In hope that scientific nomadism may turn out to be meaningful after all

Reflexive nomadism

Laying claim to reflexive nomadism and making coherent migration choices in selecting research targets is not tantamount to falling into the trap of the ‘biographical illusion’ (Bourdieu, 1986). The vast majority of my research has concerned five subject areas. The first of these is symbolic power: how cultural goods and habits can support or challenge visions of the social world. This was the core question of my PhD, completed in 1981 and published in 1985. It questioned the nature and forms of the presence of a political vision of the world in French spy fiction. Sales of these paperbacks were huge, in the 1950s and 1960s. Why were these authors, not all of whom were strongly committed to any political position, able to share a converging, conservative and sexist vision of the world? How could we explain the success of these books among working-class audiences, who usually were not readers of books, and were not politically conservative en masse? These questions – on the ideological meanings of culture, the role of languages and codes in carrying them, and the social conditions of their production and reception – can be found sprinkled through many of my texts, like a trail of breadcrumbs.

A first move, still in the realm of ‘communication’ and symbolic production, saw me focusing on the study of debate in the public space, (François and Neveu, 1999), and more specifically on the analysis of political journalism (Neveu, 1993, 2000; Kuhn and Neveu, 2002). Because it concerned journalists, Public Relations and politicians, this move was also an invitation to work on the (untranslatable) category of métier. In French, this word includes both the ideas of profession or role and the nature of the skills and dispositions needed to master a role. This research field has represented the main strand of my publications since 1989, though it has trailed off a little since the 2010s. A third research space was to develop from the late 1990s: social movements – which remains central to my work 25 years on (Neveu 2019 [1996]; Fillieule and Neveu, 2019). I had been craving these issues for a long time; this was linked to years of activism in leftist movements, where I was haunted by the question of why there are so few revolts and collective actions against a social order that I still look on as incredibly unjust. I was able to open up this research field when a great wave of mobilization in 1995 coincided with my arrival at Science-Po Rennes. There I was at last, at the age of 40, able to choose my teaching topics, after having been prisoner for years, at the law faculty, of teaching constitutional law, which was more something like a territory that political scientists had to occupy against their colleagues from public law than an intellectual passion. Though with a much smaller volume of publications, I was later to make a further move. As I’ll explain shortly, this could probably better be defined as working with and on ‘gender’ than being a card-carrying member of the Gender Studies’ club. Working with a colleague from my research team, the first French textbook on the sociology of gender was published (Guionnet and Neveu, 2004). Finally, and most recently, I have slipped into the sociology of public problems (Neveu, 2015) and the under-investigated question of their globalisation (Neveu and Surdez, 2020).

It would be misleading to claim that all my research could be neatly sorted into one of these five boxes. Academic life is also a story of unanticipated publication proposals and research team projects. The very first article I co-published, in 1984, studied Chirac’s Gaullist RPR party. The topic was far from my spontaneous interests, but did offer a young professor with a very limited publication list ² an opportunity to gain visibility. Nevertheless, the vast majority of my articles can be linked at least to one of these five themes. And although some focus more on the discussion of major authors or paradigms (Mattelart and Neveu, 2003; Bonny et al., 2003; Benson and Neveu, 2004, Neveu in Bernardin, 2022), there is always an empirical aim. How could I find the tools that would lead me to a better analysis of my research objects? These moves between objects and research fields are the result of a deliberate choice that I will try to justify. This choice is also the product of a habitus that includes a powerful libido sciendi – a lust for knowledge. This ‘will to know’, to parody Foucault, soon left me very reluctant to put up academia’s disciplinary jingoisms, which are two-faced. One face shows a lack of curiosity or openness to the contributions of other disciplines, while the other shows confinement within the hyper-specialization of a sub-discipline. There is no doubt that the level of internationalisation and methodological skills among the young political scientists of 2020 is often higher than it was for full professors 40 years ago. But the other side of the coin is often hyper-specialisation in a sub-field (public policy, local studies). Specialisation can thus operate as an invisible prison ³ , trapping the researcher in his small network or her niche journals. The logic of such microcosms is that reading even the less innovative paper of a member of the tribe matters more (we must keep up to date!) than revisiting and using the well-thumbed, dusty classics of social science.

A healthy academic rule of thumb should be that a researcher needs at least two kinds of research objects. To conclude on nomadism, it is worth saying that it can have a range of expressions. Like a game of hopscotch, my own experience of nomadism worked by moving between objects having contiguities or similarities. Other researchers – I might mention Olivier Fillieule or Sidney Tarrow – have had a pivotal object, in their case social movements – and from these social movements stem openings and explorations on issues as diverse as biographical trajectories and gender, policing and revolutions, political parties and cycles of contention.

