Simultaneous, Not Sequential

In a consular emergency, time is not determined by response speed. It is determined by information routing — whether the relevant departments inside a mission and the stakeholder ministries at home receive the right case information simultaneously, or sequentially and too late.

That is an operational claim, and it is unusual in this field because it is falsifiable. It says the binding constraint is not capability, funding or will. It is architecture. And the strongest evidence for it is not private experience; it is the way international law has had to be written.

The obligation is already a timing obligation

Article 36(1)(b) of the Vienna Convention on Consular Relations imposes three linked duties on a receiving state. If a detained national so requests, the competent authorities must inform the consular post of the sending state that he has been arrested or detained. Any communication addressed by the detained person to the post must be forwarded. And the authorities must inform the person of these rights.

All three are governed by the same two words: without delay. The drafters did not specify a standard of assistance, a quality of access or a minimum of consular resource. They specified when information must move.

The International Court of Justice has twice had to say what that means. In LaGrand (Germany v United States), judgment of 27 June 2001, the Court found by fourteen votes to one that the receiving state had breached the Convention by failing to inform two detained nationals without delay of their rights under Article 36(1)(b), thereby depriving the sending state of the possibility of rendering timely assistance. Two features of the judgment are worth more than the outcome.

The first is what the Court held to be immaterial. It did not matter whether the detained nationals would have sought consular assistance, whether the sending state would have provided it, or whether the verdict would have differed. The injury was the loss of the possibility. International law, in other words, treats a routing failure as complete at the moment of the failure, independently of what the information would have produced.

The second is how the sending state eventually found out. Its consular post learned of the case from the detained nationals themselves, roughly a decade after the arrests. The failure was not partial. It was total, and — this is the operative point — it was invisible to the party that needed the information. A sequential system that stops at the first hop does not signal that it has stopped.

In Avena (Mexico v United States), judgment of 31 March 2004, the Court addressed the trigger and the remedy. The duty to inform arises once there are grounds to believe the person is a foreign national, and the Court observed that the notice could be given at the same moment as the reading of rights on arrest — that is, at the earliest node, without waiting for the question to be raised. The remedy of review and reconsideration, first identified in LaGrand, was held to require judicial process and a case-by-case assessment of actual prejudice, rather than a general clemency gesture.

What the law does not reach

Article 36 governs one hop: the arresting authority to the consular post. Everything after it is unregulated.

The post must reach the head of mission. The mission must reach the consular department at home. That department must reach whichever ministry actually owns the outcome — foreign affairs, justice, interior, sometimes a prime minister's office. Someone must reach the family. The family must reach counsel, and counsel must be admitted in a jurisdiction where they can act. Each of those is a discretionary handoff with its own queue, its own working hours and its own view of what is worth passing on.

This is where the failure occurs, and it does not look like failure from any single position within it. Each node acts reasonably and within time. The aggregate is late.

The architecture argument

The instinctive response to a late system is more capacity. That response is wrong here, and the reason is structural rather than managerial: in a sequential architecture, an additional person is an additional handoff. Capacity lengthens the chain it was meant to shorten.

The variable that determines the outcome is not how fast any node moves. It is how many nodes must move before the last one learns. A sequential system's elapsed time is the sum of its hops and it degrades under load, because every queue lengthens at once. A simultaneous system's elapsed time is the slowest single node, and it degrades gracefully, because the queues do not compound.

That is the whole of it, and it is why the fix is architectural. Hold one case picture. Establish, at the first hour, which parties must know, which must merely not be surprised, and which must not be told at all. Deliver to the first group at the same moment rather than in order of seniority. Record what was sent and when, because in a dispute about lateness the only defensible artefact is a timestamp.

The part that is uncomfortable

Sequence is not an accident of bureaucracy. It is a control.

Institutions route sequentially because each hop is an authorisation point. Someone decides what is passed upward, how it is characterised, and what is left out — and that discretion is often the only protection an official has. Simultaneity removes it. A ministry that receives the same case picture as a mission at the same instant loses the ability to shape what its own leadership sees, and it will resist that loss whether or not the resistance is articulated.

This means the coordination problem is not a communications problem with a technical solution. It is a problem about who is permitted to hold a version of events. It also explains why the function is usually performed, when it is performed at all, by a party outside every chain of command — someone with no promotion to protect and no departmental position to defend, whose only standing is that all sides find it useful to have them hold the picture.

A reader may reasonably disagree with this. The counter-position is that discretion at each hop is a feature: it filters noise, prevents premature escalation, and stops a mission from committing a ministry to a posture it has not agreed. That is true, and it is the reason the sequential model persists in matters that are not urgent. The argument here is narrower — that in a live consular emergency the filtering function is worth less than the elapsed time it costs, and that the two cannot be had together.

What a private party can and cannot do

It cannot compel a state to receive anything. It cannot confer rights under the Convention, cannot substitute for consular authority, and cannot make a ministry act on a timetable it has not accepted. Where a state will not move, better coordination produces a faster and better-documented failure, and it is worth saying plainly that this is sometimes the whole of the result.

What it can do is narrower and, in practice, decisive. It can keep a single case picture that every authorised party reads at the same time. It can identify the nodes that must know before the ones that merely want to. It can connect legal capability to the right jurisdiction without the delay of sequential engagement. It can ensure that the depth of an institution is present from the first hour rather than the third week. In short: it can convert a chain into a set.

The obligation in Article 36 was drafted around the same insight, in 1963, and it stopped at the first hop because that was the only hop a treaty could reach. The rest of the routing has always been someone's job. Usually it is nobody's.

The function described here is set out under Embassy Crisis Management & Consular Operations and Private Foreign Secretary; the standing information architecture that supports it is the Integrated Incident Command System.