Economic sociology these days seems preoccupied with “performativity.”  That attention is well-deserved because the notion that the use of a model can improve its predictive fit is a powerful idea.  (For a graphic representation of the concept, see the excerpt by Brooke Harrington.)  But perhaps it’s time for a corrective look.

Graduate seminars are a useful barometer of when a concept starts to run past its purposes.  Some decades ago it was “socially constructed,” next “embeddedness” took its turn, and then for a brief while “boundary object.”  (I can’t begin to list all the things that students told me were boundary objects.)  Today it’s “performativity.”  The problem was not just that the terms were used, but that instead of being asked as a question (“how is it socially constructed and why does that matter?”) reference to the concept becomes the quick answer to almost any kind of question regardless of how far away from the original idea.  People who have experience leading graduate seminars will recognize what I’m talking about.  There’s a pretty good discussion going on, and just when things are getting perplexing and difficult (i.e. good, as in potentially productive), someone voices the solution (“It’s a clear case of performativity, right?) and it suddenly seems to close the conversation.

With all the talk about economists performing the economy and models performing markets, perhaps we are losing site of skilled performance.  Charlie Smith addresses such skilled performances in a recent paper, “Coping with Contingencies in Equity Option Markets” (forthcoming in The Worth of Goods:Valuation and Pricing in the Economy, edited by Patrik Aspers and Jens Beckert, Oxford University Press).  Smith distinguishes among three strategies for confronting uncertainties.  The first, “Making Sense” entails imposing some kind of narrative account on events initially experienced as chaotic.  “Routines,” the second strategy imposes some sort of behavioral order.  The third, “Acting Sensibly,” the least common but the most interesting, doesn’t seek to impose an overall order.  It is a method for handling the contingent and the disorderly.  In this rich study of option traders, Smith analyzes the practices of juggling multiple rules of thumb in managing option positions.

Daniel Beunza and I take a similar analytic approach when moving from how models make markets to understanding how traders actually use models.  We conduct an ethnographic account of merger arbitrage as a reflexively skilled performance, with reflexivity socially distributed inside and outside the trading room.  “Models, reflexivity, and systemic risk: a critique of behavioral finance,” available here, is a paper with a twist.  When the system lacks requisite diversity (or dissonance), the unintended outcome of the attempt to deal with the fallibility of financial models is systemic risk.

I’ve also been exploring these themes in a presentation called simply Performance.  It’s one of several ‘silent lectures’ — a genre I’ve been working on for my undergraduate teaching.  They’re intended to evoke the multiple, intertwined meanings of some basic concepts. As I’ve discovered in my teaching, sometimes silence on my part is the best way to stimulate students to become active in the process of making associations.  Each silent lecture is a PowerPoint slide show, typically with about 100 images and few words.  The presentations are self-timed.  Load the PowerPoint, click ‘view slide show,’ and it runs itself, usually under 10 minutes.

I’ve found silent lectures to be a very effective way to stimulate discussion, even in a relatively large lecture course with as many as 80 students. There’s clearly no “right answer” (a dagger that kills good discussion) and the students get that right away.  One can sense that there’s lots of energy in the room, people really want to talk, and so I just get out of the way and listen in amazement at the connections they’re making.

Silent lectures are available for download on my website here.  Please feel free to use them in your teaching.  I ask only that you credit the source and that you send me an email to let me know that you’ve used one.  Of course, I’d appreciate feedback:  I continue revising and tweaking them.  But I know that folks are busy.  It’s enough just to drop me a one line note (“used your slide show in my course”) so I can track usage.

Performance opens with musicians, dancers, and actors performing.  Then athletes.  Audiences applaud.  We see judges, report cards, stock tickers, measurements, Top Ten lists, and online ratings of restaurants, movies, politicians, and professors. Various performance venues, material and virtual — from stadiums and lecture halls to surgeries and Second Life.  Students get it that the images of the CEOs of the big automakers and investment banks testifying before Congress refer to a kind of performance about performance.

