orgtheory.net

you probably shouldn’t mention bruno latour to ezra, definitely not

I  noticed an issue flare up over the last two months – the hatred of science and technology studies (STS). This first popped up when Nick Rowland was guest blogging. He wrote a post about technology as a sort of field of action rather than a specific object, a new idea is STS. Ezra responded with this comment:

Hmm… with all due respect, seems to me that I should be able to understand this post. But I cannot. Then again, I confess not to understand the prior “waves” of STS either. In general, it seems to me that one should evaluate a theory (if that is what STS practitioners produce) based on whether it can it account for more facts with as minimal a conceptual apparatus as possible, where such an apparatus obviously needs to be coherent. But STS seems to be more trouble than it is worth on all three of these scores– (a) not clear how it does in accounting for facts (in part because practitioners generally seem allergic to establishing facts independent of theory [of course, facts are theory-laden but there must be some establishment of facts independent of theory if the theorist wants to convince us to accept her theory rather than some other); (b) proliferation of concepts [often with an obvious attempt at sexy labels; e.g, ‘agora’; ‘mangle of practice’]; and (c) [consequently] very hard to understand. I confess though that I do not read much STS because it long ago seemed to fail miserably, at least by criterion (c). Though I’m open to the suggestion that I’m just not smart enough to understand STS theory or that there is better work that I’m not aware of. (I’m not particularly open to the possibility that my criteria are inappropriate). FWIW.

Later, in a discussion of performativity, Ezra also played the skeptic:

But I did not hear why the emerging research you described does any better. Is there something that it can explain that existing theory cannot?

So what is it about STS that is so infuriating? I think it’s this: Most STS folks openly reject the normal science model where you stick with a variable or issue and then work on it. The normal science model allows you to clarify, simplify and then expand. Think “human capital” or “structural holes.” It’s easy for outsiders to quickly grasp the concept and for newcomers to use it.

STS works with a continental model of knowledge – thick description and, as Ezra points out, proliferation of concepts. Rather than hammer a variable home, they jump from interesting case to case. They revel in new phenomena. It’s very ecological – new empirical observations suggest new ways of combining concepts, creating a mangled web of theory. Of course, this is very anti-normal science. You end up with a cabinet of curiosities than a deep and precise knowledge of a specific issue.

I’ve always been tolerant of STS, I was on Nick’s dissertation committee after all. There’s something deceptive about technology and science. When it works, it looks simple and it’s easy to take it for granted. This has often been true of academics who study science, especially in the pre-Kuhn/Latour era. The odd concepts of STS shake me out of my complacency and see things in a new light.

Advertisements

Written by fabiorojas

July 5, 2010 at 12:20 am

44 Responses

Subscribe to comments with RSS.

  1. This debate is also very active in Management Science and Information Systems Research. For an overview see, for example, Orlikowski, W.J., and Scott, S.V. “Chapter 10: Sociomateriality: Challenging the Separation of Technology, Work and Organization,” The Academy of Management Annals (2) 2008, pp 433-474.

    I, usually, tend to stick more to the “not”-sociomaterialist, for example, Markus, M.L., and Silver, M.S. “A Foundation for the Study of IT Effects: A New Look at DeSanctis and Poole’s Concepts of Structural Features and Spirit,” Journal of the Association for Information Systems (9:10/11) 2008, pp 609-632.

    “Sociomateriality” is an interesting concept, but I have to agree with Ezra: what, iltimately, can I do with it regarding knowledge claims?

    Like

    Christoph Rosenkranz

    July 5, 2010 at 7:45 am

  2. I think part of the problem is substituting ‘STS’ for ‘ANT’. I am an STS-er: I publish in the journals, go to the conferences etc etc, but I have a lot of sympathy for Ezra’s comments as applied to Actor Network Theory.

    While the broad programmatic statements about ANT are intriguing (see Latour’s “Reassembling the Social”) in their boldness, when you get to the specific examples they very rarely deliver in terms of category shattering analysis. They are usually either good, detailed case studies (like lots of other work in more traditional SSK) or over-complex, jargon-laden obfuscation.

    But there’s lots of STS that ISN’T ANT: in fact lots of stuff that calls itself ANT isn’t ANT (it’s just SSK with a sprinkling of words like ‘network’, ‘actant’, and ‘translation’). Reading proper ANT should be a genuinely weird experience (see Callon’s ‘Scallops of St. Briac Bay) since it is meant to challenge our taken-for-granted-categories (Actor/object, nature/science etc). This doesn’t mean that some ideas from ANT can’t be usefully adopted by other approaches (I’m a fan of Law’s ‘heterogeneous engineering’), but a proper discussion of STS needs to acknowledge that ANT is not the dominant approach in this area.

    Like

    Adam Hedgecoe

    July 5, 2010 at 10:34 am

  3. Adam – If ANT isn’t the dominant approach in STS, then what is? I think a lot of us outsiders to the field have begun to equate STS with ANT.

    From my view, ANT’s big problem is its extraordinary insularity. If the only people who are fit to evaluate your theoretical contribution are other ANTers, then there might be a problem with your theory.

    Like

    brayden

    July 5, 2010 at 1:57 pm

  4. I’m going to jump on the hater bandwagon. The “interlude” chapter in Reassembling the Social was really all I needed to show me that STS wasn’t worth the opportunity cost.

    Like

    joshmccabe

    July 5, 2010 at 2:29 pm

  5. Brayden, I think that for many people there’s an overlap between ANT and STS: but looking at Ezra’s provocative post linked to above, it’s interesting to see that s/he’s tlaking about Donald MacKensie who, last time I looked, _did not do_ ANT.

    Personally, I think there’s a lot of ‘Latour-lite’ out there in STS: dropping occasional terms (like ‘network’) and making claims in the introduction and conclusion (about, for example, ‘This paper shows how non-human actants shaped the development of XXX science/technology’) that are not supported by the actual data in the paper as a whole.

    If we’re not doing ANT, what are we STS-ers doing?: some varient on SSK I think, more-or-less constructivist analysis.

    I would be interested to know what people _outside_ STS think are examples of ‘typical’ STS-as-ANT. I sit on the board of Social Studies of Science, and I just don’t see a field dominated by ANT (perhaps there’s an interesting question on the insider/outsider views of an academic [sub-]discipline.

    Like

    Adam Hedgecoe

    July 5, 2010 at 2:38 pm

  6. In Information Systems Research lots of STS-like studies that do not use ANT apply Adaptive Structuration Theory (AST) by DeSanctis and Poole (DeSanctis, G., and Poole, M.S. “Capturing the Complexity in Advanced Technology Use: Adaptive Structuration Theory,” Organization Science (5:2) 1994, pp 121-147). AST builds on Giddens but, admittetly, is controversly discussed, for example, see Jones, M.R., and Karsten, H. “Giddens’s Structuration Theory and Information Systems Research,” MIS Quarterly (32:1) 2008, pp 127-157.

    There is also a very nice recent special issue of the CJE on technology (http://cje.oxfordjournals.org/content/vol34/issue1/index.dtl), which also gives a nice introduction to SCOT (social construction of technology) approaches, for example, the STS reader from 1985 and the work of Bijker as an alternative approach to ANT.

    Information Systems Research heavily adopts social theories in order to study IT effects on organizations and vice versa. For example, MIT’s Wanda Orlikowski is one of the proponents of sociomaterialiy, entanglement in practice etc. A good overview of the usage of social theories in Information Systems Research is Mingers, J. and L. P. Willcocks (2004). Social Theory and Philosophy for Information Systems, Wiley.

    Like

    Christoph Rosenkranz

    July 5, 2010 at 3:19 pm

  7. “From my view, ANT’s big problem is its extraordinary insularity. If the only people who are fit to evaluate your theoretical contribution are other ANTers, then there might be a problem with your theory.”

    ANT at its core is a mix of ethnomethodology, semiology and hermeneutics. If you don’t like ethno you’ll not like ANT.

    Like

    Guillermo

    July 5, 2010 at 3:33 pm

  8. And disclaimer: I’m not a fanboy of ANT either.

    Like

    Guillermo

    July 5, 2010 at 3:34 pm

  9. hmmm…My sense is that what’s “wrong” with STS is what’s “wrong” with a lot of organizational theory. There are no well-characterized phenomena to study (case studies are not phenomena, although they can lead to their discovery). Instead what you see is a proliferation of schematic “concepts.” In science, “progressive” cosmopolitan collectives build on phenomena and are pragmatic towards concepts and models (the one that provides a more satisfactory account of the phenomenon wins); “oppositional” enclave-based scientific movements deplore phenomena and focus on “authorial” concepts. Authorial concepts are those that are inherently associated with an author, who thus “owns it” (e.g. Latour –> actant). Because oppositional enclaves tend to be organized around “groups” and authorial charisma scientists do not compete to offer the best explanation of an explicitly recognized phenomenon (as in the Mertonian ideal type) but instead the incentive structure is biased for leaders of different subtribes to provide their own authorial concept. The reason is that without your “own” concept, you don’t have a reputation.

    Thus, conceptual proliferation in oppositional enclaves is simply the outcome of a more basic underlying socio-structural logic. This contrast to the Mertonian ideal-type collective (which actually attaches names of authors to effects not concepts), which favors the proliferation of different explanations of the same (small) set of (clearly characterized) phenomena. Because authorial concepts, like all concepts, have the general structure (A as X) where X is the concept and A is some empirical setting, “progress” in oppositional collectives follows a “turf-war” model as you extend your concept to cover other realms, in particular realms that were previously covered by your rival’s own authorial concept (e.g. B as X and not Y, C as X and not Y, etc.)

    This is not to say that well-characterized phenomena are absent from STS. Obviously one interesting phenomenon to emerge out of the STS enclave is what has become known as performativity (although Ezra might disagree as to whether this phenomenon is well-characterized), and not surprisingly, this is where STS has been the most influential outside of the enclaves. If you prefer the Mertonian model of science (and I’m going out on a limb here and presume that Ezra–who is a boy–prefers this model) then the good thing about phenomena (as opposed to let’s say “concepts”) is that they don’t carry authorial copyright, so that anybody can take a “shot” at (accounting for) them. Phenomena-centered science thus leads to cosmopolitan, interdisciplinary endeavors that follow more closely the “progressive” line traditionally associated with normal science. Given a phenomenon as a reference point, we can make a pretty good estimate of where we stand in relation to the immediate past.

    Take for instance the phenomenon of “boom and bust” in industries (or organizational imprinting). You’d be crazy (or just uniformed) if you didn’t realize that we are in a better explanatory position in relation to these phenomena than we were let’s say a three decades ago. However, given a “concept” (let’s say “social structure”) no such “progressive estimate can be given, instead it seems as if we are stuck in second gear. Authorial concepts-centered science instead leads to either Spenglerian “coming crises” types of analyses or cynical, “nothing matters, there’s never progress, everything goes in cycles”(e.g. Abbott in Chaos of Disciplines) types of takes regarding the health of a given field.

    It is obvious that STS is a “concept-based” enclave and the “system of thought” that governs most STSers is based on this make-a-“contribution”-by-offering-a-new-concept model. A typical example is Callon, who in a recent paper (2007: 140) argues that “Relying upon the anthropology of science and technology, we can define economic markets as socio-technical arrangements or agencements (STA) whose functioning is based on a set of framings concerning not only goods and agencies but also price-setting mechanisms(1).” Then if you have the patience to go to footnote one we get the definition of agencement:

    An agencement is a combination of material and technical devices, texts, algorithms, rules, and human beings with their various instruments and prostheses. I discuss elsewhere the reasons why I prefer the French term agencement (which unfortunately has no equivalent in English) to assemblage or arrangement (Callon, 2007). Agencements denote socio-technical assemblages when they are considered from the point of view of their capacity to act and to give meaning to action. By defining markets as STAs we emphasize the fact that they are simultaneously malleable and capable of actions.

    At this point, of course Ezra has already thrown the paper in the trash. Callon’s strategy here follows the template offered above (markets as agencement) But the main point is that throughout this paper, you will not find the usual Mertonian goodies; e.g. a characterization of a phenomenon, a specification of our ignorance regarding the mechanisms that generate it and the proposal of a new model of these mechanisms that does a better job of accounting for the phenomenon than other competitors, but simply more definitions, and stylized descriptions of sites that could be thought of as “agencements.” This is the style of thought characteristic of STS.

    We can go hoarse trying to evaluate whether this is or is not a good strategy, but for now I’ll simply play the good anthropologist and say that it is “different” (I have been known to play both sides of this field) but in addition (and now putting on my more specific Mary Douglas hat) it should be noted that it is also exquisitely attuned to the social context within which it makes sense (enclaves where charismatic authority based on authorial concepts is the most natural reputational currency). Since, the main issue after reading this paper is that you come out with the impression that: agencement –> Callon so that you have to cite Callon (2007) if you want to use the concept (the same of course goes for agora, actor-network or what have you).

    Like

    Omar

    July 5, 2010 at 7:44 pm

  10. Omar: Can’t believe you’re letting that sit in the comments…

    Like

    fabiorojas

    July 5, 2010 at 7:52 pm

  11. A related, secondary, phenomenon in authorially-dominated settings is the appeal to some supposed past-master or other as a means of building one’s own reputation. If you can’t get yourself directly associated with a concept, the next best thing to do is to insist the field must rebuild itself by way of some neglected past figure whose work provides a hidden key to Where Theory Went Wrong. Tarde is presently getting this treatment from Latour, as it happens. (Expect a Tarde revival in English a few years from now, as a consequence.) Efforts to sociologically reheat Dewey, Husserl, and even Kant himself have been made by various people recently, too.

    I’m also reminded of this thing on Freakonomics, as the differing competitive dynamics in Econ vs Soc exemplify Omar’s contrast: “Perhaps ironically, economics provides a collective solution to its problems: Reflections on its own foundations are the job of a small cadre of methodologists and philosophers, and the subject of occasional, cathartic ‘We Can’t Go On Like This’ speeches by Presidents of the AEA or Nobel Laureates. Meanwhile, it is business as usual for everyone else. Sociology, by contrast, encourages its graduate students to fret about foundational questions, and competition between theorists takes the form of alternative visions for the whole field built up from first principles.”

    Like

    Kieran

    July 5, 2010 at 8:05 pm

  12. Would a Tarde revival resulting from Latour’s entrepreneurial efforts constitute a performative auto-confirmation of The Laws of Imitation?

    Like

    Omar

    July 5, 2010 at 8:16 pm

  13. Exactly. And once Princeton makes that book available as a PDF in the iBooks store (and it thereby becomes an actant), the circle will be complete.

    Like

    Kieran

    July 5, 2010 at 8:22 pm

  14. Omar, really interesting comment!

    Out of curiosity, have you read Laboratory Life? I ask because of the extensive discussion of the creation of phenomena in laboratories and all the politics and processes surrounding such creations. In some ways, the whole book is about a fight between two labs to be the first to successfully produce a phenomenon which will then be taken up in other disciplines and whose authorship will then be ignored.

    Similarly, Ian Hacking (neither an STSer nor not an STSer but mostly just a very smart empirically-minded philosopher) has an excellent book called Representing and Intervening, wherein he argues that philosophers and science studiers alike should focus more on the way that scientists create new phenomena and less on whether or not their representations (theories) are true. What’s fascinating about Hacking’s account, and Latour and Woolgar’s, is that both emphasize very strongly the capacity to intervene with the identification of phenomenona – in other words, phenomena are made, not merely identified. The laboratory is a key site for both lines of thinking precisely because a laboratory is a place where scientists can intervene in the world in very peculiar and novel, but reproducible, ways, and thus produce new phenomena.

    So, let’s start with your premise (which I agree with) that sociology in all its various guises needs to focus more on “characterizing the phenomenon” (ala Merton) and thus on creating theories and concepts to explain pseudofacts. Then let’s add Latour & Woolgar and Hacking into the mix, and their emphasis on intervention over representation, because interventions are how we make new phenomena. How then do sociologists study the world? Is the only option to go all developmental economics, and turn whole villages into your experimental sites, and whole nations into laboratories (an analogy which may or may not hold up well)? Are opinions that are revealed to be consistent in repeated surveys, ala the GSS, examples of phenomena in Sociology?

    (PS: I’m pretty sure that “agencement” and the various other similar terms go way back in French social theory, at least to Foucault and Deleuze – so I don’t think the point of that article is to introduce a new term, though one of its goals is to push people away from one term towards another, but both in reference to a well-established concept.)

    Like

    Dan Hirschman

    July 5, 2010 at 8:47 pm

  15. For me the STS was kind of interesting as a form of “critical sociology” (see Burawoy speech on ‘public sociology’) – questioning the dominant approaches to technology-related explanations. In Ezra’s terms, the question is not so much providing greater economy in explanation but pointing out inherent inadequacy and insufferable simplicity of existing treatments of technology (as something that is not social).
    The field has since gone horribly wrong — becoming a clique which exploit politics of publishing in ways that creates jingostic nonsensical descriptive explanations that do not accumulate with other sociological work, and oftentimes not even within the field (= everything already said about ANT). But to some extent, Andrew Hargadon’s work on Edison and other ‘social engineering’ in relation to technologies builds (however indirectly) on the empirical and theoretical work in the STS tradition.
    STS is at its best like ‘philosophy of technology’ – not really theory but metatheory that helps mainstream sociology to incorporate technology and science into our core theories in a productive way.

    Like

    Henri

    July 5, 2010 at 9:13 pm

  16. Hi Dan, yes, I’m a big fan of Representing and Intervening (quite possibly the best Philosophy of Science book from the 1980s). I have never grappled with the entirety of Laboratory Life, although it (along with The Microbes [aka. “The Pasteurization of France”]) is in my perpetually large queue of social science classics that I should read some day (I have read “We’ve Never been Modern” though). I’m familiar with the standard excerpts of LL from the usual STS anthologies, and I’m also familiar with the countless descriptions of the key arguments from the secondary literature.

    The idea of the production of phenomena through intervention I think was a very important and refreshing way to finally get rid of standard representationalist accounts of the role of theories inherited from normative Philosophy of Science and to strike a deep blow against various forms of empiricism and instrumentalism. That said, I think the early classic STS studies were way too “lab” centered for us to draw any lessons for sociology. For the fact is that (the bulk of) sociology will forever remain a non-experimental science, and outside of Orwellian (or Saint Simonian) nightmares, “intervention” will be strictly limited to (kind of boring) Heckman-type of program-effectiveness analysis. Taking advantage of natural experiments (Freaka-soc), as well as doing more field experiments (a la Devah Pager) will also help in this respect. Also, let us also not forget that there is an experimental tradition in sociology (so-called “group processes”) and there phenomena (e.g. such as the creation of “status beliefs” or the generation of a “minimal group” complete with in-group biases) are “produced” unproblematically and with impunity, just like Hacking would predict. This is also the area of sociology that functions most like “normal science” with over-time cumulation of findings, increasing exactitude in the explanation and the prediction of phenomena, etc.

