Karl Weick: An Actor-Network Theorist?

It is probably too great a stretch to call Karl Weick an actor-network theorist.

But I was struck by the resonance between his latest Academy of Management Review paper on The Lost Bank: The Story of Washington Mutual, the Biggest Bank Failure in American History and some notable actor-network theory concepts.

For an organization to act, its knowledge must undergo two transformations: (1) it has to be textualized so that it becomes a unique representation of the otherwise multiply distributed understandings, and (2) it has to be voiced by someone who speaks on behalf of the network and its knowledge (Taylor & Van Every, 2000: 243)…

The intertwining of text and conversation turns circumstances into a situation that is comprehensible and that can then serve as a springboard for action. If you envision WaMu as a network of multiple, overlapping, loosely connected conversations, spread across time and distance, collectively, the network “knows” the bank is failing, but that understanding is more complicated than any one node can reproduce. The distributed organization literally does not know what it knows until macro actors articulate it. And therein lies a problem. Actors who speak on behalf of the distributed organization have conversations, but the texts neither persist nor spread (Weick, 2013: 320-321).

There are a number of connections suggested by this vignette. For instance, in Science in Action: How to Follow Scientists and Engineers Through Society, Latour (1987: 227ff) introduces the concept of immutable mobiles, textualizations in Weick’s terms.

These are charts, tables, maps, figures: inscriptions of any kind that facilitate travel over time and space while retaining their form and shape. All are aimed at allowing the center to act at a distance (see also Law on “Actor Network Theory and Material Semiotics“). Importantly, immutable mobiles are also combinable. In this way, they have a tendency to become detached from their origins.

Although much has been made of the influence of Serres on Latour, as with moments of inversion, I cannot help but think the idea of immutable mobiles is already prefigured in Husserl. In fact, these two concepts themselves seem bound up with one another. Is it that immutable mobiles are constituted in moments of inversion?

Another obvious connection is to Callon and Latour’s (1981) paper on macro-actors. For them the macro-actor speaks for the assembled network. Weick seems to say as much, suggesting that WaMu’s failure may have been a failure of enrollment (see Callon 1986). For instance, Weick (2013: 32) describes WaMu’s chief risk officer “as an example of a macro actor unable to give voice to a larger set of wary conversations.”

Sociomaterial Networks and Moral Agencements

A good friend of mine recently sent me this TEDx talk in which Nitin Nohria, Dean of Harvard Business School, explores what he calls moral overconfidence and argues for the practice of moral humility as an antidote.

According to the talk’s abstract: Whenever we see examples of ethical or moral failure, our knee-jerk reaction is to say “that was a bad person.” We like to sort the world into good people who have stable and enduringly strong, positive characters, and bad people who have weak or frail characters. So why then do seemingly good people behave badly?

The centerpiece of Dean Nohria’s talk is the Milgram Experiment, which is typically argued to show that — given a strong enough situation — even “good people” will do “bad things.” More particularly, following Stanley Milgram’s own interpretation, most consider the experiment as showing the potentially dangerous consequences that may result from blind obedience to authority.

In light of my own research on values work, it seems the entire line of inquiry may be a false start — it presupposes from the beginning that good and bad are individually located. An alternative interpretation of the Milgram Experiment might start by taking notice of the many heterogeneous social and material actors that were required to be enrolled in the performance of “bad things.” Yale University, newspaper advertisements, experimental designs, subjects, confederates, experimenters, lab coats, electricity, shock machines, voltages, vocabulary tests, payments. In short, the experiment requires the enrollment of an ensemble of sociomaterial actors. If any of them had resisted, the experiment might have “failed.” So why is the actor at the end of the network the one to blame?

Such an interpretation is broadly consistent with actor network theory, in which the explanation for action can no longer be reduced to individual agency. In fact, such attributions are themselves part of what is in need of sociological explanation. What if the Milgram Experiment says more about the culture in which it is located than it does about the subjects it tested? After all, what kind of society is required for test subjects to be held responsible for the actions of an entire network, without which their performances could not have gone off? One can well imagine alternative societies in which different conclusions might have been drawn from the very “same” experiment.

In other words, we need to pose a more fundamental question. As Latour puts it, where is the morality? Is it in me, or in the objects? After reflecting on automobiles, seat belts and police officers, he concludes that morality is located in a network of humans and things. Networks make me (im)moral. Rather than an individual attribute, the definition, recognition and performance of good and evil are the result of moral agencements; moral agency is sociomaterially constituted.

See: Bruno Latour, 1992, ‘‘Where Are the Missing Masses? The Sociology of a Few Mundane Artifacts,’’ in Wiebe E. Bijker and John Law, eds., Shaping Technology/Building Society: Studies in Sociotechnical Change, MIT Press, pp. 225-258.