Comments on Market Devices – round two
September 4, 2008
The discussion following the Market Devices session in AoM in August continues. Again, I got more sharp and thought-provoking comments from Bruce Kogut, as well as from Daniel and Martha, here on the blog. Below, I am trying to answer these queries/challenges
Organisational and scientific accuracies do not represent two distinct types of knowledge, but instead the latter is sometimes an overlrapping sub-set of the former. That is, if actors and actants connect in a manner and shape that bring about consensus with regard to the model’s usefulness (in the sense of problem-solving capabilities) then it can be said that the model is ‘organisationally accurate’. That structural coalition of actors/actants can take place within and around the academic community and then we would say that ‘the model has been proved to be scientifically valid’. Similarly, such nexus of connections may evolve in a different setting, such as in financial markets. In that case, we would say that ‘the model has proved to be operationally efficient’.
Now, this conclusion leads me to the ‘essentialist’ question, as I think we can call it. That is, the notion that Daniel refers to when he says that maybe ‘there is something’ in the model that triggers or aids the cascade of events that leads to performativity. Bruce Kogut, in fact, put it very nicely by saying to me, and I paraphrase, that no matter how many connections one would have with regard to the cold fusion theory, it will not ‘become accurate’. This brings us to the question that Martha posed about tests of validity (“according to what test is Black-Scholes-Merton model considered accurate?”). The answer here is that a model that will become performative is the model around which an effective and robust coalition of actors would emerge. Would the model or theory have to be scientifically accurate for such a coalition to crystallize? Not necessarily. In fact, it is possible to image a scenario according to which cold fusion theory becomes commonly accepted. It may happen, for example, if the applications of that theory would help to solve some problems (just as we saw in the Black-Scholes-Merton case).
Still, one may ask what could be the conditions necessary for a theory to be performative. I provide a rudimentary answer here, but I increasingly believe that that performativity, at least in its present form, is mainly a retrospective analytical tool. It helps us to explain and interpret historical events. It is, however, not very good at predicting the unfolding of such events.