The Decision
as an Object of Executive Piloting.

What was entrusted along with the power to decide?

We are told very well to whom the power to decide should be given. But has anyone made explicit what there is to decide?

Let us look for a moment at the situation from the point of view of the person who receives this power. They may be delighted. They see recognition of their skills, a career step, a chance to contribute more to strategy. Then a question occurs to them: “This is wonderful. But when you entrust this decision to me, are you also entrusting me with the responsibility you bear as the executive? That one, I am not sure I want to take on.”

This question accuses no one. It shows that both parties can sincerely welcome the delegation without speaking of the same thing. The executive thinks the decision has been distributed. The colleague thinks they have received a circumscribed power, attached to their function. Both readings can coexist until the moment an account has to be given.

The colleague answers for their own acts, under the applicable regime. What they do not automatically receive is the overall position of the person who leads.

The question therefore precedes that of delegation. Before knowing what can be passed on, one must know what the mandate received contains: what its holder must answer for, what it falls to them to decide, what they may in turn entrust without giving up the exercise of their own office. The question recurs, moreover, at every level: the N-1 who delegates to an N-2 asks it in the same terms, for a more circumscribed mandate. This text remains focused on the executive decision.

“Who decides what?” is already a question of distribution. “Decide what?” obliges us to identify the content of the object being distributed. One can have defined the authorities, the thresholds, the approval circuits and the escalation arrangements without having made explicit what constitutes the executive decision to be preserved.

This is the question this text follows.

A Lawyer's Question

In law, a decision that is called into question is not first a matter of organisation. One must identify the person to whom it is attributable. Only then does one examine the circumstances that may confirm, share, displace or exclude that attribution, under the applicable regime. The law does not begin by asking how many people were around the table.

The starting point is therefore that decision-maker, the one to whom the decision will be attributed in the first place; it is the organisation built around them that is being questioned. Not the other way round.

Let us concede everything to modern ways of running a company. Trust, autonomy, decisions taken as close as possible to the information, fewer pointless approvals. We are probably at the peak of ways of administering a company to achieve a vision and results under the best conditions. The point is not to wish for an exhausted executive who wants to control everything. All of this is conceded.

But someone at the back of the room raises a hand: “And responsibility? Is it included in the transfer?” When the executive entrusts a decision to the general counsel, the compliance officer, the finance director, do they also transfer what they will have to answer for before their board, their shareholders, a regulator or a court? And will whoever set the procedures, laid down the prohibitions and built the approval circuits come to answer for what those choices did to the company's trajectory?

Were the answer yes, the question of what remains with the executive would not arise. If it is no, the question deserves more than an invitation to trust.

That is the asymmetry. The capacities to decide are distributed. The exposure of the person who leads is not necessarily distributed with them. This is not a criticism of delegation: it is a question that delegation does not suffice to resolve. And the executive's reluctance to delegate may then cease to pass for a lack of trust. It may be the rational perception of a risk they have not been taught to pilot.

Attribution does not mean that every legal consequence will be personally attributed to the executive. The levels differ: the company's liability, the duty to account for its management, a possible personal liability, depending on the circumstances and the applicable regimes. They may nevertheless meet around the same decision trajectory. No general theory of liability is needed. The starting point is the position of the person who, by reason of their function, will be called to answer for the decision, even before one examines the circumstances that would allow attribution to be displaced. It is this initial exposure that leads us to question the limits of delegation.

This question prompts another, which structures everything that follows: which decision?

There are two, and they are not two categories of people. There is the decision the executive must formulate, the one whose destination they carry: it is its content that must be made objective. And there are the decisions the organisation is empowered to take around it, real, legitimate, at times autonomous, and whose effects will determine what becomes of the first. It matters little here whether a decision is taken alone or by several. What matters is who must be able to answer for it as their own, and how they hold what they have decided when other decisions, legitimate ones, run through it.

Where Is the Decision?

We looked for the decision in theories of organisation.

We examined several of the major ways in which the company has been conceived: structure, hierarchy, delegation, coordination, autonomy, empowerment, accountability, flatter organisations, the distribution of authority, and finally decision rights. It was neither an exhaustive academic review nor an exercise in handing out good and bad marks. A single question guided the reading: where is the decision?

As one moves through these models, many things become objects. Roles, responsibilities, interactions, processes, performance, authorities, rights to decide. Each generation of theories makes visible something that its predecessor left in the shadows.

