Every Maneuver Ends

Political work that never finishes has usually stalled or grown. A stalled attempt waits on someone, and goes on waiting after everyone has stopped expecting an answer. A growing one does the opposite. Every time the decision-maker hesitates there is one more argument to make; the aim shifts a little to meet the last objection; the person who could help needs something from someone else, who needs something from a third; and when the proposal finally lapses, it comes back next year under a new name. Stalled work at least looks stalled. Work that grows looks busy the whole time, which is how it runs for years.

The Lam System for Political Interaction promises that neither happens to a maneuver it records. Every maneuver reaches an end in finitely many moves, and at no point short of the end is it stuck. What the System Guarantees states the promise beside the others. It holds for any maneuver in which four numbers are finite, and all four belong to the actor.

Four ways to try again

A cycle, left to itself, is short. It runs through its phases in a fixed order, from interest to flow back, and each step reads what the step before it wrote. Only four moves can make it longer or add to it, and each is a way of trying again.

When the target's reading of a lever does not come out as projected, the actor can declare another lever, which adds another pass through the middle of the cycle. When a lever's gate fails, the actor can escalate, which adds a cycle one level down. When the target cannot be brought to the aim at all, the actor can revise it, which runs the opening of the cycle again against a new aim. And when a cycle has ended, short or in success, the actor can open another beside it. How many levers, how many revisions, how deep and how many cycles: those are the four numbers.

They are every route the system has for trying again. Escalation never retries a failed sub-cycle on its own, and a failed reading never replays the lever that produced it. Each growth described above is one of the four left without a limit. The one more argument is the lever count, and the shifting aim is the revision count. The chain of people who each need something from someone else is the depth, and the proposal that returns next year is the count of cycles.

Why the system sets no number

The system could cap any of the four, and it caps none. Its grammar allows another lever under any objective, another revision of any aim, a sub-cycle under any lever and a new cycle under the agenda, as many times as the actor wants. The promise holds for any finite choice of the four, and it says nothing about what the choice should be.

A cap would be easy to write, and it would refuse some maneuvers that are perfectly sound. An actor who needs a fourth lever, or a third level of escalation, is making no error by needing it, and a system that stopped them would be describing a smaller maneuver than the one they ran. It would also misstate who ended the maneuvers it stopped. A maneuver halted by a cap ends because a number ran out, and the actor's own decision, that one more lever was not worth it, would never be written, because the system had made it for them. With no cap, a maneuver that stops short of success stops because the actor wrote a moot or an abandon at a particular point, and that point sits in the record with everything that led to it.

Illustrative. The people and the company are invented. Corinne Abara is head of procurement at a frozen-food manufacturer, and she wants the chief financial officer, Declan Moorcroft, to approve moving all packaging to one supplier. His adverse alternatives are to keep the three suppliers the company uses now, or to put the contract out to bid. Before she starts, Corinne settles her own numbers: two levers at most, no revision, one level of escalation, and one attempt this year. As illustrative arithmetic, they put a ceiling on the attempt before it begins: at most three cycles, the one with Declan and one beneath each of her two levers, with at most six levers among them and an aim set three times, once in each cycle, since she allows no revision.

Declan reads her first lever, a savings analysis she holds, in a way she did not project, and the perception fails: he sees a single supplier as a concentration of risk, whatever it saves. Her second lever answers that reading with a written commitment from the supplier to produce from two plants, and its gate fails because she does not hold the commitment. She escalates to the supplier's account director, Marek Halloran, and spends her volume forecast showing him the business at stake. He will commit a second plant only for a longer contract than her authority allows, and getting the longer term would mean another level down, with Declan himself.

Corinne is at her own limit on levers and on depth. She abandons the sub-cycle, and back at the parent's failed gate she abandons the cycle. In most companies the proposal would now sit with finance for a year. In the record it has ended at a definite entry, having drawn two levers, no revisions, one level and one cycle. The numbers themselves were never written down. What the record shows is where she stopped, and that she stopped by choice.

Nothing runs backwards

With the four numbers finite, the reason a maneuver ends is close to plain counting. Every move the system writes takes the maneuver one step through a finite amount of work: a phase of a cycle whose phases are fixed, or one of the levers, revisions, levels or cycles the actor allowed. No move gives any of that back. A handback returns the parent to the gate where it stopped, which is where it already was, and the gate tests again. A sub-cycle that has closed stays closed. A revision does run the opening of a cycle a second time, and that is why it is one of the four: it comes out of an allowance the actor set.

The fourth number is the easiest to miss, because it sits a level above the maneuver. Every cycle in the record can end, each of them properly, and an actor can still never stop, by always opening one more. The first three numbers bound each maneuver. The fourth bounds the run of maneuvers on the agenda, so that the actor's whole engagement has an end as well. When every cycle has finished and the actor has opened as many as they meant to, no move is available at all, and no other position in the system is like that.

Never stuck, which is not the same as moving

The second half of the guarantee is that at every point short of the end, some move is open. Almost everywhere it is a forward move: the next phase, a gate tested again, a handback, a new cycle while the actor's count allows. The forward moves can run out in two places, a lever at the deepest level the actor will go that needs a resource they cannot get, with no other lever beside it, and a parent whose sub-cycle ended without delivering. At both, mooting and abandoning remain open, and abandoning is open everywhere. Abandoning follows why that one move, with no condition on it, carries the whole promise at those two points.

What the guarantee says is that a move exists. It cannot make the actor take one. A maneuver can sit for months at a failed gate with an abandon available throughout, and the record will show it sitting, with the cycle that waits and the decision it waits on. Nor does the guarantee count time. A slow target adds months and no moves, so a maneuver with two levers and no escalation is finite however long the target takes. Of the seven shapes in the figure below, four can run on without end, and in each of the four one of the actor's numbers is unset.

An end is not a success

A maneuver can end in three ways: it closes, it is mooted, or it is abandoned. The guarantee covers the three alike and promises nothing about which one a maneuver will reach. Corinne's maneuver, ended by two abandons, has ended as surely as one that closed with the contract signed.

A record needs exactly this promise, because it is read later, and a cycle with no end can be read only up to the point where it trails off, with its full cost unknown. Because every cycle reaches one of the three ends, every cycle can be read through to how it finished, and the failed attempts can be studied as completely as the successful ones. One way of trying again is not yet fully written: a parent that wants another route to a resource after its sub-cycle has failed. Giving up is open there, so the guarantee holds, and the retry is among the questions the series takes up under What the System Cannot Yet Represent.

The promise holds whatever numbers an actor's choices turn out to be, so it holds alike for an actor who chose them in advance and one who did not. The first decided how much trying again the matter was worth while the matter was still clear. The second decides it one lever at a time, each time just after a refusal, which is when the case for one more attempt looks strongest.