Exploring the social without a separate domain for religion

In post-secular societies—after secularisation—it may increasingly be the case that the connecting and structuring of religious matter is done outsidedesignated religious sites and without appointed religious experts. The centres of calculation have changed and so the connections between these are different. The former ways of translation and ordering are transforming into new ones. By exiting the designated sites religious matter has found new freedom with the new associations and inventions in the processes of translation. Less control leads to more heterogeneous agencies and facilitates the mobility of religious materials. This less controlled mobility of religious actants can also produce an apparent increase of religious matter, but this does not necessarily mean the return of religion. In any case, this increased plurality combined with increased mobility calls for perspectives which can recognise novelty, andnot just in comparison with previous states of affairs. Actor-network theory (ANT) is about tracing the webs of associations between myriad actants whose collective actions produce what we call ‘society’. Dismissing the notion of ‘the social’ as a kind of ‘stuff ’, ANT insists that sociology should focus on the interactional processes—the circulation of ‘the social’ among human and non-human actants—collectively assembling emerging states of affairs.


Introduction
As Bruno Latour argues in his book Reassembling the Social: An Introduction to Actor-Network-Theory (2005), the problem with much of contemporary sociology is that it has become too 'good' at what it does.Sociology has become too easy.One description of the problem is that many sociologists seem to think that what they should be doing is finding powerful umbrella concepts that can be used to represent or account for some widely occurring phenomenon.For example, a researcher finds that no established concept manages to capture their findings and so they coin another term as a kind of shorthand for what is described in detail in a report.Another sociologist is impressed with the concept and consequently applies it in his/her work, but to somewhat different circumstances.After a while, the concept gets used as an explanation for a wide range of seemingly similar phenomena and students begin asking their professors whether the concept might be suitable as a 'framework' for their study.What is most troubling about this development is that now, instead of the concept being used as a shorthand reference or allusion to the original account, it begins to be used as an explanation.It has subtly transformed from a shorthand marker to a 'social force' which sociologists invoke to explain given states of affairs, where to 'explain' entails invoking as the very force 'making things happen' .While initially, the concept may have referred to a complex interaction among a number of collaborating factors, this complexity has now been black-boxed into the concept which in effect substitutes and obscures that collaboration.As Latour puts it: What begins as a classical and fully respectable search for an explanation ends up replacing the explanandum with the explanans.While other sciences keep adding causes to phenomena, sociology might be the only one whose 'causes' risk having the strange effect of making the phenomena they are supposed to explain vanish altogether.(Latour 2005: 100.)Actor-network theory (ANT) is about tracing the webs of associations between myriad actants whose collective actions produce what we call 'society' .Dismissing the notion of 'the social' as a kind of 'stuff ' , ANT insists that sociology should focus on the interactional processes-the circulation of 'the social' among human and non-human actants-collectively assembling emerging states of affairs.In order to achieve such a task, Latour insists that we need to set aside a number of widely held presuppositions.In order to be able to produce accounts which trace the assembling of the social, what is needed is the exact opposite of what Latour criticises sociology for: we need to increase the number of recognised actors/actants involved in this on-going assembling instead of decreasing it (which is what we do when we replace the actors/ actants with more and more encompassing 'social forces').Only by acknowledging the influence of every thing that is involved in making a difference to states of affairs can we actually account for complexity without recourse to 'social explanations' .The move is thus in a totally different direction: we need to get better at describing in detail the movement of the social in complex associations (and in order to do that we need to acknowledge/recognise the associating actants), instead of 'explaining away' this complexity by replacing it with sociological concepts.
ANT is also referred to as a 'sociology of translation' or a 'sociology of association' , which are in many ways better names than actor-network theory.ANT sounds technical and somewhat intimidating.One problem is that the 'network' part sounds as if mapping networks would be the main point.It is not.Deploying actants which form a network is needed in order to be able to start tracing how 'the social' circulates within these networks and hold them together.So, what is central is really tracing the associations/the movement of 'the social'; this tracing eventually generates a 'network' .Actants are provided with their unique agency not by macro structure, but by the network(s) which they performatively partake in generating.The point about 'network' should perhaps mainly be understood in terms of a counter-argument against the notion of 'context' .While this notion has been used-and rightly so-to emphasise that there are important distinctions to be made which are contextually contingent, to really take seriously this point entails abandoning the vague use of 'context' as a container.When 'context' is referred to as a container its relevance is reduced to one relation only, that between 'thing' and 'context' .This amounts to saying that a thing is influenced by where it is and this is usually invoked in order to account for various differences in a commonsensical but impossibly non-specific way.A network perspecive on the other hand distinctively recognises the need to identify what the context is made of and how those multiple 'constituents' interact with whatever is perceived to be 'in' the context.When focusing on the agency of an isolatable 'agent' , instead of situating it in a 'context' ('society'), we need to distinguish between the various actants which contextualise an actor's ability to act; we should presume that it is an actor in a network in need of tra cing-insisting that what it is doing can only be understood by considering the various actants which are 'making it act' ('making' here signifying 'causing to act' or 'exerting an influence' without fully determining the action).For that, 'context' does not suffice.The context (as container) cannot account for how the actor is made to act.Instead it serves to strengthen the temptation to maintain a given boundary between a subjective 'inside' and an objective 'outside' (that is: context).Furthermore, tracing the 'network'-describing the associations that eventually generated an 'effect'-provides an account, which, in contrast to referring to 'social forces' or hinting at the significance of 'context' , is a contestable account.'Context' is thus interesting only when seen as myriad interactions serving to collectively contextualise.Nothing is 'in' the context as much as part of a process of collective contextualisation.
In ANT we are primarily interested in describing this complex actionthe actions and interactions and the circulation of actants from site to site, mobilising other actants in collectively assembling society.Rather than looking at society as a 'something' , we are looking at it as a 'doing' , where the 'work' needed to generate and maintain as well as alter structure is in focus and where this work is not done solely by human beings.Thus, actants are not perceived in terms of substance, but in terms of 'action displacers' .Actants are 'things' which make a difference in a course of action.For our purposes in describing the circulation of the social, what things 'are' is of less importance than how they 'emerged' , how they are enabled and how they are entangled.Contrary to much of the criticism levelled against ANT, we do not think that one is necessarily reducing objects to their relations in emphasising the significance of their relations for their ability to act.In terms of agency and action, the question of what acts (in the sense of its 'core' or 'essence'), it seems to us, can be allowed to remain out of reach and indefinable.As far as we can see, agreeing with Graham Harman when he says: 'The object is a dark crystal veiled in a private vacuum: irreducible to its own pieces, and equally irreducible to its outward relations with other things' (2011: 47), does not automatically compel us to feel troubled (although we humbly grant that we may simply be failing to grasp the full significance of his critique).Whether or not we claim that Latour 'denies' or 'refutes' any hiding 'essence' beyond relations or claim that such an essence is conceptually out of reach, the point is that our focus is on tracing processes and describing complex action, not on 'substance' or representation.
As soon as we recognise that this collective 'work'-a word used here to emphasise the need to consider the vitality of actants-is full of tensions, negotiations, swerves, reactions and so forth, it becomes obvious that conceptual 'social forces' and references to 'context' do not suffice to account for very much.Instead, too often they become excuses for ignoring complexity: indeed, why do the arduous work of following the actors involved in collectively generating certain states of affairs if one can simply invoke a concept as explan ation; if one can simply continue to re-establish one's supposed lack of naivety by constantly acknowledging that 'well, of course, in reality it is a bit more complicated'?There are many answers to such questions.Some of them have to do with the bureaucracy around research grants and criteria set for 'good research' (e.g.what does it mean to have contributed with something 'new'?) and some of them have to do with an over-eagerness to render 'phenomena' comparable too quickly.

The critique of 'micro' vs. 'macro'
Society emerges through a multitude of actants making each other do things.No single actant (be it human or non-human, 'material' or 'discursive') can singlehandedly determine action; all are enabled by their associations with other actants, thus rendering agency always only partially in the control of a single actant.This is the case on all 'levels'-which is why ANT insists on limiting the use of concepts like 'macro ' and 'micro' and 'local' and 'global' to being only shorthand references and rhetorical allusions.In tracing associations, it makes no sense to focus only on 'local' as opposed to 'global' aspects or vice versa.Instead, some of the relevant associations come from far away and others are 'close by' .Similarly, in many cases some of the actants' influence stems from hundreds of years ago, while others might be considerably 'younger' .Remember that we are trying to understand some more or less specific states of affairs.Therefore, to really be open to finding something new or surprising, we should not decide beforehand what kinds of actors to consider, or in what 'domain' they 'reside': instead, we need to be open to what the trace reveals.As Latour notes, 'If the analyst takes upon herself to decide in advance and a priori the scale in which all the actors are embedded, then most of the work they have to do to establish connections will simply vanish from view' (2005: 220).A part of this work is for instance about localising (making 'local') something distant/global.The point is not one of denying 'macro' influences by insisting that 'everything is micro' .Rather, the problem lies in the practice among social scientists of jumping to 'the macro' without a localising chain; once this problem is taken seriously, 'the macro' as a separate 'field' is rendered superfluous and distractive.Again: ANT does not want to delimit or presume to foresee what kinds of actants should be focused on, only to shift focus onto the productive associations.Contrary to misconceptions, one is not obliged to become blind ('neutral') in terms of choosing what to research.We can, after all, safely form some preliminary notion of what we want to study, without allowing that notion to dictate the actual process.ANT merely insists that unless we want to revert to social explanations and the safe (but predictably platitudinous) practices of checking compatibility with predefined frameworks and hypotheses, we need to compose descriptions in which the ability/agency of an actant to do something is accounted for by the influence of other actants.This is in stark contrast to a tendency to ignore actants which are deemed to be 'outside our expertise' , thereby 'legitimising' the reverting again to explanations invoking the all-powerful 'ghost-actants' of the 'macro realm' .Crucially, focusing on the circulation of the social allows for us not to have to decide between 'social forces' (structure) and 'subjective intentions' (agency, psychology) as explanations for action.We will say more about this in the following two sections.
Why does a certain institution or group have certain characteristics?ANT looks at this in terms of 'how' .How did it become like that?What was involved?And then, crucially, no single actor could by itself achieve its present 'form' or figuration.Thus, the intentions of, say, Joseph Smith for the future of the LDS Church are not entirely irrelevant, but almost.Neither can we claim that he is the originator (ex nihilo).He may have 'started' something (what made him do it?),but as soon as he did, a swarm of actants intervened and displaced his intentions.Where did his intentions come from?Who feels comfortable claiming that they came from 'within'?What the LDS Church eventually 'became' is the result of a complex-and traceable-web of 'things' making each other do things.The point here is not only about the history of Mormonism.It is a point about action and group formation: that when tra-cing specific developments it is always a question of considering the circulation of multiple actants and the slightly displacing effects they have on one another.'Displacing' does not here entail displacing something 'original' , but is more akin to the notion of every beginning being only 'a point of departure' from previous states of affairs towards an often unpredictable future.As Jane Bennett points out, 'agency is also bound up with the idea of a trajectory, a directionality or movement away from somewhere even if the toward which it moves is obscure or even absent' (2010: 32).No actant can act alone; its avenues of action are contingent upon other actants and this dependence on other actants 'displaces' its action (although it also enables it).Such an understanding of 'interference' entails that interference is never inherently 'good' or 'bad' .Instead, it means that 'alliances' make a difference to those allied: they 'displace' one another.As for contemporary states of affairs, the question is how stability is maintained.In other words, 'although groups seem to be already fully equipped, ANT sees none existing without a rather large retinue of group makers, group talkers, and group holders' (Latour 2005: 32).This involves practices of self-definition, often via negation.'It is always by comparison with other competing ties that any tie is emphasised.So for every group to be defined, a list of anti-groups is set up as well. ..[A]ctors are always engaged in the business of mapping the "social context" in which they are placed, thus offering the analyst a full-blooded theory of what sort of sociology they should be treated with.' (Latour 2005: 32.) Considering that whenever we want to achieve something, actants swarm to displace our direction (a 'direction' made possible only by having a point of departure from which to depart), it is really quite astonishing that stability ('social structure') is ever achieved.Yet it is.Rather than describing already stable groups or institutions-a feat which 'standard' sociology accomplishes rather well-ANT researchers are interested in studying processes of destabilisation (network breakdown) and stabilisation (network composition).Human intentions cannot explain it, nor can a self-perpetuating 'social structure' .But the description of the tensions, reactions and displacements between actants that gradually configure into constellations can.As Latour puts it: For sociologists of the social, the rule is order while decay, change, or creation are the exceptions.For the sociologists of associations, the rule is performance and what has to be explained, the troubling exceptions, are any type of stability over the long term and on a larger scale.It is as if, in the two schools, background and foreground were reversed.(Latour 2005: 35.)

From scale to scaling
What is interesting and which serves to clarify the specific relevance of ANT for the study of religion is that despite our criticism of the (over)use of explanatory concepts in sociology, we cannot ignore the circulation of these concepts.Having been introduced they now circulate as actants, mobilising other actants.So, even if they are failing as good explanations, they are making a difference and should not escape our attention.What Latour manages to illustrate is that social scientists have been busy standardising and categorising the world and although ANT is critical of this endeavour, the nature of the criticism is actually to treat sociologists as metaphysicians; to treat (or acknowledge) everyone as metaphysicians, while refraining from subjecting them to our meta-analysis.By not imposing our standards or supposedly 'finding' a match between 'theory' and 'expressions of it in practice' , we can shift our attention to the processes of standardisation/stabilisation, the use of standards in performance, communication and interaction.As Latour puts it: 'Group delineation is not only one of the occupations of social scientists, but also the very constant task of the actors themselves.Actors do the sociology for the sociologists and sociologists learn from the actors what makes up their set of associations.' (Latour 2005: 32.)For instance, while it is not only contrivedleading to never-ending academic disputes about 'accurate' definitions-but also ethically problematic to study 'homosexuality' as a container-category ('what are homosexuals like?'), the processes of contestation, reactions to misrepresentation, victim versus perpetrator rhetoric, debates about 'inclusion' and 'exclusion' , in short the processes of standardisation, are fascinating and productive.'For scientific, political, and even moral reasons, it is crucial that enquirers do not in advance, and in place of the actors, define what sorts of building blocks the social world is made of ' (Latour 2005: 41).Once we stop presuming 'identities' and 'categories' to be merely the outcome of some fixed 'structure' and instead begin identifying the controversies involved and what they produce, massive amounts of research are opened up.In much of such research, concepts coined by social scientists play a remarkable role, not unlike that of concepts coined by religious innovators.They churn out standards against which and in reference to which actors identify themselves and scales with which to evaluate and differentiate themselves and others.
It is no use saying that those categories are arbitrary, conventional, fuzzy, or, on the contrary, too sharply bounded or too unrealistic.They do solve practically the problem of extending some standard everywhere locally through the circulation of some traceable document -even though the metaphor of a document might dim somewhat.(Latour 2005: 230.)While we can and often should contest categories, getting rid of categories is not the solution.The solution is to shift our position in regard to categories.The solution is to pay attention to the categorising and localising activities of those we study (and to trace the circulation of some of those categories to sociology departments, others to the rhetoric of a politician or a mullah, etc.).A crucial argument in ANT is thus that it is only when we stop categorising that we can become sensitive to and provide room for the fascinating processes of categorisation among those we try to understand.It is easy to see how relevant this is for the study of religiosity.Rather than arguing among scholars about which category is the 'right' one to use, it is far more interesting to consider how categories enable individuals and groups to act by positioning themselves in relation to others (via categorisation) and in terms of eliciting responses from those categorised by them.The trap we need to carefully avoid is making the claim that some categories are 'socially constructed' while others are real, thereby becoming sidetracked from describing the interplay among multiple and often competing productions of standards, of 'value meters' , of 'downloadable individualisers/subjectivators' .Social explanations are thus relocated into the traceable 'making and dissemination of standards' (Latour 2005: 240), along with religious ones.From an ANT perspective it is not interesting to ask whether or not a certain proposition is true.What is interesting is what difference it makes-how it mobilises other actants.This point is crucially important for the study of religious practices (as inextricably entangled with non-religious ones).We should not distinguish between 'fact' and 'fiction' , but focus instead on how both facts and fictions circulate and therefore participate in structuring processes.
So the point is not a humanist argument about being outraged at 'pigeonholing' , but rather a point about how a focus on attachments reveals that it is not a question of whether or not we are 'attached' ('enslaved') or 'detached' ('liberated') but of whether the 'ties that bind us' serve to provide us with subjectivity or if they 'subjectify us': It is not the case that some powerful people unfairly 'pigeon-hole' other people whose 'ineffable interiority' is thus ignored and mutilated; rather, the circulation of quasi-standards allow anonymous and isolated agencies to slowly become, layer after layer, comparable and commensurable-which is surely a large part of what we mean by being human.This common measurement depends, of course, on the quality of what is transferred.The question is not to fight against categories but rather to ask: 'Is the category subjecting or subjectifying you?' (Latour 2005: 230.)The strings, linkages, plug-ins, 'individualisers' , 'subjectivators' , 'competenceproviders' , and all other actants that circulate in us (as actor-networks) are, in short, enabling as well as constricting.An 'individualiser' can also transform from being primarily enabling to becoming primarily constricting.Being 'attached' is thus necessary for action-indeed for survival-and the actual challenge of 'liberation' is to liberate ourselves (as individuals as well as scholars) from the notion of mastery; the view of causality in which one thing can fully determine another.In networks, an actant is made to act by others, but they do not determine that action, rather, they elicit an (often surprising and unpredictable) response, or they displace the previous action of another actant .Thus, . ..freedom becomes the right not to be deprived of ties that render existence possible, ties emptied of all ideals of determination, of a false theology of creation ex nihilo.If it is correct that we must replace the ancient opposition between the attached and the detached with the substitution of good and bad attachments, this replacement would leave us only feeling stifled if it were not supplemented and completed by a second idea, i.e. the deliverance from mastery altogether: at all points of the network of attachments, the node is that of a make-make, not of a make nor of a made.That at least is a different project of emancipation, which is as vigorous as the former but much more credible because it obliges us not to confuse living without control with living without attachments.(Latour 2010: 59.)We do not need to make value judgments about religious beliefs in accounting for the difference they make.Angels, religious principles and slogans, concepts like 'Islamophobia' , crosses and icons make people do things and should therefore be recognised as bona fide actants insofar as they are influential in a given tracing.That is, if we for instance interview someone and they state that they went to church last Sunday 'to feel the Holy Spirit' , we should not posit the 'belief in the Holy Spirit' inside the person who is interviewed.Instead, we should recognise that what made the person go to church was the Holy Spirit.The question is how?What provided the Holy Spirit with that ability?Another question is: 'what path-through which mediators-did the Holy Spirit take in eventually coming to make the person go to church and how did its agency change along the way?' Endeavouring to answer such questions without reverting to a social explanation potentially leads to a 'good description' in the form of information about the circulation of the social.If someone tells us that they are made to feel queasy by the presence of a cross on their classroom wall, we should ask: how was the cross able to make this person feel queasy?If someone says that they refrain from criticising religious groups because they are afraid of being called 'religiophobic' , then, again: how was this concept able to stifle them?What were the actants swarming towards it, making it stifle someone?Or, to use Karen Barad's (2007) vocabulary: what collection of 'things' made up the material-discursive apparatus in which the agency of the cross emerged?What pseudo-agents 'came together' to form an assemblage in which 'agential cuts' within the assemblage provided seemingly independent actants with a more or less specific agency?A central point in ANT is to treat actants as mediators instead of intermediaries.While an intermediary is something that faithfully transports something without modifying it, mediators transform what goes 'through' them; they influence it so that it is 'translated' as it travels through mediators.In some cases this is obvious: depending on the messenger, the message is different (mobilises differently) even if it were stated in the same way verbatim.In order to be able to trace the circulation of religious principles, angels, deities and religious 'objects' and how they translate as they travel, as well as serve as conduits or vectors for other mediators to travel, we need to resist the temptation to keep them 'inside' so-called 'believers' .Only then can we account for what difference religious beliefs make.Whether or not we want to embrace the notion that religion has 'returned' in somewhat novel forms, as some understand the concept of 'post-secularity' , the relevance of religion in the lives of the non-religious or differently religious cannot be adequately accounted for as long as we presume that only 'believers' deal with religious entities (in their 'inner world').The circulation of non-human religious actants in more and more surprising situations testifies to this need to acknowledge them as co-habitants, alongside facts, attitudes, our neighbour's incessant and outrageously loud piano playing, and so forth.They are all actants making us do things, contextualising our agency.

Is religion good or bad?
As mentioned, ANT can be understood as an effort (or many efforts) to avoid the problems with 'context' .'Context' is seen as a cop-out that allows us to neglect tracing the links, attachments and highly varied circulating entities, which facilitate our ever-changing agency.The seemingly never-ending debate about agency vs. structure, macro and micro, has to a large extent been about the uneasiness we feel at rendering individuals as the effects or 'puppets' of 'social structure' , while feeling similarly uneasy-not to mention naive -when talking about 'individual freedom' .Latour addresses this problem throughout the book and in previous (in Law & Hassard 1999) and later (Latour 2010) works, insisting that it is because we have failed to sufficiently acknowledge that 'ties' are both enabling as well as constricting and because we have presumed that an incentive to act comes either from 'outside' or from an 'inside' , or some strange division of labour (part 'inside' , part 'outside') that this problem persists: The gravest consequence of the notion of context was that it forced us to stick to double-entry accounting so that whatever came from the outside was deducted from the total sum of action allotted to the agents 'inside' .With that type of balance sheet, the more threads you added in order to make you act from the outside, the less you yourself acted: the conclusion of this accounting procedure was inescapable.And if you wished, for some moral or political reason, to save the actor's intention, initiative, and creativity, the only way left was to increase the total sum of action coming from the inside by cutting some of the threads, thus denying the role of what is now seen as so many 'bondages' , 'external constraints' , 'limits to freedom' , etc. Either you were a free subject or you lived in abject subjection.(Latour 2005: 215.)If we instead acknowledge that 'you need to subscribe to a lot of subjectifiers to become a subject and you need to download a lot of individualizers to become an individual' (Latour 2005: 216), we can approach religious practices as both 'competence-providing' as well as delimiting without needing to stress either over and above the other.In other words, if we make every single entity populating the former inside come from the outside not as a negative constraint 'limiting subjectivity' , but as a positive offer of subjectivation. ..the former actor, person, individual -whatever its name -takes the same star-shaped aspect we have observed earlier. . . .It is made to be an individual/subject or it is made to be a generic nonentity by a swarm of other agencies.(Latour 2005: 213.)The point is that once you 'subscribe' to a given 'competence provider' (e.g.learn [download] academic language use in order to participate in academic discourses, or learn to associate 'religiously' with others of your congregation thereby 'becoming' an academic or religious) the actants in the network, in providing you with this competence also make you do things.Even if we were to insist that 'entering' is one's own incentive (as if nothing 'beyond us' compelled us), what one enters is a network that 'opens up' only by making one 'do' things (participate, resist, swerve etc.).But that is true of almost any prolonged 'identity-providing' endeavour (we dream of having a family, career, dog, etc.only to find that when we have it, we are entangled in a web which makes us do things that surprise us).'In doing away both with ungraspable subjectivity and with intractable structure, it might be possible to finally place at the forefront the flood of other more subtle conduits that allow us to become and to gain some interiority' (Latour 2005: 214).Subscribing makes you entangled, but crucially, freedom does not mean 'unattached' .
. ..to possess is also being possessed; to be attached is to hold and to be held.Possession and all its synonyms are thus good words for a reworked meaning of what a 'social puppet' could be.The strings are still there, but they transport autonomy or enslavement depending on how they are held. . ..when we speak of actor we should always add the large network of attachments making it act.As to emancipation, it does not mean 'freed from bonds' but well-attached.(Latour 2005: 217-18.)

Critical sociology
At a recent conference of the European Association for the Study of Religions, in Budapest, Kocku von Stuckrad ended his keynote address by expressing concerns over the misrepresentation of religion by politicians and journalists.He suggested that scholars need to get more engaged in correcting these misrepresentations and oversimplified portrayals.W. F. Sullivan (2005) has expressed concerns about the entanglements between law and religious freedom, by highlighting the difficulties in determining what exactly should be protected by 'the freedom of religion' and who (politicians?judges?) should have the authority in determining that.As soon as we question the logic of using some supposedly reliable measure of 'orthodoxy' as a point of departure, the issues appear inescapably (and predictably) controversial.Latour devotes the last chapter in Reassembling the Social to considering the political relevance of ANT.He does this in part by distancing ANT from what he calls 'critical sociology' .He notes that in this 'school' , social ties are emphasised, but in a way that is highly problematic: I have given [the] label [critical sociology] to what happens when you not only limit yourself to the calibrated social repertoire and leave aside the objects, as the other schools are often tempted to do, but when you claim in addition that those objects are made of social ties.This trend is rendered all the more worrying once the indignant reactions of the actors themselves are taken not as a sign of the danger of such a reduction, but as the best proof that this is the only scientific way to proceed.If the objects of study are made of social ties, namely what earlier social scientists have taken to be part of the official repertoire, and if you cut off the only source of falsification, that is, the objections of those that have been 'explained' , then it's hard to see the compatibility with ANT.Whatever its claims to science and objectivity, critical sociology cannot be sociologyin the new sense that I proposesince it has no way to retool itself to follow through on the non-social elements.When faced with new situations and new objects, it risks simply repeating that they are woven out of the same tiny repertoire of already recognized forces: power, domination, exploitation, legitimization, fetishization, reification. . ..The problem of critical sociology is that it can never fail to be right.(Latour 2005: 248-9.)While it is important that we choose to study matters of concern, our engagement in debates concerning religion should not be about elevating ourselves to the level of experts who alone can correctly determine what religion really is.The political relevance of ANT lies instead in providing accounts which render the entanglements of religious actants with others of concern more accessible.Accounting for complexity by describing it, instead of merely alluding to it, highlights the 'small steps' necessary for the composition and maintenance of powerful structures.What would then make a scientific point of view more credible than that of an uninformed politician or journalist is an explicit focus on mapping controversies, that is, delineating the multiple (and often conflicting) matters of concern in terms of what they are about and how all parties in controversies position themselves in relation to one another.Such perspectives would need to deliberately distance themselves from all attempts at identifying, say, 'enemies' of (modernist) 'progress' .The benefits of such endeavours in terms of positive change and political relevance are evident, while leaving the assessment of necessary 'corrections' to the actors whom the accounts concern, or to activist endeavours outside our academic pursuits.
If what is to be assembled is not first opened up, de-fragmented, and inspected, it cannot be reassembled again.It does not require enormous skill or political acumen to realize that if you have to fight against a force that is invisible, untraceable, ubiquitous, and total, you will be powerless and roundly defeated.It's only if forces are made of smaller ties, whose resistance can be tested one by one, that you might have a chance to modify a given state of affairs.(Latour 2005: 250.)Politicians provide descriptions, journalists provide descriptions and critical sociologists provide descriptions: each of them dramatising their descriptions by condemning some things and cherishing others.What is striking is that there is usually an identifiable ideological agenda.What is the ideological or political agenda of ANT?To provide (con)testable, traceable and interactional descriptions without additional, superimposed explanations referring to non-testable 'forces' , thereby opening up the processes so that, if need be, they can be altered (instead of merely bemoaned).The point is not that the existence of 'hegemonies' or oppression is disputed or not taken seriously.The point is that invoking 'oppressive mechanisms' and 'domination' as explanation foregoes the work of describing the multitude of productive (or producing) controversies involved.Invoking them unduly mystifies problems (by insisting on a force that is far too large and impossibly abstract to combat), while detailed descriptions provide multiple points of intervention.
If there is no way to inspect and decompose the content of social forces, if they remain unexplained or overpowering, then there is not much that can be done.To insist that behind all the various issues there exists the overarching presence of the same system, the same empire, the same totality, has always struck me as an extreme case of masochism, a perverted way to look for a sure defeat while enjoying the bittersweet feeling of superior political correctness.(Latour 2005: 252.)While we have thus tried to argue with Latour that referring to 'social forces' as explanation entails introducing 'ghost-actants'-actants which are part of the description only because the researcher has inserted them there-thereby rendering descriptions messy (mixing up traceable interactions with nontraceable ones), the question becomes: OK, but what then should we focus on?There has to be some differentiation between more significant and less significant locales for research.While someone might agree that the social is what circulates, he/she might easily add: 'So what?We can't research everything!We have to delimit our research somehow.Doesn't it matter what we study as long as we refrain from explaining with predefined frameworks?' The answer is that yes, there are 'locales' which appear to be more significant than others for understanding the distribution and circulation of actants which are of specific concern due to their entanglement with religion.The point is that by resisting the temptation to presume that we already know where those are we can become open to finding that their distribution follows surprising paths and that they are especially dense in often surprising nexuses.In a similar vein as Donna Haraway's emphasis on 'situated knowledges ' (1988), Latour encourages researchers to find those locales where perspectives and templates for practices are generated and in which they make most sense.

Centres of calculation and oligoptica
Latour speaks of three moves that are necessary for the exploration of how and by whom, or what, significant connections are made and sustained.The first move underlines the importance of particular geographic locations (2005: 173-90) and aims at identifying the principal structuring and ordering sites.These are locations that are by some various means densely connected to a number of other locations, and through which a significant number of actants travel.Through the connections the locations offer focused but narrow views of the connected whole.Their function localises the global through the action of the actor-networks, or, actants that are connected to other, remote, actants.Latour describes these places as centres of calculations (1987: 215-57) and later oligoptica, contrasting them with Foucault's panopticon (2005: 181).They are not places from where one can observe everything, like a panopticon.Rather, they are places that see a little, but, because of the focused interest and participations of the connecting actor-networks, what they see, they see well.As long as the actor-networks maintain the connection through continued activity, the connections are stable, and through these narrow connections these sites not only connect with other sites, but also structure the connected whole by way of translating it.The translations that are performed in these sites are innovations, based on anticipation-a form of calculation-of the reactions of other actants to the materials circulated through the site.
Latour's second move 'transforms every site into a provisional endpoint of some other sites distributed in time and space; each site becomes the result of the action at a distance of some other agency' (2005: 219).These distant agencies influence the translations and the purpose of this move is to track the practices by which they hold that influence.Agency, then, is an ability to initiate and/or influence the calculations occurring in the centres of calculation.However, as John Law points out, this agency is tangled in a dense material-semiotic network, from which a detachment of individual actants is impossible (2004: 68).Because of this entanglement, agency is an effect of the action of multiple heterogeneous networks.Further, what is seen as subjectivity is in the same way a result of translations performed by such networks, or, to quote Latour, 'subjectivity is not a property of the human soul but of the gathering itself ' (Latour 2005: 218).What the ANT model emphasises is that translation is a contingent and localised process-the anticipations are made there and then-and because of this the product of translation is variable and stable only insofar as all the parameters affecting the translation remain the same (which may not in fact be entirely possible).
Finally, even though translation implies transformation it also implies equivalence and representation of something remote by something local.Hence, translation also generates structure and order by generating devices, agents, institutions, or organisations (Law 1992).The third move recreates or reconstructs the kinds of orders and structures the translations aim at establishing.Because the processes of translation tend to create spatial ordering, the final ANT move produces a map of these orders, including the locations of the sites where the ordering takes place and the topography of the conduits through which these locales are connected with each other.The ANT strategy proposes abandoning container metaphors when exploring the social world, positing that the laws of the social do not gather, encompass or cover; they circulate, format, standardise, and coordinate (Latour 2005: 246).So, in the reconstruction of the orderings the aim is not reconstruction, but instead the depiction of mobility, the conduits, and the points of translation.

Conclusion
In post-secular societies-after secularisation-it may increasingly be the case that the connecting and structuring of religious matter is done outside designated religious sites and without appointed religious experts.The centres of calculation have changed and so the connections between these are differ-ent.The former ways of translation and ordering are transforming into new ones.By exiting the designated sites religious matter has found new freedom with the new associations and inventions in the processes of translation.Less control leads to more heterogeneous agencies and facilitates the mobility of religious materials.This less controlled mobility of religious actants can also produce an apparent increase of religious matter, but this does not necessarily mean the return of religion.In any case, this increased plurality combined with increased mobility calls for perspectives which can recognise novelty, and not just in comparison with previous states of affairs.ANT does not provide easy solutions, but assists us by clarifying what it is that we need to avoid so that we can proceed with regained confidence, even after poststructuralist critique.