But as the models objectify structure, roles, coordination and autonomy, the decision appears sometimes as a power to be allocated, sometimes as a process to be organised, sometimes as a responsibility to be distributed. It appears far less as the material of a particular craft: deciding. We know how to say who may decide. We know how to say who must be consulted, who executes, who answers. It is much harder to speak of what the decision itself becomes.

Conceding Everything to Decision Rights

This text is not written against decision rights. On the contrary, let us concede everything to them.

Two examples among the most widely used frameworks for organising roles. The RACI matrix sets out, for each task, who is responsible, who is accountable, who is consulted and who is informed. The RAPID model, proposed by Paul Rogers and Marcia Blenko in the Harvard Business Review in 2006 under the title “Who Has the D?”, goes further: it distributes the roles within the decision itself, among the one who recommends, the one whose agreement is required, the one who executes, the one who provides input, and the one who decides.

Let us suppose that all of this is implemented perfectly. That information is brought closer to those who know, that unnecessary escalations up the hierarchy are avoided, and that competent people genuinely hold the right to decide well within their field, without constantly asking permission. Let us even take seriously the requirement of defining the decision clearly before distributing the roles around it.

Very well. Who formulates it? According to what criteria? Starting from what strategic destination? And what, then, falls within the executive's own work?

Before distributing rights to decide around a decision, where does the object being distributed come from? Who turns a strategic destination into a decision formulated sufficiently for the organisation to work on it? This is where the decision-maker's own expertise reappears, which the models presuppose more than they examine.

And a second question follows: what do all the legitimately distributed decisions do to that initial decision? These frameworks say perfectly well who holds the right to decide what. They do not necessarily suffice to show, over time, what these successive decisions do to the executive decision that set them in motion.

This is not a failing. They answer to their object. We are only asking what remains to be thought once we have granted them their success.

What Delegation Does Not Make Disappear

Let us return to the beginning.

Imagine an executive alone. No expert functions, no directors, no committees, no approval circuits. They have a business to run, decisions to take, obligations to meet. They answer for their business. They cannot answer for an unmet obligation by saying that they did not know how to meet it, or that they had no one to help them. They may be faulted for not having surrounded themselves with the necessary skills; those skills would not have acted in their place, for they remain responsible.

Let us now add one employee. Then ten. Then specialised departments, committees, outside advisers. Let us entrust them with decisions, resources, margins of autonomy.

What have we changed? The organisation of how the business is carried on. Not, by that operation alone, the function of the person who leads it.

The marketing director can decide on a campaign without asking for authorisation. They can be very autonomous. They are not thereby the holder of the mandate of the person who leads: their autonomy is exercised within an organisation whose destination they do not determine under their contract.

The executive does not become any less an executive as others become capable of deciding without them. At what point would the multiplication of capacities to decide have changed the nature of their mandate? At ten employees? At a thousand? At a million?

This is what delegation reveals, by contrast. Theories of decision distribution illuminate the recipient of power: who should decide, at what level, closest to which information. They illuminate far less the person whose function is to hold together what has been distributed: what they must continue to govern when others decide.

Let us now suppose that all goes well, in the sense that everyone has done their work properly. Everyone has exercised their remit. Everyone can say: I did what had been delegated to me, in accordance with what had been delegated to me. I committed no fault.

Yet harm occurs, or an obligation has not been met, and a judge, a third party or a regulator asks to whom it is attributable. This is not to say that every disappointing outcome personally engages the executive. It is to observe that the absence of fault in each function does not guarantee that the organisation has answered for its trajectory.

Imagine there were ten of them around the table that day. Ten who took part, ten with an opinion. The person to whom the decision is attributable in the first place cannot simply say: there were ten of us. The other nine do not spontaneously come forward to say: leave them alone, we were all there. Collective participation does not, by itself, determine primary attribution. Whether responsibility is then shared, or deflected onto someone else, is another question. One must first defend oneself against the primary attribution.

Each delegate will defend themselves, rightly, by invoking what had been delegated to them: the limits of their mandate, of their employment contract, of their delegation. And the executive does not build their mandate on the hope of being able to say, after the fact, that they had delegated.

No one organises their company in order to be able to explain, after harm has occurred, why they would not have to answer for it. One organises it so that it does not happen.

That is why the level of delegation is not an abstract answer. It is defined case by case, according to the height at which the executive consents to delegate. But there is one thing that delegation does not do: it does not, by itself, transfer the function of answering for the whole.

The executive who hesitates to delegate is readily placed among those who want to control everything. Sometimes that is true. But they may hesitate for an entirely different reason: they do not know how to organise what will remain of their own decision once powers have been distributed. This reluctance is not necessarily irrational. Nor does it call for a compromise between centralisation and decentralisation, which would merely move a slider.

The question is qualitative. Which decision is entrusted? What is its content? What destination must it serve? How far can the person who receives it modify it? When must it come back up? Who arbitrates, and how are the arbitrations retained?

Delegation distributes the exercise of certain decisions. It does not yet say what it means to lead. Once the capacities to decide have been distributed, what becomes of the decision that must run through their exercise?

A Pixelated Organisation

Let us suppose that everything we have just conceded is realised. Capacities to decide have been distributed. Each is legitimate, well placed, close to the information. Each knows what it can settle.

The organisation that results looks less like a pyramid than like an image made of pixels. Each pixel is a clear decision centre: a domain, limits, someone who answers. Seen up close, the image is extremely precise.

In this image, the executive is also a pixel. One of the decision centres of the architecture, among many others. And the question changes in nature. Great care has been taken to define the decision domain of each of the others. Has the domain of the person who leads the whole been defined with the same precision? What falls to them to decide that cannot fall to anyone else? What must come back to them whenever it touches the destination? Under what conditions does what they have entrusted come back to them?

Decision rights recognise in each holder a domain in which they decide and pilot their decisions. Let us apply exactly the same principle to the executive. Their domain nonetheless often remains less explicitly defined: theories of organisation readily describe their functions — giving a vision, allocating resources, mobilising — and more rarely an object of decision described with the precision reserved for others.

This content has nevertheless been encountered above: it is the decision they formulate, the one whose destination they carry. Their own domain begins there. It has one particularity, which is not posited at the outset but observed: this decision is not realised in any one domain in particular. Its realisation runs through those of other holders, whose decisions, legitimate ones, determine its trajectory.

This allows us to distinguish three situations that are easily confused. Piloting the decisions of a domain is the normal exercise of each holder. Piloting an executive decision whose realisation runs through several domains falls within the executive's domain: what is held is the destination and its arbitrations, not the content of the decisions of those who contribute to it. Intervening in the decisions entrusted to those domains, or unduly controlling how they are carried out, belongs to a different logic. Yet as long as the content of the executive's own decision remains implicit, piloting its trajectory risks being mistaken for interference, even for micromanagement.

It is this boundary that normative constraint makes particularly visible.

Why This Question Becomes Pressing Under Normative Constraint

This is where the entry through law becomes necessary.

The company operates in an environment increasingly entangled with norms. To deal with them, it distributes expert roles: Legal, Compliance, Risk, Finance, Data, and others. These functions do not merely produce opinions. They translate their subject matter into policies, processes, controls, thresholds, procedures, frameworks, and decisions.

At the same time, contemporary models of organisation have good reasons to give those who hold expertise and information a real capacity to decide.

These two movements are individually rational. It is their meeting that interests us. An expert function can take a decision that is perfectly justified by its own frame of reference. That is not yet to say that it is good for the company. One must still know what it does to the project for which that function exists in the company.

One Segment: Legal

It is only here that we take a segment, and it is legal.

Not because Lead or Follow is a theory of the legal function. But because experience of legal functions opens up here a field of observation: a place where one can look, concretely, at what becomes of a decision in the company.

What does one see there? Let us posit, for the demonstration, the hardest case for Lead or Follow. A strategic trademark is not renewed: let us take seriously the hypothesis that a competent person, holding the right to decide, did in fact decide not to renew the registration, and that this decision was rational at the time it was taken. An arbitration agreement had set the forum, then an equally competent person decides not to invoke it when the summons arrives. A decision is not necessarily missing. There may on the contrary be several, perfectly identifiable. The question is what they did, successively, to the trajectory the company was pursuing.

A general counsel is recruited for their pedigree, without their mission having been defined with regard to the trajectory: a function was recruited, not a line of command. And the “suffered” cost also changes status. It is not necessarily the sum of decisions that no one took. It can be far more disturbing: the product of decisions that perfectly authorised people did in fact take.

These are different situations. They nonetheless have something in common: a decision was taken, or could no longer be taken, without the place where this happened necessarily corresponding to the place where we would spontaneously have looked for the decision-maker. This is what Corpus I and II document: the centre of gravity of a decision can shift, and margins can disappear before the executive has been able to arbitrate.

Around this, the arrangement that makes all of this possible without any actor having done their job badly: a professional language, frameworks, metrics specific to the function, recruitment among peers, outside counsel, an expected output. The question “how can A pass through this constraint?” imperceptibly becomes “what ground does our own frame of reference allow us to hand back to the organisation?” What comes back can be perfectly legal in nature without yet being entirely A.

These manifestations are not generalised to all functions. Legal is a field of observation.

To go further on this ground: Corpus I — The Alternative Centre of Gravity of Decision · Corpus II — The Decision Stripped of Its Margins.

A Reasonable Practice Becomes a Mandatory Passage

On this ground, one mechanism deserves close observation. It was observed in connection with the confidentiality of advice given by in-house counsel.

Take a lawyer recruited ten years ago. They arrived young, competent, keen to fit in. They learned the organisation's codes, its projects, its actors. They made themselves useful: they answered questions, produced opinions, flagged risks, and were consulted earlier and earlier. Ten years on, they are unavoidable. They did not seek this centrality. But the organisation gradually integrated their involvement as a natural step, a prudent one, then an expected one.

The decision has not changed hands. It has changed process. And the process has a mandatory passage.

This passage is not formal. It appears on no organisation chart. It settled in by sedimentation: the repetition of reasonable behaviour, the valuing of prudence, the accumulation of precedents in which consulting before acting had avoided a problem. No formal power was assigned. A power was constituted. And counsel can become, in some organisations, a de facto validator, not because they decided so, but because the organisation has learned not to move forward without them.

A question then deserves to be asked, one the organisation rarely asks itself: who decides that a strategic question must be turned into a written, documented and retained consultation? This transformation is not neutral. Expertise does not merely produce an answer. Its practices alter the form of the decision, its circulation and its memory.

Finally, two recommendations that resemble each other must be distinguished. One is formulated according to the criteria of the profession: what is compliant, what is risky, what is defensible. The other is situated in relation to the strategic trajectory: what this does to A. The first can be excellent without being the second.

No legal effect is drawn from this here. What is retained is what it shows about the decision and its organisation: formally distributed powers are not the only powers; there are those that practices constitute. And the decision may then come to be authorised by the person who reads and interprets the norm, rather than taken by the person who bears strategic responsibility.

A Method from the Thesis

Legal is not the subject. It is a field. The method for observing the decision there was not discovered in this field: it had been learned earlier, in a law thesis. It does not observe the same phenomenon. It provides the way to constitute an object and to follow it.

That thesis dealt with the fate of the creditors of a debtor in difficulty, in OHADA law in the light of French law, and with their equality. To carry it out, a standard was needed: an object from which to observe together creditors whom everything else differentiates. The unsecured creditor (the créancier chirographaire) imposed itself naturally. The conclusion says so: “The unsecured creditor was to be our guide because he has only the guarantee of the general right of pledge and that prerogative is the lowest common denominator among all creditors.”

Why that one? Because it is defined by what it lacks: it is “the creditor who has only the guarantee of the general right of pledge”. Remove every particularity — security interests, privileges, ownership, special mechanisms — and what remains is that. The unsecured creditor made it possible to start again from what is common to all creditors, before statutory differentiation.

The movement is the same here. Legal does law, Finance does finance, Risk does risk, Data does data. One will not find the company by looking for what these functions have professionally in common.

Expert functions are not the purpose of the corporate project. The company does not exist to make Legal, Risk, Finance, Compliance or Data work together. Their presence, their expertise and their performance make sense only in light of what the company seeks to accomplish: strategy does not exist to give work to expert functions; it is expert functions that serve the realisation of strategy. They belong to the same company: this is their common condition. And their interventions contribute, each from its own expertise, to the decisions through which this company pursues its strategy. The executive decision is the place where this strategy becomes something the organisation can act upon. It thus constitutes the common object from which their interventions can be observed together, without erasing what differentiates them. The centre is not the functions. The centre is what the company had decided to do.

But finding the common object was only the first half of the gesture. The thesis did not say: we finally have the right box, let us stop there. It then showed what life does to this object. The introduction posed two questions, one at the start, the other at the finish. Are all creditors, at the start, unsecured creditors? And are those who are so at the start, at the finish, actually held to equality? At the start, there was only “an infinitesimal quantity” of unsecured creditors exposed to the rule. At the finish, “it is hard to identify an unsecured creditor in the collective proceeding”. And the thesis concluded: “So many justifications that led us to doubt our guide.”

Here is what this means. A category is a photograph. It says something true at the instant it is taken: at t0, this creditor is unsecured, or holds a pledge, or a mortgage. But it is made to say something it should not about the film. Someone smiles in a photo; we deduce that they are happy, then interpret everything that follows from that state. The information is accurate; it is the inference that is not. The photograph says neither what precedes it, nor what produces it, nor what follows it.

Between the photo at the start and the one at the finish lies the life of the relationship with the debtor. And that life is not a backdrop. Each decision there produces the situation in which the next will be taken. Taking a security does not simply address a risk that was already there: it can create a cost, formalities, monitoring or renewal obligations and, when the risk materialises, open a phase of enforcement or litigation sometimes longer than the relationship it was meant to secure. Whoever decides next no longer decides in the world in which the previous one decided. They decide in a world partly produced by the previous decision.

An asset thought to be attachable has become, by contract, the property of another. A creditor has lost their security along the way, or holds it without knowing how to enforce it, or enforces it very well and it is all they have left. Having a right and exercising it are already not the same thing. And even exercising it well does not guarantee that the final position resembles the one the initial classification let one imagine.

Then comes the moment of truth: the collective proceeding arrives and says, in essence, now let us see who is really who. It is what forces equality to be measured, and therefore forces us to look at what the position has become.

One can distribute rights among categories. One cannot confine within them the life of the object on which they are exercised. In the thesis, it was the creditor's position, from the life of the relationship to the test of the collective proceeding. In Lead or Follow, it is the decision in an organisation that lives.

The double movement is therefore this: find the common object in order to be able to observe, then refuse to freeze it in the opening photograph and follow its life.

A Baton

Someone formulates and decides A. A is a destination entrusted to the organisation. DECIDE A fixes the initial photograph: this is A.

Let us take a baton, and watch it circulate. The baton is not a box. It is the object we follow while the decisions taken about it produce their effects. They commit resources, burden means, close some possibilities, open others, and reduce or displace certain branches of its trajectory. Whoever decides next therefore no longer receives exactly the object that the previous decision-maker received. They decide in a situation that the previous decisions have, in part, produced.

It meets Legal, Risk, Finance, Compliance, Data, Procurement, operations, depending on the file. Each exercises their craft. Each may hold perfectly legitimate decision rights. Each may work perfectly well.

The result may be A. Very well.

A may also have to become B, because a constraint or new information leads to a new decision, an identifiable one, that someone takes. Also very well.

But a third possibility deserves to be made observable: can A gradually become B, without any identifiable decision ever having substituted B for A?

It is this possibility that the baton lets us question.

A Carrot

The organisation often knows very well how to examine the various treatments. Has Legal worked correctly? Risk? Finance? Compliance? Were the procedures followed? Were responsibilities assigned? Do the results conform to the metrics?

In other words, we know how to look at the segments.

But if we set out with a baton and find ourselves measuring a carrot, another question arises: at what point did the baton become a carrot?

The carrot is not the result we postulate. It is the hypothesis that following the decision must allow us to detect. We had A in the first image, we have B in the last, and we look at what happened between the two. We no longer try only to know who had the right to touch the baton. We look at what each of these touches did to it.

The quality of each of the segments does not, by itself, answer the question of whether what comes out is still what went in.

Decision rights can be perfectly accurate at t0. Everyone is in the right box. Legal holds such a decision right, A has been decided: these are photographs, and they are accurate. But their photograph is insufficient not only because life moves on. It is insufficient because the very exercise of the rights it distributes continually transforms the object on which the following rights will be exercised. If, at the finish, we photograph B, the question is no longer only: “who had which right to decide?” It becomes: “Show me the film that takes A to B.”

If this film contains a moment when someone actually decided that A becomes B, fine. Otherwise, B may be made up of decisions taken without B itself ever having been decided. This presupposes neither oversight, nor a void of responsibility, nor a bad actor. All the decision rights may function. Everyone may decide. And the baton may still arrive a carrot.

The subject is therefore more general than law. How does an executive decision cross functions that legitimately hold their own expertise and their own capacities to decide, without ceasing to be an object that can be piloted?

A Reality Test from the Law

Let us return to the question posed at the beginning. We looked for the decision in models of organisation. Coming from law, we observe something much more brutal.

When an act, a fact or a decision becomes disputed, enforceable against someone, or capable of engaging liability, the judge, the regulator or the third party raising the grievance is compelled to reconstruct the film. Who could decide. Who decided. Who knew. Who should have controlled. What had been delegated. What had been arbitrated. To whom the consequences are attributable.

The organisation could live with its photographs. The one who comes to reconstruct cannot. They arrive and say: now let us see who actually decided what.

It is at the foot of the wall that the mason is recognised. It is when the decision is reconstructed that the person who carried it is identified.

Where Lead or Follow Builds

The legal path therefore leaves three distinct contributions. The thesis provides a method: find the common object in order to observe, then never confuse the classification of a situation with the trajectory that produces its result. The practice of legal functions provides a field for observing the decision in the company. And the law's reasoning on attribution provides a reality test: when something becomes disputed, the judge, the regulator or the third party is obliged to reconstruct what really happened and to attach consequences to acts, powers, decisions and persons.

The law does not supply Lead or Follow's model. It allows us to recover its object.

If the decision can become concrete enough, after the fact, to be reconstructed, enforced and attributed, it is legitimate to ask why it should not be treated with the same materiality while it runs through the organisation.

The distribution of decisions does not relieve us of piloting the trajectory of the common object on which they produce their effects.

The decision already exists as an object of attribution. Let us take it seriously also as an object of piloting.

(Re)Installing Command

Let us retrace the path. We started from the person to whom the decision is attributed in the first place. We observed that capacities to decide are distributed without the exposure of that person being distributed with them. We saw what the decisions of others can do to theirs. This points to a risk specific to the exercise of executive direction: losing control of the trajectory of a decision for which one remains answerable, even though the powers needed to realise it have been legitimately distributed. The mechanisms the company has built against legal, regulatory, financial or operational risks have their usefulness. They may leave this one barely visible. The proper organisation of functions and decisions does not necessarily suffice to cover it.

The photograph, the film, the baton and the carrot are not the result of this research. They are the instruments that make this object visible.

Following A is not enough. Observing what the organisation does to a decision does not, in itself, make it pilotable.

Lead or Follow therefore works on something else: (re)installing the conditions in which the executive exercises their own capacity to decide, in an organisation that has distributed many other capacities to decide. Concretely, they must learn to do three things. We name them DECIDE A, HOLD A and EXECUTE A.

Distinguish: DECIDE A. The executive formulates the destination, then determines what, within that destination, is theirs to decide and what can be entrusted. This distinction is not an organisation chart. It is made decision by decision. What, here, commits the destination? What belongs to an expertise that can settle without altering it? What, if entrusted to a function, would in reality be a part of the destination changing hands? Without this distinction, one does not distribute decisions: one distributes pieces of A.

Circulate: HOLD A. The executive holds the circulation of the decision. Under what conditions does what has been entrusted come back to them? What information, what change, what threshold triggers an escalation? Who arbitrates, and where is the arbitration recorded, so that the next person to decide knows what was settled and why? This is also where the question that The Shift brought to light is settled: in what form, with what memory, is what is asked of an expertise handled, and who decided that? And A may become B; sometimes it must. But then someone decides B, by name, and that decision has its memory. HOLD A is not meant to repair bad organisations: it answers what the good distribution of decisions does not resolve on its own.

Mandate: EXECUTE A. The executive organises the exercise of the distributed expertises, without taking back either their expertise or their capacity to decide. This runs through the mandate given to them. Asking a function for an opinion is receiving a recommendation formulated according to its criteria. Also asking it to situate that recommendation in relation to the trajectory, that is, to say what it does to A, is to give it a more exact mandate, without taking anything from its autonomy. The executive does not contest the expertise. They keep the arbitration between a recommendation that is excellent by the standards of the craft and what the destination requires.

Decision-making under normative constraint is the most demanding test. The executive cannot substitute for the expert. But expertise must no more substitute for the strategic destination. Neither should prevail by default.

The aim is therefore neither to remove the functions' autonomy nor to concentrate all decisions at the top. It is to organise the executive's decision domain while the others legitimately exercise theirs.

The film says what happened: it is an instrument of knowledge and proof. Command, for its part, acts while the decision is being transformed. It arbitrates, maintains the destination, deliberately modifies it, or refuses to let it be modified by default. Knowing how to show the film from A to B is useful. Having held A, or decided B, is what counts.

You have organised the domains in which others can decide. How have you organised your own?

To go further on the architecture: Corpus III — The Architecture of Command · Corpus IV — The Migration · Who decides what? · Pilot what, in fact?.