Summary

  • AFRINIC's own report of its 21 May 2009 meeting records counsel's view that the existing election regulation still permitted a candidate to vote because the Board had not adopted the resolution needed to amend or abrogate that regulation.
  • The ruling separated four things that institutional shorthand often collapses: a changed bylaw, the power to amend, an authorized Board act, and the amended rule actually applied at an election.
  • The strongest objection was serious: Alan Barrett believed newer bylaws had already clarified that candidates could no longer vote. Yet a sound policy preference could not supply the missing amendment counsel identified.
  • The episode does not establish that any candidate voted for himself, affected an outcome, or possessed a universal legal right. It establishes what AFRINIC's meeting record says governed that private corporate election at that moment.
  • The present implication is an evidence test, not a judgment on pending disputes: anyone acting through a purported Board, a court-appointed Receiver, or a coalition of supporters must identify the exact valid instrument, actor, scope, date and effective text behind each consequential act.

The question arrived too late

Election rules do their most important work before anyone reaches for a ballot. They tell participants who may act, which office has power to change the conditions, when a change takes effect and what everyone is entitled to rely upon. If those questions remain open until the meeting floor, the administrator is no longer simply applying a rule. The administrator is being invited to decide which version of institutional reality will count while the event is already under way.

That was the pressure point exposed during AFRINIC's tenth public-policy meeting on 21 May 2009. The organisation's report records Alan Barrett raising a point of order during the Board-election calling process. He understood that Mark Elkins had a right to vote in his own capacity and wanted to know why. Returning officer Sunday Folayan answered that he understood candidates to be eligible to vote.

The exchange immediately revealed that there was more than one kind of understanding in the room. Barrett said past practice had permitted candidates to vote in their own capacity, but he believed a newer set of bylaws had clarified that candidates could no longer do so. Adiel Akplogan responded from a different institutional reference point. According to the meeting report, he said the published election process provided for candidate voting, placed candidates on the eligibility list with one vote, had been public for a long time and had not been changed. It could not, he said, be changed at that moment.

Folayan acknowledged that the process had not been fixed and would need attention in the future. That admission matters because it stripped the dispute of an easy answer. AFRINIC could not simply say that one side had imagined the old practice. Its own officials were confronting a visible mismatch among constitutional expectation, published process and the act that had never been completed.

The decisive intervention came from AFRINIC counsel Ashok B. Radhakissoon. As the report records his interpretation, the bylaws could not change the election rules by implication. The newer bylaws empowered the Board to change the electoral regulation, but that power had to be exercised through a Board resolution and an amendment. Counsel said that had not been done. Until a resolution abrogated the existing rules, those rules continued to apply.

The election then proceeded. That immediate procedural consequence is enough to show that the ruling was not an academic seminar about drafting. It resolved which rule the meeting would administer. Nothing in the available record, however, supports the much larger claim that Elkins actually voted for himself, that such a vote affected any outcome, or that candidate voting is a universal entitlement. The institutional fact is narrower and more valuable: AFRINIC's own meeting report identifies the absent Board act as decisive.

Four legal states, not one changing mood

The 2009 exchange is easy to misread if every document carrying the word “rule” is treated as interchangeable. It becomes clear when the states are separated.

First, there was an existing election regulation or published election process. Akplogan's recorded explanation was that it listed candidates as having one vote and had not been changed. The report does not reproduce that document or give its exact version. It nevertheless records the institution relying on it as the operational rule then in force.

Second, there was a newer bylaw arrangement. Barrett understood it to have clarified that candidates could no longer vote. Counsel's interpretation was different in a crucial respect. He treated the new bylaws as granting the Board power to change the electoral regulation, not as silently rewriting that regulation by their own force.

Third, there should therefore have been an act by the empowered body: a Board resolution exercising the newly conferred power. A power to amend is not the amendment itself. It identifies who may perform the act; it does not pretend the act has already occurred. Counsel's recorded statement was that the Board had not passed the needed resolution.

