Renewals in science studies and history of science

At first glance, the history of science might appear to be in a difficult position: its subject – ‘science’ – is constantly being deconstructed and redefined or, more accurately, dissolved. The initial strength of social studies of knowledge was that it questioned the notion itself, demonstrating that it was useful to abandon it as a general category, that it was heuristic no longer to think about ‘science’ but about ‘the sciences’ and ‘science practices’. Social studies of knowledge led us to consider the endless overlapping of sciences with other social practices, with economic, political and military worlds. We have learned that ‘science’ is not a sufficiently differentiated and stable entity to be considered in itself, that it is a set of ‘doings’ that are material, social and discursive, and that are inserted into worlds without any given limits.Footnote ⁴ Whilst we can, of course, define the characteristics proper to certain subsets – the physical sciences of the Cold War, for example – the determination to declare the unity of sciences, which was so strong fifty years ago, appears to have lost justification and utility. Nowadays we study fields of action without any a priori autonomy – and we are therefore seeing the disappearance of the closed object that ‘science’ once was.Footnote ⁵ Paradoxically this means greater visibility of scientific activities and their appropriation by other fields. If ‘science’ is no more, this does not mean that scientific activities are not gaining strength beyond science studies – in transnational studies, for example.

More precisely, science studies have taught us that science is of a human order, that it is created by men and women made of flesh and blood and not by God or the gods. The evidence that science produces cannot be absolute; it is based on simplifications and on initial framings – even if it produces results that are effective and interesting. Of course, this formulation must not be taken as a ‘relativist’ proposal – rather its aim is to state the human and social nature of all knowledge, and to draw relevant conclusions.

Science studies have also taught us that science is not ‘spurred on’ by its own logics. The issues they examine, the ways they ask their questions and address them, depend on considerations which vary not only in accordance with epistemological choices (not all sciences use the same standards of proof), but also in accordance with institutional situations, depending on where the knowledge is produced, on the nature of the backers and their intentions, on the networks in which the sciences are put to work, and so on.

A third lesson was that modern science is not only a mode of knowledge but also an activity with a practical vocation. Through controlled experimentation; systematic observation; and the use of numeric correlations, mathematics and modelling, this form of knowledge makes it possible to master phenomena, nature and things, mankind and social matters – an aim that previous forms of knowledge had not often possessed, or necessarily desired. To put it another way, the ‘modern science’ form is a way of addressing issues that allow more effective action in the world.Footnote ⁶

This operational capacity has often led to science being linked to powers of all types. It has always been closely monitored by the powers that be, with scholars, practitioners of the arts and engineers offering their services to kings, to democratic and authoritarian states, to business and military leaders. Through statistics, scientific management, social engineering and operational research, modern science also quickly became a way to govern. Historically, it has always been a way of helping those in power to have a better understanding of the world and thus to better manage both nature and society.Footnote ⁷

With science studies, modern science is no longer seen as a neutral vehicle of progress. Like any other area of human study, it is founded on values and categories that impregnate its language, its tools and its results. What science produces is part of the world and of human and social concern. Because of its organic link with technical and economic development, modern science has never been universally recognized as a good in its own right, and it has often been contested and criticized. Modern science produces results and techniques that help answer the questions asked by individuals and societies, but the solutions it provides can also be seen as the source of the very problems that are encountered.

Finally, we have learned that modern science has never worried too much about the consequences of constantly opening Pandora's box. Scientific and technical institutions produce results (texts as technologies) that are made available through publications and markets. Products and knowledge penetrate societies via consumption and usages, but without social, moral or environmental implications having been taken into account from the outset. In other words, science and technologies must be considered as coming at a cost, as fostering certain ways to the detriment of others – as having their own politics. So we should not be surprised by the reaction of the social realm or by the questions it asks, and has been asking for almost two and a half centuries, about techno-industrial science.Footnote ⁸

Renewal in transnational history

Phenomena of the same order and magnitude are at work in transnational studies. Like ‘science’, the ‘nation’ also lost its aura of obviousness. In one sense, one might say that this point is not really new. Study of the nation state has always presupposed exchanges, trade and international relations, war, armed peace and diplomacy, supra-national agreements and organizations – from scientific congresses to the League of Nations and the Red Cross. It has always presupposed links with the worlds of production, with metrology and standards – a standardization that science has always valorized. To contest newness, one can also call upon the legacy of subaltern, cultural and feminist/gender studies.

That would be a little too unfair, however, and the constitution of the transnational sphere as an explicit target created a significant shift. It drew attention away from mere ‘international’ issues (in the sense of international relations) towards the flows themselves, towards the circulation of people and artefacts, towards intermediaries and tangled relationships. It focuses more explicitly on what is emerging, what is new in the interstices of encounter – and less on the shifts first initiated by the centres of calculation (to use Bruno Latour's famous formulation) based in the global North. It focuses more easily on the fringes, on the ‘peripheral’ spaces, on the professions and groups who live off encounter and negotiation, who transform worlds through the connections they build, who have the possibility of playing a new role in these diverse scenes.Footnote ⁹ It therefore makes it possible to return to, and contest, the unidirectional vision of the manufacture of worlds – to revisit, for example, the process of colonization.

Transnational studies have also forced us – even if they were not alone – to play more consciously with scales, from the most micro (human agency) to the most macro (systemic forces).Footnote ¹⁰ They invite us to move more systematically from eco-systemic constraints to ‘material survival mechanisms’, from ‘demographic and social reproduction systems’ to ‘social and gender relations, cultural patterns’ and ‘political command systems’.Footnote ¹¹ In terms of research programmes, that means a large array of possibilities. It means studying how globality is built in very local settings (in Calcutta around 1800, for example), how global networks of consumption are built and lead to new knowledge and products – but also, on another scale, how ‘continents’ evolve; how ‘empires’ develop, survive and die; how macro-forces like geography or ecological systems shape trajectories of groups, regions and people.Footnote ¹²

This frame could also lead to astute analyses of the para- or transnational institutions which act on a global scale from the outset. This was what historians of the early modern period often did with regard to the Jesuits or trade companies.Footnote ¹³ To take contemporary examples and quickly present the work we are doing in Paris and Strasbourg, this leads to the long-term analysis of international expert committees (on radioactivity or pesticides), authorities that do not emanate from the state, but whose recommendations weigh upon economy, societies and states' sovereign powers. The same remarks apply to studies of ‘global’ experts, examining their tools, the way in which they construct their autonomy away from national powers, and how they came to standardize the lives of entire populations through the transfers of persons, categories and dispositifs. Regarding risks, for example, assessment tools and notions (effets sans seuil, risk factors, cost–benefit analyses, monetization of all types of prejudice) came into circulation among nuclear engineers, clinicians, toxicologists and the RAND Corporation, before being appropriated by economists, becoming ‘universally reliable’ and making it possible for ‘risk societies’ to become the reality of our modern and postmodern times.Footnote ¹⁴

This period also saw the appearance of a variety of intersections between scientific expertise, social critique, political demands and regulations, between state authorities, experts from all disciplines and local and global NGOs. These arrangements vary greatly, from the drafting of standards or lists of chemicals and food additives by European or American experts to climate negotiation. The latter is torn apart between the output of the various IPCC committees, debate in the public arena, and the dynamics of the political parties' annual conferences that oscillate between ‘core’ and ‘fringe’ – between official negotiation and the bubbling activity of the side events.Footnote ¹⁵

In Paris, we also concentrate on the organizations that try to directly manage ‘globalization’, institutions that set themselves the task of defining the norms and standards of the global world. This is the case of the OECD, which could be described as the super-think tank of liberalization and which determines the universal benchmarks that are required for business, companies and societies to operate properly. This is the case of the World Bank, which since the year 2000 has defined itself as the Bank of Knowledge that will eliminate poverty by promoting sound knowledge and good governance. This is the case of the WTO and its jurisdictions, an institution that claims to be the sole organization competent to regulate international business and trade – and that, in so doing, decides what ‘sound science’ is. But it is also the case of regional political bodies, such as the European Community, institutions to which the state delegates considerable powers and which actively redefine economies, territories and populations. The European Community acts as a regulatory state that is no longer sovereign, a state whose main aim is to create homogeneous ‘technological zones’ and which in this manner defines new modes of living together. By adopting norms, by introducing measurement mechanisms, assessment tools and competition procedures, it transforms social relations – working relations, for example.Footnote ¹⁶

We have finally become most interested in the recent explosion of the global forms of government promoted by economic actors. They consist of ‘volunteering’ (for example, committing to social and environmental responsibility) and the constant creation of soft law; in ‘round tables’, ‘multi-actor arrangements’ most often under the control of a few large multinational firms, like around palm oil; in ‘public–private partnerships’, for example, in world public health and the case of the Bill & Melinda Gates Foundation superseding older forms like the Rockefeller Foundation; in drafting norms (best available techniques, environmental impact assessments) for establishing polluting industrial complexes, norms certified by institutions like the EU; in ‘labels’ and the constitution of self-standardizing domains (for example, organic or fair-trade agriculture); and, of course, in états de fait and the rampant breaching of law.Footnote ¹⁷

The debates as epistemological debates

Debates first take the form of debates on the epistemology of social sciences, and of sociology and political sciences in particular. The requirement on the one hand, as Latour's second symmetry demands, is to refuse the notion of causality so dear to the strong programme, to treat all actants equally, on a kind of ‘individual’ basis, and not to particularly specify human agency. On the other hand, the idea is to recognize that even though microbes, objects and techniques are a part of interactions and do count, the way in which humans organize their relationships raises specific problems – notably because they imply their responsibility.Footnote ¹⁹ In other words, on the one hand, it is postulated that action cannot be reduced to social identities, which are always too simplistic in any event. On the other hand, it is accepted that hypotheses of this type can be considered because ‘dominant’ social groups might be quite stable historically speaking, because they are careful to reproduce themselves and protect their perimeter, and because the conduct of the populations in these organized groups is well documented.Footnote ²⁰ The former accuse the latter of sociologism, the latter deem that the works of the former render deliberate and organized power relations invisible. The former state that the latter live off ready-made macro-categories and that this prevents them from seeing how worlds evolve, the latter surmise that the former are often dogmatic and only see part of the reality. The former advocate networks, a bottom-up politics – and even multitudes;Footnote ²¹ the latter offer a more clinical description that focuses on the ordinary nature of instituted relationships. Of course, in practice, the differences are often watered down and the answers combined – which does not mean that it is uninteresting to consider these theoretical claims with care.

Within the first framework, the social world is said to be made up of actors who interact, of actants who hybridize themselves and co-construct worlds that are constantly remade. The pluriverse, to use Bruno Latour's apt expression, thus appears quite open, infinitely malleable, which constitutes an optimistic account of matters. This frame and the political philosophy that accompany it have the advantage of offering fine-grained pictures, of showing how ‘actors’ make sense of their everyday lives and actions.Footnote ²² When well used – I mean when ‘actors’ are given a dense historical ‘texture’ – the account is often of extreme interest. In contrast to this political philosophy, other people have defended stances that focus on the limits of these recompositions. They take for granted the existence of structural inequalities in any social form, and of competitive logics, for example with or between systems (in the Habermasian sense of the termFootnote ²³). Money is the medium of economic systems, and control of power is the medium of political systems, whereas words and the duty to provide justification regulate the multi-centred life of the public space. There is not one logic of interaction but several that are quite different and in competition. Thus one cannot avoid a differentiated analysis of phenomena. Or at the very least, if this dimension is not tackled head-on, it might render invisible an essential dimension of social relationships.

In global and transnational studies, similar contrasts are described, for example, by Eric Vanhaute. He claims that if we want to focus on connections, encounters and circulations of all forms, which are essential since societies develop, grow and change through multiple forms of interaction, then this should be done by considering ‘actors’ of various kinds (people, of course, but also ‘institutions, societies and human systems’) and by additionally broaching analysis from the other end, by taking societal (economic, social, cultural) structures as units of analysis (since they condition human actions and historical development). If institutions and societies are the outcome of human choices, this follows from circumstances that are given and transmitted from the past. Connections are created and redefined in a world that is not amorphous, he insists, and inequality often defines the direction and the impact of connections. They have thus to be analysed – and analysed ‘within, rather than next to, a systems-perspective’.Footnote ²⁴

This is, of course, a long-standing debate in social sciences and each postulation has its raisons d'être. It is indeed certain that, as Garfinkel taught us, human beings are not ‘cultural dopes’, that they are not ‘determined’ from the outside by their ‘social situation’ and that they are not without ‘agency’.Footnote ²⁵ Nevertheless, this does not mean that the world is not structured and always-already constraining, and that ‘actors’ are equal with regard to the capacity to construct or deconstruct worlds.

One specific point deserves attention and I will elaborate on this. Some people – the best example is Bruno Latour after his revisit of Gabriel Tarde – proceed as if there were a kind of privileged epistemological level from which, if one proceeded in the right manner and used modelling and cartography technologies, one might reconstruct the complexity of social worlds. More precisely, the idea is to numerically model the emergence of social structures. But, instead of starting from atoms endowed with the smallest number of properties so as to obtain emergent complex structures, the idea is to begin with already complex ‘monads’ and then to detect how they ‘collaborate’ to form aggregates.Footnote ²⁶ This project is most worthy; it may appear reasonable – but it might be practical only to gods. I do not wish to employ a facile authoritative argument, but I would like to come back to the exchanges that started forty years ago between condensed-matter physicists and their high-energy colleagues. I know that Bruno Latour, following Tarde, insists on the fact that people are not interchangeable atoms and must be taken and modelled from the outset as complex entities, but this only adds another level of complexity to the debate between physicists.

The thesis defended by condensed-matter physicists in the 1970s is that each level of analysis has its own effective conceptual set and that the idea of a ‘reduction’ of solids to elementary particles will never be fully operational. Philip Anderson stresses that condensed-matter physicists are right to start from the ontologies that appear at their level, that it is interesting to try and build bridges between levels (even if that will never fully succeed, except in some simplified cases), but that there is no ‘fundamental’ level.Footnote ²⁷

That leads me to make three remarks. First, it is not necessary to wait for the ‘reduction’ – even if I understand why social theorists are most interested in such a project. The molecular biologist is right to start with the DNA molecule and the historian from the institutions (a company, an expert group, etc.) that populate the sources, because this level of aggregation (these simplifications) can be adapted to their objectives. However, playing with different scales is always an interesting research programme since it opens new connections. Physicists doing this kind of modelling today, for example, have found that the complex systems they succeeded in computing from ‘elementary’ particles tend to evolve towards stable points, and that not all evolutions are likely. Certain parameters seem to weigh more than others, certain ‘operators’ make systems converge to new states, to the emergent properties that have been identified, which shows how such programmes may be fruitful.Footnote ²⁸ My third remark bears on how physicists work at this interface. They do not try and solve the full set of equations – a formidable task unlikely to be performed – but to perturb the simple system they have built and to consider the effects, the new forms of structuring that are induced.

Bruno Latour was thus most right in Irreductions: global reorganizations do not require modifying the whole ‘structure’, but identifying the local operators that make the billions of actants align themselves. This comforts those who study the way societies are recomposed through the deployment of social technologies – benchmarks activated by certain people as tools of government, for example – but waters down hopes of solving problems through exhaustive calculations.Footnote ²⁹

More generally, continuing the discussion with Bruno Latour and his unconditional defence of Gabriel Tarde against Durkheim,Footnote ³⁰ it seems hard to believe that the former was the only one to have properly posed the problem – simply because this possibility does not exist – and that it was, above all, a matter of having the initial and right categories. Tarde grasped something important when describing the world as ‘a web of inter-spiritual actions’ from which ‘truly new actions which are impossible to predict’ spring.Footnote ³¹ This does not imply, however, that Durkheim was saying nothing of interest when examining the issue from another angle. Indeed, from a historian's point of view, his ‘victory’ over Tarde at the turn of the century (the fact that he got more followers) can be understood as due to the fact that the social realm considered itself to be made up of groups and classes in conflict and in negotiation, and that in the name of a new cité of civic justice,Footnote ³² a new type of collective law – social law – was introduced. But this means neither that he or Mauss (or Halbwachs or others) were the only ones to be right, nor that they were right ‘in general’. In his constructions perhaps Durkheim was more faithful to the society around him and to the way it re-created itself, but this in no way means that he saw all that was relevant or that he said everything that could be said. In the same way, actor-network theory effectively grasps an important part of the innovations that define the world of the last few decades, and it is in this respect that it counts. That does not mean, however, that it is capable of seeing everything, or that it grasps the essence of things and defines the only correct method. From a theoretical point of view, no one can decide what ‘reality’ is made from.

I therefore suggest that we continue to exercise caution with regard to methodological prolegomena that claim to offer privileged views. I think it is more reasonable to admit that what we have to understand is too complicated, that what matters is what the different approaches have to offer of interest – that it is and will be impossible ever to guarantee the superiority of any one epistemological view over another. Not that they are all of equal merit, far from it, and we must be wary of all the adulterated merchandise. But we will never be in God's position and it is the multiplicity of framing, scales, results and values that guarantees that we might understand anything worthy. Given that all knowledge cannot but be partial, betting on the multiplicity of approaches is the only creative policy.

The debates as ontological debates

Nowadays debates also explicitly revolve around the question of which ontologies should be used. To describe the world one must state of what it consists, of what reality is made. Because the category of ‘actor’ is often taken as obvious nowadays, notably in theoretical discourses, the simplest thing to do is to dissect an injunction which has been all the rage since the 1970s: ‘just follow actors!’ This principle, in part inherited from ethnomethodology and interactionist theories, has undoubtedly been highly productive as it has forced us to focus on how individuals assess situations and act, and on how emergent situations come to the surface. But it has its weaknesses and it conveys highly specific ontologies that deserve to be considered. Now that this research programme is no longer under threat, it is possible and perhaps urgent to examine the way in which it restricts our thinking.

So, let me comment on this injunction. First there is the issue of which actors should be followed – because ‘actors’ never ‘give themselves over’ to being followed, because their number is quasi-infinite, because it depends on how we divide up our scene of inquiry. In a strong sense, we never follow actors. Rather we select them, we select entities that will play a role on the stage of our story (or in our numerical simulations), we choose the ones who will be pertinent to what we care about. Symmetrically, we ignore many others – and, explicitly or not, our choice reveals what is at stake for us and frames the conclusions that can be drawn from our story.

We also make actors move and speak in particular ways, we give meaning to their actions by putting them into scenarios – just like Pasteur made microbes act according to rules he shaped during his laboratory work. Not that we can impute anything to the actors we have selected – the situations we analyse and the data we gather are resistant to totally free interpretations. But, as science studies have demonstrated again and again, we have freedom in defining the properties we attribute to actors – and Pasteur's and Koch's microbes, as historians have shown, were not the same. Like Boyle – to use another well-known example – when claiming that we follow actors, we de facto behave as if we were mere modest witnesses – thus hiding our own agency. In short, we select actors because we have to simplify events; because, as human beings, we are limited; because our job is to tell an understandable and finite story; because we have a certain idea of what understanding (or explaining) means; because we have a certain vision of what is decisive in the case we study.Footnote ³³

Also worthy of comment is the very notion of ‘actor’, the unity that it supposes, the social ontology that it suggests. ‘Actor’ is today a polysemic term. For example, it could apply to a world in which everyone plays, or is expected to play, a role, a world where individuals must put themselves onstage, which is a notion at the core of current reality shows and advertising (such as the Reebok slogan that has long been ‘I am who I am’). It could apply to a world made up of proactive individuals who take responsibility for their lives and who are free to construct themselves, which echoes liberal beliefs and injunctions.Footnote ³⁴ It could refer to ‘rational actors’ that continuously make cost–benefit assessments and optimize their interests – or to ‘strategic actors’ in an essentially agnostic world that try to force others to fall into line, which is at the core of the initial version of actor-network theory.Footnote ³⁵

When one considers such hypotheses, it is hardly surprising that the world appears somewhat isotropic and poorly differentiated. But it may be that the social realm is more structured and that institutions and groups, as such, are not without importance when considering how societies function. If the interactionists are right, says Florent Champy, and therefore if there is no point in becoming attached to these ‘macro-realities’, then the threats which, within the framework of liberal policies, hang over the autonomy of some of them (doctors, academics or magistrates, for example) ‘only call into question undue advantages’. But, on the other hand, if these people offer ‘specialist knowledge useful for the resolution of certain problems’, and if they convey different conceptions of social justice, then the threats that hang over them may lead to the disappearance of forms ‘of rationality other than trade and managerial logics’, to the removal of interesting resources and to weakening the defences available to people and citizens.Footnote ³⁶

We must therefore conceive more differentiated kinds of ‘actor’. One interesting case is that of institutions, a subject in which Luc Boltanski has recently become interested.Footnote ³⁷ Institutions, says Boltanski, can be described as ‘beings without bodies’, as durable entities that transcend those who speak in their name. They are entities that are entitled to say, in the social world, what is true, what is just, what is unavoidable – what ‘reality’ is. Institutions exist to confront and to legislate on the infinite variety of social interests, conflicts and values, to reaffirm the ‘right’ order in a context of contradictory interests. Institutions exist to guarantee the durability of life in society – and they matter because they have the wherewithal to back their policies – for example, legal decisions or technical standards, all of which do enact reality. In most ordinary situations, they have an arbitration capacity that cannot be resisted; they have the power to impose solutions, if needed, and to subside debates. Of course, many (most, all) ‘other actors’ try to impose their own norms and visions of reality and contest institutions' visions, but the latter have, in most circumstances, high(er) power in shaping situations.

More centrally, institutions do not interact in the same way as individuals made of flesh and bone – even if they are incarnated by people, even if they only speak through them – and they have their own temporality. A judge handing down a court decision is not talking and ‘interacting’ in the same way as he would talk privately. Institutions cannot be reduced to the humans that represent them; from here follows the question of rituals, essential to their identification and perpetuation. Institutions persist beyond the longevity of their representatives (as in the traditional saying of ‘the King is dead, long live the King’) and they often have a superior capacity of resilience (as we say today). In short, institutions are ‘factish’, in Bruno Latour's sense; they are made up fictions that are, however, very real, fictions that people may attempt to undermine and denaturalize, but which remain endowed with strong and specific material, social and discursive forms of action, interaction and performance. Ontologically, this specificity dissuades us from treating them as another undifferentiated ‘actor’.Footnote ³⁸

To broaden our vision, we might finally turn to the Italian microstoria tradition of the 1970s.Footnote ³⁹Microstoria also examines things at a micro-scale, whilst suggesting, however, that we focus on situations rather than on actors. Indeed, its basic injunction is not to follow actors, but to choose and consider the configuration that will allow us to get to the heart of the issues that are of importance to actors – and to us. In the manner of E.P. Thompson, microstoria sees the world as being structured by lasting oppositions that compel ‘actors’ and as driven by a variety of logics. It focuses on local scenes, on the way in which people, groups and institutions deal with situations, but it considers essential both the emergence of what is new and the creation of historical irreversibilities. It suggests that analysts examine the range of options facing their ‘actors’, the range of possibilities that they are open to and those that they reject. Microstoria suggests that the analyst circulates between this construction (experimental, one might say, this word being so fashionable nowadays) and what the sources allow to be stated. Finally, microstoria branches from the hypothesis that there is a systematic incoherency of normative universes in every society, both between groups and people and in every person or belief system.Footnote ⁴⁰ Thereby the variety in the values and norms that guide action means that we cannot state ‘the mechanism’ of social encounter, the logic that governs the choice of actors.Footnote ⁴¹ In short, microstoria broaches a world made of individuals, groups and institutions that are always, and already, in highly structured universes where not everything is possible, where everything is not permanently or freely reconfigured.

The debates as debates on the social, moral and political stances of analysts

The debates have also taken the form of debates focusing on stances – social, moral and political – that analysts should adopt. One way of viewing this debate is to once again use Bruno Latour as a reference and to see what he says about what he calls la critique, a stance that he strongly criticizes. La critique, which is not often referenced with any real precision in his work so we do not always know who is targeted, is put forward as always harping on the same subjects, as referring to categories that are ready-made and far too loose to be effective – and as being destructive since it undermines the social body by constantly re-creating divisions. La critique relates to something ritualistic, to categories designed to decry (and thus prolong discord) and it allows neither analysis nor calm description that is true to the complexity of the world. Opposition is made between the obligation to only mobilize that which is mobilized by the actors themselves and those who adopt stances ‘from above’, which are necessarily reductionist, approximate and of little interest. One must follow the networks that ‘actors’ constantly create and jointly initiate; one must not stand on the outside and mobilize categories and values of which the actors are unaware.Footnote ⁴²

From a theoretical point of view, this question was central to science studies because it represented their starting point. Indeed, it is at the heart both of the strong programme and of the analyses of controversies. It was initially a case of symmetrizing the stance that one adopts, of asserting a sceptical posture as a rule of method. However, from the outset this rule could be interpreted in two ways: either by putting forward an axiologically neutral stance, or by presenting a way of undermining the authority of the victorious scientists and of giving the losers another chance. In the first case, one takes a neutral and purified stance (a scientific stance); in the second, analysis of the controversy is used as a critical device that reveals the injustice of what the victors say. In both cases we consider how an arrangement (or possibly a consensus) emerges, but the first focuses on the relativity of the propositions, whereas the second reveals the arbitrary nature of the construction that gains acceptance and the forces that make it possible to silence the opposition.

The most successful development of a stance that refused to settle only for the ‘symmetrizing constructionist programme’ was that achieved by feminists and subaltern studies in the 1980s, just as it is now that of those who consider the complexity of transnational relations. The structure of the arguments, which I take from Donna Haraway, is well known. Her work is founded on a tension; the contradiction between the importance of demonstrating the constructed nature of all things, the richness and the heuristic nature of symmetry – and the (live) facts of domination, injustice, invisibility and alienation. When one is caught up in these situations, which are more often than not justified by legitimate forms of knowledge, it is important to produce counter-knowledge, to elaborate one's alternative knowledge – to end up with solid and well-founded analyses that make it possible to loosen constraints and to exist in an autonomous manner. In other words, the question of knowledge is not only an epistemological question, but also a political one because the question of objectivity has to do with the very possibility of emancipation.Footnote ⁴³

Therefore those who try to escape from what is forced upon them on a daily basis tend to refuse the neutralization that one could derive from symmetrizing discourses. When it exclusively stresses that knowledge is distributed and forever local, the constructionist stance disarms reflection on what is endured, it makes most people powerless. Those who are ‘at the bottom’ need knowledge that will enable them to understand what lies behind the truths that are held up before them as ‘unmarked’ knowledge. They need knowledge (in the full sense of the word) that will allow them to make sense of the situations in which they find themselves, that will lead to better accounts of the world. ‘So I think’, writes Haraway, that

This means that there is no inconsistency between searching for the truth and having explicit a priori positions, between a desire to produce ‘good knowledge’ and a systematically critical approach articulated on social preoccupations, between a desire for objectivity and the active mobilization of our own situatedness. Indeed, claiming to know the way problems have to be tackled is not just an illusion, it is an epistemological error that always has moral and political consequences. We need, on the contrary, to fully accept that any knowledge-production process is of and in this world, to recognize this as a banal fact, and to assume the consequences. No one really knows how to set out these requirements and to resolve this formidable paradox, but the issue – a highly practical and political one – forces us to recognize the problem, to live with it and to act responsibly.

Behind these debates – and this is how I will conclude this section – lies the question of the messages that our texts pass on to our readers. I am not referring so much to what we intend to say, but to what our words convey, the promises or fears that they imply. For example, the world that emerges through the vocabulary of actors endowed with agency who constantly hybridize and co-construct worlds that show great reversibility, has a specific flavour. If we accept – and I cannot see how we cannot – that everything is effectively ‘co-produced’ through ‘complex’ and ‘hybrid’ ‘arrangements’ mobilizing many ‘agents’ – and if we also note that the formula has become hegemonic and that it relates to a relatively happy world – it might be time to consider the dark side of it, and to consider what it renders invisible, what it makes us forget.

My feeling is that the world created by these notions and by their repetition tends to make us forget the limits of agency available to certain actors; above all, it makes us not consider this an important issue that justifies our attention. I believe that the world created by these notions and by their repetition is helping to invisibilize the major constraints weighing on certain populations, that the unilateral promotion of these notions has a politics. My feeling is that it instils the idea that what is new, emerging, mobile is a good in itself, that the desire for change, adaptation and self-transformation is positive – whereas understanding the reproductions, repetitions and permanence is no longer a vital issue. My feeling is that this exclusiveness de facto euphemizes the violence and the variety of situations, that it renders them less visible, that it caps the always-already structured complexity of the world. My impression is that this vocabulary suggests a far too uniform image of the dynamics of our natures–cultures – in short, that it constructs a world that is too undifferentiated, too non-specific, blind to vast continents of questions and situations. There was a historical advantage to no longer systematically starting with major social structures – but this does not mean that these realities do not exist or that they are of little importance. The issue is thus that of the tools that we can use to integrate these questions and postures – and this text is an invitation to reintroduce this complexity and these differentiations.