An executive's plan for a contested decision is usually a list of actions: brief the chief operating officer, get finance to sign the number, square the depot heads. Each item is something the executive will do. The Lam System for Political Interaction writes a lever differently. Its word for one is intention, the specific action an actor chooses in order to move the target's decision, and the intention is written as its intended effect: how the target will compare the options once the lever has landed. What the actor physically does is written later, at execution, and what the action needs is written separately, as required resources.
Writing a lever this way changes what a plan is. A lever written as its effect is a forecast about someone else's judgment, open to being wrong in a way an action never is. An action either happens or does not. A forecast can come true or fail, and only a forecast gives the actor anything to learn from when the target reacts.
The constituents of a maneuver
Structure. Each line joins a constituent to one it contains.
The map shows what contains what in one actor's maneuver, from the actor down to what flows back when the cycle closes. It illustrates the structure only.
Actorα / a · A name only, performed by no oneThe person who opens the cycle and acts within it.
AgendaΑ / A · A name only, performed by no oneThe actor's long-running aim, which holds every cycle the actor opens.
Interestι / i · Performed by the actorWhat the actor stands to gain from the cycle.
- Detectionδ / d · Performed by the actorThe actor finding out who controls the resource wanted.
Targetτ / t · A name only, performed by no oneThe person whose decision completes the cycle.
Decision MatrixΔ / D · A name only, performed by no oneThe options open to the target, each weighed by what it offers and what it asks.
Preferred Positionπ / P · A name only, performed by no oneThe option that leads to the outcome the actor wants.
Adverse Alternativeψ / X · A name only, performed by no oneAn option that leads away from the outcome the actor wants.
- PerceptionΓ / G · Performed by the targetHow the target compares the options' worth before choosing.
- SelectionΣ / S · Performed by the targetThe target's choice among the options.
Objectiveω / o · Performed by the actorThe direction the actor wants the target's choice to take.
IntentionThis essayγ / g · Performed by the actorA specific action the actor chooses to move the target's options. A cycle holds one or more.
- Required Resourcesθ / q · Performed by the actorWhat the actor must hold to carry out the intention.
- Executionε / e · Performed by the actorThe actor carrying out the chosen action.
- A sub-cycle's interest, when the gate fails↺ Interest, under AgendaWhen the actor lacks a resource the intention needs, a sub-cycle opens whose interest is exactly that resource. It is an ordinary interest, drawn here a second time.
OutcomeΩ / O · Performed by the cycle itselfThe state the cycle closes in, one of the futures that were possible.
- Obtained ResourcesΦ / Q · Performed by the actorWhat the actor gains when the cycle closes.
- Obtained UtilityΛ / U · Performed by the targetWhat the target gains from the outcome.
- Current ResourcesΘ / R · A name only, performed by no oneWhat the actor holds now, counted.
What a lever is for
The objective says which selection the actor expects the target to make. The target makes it on the strength of how they see the options: what each one offers and what each one asks, netted into a sense of which is worth more. The intention is the link between the two. It names the change in that view that would make the target choose as the objective projects, and it implies an action that alters some part of the target's options, raising a benefit or lowering a cost, so that the target weighs them again.
In the running example Lena Ortiz has two levers on her freight company's chief operating officer, Daniel Okafor. The first is to show him that a single dispatch desk brings down overtime, a number he has taken heat for, which raises what approving the move offers him. The second is to get Priya Natarajan, the controller, to put her name to the savings figure, so that the number he would be backing carries finance's name and costs him less to stand behind.
Levers are the actor's alone to declare. A cycle may hold one or many, and each gets its own pass through the middle of the cycle before the passes converge on one selection. Every lever's address runs through the objective, the forecast it exists to make true.
The symbol and one line
The intention's operation is the lowercase Greek gamma, γ, and as an address it is the Latin g. Each lever is one instance, so the letter is lowercase, and a cycle with several numbers them: Lena's are the first and the second. Her first lever, in the system's notation:
Read aloud: the actor performs the first lever under the objective of the only interest, and it is equivalent to the target's perception of his decision matrix, as projected, coming out as follows. The utility of the preferred position, as projected, exceeds the utility of the first adverse alternative, as projected. In Lena's terms: once the overtime case has landed, Daniel will see approving the move as worth more to him than leaving dispatch with the depot managers.
The line names the comparison the lever is meant to win: the preferred position against one particular alternative. Lena's second lever is the one she aims at the other alternative, the automation program, which can promise savings too; the signed figure is meant to make hers the savings finance stands behind. The line also puts the target's perception on the far side of the triple bar, so that, as with the objective, the actor's declaration consists of something the target will do. What it leaves out is the action and the means. It mentions no meeting, no data and no standing with Daniel. Those come later, at addresses under the lever: its required resources and its execution.
The effect, the means and the act, kept apart
Writing the three apart is what lets each fail on its own. The required resources are tested at the gate before anything is done, so a lever that needs a sign-off the actor lacks is stopped on paper, before the meeting. The execution records the act and the resources it spent. The perception that follows is set beside the lever's projection. If the overtime case had been put to Daniel exactly as planned and he still rated the depot managers' option higher, the record would show an act carried out and a forecast that missed. That is a different failure from a lever stopped at the gate for want of a resource, and it calls for a different correction.
A plan written as a list of actions merges all three. "Brief Daniel on overtime" succeeds the moment the briefing happens, and nothing in it says what the briefing was supposed to change.
Each lever is new
A lever can be declared at several points: at the outset, straight after the objective; after another lever has been declared; after a lever has been executed; or after a perception has failed. Which of these an actor uses is a matter of stance, how they handle the uncertainty about whether a set of levers will be enough, and the system receives whichever stance the actor takes. How many levers to declare in all is also the actor's choice.
Whenever it is declared, a lever is fresh. It takes the next number, and no number is ever declared twice. If the overtime case had missed and Lena wanted to try a sharper version, the sharper version would be her third lever, and the first would stay on the record as a forecast that failed. An actor's memory works the other way, merging the failed version into the one that eventually worked until the lever seems to have been right all along. The record keeps each lever as one forecast, made once, with its own resources, its own act and its own reading.
Whose comparison a lever moves
A lever always works on the target's comparison. "Get Priya's name on the savings figure" sounds like a lever on Priya. In Daniel's cycle it is a lever on Daniel: its projected effect is on how he weighs approving the move, and Priya appears only as the holder of the resource it needs. Moving Priya is a separate decision, hers, and it needs a cycle of its own, which is what the failed gate opens. Inside that sub-cycle Lena has a lever whose target is Priya: she walks Priya through how the savings figure was built, and that spends her standing with Priya.
The same test sorts the rest of an executive's list. Squaring the depot heads is no lever in Daniel's cycle, because their cooperation is their decision, which makes them another target and the step another maneuver. A step aimed at anyone other than the target is either a resource some lever needs or a cycle of its own.
Which of these is a lever in Lena's cycle with Daniel
Illustrative. The people and the company are invented. Steps from Lena's plan, and one she might have added: guess which ones are levers on Daniel's decision, then select each to see where it belongs.
Select each item to see the verdict.
The first lever. It projects that Daniel will rate approving the move above leaving dispatch with the depot managers.
The second lever. Its projected effect is on how Daniel weighs approving; Priya appears only as the holder of the sign-off it needs.
A required resource of the second lever, declared under it and tested at the gate before the lever is played.
The execution of the first lever: the act itself, written once the gate passes, spending the overtime data and some of Lena's standing with him.
A lever, but in the sub-cycle, where Priya is the target. It works on Priya's comparison, and Daniel's cycle has no place for it.
Their cooperation is their own decision, so they would be another target and this would be another cycle.
The objective: the selection every lever in the cycle is meant to bring about.
A lever, but a new one, Lena's third. The first would stay on the record as the forecast it was.
How it goes wrong
A lever can win a comparison that does not matter. Suppose Daniel never seriously meant to leave dispatch with the depot managers. The overtime lever would then beat an option that was never live, its perception would come back as projected, and the comparison that decides the matter, the desks against automation, would be untouched. The record would show a lever that worked and a selection that went the other way.
A lever can also work on something the target does not weigh. Raising a benefit helps only if the target counts it. Had Daniel cared little about the overtime criticism, the first lever would alter his options in a direction he does not value, and the perception would fail with the case delivered perfectly.
A plan can carry too many levers. Each one is another pass through the cycle and another set of resources to find and spend, and a lever whose resource belongs to someone else tells that person what the maneuver is attempting. The system leaves the count to the actor, and the record shows afterward how many levers each maneuver took.
The last error is treating a lever as revisable in place. An actor who adjusts the overtime case midway and keeps calling it the first lever has merged two forecasts, and the reading that follows belongs to neither.
What the record teaches
Kept lever by lever, the record shows which comparisons an actor can move, with which resources, on which targets. It can also show which alternatives an actor's levers keep losing to. An executive whose levers reliably beat doing nothing and reliably lose to a rival program has a pattern no single maneuver would reveal, and it points at how they choose levers, since the record shows those levers were played as planned.
Whether a lever will land is the actor's judgment before the act and the target's response after it. The record holds the forecast in between, written before the act and in the target's terms, so that when Daniel's perception comes in it can be set beside the forecast it answers.