Skip to content

Open proposals and questions

Everything currently open for comment in the Foundry: proposed changes to the site and the ontologies, and open questions and tasks. You do not need a GitHub account to take part: see how to participate. Links to GitHub open the full record.

Updated 16 November 2026, 22:17 UTC.

Open for decision (4)

These proposals are being decided now. Voting members can sign in and vote below, use the ballot link emailed to them, or vote on GitHub. The count is updated within minutes of each vote. See Governance.

Review procedure: who decides each step, and an automated status check

Substantive change, open for decision since 4 November 2026. Substantive change: adopted after a comment period of at least 14 days, once every Operations Committee member approves and a majority of the voting members who vote approve.

Status: Waiting for comment period: 2 days left; approval from William Mandrick.

Voting member Vote How
David R. Koepsell (committee) Approve on website
William Mandrick (committee) not yet voted
Barry Smith (committee) Approve on website
Read the full proposal

Clarifies how registry entries move past draft, and automates the checking.

The clarification (docs/sop/review.md)

Step Requirements Decision
draft → candidate the entry names the release (release: {version, tag}); the repository is public; the tag exists; lof-{id}.owl is in the tagged release; automated checks pass on the tagged commit routine
candidate → accepted the same, for the release reviewed; two Ontology review issues of that release recommend Accept, from two reviewers, one a committee member and the other not a maintainer; the entry records review: {date, issue} substantive
any → deprecated the successor, if any, named in replaced_by substantive

Also stated: - While the Editorial Working Group has no lead, the Operations Committee assigns reviewers. - Reviews are due within 30 days. - Reviews of an earlier release do not count. - No other status change is allowed (for example, draft straight to accepted).

The metadata policy and the registry schema gain the optional release field.

The automation (.github/workflows/registry-status.yml, scripts/registry_status.py)

  • On a pull request that changes an ontology's status: it checks the requirements for that step and reports them as a checklist in one comment that updates itself.
  • When all are met, it adds decision: routine or decision: substantive, which opens the vote: ballots, the website, and the bot's adoption, as for any decision.
  • If a requirement stops being met, it removes the label. Draft pull requests are never opened.
  • On a review issue being filed, edited, or closed: it rechecks the acceptance pull request for that ontology.
  • On a push to main that makes an ontology a candidate: it opens a review tracking issue, which asks the committee to assign two reviewers, links the review form, and sets the 30-day due date.

It uses pull_request_target, so it runs from main and reads the pull request's files through the API; a proposal's own code is never run.

Tested

  • 14 checks against a simulated GitHub:
  • missing release, failing checks, then passing (opened as routine);
  • a regression removing the label;
  • a draft not opened;
  • too few Accept reviews; valid and invalid reviewer pairs;
  • reviews of another release not counting;
  • a review field pointing elsewhere;
  • draft → accepted refused;
  • a non-status edit left alone;
  • the tracking issue opened once.
  • CI: the test runs on every pull request (added to check.yml).
  • Site: builds with --strict.

Labels review (which the review form already applied, but which did not exist) and review tracking have been created.

Deciding this. It amends a procedure, so it is a substantive change under the Governance page. The automation only acts once this is merged, because pull_request_target runs from main.

See the exact changes

comment without GitHub · comment on GitHub

Adopting principles: public comment, default adoption, member vote

Substantive change, open for decision since 4 November 2026. Substantive change: adopted after a comment period of at least 14 days, once every Operations Committee member approves and a majority of the voting members who vote approve.

Status: Waiting for comment period: 2 days left; approval from William Mandrick, Barry Smith.

Voting member Vote How
David R. Koepsell (committee) Approve on website
William Mandrick (committee) not yet voted
Barry Smith (committee) not yet voted
Read the full proposal

Adds a procedure for adopting new principles and changes to principles, decided by the Foundry's members after public comment.

The procedure (new page About > Adopting principles): 1. A coordinator labels a pull request that changes a principle decision: principle. 2. 14 days of public comment. Anyone can comment, on GitHub or through the website. Non-members' comments are advisory. 3. Adopted by default if no member objects in that time. 4. If a member objects, members vote for 14 more days. The principle is adopted if more approve than object, with at least 5 votes cast (abstentions count toward the quorum). Otherwise the bot closes the pull request with the result recorded.

Committee members vote as members, one vote each.

What changes in the voting system: - community/decision-rules.json gains a principle kind with electorate: members. Routine and substantive changes are unchanged: still decided by the voting members. - Ballots, sign-in, and notices reach all members, not only voting members. Members' addresses come from the voters tab, plus members recorded by the digest in the subscribers tab, that is, people who joined through the website. On other kinds of proposal, a non-voting member who signs in is told they can comment but not vote. - The site shows a principle's phase (public comment, member vote, or closed), and objections so far. The weekly digest reminds every member who can vote. - A failed vote is recorded and the pull request is closed.

Tested: - Same count: the website and the bot agree on 1,500 random cases across all three kinds. The original 400 routine and substantive cases are unchanged. - Principle flow, 12 of 12 checks against simulated GitHub and Google services: - default adoption; - ballots to all five members; - a non-voting member voting on a principle, but not on a substantive change; - an objection leading to a member vote; - adoption 4 to 1 with quorum; - failure without quorum, closed with notices. - Earlier suites all still pass, and sign-in and voting were rechecked in a browser. - The site builds with --strict.

Deciding this proposal. It changes how decisions are made, so it is itself a substantive change: label it decision: substantive. It needs the 14-day period and approval by all three committee members.

After it is adopted: - Paste the updated Voting.gs and Digest.gs into Apps Script and publish a new version. - The website changes deploy on merge. The new rule takes effect once decision-rules.json is on main. - The label decision: principle already exists.

See the exact changes

comment without GitHub · comment on GitHub

Governance: voting members and website ballots

Substantive change, open for decision since 19 October 2026. Substantive change: adopted after a comment period of at least 14 days, once every Operations Committee member approves and a majority of the voting members who vote approve.

Status: Waiting for approval from Barry Smith.

Voting member Vote How
David R. Koepsell (committee) Approve on website
William Mandrick (committee) Approve on website
Barry Smith (committee) not yet voted
Read the full proposal

Amends the Governance page so that: - voting members are the Operations Committee and the working group leads; - routine changes need one committee approval and no objection from a voting member; - substantive changes keep the 14-day period and all committee approvals, and also need a majority of the voting members who vote; - votes may be cast either on GitHub or with the personal ballot link emailed through the website, with equal weight.

This is a substantive change: it needs approval from all three founding coordinators, after a 14-day comment period announced in Discussions.

See the exact changes

comment without GitHub · comment on GitHub

Pattern review 0.2.0: adopt proposals from #6 and #8 to #12

Substantive change, open for decision since 19 October 2026. Substantive change: adopted after a comment period of at least 14 days, once every Operations Committee member approves and a majority of the voting members who vote approve.

Status: Waiting for approval from Barry Smith.

Voting member Vote How
David R. Koepsell (committee) Approve on website
William Mandrick (committee) Approve on website
Barry Smith (committee) not yet voted
Read the full proposal

Adopts Bill's proposals in #6, #8, #9, #10, #11, and #12 on the site pages, as discussed in those issues.

Pages changed - ODP-001 to 004 pages updated to 0.2.0: new terms, resolved open questions marked as resolved. - New: Alignment (the D-Acts and CCO bridges) and Changes in 0.2.0 (each proposal, what changed, what was not adopted and why, and four questions for the next meeting). - Related initiatives: adds D-Acts, FOLIO, and SALI.

Main changes to the patterns - ODP-001: performed by means of and evidences (documents do not act); new kinds of legal content: legal norm, legal definition, legal designator. - ODP-002: role modification act; a legal role has exactly one bearer. - ODP-003: legal positions are BFO roles (resolves the Option A / Option B question), aligned with the D-Acts deontic roles; new succeeds for assignment and transfer of office. - ODP-004: legal basis for, derives authority from, delegates legal authority to, all inferred from the chain of authority (8 U.S.C. 1103(a)(3) to 8 C.F.R. 103.1 example).

The OWL files and both bridges were tested with HermiT, separately and together (consistent, coherent, OWL 2 DL). They are not in this PR yet; where they should live in the org is one of the questions for the meeting.

Nothing here is final until Bill and Barry have commented.

See the exact changes

comment without GitHub · comment on GitHub

Proposed changes (10)

A proposed change is a set of edits awaiting review. Drafts are still being prepared; the others are ready for decision.

Review procedure: who decides each step, and an automated status check

Proposed by dkoepsell on 4 November 2026 · ready for review · comment without GitHub · comment on GitHub

Clarifies how registry entries move past draft, and automates the checking. The clarification (docs/sop/review.md) Step Requirements Decision --------- draft → candidate the entry names the release (release: {version, tag}); the repository is public; the tag exists; lof-{id}.owl is in the tagged release; automated checks pass on the tagged commit routine candidate → accepted the same, for the release reviewed; two Ontology review issues of that release recommend Accept, from two reviewers, one...

Full description

Clarifies how registry entries move past draft, and automates the checking.

The clarification (docs/sop/review.md)

Step Requirements Decision
draft → candidate the entry names the release (release: {version, tag}); the repository is public; the tag exists; lof-{id}.owl is in the tagged release; automated checks pass on the tagged commit routine
candidate → accepted the same, for the release reviewed; two Ontology review issues of that release recommend Accept, from two reviewers, one a committee member and the other not a maintainer; the entry records review: {date, issue} substantive
any → deprecated the successor, if any, named in replaced_by substantive

Also stated: - While the Editorial Working Group has no lead, the Operations Committee assigns reviewers. - Reviews are due within 30 days. - Reviews of an earlier release do not count. - No other status change is allowed (for example, draft straight to accepted).

The metadata policy and the registry schema gain the optional release field.

The automation (.github/workflows/registry-status.yml, scripts/registry_status.py)

  • On a pull request that changes an ontology's status: it checks the requirements for that step and reports them as a checklist in one comment that updates itself.
  • When all are met, it adds decision: routine or decision: substantive, which opens the vote: ballots, the website, and the bot's adoption, as for any decision.
  • If a requirement stops being met, it removes the label. Draft pull requests are never opened.
  • On a review issue being filed, edited, or closed: it rechecks the acceptance pull request for that ontology.
  • On a push to main that makes an ontology a candidate: it opens a review tracking issue, which asks the committee to assign two reviewers, links the review form, and sets the 30-day due date.

It uses pull_request_target, so it runs from main and reads the pull request's files through the API; a proposal's own code is never run.

Tested

  • 14 checks against a simulated GitHub:
  • missing release, failing checks, then passing (opened as routine);
  • a regression removing the label;
  • a draft not opened;
  • too few Accept reviews; valid and invalid reviewer pairs;
  • reviews of another release not counting;
  • a review field pointing elsewhere;
  • draft → accepted refused;
  • a non-status edit left alone;
  • the tracking issue opened once.
  • CI: the test runs on every pull request (added to check.yml).
  • Site: builds with --strict.

Labels review (which the review form already applied, but which did not exist) and review tracking have been created.

Deciding this. It amends a procedure, so it is a substantive change under the Governance page. The automation only acts once this is merged, because pull_request_target runs from main.

Adopt principles: stewardship

Proposed by dkoepsell on 4 November 2026 · draft, still being prepared · comment without GitHub · comment on GitHub

Proposes adopting 4 of the Foundry's founding principles: how ontologies are versioned, documented, maintained, and shown to be used. Principle Title Summary --------- LOF-P-006 Versioning Each release carries a version IRI and version number, and every earlier release remains resolvable. LOF-P-013 Documentation The ontology has a...

Full description

Proposes adopting 4 of the Foundry's founding principles: how ontologies are versioned, documented, maintained, and shown to be used.

Principle Title Summary
LOF-P-006 Versioning Each release carries a version IRI and version number, and every earlier release remains resolvable.
LOF-P-013 Documentation The ontology has a public repository with documentation and an issue tracker.
LOF-P-014 Responsiveness and maintenance Maintainers respond to issues and keep the ontology aligned with its dependencies.
LOF-P-016 Documented use The ontology has documented users, or a credible plan for use.

What adopting means. The principle pages are kept as they stand; their Draft for ratification notice becomes Adopted. To propose changing the wording, comment here, or open a separate proposal.

How this is decided. This follows the procedure proposed in #52 (Adopting principles), itself now under decision: 1. Once #52 is adopted, this proposal is labeled decision: principle, which opens 14 days of public comment. 2. It is adopted by default unless a member objects. 3. If a member objects, members vote for 14 more days: adopted if more approve than object, with at least 5 votes cast.

Until then, comments are welcome from anyone, here on GitHub or through the website's comment without GitHub links, which also appear on each principle's page.

The founding principles are proposed in six groups, so that an objection to one group does not hold up the others.

Adopt principles: legal specificity

Proposed by dkoepsell on 4 November 2026 · draft, still being prepared · comment without GitHub · comment on GitHub

Proposes adopting 2 of the Foundry's founding principles: the two principles particular to law: stating the jurisdiction a term belongs to, and recording the source of legal authority. Principle Title Summary --------- LOF-P-010 Jurisdictional transparency Classes specific to a legal system or jurisdiction are marked as such and are not presented as universal....

Full description

Proposes adopting 2 of the Foundry's founding principles: the two principles particular to law: stating the jurisdiction a term belongs to, and recording the source of legal authority.

Principle Title Summary
LOF-P-010 Jurisdictional transparency Classes specific to a legal system or jurisdiction are marked as such and are not presented as universal.
LOF-P-017 Locus of authority Each ontology has a single named point of contact responsible for it.

What adopting means. The principle pages are kept as they stand; their Draft for ratification notice becomes Adopted. To propose changing the wording, comment here, or open a separate proposal.

How this is decided. This follows the procedure proposed in #52 (Adopting principles), itself now under decision: 1. Once #52 is adopted, this proposal is labeled decision: principle, which opens 14 days of public comment. 2. It is adopted by default unless a member objects. 3. If a member objects, members vote for 14 more days: adopted if more approve than object, with at least 5 votes cast.

Until then, comments are welcome from anyone, here on GitHub or through the website's comment without GitHub links, which also appear on each principle's page.

The founding principles are proposed in six groups, so that an objection to one group does not hold up the others.

Adopt principles: scope and coordination

Proposed by dkoepsell on 4 November 2026 · draft, still being prepared · comment without GitHub · comment on GitHub

Proposes adopting 3 of the Foundry's founding principles: how each ontology states its scope, how overlap between ontologies is avoided, and how they align with existing legal standards. Principle Title Summary --------- LOF-P-008 Clearly delineated scope The ontology has a stated scope and a set of competency questions. LOF-P-009...

Full description

Proposes adopting 3 of the Foundry's founding principles: how each ontology states its scope, how overlap between ontologies is avoided, and how they align with existing legal standards.

Principle Title Summary
LOF-P-008 Clearly delineated scope The ontology has a stated scope and a set of competency questions.
LOF-P-009 Orthogonality Member ontologies do not duplicate one another's content.
LOF-P-015 Alignment with existing legal standards The ontology documents its relationship to established legal information standards.

What adopting means. The principle pages are kept as they stand; their Draft for ratification notice becomes Adopted. To propose changing the wording, comment here, or open a separate proposal.

How this is decided. This follows the procedure proposed in #52 (Adopting principles), itself now under decision: 1. Once #52 is adopted, this proposal is labeled decision: principle, which opens 14 days of public comment. 2. It is adopted by default unless a member objects. 3. If a member objects, members vote for 14 more days: adopted if more approve than object, with at least 5 votes cast.

Until then, comments are welcome from anyone, here on GitHub or through the website's comment without GitHub links, which also appear on each principle's page.

The founding principles are proposed in six groups, so that an objection to one group does not hold up the others.

Adopt principles: terms: identifiers, definitions, and names

Proposed by dkoepsell on 4 November 2026 · draft, still being prepared · comment without GitHub · comment on GitHub

Proposes adopting 3 of the Foundry's founding principles: how terms are identified, defined, and named. Principle Title Summary --------- LOF-P-005 Persistent identifiers Every entity has a persistent IRI in the https://w3id.org/lof/ namespace, and IRIs are never reassigned or deleted. LOF-P-007 Textual definitions Every class and...

Full description

Proposes adopting 3 of the Foundry's founding principles: how terms are identified, defined, and named.

Principle Title Summary
LOF-P-005 Persistent identifiers Every entity has a persistent IRI in the https://w3id.org/lof/ namespace, and IRIs are never reassigned or deleted.
LOF-P-007 Textual definitions Every class and relation has a precise, non-circular textual definition.
LOF-P-012 Naming conventions Labels are unique, singular, and follow a consistent lexical style.

What adopting means. The principle pages are kept as they stand; their Draft for ratification notice becomes Adopted. To propose changing the wording, comment here, or open a separate proposal.

How this is decided. This follows the procedure proposed in #52 (Adopting principles), itself now under decision: 1. Once #52 is adopted, this proposal is labeled decision: principle, which opens 14 days of public comment. 2. It is adopted by default unless a member objects. 3. If a member objects, members vote for 14 more days: adopted if more approve than object, with at least 5 votes cast.

Until then, comments are welcome from anyone, here on GitHub or through the website's comment without GitHub links, which also appear on each principle's page.

The founding principles are proposed in six groups, so that an objection to one group does not hold up the others.

Adopt principles: grounding and logic

Proposed by dkoepsell on 4 November 2026 · draft, still being prepared · comment without GitHub · comment on GitHub

Proposes adopting 3 of the Foundry's founding principles: how every ontology is grounded in BFO, how legal classes state their BFO category, and the logical checks every release passes. Principle Title Summary --------- LOF-P-003 Common upper ontology The ontology is aligned to Basic Formal Ontology, and every class descends from a BFO class....

Full description

Proposes adopting 3 of the Foundry's founding principles: how every ontology is grounded in BFO, how legal classes state their BFO category, and the logical checks every release passes.

Principle Title Summary
LOF-P-003 Common upper ontology The ontology is aligned to Basic Formal Ontology, and every class descends from a BFO class.
LOF-P-004 Explicit grounding of legal entities Each top-level legal class states which BFO category it falls under, and why.
LOF-P-011 Logical well-formedness Each release is logically consistent and contains no unsatisfiable classes.

What adopting means. The principle pages are kept as they stand; their Draft for ratification notice becomes Adopted. To propose changing the wording, comment here, or open a separate proposal.

How this is decided. This follows the procedure proposed in #52 (Adopting principles), itself now under decision: 1. Once #52 is adopted, this proposal is labeled decision: principle, which opens 14 days of public comment. 2. It is adopted by default unless a member objects. 3. If a member objects, members vote for 14 more days: adopted if more approve than object, with at least 5 votes cast.

Until then, comments are welcome from anyone, here on GitHub or through the website's comment without GitHub links, which also appear on each principle's page.

The founding principles are proposed in six groups, so that an objection to one group does not hold up the others.

Adopt principles: openness and credit

Proposed by dkoepsell on 4 November 2026 · draft, still being prepared · comment without GitHub · comment on GitHub

Proposes adopting 3 of the Foundry's founding principles: who may use Foundry ontologies, in what format they are released, and how contributors are credited. Principle Title Summary --------- LOF-P-001 Open The ontology is openly available and may be used by all without constraint other than attribution. LOF-P-002 Common format The ontology is...

Full description

Proposes adopting 3 of the Foundry's founding principles: who may use Foundry ontologies, in what format they are released, and how contributors are credited.

Principle Title Summary
LOF-P-001 Open The ontology is openly available and may be used by all without constraint other than attribution.
LOF-P-002 Common format The ontology is released in OWL 2, with a stable primary release file.
LOF-P-018 Attribution Contributors to the ontology are credited, and it can be cited.

What adopting means. The principle pages are kept as they stand; their Draft for ratification notice becomes Adopted. To propose changing the wording, comment here, or open a separate proposal.

How this is decided. This follows the procedure proposed in #52 (Adopting principles), itself now under decision: 1. Once #52 is adopted, this proposal is labeled decision: principle, which opens 14 days of public comment. 2. It is adopted by default unless a member objects. 3. If a member objects, members vote for 14 more days: adopted if more approve than object, with at least 5 votes cast.

Until then, comments are welcome from anyone, here on GitHub or through the website's comment without GitHub links, which also appear on each principle's page.

The founding principles are proposed in six groups, so that an objection to one group does not hold up the others.

Adopting principles: public comment, default adoption, member vote

Proposed by dkoepsell on 4 November 2026 · ready for review · comment without GitHub · comment on GitHub

Adds a procedure for adopting new principles and changes to principles, decided by the Foundry's members after public comment. The procedure (new page About Adopting principles): 1. A coordinator labels a pull request that changes a principle decision: principle. 2. 14 days of public comment. Anyone can comment, on GitHub or through the website. Non-members' comments are advisory. 3. Adopted by default if no member objects in that time. 4. If a member objects, members vote for 14 more days. The...

Full description

Adds a procedure for adopting new principles and changes to principles, decided by the Foundry's members after public comment.

The procedure (new page About > Adopting principles): 1. A coordinator labels a pull request that changes a principle decision: principle. 2. 14 days of public comment. Anyone can comment, on GitHub or through the website. Non-members' comments are advisory. 3. Adopted by default if no member objects in that time. 4. If a member objects, members vote for 14 more days. The principle is adopted if more approve than object, with at least 5 votes cast (abstentions count toward the quorum). Otherwise the bot closes the pull request with the result recorded.

Committee members vote as members, one vote each.

What changes in the voting system: - community/decision-rules.json gains a principle kind with electorate: members. Routine and substantive changes are unchanged: still decided by the voting members. - Ballots, sign-in, and notices reach all members, not only voting members. Members' addresses come from the voters tab, plus members recorded by the digest in the subscribers tab, that is, people who joined through the website. On other kinds of proposal, a non-voting member who signs in is told they can comment but not vote. - The site shows a principle's phase (public comment, member vote, or closed), and objections so far. The weekly digest reminds every member who can vote. - A failed vote is recorded and the pull request is closed.

Tested: - Same count: the website and the bot agree on 1,500 random cases across all three kinds. The original 400 routine and substantive cases are unchanged. - Principle flow, 12 of 12 checks against simulated GitHub and Google services: - default adoption; - ballots to all five members; - a non-voting member voting on a principle, but not on a substantive change; - an objection leading to a member vote; - adoption 4 to 1 with quorum; - failure without quorum, closed with notices. - Earlier suites all still pass, and sign-in and voting were rechecked in a browser. - The site builds with --strict.

Deciding this proposal. It changes how decisions are made, so it is itself a substantive change: label it decision: substantive. It needs the 14-day period and approval by all three committee members.

After it is adopted: - Paste the updated Voting.gs and Digest.gs into Apps Script and publish a new version. - The website changes deploy on merge. The new rule takes effect once decision-rules.json is on main. - The label decision: principle already exists.

Governance: voting members and website ballots

Proposed by dkoepsell on 19 October 2026 · ready for review · comment without GitHub · comment on GitHub

Amends the Governance page so that: - voting members are the Operations Committee and the working group leads; - routine changes need one committee approval and no objection from a voting member; - substantive changes keep the 14-day period and all committee approvals, and also need a majority of the voting members who vote; - votes may be cast either on GitHub or with the personal ballot link emailed through the website, with equal weight. This is a substantive change: it needs approval from...

Full description

Amends the Governance page so that: - voting members are the Operations Committee and the working group leads; - routine changes need one committee approval and no objection from a voting member; - substantive changes keep the 14-day period and all committee approvals, and also need a majority of the voting members who vote; - votes may be cast either on GitHub or with the personal ballot link emailed through the website, with equal weight.

This is a substantive change: it needs approval from all three founding coordinators, after a 14-day comment period announced in Discussions.

Pattern review 0.2.0: adopt proposals from #6 and #8 to #12

Proposed by dkoepsell on 16 October 2026 · ready for review · comment without GitHub · comment on GitHub

Adopts Bill's proposals in #6, #8, #9, #10, #11, and #12 on the site pages, as discussed in those issues. Pages changed - ODP-001 to 004 pages updated to 0.2.0: new terms, resolved open questions marked as resolved. - New: Alignment (the D-Acts and CCO bridges) and Changes in 0.2.0 (each proposal, what changed, what was not adopted and why, and four questions for the next meeting). - Related initiatives: adds D-Acts, FOLIO, and SALI. Main changes to the patterns - ODP-001: performed by means of...

Full description

Adopts Bill's proposals in #6, #8, #9, #10, #11, and #12 on the site pages, as discussed in those issues.

Pages changed - ODP-001 to 004 pages updated to 0.2.0: new terms, resolved open questions marked as resolved. - New: Alignment (the D-Acts and CCO bridges) and Changes in 0.2.0 (each proposal, what changed, what was not adopted and why, and four questions for the next meeting). - Related initiatives: adds D-Acts, FOLIO, and SALI.

Main changes to the patterns - ODP-001: performed by means of and evidences (documents do not act); new kinds of legal content: legal norm, legal definition, legal designator. - ODP-002: role modification act; a legal role has exactly one bearer. - ODP-003: legal positions are BFO roles (resolves the Option A / Option B question), aligned with the D-Acts deontic roles; new succeeds for assignment and transfer of office. - ODP-004: legal basis for, derives authority from, delegates legal authority to, all inferred from the chain of authority (8 U.S.C. 1103(a)(3) to 8 C.F.R. 103.1 example).

The OWL files and both bridges were tested with HermiT, separately and together (consistent, coherent, OWL 2 DL). They are not in this PR yet; where they should live in the org is one of the questions for the meeting.

Nothing here is final until Bill and Barry have commented.

Open questions and tasks (20)

ODP-002: uptake as a part of the act (Reinach's Vernehmung)

Raised by dkoepsell on 22 October 2026 · 0 comments · question · comment without GitHub · comment on GitHub

Raised at the 21 October meeting (Bill, on Reinach's promise; see the layer schema slides, version 2, slide 6). The point. For Reinach, a promise is a social act that must be taken up (vernommen) by the person it is addressed to. Without uptake, no claim and no obligation arise. Bill's diagram of a promise between John and Sue shows: - the act of promising, in which both take part; - the promisor and promisee roles; - a claim borne by Sue and an obligation borne by John, each dependent on the...

Full text

Raised at the 21 October meeting (Bill, on Reinach's promise; see the layer schema slides, version 2, slide 6).

The point. For Reinach, a promise is a social act that must be taken up (vernommen) by the person it is addressed to. Without uptake, no claim and no obligation arise. Bill's diagram of a promise between John and Sue shows:

  • the act of promising, in which both take part;
  • the promisor and promisee roles;
  • a claim borne by Sue and an obligation borne by John, each dependent on the other.

This matches the 0.2.0 patterns closely:

In the diagram In the patterns
claim and obligation, each dependent on the other ODP-003 correlative positions (Option A: positions inhere in their holders)
promisor and promisee roles ODP-002 roles brought about by the act
the words on a napkin ODP-001 performed by means of, or evidences (#8)

What the patterns lack is uptake.

Proposal for discussion. Uptake is not normative reasoning. It is the addressee's grasping of the act: in Mental Functioning Ontology terms, a cognitive process. So it belongs in ODP-002 and ODP-001, not in the Normative Reasoning Ontology:

  • an uptake is a process in which the addressee of a social act apprehends it;
  • a social act addressed to someone (a promise, a command, a notice) has part some uptake by that addressee, or is completed only when one occurs;
  • the claim and obligation (ODP-003) arise from the act with its uptake.

Questions.

  1. Is uptake a part of the act, or a condition of its legal effect? The latter fits acts that take effect on receipt, such as notices, or on publication.
  2. Which acts need uptake? Promising and commanding do. Enactment, publication, and a will plausibly do not.
  3. Software parties. When contracts are formed or accepted by software, uptake cannot require a brain. This is the same issue as #40: uptake should be a process with a cognitive kind, not only a mental one.

NRO and NAO: reasoning before a breach, accountability after it

Raised by dkoepsell on 21 October 2026 · 1 comment · question · comment without GitHub · comment on GitHub

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology. The NRO models the forward-looking half of normative life: an agent working out what ought to be done. NAO models the backward-looking half: a norm is breached, the breach leaves something behind, and someone answers for it....

Full text

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology.

The NRO models the forward-looking half of normative life: an agent working out what ought to be done. NAO models the backward-looking half: a norm is breached, the breach leaves something behind, and someone answers for it. Proposed links:

  • Capacity. NAO's accountable agent bears some normative reasoning capacity: a disposition realized in normative reasoning processes. This gives NAO's standing and counterfactual-control conditions a cognitive basis. Tested. A draft bridge between the NRO, NAO, LOF-ODP-001 to 004, the Common Core Ontologies, and the Mental Functioning Ontology was checked with HermiT; results are quoted below. The files will be posted once the Foundry settles where bridge files live. In NAO's Reset Button example, the human (Reyes) and the permanent AI (UNITY) are both classified as normative reasoners, and the resettable systems are not. This is coherent only with the change proposed in the issue on MFO mental processes.
  • Judgment and breach. An NRO normative judgment that some conduct breached a norm is about a NAO normative breach. Accountability responses (apology, restitution, sanction) follow such a judgment, but the breach does not depend on it.
  • Interpretation. The NRO's norm interpretation process answers the question left open in #31: how to record that someone interpreted a provision as expressing a given norm. The interpretation is a process, with the interpreter as participant, the provision as input, and the norm as output.
  • Conflict. The NRO's norm conflict resolution process is where a typology of normative contradictions fits, for example the one in A Structural Ontology of the Law. The NRO names the process; the typology classifies the conflicts it resolves. In the bridge it requires at least two norms as input.

New ontology request: Normative Accountability Ontology (NAO)

Raised by dkoepsell on 21 October 2026 · 2 comments · enhancement registry · comment without GitHub · comment on GitHub

New ontology request, following the new ontology request SOP. Declaration of interest: NAO is David Koepsell's ontology, and the request comes from him as a founding coordinator. Field Value ------ Title Normative Accountability Ontology (NAO) Requested prefix NAO Jurisdictions general Status requested draft License CC BY 4.0 Contact David R. Koepsell (drkoepsell@tamu.edu, @dkoepsell) Scope What it is to be held to account:...

Full text

New ontology request, following the new ontology request SOP. Declaration of interest: NAO is David Koepsell's ontology, and the request comes from him as a founding coordinator.

Field Value
Title Normative Accountability Ontology (NAO)
Requested prefix NAO
Jurisdictions general
Status requested draft
License CC BY 4.0
Contact David R. Koepsell (drkoepsell@tamu.edu, @dkoepsell)

Scope

What it is to be held to account:

  • a norm is breached, and the breach is someone's conduct;
  • the breach leaves something behind in those it connects;
  • someone answers for it, through apology, restitution, sanction, or other responses.

NAO covers legal, moral, professional, and institutional norms. One distinction is central: between what can be reset and what carries its own record. That distinction is what lets NAO say who answers when an AI system or other artifact does wrong.

What exists today (version 0.1.0, draft)

NAO is complete and tested, not a plan. It has five patterns:

Pattern Covers
P1 norm, breach, attribution
P2 residue and record
P3 capacities and answerability
P4 responses
P5 the Irreversible Accountability Test

It is published as two modules:

  • NAO core: 23 classes and 12 relations, with no commitments beyond BFO 2020 and LOF-ODP-001 to 004.
  • NAO-IAT: the Irreversible Accountability Test as two exclusion axioms, kept apart so the vocabulary can be adopted without the thesis.

It comes with:

  • A worked example of three scenarios from the classroom demonstration The Reset Button: a physician, a resettable diagnostic system, a permanent AI, and an autonomous probe. A plain-language report states who answers for what. It also finds an accountability gap: a breach no one answers for.
  • Checks: HermiT consistency, SPARQL quality checks, a ROBOT report, SHACL shapes with readable messages, and three thesis tests. Each test is inconsistent with the IAT module and consistent without it. All of these run in CI.

Competency questions

  1. Which norm did this conduct breach, whose conduct was it, and whom did it wrong?
  2. Is this breach a repeat of an earlier one?
  3. What does a breach leave behind, and in whom? What records exist about it, and can they be changed?
  4. Who answers for this breach, and through which responses?
  5. Is answerability for an artifact's breach displaced to whoever deploys it? Which breaches does no one answer for?
  6. Which agents satisfy the five conditions of the IAT, and which conditions do the others fail?
  7. Is a moral answerability codified as a legal position, and does the breach trigger it (LOF-ODP-003, ODP-004)?

Fit with the Foundry

  • Layer. NAO belongs in the general normative layer proposed in #38, beside the Normative Reasoning Ontology. It covers what happens after a breach; the NRO covers reasoning before one (#43).
  • Norm. NAO's norm is the broad class proposed as the hub in #39. LOF-ODP-001 legal norm sits under it, and both sit under the CCO prescriptive information content entity.
  • Legal patterns. NAO imports LOF-ODP-001 to 004. A breach that triggers a codified legal position is inferred to be an ODP-004 triggering occurrence.
  • Tested with the NRO. In the NRO and NAO bridge (#40, #43), NAO's thesis module and worked example stay consistent. The human and the permanent AI are both classified as normative reasoners; the resettable systems are not.

Overlap check (LOF-P-009)

Topic Who covers it
Legal remedies ODP-004 remedial position. NAO does not redefine it.
NAO responses (apology, restitution, sanction) NAO: processes that realize answerability. They may realize a remedial position but are not one.
Legal positions ODP-003. NAO's codified as points into ODP-003 rather than duplicating it.
Breach by omission ODP-004 undischarged during. NAO's open question 4 proposes adopting it.

Decisions for the Foundry

  1. Register at draft status. This is a routine change under the Governance page, once the scope discussion here is settled. The registry entry is ready in the linked pull request.
  2. Repository. Host NAO in the organization as lof-nao (the SOP's default), or link the author's own repository.
  3. Identifiers (LOF-P-005). NAO's IRIs are currently https://sool.davidkoepsell.com/nao/NAO_nnnnnnn. P-005 requires https://w3id.org/lof/NAO_nnnnnnn. NAO is a draft no one else yet depends on, so migrating now is cheap: new IRIs in 0.2.0, with each old IRI deprecated and pointing to its replacement.
  4. The thesis module. Should NAO-IAT be registered with the core, as an optional module, or kept outside the Foundry as the author's own theory? Keeping it apart is NAO's own design, so either works.
  5. Open questions to settle in review. These come from NAO's specification:
  6. the BFO category of moral residue;
  7. the category of normative standing;
  8. a weaker, time-indexed version of thesis 2;
  9. breach by omission;
  10. whether accountability gaps should be validation errors.

Layer schema: a general normative layer between the reference layer and the Legal Core

Raised by dkoepsell on 21 October 2026 · 3 comments · question · comment without GitHub · comment on GitHub

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology. The point. The schema places the Normative Reasoning Ontology in the domain layer, beneath the Legal Core Ontology. But the NRO covers moral, military, organizational, and regulatory reasoning, not only legal reasoning. NAO likewise...

Full text

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology.

The point. The schema places the Normative Reasoning Ontology in the domain layer, beneath the Legal Core Ontology. But the NRO covers moral, military, organizational, and regulatory reasoning, not only legal reasoning. NAO likewise covers moral and professional norms as well as legal ones. Neither is a specialization of a legal core.

Proposal. Add a general normative layer between the reference layer and the Legal Core, holding what is common to all norms:

  • norms;
  • breaches;
  • normative reasoning (NRO);
  • accountability (NAO).

The Legal Core would then specialize it: a legal norm is a kind of norm, and legal reasoning is a kind of normative reasoning. This is how LOF-ODP-001 and NAO already relate: NAO's norm includes legal content, and moral or institutional norms that are not legal content.

A small correction to the diagram. The Document Acts ontology is drawn as a module of the Common Core Ontologies. Its terms are in the IAO namespace (IAO_00210xx, as noted in #10 and #11), and it is released separately. It may be clearer to show it as a separate import alongside the CCO.

NRO: normative reasoning as an MFO mental process excludes decision-support systems

Raised by dkoepsell on 21 October 2026 · 0 comments · question · comment without GitHub · comment on GitHub

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology. The problem. The slides name autonomous decision-support systems as a main use of the NRO. But they place normative reasoning process under MFO thinking with language, and so under mental process. The Mental Functioning Ontology...

Full text

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology.

The problem. The slides name autonomous decision-support systems as a main use of the NRO. But they place normative reasoning process under MFO thinking with language, and so under mental process. The Mental Functioning Ontology defines a mental process as "a bodily process that occurs in the brain". Every bodily process has some organism as participant at all times (OGMS_0000060: has participant at all times some extended organism). As proposed, an artifact cannot engage in normative reasoning at all.

Tested. A draft bridge between the NRO, NAO, LOF-ODP-001 to 004, the Common Core Ontologies, and the Mental Functioning Ontology was checked with HermiT; results are quoted below. The files will be posted once the Foundry settles where bridge files live.

Case NRO as proposed NRO revised
A decision-support system (NAO's DiagnoBot) evaluates whether an action is permitted inconsistent consistent: a normative reasoning process, not a mental one
A judge reasons, thinking with language consistent consistent: inferred to be a mental normative reasoning process

Proposal.

  • Make normative reasoning process a BFO process.
  • Define mental normative reasoning process as normative reasoning process and MFO mental process.
  • Add computational normative reasoning process for artifacts.

NAO takes the same line: what makes an entity answerable is a disposition, not a substrate.

Also noted. The MFO uses BFO 2 temporalized relations: has participant at all times (BFO_0000167) and inheres in at all times (BFO_0000052). A stack built on BFO 2020 needs these mapped to BFO_0000057 and BFO_0000197. The bridge does this.

One class for "norm" across the stack

Raised by dkoepsell on 21 October 2026 · 1 comment · question · comment without GitHub · comment on GitHub

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology. The problem. Once the schema's layers are combined, there are five near-duplicate classes for a norm: Ontology Class Placement --------- CCO Prescriptive Information Content Entity (ont00000965) information content entity NRO...

Full text

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology.

The problem. Once the schema's layers are combined, there are five near-duplicate classes for a norm:

Ontology Class Placement
CCO Prescriptive Information Content Entity (ont00000965) information content entity
NRO (slides) stated / documented norm information content entity
NAO norm (NAO_0001022) generically dependent continuant
LOF-ODP-001 legal norm (ODP001_0001005) legal content
D-Acts deontic classes information content entity

Proposal. Make the CCO prescriptive information content entity the hub, and place the others relative to it:

  • CCO prescriptive information content entity, the hub;
  • NAO norm, the broad class: anything that prescribes, permits, or prohibits conduct, including unwritten moral and customary norms;
    • LOF-ODP-001 legal norm, already placed under CCO process regulation by the LOF and CCO bridge;
  • NRO stated norm, a defined class: a norm that some material entity, such as a document, is a carrier of.

Tested. A draft bridge between the NRO, NAO, LOF-ODP-001 to 004, the Common Core Ontologies, and the Mental Functioning Ontology was checked with HermiT; results are quoted below. The files will be posted once the Foundry settles where bridge files live. With these axioms, a regulatory provision carried by an Official Gazette is inferred to be a stated norm, a NAO norm, a legal norm, and a CCO prescriptive ICE. The whole is consistent.

Open question for the CCO community. Is an unwritten norm, held only in minds and practices, a CCO information content entity? NAO assumes so; the CCO treatment of concretization may not.

NRO: one classification axis, defined domain classes, and relations

Raised by dkoepsell on 21 October 2026 · 0 comments · enhancement · comment without GitHub · comment on GitHub

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology. 1. Two axes of classification. Slide 3 divides normative reasoning by domain: legal, moral, regulatory, organizational. Slide 5 adds a second division, by task: applicability, interpretation, obligation, permission and prohibition...

Full text

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology.

1. Two axes of classification. Slide 3 divides normative reasoning by domain: legal, moral, regulatory, organizational. Slide 5 adds a second division, by task: applicability, interpretation, obligation, permission and prohibition assessment, conflict resolution, justification, judgment. Asserting both gives multiple inheritance, which the BFO and OBO conventions avoid. Regulatory reasoning is also defined as "a legal or organizational normative reasoning process", overlapping its own siblings.

Proposal.

  • Assert the task axis.
  • Define the domain classes by the kind of norm that is an input. For example, legal reasoning process is a normative reasoning process that has as input some legal norm.
  • Make regulatory norm a kind of legal norm. Regulatory reasoning is then inferred to be a kind of legal reasoning, with no overlap to assert.

Tested. A draft bridge between the NRO, NAO, LOF-ODP-001 to 004, the Common Core Ontologies, and the Mental Functioning Ontology was checked with HermiT; results are quoted below. The files will be posted once the Foundry settles where bridge files live. The inference holds. An obligation assessment whose input is a regulatory norm is classified as legal reasoning, regulatory reasoning, and obligation assessment.

2. Relations. The slides' "input" is not a BFO relation. Proposal: use has input and has output from the Relation Ontology (RO_0002233, RO_0002234).

3. Judgment. Normative judgment process is defined as "an act of judging", not in the genus-and-difference form of its siblings. Proposal:

  • a normative judgment process is a normative reasoning process whose output is a normative judgment;
  • a normative judgment is CCO descriptive information content;
  • a breach finding is a judgment that is about some normative breach (NAO).

The finding does not create the breach. The breach exists whether or not anyone judges it.

NRO: add an authority assessment process (powers and competence)

Raised by dkoepsell on 21 October 2026 · 0 comments · enhancement question · comment without GitHub · comment on GitHub

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology. The gap. The NRO assesses obligations, permissions, and prohibitions. The most common legal question is often different: was this actor authorized, or competent, to do this? That concerns powers and competence, Hohfeld's...

Full text

Raised by the "LOF Layer Schema" slides (21 October 2026), which propose the Foundry's layer structure and a Normative Reasoning Ontology (NRO), compared with the Normative Accountability Ontology (NAO). Declaration of interest: NAO is David Koepsell's own ontology.

The gap. The NRO assesses obligations, permissions, and prohibitions. The most common legal question is often different: was this actor authorized, or competent, to do this? That concerns powers and competence, Hohfeld's second-order positions, which LOF-ODP-003 includes and on which LOF-ODP-004's chain of authority is built.

Proposal. An authority assessment process: a normative reasoning process that has as input some legal content that grounds a power (ODP-003 grounded in). It is the reasoning counterpart of ODP-004's legal basis for and derives authority from.

Tested. A draft bridge between the NRO, NAO, LOF-ODP-001 to 004, the Common Core Ontologies, and the Mental Functioning Ontology was checked with HermiT; results are quoted below. The files will be posted once the Foundry settles where bridge files live. An assessment whose input is Article 22 of Brazil's General Telecommunications Law, which grounds the Anatel Board's rule-making power, is classified as an authority assessment. The example comes from #35.

A question the test raised. That same assessment is not classified as legal reasoning, because its input is legal content that grounds a power, not a legal norm. Should legal reasoning process be defined by any legal content as input, not only legal norms?

New pattern: legal content in time (publication, entry into force, repeal, versions)

Raised by dkoepsell on 21 October 2026 · 0 comments · enhancement · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 8 and 9. The gap. None of the four patterns says when legal content is in force. Danilo's Anatel Resolution 765/2023 example separates, each with its own date: - the act (the Board's vote, 26 Oct 2023) and the signing (6 Nov 2023); - publication (10 Nov 2023), and two corrective republications, which under LINDB Art....

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 8 and 9.

The gap. None of the four patterns says when legal content is in force. Danilo's Anatel Resolution 765/2023 example separates, each with its own date:

  • the act (the Board's vote, 26 Oct 2023) and the signing (6 Nov 2023);
  • publication (10 Nov 2023), and two corrective republications, which under LINDB Art. 1 § 3 restart the waiting period;
  • entry into force: some articles at once, the rest postponed by a later Board decision to 1 Sep 2025;
  • revocation of the previous regulation on that date, and in Oct 2026 the partial revocation of Arts. 42 to 85 only.

Validity is an interval, opened and closed by other acts, and it can differ article by article.

Proposal. A new pattern (LOF-ODP-005), covering:

  • an in force during relation between legal content and a temporal region, at the level of single provisions;
  • publication as an occurrent distinct from the act that establishes the content;
  • acts whose object is another text's validity: postponement, repeal, revocation, and amendment;
  • point-in-time versions, aligned with Akoma Ntoso and ELI rather than invented.

Questions.

  1. Is publication a legal act, or an occurrent that conditions the effect of one?
  2. Does a repealed provision cease to exist, or continue to exist but stop being in force? This matters for LINDB Art. 6; see the companion issue on vested rights.
  3. How should content that lapses with no terminating act be handled, such as a provisional measure not converted within 60 + 60 days (Constitution Art. 62)?

Distinguish provision (text) from norm, and model the structure of legal texts

Raised by dkoepsell on 21 October 2026 · 0 comments · enhancement · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 3, 4, 10 and 11. The gap. ODP-001 makes legal norm a subclass of legal content, so the wording of an article and the norm it expresses are the same kind of thing. The deck argues, following Humberto Ávila, that text is not norm: one article can contain several norms, and one norm can draw on several articles. The...

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 3, 4, 10 and 11.

The gap. ODP-001 makes legal norm a subclass of legal content, so the wording of an article and the norm it expresses are the same kind of thing. The deck argues, following Humberto Ávila, that text is not norm: one article can contain several norms, and one norm can draw on several articles. The passage from one to the other is interpretation, which is where the hardest questions lie (slide 19).

Proposal.

  • Distinguish a legal provision (a unit of text: an article, a paragraph, an item, a clause) from a legal norm (what it requires, prohibits, or permits). Both are information content. Relate them by an expresses relation that is many-to-many.
  • Model the parthood of text units fixed by statute. In Brazil, Complementary Law 95/1998 Art. 10 sets out the hierarchy: Part > Book > Title > Chapter > Section > Subsection > Article, and within an article, caput > inciso > alínea > item, plus parágrafo. Headings are already covered by ODP-001's legal designator: they group and name, and only articles carry norms.
  • Do not invent the structural vocabulary. Align with Akoma Ntoso, and with LexML Brasil, Brazil's standard for legal identifiers and markup, so that each unit can carry a persistent identifier under LOF-P-005.

Questions.

  1. Is a norm a separate entity, or should the LOF model only provisions and the positions they ground (ODP-003), leaving norms implicit?
  2. How should the LOF record an interpretation, that is, a claim that a provision expresses a given norm, made by someone, at some time?

ODP-001: content that is asserted or requested, not established

Raised by dkoepsell on 21 October 2026 · 0 comments · enhancement question · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 15, 16 and 17. The gap. ODP-001 defines legal content as content established by some legal act. Several kinds of content in legal documents are not established in that sense: - Allegations. A complaint and a defence make contrary claims about the same event. The ontology must keep "the document says that X" apart from...

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 15, 16 and 17.

The gap. ODP-001 defines legal content as content established by some legal act. Several kinds of content in legal documents are not established in that sense:

  • Allegations. A complaint and a defence make contrary claims about the same event. The ontology must keep "the document says that X" apart from "X happened".
  • The report and reasoning of a judgment. Under the Brazilian Code of Civil Procedure, Arts. 489 and 503, only the operative part (dispositivo) has the force of law. The report describes and the reasoning argues.
  • A requested decision. An appeal argues for a decision that was never issued. Bills, draft regulations, and this repository's own proposed changes are the same kind of thing.

Proposal.

  • Add descriptive and argumentative content related to its subject by is about. The CCO bridge in 0.2.0 already imports IAO is about and the CCO descriptive and directive information content classes.
  • Add a class for proposed legal content: directive content that some act asks to be established, but that has not been.

Note. Filings create no norms, but they are still legal acts with legal effects. They generate procedural positions: the court's duty to decide, the defendant's burden to answer, and the presumption that uncontested facts are true (CPC Art. 341). These fit ODP-004 as acts that generate positions.

ODP-003: positions that survive a change in their ground (vested rights, res judicata)

Raised by dkoepsell on 21 October 2026 · 0 comments · question · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slide 9. The question. LINDB Art. 6 says that a new law has immediate and general effect, but respects: - the consummated legal act: one completed under the law in force when it was performed; - the vested right: a right its holder may already exercise; - res judicata: a decision no longer open to appeal. Under ODP-003...

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slide 9.

The question. LINDB Art. 6 says that a new law has immediate and general effect, but respects:

  • the consummated legal act: one completed under the law in force when it was performed;
  • the vested right: a right its holder may already exercise;
  • res judicata: a decision no longer open to appeal.

Under ODP-003 (Option A, adopted in 0.2.0), a legal position is a role grounded in legal content. If that content is repealed, does a vested right lapse? The law says it does not.

Two readings to choose between.

  1. grounded in records grounding at the time the position came into existence, and the position's continued existence does not depend on the content staying in force; or
  2. the position persists because some other content, here LINDB Art. 6 itself, grounds its survival.

A Hohfeldian account. A vested right, and res judicata, can be modeled as an immunity of the holder, correlative to a disability of the legislature or the courts to alter the position. That is exactly what ODP-003's second-order positions are for. It would make a good worked example for them.

Suggested competency question. Which positions held by X remain in force after content C is repealed, and on what ground?

Bring a walk-through of your legal order

Raised by dkoepsell on 21 October 2026 · 1 comment · help wanted worked example · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Danilo's proposal follows one legal order from its constitution down to a single dispute, and asks at every step: what is the act, what norm results, what text expresses it, and what document carries it? Each step isolated something a legal ontology has to represent, and several exposed gaps in the current patterns (see the...

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal.

Danilo's proposal follows one legal order from its constitution down to a single dispute, and asks at every step: what is the act, what norm results, what text expresses it, and what document carries it? Each step isolated something a legal ontology has to represent, and several exposed gaps in the current patterns (see the related issues).

The invitation. Bring the same walk-through from another legal order: a common-law system, another civil-law system, EU law, international law, religious law, or any order you know well. Comparing walk-throughs is the best test that the Foundry's patterns are neutral between legal systems (LOF-P-010).

A suggested shape.

  1. A chain of general norms: for example constitution, then statute, then regulation. For each link, give the act, the text, and where the text is published.
  2. One concrete case: for example contract, non-performance, claim, decision, appeal. It may be invented, as long as it follows the real rules.
  3. At each step: the act, the norm, the text, the carrier, and what the step creates (positions, a decision, nothing).

Post it as a comment here, or through the website's Take part without GitHub form. Slides or a document are fine; a link will do.

ODP-003: multiple grounds, and one claim against several debtors

Raised by dkoepsell on 21 October 2026 · 0 comments · question · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 13 and 20. 1. More than one ground. In the deck's mobile service contract, clause 9 ("the consumer may terminate at any time, without charge") restates Anatel's consumer rights regulation, Art. 4, XVI. The consumer's position is then grounded both in the regulation and in the contract. ODP-003 should say explicitly...

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 13 and 20.

1. More than one ground. In the deck's mobile service contract, clause 9 ("the consumer may terminate at any time, without charge") restates Anatel's consumer rights regulation, Art. 4, XVI. The consumer's position is then grounded both in the regulation and in the contract. ODP-003 should say explicitly that grounded in may have more than one value, and whether the two grounds differ in kind: the regulation grounds the position, while the contract binds it to named parties.

2. Joint and several liability. Under the Consumer Protection Code (Art. 7, sole paragraph, and Art. 18), every supplier in the chain is jointly and severally liable to the consumer. ODP-003 defines correlative of as a symmetric relation between two positions, "each of which exists if and only if the other does". One claim matched by the duties of several suppliers does not fit that one-to-one definition. The options:

  • one claim per debtor, each with its own correlative duty, plus an axiom that discharge by one discharges all; or
  • a correlative relation that is one-to-many.

3. A note on the termination right. The deck pairs "Maria's right to terminate" with "the provider's duty to stop billing". In Hohfeld's terms the right to terminate is a power, correlative to the provider's liability. The duty to stop billing comes into existence only when the power is exercised, through ODP-003 exercised in and ODP-004 generated by. This could be the pattern's worked example of the difference between first-order and second-order positions.

ODP-004: add the Anatel chain as a second, civil-law worked example

Raised by dkoepsell on 21 October 2026 · 0 comments · documentation enhancement · comment without GitHub · comment on GitHub

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 12, 14 and 18. ODP-004's chain of authority is illustrated with a U.S. example: 8 U.S.C. 1103(a)(3) to 8 C.F.R. 103.1. Danilo's deck supplies a real civil-law counterpart, with the act, text, and carrier documented at each link: Link Act Text Carrier ------------ Constitutional Amendment 8/1995 promulgation by the...

Full text

Raised by Danilo Delôgo Tavares's proposal "What a legal ontology must tell apart" (Anatel), a walk-through of the Brazilian legal order from the Constitution to an appeal. Slides 12, 14 and 18.

ODP-004's chain of authority is illustrated with a U.S. example: 8 U.S.C. 1103(a)(3) to 8 C.F.R. 103.1. Danilo's deck supplies a real civil-law counterpart, with the act, text, and carrier documented at each link:

Link Act Text Carrier
Constitutional Amendment 8/1995 promulgation by the Boards (Mesas) of both Houses Art. 21, XI: a law shall provide for "the creation of a regulatory body" Official Gazette, 16 Aug 1995
Law 9,472/1997 enacted by Congress, with presidential assent Art. 8 creates Anatel; Art. 22, IV empowers its Board to "issue rules" Official Gazette, 17 Jul 1997
Anatel Resolution 765/2023 vote of the Board of Directors, after a public consultation annex, Art. 4, XVI: the consumer's right to terminate at any time, at no charge Official Gazette, 10 Nov 2023

Proposal.

  • Add this chain to ODP-004 as a second example, so that legal basis for, derives authority from, and delegates legal authority to are shown working in a codified system as well as a common-law one.
  • Continue it into the deck's invented consumer case: contract, non-performance, judgment. That exercises the remaining nodes: triggering occurrence, undischarged during, remedial position (Civil Code Arts. 389 and 475), and the individual norm in the operative part of a judgment.
  • Note the role of the bar on self-help (Criminal Code Art. 345). The injured party holds a remedial claim but no power to enforce it herself; enforcement runs through a court's power. That is a clean example of a remedy that requires another authority's act.

Help from anyone who knows the Brazilian sources is welcome, especially Danilo.

Map an Akoma Ntoso or ELI element to the design patterns

Raised by dkoepsell on 17 October 2026 · 0 comments · good first issue help wanted · comment without GitHub · comment on GitHub

LOF-ODP-001 separates legal acts, legal content, and legal documents, and adds legal norms, definitions, and designators. Akoma Ntoso and the European Legislation Identifier (ELI) distinguish similar things (work, expression, manifestation; references and identifiers). How to help: pick one Akoma Ntoso or ELI element and say which ODP-001 term it corresponds to, with a link to its documentation. A small table in a comment is ideal. Familiarity with either standard helps; no OWL needed.

Full text

LOF-ODP-001 separates legal acts, legal content, and legal documents, and adds legal norms, definitions, and designators. Akoma Ntoso and the European Legislation Identifier (ELI) distinguish similar things (work, expression, manifestation; references and identifiers).

How to help: pick one Akoma Ntoso or ELI element and say which ODP-001 term it corresponds to, with a link to its documentation. A small table in a comment is ideal. Familiarity with either standard helps; no OWL needed.

Suggest an existing legal ontology or vocabulary for the registry

Raised by dkoepsell on 17 October 2026 · 0 comments · good first issue help wanted · comment without GitHub · comment on GitHub

The Foundry's registry should list legal ontologies and vocabularies that already exist, whether or not they are BFO-based, so that we can align with them rather than duplicate them. How to help: add a comment with - the name and a link - its license - what it covers (for example contracts, legislation, courts, compliance) - whether it is built on an upper ontology (BFO, DOLCE, UFO, none) One comment per ontology is fine. No GitHub or ontology expertise needed.

Full text

The Foundry's registry should list legal ontologies and vocabularies that already exist, whether or not they are BFO-based, so that we can align with them rather than duplicate them.

How to help: add a comment with - the name and a link - its license - what it covers (for example contracts, legislation, courts, compliance) - whether it is built on an upper ontology (BFO, DOLCE, UFO, none)

One comment per ontology is fine. No GitHub or ontology expertise needed.

Contribute a worked example from your jurisdiction

Raised by dkoepsell on 17 October 2026 · 0 comments · good first issue help wanted · comment without GitHub · comment on GitHub

The design patterns are tested on worked examples: a residential lease, a change of command, and a statute with its implementing regulation (see the design patterns). We need examples from other legal systems and areas of practice. How to help: describe a short, real case in plain language in a comment: who did what, under which law, and what rights, duties, or roles resulted. For example, an appointment of a guardian, a contract assignment, or...

Full text

The design patterns are tested on worked examples: a residential lease, a change of command, and a statute with its implementing regulation (see the design patterns). We need examples from other legal systems and areas of practice.

How to help: describe a short, real case in plain language in a comment: who did what, under which law, and what rights, duties, or roles resulted. For example, an appointment of a guardian, a contract assignment, or a power delegated by statute. A coordinator will help turn it into a test case. Civil-law, common-law, and international examples are all welcome.

Review one definition in a design pattern

Raised by dkoepsell on 17 October 2026 · 0 comments · good first issue help wanted · comment without GitHub · comment on GitHub

Clear definitions are the heart of each pattern. Pick a single term from any of the four design patterns, such as legal act, legal role, claim, or legal basis for, and tell us: - Is the definition clear to someone outside ontology? - Does it fit how the term is used in your legal system? - Is there a case it gets wrong? Comment here or at the foot of the pattern's page. Short comments are useful.

Full text

Clear definitions are the heart of each pattern. Pick a single term from any of the four design patterns, such as legal act, legal role, claim, or legal basis for, and tell us:

  • Is the definition clear to someone outside ontology?
  • Does it fit how the term is used in your legal system?
  • Is there a case it gets wrong?

Comment here or at the foot of the pattern's page. Short comments are useful.

Add a reading to the Literature pages

Raised by dkoepsell on 17 October 2026 · 0 comments · good first issue help wanted · comment without GitHub · comment on GitHub

The Literature section starts with Adolf Reinach. We would like pages or entries on Hohfeld and legal positions, document acts and social ontology, and existing legal ontologies. How to help: comment with a full citation and a link to a lawful copy, or open a pull request editing a page under docs/literature/ (see Contributing).

Full text

The Literature section starts with Adolf Reinach. We would like pages or entries on Hohfeld and legal positions, document acts and social ontology, and existing legal ontologies.

How to help: comment with a full citation and a link to a lawful copy, or open a pull request editing a page under docs/literature/ (see Contributing).

Discussion