Summary

  • Multiple roles are normal in specialist governance and often valuable: experienced entities connect operational reality, policy history and institutional constraints. The risk begins when role count is mistaken for independent support.
  • Disclosure must be intervention-specific. An employer line alone cannot show whether a person speaks personally, for a client, for a working group, from an advisory seat or as a chair performing a procedural duty.
  • Participation analysis should preserve four units—person, organisation, formal role and claimed mandate—and map common control and time without assigning opinions by association.
  • Strong safeguards combine current statements of interest, spoken role declarations, public decision traces, recusal where power conflicts, privacy and safety exceptions, and aggregate concentration reports rather than blacklists or vote discounts.

The same voice can enter the record through several doors

Imagine a familiar conference intervention. A network executive approaches the microphone. She also co-chairs a working group and occupies an advisory seat in another body. Her employer operates infrastructure affected by the discussion. Her working group has previously considered the problem. Her advisory committee may later issue a formal view. When she speaks, how many voices have entered the record?

The honest answer depends on what she says she is doing. She may offer personal technical experience, communicate an authorised employer position, explain the working group's agreed conclusion or perform a chair's duty to clarify process. Those acts can all be legitimate. They are not automatically independent evidence from four constituencies.

Internet governance makes such overlap likely. The subject is specialised, volunteer time is scarce and experienced people are repeatedly asked to serve. An operator who understands routing may become a policy author, chair, reviewer and liaison precisely because the community trusts that competence. Removing everyone with several roles would destroy institutional memory and leave power with people whose connections are merely less visible.

The governance problem is therefore not “multiple hats” as moral failure. It is ambiguity. A transcript may list one affiliation while the speaker relies on another role's authority. A consultation summary may count an advisory statement, a working-group contribution and several employees as separate support even when they share organisational control or authorship. A chair may move between facilitating consensus and advocating substance without marking the transition.

Role clarity is a method for preserving useful overlap while preventing representational inflation. It asks who is speaking, in which capacity, under what mandate, with which relevant interests and at what stage of authority. It does not decide whether the argument is right.

Four units must remain visible

Participation records commonly collapse four units. The person is the human contributor. The organisation is an employer, client, member, funder or institutional home. The formal role is an office such as chair, liaison, council member or adviser. The mandate is the scope of authority to communicate or decide for another principal.

One person can have several organisations over time and several simultaneous roles. One organisation can employ many speakers. One role may carry a narrow procedural mandate but no authority to state a substantive position. One association representative may hold a documented mandate reflecting consultation with hundreds of members, while ten employees of a company may all speak personally.

Counting people captures human participation but can exaggerate independent organisational breadth. Counting organisations can erase disagreement and individual expertise. Counting roles can make one person's repeated service look like several constituencies. Counting formal statements without tracing authorship can miss that the same small group shaped them.

No single deduplication solves this. A defensible report should present layers: unique people, declared organisational groups, formal offices and claimed representational mandates. It should identify known overlap and unknown data. The layers diagnose concentration; they do not allocate voting weight.

This distinction also protects contributors. A entity can say “I work for this organisation, but I speak only for myself.” The record then provides relevant context without attributing the view to the employer. Conversely, a person claiming to speak for a membership association should identify the authority and consultation behind that claim rather than relying on the association's name.

Existing rules already recognise role ambiguity

The issue is not theoretical. RIPE's working-group chair description says chairs and co-chairs should endeavour to clarify whether they speak for themselves, their employers or the working group. The sentence recognises three capacities inside one person and places responsibility at the moment of participation.

ICANN's adopted Community Entity Code of Conduct discusses existing GNSO Statement of Interest questions. These include current employer and position, financial and representational relationships, material interests, arrangements concerning participation and working groups or other chartered teams. The breadth reflects a sensible insight: a job title alone cannot explain relevant influence.

The ccNSO Statement of Interest page describes self-declarations as tools for transparency, fairness and accountability where interests could influence views or decisions. ALAC profiles identify terms, selection bodies and current declarations for members and appointees.

These mechanisms are valuable foundations, but forms can become static biographies. A declaration updated annually may not show a new client, a temporary drafting role or the capacity used in a particular intervention. Conversely, a long list of affiliations may imply that every one influenced every statement.

The next step is not maximal disclosure. It is role-specific disclosure connected to the act of governance. A concise declaration at the microphone or top of a submission can identify the relevant capacity while the fuller statement provides background.

Affiliation is context, not ownership of speech

Institutional audiences often hear an employer name and assign the person's words to the company. That shortcut can be wrong and harmful. Employees may participate in personal time, disagree with corporate policy or lack authority to bind the organisation. Employers may prohibit public association with a controversial view.

At the same time, employment can shape access and perspective. It pays salary, supports travel, provides deployment experience and creates material interests. Pretending it is irrelevant can hide structural concentration. The solution is a two-part statement: relevant affiliation and claimed capacity.

“Employed by X; speaking personally” is meaningful. It lets listeners assess experience and potential interest without manufacturing an organisational endorsement. “Speaking for X under responsibility for its network operations” makes a different claim that can be checked. “Working for X and serving as chair; answering only about procedure” narrows the authority further.

Records should preserve the speaker's own declaration. Institutions should not infer that every employee shares a position or count a personal intervention as a corporate submission. Where evidence suggests coordinated organisational advocacy, analysis can report common affiliation without asserting hidden instruction.

This principle is essential for junior staff and people in restrictive environments. Forced public association can expose employment or safety risks. A confidential channel or broad sector disclosure may provide enough context. Role transparency should increase accountability without making participation conditional on personal vulnerability.

Working-group office changes the meaning of ordinary speech

A working-group chair does more than add another affiliation. Chairs manage queues, frame questions, interpret objections and may declare consensus. Their procedural acts can shape which substantive positions become visible. When a chair advocates, listeners need to know the chair has switched capacities.

The switch is not inherently improper. Chairs often possess deep knowledge and remain members of the community. A rule of complete silence could deprive debate of useful evidence. But the person should yield facilitation where advocacy creates a real conflict, or clearly mark the intervention and allow another chair to manage it.

Co-chair structures help when responsibility can transfer. Meeting records should identify who chaired each agenda item and who assessed consensus. If a chair authored or strongly advocated a proposal, another authorised person should handle the decisive assessment. Recusal protects both the process and the chair from an impossible dual role.

Working-group statements also need provenance. Did the group formally agree, did chairs summarise discussion, or did one officer write an interpretation? “The working group believes” should link to the method and record. Otherwise an office becomes a megaphone that turns personal judgment into collective voice.

These practices do not require legalistic meetings. A brief oral declaration and clear minutes are enough in most cases. The goal is to stop procedural authority from lending invisible extra weight to a substantive position.

Advisory seats carry a bounded mandate

An advisory committee member may be selected through a region, nominating body or constituency. The seat carries duties defined by governing documents. It does not necessarily authorise the holder to speak for every person in the geographic or public category associated with the seat.

When the member contributes to another forum, three possibilities arise. The person may state an adopted advisory position, explain committee discussion or offer a personal view informed by service. Each is legitimate if labelled. Only the first should be counted as a formal institutional statement, and even then the record should link to adoption.

The selection source matters. An appointee chosen by a nominating committee has a different chain from someone elected by eligible members. Neither chain should be exaggerated. A regional label can ensure geographic diversity without constituting a regional electorate. An advisory office can create responsibility without a general public mandate.

Term dates matter too. Former office-holders often retain authority in social perception long after their formal role ends. Profiles and transcripts should distinguish current, former and acting positions. A former chair's expertise remains valuable; the title should not silently travel with the person as current institutional authority.

The bounded-seat principle allows advisory participation to remain influential. Formal advice can carry collective reasons and a documented process. Personal expertise can stand on evidence. The distinction prevents both from borrowing legitimacy from the other.

Consultants make the organisational graph harder

Consultants may work for several clients while maintaining independent practices and community roles. Publicly naming every client can violate confidentiality. Naming none can conceal material interests. A simple employer field cannot describe the relationship.

Disclosure should focus on relevance. A consultant can state that they advise one or more organisations materially affected by the issue, whether they are authorised to represent a client, and whether confidentiality prevents naming. A protected conflict reviewer can receive more detail where the role has decision power.

Listeners should not assume that consulting creates instruction. Consultants often hold their own views. Nor should a consultancy name be counted as an independent constituency each time the same underlying client interest appears through employees, trade groups and advisers.

Aggregate analysis can group disclosed representational relationships without exposing confidential identities. For a consequential decision, an independent reviewer can test whether apparently diverse submissions share a common principal. Public findings can state concentration ranges and limitations.

The standard should be proportional. An ordinary commenter offering technical evidence needs less disclosure than a chair assessing consensus or a member voting on procurement. Power, not curiosity, should determine the depth of review.

Associations can compress mandates—or merely claim them

Membership associations complicate headcounts in the opposite direction. One person may carry a genuinely broad mandate after consultation with many members. Counting only the speaker understates the institutional process. But an association staff statement may also reflect internal judgment without member approval.

A submission should explain its mandate: board-approved, member-balloted, committee-developed, staff-authored after consultation or personal. It should identify the eligible membership denominator and participation where safe. This does not determine correctness, but it lets readers understand the chain.

Association and member submissions may legitimately coexist. Members may disagree with the collective position or emphasise distinct effects. Deduplication should show overlap rather than deleting either voice. A report might say that support included one association statement developed through a stated process and several members, some speaking independently.

The association should also disclose whether a small subset dominated drafting. Again, this is not a reason to reject the argument. It calibrates claims such as “the industry supports.”

Mandate provenance turns organisational scale into auditable context. Without it, a single logo can be treated as thousands of voices while ten employees are treated as ten organisations. Both errors disappear when the record separates person, organisation and authorisation.

Staff support can create invisible authorship

Secretariats provide research, meeting support, editing, legal analysis and institutional memory. This work is necessary. It can also shape agendas and text while public narratives credit only volunteer bodies.

Documents should distinguish community authorship, staff drafting under direction and staff recommendation. Version histories can show who requested changes without exposing routine editorial labour. Legal or implementation advice should identify assumptions and the authority of the final decision-maker.

Staff attendance should not be counted as independent community support. Nor should staff expertise be discounted because it is employed. The record should identify when employees answer operational questions, facilitate, advise or advocate an organisational position.

Where staff support one side of a contested issue more extensively because that side holds a formal office, equal process may require assistance for material objections or an independent assessment. Resource asymmetry can determine which view appears polished and feasible.

Transparent authorship protects the institution. It avoids blaming volunteers for text they did not control and prevents community branding from hiding executive choice. The issue is accountability, not suspicion of professional support.

Role switching needs a grammar

Entities cannot recite a biography before every sentence. Meetings need a concise grammar: “personal,” “for employer,” “for client,” “as chair on process,” “reporting group position,” or “as liaison.” Written systems can provide equivalent labels.

The grammar should be optional for interventions where capacity is obvious and required where a person invokes institutional authority or holds a decisive office. Chairs can prompt gently: “Which capacity are you using?” The correction should not become a rebuke.

Minutes should capture the declaration with the intervention, not only in a separate roster. Searchable links to current statements of interest can provide detail. If the role changes during discussion, the record should change too.

Machine analysis can then count roles without guessing. It should not classify sentiment or infer undisclosed relationships. Human review remains necessary for ambiguous statements and context.

A shared grammar across ICANN and RIR forums would help entities who cross institutions. It need not impose one conflict policy. It would make ordinary disclosure interoperable and reduce the social cost of clarifying a hat.

Time is a dimension of every affiliation

Affiliation graphs become inaccurate quickly. People change jobs, clients, offices and committee terms. A current profile applied to a five-year-old transcript can rewrite history. A historical employer line applied today can misstate a person's interests.

Every relationship should have a start, end and “known as of” date. Statements of interest need visible revision history. Meeting records should preserve the declaration made at the time rather than dynamically replacing it with a current profile.

Cooling-off concerns also arise. A former employee may retain knowledge and relationships without current financial interest. A newly appointed adviser may have participated in drafting before taking office. The graph should show sequence without assuming influence.

Retrospective research must avoid reconstructing sensitive career histories beyond public governance need. The objective is to explain decision conditions, not permanently monitor volunteers.

Time-aware records support fair correction. A entity can update a current interest without erasing what was disclosed when a decision occurred. Readers can see both the historical context and later change.

Common control is more relevant than matching names

Several brands may belong to one corporate group, while similarly named organisations may be independent. Counting registration strings can therefore overstate or understate organisational diversity.

For aggregate governance analysis, a reviewed ultimate-control field can group publicly documented parent relationships. The method should record uncertainty, joint ventures and changes. It should never infer ownership from domain names alone.

Common control does not prove coordinated speech. The report should say that contributors share an organisational group, not that they conspired. Their evidence may be independent and technically valuable. The information helps readers assess breadth.

Conversely, independent organisations can coordinate through an association or campaign. Claimed representational relationships belong in a separate edge. The graph needs typed connections rather than one undifferentiated affiliation link.

Public reports can present concentration bands and network diagrams stripped of personal identity. Detailed linkage should be available only to authorised reviewers under retention and correction rules.

An affiliation graph is a map, not a verdict

A practical graph would use nodes for people, legal entities, organisational groups, formal bodies, roles and submissions. Edges would record employment, client representation, membership, appointment, authorship and claimed speaking capacity. Each edge would carry time and provenance.

The graph could answer bounded questions. How many unique organisational groups contributed to a proposal? How many formal statements shared authors? Did the chair also author the text? Were several apparent constituencies connected through one disclosed principal? It could not answer whether a entity was sincere or an argument correct.

Unknown should remain a valid state. Not every relationship is public or relevant. Confidence labels can distinguish self-declaration, official office and independently verified corporate control. Inferences should not be mixed with declarations.

Publication should favour aggregates. An individual graph can create security and harassment risk. Access to detailed data should be purpose-limited, logged and reviewable. Entities need correction and appeal.

Used modestly, the graph prevents role multiplication from becoming constituency multiplication. Used aggressively, it becomes surveillance. Governance must define the boundary before collecting the data.

Deduplication should produce several counts

The phrase “deduplicate entities” suggests one correct number. In reality, a report needs several counts: unique people, unique declared employers, unique controlled organisational groups, unique formal roles, unique authorised submissions and unique affected sectors.

Each count should state its question. Human breadth asks how many people participated. Organisational breadth asks how many independent groups appeared. Institutional breadth asks how many bodies adopted positions. Mandate breadth asks how many principals authorised representation. None replaces the others.

Reports should show overlap. A Venn-like aggregate may reveal that several people, groups and roles converge. It should avoid a single adjusted score that hides methodology.

Repeated voices can still add information. An engineer and lawyer from one company may provide different evidence. Deduplication should not delete contributions; it should prevent their number from being narrated as independent constituency support.

This multi-count method is especially important for consensus. Chairs can assess reasons while reporting that apparent support came from a smaller number of organisational groups. The conclusion remains qualitative but better informed.

Speaking twice is not voting twice

Open discussion often permits unlimited relevant interventions. Formal elections usually restrict ballots. Confusing the two creates unnecessary hostility toward active contributors.

A person who speaks in a working group and later advises a committee has not necessarily voted twice. The second body may consider the first record and exercise separate authority. The issue is whether the same contribution is presented as independent validation at each stage.

Decision traces should show lineage. If an advisory statement relies on a working-group conclusion authored by overlapping people, say so. Independent review means more than a new letterhead; it requires reviewers able to test assumptions and disagree.

Where a formal vote occurs, eligibility and conflicts govern ballots. Where rough consensus occurs, the chair evaluates reasons, not one-person-one-vote arithmetic. Role-aware concentration informs confidence without importing electoral rules.

The title's question is therefore rhetorical. The concern is not literal duplicate ballots but duplicated legitimacy: one interest appearing as employer evidence, community consensus and advisory approval without visible overlap.

Recusal belongs where roles meet power

Disclosure alone may be limited public evidence when a entity can decide a matter affecting a material interest or review their own work. Recusal transfers that act to another authorised person while preserving the entity's ability to provide evidence where appropriate.

Triggers should be defined in advance: direct financial interest, authorship combined with consensus assessment, appeal review of one's decision, or other specified conflicts. Vague standards invite selective enforcement.

RIPE's principles for chair remuneration require disclosure of potential conflicts tied to leadership roles and attention to undue influence. The principle is broader than remuneration: leadership legitimacy depends on visible handling, not merely disclosure.

Recusal records should identify the decision, reason category and replacement authority without exposing unnecessary private detail. Excessive recusal can also disable volunteer institutions, so proportionality matters.

The strongest design uses multiple officers, independent appeal and documented handoff. It does not expect one person to be simultaneously advocate, chair, judge and reviewer.

Privacy and safety are part of legitimacy

Calls for transparency can privilege people secure enough to disclose. Employees may face retaliation. Activists may operate under hostile governments. Consultants may owe lawful confidentiality. People may discuss abuse affecting themselves.

Institutions should collect only information necessary to understand relevant power. Broad sector disclosure, confidential review or pseudonymous contribution may suffice. Public raw data should not expose home location, private clients or inferred beliefs.

Small-group aggregates can re-identify. Reports should suppress cells and avoid combining tables that reveal individuals. Retention should be limited, and access logs available for oversight.

Safety exceptions should not become an invisible exemption for powerful interests. A trusted reviewer can confirm that a material relationship exists and describe its category without naming the party. Decision-makers can apply recusal privately with a public reason class.

Transparency and privacy are not opposites. Both constrain institutional power: one prevents hidden influence; the other prevents the institution from turning participation into personal exposure.

Newcomers experience role ambiguity as hierarchy

Regular entities know who is a former chair, employer representative, staff expert or influential author even when the record does not say. Newcomers see a flat microphone list and miss the social authority behind names.

Role declarations level this informational field. They show why one intervention receives an immediate response and help newcomers identify decision routes. Public organigrams, term dates and chair powers reduce dependence on insider knowledge.

The language should remain accessible. A string of acronyms can reproduce exclusion. Session pages can explain each role and whether it decides, advises or facilitates.

Newcomers should also be able to speak without a prestigious affiliation. Evidence must be evaluated on substance. A role-aware record should expose institutional power, not create a caste system in which unaffiliated voices appear weak.

Mentoring can teach how to declare capacity and challenge a summary. That practical literacy is more important than memorising every body.

Diversity metrics can overcount institutional independence

A panel may be geographically and demographically varied while members share one employer, professional network or selection pathway. Conversely, organisationally independent entities may look similar in a photograph. Descriptive and institutional diversity measure different risks.

Both matter. Geography and lived experience can reveal neglected effects. Organisational independence reduces common-interest concentration. Neither proves mandate.

Reports should present dimensions separately and avoid a composite diversity score. Sensitive identity disclosure must remain voluntary and aggregate. Organisational and role data should follow relevance and privacy rules.

Selection bodies can use the information to avoid repeated concentration, but they should not treat people as interchangeable category tokens. Competence, independence, experience and affected perspectives need reasoned balancing.

The aim is not a perfectly representative room. It is a room whose limitations are known and whose authority is calibrated accordingly.

Formal statements need an authorship trail

A committee statement can appear as one institutional voice even when one person drafted it and a small quorum approved it. That may be entirely valid under the rules. The process should be visible.

Publish the call for input, draft author or drafting team, material revisions, approval method, participation and dissent where rules permit. Link underlying discussions. Staff editing should be identified at an appropriate level.

Authorship overlap across bodies is particularly relevant. If the same person writes a working-group recommendation and an advisory endorsement, the second review should show independent scrutiny rather than mere adoption.

This does not require exposing every edit. A provenance note can state who led drafting and how approval occurred. The statement's authority then rests on the body's process, not an illusion of spontaneous collective authorship.

Readers can also distinguish formal advice from a chair's letter or liaison report. Precise document types prevent one voice from being counted twice under institutional names.

A minimum role-aware record

For consequential work, each entity with formal power should maintain a current statement of interest with revision history. Each intervention invoking authority should carry a capacity label. Each meeting record should identify chairs by agenda item, authorship, recusals and the method used to adopt outputs.

Annual aggregate reporting should show unique people, organisations, groups and roles; concentration of authorship and leadership; multiple-role prevalence; and missing or stale declarations. It should not publish individual rankings.

Decision pages should trace statements across bodies and identify overlap. Independent review should test methods, privacy and corrections. Entities should have a route to challenge misattributed capacity.

These requirements scale with power. Casual discussion needs little overhead. Consensus declaration, formal advice, elections, appeals and binding decisions need more.

The standard turns disclosure into usable accountability. A pile of biographies cannot show who spoke twice; a role-aware decision record can.

Appeals must not reproduce the original overlap

Role clarity matters most when a decision is challenged. An appeal that moves from one officer to a collective containing the same officer, close co-authors or people acting under the same organisational interest may provide a new procedural label without independent review. The appeal record should therefore identify the original authors, decision-makers and relevant relationships, then show who was excluded from reconsideration and why the remaining body had authority.

Independence is contextual rather than absolute. A reviewer may know the entities and still judge fairly; specialist communities cannot import strangers for every dispute. The practical test is whether the reviewer has a material interest, exercised the contested power, authored the decisive finding or is otherwise committed to defending their own act. When the pool is too connected, an external chair or independently appointed panel can narrow the problem without taking over ordinary community governance.

The appellant also needs role clarity. A company challenging a result, an individual member questioning procedure and a formal constituency invoking a charter right present different mandates. Their arguments should receive substantive assessment, but the remedy and standing may differ under published rules. An affiliation should not be used to dismiss an appeal, just as a prestigious office should not guarantee success.

Appeal outcomes should disclose the issue, applicable standard, recusals, evidence considered and reasons. A decision trace can then show whether a second institutional voice represented genuine scrutiny or the same network speaking again. This is the point at which deduplication becomes a safeguard rather than an academic count.

Funding creates another edge in the graph

Travel support, employer time, secretariat assistance and project grants determine who can sustain participation. Funding does not prove control, but omitting it can make institutional independence look greater than it is. Disclosure should focus on support materially connected to the work: who paid for participation, whether the funder set an expected position and whether continued support depends on an outcome.

Routine employer salary is usually captured by affiliation. Sponsored travel or a funded research project may need a separate declaration where relevant. Fellowship support should not be treated as instruction; programmes typically fund access and learning rather than policy positions. The graph must distinguish support from mandate.

Funding transparency should avoid a class penalty. Self-funded entities are not automatically more independent, and people needing travel support should not carry a presumption of institutional loyalty. Wealth can hide influence as effectively as sponsorship can reveal it. The question is whether a relationship creates a relevant incentive or claim of representation.

Aggregate reporting can show how participation is financed across broad categories—employer, institution, programme, self or mixed—without exposing personal income. Decision-makers with direct financial conflicts need deeper review. Ordinary attendees need only proportionate context.

Adding funding as a typed, time-limited edge helps explain repeated presence without converting support into suspicion. It also lets boards see whether a nominally open process depends on a narrow set of employers willing to pay for governance labour.

Measurement itself needs institutional separation

The body whose legitimacy is being measured should not have unlimited discretion to classify entities and declare its own diversity. Staff may responsibly collect records, but methods, corrections and aggregate claims need review independent of communications incentives. Otherwise categories can be chosen to maximise apparent breadth.

A small measurement panel could include privacy, community and methodological expertise. It would approve definitions, audit samples and publish limitations. It would not investigate political beliefs or decide policy. Its authority would stop at data integrity.

Entities should know what is collected before contributing, how long it remains, who can access linkage data and how to correct errors. New uses should require notice rather than relying on a broad consent box from conference registration. Public tables should be tested for re-identification and should never expose sensitive clients through combinations of role and geography.

The panel should publish disagreements about method. If common-control grouping is uncertain, show a range. If mandate declarations are incomplete, do not impute them. If a platform change breaks comparison, mark the break. Honest uncertainty prevents the measurement function from becoming another advisory seat that silently amplifies the institution's preferred story.

Independent measurement completes the article's central rule: no role should validate itself merely by appearing under a second name. The same discipline applied to speakers, chairs and advisers should apply to the people who count them.

Conclusion: count connections without cancelling contributors

Internet governance depends on people willing to serve repeatedly. Their overlapping roles carry knowledge across technical, corporate and public-interest domains. That density can make institutions work. It can also make a small professional network appear as several independent layers of approval.

The answer is neither suspicion nor forced turnover. Preserve person, organisation, role and mandate as separate units. Ask speakers to identify relevant capacity. Make statements of interest current and time-aware. Show authorship, handoffs and overlap across bodies. Use recusal where a person would exercise decisive power over their own interest or work. Protect privacy and safety.

Then interpret the record modestly. Several interventions from one organisational group can contain valuable evidence without becoming several constituencies. One association statement can carry a broad mandate if its process supports the claim. A chair can contribute substance if facilitation transfers. An adviser can speak personally without turning the seat into a borrowed mandate.

Who speaks twice is less important than whether the institution tells us which voice we are hearing. Once that is visible, arguments can stand on evidence, offices on defined authority and community claims on a traceable process rather than multiplication by affiliation.

That clarity also makes disagreement safer: contributors can challenge an institutional position without attacking the individual, and institutions can correct a role claim without erasing the evidence that person supplied.

Sources