    But the main point is that there are a lot of sciences that are not experimental (evolutionary biology and cosmology jump out in this respect), and these sciences have been rather ignored in STS (although an early classic paper by Garfinkel concerned pulsar-detection). Stanley Lieberson argued in a recent ARS piece that these sciences (and not experimental Physics) should be the “model” for sociology. Scientific activity in these fields is no less phenomenon-centered than in other fields, so I would question the idea that you have to have direct control over the production of your phenomena to have a phenomenon-centered science. For instance, organizational ecology is a phenomenon centered research program and it is largely non-experimental.

    Also (and this is something that I’ve gotten to observe first hand in my last three years at ND), even when physicists confront non-experimental phenomena (e.g. degree-distributions in networks or small-world structure), they do it in a phenomenon-centered way just like Merton would predict: they characterize first (measure and fit distributions across different types of networks so that the phenomenon is deemed to be of sufficient generality) and then they build generative models to explain it (Barabasi and Albert’s 1999 and Watts and Strogatz 1998, both classics, are the exemplars here). As Joshua Epstein has noted, the phenomenon is deemed explained only when you have “grown it” somehow (economists are still not very good at the growing thing).

    I think the fact that you can identify a “genealogy” of the term agencement on some previous authority is actually consistent with the argument. In Mertonian phenomena-centered science, historical “priority” over discoveries is essential and a lot of acrimony and energy is spent in settling priority disputes (Merton had a famous “functionalist” explanation for this). In authorial concept-centered enclaves, priority disputes are not that important, and in fact (as Kieran notes) connecting your concepts to a distinguished set of previous authorities is encouraged. The main thing is that right now the concept of agencement is Callon’s and he gets to infuse it with whatever contemporary connotations he wants (including claims that you can’t really say “arrangement” because the French word has a certain je ne se quois (whatever that last thing is).

    Like

    Omar

    July 5, 2010 at 11:13 pm

  17. Omar,

    Thanks for clarifying and responding! As a quick follow-up, I think there has been a lot of recent STS work on “non-experimental” sciences, some of which is quite fantastic. I’m thinking here of Naomi Oreskes stuff on continental drift and more recently climate science, and especially Paul Edwards’ new book “A Vast Machine” on the history of climate and weather data and models. Edwards talks at length about the differences between experimental and “historical” (I think is the term he uses) sciences, including cosmology. It’s a great read, and the analysis of what data mean to such a science are useful for sociologists, I think.

    Like

    Dan Hirschman

    July 6, 2010 at 12:02 am

  18. Hmm… I go away for a few days and all of a sudden have an urge to sing Carly Simon (I’m so vain, I probably think Fabio’s post was about me). My main reaction is to just say amen to what Omar (and some others) wrote. And here are a few quick thoughts:

    * In those comments that Fabio quoted, I was really just reacting to the posts. I claimed ignorance of STS (I’m more familiar with performativity) and asked to be disabused of my impressions. I’m still waiting, I guess.

    * I plead guilty to Omar’s charge of being a Mertonian– and especially in a *normative* sense. (I’m not sure that I should plead guilty to his charge that I am a boy though. Sometimes I feel quite girlish! :->) That is, it is fine to take Merton apart as a model of how science actually operates. But what better model do we have for how science *should* operate? All too often, people confuse *is* and *should*. Sure scientists often violate Mertonian norms. But should they/we?

    * Part of what seems corrupt about the intellectual stance that Omar nicely called an “oppositional enclave” is that they do not say that they are playing the game that Omar described. For example, Callon uses the term “definition” for what is in fact obfuscation. And I do not believe that they readily admit that the empirical work is just there to dress up a concept. Rather, the conceit is that they really are providing persuasive evidence for a coherent argument.

    * The last two points lead me to a different place from Omar– that is, I just cannot accept this stuff as just “different.” I simply have no use for obscurantism.

    Like

    ezrazuckerman

    July 6, 2010 at 6:17 am

  19. Omar’s post on “oppositional enclaves” is really interesting, but again I think it’s interesting that the example he gives of ‘STS’ comes from a strong ANT-er (Callon): just because ANT tends to ‘make-a-”contribution”-by-offering-a-new-concept model’, doesn’t mean that the rest of STS does.

    I am alos unconvinced by the claim that STS does not have “well-characterized phenomena” to work on: how about ‘the closure of scientific controversies’, or ‘the public reaction to controversial technologies’, both of which are classic phenomena which have been (over?) studied in STS.

    The picture of STS being presented here is both too homogenous (there are serious differences between traditional SSK and ANT – see Bloor’s “Anti-Latour” for an example) and too fractured (a lot of STS-ers feel that Mertonian norms are held by the scientific community, even if they are often brocken).

    Ezra, I suggest you have a look at Harry Collins’ work (‘Changing Order’ is the best place to start) for an example of non-ANT STS that does not go down the obscurantism route.

    Like

    Adam Hedgecoe

    July 6, 2010 at 9:33 am

  20. Thanks Adam, both for being a voice of sanity and for the suggestion of where to start. Nothing of yours to suggest that we read? :-)

    Like

    ezrazuckerman

    July 6, 2010 at 12:32 pm

  21. Ezra, not often I get called the voice of sanity, but thanks. I think.

    Part of my concern is, I guess, that I have been around long enough to sit through the ‘science wars’ (Sokal, Levitt etc etc): it’s bad enough scientists picking a fight with STS and tarring us all with the same brush (“STS-ers are all wacky post-modern relativists who say there’s no such thing as a fact”). But when fellow sociologists start to say the same thing…

    I’m far too modest (ha!) to reccomend my own work, but do feel there’s a lot of great ideas in the early canonical work in SSK (Barnes, Bloor, Collins etc.) and slightly later work (huge fan of Mackenzies ‘inventing accuracy’) that gets overlooked as ‘part of the background’ but which still has much to teach us.

    Oh yeah, another ‘phenomenum’ examined in STS: The way increasing proximity to the creating of technical knowledge increases unceratinty about that knowledge.

    Like

    Adam Hedgecoe

    July 6, 2010 at 2:19 pm

  22. This is probably not too central, but a quick comment on Omar’s note. It was Greimas (a linguist) and not Latour who came up with the actant concept, no? Latour then just applied it to Science and Technology studies. But now the concept seems to be associated with him. So I suppose it is possible to “own” a concept that was produced by someone else. I wonder how common that is. Maybe very common.

    Like

    Ari Adut

    July 6, 2010 at 3:50 pm

  23. “It was Greimas (a linguist) and not Latour who came up with the actant concept, no?”

    A semiotician, yes. Latour is in fact an intellectual heir of a group of French intellectuals (Greimas among them) who started criticizing structuralism in the sixties. It was structuralism as understood in the French intellectual tradition, not American structural-functionalism of course.

    Like

    Guillermo

    July 6, 2010 at 4:49 pm

  24. So I suppose it is possible to “own” a concept that was produced by someone else. I wonder how common that is. Maybe very common.

    If you believe Stigler’s Law of Eponymy, it is not only common but inevitable.

    Like

    Kieran

    July 6, 2010 at 6:08 pm

  25. As Dan noted above, I’d cast my vote for everyone to page through Paul Edwards’ new book on the evolution of climate science. The book is interesting because really it’s a history of data management and computational modeling, discussing at length the creation of an information infrastructure to collect, analyze, and store reams and reams of climate and weather data going back now over a century plus. If anything, now the discussion has switched from how can we further lambaste STS to how we can help sociology, the book is a go-to manual how to scale up a science from a handful of observations taken at random to empirically rich, predictive science.

    Perhaps the most interesting “finding” is what Edwards calls “infrastructural inversion”, an on-going process of continual introspection and reflexivity over methods of data collection, questioning and investigating at length the relationship between artifact and real variation (for example: the move in US weather stations from airports to less open, often shaded areas created an historical drop in temperatures due to less wind). As Edwards argues, more than simply creating a robust infrastructure to collect and analyze data, this renewed self-doubt contributed greatly to creating new standards and harmonizing the remarkably uneven output of weather stations around the world. One could imagine a more robust International Social Survey Programme operating in a similar way, generating and enforcing new standards of social science data collection, permitting perhaps a truly global science–or at the very least, better cross-national comparisons.

    Like

    dr

    July 6, 2010 at 9:38 pm

  26. Again, this has little to do with the main theme of the post, but here is another example of succesful reappropriation of someone else’s concept: didn’t Bourdieu also “steal” the habitus concept from Elias?

    Like

    Ari Adut

    July 6, 2010 at 11:36 pm

  27. Wow, what an outrageously interesting post and discussion, particularly for the summer! My only contribution is to flag that, whether or not he is “open to the possibility,” I think Zuckerman’s criteria are inappropriately narrow (epistemologically if not ontologically) and therefore his hatred of ANT (to which I am largely sympathetic) is nearly tautological.

    Liked by 1 person

    andrewperrin

    July 7, 2010 at 1:27 pm

  28. Andrew: I can’t resist asking for you to elaborate on why you think I’m so narrow and how hatred (a word I did not use but am not altogether uncomfortable with) can be tautological. Btw, unclear if you are sympathetic with ANT or my purported hatred of it.

    Like

    ezrazuckerman

    July 7, 2010 at 4:48 pm

  29. Fair enough! I do not think you’re narrow, I think your criteria for theoretical success are. Specifically: both ANT and some of its theoretical ancestors (e.g., Foucault and some Frankfurt School) argue specifically, and non-trivially, that “facts independent of theory” cannot be ascertained, since ascertainment is itself representational and, therefore, theoretical. (NOTE: this is a distinct question from whether such facts “exist.”) So if you adopt ab ovo the criterion that a theory must account for theory-independent facts, any theory that rejects the capacity to ascertain these facts must of necessity be rejected, not because of the content of the theory but because its foundational claims are incompatible with the criteria. Hence the sort-of tautology :)

    Sorry for the unclear modifier – I meant that I am largely sympathetic to your hatred of ANT.

    Like

    andrewperrin

    July 7, 2010 at 6:16 pm

  30. As a quick note – You don’t need the Frankfurt School or Foucault for the theory-dependence of facts. The “Duhem-Quine Thesis” on underdetermination and background assumptions leads to Kuhn and a different take on the theory-dependence of all observations. Zammito’s A Nice Derangement of Epistemes covers this history, and puts the science-studies tradition(s) (SSK, ANT, etc.) in dialog with mainstream philosophy of science (Popper, Lakatos, etc.). Both broad camps agree on the theory-dependence of facts though, they just disagree on how much that changes how we can and should think about science.

    Like

    Dan Hirschman

    July 7, 2010 at 6:22 pm

  31. Thanks Andrew. I guess I’d say that yes I adopt “ab ovo” (had to look that one up!) the criterion that a theory must account for theory-independent facts. And so yes therefore I reject out of hand any position that says that there is no access to facts independent of theory. I’m not sure what makes that tautological though.

    BTW, I am quite aware that all facts are theory-laden. But that doesn’t mean that there aren’t some facts that we can agree to independent of theory. (This is why, I think, we are constantly talking about the weather– best way to connect with others in that it establishes that we independently are accessing the same facts). What I find maddening about the epistemology adopted by “ANT and its theoretical ancestors” is that if you assert that it is impossible to talk of facts independent of theory then it follows that you cannot advocate for one theory over another. Conversely, any time we try to persuade others of the rightness of our theory, the implicit premise must be that there is some access to common facts such that I can persuade you that my account of those facts is better the theory you previously favor.

    Like

    ezrazuckerman

    July 8, 2010 at 12:49 am

  32. if you assert that it is impossible to talk of facts independent of theory then it follows that you cannot advocate for one theory over another.

    I don’t think this follows at all, unless you assume that the only way to “advocate” for one theory over another is by reference to independently agreed-upon facts. But this is precisely what’s at issue. A pragmatist, for instance, might advocate for a theory on the basis of its usefulness. Others might be prepared to disregard or deny the relevance or reliability of particular facts on the grounds that they have a general theory that, while temporarily unable to account for this or that anomaly, is more elegant or parsimonious and thus more likely to be true.

    Conversely, any time we try to persuade others of the rightness of our theory, the implicit premise must be that there is some access to common facts such that I can persuade you that my account of those facts is better the theory you previously favor.

    Again, while disagreements may involve reference to various empirically observable phenomena, and may require shared standards of evidence and argument to profitably be adjudicated, to call those phenomena “facts” or those standards “common facts” ex ante seems like begging the question. E.g., we might agree that, as a matter of observation, something seems to be burning or rusting, but is the underlying fact that oxidation is happening or that the air is being dephlogisticated?

    Now you will want to say “Aha! But we agree on the fact that something is burning!” But that doesn’t follow. God knows I’m no pragmatist, but when you say “that doesn’t mean that there aren’t some facts that we can agree to independent of theory” you’re not that far from Quine’s “web of belief”. Saying “there are facts independent of theory” is quite different from saying “there are facts we can agree are independent of theory, for present purposes of resolving this disagreement”. That seems quite pragmatic, and it would be strange — or far from what you originally seemed to have in mind — to say that it’s this mutually agreed-upon discursive convenience that the “facts” consist in. The core issue is very much about this slippage, if that’s what it is, between an ontological state of affairs and an epistemological agreement about how to go on.

    Like

    Kieran

    July 8, 2010 at 1:40 am

  33. Hi Kieran.

    Re your first reaction: I don’t see how your proposed alternatives to my criterion really are alternatives. If the pragmatist wants to argue on the basis of utility, that will win someone else over only insofar as there is some commonly understood practical project that they are referencing. Otherwise, when you say that your theory is more useful than mine, why should I care? And I’m not aware of anyone who would say that a parsimonious or elegant theory should be acceptance without regard to how well the theory explains data. Theories are (or at least, should be) promoted on the basis of a balance of parsimony against explanatory power. And elegant theories are typically promoted with the argument that it may not yet have data supporting it, but it eventually will (with the implication that the theory should be rejected if support never comes)

    Re your second comment: I am reasonably comfortable with “there are facts we can agree are independent of theory” though I would say that ‘discursive convenience” and “moving on” are not trivial matters since it is difficult to get someone to agree to a set of facts if those facts are more consistent with our preferred theory rather than theirs. This means that the facts that have any hope of being established are those that can be verified independent of the competing theories. Whether those facts are “really” true seems relatively unimportant though. They just be real to each side in the debate.

    Like

    ezrazuckerman

    July 8, 2010 at 7:44 am

  34. I agree with Kieran, who has (as usual) stated the case more elegantly and convincingly than could I!

    Like

    Andrew Perrin

    July 8, 2010 at 2:02 pm

  35. Ezra, pragmatists consider ‘parsimonious account of observable facts’ as a pragmatic criterion. Indeed, it is quite common mistake for people to equate ‘pragmatic force’ in relation to some practical project since most pragmatist philosophers consider coherence to be the most important pragmatic criteria for knowledge. (I would consider coherence theory of truth to be inherently a pragmatist theory, although some disagree)

    Disagreements on whether facts exist independent of theory is down to misuse/misunderstanding of terminology. If one considers all shared lay knowledge as theories, then there is surely no way we can verbally express facts without some shared knowledge structure that ‘theorizes’ how words should be used.

    This theory-independence is a bit like the argument whether reality is socially constructed or not — people just hold different meanings for the words theory/reality in order to maintain a bogus disagreement.

    Like

    Henri

    July 10, 2010 at 3:11 pm

  36. […] episode in the new “culture wars” within economic sociology. I’m referring of course to Zuckerman’s attack on performativity for obscurantism, ontological relativism and theoretical weakness. Over the years, attacks on […]

    Like

  37. I recognize what I know of STS in Adam’s posts. The only nuance is, I would add Martin Kusch to Barnes and Bloor on the SSK side, and put Collins close to Latour (Adam, check Bloor’s ‘Anti-Latour’, there is a line where Bloor criticises Collins for having a line in ‘Changing Order’ about scientific phenomena being only ‘pictures in a fire’).

    Like

    Fr.

    July 13, 2010 at 5:18 pm

  38. Adam Hedgecoe made the point about the differences between STS in general and ANT. However, it would be important to state again a few things about social studies of finance.

    First, the claim that performativity is ‘just like’ Merton’s self-fulfilling prophecy was made, time and time again by none others than Donald MacKenzie and myself. I presented the paper about the Black-Scholes formula and CBOE about a dozen times and each time, without fail, I stated that the fundamental process at the basis of performativity is not different from self-fulfilling prophecy, but the innovation is that we show _in detail_ how self-fulfilling prophecy comes about, how it becomes institutionalized, under what conditions it is reversed and we expose the crucial role that non-human actors play in the process.

    Second, we hear time and again the claim that social studies of finance is ‘sympathetic to the markets’ and it SSF researchers analyze market devices in such great detail that they actually ‘go native’. The reason behind the detailed analysis is very different, of course. Understanding the intricacies of the analyzed subject is a crucial part of the strand of sociology of science from which the social studies of finance. This intellectual landscape includes, for example, Leviathan and the Air-Pump (Steven Shapin and Simon Schaffer), Andrew Pickering’s work on particle physics, Harry Collins’ works on the detection of gravitational waves and Donald MacKenzie’s research on missile guiding systems. This is very partial list, of course, but the similarities between the pieces are obvious. All these works are empirical, they strive to understand the techno-scientific knowledge embedded in the machines/devices studied and, critically, they identify and analyze the reflections of the larger political, social and cultural circumstances where these technologies were developed. This last point should be made again, for the benefit of the ‘attackers’: if you want to gain better understanding of the processes through which the financial crisis came about then you really have to understand the nuts and bolts of the business. And, no: simply saying that ‘it’s all greed’ or that ‘the models are always wrong’ (or their more elaborate, but not less simplistic, academic equivalents) is not nearly good enough. Instead, by understanding the minute details of, say, how credit derivates are designed we can shed new light on the macro conditions of neo-liberalism and gain better understanding of the crush.

    Like

    yuvalmillo

    July 13, 2010 at 11:02 pm

  39. Fr., when I see Harry Collins later on this week (he’s a colleague of mine in Cardiff) I’ll suggest to him that he’s intellectually ‘close to latour’ and watch the fireworks. But you are right in that there are clear, if fine grained, differences between the Bath and Edinburgh Schools of SSK.
    a.h.

    Like

    Adam Hedgecoe

    July 14, 2010 at 7:39 pm

  40. I should have said ‘intellectually closer’—but again, that is only because on his defense of idealism in Changing Order (check Bloor’s article for the precise reference and interpretation, or check the book directly, page 16).

    Like

    Fr.

    July 15, 2010 at 12:31 am

  41. […] l’ami François-du-blog-d’à-côté, je découvre une discussion intéressante chez orgtheory.net. Tout commence à partir d’un commentaire désabusé d’Ezra Zucherman (Associate […]

    Like

  42. […] has immediate relevance to many disciplines, and to STS specifically, as it has come under more criticism than expected in the last […]

    Like

  43. […] Guest blogger emeritus Nick Rowland has joined a new group blog, Installing (Social) Order, run by Antonia Langhof, Jan-Hendrik Passoth, and Hendrik Vollmer. The purpose is to explore emerging themes in social studies of technology and science. Ezra, please sit down. […]

    Like

  44. […] guest blogger, I took some heat in the name of Latour and ANT, so much so that Fabio joked that Latour ought not be mentioned in certain […]

    Like


Comments are closed.

%d bloggers like this: