NG You are widely known to be a critic of the idea of the social. This position seems closely tied to your refusal to be either modern or postmodern, and to your proposal that we have, in fact, never been modern. But how exactly is your rejection of the social tied to your rejection of 'the modern'? And beyond this, how might we conceive of the social sciences in general, and sociology in particular, without placing the social at the center of our analysis?

BL From very early on, since science studies started, I have not considered the social to be at the center of sociology, and from this starting point I slowly developed an argument about the anthropology of modernity. So, it actually goes the other way: because I started in science studies I realized that the social was not at the center of sociology but rather what I call association. The ety­mology of these two terms is the same - the word socius which gives rise to the word 'social' is the same as 'association' - but 'association' leads us in a very different direction. It was after realizing the difficulty of my arguments in sci­ence studies for sociologists and also for philosophers of science that I was led to dig further and further in the argument around the modern. So, it is because I think the social is the wrong focus for the social sciences that I have been led to anthropology.

NG This position comes out very strongly in your book We Have Never Been Modern (Latour, 1993), which you call an 'anthropology of science'. Perhaps the key distinction made in this work is between purification (the act of con­structing 'distinct ontological zones' of culture and nature or of humans and non-humans) and translation (the creation of hybrids of nature and culture and with this new types of beings, including objects). But it is hard to grasp the connection between these two practices. For example, you say that the work of purification led to the proliferation of hybrid objects, with the implication that the modern project, which is based upon the purification of different dualisms and categories, could never actually begin without destroying itself. But how is this so? How is it, to use your own words, that: 'the more we forbid ourselves to conceive of hybrids, the more possible their interbreeding becomes' (Latour, 1993:12)? And conversely, why is it that in 'premodern' societies hybridization cannot begin without an accompanying belief in purification?

BL The sentence you quote is a proposition I made for empirical research in comparative anthropology. It is still being investigated. But first you have to understand the difference between purification and hybridization. The best thing is to take an example, say of the bio-ethical discussion around stem or embryo cells. If you interview biologists they will talk about law, ethics, reli­gion, molecular biology, computers, and so on. You will obtain a long, long list of heterogeneous elements that will be linked together in a sort of seamless web. But at the end of the interview the biologist might say 'I am just doing science and I am not concerned with ethics or politics'. And here lies the difference between purification and hybridization. The first proliferation of heterogeneous entities is what I call 'hybridization'. It is the number of things necessary to do any good research in embryo or in stem cells. The other one is a gesture of cutting, a gesture that covers, hides, ignores and externalizes the work of hybridization. Now it turns out - and this is a very difficult argument that I am not sure I can prove - that this is absolutely the opposite of the premodern attitude, which would be very worried about hybrids and very explicit about forbidding them. What is peculiar to the modem Constitution, as I call it, is that it gives a lot of freedom to those who are able to say simulta­neously 'I do all these things with embryos, with marketing, ethics, religion' and also 'I am not concerned by this proliferation of linkages as I am just doing my little science in my laboratory'. What I have been interested in for all these years is the source of this freedom - its creativity, its energy, and its juvenile enthusiasm - which is provided by this built-in irresponsibility, so to speak. But this has a price. The price is that it is very difficult to do an anthropology of the modem because of this double-take. This is what is meant by the sentence you quoted, and it might not have been fully validated yet but I still think it is an interesting proposition: the basic linkage between the two processes is that the more you ignore hybridization, the more you multiply hybrids.

If you look at the work of Ulrich Beck it is interesting to see his plea for reflexive modernization as precisely a plea to get rid of this ignorance, of this double-take, and this is what he calls 'risk'. That is why I am so interested in Beck, because he points exactly to the point where risky objects, or what I would call 'disheveled' or 'quasi-objects', are precisely what we have to portray explicitly. But, of course, his position as well as mine has a political price, which is just the opposite of the one not paid by the modernizers, which is that we have to slow down. We have to slow down innovation to obtain due processes in the political representation of innovation. I am not sure my proposition is validated anthropologically but it is validated politically, because, politically, it is perfectly clear that it is this double bind of the modern which forbids what we call 'reflexive modernity', or what I call 'non-modernity'.

NG You say that because the quest for purification necessarily engenders acts of translation the modern can never actually begin. But why define the modern simply in terms of purification? What is the theoretical motivation for doing so?

BL Purification is simply one of the many mechanisms making up the anthropology of ourselves. There are endless ways of defining modernity, but my take, because I was coming from science studies, was to use the position that science is a good tell-tale to decide who is modern and who is not: tell me what you think of science, and I will tell you who you are. Purification is one way to divide between natural, dead, mute matters of fact which are simulta­neously real and meaningless, and, on the other hand, humans with intentions, subjectivity, intentionality, interests, who have many things in their heads but which are all unreal and meaningful. Strange system, is it not? Of course, it is simply a Constitution because the world itself is not really made of boring matters of fact, on the one hand, and of acting intentional subjects, on the other. In the world these things do not exist, so they have to be made out of a political Constitution. It seems to me (although this is not really history but more of a proposition) that a good test to know how modern someone is is to see how much of a distinction he or she will accept between matters of fact and intentional humans. So far, it has worked very well for finding my way through the social sciences, which are massively modern. It is a good litmus test.

NG It seems that one of the things you are particularly keen to do is to overcome all conceptual dualisms, particularly those relating to fictional divi­sions between nature and culture. But could it be that because you set up your initial problem in terms of certain dualisms (nature and culture, or purification and translation) you can never in fact free yourself from these oppositions? For example, in talking of nature-culture hybrids do you not in some sense lend credibility to the actual existence of nature and culture, thereby reinforcing rather than overcoming an underlying dualism?

BL I have nothing against dualisms, it is just that as social scientists we must find interesting dualisms, not the ones that render our subject matter impos­sible to study. For instance, I am very interested in another dualism (which I have introduced in my book Politiques de la nature (Latour, 1999a)): the dif­ference between the number of entities to be linked together (i.e. how many are we?) and the question of sharing a common world (i.e. can we live together?).

The dualism between these two absolutely essential questions is very important for me, so I do not take a position against duality per se. And if I say that a fact/ value or human/non-human dichotomy does not work, it does not mean to say that I am against all dichotomies. Rather, it is because I am trying to get rid of the ones that have been made to render the modern unstudiable, that is all.

But things change fast here, and there is now very good work by anthro­pologists like Philippe Descola on nature and culture, so that this dichotomy is no longer as powerful as it was when I was writing We Have Never Been Modern. And in playing with the association between humans and non-humans, I am trying to do something different from distinguishing between subject and object, but I agree it is difficult as people constantly come back to the old dichotomies. It is my readers who reinforce these dualisms not me, but maybe this is because I don't do enough work!

NG One key thing that comes out of this approach is a theory of the object. This theory has a distinctly political edge. For example, you talk of a 'parlia­ment' or 'democracy of things'. Why is this?

BL First, because of the etymology of the word thing, which also means assembly or Ding in German and old English. All 'things', so to speak, have started in a political assembly of some sort, in a quasi-judiciary state of affairs. This is why I use the term 'state of affairs' rather than 'matter of fact', and also for the simple reason that objects have always been involved in politics under the aegis of Nature. Nature is not an obvious ontological category; it is a highly elaborated and controversial way of doing politics. When things were convoked or assembled as Nature, for instance the gene in socio-biology, they have always been part of a political process. The problem was that it was completely implicit and it was the peculiar political function of Nature to do the job outside of the political domain, which remained limited to human intentions and representations. What I have done is simply to ask: if they are political anyhow then what is the due process? Let us imagine the due process for these things. I am not politicizing these non-political matters of fact, for they are already political states of affairs; I am bringing them back into normal due process, which I think is the responsibility of intellectuals.

NG You have also spoken of quasi-objects, and more recently, recalcitrant objects. But is there necessarily anything new about seeing objects as hybrid social and a-social forms? In your conversations with Michel Serres (Serres and Latour, 1995: 200) you complain that social science has been 'obsessed by subjects alone, by people interacting among themselves, and never speaks of objects per se'. But is this exactly true? In the work of Karl Marx, for example, there is the idea that capitalist production fetishizes objects (in the form of commodities) by investing in them a particular set of social relations, while at the same time reifying humans by making them more object-like. In this way it would seem that sociology has spoken about objects since its inception. Would you disagree with this viewpoint?

BL Commodities, fetish, reification! The three words I have combated most ... Yes, sociology has spoken about objects, but so badly! The notions of fetishism and commodities are among the worst things that have happened to sociology to understand economics, capitalism and objects. The notion of fetish is exactly the sort of illusion social scientists possess about the illusion of others and is exactly what has made the anthropological study of markets and goods so impossibly difficult.

Of course, in Marx there is an attention to materiality that is very important. But all the material elements are linked together to make an infrastructure. So the objects are there but they take the role of a vast infrastructure whose causal forces are largely exaggerated, because they are either too powerful - they cause social relations to happen - or far too weak - they are simply congealed, frozen or reified social relations. Objects have never had a chance in the social sciences because either they are too powerless (and this is exactly the notion of fetish where they are supposed to be just that onto which we project human inge­nuity), or too powerful (and they make you do things causally). I don't know of any social scientist, except Simmel perhaps, who has done any interesting work on neither all-powerful nor powerless objects. They always count for too much or not enough.

Object agencies are never focused on, which is not surprising, of course, because social scientists had other tasks, especially emancipatory tasks that blinded them to the interest of the object, and it was not their priority. But in science studies, objects are everywhere, not as simply 'reified' but, on the contrary, as controversial states of affairs. And the traditions of technical determinism or technical push have offered no resistance to the fieldwork we have done in technical laboratories or technological projects. It is as simple as that. Yes, the social sciences have spoken of objects, but mainly as superficial screens on which to project the social.

NG Part of your approach to the study of objects has been to develop a physical sociology to sit alongside, or be incorporated into, what you call 'social sociology' (Latour, 2000). But what do you mean by 'physical' and 'social sociology'? And what is the new 'political situation' that you say is lending to the weakening of divisions between natural and social scientists?

BL That was partly a joke, of course. I was comparing sociology with anthropology, which has the chance, not in Britain but in the United States, of having physical and cultural elements side by side, just as there is physical and human geography, and in psychology neurobiologists and psychoanalysts sit on the same committees. But sociology is different because of what Zygmunt Bauman calls its 'legislative pretension' - to solve the social question by shortcutting political process - and so has never been very interested in having a physical part. This would not be much of a problem, after all sociologists can do all the things they want and I don't care much, if the situation had not changed quite a bit since the nineteenth century. Today the problems are with the state of affairs: the number of elements that are hybrids and in need of a political voice have multiplied. And it is extremely annoying to have sociolo­gists of the social type saying 'well I'm just dealing in the symbolic dimension of the object but, of course, I don't touch the object and I let you, the biologist, economist, geneticist, physicist, do that part of the work', because it means that you deprive yourself of the entry into political processes with any sort of weight. And that is why I fight against this object-less sociology and against the dis­regard for science studies - especially in Britain, which is paradoxical because this is where it was invented - because I think it has become politically detri­mental.

NG Recent theories of the object have also been tied to ideas of networks and flows. But you seem to have dropped this language from your writings. For example, you have objected to using the term 'network', saying that it now means 'transport without deformation' and 'instantaneous, unmediated access to every piece of information', or in other words the very opposite to what you originally meant (Latour, 1999b: 15). In place of this, you talk of the 'topology of the social', and of the transformation of 'the social from what was a surface, a territory, a province of reality, into a circulation' (Latour, 1999b: 19). What do you mean by the idea of this social as circulation?

BL I don't use the word network because of what you say, but I still think it is useful. I have changed my mind since this paper you quote, because now I am using the work of Gabriel Tarde. There were clearly two traditions at a beginning of sociology: one a sociology of the social, and the other one the sociology of mobilities, transfers and what Tarde called 'imitative rays' (which is not a very good term but is really the translation of the equally bad term 'actor-network'). So, there were already at the origin of sociology, at least in France, two traditions. One of them saw the social as a special part of reality, different from geology, biology, economics and so on, and another one saw very well that what counts in the social is the type of connections that are made.

In this view, the social is not a homogeneous domain of reality composed of social elements, but a movement between non-social elements - a piece of law, laboratory practice, etc. - connected in certain ways. What I have been doing together with Michel Callon and John Law around the word network is to revise or revive this second tradition. It turns out that network is also a term used for sewage, telephones and the Internet, so I tried to play at one point with the word work-net, but it didn't function very well. What is important in the word network is the word work. You need work in order to make the con­nection. It simply means that we designate what is not already there as a sui generis reserve of forces, while the sociologists of the social, the other guys, invoke it under the name of society. That is the difference between the two arguments. But I still think that the word network can be used.

NG In view of this, how do you now approach concepts such as 'local' and 'global', which, you say, 'work well for surfaces and geometry, but very badly for networks and topology' (Latour, 1993: 119)?

BL That is precisely the earliest and in my view the most important aspect of our social theory, and one dating from a 1981 paper I wrote on the Big Leviathan (see Latour and Callon, 1981). But I have never met any social scientist that has understood it. Scale is not one of the things that the sociologist should decide but what actors themselves produce in scaling or measuring up one another. This means that the 'local' and 'global' are two totally implausible departure points, not because they have to be 'dialectically reconciled' as in the notion of habitus, but because they simply don't exist. The social world is flat or, if you wish, dimensionless. I have tried to demonstrate this in a book I did on Paris, but there is nothing to be done: sociologists have the practice of zooming from macro to micro built in their skulls, it seems, and so the most important features of social theory - that scale is produced, sites are localized, and the global is always localized in highly connected loci - simply escape them.

NG In addition to seeing the social as circulation, you have also defined the social as a part rather than as a whole. By this you seem to mean that the social does not exist at different levels of analysis (macro or micro), for 'the big is never more than the simplification of one element of the small' (Latour 2002: 123). This idea is developed from the work of Gabriel Tarde, in particular his writings on monadology (see Latour, 2002). What interests you in the work of Tarde, and how might his work lend to an analysis of the social?

BL Tarde is the inventor of sociology just as much as Comte, Spencer and Durkheim, except he has been kicked out because he has been accused of psychologizing everything - which is the exact opposite of what he says. 'What he calls sociology is interpsychology, that is, the attention to all those circulating entities that for him are what the social is made of, and he never said a thing on intrapsychology. Basically he was an associationist very early on. He has this extraordinary idea that what is blocking the whole interpretation of the social is the macro and micro distinction. I have made this argument for years but I found the same thing in Tarde and that is why I am so interested in him. The macro is just a slight amplification or standardization of the micro. Organiza­tions are not a pyramid or a sphere but the slight provisional amplifications of the variations of the micro, and that is why doing a monadology, doing local fieldwork, is as interesting, or in view ofTarde more interesting, than gathering statistics. To make his point Tarde goes into Leibniz and develops a mon­adology - which influenced Deleuze very much by the way. It is hard to swallow for social scientists because what Tarde did not do was to break with philosophy, and, of course, for sociologists this is a sin because the idea is that you have to break with philosophy in order to be scientific. Tarde does not make this distinction, and nor do I.

NG You have spoken of the revenge ofTarde over Durkheim, or the idea that 'society explains nothing but has to be explained' (Latour, 2002). Does this apply to the social too, or are 'the social' and 'society' two quite different things?

BL Society does not exist - this is Tarde's position and is my position too, as well as Mrs Thatcher's ... You do not need to add society to anything to provide a social explanation. The social (as society) explains nothing. It seems to explain something only in the very perverse view of social scientists - the one held by Bourdieu especially - that it is a necessary illusion, that if we were to reveal the illusion behind market forces, art, science, etc., people would be so blinded by truth that you would need to cover it up with a veil of illusions. But as I have said many times, the ones who need the veil of illusions are the sociologists, not the actors themselves. It is like ether in physics at the end of the nineteenth century. Just as physicists at the tum of the last century learned to do without ether, social scientists can learn to do without the social understood as society, but not the social understood as association, of course, because that is our business. The notion of society is the remnant of trans­cendence in social sciences that do not care for religion. Society is their reli­gion, their last transcendence, and plays the same role of God as in some forms of very bad theology.

NG So society is a transcendental or quasi-religious concept?

BL Yes, it is exactly that. It does nothing but reassures, gives moral comfort, and allows the sociologist to have an overview. It does moral things but has no empirical grasp. And that is why I am fighting it.

NG But how is what you do so different?

BL The responsibility of the social scientists, in their view, has always been to enlighten the actors, and to emancipate them. So, whenever it becomes a bit complicated philosophically because actors do all sorts of bizarre things - they define scale, they define time, they define subjectivities, they go into very bizarre, complex arguments about divinities, about market forces, etc. - what do the social scientists do? They ignore all these things because they want to free these actors from their ignorance. If someone believes in God, they say we know that God does not exist, and so belief must be projected onto something else. The actor might insist that he believes in God, but this would fall on deaf ears with the social scientist, and the ontology of God is thrown out of the picture. The sociologist would do the same thing with law (just a packaging of social forces), with technical objects (they are just there to project social for­ces), with scientific nature, and so on. Every time they have troubling data, they throw them out and replace them with the all-purpose social. It is the only discipline that does so, and all, of course, in the name of scientific method!

NG How might we get around this problem?

BL I am an empiricist. I am trying to follow what the actors do. And if they do and say complicated things, like if they say that God made them act, then I take God in the picture very seriously and I follow to the bitter end what sort of ontology that sort of God involves, rather than snugly bracketing out the 'real force of God' because 'obviously' we 'know' that 'He does not do a thing'. This method has been carried out recently in two marvellous studies by French scholars Albert Piette and Elizabeth Claverie, and in their hands God becomes a very important actor; yes, God does things and is no longer the sort of obvious illusion which the sociologist's task is to make evaporate. These two research programs are very, very different.

NG Further to this, you also make the distinction between collectives and societies (Latour, 1999c). What is the purpose of this distinction? And how might it connect to a theory of the monad?

BL The first assembly in the modem Constitution is that of Nature, con­sidered as non-political, and the second is Society, with this one and this one only being explicitly political, legal and moral. If my argument is followed, and if incontrovertible matters of fact are becoming more and more controversial states of affairs, we are entering a risk society or second modernization, and so the divide between the two assemblies has become moot. The social, as I have just said, is not the right object to study because there is nothing in the world homogeneous enough to compose it. You cannot divide any object - a tape recorder, a gene, etc. - into its natural or symbolic components any more. So the question is: how do you assemble them again and anew? And this is what I call a 'collective'. But what is important in the collective is the verb collect. How do you collect?

I have elaborated this argument much further because I am interested in the task of collecting, and in the two questions I mentioned before: one of them being how many are we, and the other, how can we live together? In my view, it is this convocation of a collective that is now the duty of the social sciences and what I call politics: the progressive composition of the common world. This is a very different political duty from the past, which was a mixture of emancipation and scientism. I am trying to avoid these two things and imagine a destiny for the social sciences that breaks with the modem, because sociology is especially modernist. In fact, I am simply trying to understand why so much social sci­ence is pure trash. It is always a mystery to me: how do all these interesting people on interesting subjects manage to write all this absolutely boring, uninteresting stuff? It is not: why are the social sciences not 'really' sciences? That is the obsession of the epistemologists. I am simply trying to understand why they are just plain bad. Why are they empirically bad? Why are there so few empirical, first-hand field studies? That is what I am interested in. One possible solution, I don't claim it is the only one, is that whenever it becomes compli­cated philosophically, metaphysically or ontologically the social sciences simply stop and shift to their emancipation mode. They say: 'this is too complicated, let us be relevant'. In this way, they constantly thwart the effort of description because of an appeal for emancipation or a political relevance which, of course, they do not have because they are locked in their campuses, like everyone else. But I am trying to imagine another take, which would be, first, more scientific - actors would be allowed to deploy their metaphysics and their ontologies - and, second, more political: the collective has to be convoked. Like Isabelle Sten­gers, I portray the social scientists as the diplomats of this collective, and then ask: how do we do this?

NG As for the second half of my question, if we connect things or monads together do they not in some way lose their singularity?

BL We don't know this before we have tried to assemble the collective and pay the full price of this convocation. If we wanted to get all the Gods together they might not want to lose their singularity. This is why the notion of diplo­macy is so important: to record this singularity as far as possible and to avoid as much as possible the collection being made cheaply. For example, Nature was one way of making the collection cheaply, and so was Society.

NG How might we proceed to collect things then?

BL This is the object of my work on the politics of nature (Latour, 1999a). I did a whole book on the 'parliament of things' which asks how it is organized. It is pure political philosophy but it is a very important question. It is perhaps the most important question for the social sciences, one that was given too fast a solution at the tum of the last century by the Durkheimians through a mixture of emancipation and scientism. But these questions have, of course, become very interesting again and very difficult, and this is what I am concerned with now.

NG Apart from this, you have also been involved in the Iconoclash exhibition at Karlsruhe. Does this mark a change of direction for you? In your accom­panying piece to this exhibition - 'What is Iconoclash' - you say that the show aimed to present 'images, objects, statues, signs and documents in a way that demonstrates the connections they have with other images, objects, statues, signs and documents' (Latour and Weibel, 2002). This seems to follow on from your general interest in networks and quasi-objects, but what is your interest in images here? More specifically, why are you interested in the clashing rather than breaking of images (iconoclash rather than iconoclasm)?

BL It was a very big undertaking. There is no fast link with the social science arguments we just made, it is linked rather to another interest of mine - doing an anthropology of a critical gesture. As I got into this argument about the social, I got more and more interested in why critique is the normal way to behave for Left social scientists, having read Marx and Benjamin. If they study so little it is because they are so busy debunking. So, I became interested in the destruction of images in the anthropology of this very gesture of breaking fetishes. And, of course, there is a long iconoclast tradition in religion: Pro­testantism, Byzantinism, Catholicism against the others, and so on. There is an absolutely enormous tradition which is extremely interesting and about which I know much more since this exhibition. There is also the case of modern art, which is also engaged in a very strong iconoclastic tradition. And finally, there is the argument in science about theory against intuition, and also the very powerful right science provides for 'debunking' beliefs. So, there is an enor­mous domain which we mapped out in this exhibition and the catalogue. This domain is a clash rather than a clasm.

I am not interested myself in being an iconoclast - although I have been accused by some scientists of being one! - but I am interested in using ico­noclasm as a topic rather than a resource, and to try to say something about critique that is not critical. Luc Boltanski calls this 'a sociology of critique' instead of a 'critical sociology'. I think critique has emptied itself and its forces have been weakened. Again, this corresponded with a belief in emancipation, with a certain definition of the social sciences, a certain confidence in political processes, and a certain belief in inevitable modernization. But none of these things are still there, so all of these mental armaments of social science have to be retuned. This is what I tried to do in this exhibition with my colleagues, which was a marvelous experience of trying to spatialize complex arguments. And, by the way, I am doing another one in 2004 on the question of assembling the collecting of things. This is because the next question, after the question of critique, is 'how do we now do the work of assembling the collective?', which was exactly your previous question. How, for example, do you compare pro­cedures in different domains, from scientific congresses to Palabres in Africa, at the very practical level?

NG What do you think sociologists might learn from an exhibition such as Iconoclash?

BL Sociologists have stopped learning from those sorts of places a long, long time ago! They have deprived themselves of philosophy, literature and art. There was one sociologist in the exhibition - Luc Boltanski - but he is the best and most innovative social scientist in France. Apart from him I do not think there was anyone else. The exhibition itself was thick with things and images and questions of theology (which, of course, social scientists despise or 'socially explain'), science (which they either worship or love to hate), and art (which they constantly 'socially explain'). But when you 'socially explain' things you are not interested in things anymore. This exhibition was not made for sociologists but for the general public who could reject the belief in belief they had been accused of holding! I wanted to call this exhibition at some point the 'revenge of the Philistines'! At last, a breathing space with no 'social explanation' ...

NG Finally, what continues to attract you to the discipline of sociology? And in which direction would you like to see the discipline move in the future?

BL I like sociology, but I don't like sociologists! There is something deeply disheartening in the way sociologists think they have the right not to look at their data. My view of sociology is influenced, of course, by the fact that I live in France, which has been completely swamped by the worst possible Durkhei­mianism, which is Bourdieu's 'critical' sociology- that is, the least critical of all sociological imperium. But if you look at Boltanski's work (he used to work with Bourdieu) it is fabulously interesting sociology. Even in France there are very, very good sociologists, but they work on completely different paradigms from that of critical Durkheimianism. And this does not mean that they work in my little paradigm of the sociology of association. There are lots of domains of sociology that are very interesting.

So yes, in spite of all that I have said, I believe in the discipline. Like Comte I even believe it is the Queen of the sciences! I think the discipline as a collective has an essential role to play, exactly as important as Comte imagined - in a very different way but just as important. After all, what Comte wanted was for sociology to do this diplomatic work. Of course, he was a complete madman and the argument for sociology being the science overseeing all other science was ridiculous. But the idea he had of sociology being a master of ceremonies of the collective (to use my term) was right. So, I believe just as much in sociology as Comte did, but for very different reasons.

I think sociology is very important, but I don't understand why it is con­stantly limited. One of the limits, I know, is coming from its bizarre ideas about science, but this, I think, we have overcome to some extent through science studies. Another limit is coming from the idea of the social. I think that the sociology of associations is overcoming that defect as well. So I think we are making some progress. What is missing now might be the new type of numbers, the new type of data, because you cannot always stay at the limit of qualitative data. You have, at some point, to be able to get a different type of numbers for the same tasks that statistics used to give sociology at the beginning of the last century. This is the limiting factor now, but I believe a lot in digitalization, which highlights or materializes social connections in a way that may be very productive for giving qualitative sociology its quantitative arm. So, I am very positive for sociology, but it cannot remain stuck in the 1950s or in the deconstructed ruins of Marxism. It cannot continue to use a destitute reper­toire which, while important at the beginning of the twentieth century, between the wars and for reconstruction after the war, has now used itself up.