The Loop Closed

This volume has established three results, and it has established them separately, each in its own register, each standing on its own proof. The first result concerned the record: that the value produced by a specific combination of contributors appears in no contributor's entry, so that the instrument which allocates by entries is blind, formally and not accidentally, to the joining that produces the surplus it allocates. The second result concerned the body: that what the instrument cannot capture does not therefore go uncarried; it is absorbed, continuously and invisibly, by identifiable people whose absorbing is precisely what keeps it out of the record, so that the instrument's picture of a smoothly running operation is, read from the other side, a map of where the uncarried is being carried. The third result concerned the operator: that advancement runs through the instrument, that those advanced are those the instrument reads well and who read it well, and that the population of decision-makers therefore converges, cycle by cycle and without anyone intending it, on operators calibrated to the record rather than to the substrate the record is about.

Each result, taken alone, describes a defect. Taken together, they describe something with a different character entirely: a loop, and a loop of the self-sustaining kind. The purpose of this bridge is to close it in front of the reader, once, slowly, at the scale of an ordinary quarter, so that when the next volume composes the machinery upward into the economy, the reader carries not three findings but one mechanism.

Begin where the volume began, in a quarterly review. The instrument is on the table: the dashboard, the per-role ledger, the targets against actuals. An operator sits behind it, and the operator is not a caricature; she is diligent, numerate, and sincere, and the volume's third result says she is behind that table partly because she is the kind of person for whom this instrument feels adequate, having been advanced through a sequence of exactly such tables. The numbers before her show a team whose recorded output has risen. The first result says the numbers cannot show the couplings that produced the output, and so cannot show that two of those couplings have quietly thinned since the last review, one because a person who used to translate between two functions was reallocated to a recordable project, one because a channel of informal checking was priced as duplicated effort and closed. The second result says that the thinning is not yet visible as cost because a third person has been absorbing it, working the seams between the functions in hours that appear in no entry, and that this person's record, by exact consequence of the absorbing, looks unremarkable.

Now the review acts, because the instrument is allocative and a review is an allocation. The operator, reading the record well, does what reading the record well entails. The unremarkable absorber is passed over; the recordable performers are advanced; the released duplication is booked as efficiency; and next quarter's targets are set from this quarter's improved baseline. Every one of these acts is correct in the instrument's vocabulary, and every one draws down the substrate in the vocabulary the instrument does not have. The absorber, passed over, begins to withdraw the absorbing, or leaves, and the withdrawal will surface two or three periods hence as a threshold event the record will attribute to whatever is nearest when it crosses. The advanced performers, selected for their calibration to the record, staff the next layer of reviews. The improved baseline hardens the extraction into the plan.

And then the loop's third arc, the one this volume proved last, completes the circuit. Somewhere in the organisation, someone sees it. She writes it up. To be heard in the review, the writing-up must speak the review's language, and so it arrives as a proposal: a new metric for collaboration, a retention index, a dashboard for the invisible work. The proposal is welcomed, because the instrument can always extend its schema, and within two cycles the invisible work has entries, targets, and a gamed distribution of its own, and the person who saw it has been advanced for the contribution, into the population of operators, where her seeing will henceforth be exercised through the instrument that absorbed it. The correction has not been suppressed. It has been metabolised, and its metabolism is recorded, in the only vocabulary present, as the system improving.

That is one full turn of the loop, and the reader should notice what completed it: nothing. No decision was required beyond the decisions the instrument itself proposed; no actor deviated from sincerity; no information was hidden. The loop closes through the ordinary competent operation of the machinery this volume has described, which is why the volume has insisted, against the consoling alternatives, that the machinery and not the people is the object of the diagnosis. Replace every operator in the scene with a better one and the loop turns again, slightly more smoothly, because a better operator reads the record better, and reading the record better is the loop's fuel.

The reader should also notice what the turn consumed. One coupling, one channel, one absorber's willingness, one critic's outsideness: four stocks, none of them entries, each of them replenishable only on a cycle longer than the quarter that spent them. The instrument's ledger for the same period shows output up, efficiency up, engagement addressed, innovation in the pipeline. Both descriptions are accurate, and they are accounts of the same events in the two vocabularies this series asks the reader to hold at once; the distance between them, compounded across quarters and layers, is the quantity the next volume will follow upward: out of the firm, into the price system that aggregates a million such ledgers, and into the question of what an economy is doing when it grows by the first vocabulary while spending by the second.

The organisation, this volume concludes, is not a site where extraction sometimes happens. Under an archival regulator, the organisation is a machine whose ordinary operation is extraction read as performance, staffed by a selection process that renews the reading, defended by an absorption process that converts every account of the machine into a component of it. The machinery is now assembled. What remains is to watch it run at scale.