism of the Kabbalah), with it Benjamin sought to achieve a unique convergence of Platonism and Kantianism. Ideas were nothing but the empirical phenomena, and at the same time, like ‘constellations’. In Plato, the ideas as truth appear in the phenomena. Benjamin’s theory was an inversion of Platonism: the phenomena appear as truth in the ideas, so that the ‘dignity’ of their particularity is maintained. Through his research, Benjamin was also aiming to ‘illuminate’ the idea of German Baroque Drama through ‘constellations’ of empirical materials, constituted mainly by quotations – not from just the best examples of dramas, but even from the less accomplished and neglected. ‘Juxtaposing the extremes’ was Benjamin’s guiding principle, which means discovering not only the similarity between opposites, but also the connecting links (‘the inner logic’) between seemingly unrelated elements of a phenomenon.

Far from being a merely mystical, dogmatic practice, Benjamin invented a new method, destined to attract Adorno’s attention: ‘illumination’, ‘juxtaposing the extremes’, ‘unintentional truth’, constructing constellations out of empirical material phenomena – these were the cornerstones of Benjamin’s paradigm. Benjamin’s conception of truth as a ‘constellation’ reveals an evident ‘elective affinity’ with Goethe’s concept of Urphänomen (primal phenomenon), developed especially in his writings on the ‘morphology of nature’. Here Goethe was trying to lay the foundations for a new form of ‘qualitative’ natural sciences grounded in physiognomy. In his naturalistic studies, Goethe was concerned with the possibilities and constraints of what can be known through experience, and what can be deduced by our reasoning. ⁹ This approach also lends itself in a highly congenial manner to a ‘representation’ of metropolitan reality, which is, by its very nature, fragmentary and discontinuous. Benjamin’s intention in the Arcades Project was to create a physiognomic representation ‘out of the facts’ with the complete elimination of theory (what Goethe attempted in his morphological writings), and this time of a very material reality: urban commodity culture.

In his metropolitan writings in general (on Naples, Moscow, Berlin), and particularly in the Arcades Project, Benjamin proffers a fascinating method for interpreting city culture: ‘Method of this project: literary montage. I needn’t say anything. Merely show… the rags, the refuse – these I will not inventory but allow, in the only way possible, to come into their own: by making use of them’ (Benjamin, 1999: 460). What we have is a physiognomy of the urban context, whose principal protagonist is to be found in the flâneur. Just as the flâneur’s stroll is a meandering without any precise destination – abandoning himself to the labyrinth of the city, following its lures and hidden attractions – so the same can be said of the construction of Benjamin’s urban texts. The experience of the metropolis is not represented in a coherent or systematic manner, but rather through snapshots of the moment – fleeting spontaneous images of everyday life. ¹⁰ Like the flâneur who ‘wanders around aimlessly’ along city streets, roaming through his own district and its most hidden monuments, Benjamin (1999) himself aimed, in the Arcades Project, to highlight ‘the expressive character (Audruckscharakter) of the earliest products and the first shapes of industrial architecture, of the first machines, but also of the first great department stores, advertising hoardings, etc.’ (p. 574). The fascinating, albeit highly problematic, ‘social physiognomy’ delineated by Benjamin thus proposed to read and interpret the city by starting out with its physical and material characteristics (the architectural shapes, the spaces, the billboards, etc.). In a totally eclectic manner, Benjamin considered the method that he had devised as related to historical materialism:

Marx lays bare the causal connection between economy and culture. For us, what matters is the thread of expression. It is not the economic origins of culture that will be presented, but the expression of the economy in its culture.

However, the first part of this study – entitled The Paris of the Second Empire in Baudelaire (Benjamin, 2003 [1938]) – was destined never to be published in the journal of the Institute for Social Research that had commissioned it, and was subjected to a barrage of particularly harsh criticism by Adorno. In Adorno’s vision, the interpretation of cultural phenomena can be undertaken only through the mediation of the global historical-social process. Any ‘immediate inference’ regarding links between economic and spiritual phenomena endows phenomena with precisely the type of spontaneity, concreteness and compactness they have lost in the capitalist context. Adorno also furnishes a summary of his criticisms, together with a definition of Benjamin’s method, which despite its stinging criticism, is undoubtedly insightful:

To express this another way: the theological motif of calling things by their names tends to switch into the wide-eyed presentation of mere facts. If one wanted to put it rather drastically, one could say your study is located at the crossroads of magic and positivism. This spot is bewitched. Only theory could break this spell …

Nevertheless, Benjamin was focusing not so much on Hegelian dialectics as on the idea of re-actualising his mimetic theory of language in light of the cinematic arts and, above all, the new avant-garde movement of surrealism. The flâneur is the urban physiognomist, the detective (as Benjamin himself observed, detective stories arose first precisely during this period in metropolitan settings), hunting for ‘the utopia that has left its trace in a thousand configurations of life, from enduring edifices to passing fashions’ (Benjamin, 1999: 4–5). The conjectural method of Benjamin’s ‘social physiognomy’ was therefore substantially different from the ‘dialectic mediation’ that forged Adorno’s line of reasoning. Only by referring to the technique of cinematographic montage could Benjamin’s approach be put into effect. In this perspective, Benjamin’s work was (and is) to be read as a city-as-text, which seeks to transfer into words the fragmentary and discontinuous character of the metropolitan experience, made of shocks, collisions, sudden changes of direction. The flâneur’s gaze is that of the photographer and the cameraman, which manages to unite, paradoxically, logos and poetry – the poetry-making word and the cognisant word.

Adorno’s constellations in sociology

In his profile of Benjamin, Adorno observed, ‘it was Benjamin’s intention to renounce all open explanation and to allow the meanings to emerge solely through the shock-like montage of the material. Philosophy was not only to catch up with surrealism, but become surrealist itself’ (Adorno, Charakteristik Walter Benjamins, 1950: 14; quoted in Buck-Morss, 1977: 134). Moreover, surrealist aesthetics sought to undermine Benjamin’s urban physiognomy. For Adorno, surrealism fused science and art by eliminating their differences (theory and concept in science, the logic of form in art), and Benjamin tried to fuse art with philosophy in the same manner.

Despite his harsh criticism, Adorno was always attracted by Benjamin’s method, ¹¹ and it is not by chance that he quoted it many years later in the dispute with Popper. In contrast to Benjamin, he tried to formulate his own model of physiognomy, drawing inspiration not from art imagery, but from musical composition. As a composer and musicologist, he clearly found his own experience to be prototypical of cognitive experience in general. However, Adorno remains firmly convinced that aesthetic models, whether in music or art imagery, cannot carry the whole weight of rational practice. Aesthetics can provide a corrective measure for positivism and pseudo-scientific rationalism (which Popper was also right to denounce in the PD), which did violence to the individuality of objects by consuming them within an abstract conceptual schema. Philosophical interpretation cannot go beyond the immediate appearances of reality without the theory and concepts developed by the sciences. For Adorno, science and art, concept and image, analysis and expression form the two poles of rationality.

In his Negative Dialectics (1983 [1966]) – a kind of theoretical testament – Adorno offers a practical example of his conception of the relation between art and (sociological) knowledge, commenting Max Weber’s theory. According to Weber’s definition, the ideal type is an instrument for observing social reality that ‘is formed by the one-sided accentuation of one or more points of view’, according to which ‘concrete individual phenomena are arranged into a unified analytical construct (Gedankenbild)’. Those familiar with Weber’s work know that ‘capitalism’, ‘the spirit of capitalism’ or ‘the Protestant ethic’, as well as ‘charismatic power’ or ‘bureaucracy’ are constructs of this sort. It should, however, be stressed that ideal types are in part ‘utopian’: they are a methodological ‘utopia [that] cannot be found empirically anywhere in reality’ (Weber, 1997: 90), even if they are stimulated by the existence of empirical facts. Keenly aware of its fictional nature, the ideal type never seeks to claim its validity in terms of a reproduction of, or a correspondence with, social reality. Its validity can be ascertained only in terms of adequacy. Adorno, in other words, comes to evoke Weber as an unwitting practitioner of the ‘constellation method’:

When Weber, in his treatise on the Protestant ethic and the spirit of capitalism, raised the question of defining capitalism, he – in contrast with current scientific practice – was well aware – as previously only philosophers such as Kant, Hegel, and Nietzsche had been – of the difficulty of defining historical concepts.

Instead of resorting to the delimiting procedure of definition – in adherence to the taxonomical schema genus proximum, differentia specifica traditionally used for biological categories – Weber asked that sociological concepts be ‘gradually composed [from] individual parts to be taken from historic reality. The place of definitive conceptual comprehension cannot therefore be the beginning of the inquiry, only the end’. ¹² The problem for Adorno is not one of finding a procedure that avoids definitions; rather, it is a criticism of a conception of language that regards knowledge as a mere adequatio intellectus rei. Adorno noted that even the purest, most precise use of language and neutral definitions is more than the pure definitional language of science – it must have something of the representational language of art (specifically of musical composition). Adorno, following Benjamin, believed that modern natural sciences are for humanity the minimum level of experience, based on a mechanistic conception of time and space. Lived experience is not what science considers experience. This ‘reified’ form of experience is the consequence of the nominalistic conception of language as pure ‘signifier’, without any necessary relation with the ‘signified’. In the introduction to the PD, Adorno recalls his conception of language as ‘expression’ (Ausdruck): ‘the closer [every linguistic form] follows the objective circumstances, the more it surpasses mere signification and comes to resemble expression’ (Adorno, 1976a [1969]: 35). ‘Expression’ refers to the mimetic faculty of perceiving and reproducing the similarities of surrounding nature. Thus, in Benjamin’s view, the origin of language does not spring from a casual relation between word and thing. Words are ‘names’ and the human individual who denominates things may or may not grasp their essence: the denomination is a sort of translation of that which has no name into the name, the translation of the imperfect language of nature into the language of man. The culture of mankind, like language more generally, therefore originates from this immediate, ‘mimetic’ relation with nature, whose elementary forms it tends to reproduce: the ‘expression’ is systematically linked to the stimulating qualities of the environment.

Even when Max Weber appears to adhere to the traditional ideal and theory of science, aspects of the ‘constellation method’ are not lacking. In his most mature works, such as for instance the posthumous Economy and Society, which seems at times to suffer from a glut of verbal definitions borrowed from jurisprudence, a closer look reveals the ‘constellation method’, which operates ‘by gathering concepts round the central one that is sought, [in an] attempt to express what that concept aims at, not to circumscribe it to operative ends’. ¹³ What Adorno wishes to accomplish here is not a complete interpretation of Max Weber’s methodology, but to highlight traces of a conjectural method in Weber’s work. Although Weber can by no means be construed as supporting physiognomy, his methodological ‘cornerstone’ – the concept of ideal type – implies abduction and figurative representation far more than is commonly recognised.

There are at least two major sociological research studies where Adorno puts Benjamin’s physiognomy (or constellation method) to work. One is Adorno’s research studies on music, one of which (On Jazz, in Adorno, 2002 [1936]) ¹⁴ is explicitly mentioned in the Introduction to the PD (Adorno, 1976a [1969]). The other is the famous study on The Authoritarian Personality (1950). In both cases he was able to translate the philosophical influence of his debate with Benjamin into empirical research terms. The innovative quantitative method of The Authoritarian Personality [AP] consisted of developing a set of questionnaire items (the famous F scale), which, instead of treating opinions as isolated data, recorded clusters of opinions identifying the presence of each of these elements. When the elements existed in correlation, they were considered to form a latent structural pattern of the potentially fascist personality. The idea that a cluster of elements, which on the surface appeared to be unrelated and irrational (in this case, responses to an opinion questionnaire), could be rearranged in various trial combinations (the final F scale was the product of many such arrangements) until they fell into a configuration with an inner logic that could be read as meaningful (here the structure of the authoritarian personality) fully paralleled the method of constructing constellations borrowed from Benjamin. Actually, AP can be seen as a socio-psychological representation of the ‘idea’ of fascism, whereas Benjamin tried to write das Passagenwerk as a representation of the idea of commodity fetishism and, before that, the Trauerspiel book as a representation of the idea of German Baroque Drama. At the same time, Adorno tried to avoid what he actually criticised in Benjamin, that is, the lack of theory: all the elements of the AP constellation were related to a general theory of anti-Semitism, and their interpretation was in all cases mediated by that theory.

Like Benjamin, who wanted to construct the ‘idea’ of Trauerspiel starting with the ‘extremes’ (which meant examples of such dramas that are not perfectly developed), Adorno also tried to find a means to correct and enrich the merely quantitative pole of the study. Certain subjects, particularly high- and low-scoring ‘extremes’, were chosen for in-depth interviews, and the task of interpreting the significance of the interview material was – not surprisingly – assumed by Adorno (see Adorno et al., 1950: 603–783). Adorno conducted these interviews ‘physiognomically’.

He focused on the contradictions and logical gaps in the interview responses, and his interpretations of these gaps, as well as the language imagery and even the body gestures of those interviewed, were used to verify the previous theoretical assumption.

The aim of the qualitative research in AP was to develop, through close scrutiny of the interview material, a phenomenology that both verified and give concreteness to the theory of anti-Semitism by concentrating on seemingly insignificant textual details. Adorno’s method of considering ‘unique’ and ‘extreme’ statements, rather than the more typical, quantifiable responses, gave his interpretative fantasy free reign at the same time that the consistency of the theoretical frame provided for exactness, ‘a safeguard against arbitrariness’ (Adorno, quoted in Buck-Morss, 1977: 183).

Conclusion

In this article, we have tried to address physiognomy both as a method and as an object. On one hand, we try to establish a physiognomy of the PD by focusing on apparently secondary details that can provide some insight into some of the neglected aspects of the dispute. On the other, physiognomy is viewed as that implicit practice that reveals some differences between Popper and Adorno, between critical theory and critical rationalism, regarding the concrete way of doing sociology and sociological research. Physiognomy evokes that ‘concept of unregimented experience’ that underlies the scientific practice that Schelsky saw as one of the central points of controversy between the two contenders (as mentioned by Adorno, 1976a [1969]: 57). Popper believes that deductivism and logical inference can help explain a fact or regularity among phenomena. A theory, a deductive system, enables explaining phenomena by connecting one fact logically with other facts. On the other hand, Adorno is sceptical of using formalistic logic and systematic methods because he sees in them the danger of bureaucratising sociological study and distancing it from the contradictions of social reality: ‘methods do not rest upon methodological ideals but rather upon reality’ (Adorno, 1976b [1969]: 109). In physiognomy and its abductive practice, he does not establish a new foundation for sociology, but a cognitive style that moves away from the linear thinking of deductive logic. Abduction implies ‘displacement’, moving from conventional to unconventional reality: ‘without the broken, the inauthentic, there can be no knowledge which might be more than an ordering repetition’ (Adorno, 1976a [1969]: 35).

Ginzburg went beyond Adorno’s conception, showing that physiognomy can be considered part of a paradigm called ‘conjectural’, which relies on a more flexible model of ‘rationality’ and scientific rigour. Physiognomy is one of those ‘mute forms of knowledge in the sense that their precepts do not lend themselves to being either formalised or spoken … No one learns to be a connoisseur or diagnostician by restricting himself to practicing only pre-existent rules’. The search for highly formalised rules of deductive logic can be of little or no interest to those who are involved in ‘forms of knowledge most linked to daily experience – or, more precisely, to all those situations in which the unique and indispensable nature of the data is decisive to the persons involved’ (Ginzburg, 1989: 125). The historian, as well as the sociologist, may be more interested in abduction as a form of practical reasoning. The physiognomic paradigm is hard to formalise and to find expression for – in different ways – in the sociology of Georg Simmel, the historical method of Carlo Ginzburg, and in Benjamin’s social physiognomy; it was practiced by Adorno himself in his empirical studies on Jazz and AP.

In summary, the following general conclusions can be drawn: physiognomy is a form of knowledge centred on the individuality of the phenomena; physiognomy focuses on marginal, secondary, even ‘eccentric’ phenomena that cannot be incorporated in a pre-existing categorical network or taxonomy. Even if they are ‘peripheral’, gaining insight into their make-up can shed some light upon ‘what counts as the core domain, but which often is not’ (Adorno, 1976b [1969]: 110); physiognomy is more a practice than a theory: it is bound to the concrete. Its rules are hard to formalise or enumerate; physiognomy is often figurative; it illustrates the truth by using examples and collecting phenomena, rearranging them into ‘constellations’, so that their relations illuminate previously unknown areas of reality (like Holmes’ method); physiognomy is related to the senses (sight, smell, taste) and experience, even if this experience is unlikely to yield a replicable experiment. After more than half a century after the PD, we can say that the topicality of Adorno’s sociological thinking lies more in his efforts to unify abstract theorising, object, theory and aesthetic representation, than in his explanatory analysis of capitalist society. In contrast to Benjamin, Adorno did not refuse theory (science and philosophy) as such, preferring art as the only way to reach the truth. Moreover, he uses art and aesthetics to arouse thought when it seems to have become conventional and aged.