Fourth, there should have been amended or abrogated election text with an effective date capable of being published and consistently implemented. Only then could an eligibility record and meeting procedure reflect a rule different from the old one. On the account given in the room, this last state had not been reached because the preceding authorized act had not occurred.

The chain can be stated without legal ornament: identify the operative regulation; identify the bylaw power; identify the body entitled to exercise it; find its quorate, dated resolution; find the resulting text and effective date; then apply it. Each link answers a different question. What is the rule? Who may change it? Did that actor actually decide? What did the decision change? When did participants become bound by the new version?

Institutional language often compresses this chain because compression is convenient. People say “the bylaws changed,” “the Board can do it,” “everyone understood,” or “this is how we now operate.” But a changed bylaw, a capacity, an expectation and a repeated practice are not the same state. The 2009 ruling preserved those distinctions in a few minutes of floor discussion. It is a compact lesson precisely because the missing link was so easy to name.

Why custom could not do the Board's work

An organisation needs habits. Staff must be able to perform routine tasks without reopening foundational questions at every step. Members also rely on settled practice because a process that changes unpredictably is costly to use. The problem begins when operational habit is asked to do constitutive work: to create, remove or narrow a right that the written structure assigns to a particular decision-maker.

Habit has no minutes of its own. It does not prove quorum. It does not identify a proposer, a vote, a conflict or an effective date. It rarely tells an affected participant when the old rule ceased to apply. Because habit leaves these matters blurred, it gives discretion to the person standing closest to implementation. A staff member, returning officer or meeting chair can present a preference as settled simply because the event cannot wait.

That discretion is especially dangerous at an election. A change to eligibility is not a housekeeping adjustment after the fact; it shapes the body that selects officeholders. If an official may invent the applicable rule at the point of enforcement, the same visible conduct can be accepted for one participant and rejected for another without an auditable source of authority. Even a well-intentioned correction becomes vulnerable because nobody can distinguish reform from selective administration.

The insistence on a resolution is therefore not reverence for paperwork. It is a way of allocating responsibility. A resolution tells members which body owned the choice. The amended text tells them what the choice was. Publication tells them when reliance became fair. Consistent implementation tells them whether the institution respected its own declared boundary. Formal action turns power into something that can be examined rather than merely felt.

The 2009 ruling also protects reform. It does not freeze every old procedure forever. If candidate voting was considered unsound, the Board could exercise the power counsel said the bylaws gave it. The institution could debate the change, resolve it, write it and publish it before the next point of application. Requiring that sequence does not prefer the substance of the old rule. It prevents the official who dislikes an old rule from replacing the authorized reformer.

That is why “everyone knew” is an inadequate answer. Everyone may have shared a policy expectation and still lacked an amendment. “The new bylaws intended it” is also inadequate where, on counsel's recorded reading, those bylaws delegated the next act instead of performing it. “The meeting had already begun” explains pressure; it does not create power. Necessity can justify keeping the existing process moving while uncertainty is addressed. It cannot enlarge the office of the person who happens to be holding the timetable.

The strongest case against candidate voting

Candidate self-voting can appear to present an obvious conflict. A participant seeking office may have a direct personal interest in the outcome, and an institution may reasonably want to prevent even the appearance that a candidate helped decide his own selection. Barrett's intervention deserves to be treated as the strongest contemporaneous case, not as an inconvenience to be dismissed. He understood the new bylaws to have resolved the issue and argued, in substance, that an election process should not contradict the organisation's superior constitutional arrangement.

That argument has institutional force. A regulation should not be allowed to defeat a genuinely self-executing higher rule. If the bylaw text itself had expressly and immediately prohibited candidate voting, the analysis might have focused on direct hierarchy rather than delegated amendment. But the actual texts are not available in the evidence for this article, and AFRINIC counsel did not describe that as the structure before him. His recorded interpretation was that the new bylaws empowered the Board to alter the regulation and that the Board had not completed the alteration.

The difference between those positions cannot be solved by announcing which policy sounds cleaner. It depends on the architecture of the instruments. Did the bylaw itself make a complete rule, or did it identify an authorised body and method for making the rule? Counsel answered that question at the meeting. This article can report his answer and analyse its institutional logic. It cannot independently certify his interpretation as a final statement of Mauritian law because the underlying bylaw, regulation and complete resolution register are not available here.

That limitation strengthens rather than weakens the procedural lesson. When constitutional texts are capable of generating opposing good-faith understandings, the institution should make the operative chain easier to inspect before the event. A versioned regulation, a dated resolution and a notice of effective change would have converted a clash of memories into an answer visible to all participants. The absence of those materials left the meeting dependent on an oral interpretation under time pressure.

There is also an important restraint on the conclusion. The report does not establish whether every candidate was also a member, what exact legal basis counsel had for language the report associates with a fundamental right, or whether any candidate actually exercised the disputed entitlement. A private election rule can confer a right within the corporate process without creating a public or universal right. The episode is not a referendum on the wisdom of self-voting. It is evidence that changing the operative rule required the actor and instrument the institution itself prescribed.

What AFRINIC's record proves—and what it cannot

Official records are indispensable because they preserve what an institution said and did. AFRINIC's meeting report is the sole contemporaneous primary account here of the exchange among Barrett, Folayan, Akplogan and Radhakissoon. It records the dispute, the officials' understandings, counsel's interpretation and the decision to proceed. Without it, the institutional act would be far harder to reconstruct.

The report is not a court judgment. It does not authenticate its own legal correctness merely because AFRINIC produced it. Its grammar and spelling are imperfect, and no audio has been located against which its wording can be checked. The underlying 2008 bylaw text invoked on the floor is absent. So is the election regulation and its version history. There is no comprehensive Board-resolution register in the record used here. The conclusion that the amendment had not occurred therefore rests on counsel's recorded statement and the surrounding exchange, not on an independent search of every possible corporate record.

The source also now presents an evidence-preservation warning. A direct request for the indexed first-party meeting report returned a missing-page response on 11 August 2026, although the report's extracted text remained available through a research index. A broken current link does not mean the 2009 record never existed. It means durable institutional accountability should not depend on a single live location decades after the event.

The appropriate evidentiary posture is exact. Attribute Barrett's understanding to Barrett. Attribute Akplogan's explanation to Akplogan. Attribute Folayan's acknowledgement to Folayan. Describe the decisive position as counsel's meeting interpretation. State that the election proceeded because the report says it did. Do not convert any of those statements into proof about an individual candidate's conduct or about the legal status of current officeholders.

This precision matters because official language can easily be made to carry more than it proves. When an institution calls an election successful, says directors have resumed duty, or describes collaboration with a Receiver, those statements establish its position and its conduct. They do not, by repetition, settle a challenge to the underlying authority. The record is evidence of the institution's claim, not authorship of the legal reality it records.

A private election inside a private coordinator

AFRINIC's election rules govern the selection of officeholders inside a private company. If validly made and applied, they can authorize people to exercise the bounded corporate functions assigned to those offices. They cannot turn the organisation into a legislature for Africa, a regulator of networks, a police force, a punitive authority, a confiscator, an owner of number resources or a court.

That boundary is not rhetorical modesty. It is the design condition under which a registry can remain trustworthy. AFRINIC maintains records and coordinates uniqueness so operators can rely on a common reference. The fact that the ledger is important does not mean the keeper created the networks, assets or lawful interests reflected in it. Registration records reality for coordination; it does not manufacture ownership or sovereignty.

The distinction also places the 2009 ruling in proper scale. Requiring AFRINIC to obey its own election instruments is not an argument that its internal electorate represents the African public. A corporate election may select a Board for the corporation. It does not create a continental demos. Participants and supporters can contribute knowledge, cooperation and confidence, but attendance and approval cannot bind absent operators, states or the public as if a private meeting had acquired sovereign mandate.

This is why a narrow procedural event deserves serious attention without being inflated. The candidate-vote dispute affected an internal corporate right. Counsel's answer showed that even inside that limited sphere, the organisation's officials could not derive power from implication when the governing arrangement called for a resolution. If the private company could not skip its own amendment step for an internal election, it plainly cannot use corporate ceremony to invent external powers the company never possessed.

The ledger should remain accurate and available while authority disputes are tested. Continuity of the record is an operational good. Yet continuity does not make the current administrator irreplaceable or expand a temporary office. A private bookkeeper may preserve records, carry out routine service and implement bounded lawful decisions. It must never treat dependency on the ledger as permission to rule the people and networks that rely on accurate bookkeeping.

LARUS supplies an important operator-reality discipline here, even though no event-specific LARUS publication is part of the 2009 account. Administrative ambiguity ultimately lands on networks that must keep serving customers, meeting contractual obligations and preserving continuity. That operational exposure is first-class reality; it cannot be reduced to background colour. At the same time, the absence of a LARUS account of this particular floor exchange must not be filled with claims it did not make.

The 2009 ruling as a present authority test

The contemporary value of the event lies in its method. For every consequential act, ask for the actor, the source of power, the instrument, the scope, the conditions, the duration and the effective text. Then ask whether the act can be kept reversible while authority remains disputed. Titles, announcements, filings, invoices, audits, institutional support and repeated operation may each provide evidence of a particular state. None is automatically the state that legally authorizes the act.

AFRINIC's update of 12 March 2026 illustrates why the distinctions remain live. The organisation described the September 2025 election as successful and said the announced directors had resumed duty. It also acknowledged continuing collaboration with the court-appointed Receiver pending formal discharge, acknowledged proceedings challenging the appointments of directors, and said legal actions had cost millions of dollars. Those statements prove what AFRINIC represented about its position and circumstances. They do not resolve the challenges AFRINIC itself acknowledged.

The wording creates a practical need for an instrument map. If a Board acts, which provision and valid corporate act establish its authority for that decision? If the Receiver acts, which exact term of the court-defined mandate covers the act, and does the continuing role change when formal discharge remains pending? If staff act, are they implementing a pre-existing delegation or relying on a disputed officeholder's direction? If supporters ask members or operators to accept the outcome, what document are they asking those parties to rely upon?

The answer cannot be “continuity requires it” without more. Operational continuity may justify preserving the ledger and routine service. It may support cautious interim administration. It does not create structural authority that a resolution, delegation or court order has not granted. Necessity is a reason to narrow consequential action and preserve reversibility, not an invitation to make contested control permanent.

NRS's formal member action of 24 June 2026 applies the same separation to present spending and authority. NRS disputed the purported Board's authority, urged members not to treat later member action as a cure for missing contemporaneous authorization, and demanded transaction-specific records showing who approved consequential commitments and on what basis. It highlighted a disclosed engagement term of USD 1,000 per hour, excluding VAT and disbursements, for C&A Law.

That figure must be used with care. It does not prove the reasonableness or total size of any bill. It does not prove that two lawyers worked concurrently at an aggregate USD 2,000 per hour; NRS expressly identified the missing records behind that unproved risk. Its importance here is structural. Engagement, work, invoice, payment, audit and later approval are different acts. A paid invoice may show that money left an account. An audit may show how the transaction was recorded. Neither fact, by itself, proves that the authorised actor approved the expenditure at the time it was incurred.

The analogy to 2009 is exact without making the events identical. A bylaw, a grant of power, a Board resolution, an amended regulation and implementation are distinct states. So are an engagement, an instruction, performed work, an invoice and a payment. In both settings, collapsing the sequence benefits whoever controls the later record because the record can be presented as proof that the earlier authority must have existed. Good accountability refuses that backward inference.

BTW's current election research reaches the same boundary from another direction. Its conclusion is that the Receiver-led September 2025 process disregarded express bylaw mechanics and that changing prescribed procedure through operational redesign, rather than through the authorised instrument, deepened legitimacy risk. That is BTW's first-class research conclusion. It is not evidence of the 2009 rule text, and this article does not turn the earlier episode into a general account of the later election.

The relevance is that an old institutional test remains available: name the exact power that permitted the change before treating implementation as proof of authority.

Supporters of present control are not disqualified from making their case. They may point to continuity, participation, institutional need, records or public confidence. But support has a defined evidentiary role. It can make cooperation easier and show that people accept a course. It cannot write a missing resolution, widen a court order or convert a contested title into a final adjudication. A supporter is not the authorising principal merely by being supportive.

Critics carry a corresponding obligation. They should not treat unresolved litigation as if it already established their preferred judgment. The evidence supports a demand for documents, precision and reversible action. It does not support declaring the current Board conclusively invalid or the Receiver conclusively discharged. The disciplined position is neither automatic recognition nor automatic condemnation. It is refusal to substitute continuity, opposition or confidence for the instrument that actually governs.

The economic cost of an unmade act

Procedural ambiguity is often described as a lawyer's problem until the institution begins spending money, entering commitments or making decisions on which networks rely. Then uncertainty becomes an operating cost. Members must hire advisers to reconstruct authority. Courts are asked to settle questions that clear records could have prevented. Staff must decide whose directions to follow. Counterparties price in the risk that a decision may later be challenged.

The NRS disclosure of the USD 1,000-per-hour engagement term gives that cost a specific edge without proving the total expense of any matter. AFRINIC's own statement that legal actions had cost millions of dollars supplies its self-described scale of litigation burden. Neither figure identifies a simple culprit. Together they show why ambiguity about who could act, when and for what purpose is not harmless institutional fog.

The burden is not confined to legal fees. When rule changes are undocumented, members cannot plan participation with confidence. When authority is disputed, suppliers and advisers face uncertainty about instructions. When a private registry allows contested control to shape consequential decisions, operators must consider whether the ledger and related services will remain dependable. Every dependency built on an act whose authorization cannot be produced makes later correction more expensive.

Formal amendment reduces those costs by creating a shared reference. It permits reform without surprise and challenge without guesswork. It also changes institutional incentives. A decision-maker required to sign a resolution, state its basis and publish its scope must own the consequences. An implementer permitted to rely on atmosphere can take power while leaving responsibility nowhere.

That asymmetry is why the missing act matters more than the apparent smallness of the 2009 issue. One candidate's eligibility may seem modest beside current litigation and financial exposure. Yet the structure is the same structure by which larger powers appear: an office claims that a desired state already exists, practice begins to reflect it, costs accumulate around it, and accumulated reliance is later offered as the reason the state cannot be undone.

The safest point of intervention is before those dependencies form. Ask for the authorized instrument before the ballot, instruction, contract or structural change. If it cannot be produced, pause what can safely be paused, preserve routine service, document the uncertainty and seek the decision from the body that actually has power. Delay may carry a cost, but irreversible action under invented authority carries a larger one.

The small case that keeps power honest

The institutional lesson of 21 May 2009 is not that an old rule deserves eternal loyalty. It is that a right changes when the authorized actor performs the authorized act. A private organisation may decide that a practice is undesirable. It may empower its Board to amend a regulation. It may then adopt and publish the amendment. What it may not do is treat those separate steps as a mood that staff can sense and enforce.

Counsel's ruling made the hierarchy concrete at the moment when convenience argued for compression. Barrett supplied the substantive objection. Akplogan identified the public process that still existed. Folayan acknowledged that the process had not been fixed. Counsel named the missing resolution. The election proceeded under the rule the institution had actually made, not the rule some participants believed it meant to make.

That sequence is valuable today because control is still tempted to present itself through visible ceremony while hiding the instrument that made it lawful. Elections can confer bounded corporate office when their rules and authority chain are valid. They cannot generate sovereignty. A Receiver can act within the exact mandate of a court. Institutional need cannot enlarge that mandate. Members and supporters can cooperate. Cooperation cannot adjudicate a pending dispute.

AFRINIC's proper strength lies in being a dependable private bookkeeper and coordinator: accurate records, narrow administration, transparent instruments, reversible interim conduct and a clean path of succession. Its weakness begins when the record is mistaken for authorship, the ceremony for mandate or repetition for authority. The rule AFRINIC had not changed in 2009 is therefore more than an archival curiosity. It is a reminder that institutional power becomes legitimate only where its source, owner and limits can be shown before the power is used.