Posts ⁂ The Postmodern Terminal >_
Metrics, checklists, and dashboards steer practice by feedback, and Jan Overwijk's category of the incommunicable names the expert judgment those instruments can neither see nor protect.
Published: 2026-09-21
Author: Adam DJ Brett
Series: Why Theory? (DPPR 916)
Tags: cybernetics, best practice, systems theory, critique, innovation
Source: DPPR 916, Why Theory Unit 10: Cybernetic Capitalism and the Critique of Best Practice
A checklist is a small piece of legislation. It says that these five things matter and the rest of what you know has no box. I will argue that the ordinary instruments of best practice (the metric, the key performance indicator, the protocol, the dashboard) are cybernetic control devices. I will show it by running Jan Overwijk's category of "the incommunicable" across a hospital discharge form. My aim is not to talk you out of systematizing, but to name what you trade away when you do.
Cybernetics takes its name from the Greek for steersman, the one at the tiller, and Norbert Wiener chose the word deliberately when he named the field in the late 1940s. His subtitle is the definition worth memorizing: control and communication in the animal and the machine. A thermostat, a nervous system, and a hand reaching for a glass of water are doing the same formal thing. Each has a goal, senses the gap between where it is and where it wants to be, and routes information about that gap back into its next move. That loop is feedback, and feedback lets a system regulate itself with nobody standing over it issuing orders.
Stafford Beer then built management cybernetics on that foundation, proposing in earnest that a firm could be steered as a self-regulating system. Niklas Luhmann pushed further: society is composed not of people but of communication, and systems like law, medicine, and education are operationally closed, each processing the world through its own distinctions alone. Only communication communicates. That principle is the hinge. Whatever cannot be put into a system's language does not exist for it.
Overwijk splices Luhmann together with the Frankfurt School's critique of instrumental reason. The old story said capitalism wants total rationalization: measure everything, eliminate the unpredictable. Overwijk's claim is subtler and worse. Contemporary capitalism has learned to valorize what it cannot compute. A platform monetizes attention it cannot model; a firm sells the "creativity" of workers it has no instrument for; volatility itself trades as an asset. The incommunicable is not crushed. It is fenced, farmed, and harvested, and it stays outside the system's own understanding the whole time. Gilles Deleuze caught the political texture of this in 1990: enclosure gives way to modulation, and the individual becomes a "dividual," a bundle of scores adjusted in real time.
Concede the gain first, and generously. Atul Gawande's case is well evidenced; the surgical safety checklist moved a body count in the right direction, and reliability is a moral achievement. Apply the hinge and three things happen at once. The measured dimension improves. The unmeasured dimension goes invisible, not because anyone judged it unimportant but because no channel carries it. And practitioners, being intelligent, optimize toward the channel that exists, which is why every field grows its own version of teaching to the test.
A hospital standardizes discharge: medications reconciled, follow-up booked, instructions given in the patient's language, teach-back confirmed, transport arranged. Readmissions fall. Then a nurse looks at a man whose boxes are all ticked and knows something is off, a flatness in the teach-back, the fact that nobody has come to see him. There is no field on the form for "I do not think this man is going home to anyone." The bed is needed, the form is complete, and the perception has nowhere to go. It is incommunicable to the system, therefore nonexistent to it, and after a few hundred repetitions the nurse learns to stop having it. The flattening is not the instrument failing. It is the instrument working as designed.
So here is the practice I would hand you: do not scrap the checklist, build it a door. Name a blameless, low-friction way for a practitioner to say "hold this, I have a concern I cannot itemize," and then defend that door from being converted into another metric. Michael Polanyi's move from tacit to focal knowledge is what makes excellence teachable; Overwijk names what the move risks, the tacit surplus that made the practice worth spreading. Systematizing is not the error. Systematizing with nowhere to put what will not fit, that is the error, writ large.
The Postmodern Terminal >_