A plea for ‘dirty’ objects

‘What you like is dirty objects!’ – this ironic but friendly comment was made by my colleague Bernard Pudal. I can’t deny that many of my research objects are not among those rated by French political science as canonical, or indeed even respectable. My MA dissertation was a research project on the ‘ideological meanings of pop music’. My PhD years were devoted to what was then termed – with a mix of recognition and condescension – as a ‘para-literature’ (Arnaud et al., 1970). At least three explanations underlie such choices; these mirror what Gobille (2008) termed the ‘calling of heterodoxy’, that is, the readiness to challenge taken-for-granted social norms that was so strong in the ‘68 generation – and strengthened in my own situation by a twofold investment in radical politics and critical social sciences. It can also be explained by my position as a complete outsider. I was a PhD student in a ‘provincial’ law faculty, when there was no such thing as ‘laboratory life’. Connections to the heavily Parisian, small world of political science were loose, at best. As in the ballad of the Cavalryman of Constanz Lake, riding boldly across the ice on the lake that he had mistaken for a road, I underestimated the riskiness of my object choices by comparing them to what was acceptable in literary studies, rather than in my adopted discipline. I should add that although political science was at the time strongly focused on classic objects (e.g. elections, government), it was also under-institutionalised, still questioning its frontiers and thus permeable to innovation.

Later, one of my few experiences of contract-funded research was a study of the Children’s Departments of the French television networks (Neveu, 1990). I was to maintain a lasting commitment to television, even co-writing a small textbook on its sociology (Le Grignou and Neveu, 2018). The editors of a classical reader on TV Studies (Allen and Hill, 2004) have written, tongue in cheek, that in the US in the 1980s, having a PhD in TV Studies was just ‘a bit above a PhD in fridge study’. It goes without saying that the weight of cultural legitimacy bestowed up TV Studies a lower status in French academia. My academic ‘criminal’ record shows that I’m a recidivist in terms of what French culture terms mauvais genres. I worked with Annie Collovald, in partnership with the Pompidou Centre Library, on intensive readers of detective novels (Collovald and Neveu, 2004). I also worked with her on the Guignols de l’info. ⁴ Most of our colleagues were watching this programme with delight, much as they watch TV series today. But although TV series are now taken seriously enough to quickly trigger a flow of publications, this was not the case with regard to this show, 20 years ago. The conclusion of one of the very few studies on the Guignols was even that they were expressing a threatening antipolitical cynicism (Tournier, 2005). My small collection of academic antiques and curiosities includes research on the memoirs of French politicians (Neveu, 1992), research on the reporting produced by ‘New’ and ‘New-New Journalism’ in the US (Neveu in Leroux and Neveu, 2017). I should mention the Goude parade organised on the Champs-Elysées in Paris in 1989 to celebrate the bicentenary of the French Revolution (Neveu, 2008), and studies on the local press. I should also mention objects that are now unanimously considered to belong to political sociology, but were, for a long time, repressed. A simple test would be to search the tables of contents of French political science journals ⁵ for studies on social movements during the 1960s-1980s. Their almost complete absence suggests the overpowering legitimism that then ruled. Voting and opinion polls, along with, voluntary associations to a lesser degree, were seen as the only reasonable expressions of vox populi.

Being suspicious of any fascination with research objects considered minor or strange can be a healthy reflex. I have more than once warned students that the issues they feel passionate about, or committed to, may not promise much in the way of exciting theoretical developments. Transforming a hobby or a passion into a research object is no guarantee of scientific results; it may be that the only output will be the attempt to seem extraordinary.

There are two yardsticks that could be used to evaluate the scientific potential of small or ‘dirty’ objects. The first of these is the reason behind their disqualification. Is it really impossible to use this object to gain theoretical knowledge and/or a better understanding of something significant about the social world? Or is it repressed by a normative vision of what is legitimate, serious or vulgar? In the French Academia, press studies focus, most of the time, on the ‘national’ (that is, Parisian) press and the news magazines. Yet 70 percent of the French daily press is made up of regional newspapers. However, in hyper-centralised France, regional means provincial – a synonym for unimportant. A variation of these a-priori disqualifications can be discerned in the certitude that some objects are so simple, so transparent that they are either self-explanatory or not worthy of reflection. Even Raymond Aron (1968), blinded by his loathing of the student revolt, dared write that it was only possible to make sense of Mai ‘68 through the Paretian ‘residues’ method: issues escaping any rational understanding. Lazy readings of Gramsci or Eco justify the idea that detective novels can only be a conservative literature, based on the restoration of order and the reader’s projection onto super-human characters. Our interviews of readers reveal the opposite results. These books are often read for their dimension of social criticism, with the heroes’ flaws being more attractive than their achievements. In the academic unconscious, there is room to question the enduring presence of visions of what ‘purity and danger’ (Douglas, 1966) really are. The use of the noble word ‘culture’ and the recognition of reflexive abilities often seem suspect when applied to those from below the upper crust. But when a researcher works on the cognitive tools used to make sense of politics and society, both everyday experiences and ‘profane’ cultures matter, and reveal their creativity (Goulet, 2010). Champagne (1990: 37–39) suggested a second yardstick, making a distinction between ‘empirical’ and, more importantly, ‘theoretical’ generalisations. There is no real ‘empirical’ generalisation to be hoped for from a case study of press coverage of somewhat violent farmers’ protests in Brittany (Neveu, 2002). Conducting the same kind of study in a different region of France would show that farmers there, are producing other goods and have different habits of collective action or weigh differently in the social structures. But this study allows a theoretical generalisation that other case-studies were to confirm: press coverage is like filo pastry in a regional daily that has tens of editions covering different regions. The farmers’ protest is reported in each and every local edition in which it develops, in the pages of the Finistère département, sometimes in the ‘regional’ news section, and in the ‘general news’, along with contributions from columnists, if it grows in importance. The coverage is both more intense and less critical in the ‘local’ or départemental news-sections closest to the action. It may become more distanced (or even critical) when written by columnists or journalists from the ‘social’ or political newsbeats, who are 200 miles away from the action. And if one compares the local/regional and national French press, it is also clear that a high level of mobilization (and of violence in particular) is needed for these protests to attract the attention of Parisian newspapers. Such facts are important for any researcher practising Protest Event Analysis. As it is visible in Figure 1, a huge number of protests stay under the radar of the ‘national’ press that generally supplies the data for comparative research. They suggest the extent to which the outputs of violence take the form of media coverage that is both more copious and (in all likelihood) more critical.

OPEN IN VIEWER

Figure 1.

Coverage of farmers protest (3–22 April 2000) by two regional dailies (Le Telegramme, Ouest-France), and two national dailies (Le Monde, Libération).

Analysis of the Guignols de l’info (Collovald and Neveu, 1996) is more than just the production of an insignificant (and soon obsolete) inventory of themes and characters in the flow of skits. Through theoretical generalisation, it is an invitation to explore the growing weight of a media culture that uses – and hijacks – ad spots, TV series and musical clips to produce narratives of politics and society. The analysis also suggests how humour develops a ‘systemic’ criticism of politicians. What gets targeted here – albeit unintentionally – is not so much politicians’ lack of virtue or honesty and more the fact that they are trapped in their interdependencies with the media – prisoners of a game whose rules they don’t control. Without jumping to the (mistaken) conclusion that the producers of the show are undercover sociologists, the programme invites us to question what a pleasant and stimulating popularization of certain results of sociological research on media-politics interdependencies might look like.

If I dare connect my research to that of greater scholars, might I add a reminder that some classics of social science have explored such matters as: poaching in the English forest, the circulation of satirical and pornographic pamphlets in the depths of the French countryside during the 18th century, and the cultural status of comic strips? Out of these ‘dirty’ or tiny research objects, Thompson (1975), Darnton (2014) and Boltanski (1975) produced theoretical analysis of major importance, making sense of the moral economy of the working classes, regime delegitimisation processes and cultural consecration. The question that remains is: ‘How to choose small objects capable of opening the door to theoretical discoveries?’ I can offer a partial check-list: Does the object of the study attract the attention or trigger the activity of a significant number of people? Do these practices or issues trigger a major investment of resources (time, money, attention)? Do they structure everyday lives as ‘community-building’ experiences? ⁶ Are the social facts that are the target of the investigation defined as a public problem requiring watchfulness or public action? Do these social facts have the aim (or objective effect) of challenging or changing social hierarchies and power balances?

To conclude on this point, it is important to bear in mind that although heretical choices of research objects can be highly scientifically productive and furnish the happiness of exploring new research fields, they also come at a price – and this may outstrip the small change or distinction benefits they can bring. Both because of its research object, and because it was inspired by the sociology of Bourdieu, L’idéologie dans le roman d’espionnage was probably the first book published by the respectable Presses de Sciences-Po not to be reviewed in the flagship Revue Française de Science Politique journal, where sectarianism (hiding in plain view behind impeccable manners) was at the top of its game, in the book reviewing business.

The rules and virtues of conceptual free trade

The internationalisation of scientific life is much more real for young colleagues in search of a job today than it was for my generation. In the 1970s, French social sciences academia was still parochial, confident in the knowledge that internationalisation meant revealing – to a planet dazzled by French culture – the contribution made by great French thinkers. Though there is still room for improvement in terms of internationalisation, ⁷ things have changed. French research is now immersed in a global academia with its anglophone journals, shared references and international conferences.

Converging reasons (including my personal Anglophilia, my quest for research capable of helping with the analysis of my object, and affinities for heterodox works) channelled me towards a growing investment in ‘British Cultural Studies’, whose impact was at the time very limited in France. Often linked to the social trajectories of their authors (Hoggart, 1957), these studies were, to use Willis’ phrase, an invitation to pay attention to profane cultures, and were inventing what Passeron termed a culturologie externe: an analysis of culture in its relationship to the perpetuation or undermining of social dominations. It was not so much the conceptual legacy of these studies (on hegemony or resistance, for example) as its methodologies that were illuminating. In what way? Through their invitation to look at the social world bottom-up, to consider the practices of everyday life and the opinions of the layperson to be worthy of analysis. Through their practical suggestions as to how the black box of media messages (narratives, songs, advertisings) could be prised open, and thus through their challenge to the invisible China wall that disconnected sociological analysis from research on signs and messages. Through their experiments, which shifted the question of reception from a space of theoretical hypothesis (Hall’s encoding-decoding model) to empirical work. Moreover, this move was also a moment of fruitful methodological bricolage, as television viewers were asked, via a TV listings magazine, to write about why they watched Dallas (Ang, 1985), combining interviews with focus groups (Brunsdon and Morley, 1999 ; Katz and Liebes, 1993).

Gaining a deeper understanding of a research flow and considering it important creates something akin to a duty to disseminate. Such was the goal of a trend analysis article I wrote with Armand Mattelard that soon became a small book. Armand’s reflections on culture had been an inspiration for me ever since his years in Chile with the Unidad Popular (Mattelart, 1974). We were careful to prevent the oversimplification that can happen as a result of jumping to conclusions, which tends to disconnect authors and concepts from their social and intellectual contexts of production. Our plan was also to supply the broadest possible ‘group photo’ that would include all the major contributors, the different generations of researchers, the dynamics of intellectual moves. The idea was to write a book that would allow its readers to gain access to a social history and conceptual synthesis of these studies while allowing us to take a critical stand concerning for instance the theoricist drift of some of these studies, or their growing distance from empirical work. The import of theories can be assessed by what they provide, and probably also by what they prevent. We are not the owners of British Cultural Studies in France. Their growing impact is good news. If other researchers are making different choices concerning the ‘must read’ or ‘best’ conceptual tools, well – such is the life of free research. But the role of a good ‘Introduction to…’ is also to suggest that certain uses of theories go beyond inventive interpretations and become hijacking. The aim of the young generation in Birmingham was to understand culture, rather than to be radical enough to terrify Oxbridge dons (Bourcier, 2003). The contribution made by historians such as EP Thompson to the rise of Cultural Studies cannot be brushed under the carpet. And since these studies created an ‘anti-disciplinary’ academic space, their aim never was the ‘rather stupid’ (Morley, 1992) idea of endlessly multiplying sub-disciplines and micro-studies (Glévarec et al., 2008).

After the initial historian-driven impulse, my work with sociologists Bonny and de Queiroz (Bonny et al., 2003) has also contributed to a better understanding of Norbert Elias’ processual sociology. By inviting those who call themselves Nobert Elias Flying Circus to Rennes for a conference, we helped amplify contributions discussing the dynamics and limits of the ‘process of civilisation’. I would like to focus on one of the most inventive of the ‘Eliasians’. In a friendly discussion with Elias, Wouters deepened the reflection on the dynamics of civilisation. One issue raised was the question of ‘self-controls’. Was it possible to describe these ‘civilised’ controls on emotions and drives as being very powerful, when so many facts were suggestive of their loosening: changes in sexual behaviours, the rise of hedonistic behaviours, celebrations of the body and of authenticity. Could such facts fit the vision of ubiquitous control of our drives? Wouters (2007) develops a notion of informalisation as ‘controlled decontrolling of emotional controls’. The self-control of the civilised is so deeply embodied that it becomes second nature. This opens the doors to a new stage – access to a ‘third nature’ where emotions and desire are more ‘present’ to conscience, where conscience governs those drives in a more reflexive way. In a nutshell, this analytical frame offers a response to naïve objections to the Eliasian vision of civilisation while also supplying a powerful tool with which to explore manners, emotions and morality. I tried to extend the explanatory power of the concept to culture. The retreat of what Bourdieu termed ‘cultural good will’ is visible in a shrinking deference to yesterday’s legitimate culture. It can also be seen in a more relaxed, less ascetic relationship to cultural goods – even when they are used to gain knowledge. Detective novels allowed us to speak of an informalisation of cultural consumption (Neveu and Collovald, in Coulangeon and Duval, 2014). The notion suggests a de-sacralization of many works of ‘classical’ culture as well as a more disinhibited and sometimes ironic relation to cultural practices, which must be entertaining and pleasant. But de-sacralization is not disinvestment. Conversely, a significant proportion of readers of detective novels claim to find in them useful political criticism of societies, sociological readings of the world, even a metaphysical understanding of human frailty. My experiment with concept-stretching triggered sceptical feed-back from Wouters, who didn’t see his informalisation in it. I would have preferred praise. But I continue to consider my use of his work to have been fruitful, based on a reasoned analogy. Concepts are not made to sleep within the pages of a textbook or be polished by endless comments. They are made to work, to be adapted, even tinkered with. As long as adaptations do not hollow out a notion’s original meaning, as long as new understanding of social facts is opened up, working on them (as opposed to just with them) is allowable.

The international dissemination of theories cannot be a one-way process. A French researcher must also publish in other languages and build the impact of the best research achieved by his or her own linguistic community. I have tried to do this – by publishing in English and by explaining to an Italian readership why French research on culture had taken paths different to those of British Cultural Studies (Neveu, 2011), for instance. But my most significant contribution is certainly the book I edited with Rod Benson, Bourdieu and the Journalistic Field (2004). It sold well, and was translated to Chinese. It became a core reference for scholars working on journalism. Saying that it balanced certain hasty receptions of On Television (Bourdieu, 1998) would indicate a lack of modesty. But we made it clear that Bourdieu could bring much more to a sociology of media and journalism than in these short papers. Our book has provided a resource to colleagues wishing to import Bourdieu’s tool kit in journalism studies. But its effect was also – and it matters – to prevent the publication of bullshit (Frankfurt, 2005). This might seem strongly worded, but there have been some astonishing re-interpretations of field theory – a ‘good’ example being, in a Cultural Studies reader, the suggestion that a family could be an excellent illustration of a field. Sans blague! Another reason our book has been useful is because its construction combined theoretical articles, case studies mobilizing concepts and contributions contextualising Bourdieu’s reflection on culture and media.

I have suggested that his statement ‘Theory travels better’(Morley, 1992: 2) should be coined ‘Morley’s law’. The suggestion is not that theory travels well, but that in the international diffusion of social science, the more a piece of research is conceptual (or should we say ‘abstract’ or ‘disembodied’?), the better it moves. One obvious reason for this comes from the fact that the empirical flesh that gives birth to concepts is way beyond reach, far from the foreign reader’s experience. Who among the readers of British Cultural Studies has ever played in a penny arcade or watched Nationwide on TV? Who in Pittsburgh or Seoul has a realistic idea of what is meant by the French State Nobility that inspired Bourdieu’s eponymous book? No text should be imported bereft of its social and academic production contexts. Loïc Wacquant’s ‘Bourdieu in America’ (in Calhoun et al., 1993) offers an illuminating case study of such processes at work in selective, biased or absurd understandings of the French sociologist in the US.

The benefits of cumulativity

Nomadism is fruitful only if at least some of the knowledge gained can be invested in the exploration of fresh territory. My first research on gender issues was about women journalists and possible ‘feminine’ ways of covering the news (Neveu, 2000). Because it combined the tools of Gender Studies with the sociology of journalistic work, it remains pertinent 20 years later. This combination allowed me to question how and why certain kinds of papers or ‘angles’ (more practical, more empathetic towards the people portrayed) could be explained by something like a ‘feminine habitus’. It also suggested that such gendered reading could be balanced by considering the fact that female journalists were also often at the bottom of hierarchies. On the health news-beat, an international conference on COVID-19 is most likely to be covered by a man. An article on a more ‘practical’, but also relatively lightweight topic such as the risks of allergy to spring pollens, or the need to be vaccinated against flu would be more likely to fall to a woman. But laying claim to specific ‘feminine’ skills or styles, whether real or imagined, could also be put to work, with editors desperately in search of ways to win back a vanishing readership. This example allows another explanation. Gender Studies has, for several decades now, been a major space for scientific innovation and the destruction of sexist prejudice. It also stands as an example of the risks of the self-sufficient insularity of ‘studies’ communities. Gender studies faces the combined risks of campus radicalism, theoricism and hyper-specialisation. Knowledge of gender relations, now firmly established and institutionalised, would probably gain in terms of both influence and productivity if it were a compulsory component of a gendered sociology, rather than the property of a small, gatekept community.

Illustrations of the benefits of knowledge cross-fertilisation can be multiplied. Having a knowledge of the logic of media and news production is helpful when exploring how they cover social movements. This coverage is often simplistic. But if sociology has other ambitions than parroting the less sophisticated deploration of activists, it could offer a better understanding of these biases, which are rarely the straightforward result of a ‘line’ emanating from either the owners or the political views of editors in chief. From the social division of work between ‘seated journalists’, suggesting frames and reporting angles from the newsroom, and on-the-spot reporters, to the weight of official primary definers (the police, for example) and the social distance between journalists and working-class protestors – these are enough to produce coverage biases. Mobilising the cumulativity of knowledge also means strengthening this practical sense in a sociologist’s habitus, expressed in the intuition of an analogy, of the possible shift of a research scheme from another research topic. A student (now a colleague) had kindly written in the acknowledgements for his PhD that some of the most useful advice I gave him as supervisor began with ‘It probably bears no relation to your research, but I believe you should read…’. Indeed! Isn’t Daniel Gaxie’s classic book (1978) on politicization and political competence partly inspired by L’amour de l’art in which Bourdieu (1966) studies the social logics of museum attendance? And in the same way, Muriel Darmon’s illuminating study on anorexia (2008) was inspired by Becker’s studies of deviance.

Another output of nomadism is that it renders the study of some topics much easier. While there’s nothing to stop young researchers from bringing innovative contributions to the study of public problems (Comby, 2015), it is much easier to piece together the variables and issues of a huge intellectual jigsaw puzzle when your many years of experience reading across a range of subjects is combined with both knowledge and know-how ⁸ . And trying to expand this approach to the international circulation of claims, frames and problems (Neveu and Surdez, 2020) soon reveals just how many more research flows or partnerships (International Relations, Policy transfers) need to be mobilized.

Building the sociological panopticon

One of the unaccomplished ambitions of my research on cultural goods has been to produce a global or panoptic understanding. To make sense of this dream as simply as possible, a complete sociological analysis of a cultural good (TV series, literary genre, press) demands three approaches.

First, a space of production must be mapped, and here, the notion of ‘field’ is especially illuminating; it invites us to take a historical approach, to follow the border-building process. Thinking with field also means developing a relational vision of producers and their competing visions of what constitutes an accomplishment. One trick of the trade that I picked up during my fieldwork on children’s television is that questioning producers on their competitors often results in useful material. Whether careful or peremptory, words expressing differences, hierarchies or disgust are meaningful. ⁹ Field analysis is also a matter of mapping the articulations between strategies aimed at the typification and legitimisation of productions as creative, and strategies aimed at maximizing their market impact. In terms of ranking symbolic capital, indicators include number of prizes (Goncourt, Booker, Pulitzer), volume and depth of catalogue, and ranking of producers and writers. Market indicators matter too: cash-flow, market share. Recruitment strategies can be used to render the relationship between managers and producers objective. Is their purpose to build a publishing house – like Rivages for the ‘noir’ novels – in which to incubate promising talents, or to behave as a launch pad for best-sellers bought with the highest bids on translation rights? At Fleuve noir (a leading publisher of spy-novels for the working classes in the 1950s and 1960s), signing a first contract didn’t necessarily mean being published. A second manuscript had to be produced immediately, as proof of ability to sustain a certain rhythm of production – often a serious hurdle to clear, for more literary accomplishments ¹⁰ . Method in action also demands a sensitivity to the clues that fieldwork silently supplies. Awaiting the return of my questionnaires during my research on spy fiction authors, it took me a while to realise that a surprising number of their letters were postmarked in the 16th arrondissement of Paris. A technical manifestation of residency in the wealthiest part of Paris – but also a sociological invitation to think about the peculiar situations of these writers as cultural producers living in a bourgeois space, far from the intellectual Rive gauche.

Thinking sociologically about cultural goods also means investigating their reception. Who is watching, who is reading? What are the diffusion networks? Studying the French spy fiction of the 1950s–1970s, the working-class profile of the readership was evidenced by the fact that they were sold en masse in railway-stations, tobacconists and department stores, whereas bookshop sales were marginal. Mediating institutions, whose aim is to intensify contact with audiences, are worthy of attention. These might take the form of festivals, online chat rooms, prize ceremonies, niche programmes on radio and television or book signings, all of which produce a dense amplification process for detective or graphic novels (Nocerino, 2020). Exploring reception also means paying attention to the wayward gratifications that make the uses of cultural goods so unpredictable, so close to de Certeau’s ‘poaching’ tactics. Thinking back to certain encounters with readers, a shared interest in the detective novel was like an umbrella covering a complex ‘colour chart’ of uses. A retired railway worker enjoyed reading about the adventures of a woman detective, investigating in the small towns of Brittany that he knew well. A professional working in import-export was reading Chinese novels, by Qiu Xialong, to get a better feel for China. But he was also using these readings to build cultural resources he could use both with his partner and in his social life. A student, facing several deaths and dramas among her family and friends, discovered in the reading of Ellroy an almost metaphysical salvation literature that explored experiences of darkness. I should say that becoming aware of the surprising complexity and breadth of activities capable of being defined by categories as simple as ‘reading detective novels in the 2000s’ or ‘being a leftist activist in the 60s’ (Neveu, in Fillieule and Neveu, 2019) left me dizzy. I feel it when I hear of comparative studies on multiples of countries or cases – mostly from data not even produced by the researcher. I am not suggesting that such research is pointless, but these data demand an extraordinary level of reflexivity and prudence, which is often skipped with a careless claim that ‘data are robust’. Perhaps. But robust enough to highlight all the complexity of social facts?

It should be obvious that cultural products are the link between the field of cultural production and the market. Yet the knee-jerk reaction of sociologists is often to consider that any ‘internal’ approach to cultural productions is the job of other tribes: semiologists, literature specialists or art historians, for example. So, it’s okay for social scientists to make a socio-history of cultural goods and identify themes and issues – but not to open the black box of structure and rhetoric! This would be like going over to the dark side – a space in which the price of deciphering signs is the disappearance of the social. Bourdieu’s work on symbolic power could even be brought into play – in particular, his remark on the pointlessness of ‘desperately searching in language what is anchored in social relationships’ (1982:14). The conclusion seems crystal clear. Paying attention to the forms and ‘internal’ peculiarities of cultural goods is not very useful when making sociology. ¹¹ I have shown in my study of spy fiction how deeply flawed such a position is. What is ‘formal’ (symbols, narrative structures, rhetoric, etc.) has a power of condensation, amplification and representation of a common sense. Because they trigger emotions and shared representations, forms embody the silent expression of ideas and evaluations which – expressed in a more explicit, more visible way – would appear as taking issue, as political statements. In one of his novels, bestselling spy fiction author de Villiers describes Russian women as vulgar and not very sexy on the grounds that ‘carrying manhole covers around Red Square all day long’, their shoulders and arms are like those of heavyweight weightlifters. Such tongue in cheek comment, backed up by a flow of images, remarks, and stereotyped characters call up a common sense, and the lesson here is that a dogmatic view of sex equality destroys the grace and powers of seduction that nature gave to women. The Danish linguist Hjelmslev coined the phrase ‘form of content’ to highlight how images and texts, both by their forms and by their explicit content, ¹² were suggesting reception programmes. These programmes can structure receptions according to both their encoding logics and the decoding resources of audiences (Hall, 1980). Any sociologist thus has everything to gain by having some basic knowledge of rhetoric – reading a bit of Barthes (2015) could fit the bill – as well as of the basics of Saussurian linguistics (inspired by Durkheim) with its central opposition between language (social) and speech (individual). In suggesting this I seek neither to annex semiology to sociology nor to consider researchers whose blindness to cultural and social differences is often striking as sociologists. ¹³ Borrowing (or poaching) semiological concepts is the price and imperative of Cultural Studies, which is inter-disciplinary by nature. However, crossing disciplinary borders also constitutes an invitation to revisit one’s own concepts. Isn’t it possible to suspect that, in some of their uses, the notions of frame and framing have become conceptual sponges, absorbing too many processes and meanings? A reflexive approach to concepts can be sharpened through an exploration of the legacy of ‘ancient’ rhetoric. This supplies a broad range of concepts and distinctions with which to study the processes and objects that routinized uses of frames tend to compact and confuse (Neveu, 2019 [1996]: 104–105). And lastly, mobilizing ‘internal’ analysis in the service of sociological analysis – isn’t this exactly what Bourdieu practices under the banner of ‘discourse analysis’ (1982: 162-239), when disassembling the rhetoric of symbolic intimidation used by Althusserian Marxists and Heidegger? To play the quotation game:

The formal peculiarities of works only reveal their meaning if one connects them on the one hand to the social conditions of their production […] and on the other, optionally to the successive markets in which they were received (1982: 165).

As for the confession – it concerns the few empirical successes of this analytical framework, which I have probably never managed to apply in a fully satisfying way. My study of spy fiction is precise as to the field of production, it is a deep dive into the study of these novels and their rhetoric. Yet it remains superficial in terms of the reception of these narratives. I could argue that such investigation would have taken more time, or that reception studies were only just emerging in 1980, but the flaw remains. In a strange symmetry, our work on detective novels is rich, in terms of the uses, gratifications and appropriations of these novels. It develops a precise field analysis of the publishing space. But it does not really supply an in-depth narratological exploration of the different kinds of novels. The research programme of L’idéologie dans le roman d’espionnage has found few echoes and little comparable research in sociology and political science (although see Thiesse, 2017). I might add that Cultural Studies monographs articulating these three dimensions are also rare (Radway, 1991; Brunsdon and Morley, 1999). How can what looks like failure be explained? Part of the answer comes from the cost of investment in readings and concepts, the multiple dimensions of investigation. ¹⁴ Allowing this approach its full expression demands the organisation of multi-disciplinary research teams. And the reality of academia, beyond the compulsory celebration of inter-disciplinarity, is that established disciplines remain strongly resistant to hybridisation. In French academia this failure is also linked to a moment (maybe a mystery) in intellectual history: the almost complete disappearance, in the space of just a few years, of the influence and creativity of the structuralist school heralded by Barthes, Eco, and Genette (Genette, 1966, 1969, 1972).

It is still worth trying to promote these analytical frames. Beyond the contributions made by these monographs to how ideologies circulate, how readers use (and poach) the contents of novels, these studies have also contributed to a general sociology of the circulation of political ideas. They have provided keys to understanding the modus operandi of narratives, their possible influences. But they have also made clear that the study of ‘political ideas’ could not be contained by the study of theoretical and philosophical texts, nor even by the language of politics and politicians. Other texts, other cultural goods, furnish the political and ideological nourishment of most citizens. A ‘social history of ideas’ (Matonti, 2012), questioning their production and reception, widening the definition of ideas worthy of academic attention, is becoming institutionalised in French academia. The study of such objects as spy and detective fiction and TV flow was part of this ground breaking approach.

The art of induction

All research is a combination of inductive and deductive approaches; both are useful. Deductive approaches start from the choice of a system of hypotheses, based on a theoretical framework. These hypotheses are put to test on an empirical material – often quantitative data. These data can be very fruitful in terms of understanding the importance of a variable or debunking prejudice. These methods do, however have some limitations that are worthy of mention. The first of these arises out of the fact that the construction of hypotheses on the first page of a research article presupposes that the right questions have already been identified. No doubt: science is cumulative and we know a great deal about a great many social facts. Still, the old Durkheimian and Bachelardian caveats, which suggest that we are often richer in prejudice than we are in sound knowledge, remain true. Another issue can arise out of the fact that testing hypotheses on data (often produced by someone else) demands both strong reflexivity and an aim of producing cumulative results. Such imperatives are not always respected. To give an example, the collections of the Journal of Communication (at least prior to Silvio Waisboard’s editorship) on the canonical issue of the effects of television provide a perfect sample of what Charles Wright Mills (1959) called the ‘grinding wheel’ effect. Numerous research articles, using sophisticated quantitative data treatment, focus on variables such as race, education, gender, income level or political orientation, to evaluate both how viewing habits are influenced by such factors, and how these uses of TV can produce effects (on learning performance, feelings of unsafety, visions of nations or professions). But – and here is the grinding wheel’s effect – making use of these results is like grabbing handfuls of sand. The knowledge they produce can indeed be very precise, yet it is often also terribly ‘local’. It supplies a patchwork of micro-explanations which could scarcely become cumulative: the more they tell us about the detail, the less they allow us to see the big picture.

Research demands deductive approaches. Yet my experience is that researcher should embrace the kind of ignorance that has more to do with an availability to follow unsign posted paths, a willingness to pay attention to unanticipated facts when the fieldwork reveals them, rather than a conceptual lack of culture. Sociological imagination is more often triggered by a puzzle than by a hypothesis. Until I was at least fifty-something, I faced a moment of panic each time I started a research project requiring significant fieldwork, a feeling of ‘I’ll never manage it’. When I did escape the blank page it was always by both drawing on a theoretical culture (being deductive) and by allowing my objects to ‘trouble’ me. I wanted to borrow the notion of the psychoanalyst’s ‘floating attention’, but I feared that misinterpretation of these words might result in the loss of the all-important oxymoronic aspect of the concept, with the adjective erasing the noun. Working with Annie Collovald on the reasons for the success of detective novels, ¹⁵ we launched a programme of interviews with passionate readers. We collected life stories, trying to understand the links between social trajectories and reading habits. I was more attached than my colleague was to a hypothesis that something like elective affinities, possibly even structural homologies, would emerge (e.g. women reading more historical detective novels, men more ‘hard-boiled’ style narratives). Nothing like this was obvious. But in discussing our first interviews we were both struck by something neither of us had anticipated. The moments of discovery, intensification (more rarely withdrawal) from reading detective novel were very often connected to biographical turning points. To quote one secondary teacher, it was a literature that became meaningful when ‘the earth starts shaking beneath your feet’ (divorce, unemployment, a move, or the death of a significant other). This correlation, present in many life stories, offered us an important interpretive key. We developed the concept of ‘escapism into the real’ – an investment in reading as an activity at once entertaining, realist and capable of boosting critical reflexivity. Paradoxically, the empirical failure of the starting hypotheses can often serve as a resource, compelling us to pay attention to what was visible but unseen. One of my major research projects, which also used life narratives, was about the activists of ‘68. As many books on 1968 in France suggest, with the enormous weight of radical (and perhaps threatening) ideologies on activists’ commitments and their expression, some questions were about the reading of the classics of Marxism, or of organisations’ leaflets. But most activists were not intensive readers of such literature, with more than one even suggesting that the working sessions on these texts were a bore. One interviewee recalled the happiness expressed by two newly-arrived young activists, their feeling of having made the right choice by signing up to a warm and action-oriented ‘anarchist’ group. They had however belonged, for several months, to a Maoist group! Findings such as these invited me to ask new questions about what it is that drives activist engagement – the power of books, the dark fascination of ideologies? Motivations such as these are probably more heartfelt among academics. Here the feelings of acting against injustice and staying true to working-class origins and the gratifications of warm sociabilities mattered more than words ending in -ism. During my research on the writers of spy fiction one puzzle entailed understanding why almost all of them shared strong conservative prejudices and beliefs. Real ideological commitment rarely seemed to be the right explanation, but gradually it appeared that the codes and templates of this kind of narrative, the feeling of being considered second (or third) rate writers by the ‘intellectuals’ and trajectories that included enforcement roles in the armed or police forces held more explanatory power than conscious ideological beliefs. One virtue of paying attention to what fieldwork reveals or suggests is that it prevents intellectualism. My point here is neither to decry conceptual ambition nor to underestimate the reflexivity of social agents; the warning targets the objectivist trap. It is wise to hesitate before defining the rationalisations or theories produced by the researcher, or borrowed from half-baked academic common sense as the cause of and motivation behind actions.

To plead for this attention, simultaneously floating and intense, is also to suggest the power of serendipity emerging from fieldwork. ¹⁶ There is no methodological spontaneism in this remark. The richer the researcher’s conceptual toolkit, the better their use of the clues offered up by fieldwork. Going to the INAthèque ¹⁷ to view the images of the Goode parade, organised in Paris on July 14th 1989 to celebrate the bicentenary of the French Revolution, I started watching this celebration on public TV network France 2. I could not say, 20 years later, why I had also asked to see the images of leading private TV network TF1, since both networks had broadcast the exact same images. But it immediately became clear that the same images were not speaking of the same show. The parade, staged by a famous advertising agent, was closer to the style of advertising clips than to any kind of historical reconstitution. Its framing of 1789 was as a moment of universal fraternity, symbolised in the show by world music; almost every dimension of social conflict had been edited out. On TF1, elderly anchor-man Zitrone – himself an almost ‘historical’ feature of 1960s TV, plucked out of retirement for the event – was spelling out a very factual and explanatory reading of the meaning of the different groups and flows. He was also visibly sceptical and on edge, having to make sense of what he also described as an opaque ‘disorder’, so far removed from the common sense or history-readers’ understanding of 1789. On France 2, left-wing journalist Serillon and TV producer Mitterrand had paired up and were making much more enthusiastic comments, praising the modernity of the event. They were in tune with the choice of a celebration centred on universalism and the bridging power of world music, far from the dusty revolutionary mythologies. Two very different understandings of the parade were thus visible. This observation also triggered the question of the homologies between reporter perceptions and the interpretive resources of the audiences of the two channels – very different in terms of education, jobs and age.

One simple trick combining the resources of induction and deduction (regardless of which data is being used) might be to investigate their logic of production. Analysis of press coverage of a farmers’ protest can result in useful figures, graphic curves on levels of newspaper coverage, data on the variation of framing, depending on news-beat. But a short trip to the scene, to meet the local reporters will supply fresh analytical tracks concerning their caution, their almost conniving coverage of violence. They can’t risk losing their union leader sources, and they can’t blame a movement that is enjoying strong local support. A report perceived as critical could mean facing physical threat from protestors.