The slide show explores the interplay of the real-time skilled achievement of effective performance and the evaluative principles of measurement.  From the bedroom to the boardroom, pharmaceutical companies and management consultants promise enhanced performance.  We live in an age of performance anxiety.

Written by dstark

June 24, 2010 at 10:31 am

17 Responses

Subscribe to comments with RSS.

  1. Is a right answer (where there is one) really “a dagger that kills discussion”? Is it not, sometimes, a knife that trims away the fat? (A dagger is not a killer in itself, but only when performed in certain ways.)

    That said, I completely agree with your view of the social/rhetorical function of the concept of performativity, embeddedness, and construction (my browser’s spellchecker recognizes two out of three … and it’s no doubt thinking of buildings and bridges in the third). I’ve seen very strange things called boundary objects by professors too, I should say.



    June 24, 2010 at 11:08 am

  2. Thanks, Thomas. Perhaps I wasn’t clear about the ‘right answer.’ It’s not the answer that is the problem. The move that kills discussion is of the following type: professor asks a question, a student answers, and the professor immediately follows up “That’s right.” I still make this mistake. But it’s good to be aware when it happens and so to try to ask questions — even better, pose problems — that really stimulate discussion. My silent lectures are one kind of strategy that attempts to move in that direction.

    And, yes, what’s with these spellcheckers? Austin wrote How to Do Things with Words in 1955. So you would think there would be enough usage of ‘performativity’ that it wouldn’t trigger a prompt.



    June 24, 2010 at 12:17 pm

  3. Another example of the spellchecker-effect here (in the comments):



    June 24, 2010 at 1:28 pm

  4. David:

    Your post calls to mind Mirowski & Nik-Khah’s take-down of Performativity (see here:

    Like you, they notice the emergence of a new fad (within constructionist “science studies”) and wonder what happened to the old fad (“Actor-Network Theory”) and who decided it was now time to change, and why. (They also accuse the performativistas of sucking up to economics; such sycophancy is appealing because the alternative is to actually admit that there is a world out there independent of our theories about it, which would in turn imply that both economists’ theories and performativistas’ theories can be wrong. And boy can they!)

    In their case though, they call out senior scholars rather than graduate students, and charge such scholars with essentially running a shell game. Grad students, after all, are at the mercy of our guidance. And where they especially need our guidance is in telling them that social science is not a form of aesthetics, such that we like theories just because they happen to appeal to us, but rather that theories must be judged on the basis of their ability to explain a set of facts more effectively and parsimoniously than other/prior theory.

    Given that, I worry about your characterization of how new research is emerging to replace performativity. I’m all for ending the performativity fad (which thankfully never became a “preoccupation” in my corner of economic sociology), but only because there was nothing in it that did a better job of explaining facts than what we already had in our theoretical arsenal (which included the Thomas Theorem; if someone can tell me what is in the idea of performativity that is not in the Thomas Theorem, I will eat my hat). 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? In particular, I’m not sure what it means to say that “we are losing sight of skilled performance.” That just sounds like an appeal to make an idea more popular than it was before (not sure I’d agree though; there is obviously a huge literature on various kinds of skilled performance). I’m missing what set of facts out there we can’t explain with existing theory (which might include theory that has not been recently popular) such that new theory is needed. And more specifically, what was performativity unable to explain that this emerging research can better explain? As I said, I’m also sick of performativity; but not because it’s yesterday’s fad, but rather because I could not see how it added to our ability to explain the facts that economic sociologists are interested in explaining.

    Interesting idea about the silent lecture, btw. I’m sure my students would appreciate it if I were more silent too!





    June 25, 2010 at 2:16 am

  5. Thanks, Ezra, for the provocation and for the link to the piece by Mirowski and Nik-Khan. I’m writing between conference sessions so I didn’t have a chance to study it, but I did note the book in which it was published. Orgtheory readers might be interested to know that this criticism of performativity was published in a volume edited by Donald MacKenzie, Fabian Munieza and Lucy Siu. I think that’s a credit to them. Researchers in science studies frequently study controversies and so it’s interesting that MacKenzie is not suppressing controversy but actually facilitating it with reference to his own project.

    The Thomas Theorem – them that’s got, get more – in the hands of Bob Merton and others was typically applied to persons. I guess that one could read performativity as extending the Thomas Theorem from persons to models. But in that reading, instead of dismissing it (“so, what’s new in that?”) one could acknowledge that this is a novel application of an idea to another object of study. We regularly credit work that makes such extensions. So, I find your reading quite interesting, although I would give it a more generous twist. My reading is a bit different: Performativity occurs when the use of a model improves its predictive fit.

    On ‘silent lectures’: it’s possible that your students might appreciate more silence on your part. But your colleagues in economic sociology would not expect that from you – nor would we want it. It’s always a delight to engage with you.



    June 25, 2010 at 12:35 pm

  6. Thanks David. I agree that it is to the credit of the editors of that volume that they included the Mirowski & Nik-Khah chapter. Would it that folks in that literature actually took on board its lessons.

    One slight correction is that I think you are confusing the Matthew Effect and the Thomas Theorem– which is “if men define situations as real, they are real in their consequences.” Merton made the TT famous in his essay on the self-fulfilling prophecy, which is very much about models (=theories) becoming more predictive simply by dint of their acceptance (of course, widespread acceptance of a theory does not always make it predictive–the real challenge is to figure out the conditions under which social theories are– and are not– self-fulfilling; a related challenge is to distinguish between cases of apparent self-fulfillment and those in which the theory was predictive simply because it was a better theory; if the performativistas had ever taken up either of those challenges [rather than just give the post-hoc label ‘counter-performativity’ when it doesn’t work], I’d be all ears]).

    BTW, the ironies in reducing performativity to the Thomas Theorem are many, including that science studies set itself up against Merton from the beginning.



    June 25, 2010 at 1:07 pm

  7. Ezra,thanks for the correction. Oh, goodness, I’m too jet-lagged to keep all the biblical characters straight. Sorry for any confusion.

    But the mistake did turn out to be revealing. If performativity were just an extension of the Thomas Theorem (your corrected version here), I can see how you wouldn’t find it insightful. The Matthew Effect (applied to models) would be an interesting extension.

    BTW, I very much agree with your criticism of ‘counter-performativity.’



    June 25, 2010 at 1:18 pm

  8. Sure thing, David. (Took me a sec to figure out the Biblical characters reference; neither name shows up in my Bible! :->)



    June 25, 2010 at 1:21 pm

  9. Ezra,

    I think you (along with David, and others) over-emphasize the self-fulfilling prophecy aspect of performativity (for a longer explanation, see this post). MacKenzie’s work makes this very easy to do – especially the stuff before his newest book, “Material Markets”, which broadens the analysis back out. Let me offer my take on what performativity gets us, what it does.

    Focusing on performativity, as opposed to some other existing approach, leads us to look for explanations in different places. It leads us to focus on the tools people use to make markets (computers, spread sheets, etc.), on the nitty gritty of the rules that those markets abide by, and how those rules are enforced, and to note that no two markets look the same (hence “The Laws of the Markets”). Performativity helps us note that no two markets look the same, and no single market looks like the market of economic theory. But it also leads us to see how many actors are working vigorously to make markets look more like the markets of economic theory*, and that sometimes they succeed (and sometimes they fail dramatically, and sometimes they do one followed by the other). Perhaps one could pay attention to all of those things using different theoretical tools and come to similar conclusions – but I don’t think anyone was, and so I credit performativity with helping us understand more than we used to.

    * Note that economic theory here too is not static – hence the elegance of MacKenzie and Millo’s historical approach to options markets, options regulation and options theory, all bundled together.


    Dan Hirschman

    June 25, 2010 at 2:38 pm

  10. Dan:

    From a quick look at your post here and then one you linked to, I don’t think we could have a productive discussion on this. I’ll just say three quick things and then not comment further on this thread:

    * If the value of performativity is just as a sensitizing device (“leads us to look”) to heterogeneity in markets, that is some pretty thin gruel. I don’t know who needed performativity for that.

    * Callon’s assertion that the economy is embedded in economics fails for the same reasons I alluded to earlier– insofar as it is true, there is nothing new or surprising in it (recall the famous Keynes line about Practical men… are usually the slaves of some defunct economist”); but it is obviously not true to a large extent (were there no markets prior to the discipline of economics? would markets not exist had economics not arisen or taken a different form?), and the real challenge is to roll up our sleeves and figure out the conditions under which it is, and is not, true.

    * It is sad and corrupt, rather than “useful” that STS denies a truth outside of (its) discourse. No doctrine could be more convenient for a scholar who wishes to retain her status.



    June 25, 2010 at 3:01 pm

  11. Dan,

    Just to be clear: Performativity as developed by Callon and MacKenzie is not simply about self-fulfilling prophecy. Callon has an elaborated statement about that. I do think it is insightful that financial models are not representations but are instead interventions. My post was calling for more attention to how models are actually used. That kind of detailed research is very different from facile claims that saying something makes it so.



    June 26, 2010 at 10:34 am

  12. I’m curious what the theoretical backdrop of this is. Are we talking Goffman? Ethnomethodology? Feminist work (like Butler’s performativity)? A pragmatist “creativity of action”? I think this would be ripe for an annual review piece — something about the different approaches to performance and the empirical work that they have spawned.

    Goffman’s dramaturgical account suggests (already existing) roles that people perform. Ethnomethodology is not about roles, and more about the importance of interaction. Butler is in between. And the pragmatists suggest performance is habits, until they break down in an interaction. Then, we have intelligence.

    In each of these, creativity exists in a different moment of the social act. Some are analytically clearer to me than others. But I’m curious about the economic sociology approach. My understanding is that it’s close to Goffman — that the way people frame (or abstract) ideas impacts future action, regardless of the empirical justification of that frame.

    There was an earlier discussion of this on orgtheory, that was quite interesting:




    June 26, 2010 at 12:34 pm

  13. Shamus,

    A delay in responding because I was traveling back to Europe (and just watched the impressive German team against England).

    Yes, on all you mention above. The intention of my post was to invite more perspectives on performance. That is, in addition to studying performance indicators, we should study the real time action of skilled performance. Goffman, ethnomethodology, and other studies of performance would be useful perspectives. My own work would be much closer to the American pragmatist tradition. In any case, I’m not sure that there is (or even should be) something that is “the” economic sociology approach.




    June 27, 2010 at 4:24 pm

  14. @Ezra

    From from denying the existence of truths outside of its own discourses STSers spend a lot of their time acknowledging and studying the multiple existing ontologies through which people cope with sensory phenomena. Although you seem pretty closed to discussion, I’d suggest you pick up Annemarie Mol’s “The Body Multiple.”


    Justin Kraus

    June 28, 2010 at 2:52 pm

  15. edit that,

    “Far from denying…”


    Justin Kraus

    June 28, 2010 at 2:53 pm

  16. I find this discussion ever so frustrating as most are content to straw man performativity and not take the time to really explore both its origins and arguments (as Kieran did wonderfully 3.5 years ago in a related orgtheory post).

    The point all along has been and continues to be more than mere self-fulfilling prophecy. It is not enough to simply imagine a world and have it be so. Instead, ideas about markets must be carefully enacted and embodied in material artifacts, tacit skills, and coordinating conventions. As I read the above post, David is pushing performativity to explain why some performers are better than others. It seems some notion of skill, which has been elaborated elsewhere in science studies literature (much of Harry Collins’s work centers on this), should enter the discussion.

    I appreciate Ezra’s realist orientation to generate theory and subject it to the crucible of falsification; at the same time, we should recognize the best examples of performativity were elaborated through rigorous historical work–a method that is often theory-guided but rarely theory-testing.



    June 28, 2010 at 3:49 pm

  17. […] 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 […]


Comments are closed.

%d bloggers like this: