performativity: the future of economic sociology?

The performativity literature has been getting lots of attention in organization theory and economic sociology. The performativity argument, essentially, is that theories — well, economic theories have received the brunt of the focus — do not just explain or describe reality but also participate in creating or constructing reality. The strong form of the performativity argument suggests that the ex ante content of theories themselves does not matter: theories could even be false and nonetheless fulfill themselves. Thus theories can, as argued by Callon, be “arbitrary conventions” ex ante. With the content of theories set aside, performativity focuses on the social, political and technological factors that drive outcomes.

Has economic sociology embraced performativity?   I’ve heard some say that the future of the field is strongly wrapped up in the performativity argument.

Now, the performativity argument might appear innocuous enough — certainly it is extremely intriguing — but here are some open questions related to performativity:

  • Is it feasible to completely throw out ex ante judgment related to the content of theories, essentially the scientific apparatus itself? Surely we can make some ex ante judgments about the falsity or truth of statements and theories? Solely focusing on the socio-political-technological machinery does not seem very satisfactory nor convincing.
  • How does the rationality of humans (the ‘performers’) factor into performativity?
  • How does performativity address (real) change and progress?
  • Why does performativity “work” for explaining economics but then the argument is not turned on the performativity program of research itself? In other words, how do performativity researchers somehow step out of the performative water that social science is presumed to be in? How are performativity scholars somehow able to be more “meta” than, say, economists?
  • What are the boundaries for theories fulfilling themselves? If the ex ante content of theories is arbitrary then there seemingly are no boundaries.
  • How are the range of ex ante possible theories accounted for in performativity research? The ex post labeling of theories as performative (which captures a bulk of extant work) and the ex post rationality of performativity scholars is one thing — but seemingly one needs to appropriately specify the constraints (informational and otherwise) and choice sets of the actors involved at the time of the ‘performance.’
  • Really — how do we distinguish between performativity and counter-performativity? What is the difference? Counter-performativity appears to question the whole validity of the perspective.

On the margin, I think the performativity argument has some merit. Sure, yes, we can point to how we, perhaps foolishly, may have adopted certain conventions or theories that in restrospect prove “arbitrary.” But, the full-blown strong version of the performativity argument, that seemingly is the norm in org theory and economic sociology, is quite problematic as it cannot meaningfully address the above questions. That said, undoubtedly there’s lots of opportunity to explore these and other issues in future work — maybe performativity indeed is the future of economic sociology! (Unless it somehow gets counter-performed.)


Written by teppo

June 2, 2009 at 12:17 pm

25 Responses

Subscribe to comments with RSS.

  1. The strong form of the performativity argument suggests that the ex ante content of theories themselves does not matter

    This is not true. The idea is that the theory is a tool as well as a description, and if it’s a really useful tool it does not need to describe the state of the world ex ante in order to take hold as a description of it. That is, if it’s a powerful calculative device or whatnot it doesn’t have to be an accurate description of the way the world was before people started applying it in order for it to become a feature of the world later. This is a very different idea from the cod-relativist bogeyman view that performativists claim any old BS can become “real” as long as enough people believe in it.

    Economic sociology has embraced performativity

    I’d say this was quite a large exaggeration.



    June 2, 2009 at 1:52 pm

  2. Amen. The power of performativity is heavily constrained by some basic facts about motivation and behavior. For example, people want more money. Companies want market share and profitability. Is a cultural element which engenders these preferences? To a certain extent, but they are hardly arbitrary. People’s desire for more money is pretty fundamental. Companies that don’t prioritize market share and profitability will find themselves replaced by companies that do.


    Michael Bishop

    June 2, 2009 at 2:04 pm

  3. How does the rationality of humans (the ‘performers’) factor into performativity?

    It depends on what you mean by rationality. If you mean reasons in heads, then it does not play a role in performativity. If you are talking about reason-giving as part of the performance, as constituting the performance itself, then rationality is an empirical part of the performance to be studied.



    June 2, 2009 at 2:05 pm

  4. Kieran: “Economic sociology has embraced sociology” —- you’re right, “embrace” is certainly hard to measure and perhaps an exaggeration. I’m making the point based on works published in AJS, awards, citations to performativity-related articles, very favorable reactions to MacKenzie’s book (Swedberg etc), recently published books, of course some ‘unidentified sources’ and hearsay (this is a blog after all), research by up-and-coming and established luminaries, comments by Fligstein, etc. But, you’re still right, embrace is too strong. (I changed it to a question.)

    On the second point —-well, there indeed you highlight the point about a much-needed boundary though some of the foundational statements in my mind are quite clear about ex ante judgment and accuracy – Callon points out the ex ante arbitrariness issue, and Barnes and Bloor point out that “there are no context-free or supercultural norms of rationality” — highlighting the ex ante assumption made by performativity and related programs of research (e.g., ANT). And, your phrasing above almost suggests that you are talking about the emergence of norms (and essentially about ‘satisficing’ related to coordination), but I’m not sure that we then necessarily need performativity.

    In terms of the ex ante accuracy of description, I agree — but note that the full range of “calculation” (rationality — the same problem shows up when making the self-fulfilling prophecy argument) seemingly does not enter into the equation when highlighting how something (one thing, among many possibilities) might have performed. Naturally, agents are also dealing with high levels of uncertainty, but presumably they subjectively do/choose the best they can.



    June 2, 2009 at 2:09 pm

  5. Much as I think performativity is an interesting concept, I’m not sure I think it will become the future of economic sociology–nor would I want it to. For one thing, it perpetuates our obsession with setting economics up as a boogie man. Economists do their thing and I think it has lots of shortcomings. But we spend so much time trying to debunk that we lose site of the real goal: articulating ideas and mechanisms that focus on power and culture and relationships rather than simply on prices in a way that captures attention.

    At the same time–and in know this statement might relegate me to the orgtheory doghouse–I have a relatively hard time seeing how the performativity argument is different from the initial iterations of institutional isomorphism. Kieran’s defense notwithstanding, my read of the stuff on the black-scholes theorem is not a whole lot different from, say, Westphal, Gulati and Shortell on TQM or Tim Porac’s work on management fads. That is, with perhaps one important exception: MacKensie’s and Callon’s arguments draw implications for how ‘ideas’ merge with power structures to produce and reproduce reality (and conformity). I think that’s an important step to make, but is it enough to become the basis for economic sociology? Well, if you think that institutional theory has already basically done that (at least as far as macro econ soc goes), then maybe. But I hope the field is broader than that.


    Sean Safford

    June 2, 2009 at 3:03 pm

  6. Was Marxism a performative theory?



    June 2, 2009 at 3:08 pm

  7. I like this performativity concept, but it needs to be made more precise. I think a temporal scheme helps to make some of this more precise. One can assume that the physical and biological constraints on behavior underdetermine how a given population will behave over time. If one then assumes that individuals can influence their neighbors, then it’s easy to get to a “local equilibrium” in which the patterns of behavior interlock. Nonetheless, a sufficient perturbation — either in terms of physical/biological constraints or in terms of a very compelling new pattern of behavior — could result in a phase transition to some other local equilibrium.

    So to Michael Bishop’s point, sure American companies prioritize profit maximization ahead of other institutional goals. But that observation is not an explanation. The explanation has to account for why the goal of maximizing profits for the benefit of equity holders should be emphasized to the exclusion of unsecured or secured creditors, &c. There are decisions like Dodge v. Cox that are so embedded in American law that we no longer question whether they are correct because we no longer think of the issues at stake as questionable!

    I think of performativity as a version of Kuhn/Quine’s web of knowledge in which the edges of the field that interact with other fields and/or accumulate observations about the physical universe occasionally find new patterns of behavior that provide for vastly more efficient local equilibria. But whether the new patterns of behavior can spread is contingent upon the structure and dynamics of the neighborhood to the agent who makes the new discovery.


    Michael F. Martin

    June 2, 2009 at 3:08 pm

  8. my point put differently, is performativity simply the reincarnation of structuration theory (except that the process has a finite beginning point with the articulation and publication of a theory of performance which gets the recursive structure-action loop going)…


    Sean Safford

    June 2, 2009 at 3:18 pm

  9. MacKensie’s and Callon’s arguments draw implications for how ‘ideas’ merge with power structures to produce and reproduce reality (and conformity).

    I don’t remember their ideas being articulated quite so clearly. My problem with performativity is that it’s not really a theory at all. It’s essentially descriptive.



    June 2, 2009 at 4:19 pm

  10. Alex asks “Was Marxism a performative theory?”

    I think its a fair question…


    Sean Safford

    June 2, 2009 at 4:36 pm

  11. The thing to remember here with MacKenzie and Callon, given their background in the sociology of science, is that ideas and theories are embodied in and enacted through circulated material devices/objects/artifacts (whatever your preferred term), not broad institutional reforms. The fact that Black-Scholes not only create a way to price options, but sell sheets that help traders calculate the beta value is imperative here. I hope that helps clear up some of the differences between neo-institutional arguments vs. STS-centric explanations.

    In terms of bigger picture questions, MacKenzie’s account, most would agree here, would be made better if he indeed focused on macro side of his story. How are derivative markets really created? What institutions need to be in place for them to work? Saying a single theory simply created the market is obviously wrong – we all know there is much more (see Fligstein or even North). But alas, this has never been the strong suit of sociologists of science, who more often than not single out micro-foundations as determinative. This is not a problem, however. Performativity may be the new microfroundations undergirding institutionalization alongside the old standbys of ethnomethodology and the new cognitive sociology (as Powell and Colyvas 2008 suggest).



    June 2, 2009 at 5:43 pm

  12. “Was Marxism a performative theory?”

    In retrospect, yes.

    So, that’s exactly the problem: we observe given realities and histories, label them performative ex post, without considering the set of alternative possibilities, subjective assessments, uncertainties, etc.

    And, say Marxism indeed was performative. At some point the theory counter-performs, right. Why? Inevitably one has to refer to some kind of hard realities and emerge from the reflexive soup that we swim in.



    June 2, 2009 at 6:45 pm

  13. ‘Inevitably one has to refer to some kind of hard realities and emerge from the reflexive soup that we swim in.’

    That sounds like a statement of faith — a doctrinal claim of philosophical realism.

    From the perspective of performativity, researchers refer to ongoing practices and processes. That calls for a different kind of philosophical position.



    June 2, 2009 at 7:01 pm

  14. Statement of faith? — in that case even performativity must have some faith since there inevitably are references to changes, to falsity and to truth (veiled references, but they necessarily are there or else the whole exercise breaks down).

    Also, the performativity of performativity is a question (point #4 above) and the mysterious counter-performativity (point #7 above) also necessarily requires some reality to step in, or?



    June 2, 2009 at 7:11 pm

  15. Jacob: Let me try again — this time more to the point. So, if Marxism is performative (reality becomes a one-to-one representation of the theory), but then it fails and does not realize anticipated outcomes (counter-performs) — don’t we have an instance of reality putting a “check” on performativity. Not sure any faith is needed there.

    I struggle with the performativity angle in the first place as even at the ex ante stage theories compete among a set of possibilities, but then there’s the ex post testing stage as well which either confirms or discomfirms theories and possible realities — again, no faith needed.

    Its interesting to me that Actor-Network-Theory (ANT) is the direct (cf. Mirowski and Nik-Khah) antecedent of the performativity argument — where strong statements about the construction of reality focused on the sciences (Latour, Woolgar, etc), but that program of research simply did not pan out given the rather extreme assumptions it made about science and its relation to reality. Now, in the social domain, undoubtedly, there are important differences —- but, the differences might be more in our lack of understanding rather than anything else.



    June 2, 2009 at 7:33 pm

  16. Sure, like I said, performativity calls for a different kind of philosophical faith–I just think it is important be clear that scholars operate from sets of presuppositions about what the world is composed of and what counts as knowledge.

    Generally, I think, a performative oriented approach is interpretivist and focused on the social construction of meaning and conduct. Someone like Herbert Blumer operated from a performative perspective, symbolic interactionism. American pragamatists are also performative focused studies.



    June 2, 2009 at 7:36 pm

  17. Jacob — If we suppose that all approaches operate from a set of assumptions that are an intrinsic part of everything they produce, and that these assumptions are a point of faith (e.g. they are fundamentally arbitrary and their content is not subject to argument from outside the assumptions), it seems like the result of any inquiry (performativity aside) is just a restatement of reality in terms of those assumptions, combined with an adjustment to the logic of the assumptions themselves. If this is the case, I can still see a point of theoretical/predictive approaches: the content of the theories may be fundamentally arbitrary, but they have a pragmatic use — they can predict future states of the systems they represent. Furthermore, the proponent of a predictive theory can argue that someone should take on their assumptions because they are useful (thus not violating our supposition that the content of the assumptions is not subject to argument). But where does this leave the purpose of strictly interpretive approaches? Since they do not predict anything, it seems like they remain just a rehashing of their initial assumptions; and, since those assumptions are furthermore said to be a point of faith, why should anyone not already in agreement with those assumptions either take on those same assumptions, or be interested in the products of the research altogether?

    (n.b. I am not opposed to interpretive approaches — just to the idea that interpretive approaches can be justified as a matter of fundamental philosophical difference.)


    Subject A

    June 2, 2009 at 10:33 pm

  18. I think it might be useful to go back to the linguistic roots of performativity. Unlike sentences that described a pre-existing state of the world, performative statements “acted” on the world instead. On the role of discourse, the feminist Judith Butler said that performativity “contests” a subject and “produces what it names”.

    I think Butler pointed out that women are born into a world where their concept of themselves are already shaped by public and private discourse. To get out of this, one can’t really just step aside and ignore it. Instead, a performative act of re-signification needs to occur. So there are two things going on here: first, the dominant way of perceiving the world we live in was, at some point in the past, “created” by a performative act, theory or utterance and is constantly being reinforced (via description) as such. Second, this dominant hegemony is being contested by new performative theories or acts.

    Successful contestation necessarily provokes a restructuring of pre-existing, dominant relations in order to deal with this threat. I believe that Marxism was performative in this regard, because its critique of capitalism forced capitalist economies to re-organise relations of production in order to ward off class war. People say Marx didn’t factor in the inertia of the middle-class, but perhaps that’s because the middle-class might never have come about without Marxism in the first place.

    And this is also why I think that successful theories, if they manage to change the way we think about things, nevertheless give rise to actions that can deviate very far from the original theoretical scope. Communism is the obvious example, and its failure does not deny the world-changing impact of Marxist theory.

    Going back to finance, pricing models did not seek to contest the dominant market ideology they were embedded in, but they did contest how risk was assessed. The models succeeded in commoditising risk and changed the social relations that used to underpin valuations of risk. Consequently, however, the hegemony of free capital markets is now under threat because of the massive failures we’ve seen over the past 1.5 years. So the unintended consequence of a meso-level performative act (pricing models) is a macro-level “counter-performativity”.



    June 2, 2009 at 10:39 pm

  19. Teppo notes: How are the range of ex ante possible theories accounted for in performativity research? The ex post labeling of theories as performative (which captures a bulk of extant work) and the ex post rationality of performativity scholars is one thing — but seemingly one needs to appropriately specify the constraints (informational and otherwise) and choice sets of the actors involved at the time of the ‘performance.’

    I’m a bit reluctant to step into this (mostly because it feels like we’ve been here before, and I’ve been pissy about grand statements about the field…), but if you like we can look at the current examples of regulation of financial markets right now to see some of this ex ante activity.

    The debate is current, over how to regulate financial markets – effectively to ask ‘what should financial markets look like going forward? And an enterprising person could track the possible theoretical visions for how mkts should look going forward, from the perspectives of financial economics (Robert C. Merton has written extensively about financial innovation and shifting to a functional over institutional view of markets); from the perspective of ‘economists in the wild’ (i.e., industry pushback); and government interventionists (who may be proto-Keynesians of some stripe).

    So, you can reconstruct the view of what markets would look like from these varied perspectives, see how regulation actually changes, and then track what happens to these markets after that.

    If performativity has problems (and I think it does), I’m not sure impossibility of contemporaneous empirical research is one of them. Your questions, though, are certainly thought-provoking.



    June 2, 2009 at 11:54 pm

  20. I agree, Subject A.

    I think sets of presuppositions are more or less useful–regardless of whether or not they offer an accurate reflection of the world.

    Usefulness is not limited to prediction, I would say. Explanation and understanding, both of which are within the capacity of an interpretive approach, are also useful given certain contexts.



    June 3, 2009 at 12:57 am

  21. To me, the interesting issue here is related to theory choice. The question is whether the descriptive or prescriptive “correctness” of a theory (arguably its scientific or political correctness) determines its uptake in the literature. It’s not either/or, of course, but a question of degree; different disciplines, and different corners of disciplines, will have differing degrees of performativity in this sense.

    We use theories to describe reality in order to prescribe for it; or we might put it this way: a theory both represents the real and performs the ideal.

    I think an important determinant of theory choice in sociology has to do with the way explanations willy-nilly assign responsibility. They implicate actors differently, and sometimes entirely fail to implicate certain actors. (They also, as in the case of Marxism, actively construct actors.)

    I have found Bourdieu’s piece on “Description and Prescription” in Language and Symbolic Power useful in thinking about these matters, as well as Steve Fuller’s Social Epistemology.


    Thomas Basbøll

    June 3, 2009 at 7:39 am

  22. Between Galileo and Newton, astonomers sought numerologic patterns and epicycles, thought of gravity as “brooms” that sweep the planets, and so on, thrashing about for a paradigm that explained all of the known facts. As the study of society, sociology enjoys a surfeit of facts.

    Some sociologists suffers from a “flavor of the week” hunger for fads: differential learning, differntial reinforcement, differential association …

    The 20+ comments here sound much like Richard Feynman’s expose of “Das Ding an Sich” — if it means whatever you want it to, then perhaps it means nothing at all.

    I say: “Post modernality envelopes grammatarian continua that nebulize dimensionalist narratives of performalistic transgressors.” And you may quote me.


    Michael E. Marotta

    June 3, 2009 at 11:51 am

  23. OK, I admit, I’ve been out-meta’d.

    But, on the performativity issue, the questions in the above post still stand.



    June 3, 2009 at 11:57 am

  24. […] liens du matin (49) * “Performativity : the future of economic sociology ?” – […]


  25. […] on the pro and cons of performativity, self-fulfilling prophecies, etc…but after Teppo’s recent post, I couldn’t resist…so here is my 2 cents on this debate.  Going through the comments that […]


Comments are closed.

%d bloggers like this: