After permission changes in a shared dataset
SN-2026-093 · Practice essay · Communities and governance
By Ivalu Petersen · · 3 pages · 1,533 words
Fictional demonstration publication. Not peer reviewed. Still North Arctic Institute is a fictional organisation. This paper is a teaching synthesis or illustrative practice document, not a report of original institute fieldwork. Published sources are identified below; illustrative examples are not measured results.
Abstract
Changed permission requires an accountable response even when a shared dataset cannot be recalled perfectly. This practice essay compares consent policy, Indigenous research-governance sources, Creative Commons terms, and data-citation guidance. A fictional controlled table released publicly without authorisation illustrates why an original breach, withdrawal from participation, requests to remove data, and discontinuation of a public download must remain distinct. The analysis follows copies, derivatives, backups, recipient notices, revised analyses, and persistent landing pages, separating reported actions from exposure that remains. It neither determines jurisdiction-specific legal duties nor promises universal erasure, and makes no claim to resolve a real community's authority. Its implication is proportionate provenance, legitimate decision routes, honest advance limits, and stewardship capacity that continues after publication and funding end.
Download the complete PDF (3 pages)
Full paper
A change of permission is a stewardship event
A dataset can travel beyond the team that collected it. Copies enter partner workspaces, figures appear in reports, and derived tables lose their obvious connection to the source. When permission changes, the institution cannot honestly promise that deleting its own download will erase every copy. It also cannot treat imperfect recall as a reason to do nothing. The task is to identify what authority and obligations apply, stop avoidable further exposure, and account for what can and cannot be changed.
This practice essay examines that task through Canadian consent and Indigenous-research policy, Inuit Tapiriit Kanatami's strategy, Creative Commons legal terms, and data-citation guidance. The case is entirely fictional. It is not a legal determination, an incident-response standard, or a claim that any community has approved these procedures. Its focus is the relationship between permission, distribution, and the enduring scientific record.
First establish what changed
Several events can sound like withdrawal while having different implications. An individual may stop participating in further research. They may request removal of previously contributed data. A governing authority may revise the conditions for a particular reuse. An institution may discover that a release exceeded the permission originally given. A rights holder may wish to discontinue offering a public download. These should not be collapsed into a single database flag called “consent revoked”.
TCPS 2 Chapter 3 states that consent can be withdrawn and participants can request withdrawal of their data or materials. It also requires explaining circumstances in which withdrawal is not possible, including practical limits after anonymisation or publication. That policy does not make every request technically reversible, but it makes advance explanation and an accountable response important.
Public licensing raises another distinction. CC BY 4.0 grants specified rights irrevocably under its terms, only to the extent the licensor has authority to grant them. Stopping a public download does not automatically cancel valid permissions already granted. Conversely, an institution cannot cure an unauthorised release merely by pointing to a licence it lacked authority to apply. Applicable law and qualified advice matter.
Fictional case: a controlled table escapes its intended route
Imagine an invented environmental project holding precise observation locations and contributor notes. An agreement permits two named analytical partners to use a defined table for one study, prohibits public redistribution, and provides a route for revising permission. A partner inadvertently puts the table in a public supplementary archive. Later, a contributor asks that their identifiable notes no longer be used. No real place, contributor, or Indigenous protocol is represented by this scenario.
Two issues now coexist. The public release exceeded the fictional agreement before the contributor's request arrived. The request also changes what future authorised use may continue. Treating everything as a new preference would conceal the earlier breach. Treating everything as a breach would fail to distinguish the contributor's present choice from the original permission. The response record needs both facts.
The institution should establish the affected records and scope through the agreed contact route, without requiring a contributor to restate sensitive reasons in a widely circulated ticket. It should involve the responsible governance, ethics, legal, and repository roles as applicable. Where a public exposure is continuing, avoidable distribution can be paused while the facts are clarified. Clarification should not become an excuse to leave a known inappropriate download available.
Follow the objects that can carry the information
A useful inventory separates the original table from copies, extracts, figures, metadata, and outputs. A figure can reveal a restricted location even after its underlying table is removed. A citation's landing-page description can contain identifying detail. A cached export may reappear when a routine workflow runs. The relevant unit of response is the information's actual path, not merely the filename first mentioned.
| Fictional object | Possible response | Residual limit to report |
|---|---|---|
| Institution-controlled download | Remove or replace the affected payload and stop automated redistribution | Earlier downloads may persist elsewhere |
| Named partner copies | Issue a specific notice under the agreement and request confirmation of the applicable action | Confirmation describes reported action, not omniscient proof |
| Public supplementary archive | Contact the repository through its correction or withdrawal process | Public mirrors and caches may not be recallable |
| Derived figure or table | Assess disclosure and analytical consequences separately | Removing rows does not necessarily reverse every derivative |
Not every retained object needs identical treatment. A genuinely non-identifying aggregate may no longer allow a person's contribution to be isolated, while a small-cell summary can remain revealing. Removing a source record may alter an estimate enough to require a correction. It may also leave a published conclusion unchanged. Both possibilities require examination; neither should be assumed because one is institutionally convenient.
Backups and retention obligations need explicit handling. A restricted recovery copy may be governed differently from an active analytical copy, but its continued existence still needs a legitimate basis, limited access, and a defined disposition. A workflow should not restore withdrawn material to public service during recovery. “We have backups” is not an explanation of whether the requested action can be honoured.
Notify without disclosing the dispute
A notice to downstream users should identify the affected version and material, the action requested or required, the authority or terms relevant to that action, and a contact route. It need not circulate a contributor's identity or private reasons. Where public explanation itself could expose sensitive facts, agree an appropriately limited description. Transparency concerns accountability, not publication of every confidential deliberation.
Recipients can be asked to identify onward transfers where their agreements support that request. Known recipients can then be contacted specifically. An open download with no access register presents a different practical limit. The institution should distinguish attempted contact, acknowledgement, reported deletion, and confirmed replacement of a controlled copy. Combining these into a single “resolved” count would overstate what is known.
A notice should also avoid claiming new legal powers. If a valid public licence already permits reuse within its scope, a voluntary request to stop is not automatically a binding revocation. Other obligations, including privacy, confidentiality, or applicable governance arrangements, may still be relevant. The relationship among them requires careful assessment, not a general statement that either licensing or consent always defeats everything else.
Keep the scholarly record intelligible without preserving harm
FORCE11's citation principles call for persistent identifiers and metadata describing the disposition of data beyond the data's lifespan. A withdrawn dataset can therefore retain a carefully limited landing page explaining that a release is unavailable and identifying an authorised replacement, where appropriate. Persistence need not mean an indefinitely downloadable sensitive payload.
The RDA recommendations for evolving data explicitly acknowledge that deletion under appropriate policies can make an earlier query no longer executable while its landing page persists. That is an important boundary on reproducibility. A repository should say when exact rerunning is no longer possible rather than quietly substitute new data under an old citation and imply that nothing changed.
Where a revised analysis is needed, release it as a documented new evidential state. Explain which broad analytical consequences changed without reconstructing protected contributions in the correction note. Keep any necessary confidential audit evidence under its applicable authority and retention rules. Scientific accountability does not require keeping every past version public, and confidentiality does not justify hiding that a published result has been corrected.
Authority and capacity continue after publication
TCPS 2 Chapter 9 recognises collective interests and the importance of research agreements, while distinguishing community engagement from individual consent. The National Inuit Strategy on Research places Inuit control and capacity at the centre of Inuit Nunangat research. Neither source gives this fictional institute authority to decide a real community's permission dispute or to impose one process across Indigenous peoples.
Agreements need institutions capable of acting after a grant closes. Responsibility should attach to a maintained role, with resources for requests, corrections, repository contact, and succession. People should not have to locate a former investigator's personal email address to exercise an agreed right. Nor should an outside institution shift the labour and cost of tracing its distribution mistake onto those affected.
Limits and implication
This case cannot settle jurisdiction-specific erasure rights, record-retention law, intellectual-property ownership, or the technical reversibility of every derivative. Fully anonymised pooled data and public dissemination can create real limits; describing a file as anonymised does not establish that it is. Requests also require context-specific interpretation where individual and collective interests intersect.
The implication is neither perfect recall nor resignation. Make the original limits honest, preserve provenance and distribution information proportionately, respond to changes through legitimate authority, and distinguish actions completed from exposure that remains. A responsible archive can acknowledge that some copies cannot be recovered while still reducing further harm and keeping the evidential record truthful.
References
- Canadian research agencies (2022). TCPS 2, Chapter 3: The Consent Process.
- Canadian research agencies (2022). TCPS 2, Chapter 9: Research Involving the First Nations, Inuit, and Métis Peoples of Canada.
- Inuit Tapiriit Kanatami (2018). National Inuit Strategy on Research.
- Creative Commons. Attribution 4.0 International: Legal Code.
- Data Citation Synthesis Group, Martone, M., editor (2014). Joint Declaration of Data Citation Principles. FORCE11.
- Rauber, A., Asmi, A., van Uytvanck, D., and Pröll, S. (2015). Data Citation of Evolving Data: Recommendations of the Working Group on Data Citation.
Keywords: permission changes; data withdrawal; downstream copies; repository stewardship.
Browse and filter the research library · Download learning guides