By Michel Callon, Pierre Lascoumes, and Yannick Barthe
While looking up another of Callon's books recently, I ran across this 2010 publication and bought it immediately. Callon has not been as well read as Latour, at least on this side of the Atlantic, but I've found his insights into actor-network theory to be extremely valuable - particularly the question of how groups form around controversies. That, in fact, is what this book is about.
In Acting in an Uncertain World, Callon and his coauthors tackle the question of how to make democracy work in an increasingly technical, specialized world. "Science and technology cannot be managed by the political institutions currently available to us," the authors charge (p.9), because scientific and technical controversies are becoming increasingly specialized while impacting more people. The general public don't have the expertise to make decisions about highly technical issues, but we also can't leave high-impact decisions up to specialists in a democracy. Such decisions include "GMOs, BSE [a.k.a. "mad cow disease"], nuclear waste, mobile phones, the treatment of household waste, asbestos, tobacco, gene therapy, genetic diagnosis" (p.9) - and we could also throw in anthropogenic global warming, a controversy that is the current Ground Zero of the problem Callon et al. discuss. Examine the discourse around that issue and you'll see the same dilemma: the public is impacted by any decisions made about global warming (particularly the rapidly industrializing Third World, which is only now realizing the basic economic benefits of industrialization, including food security and lengthening lifespans), but only specialists are deemed to have the deep expertise to make informed decisions. This skew, this disjuncture between expertise and democratic decision-making, has resulted in (among other things) hard and sometimes unfair questions about the nature and motivation of that expertise (see Climategate).
(Side note: Climate change is one of Yrjo Engestrom's examples of a "runaway object," an object(ive) of multiple activity systems that transcends and can't be contained by any particular activity system. Reading Callon's book helped me to rethink Engestrom's argument - and to wonder whether a "runaway object" is actually an attempt to introduce a notion of the public into an analytical theory that has no other mechanism for dealing with the public. More on this later, I'm sure.)
But Callon et al. don't want to replace our democratic institutions. Rather, these institutions "must be enriched, expanded, extended, and improved so as to bring about what some would call technical democracy, or more precisely in order to make our democracies more able to absorb the debates and controversies aroused by science and technology" (p.9).
In particular, Callon et al. are interested in
hybrid forums - forums because they are open spaces where groups can come together to discuss technical options involving the collective, hybrid because the groups involved and the spokespersons claiming to represent them are heterogeneous, including experts, politicians, technicians, and laypersons who consider themselves involved. They are also hybrid because the questions and problems taken up are addressed at different levels in a variety of domains, from ethics to economic and including physiology, nuclear physics, and electromagnetism. (p.18)
Callon et al. take some pains to distinguish the key issue here, uncertainty, from the "false friend" risk (p.19). In a nutshell, risk is "a well-identifiable danger associated with a perfectly describable event or series of events" (p.19). We know what might happen, and we can usually estimate or calculate its probability and develop contingency plans for dealing with it. Risk implies certainty about conditions (p.20). But uncertainty involves, as Donald Rumsfeld might say, the "unknown unknowns": or, "we cannot anticipate the consequences of the decisions that are likely to be made; we do not have a sufficiently precise knowledge of the conceivable options, the description of the constitution of the possible worlds comes up against resistant cores of ignorance, and the behavior and interactions of the entities making them up remain enigmatic" (p.21). Or, "We know that we do not know, but that is almost all we know; there is no better definition of uncertainty" (p.21). And in these cases, "the only option is questioning and debate, notably on the investigations to be launched. What do we know? What do we want to know? Hybrid forums help to bring some elements of an answer to these pressing questions" (p.21).
Callon et al. take the position that "controversies enrich democracy" (p.28); but for that to happen, we must use controversies as a mode of exploration. They allow us to see overflows, places where proposed solutions jump their defined parameters, "give rise to unexpected problems by giving prominence to unexpected effects" (p.28). "Each decision-making process requires a work of opening out, of diffusion, if only because of the need to mobilize the actors who will enable the project to be brought to a successful conclusion" - and "deciding is opening Pandora's Box by permitting actors previously held at arm's length to take part in a dynamic to which they quickly contribute" (p.30). Such sociotechnical controversies reveal the stakes and makes the network of problems visible and debatable (p.31).
An aside here: In Network I claimed that ANT doesn't really have an account of learning. But here, we (sort of) have an ANT book on learning and development - which is to say, a book on setting up dialogical spaces of hybrid forums in which controversies are defined and clarified so that participants can explore and learn (p.35). This learning "leads to the discovery of mutual, developing, and malleable identities that are led to take each other into account and thereby transform themselves" (p.35). Hybrid forums bring together specialists and laypersons, citizens and representatives (p.35).
Hybrid forums, in fact, sound a bit like Robert Jungk's future workshops, which partially inspired participatory design. These themes are familiar, but applied to scientific and technical controversies (rather than existential threats, as Jungk's work was) across societal scale (unlike PD).
To better understand hybrid forums, Callon et al. take us through the ANT concept of translation. In this context of scientific and technical development, translation involves three phases:
- From the macrocosm to the microcosm: specialists reduce the world to the microcosm of the laboratory so that they can simplify, prune, and reconfigure it for study (pp.48-49).
- In the small world of the lab: specialists turn the phenomenon into its traces and signatures (inscriptions), a chain of equivalences that can be manipulated and studied (pp.51-59).
- From the microcosm to the macrocosm: specialists return the laboratory results back to the wide world (pp.59-68).
Certainly this is a huge issue. The last several decades have worn out public faith that scientific and technical change will be universally good, and that faith has been replaced by the wary question: how will this affect us? As Callon et al. put it in the next chapter: "There is nothing ... more rigorous than a group of non-specialists who want to know why they endure unbearable misfortune" (p.80 - in the context of a community enduring unusually high rates of cancer). So we begin to see the public, or public groups, engaging in all phases of Translation: from translation 1 ("taking part in the formulation of problems," p.76) to translation 2 ("taking part in the research collective in order to broaden and organize it," p.83) to translation 3 ("turning back to the world," p.89). In each translation, we see citizens engaging dialogically with the scientific process. Take the question of whether sheep in England are being affected by the fallout from Chernobyl.
Despite the scientists' fine self-assurance, and maybe even because of it, the shepherds remain skeptical. First, because the specialists have already been wrong once and it does not seem unreasonable to think they could be wrong again. The sequel proves moreover that their fears were well founded: some months later the experts recognize that the observed radioactivity is half due to Chernobyl and half to what are discreetly called "other sources." Later because a serious analysis would have required data from before 1986. Now, despite the farmers' and their representatives' repeated demands, these data were never supplied, the administration finally acknowledging that they did not exist, implicitly admitting that it had not done its work. The cocktail of arrogant certainty, a background of secrecy, and poor work could only arouse the non-specialists' mistrust. In fact, in the farmers' opinion, the most serious thing is not so much that the experts made mistakes, or even that they botched their work, but clearly that they hid all this behind a self-assurance deriving from their status as scientists or experts. (p.92)
In this story from 1986, we might find ourselves cheering the shepherds, who stood up for their way of life, demanded accountability, and exposed scientists' secrecy, sometimes-shoddy work, and overreliance on authority. But change the date to 2009, change the issue from Chernobyl to global warming, change garden-variety stonewalling to failure to comply with FOIA requests, and change the botches to shoddy data-keeping practices, and you get last year's so-called Climategate controversy. You may or may not be cheering at this point, and you may find yourself listing the ways in which this more recent, "hot" controversy is different - but the basic issue is still the same, whether we are talking about "GMOs, BSE [a.k.a. "mad cow disease"], nuclear waste, mobile phones, the treatment of household waste, asbestos, tobacco, gene therapy, genetic diagnosis" (p.9). In all of these, "science says" is no longer enough when people's livelihoods are at stake or when people no longer trust the gatekeepers. In all of these, political decisions are being made, not democratically and publicly, but in seclusion by specialists. "If translation 1 does not reconstitute the networks of interests, translation 3 will end in failure," Callon et al. conclude dolefully (p.103).
So how do we get past this impasse? Callon et al. discuss various forms of cooperation between secluded research and research in the wild (p.125), and between the aggregation of individuals and the composition of the collective (p.131). These hybrid forums aim for some mix of delegative democracy (conducted by scientific specialists or specialized political representatives) and dialogic democracy (conducted collaboratively, by all interested parties) (p.135). Chapter 4 describes the basic ideas behind the hybrid forum, while Chapter 5 discusses how to organize them, including criteria and comparisons with similar efforts.
In all, this book was extremely valuable. It moves toward a macrodevelopmental, proactive approach to ANT that has too often been missing. It hints at an ANT account of learning. It moves in the same general direction as some of Latour's later work, but in more concrete ways. It provides a methodology for developing hybrid forums, and in doing so, helped me to connect ANT with some of the more politically oriented work in other literatures. It provides a possible point of connection with Developmental Work Research and with participatory design. And it's written in a riveting way. If you have interest in any of these aspects, please do pick up the book.
Clay: Thanks for this veritable feast of food for thought in this review and the other four new ones! Thanks also for adding more dimensions to my appreciation of ANT and your transparency about your varied struggles with the texts. You have a wonderful sense of how to make elusive understandings more accessible.
ReplyDeleteOn ANT's lack of an account of learning - I've been wondering if ANT dismisses learning from the same premise as Latour's refutation of sociology, princely accumulation of power, and scientific research in isolation. I suspect those refutations spring from Whitehead's injunction to avoid misplaced concreteness. Making a thing of learning disconnects it from intermediaries and our continual exploration without limiting preconceptions. Perhaps ANT could integrate learning (as a verb) much as it uses verbs like translating, tracing and transforming. To tie into to your insights here, learning would occur in dialogic spaces, when interrogating and interacting with any actor or actant. That infers that I'm learning as I write this comment and you're learning as you read it :-)
Aw thanks. But now I will have to put Whitehead on my reading list! I think I can fit him in sometime in late 2011 or early 2012 ...
ReplyDeleteChill dude! Googling "misplaced concreteness" will be enough of dose of A.N. Whitehead for us mere mortals. Meanwhile, I've ordered Acting in an Uncertain World through the inter-library loan system here in Colorado. As I reread your exploration of the book, I realized I have to read it pronto. Thanks again, Clay.
ReplyDeleteHi Clay - I have read your posts off and on, and each time I have found them so useful and wonderful. Thank you.
ReplyDeleteWith the warning that I am still a PhD student and have read little and know even less, I want to venture that 'learning' might imply a 'correct' thing - nugget of knowledge - already out there, whereas ANT focuses on relational ways in which something comes to be seen as 'correct' or 'knowledge' comes into being...
Thanks, Graham. I think you're right that ANT doesn't lend itself to a trajectory of learning (although I wouldn't characterize learning as conveying nuggets of reified knowledge in any case). At the same time, Callon et al do seem to be pointing us to a process or trajectory of translations that lead to developing a mutual consensus. Maybe "learning" isn't the right word - development? Clearly I'll have to revisit this question!
ReplyDelete'Learning'has become really problematic to understand. Having taught in schools for nearly three decades, this is seriously discombobulating! I was really interested to hear Estrid Sorensen present her work on the materiality of learning - where she talks of learning being not only relational but in the relations - but I am still worked up about who or what learns.
ReplyDeleteJasanoff's notion of civic epistemology works as societal learning in the sense you are talking about (mutual consensus) - but what if this mutual consensus was a really detrimental, a poor choice?
When "learning" is used as a noun, it's clear that learners are learning and their teachers, tools and friends & relations are not learning. We're acting in a certain world where things can be concretized and objectified without losing validity, credibility or value. We learn but our tools and other mediators do not. Mediated activities are a one way street.
ReplyDeleteWhen "learning" is used as a verb, these immaterial processes encompass other processes which may appear as teachers, tools and related individuals. Through the interrelations, everything is getting mediated together while evolving separately, not only the objectified learners. We're acting in an uncertain world where concretizing and objectifying things results in the loss of validity, credibility and value. The processes that can be trusted are ephemeral and elusive to empirical reductions. The outcomes of those processes emerge from complex, non-linear interactions which sponsor more wonder, curiosity and discoveries, not more certainties.
That's how I see learning in my ANT-informed, relational worldview that has evolved with Clay's very helpful writings.
Does this mean 'being' or 'acting' in an uncertain world is the same as 'learning' in an uncertain world? Or does learning imply some other specialised concept/activity/change?
ReplyDeleteIn my thesis (under examination as we speak), I conceptualised what goes on through the processes of translation, negotiation etc - in other words, relational processes - as 'knowledge practices'. I was studying policy assemblages - and that made eminent sense. Where I trip up is when I think of 'learning' on the level of an individual (image of me working one-on-one in a classroom with a kid struggling with long division comes to mind - I spent many years as a primary school teacher) - but of course, the moment I realise that 'individuals' are collectives, things begin to clarify (momentarily).
Thank you very much for the really helpful clarifications, Tom.
@Radhika- I like your connecting of translation processes with knowledge practices. I presume there is often no learning, translating or mediating processes occurring when we are being or acting in the world. I've written much on my blog about "closed minds" as a metaphor for those instances. Within that metaphor, learning does not result from opportunities to learn, but from receptivity to those opportunities. Receptivity is far more likely when we imagine living in an uncertain world inviting our wonder, curiosity and further discoveries than living in a certain world of facts, figures and objective evidence.
ReplyDeleteSimilar to your saying "individuals are collectives" I imagine individuals as getting translated (mediated, negotiated, transformed, contextualized, etc.) by immediate and remembered surroundings. The student struggling with long division is getting influenced by the particular seat in proximity to other students amidst varied activities, sounds and sights. The experience of learning long division individually could be very different next to a fountain in a plaza, on a log in a forest, in a chair at home or inside a circle where the class is observing the process.
I hope Clay weighs in on this too!
Sorry to weigh in late, all, I've had my hands full with other things today. Looks like you had a great discussion without me, though!
ReplyDeleteJust a few quick comments. We can think of learning in at least three ways.
One is the what we might call the Nurnberg Funnel model, the "drill-and-spill" model that represents a wholesale transfer of reified knowledge from one mind to another. It's best illustrated in the way Neo "learns" Kung Fu in The Matrix, by having skills directly implanted in his mind. This notion of learning is what Graham reacts against (I think). It's asocial - and as far as I can tell, untenable.
Much more popular these days is a more social model, based on folks such as Vygotsky. In this model, learning is essentially and intensely social, culturally embedded, and not reified. Barbara Rogoff's work provides a good example of this notion of learning. It comes in various flavors, some of which tend to assume an unproblematic paternalism, some of which expressly critique such paternalism. This model underpins much of the work in activity theory.
So far so good. But when we get to ANT, the focus is very different. In fact, Latour famously became fed up with asymmetric cognitive explanations for human abilities, and called for a moratorium on cognitive explanations until other ones had been exhausted. (I forget the cite, but it's somewhere in my book Network.) A few years later, he enthusiastically reviewed Hutchins' book Cognition in the Wild, which interpreted cognition within an entirely symmetrical account (with cognition stretched across humans and nonhumans).
So here, it seems that an account of ANT-based learning (as some sort of cognitive development on the individual or social level) is really not tenable. That doesn't mean that Latour thinks people don't learn. It means that ANT simply doesn't have an account of that learning, because an account of learning is not terribly useful, and probably interferes with, the account that ANT is trying to provide of how activities happen. ANT doesn't have a theory of learning for the same reason that it doesn't have a theory of color or gravity. Or at least that's more or less what I concluded in Network.
But Callon et al. do seem to be actually talking about learning. And not just in an abstract systems sense, but actual human beings, getting together, finding out what scientists say and why they are saying these things, with an endpoint of being able to understand, contend with, and amplify or critique those things. This account seems to be modeled on a well known ANT-described process - translation. So I wonder: can translation anchor an ANT-ish theory of learning? Can translation prescribe rather than simply describe? Or are Callon et al. pushing beyond the until-recently-accepted boundaries of ANT? I'm not sure of the answer, and I wonder if I am reading too much into Callon et al's account. But I like the direction the discussion is going!
Hi Clay - thanks for this - I am new to blogging but I find the format allows for specific focus and pointed discussion which is very good - though of course there are all sorts of tangents that present themselves as well.
ReplyDeleteLOTS of questions came up after reading Tom's post - for instance, how can one tell is learning has occurred or not? I feel that exapnding the notion of individual to take in the 'surrounds' is not adequate - I think I am much more comfortable with collectives working out how to go on together or solve a problem etc. (a la Stengers' ecologies of practice). That is the sense I got when doing my work on policy - statisticians and policy makers and others try to work out ways to get a grasp on their worlds and make sense of things and look for answers to policy problems. I think a clear expalanation of this is in Jasanoff's 'civic epistemologies' - collective understandings of what counts as authoritative knowledge, what kinds of proof are acceptable and trustworthy, what kinds of accountabilities ought to be demanded and so on.
Ah, now I have to read Jasanoff and probably review Stengers. So many things to do.
ReplyDeleteTom has a nice elaboration of his thoughts on his blog too. With puppets!
I would think an ANT approach to learning would begin with an ANT theory of thought, which in turn would not be limited to humans. Tom mentioned Whitehead, which would perhaps take us toward panpsychism, but I don't think that is necessary. Instead, following on the distributed cognition model, I would look for thinking (and hence learning), like agency/activity, to emerge from relations among actors.
ReplyDeleteI suppose it also depends on what is meant by learning, which is a rather nebulous term. Sometimes we use learning to mean something like sensing or recognizing. One might also call the laboratory activities so well accounted in ANT as learning activities. If learning refers to a classroom or other institutional/programmatic setting then one is looking at a fairly complex network, but I would think its activities could be approached in an ANT-like way.
@Clay Thanks Clay! Good luck with your SXSW gig on Friday.
ReplyDelete@Alex Thanks for giving us some ANT food to chew on. As I've read Latour's Reassembling the Social, he regrets his prior emphasis on actors/actants. He is seeking to flatten all privileged positions and asymmetric relations. That revision tells me to render actors/actants as insignificant as if that stops making something out of nothing. It's in this sense that I invoked Whitehead. Latour suggests viewing power dislocated from those "in power" and expertise (sociology, pedagogy, etc.) as evenly distributed among actors. Then a figure/ground reversal can occur where what's in between becomes the significant figure and objects become insignificant background. Apprehension of those interrelations begins and ends with wonder, not with categories, frameworks or hierarchies. Process becomes something while misplaced concreteness (reifications, delusional constructs, labels, etc.) gets rendered as nothing. Learning could only a verb that appreciates distributing cognition without a model of distributed cognition or framework of learning as a thing/noun.
Hi Tom - interesting - not my reading at all. We may wish away differences in power etc but I think without differences in power, there would be no assemblage at all. Besides it would be a great disservice and offence to those who suffer the consequences of differentials in power to insist that there are no differences in power. The criticism of the 'first wave' ANT (criticism I don't agree with) is that it focused on 'those in power' as in Pasteur - that the 'actors' that were 'followed' were the 'powerful', at the cost of the 'silent' or worse, 'silenced'. Later ANT studies focused on multiple modes of ordering rather than merely the one that is privileged. I don't see Latour at all as denying that inequities occur in power - what he says, I think, is that power in not an inherent characteristic of any actor - it is aquired relationally. I am not sure at all about this idea of 'focus on the relations and ignore the actors'- without actors there is no assemblage. The concept of 'wonder' to me invokes individual cognition and I cannot relate it to any ANTish ideas.
ReplyDeleteI can see that this conversation will continue awhile!
Hi Radhika - I wonder how you could say such a thing? By my wondering like that, I'm likely to discern the assemblage which translated what I wrote into what you wrote. You must have eaten some food that gave you the energy to think what to say and then to type your thoughts. What you ate is an actant and insignificant in my view. Giving you energy is a translation within an assemblage -- which I deem significant. The food exhibits asymmetric power as none of the other actants in your assemblage can give you this metabolic energy. You must have fingers and a keyboard or voice recognition software -- to have typed out your thoughts into these words. The translation of your thoughts into words on a screen is significant. What actants enabled that translation is not. The food has inferior power in the context of getting your words typed out on a screen. You must have an Internet connection for what appeared on your screen to appear here. What connection you use is another insignificant actant. Translating words on your screen to this remote location is very significant in my view. The power to transmit your words digitally is also asymmetric with food and keyboard/software which are powerless to get that done.
ReplyDeleteNow I'm wondering how I could say such a thing in response to what you wrote? To wonder about my own assemblage is symmetric with yours. Neither of our assemblages, inquiries or words are privileged over the other. My learning increases significantly when my wondering extends beyond a seemingly isolated instance. Yours may also. Wondering translates unknowns into inquiries and knowns into expanded possibilities.
I discern my similar assemblage of insignificant actants and significant translating to show up here with these words. I wonder where your assemblage ends and mine begins and whether the two are really disconnected when translating becomes the foreground to the background of separate food, computers and Internet connections. I wonder what will come of having seen and said all this? My food, keyboard and ISP may be wondering the same.
Thanks for keeping this going!
Hmm. Could we not see the 'significant/insignificant actant' as the difference between a mediator and an intermediary? 'Wonder,' it seems, is your way to describe how we apprehend/understand what has occurred - I have to think how this sits with ANT's performative/co-production accounts...
ReplyDeleteHi Radhika: Yes to how I'm using "wonder".The phrase "begin and end with wonder" is a Whitehead quote I found in "The Prince of Networks - Bruno Latour and Metaphysics" by Graham Harman. Being free of Modernism and Post Modernism leaves one open to trace connections without preconceptions, separations or models. Wonder seems like a great word for that receptivity.
ReplyDeleteThere is no difference between mediators and intermediaries in my mind. Teach me what you're discern with that difference. The possibility of nodeless networks linking empty places has captured my imagination and "insignificant actants" fits that possibility. Insignificant actants also fits my shallow grasp of Whitehead's advice to avoid the error of "misplaced concreteness". It also fits the premise of Gestalt psychology's reversing figure/ground or foreground/background to replace chronic problems with sustainable solutions. Within those contexts of mine, "significant actants" become Modernist, components of networks linking enduring objects, misplaced concreteness and a gestalt for chronic problems.
I look forward to your next thoughts.
Tom - excellent - very thought provoking. Latour (in Reassembling...) says something like 'intermediaries just sit there and do nothing' - in that they don't translate or change anything. Mediators, on the other hand, do things and leave traces.
ReplyDeleteI think the issue with significant/insignificant actants is a temporal matter - significance (or insignificance) being not inherent properties of actants, but empirically realised, emergent properties.
Thanks Radhika! Now that you mention it, I had pondered that distinction between mediators and intermediaries deeply. I eventually dismissed it as inappropriate for ANT which is why it's no longer in my mind. In my view, those non-mediator intermediaries have assemblages of interests in being that way, have been enrolled in not mediating, got translated by mediators or intermediaries to be non-mediators, etc.
ReplyDeleteTo dismiss them as impenetrable black boxes is, in itself, functioning as an intermediary. A genuine mediator would transform intermediaries by enrolling them in mediating by showing an interest in their assemblage of interests, not by slapping an " Danger- Un-mediator Inside" label on their black box.
I wasn't considering insignificance as an attribute of the actants, but rather a characterization of my perception (map, text, sociology etc) - which is, as you're saying, emergent and temporal, but facets of my outlook.
Okay - now I see where we are at cross purposes. You are focused on how you understand phenomena, on the process of understanding, on your perceptions, whereas I am focused on the phenomenon itself. I don't see the act of calling some actors intermediaries and mediators as 'slapping a label' thing at all - just a matter of tracing the emergence and understanding the participation of different actors in assemblages how things emerge in assemblages. I don't think intermediaries are dismissed as 'impenetrable black boxes' at all, but rather the understanding of their role as 'intermediary' is arrived at through empirical work. I don't see this as any different to your label of 'insignificant actor'. Curious how different labels grab us!
ReplyDeleteThis helps me a lot too. Your characterization of my approach seems accurate to me. My label of "insignificant actant" ( for your fingers, food, or keyboard for instance) is used in the context of significant translations, mobilizations, mediations or transformations. The foreground is what process comes between the backgrounded things. I run much deeper in psychology than sociology and that may distance me from objective phenomena and fascinate me with subjective perceptions, understandings and interests. This helps me sort out the strange way (in my view) that Latour characterizes enrollment.
ReplyDeleteThanks for all these clarifications, Radhika.