{
"title": "Keep the claim and its grounds together — Ontologies & vocabularies",
"text": "MODAVIS (2026). Keep the claim and its grounds together — Ontologies & vocabularies. Research notebooks, September 2026 · Connected evidence edition. https://modavis.org/editions/2026-09-r6/ontologies/evidence/",
"bibtex": "@misc{modavis_ontologies_evidence_2026_09_r6,\n author = {{MODAVIS}},\n title = {Keep the claim and its grounds together — Ontologies & vocabularies},\n year = {2026},\n month = {09},\n version = {2026-09-r6},\n url = {https://modavis.org/editions/2026-09-r6/ontologies/evidence/},\n note = {Content SHA-256: 78fc99fb49c1ca3d29c65bd21bd180845987ec66b000f0b2aceb02bb3d0e5138}\n}",
"url": "https://modavis.org/editions/2026-09-r6/ontologies/evidence/",
"sha256": "78fc99fb49c1ca3d29c65bd21bd180845987ec66b000f0b2aceb02bb3d0e5138",
"edition": "2026-09-r6",
"date": "2026-09-26",
"canonicalPayload": "{\"book\":{\"number\":\"07\",\"slug\":\"ontologies\",\"title\":\"Ontologies & vocabularies\"},\"chapter\":{\"blocks\":[{\"items\":[{\"detail\":\"A continuous worked example: readable explanations, exact records and the limits of each claim.\",\"href\":\"/notebooks/ontologies/cuntz/\",\"label\":\"Learn this through the Cuntz organ\"}],\"type\":\"links\"},{\"text\":\"The ontology distinguishes a source, a fixed snapshot of that source, a relevant fragment and the claim supported by it. It also represents conflicts and editorial decisions. A simplified research view can then remain connected to the richer account behind it.\",\"type\":\"p\"},{\"caption\":\"Evidence and decisions support a representation; an interface provides one view of that representation.\",\"kind\":\"evidence\",\"type\":\"figure\"},{\"id\":\"time\",\"text\":\"Time and context are not optional decoration\",\"type\":\"heading\"},{\"text\":\"A component’s role or an instrument’s configuration can depend on a historical state. An assertion’s applicability can depend on context. If the scope is not known, it should not silently become universal.\",\"type\":\"p\"},{\"boundary\":\"No exact acquisition date is inferred from the timestamp in this example.\",\"caseLabel\":\"Cuntz\",\"concept\":\"Provenance describes how a record came into being. Applicability describes when a claim concerns its subject. A file timestamp is a third, potentially different fact.\",\"evidence\":\"Released record\",\"expanded\":false,\"id\":\"cuntz-time\",\"scenario\":\"The same Cuntz observation carries 2020-09-16T10:33:05Z and flags identifying the result time as a workbook modification timestamp, with a source series marked 2019.\",\"sources\":[{\"detail\":\"Published manifest: identities, measurement observations, realization digests, profiles and rights. Exact examples checked against this release.\",\"href\":\"https://doi.org/10.5281/zenodo.22151203\",\"label\":\"Cuntz Positiv · VAO 0.5.0-rc.2\"},{\"detail\":\"Identity, state, configuration, evidence, assertions, context and provenance. Teaching examples explain its distinctions without inventing new normative properties.\",\"href\":\"https://doi.org/10.5281/zenodo.22126086\",\"label\":\"MODAVIS Ontology Network 0.1.0\"}],\"steps\":[{\"text\":\"The measurement activity explicitly says its timestamp is the workbook modification time, not an asserted measurement-session time.\",\"title\":\"Read the qualification\"},{\"text\":\"Keep the source-series label, the workbook timestamp and any independently documented session date as separate information. Do not turn the most precise-looking timestamp into the date the pipe was measured.\",\"title\":\"Preserve the distinctions\"},{\"text\":\"A query for files modified in 2020 can use this timestamp. A query for instruments measured during 2020 needs evidence of session dates and cannot silently use the same field.\",\"title\":\"Ask the right query\"}],\"takeaway\":\"Precision of formatting is not precision of historical knowledge.\",\"title\":\"Is the workbook timestamp the measurement date?\",\"type\":\"example\"},{\"id\":\"validation\",\"text\":\"Validate the model, review the interpretation\",\"type\":\"heading\"},{\"text\":\"OWL distinctions and SHACL constraints help catch incompatible identities and malformed data. They complement, rather than replace, source criticism and organological judgement.\",\"type\":\"p\"},{\"id\":\"evidence-chain\",\"text\":\"A source, a fragment and an assertion\",\"type\":\"heading\"},{\"headers\":[\"Record\",\"Purpose\"],\"rows\":[[\"Source\",\"Identifies the publication, collection or other source being consulted.\"],[\"Fixed snapshot\",\"Identifies the particular captured or versioned form of that source.\"],[\"Fragment or selector\",\"Locates the passage, image region or other relevant part.\"],[\"Assertion\",\"States an attributable claim with its scope and status.\"],[\"Decision and projection\",\"Records how a reviewed interpretation becomes available in a research view.\"]],\"type\":\"table\"},{\"text\":\"A stable source URL is useful but may not identify the wording that was actually read. A fixed snapshot and a precise selector make a claim more inspectable, while a source citation helps readers assess authorship and context. Fixity establishes content identity; source criticism addresses whether the content warrants the claim.\",\"type\":\"p\"},{\"boundary\":\"The example reproduces a release observation; it does not independently remeasure the pipe or supply a calibration uncertainty that the record does not provide.\",\"caseLabel\":\"Cuntz\",\"concept\":\"An observation connects a property and a result to the thing observed. A unit gives the number a physical interpretation; a source selector identifies its evidential location.\",\"evidence\":\"Released record\",\"expanded\":false,\"id\":\"cuntz-measurement\",\"scenario\":\"The published Cuntz manifest contains an observation for rank:ged:pipe:key-036 with property code GesL, value 1281, unit MilliM, and source-cell flag Holzgedackt 8!D9.\",\"sources\":[{\"detail\":\"Published manifest: identities, measurement observations, realization digests, profiles and rights. Exact examples checked against this release.\",\"href\":\"https://doi.org/10.5281/zenodo.22151203\",\"label\":\"Cuntz Positiv · VAO 0.5.0-rc.2\"},{\"detail\":\"The standard contract is separate from the Cuntz dataset content version 0.5.0-rc.2.\",\"href\":\"https://doi.org/10.5281/zenodo.22214248\",\"label\":\"VAO Standard 0.5.0\"}],\"steps\":[{\"text\":\"The number belongs to the specified pipe entity and the workbook’s GesL property. It is not a dimension of the complete organ or an arbitrary distance on the screen.\",\"title\":\"Identify the subject and property\"},{\"text\":\"The stored value is 1281 millimetres. Converting it gives 1.281 metres; keep that conversion separate from the original source value and unit.\",\"title\":\"Retain value and unit together\"},{\"text\":\"The observation links to the exact workbook realization, the measurement activity and the cell selector. A later reader can check the entry and consult the workbook’s definition of GesL.\",\"title\":\"Locate the source\"}],\"takeaway\":\"A reusable measurement needs a subject, property, value, unit and evidence location.\",\"title\":\"What exactly does “1281” tell us?\",\"type\":\"example\"},{\"id\":\"retain-conflict\",\"text\":\"Retain conflict as information\",\"type\":\"heading\"},{\"text\":\"If two sources disagree about a maker or a date, preserve both attributed assertions and record the interpretation separately. An accepted convenience view can show a selected result while still exposing alternatives and reasons. Rejecting an interpretation should not make the original source vanish, and superseded decisions should remain distinguishable from currently accepted ones.\",\"type\":\"p\"},{\"boundary\":\"This is a teaching application of the assertion model, not a claim that a new authoritative review of the historical register mapping has been completed.\",\"caseLabel\":\"Cuntz\",\"concept\":\"An assertion is an attributable claim. A review decision records how that claim is interpreted; a projection is a selected view of the reviewed knowledge.\",\"evidence\":\"Unresolved interpretation\",\"expanded\":false,\"id\":\"cuntz-assertions\",\"scenario\":\"The Cuntz thesis description and the digital inventory present different register counts. Treating the latest imported count as an unconditional replacement would erase the question the discrepancy raises.\",\"sources\":[{\"detail\":\"Published manifest: identities, measurement observations, realization digests, profiles and rights. Exact examples checked against this release.\",\"href\":\"https://doi.org/10.5281/zenodo.22151203\",\"label\":\"Cuntz Positiv · VAO 0.5.0-rc.2\"},{\"detail\":\"Identity, state, configuration, evidence, assertions, context and provenance. Teaching examples explain its distinctions without inventing new normative properties.\",\"href\":\"https://doi.org/10.5281/zenodo.22126086\",\"label\":\"MODAVIS Ontology Network 0.1.0\"},{\"detail\":\"Submitted dissertation, 11 September 2026; §§8.2.2.1 and 8.4.1, printed pp. 264–265 and 275–276. The historical description and digital inventory require a resolved mapping; the manuscript is not redistributed here.\",\"label\":\"Dominik Ukolov · Musikinstrumente im virtuellen Raum (2026)\"}],\"steps\":[{\"text\":\"Retain the four-register physical description and the five-label digital configuration, including their sources and intended subjects.\",\"title\":\"Store both claims with their scope\"},{\"text\":\"Ask whether the accounts concern comparable states and counting units. The missing Regal 8′ mapping remains an explicit review issue rather than an invented explanation.\",\"title\":\"Review the relationship\"},{\"text\":\"A catalogue summary can show a selected statement and its qualification, while an evidence view exposes the other assertion and the reason the relationship remains unresolved.\",\"title\":\"Publish an understandable view\"}],\"takeaway\":\"A clean interface can present an editorial decision without deleting the evidence behind competing interpretations.\",\"title\":\"How can two accounts disagree without corrupting the database?\",\"type\":\"example\"},{\"id\":\"context-provenance\",\"text\":\"Context is different from provenance\",\"type\":\"heading\"},{\"text\":\"Provenance explains how a statement entered the research record. Context explains where or when the statement applies. Knowing who extracted a stop assignment does not establish which historical configuration it describes. Conversely, knowing the intended configuration does not identify the source and processing steps that support it. Both relationships matter, and an unspecified context should remain unspecified rather than becoming a universal claim.\",\"type\":\"p\"},{\"text\":\"A reviewer may accept one interpretation from a source while leaving another unresolved. A blanket “trusted source” label is too coarse to express the different grounds of individual claims.\",\"title\":\"Review follows the claim\",\"type\":\"note\"},{\"items\":[{\"detail\":\"The published semantic model.\",\"href\":\"https://doi.org/10.5281/zenodo.22126086\",\"label\":\"MODAVIS Ontology Network 0.1.0\"}],\"type\":\"links\"}],\"intro\":\"An accepted fact can be useful without making its supporting evidence disappear.\",\"slug\":\"evidence\",\"sources\":[{\"detail\":\"The published semantic model.\",\"href\":\"https://doi.org/10.5281/zenodo.22126086\",\"label\":\"MODAVIS Ontology Network 0.1.0\"}],\"title\":\"Keep the claim and its grounds together\"},\"date\":\"2026-09-26\",\"edition\":\"2026-09-r6\",\"figures\":{\"evidence\":\"\"}}",
"payload": {
"edition": "2026-09-r6",
"date": "2026-09-26",
"book": {
"slug": "ontologies",
"title": "Ontologies & vocabularies",
"number": "07"
},
"chapter": {
"slug": "evidence",
"title": "Keep the claim and its grounds together",
"intro": "An accepted fact can be useful without making its supporting evidence disappear.",
"blocks": [
{
"type": "links",
"items": [
{
"label": "Learn this through the Cuntz organ",
"href": "/notebooks/ontologies/cuntz/",
"detail": "A continuous worked example: readable explanations, exact records and the limits of each claim."
}
]
},
{
"type": "p",
"text": "The ontology distinguishes a source, a fixed snapshot of that source, a relevant fragment and the claim supported by it. It also represents conflicts and editorial decisions. A simplified research view can then remain connected to the richer account behind it."
},
{
"type": "figure",
"kind": "evidence",
"caption": "Evidence and decisions support a representation; an interface provides one view of that representation."
},
{
"type": "heading",
"text": "Time and context are not optional decoration",
"id": "time"
},
{
"type": "p",
"text": "A component’s role or an instrument’s configuration can depend on a historical state. An assertion’s applicability can depend on context. If the scope is not known, it should not silently become universal."
},
{
"type": "example",
"id": "cuntz-time",
"title": "Is the workbook timestamp the measurement date?",
"concept": "Provenance describes how a record came into being. Applicability describes when a claim concerns its subject. A file timestamp is a third, potentially different fact.",
"scenario": "The same Cuntz observation carries 2020-09-16T10:33:05Z and flags identifying the result time as a workbook modification timestamp, with a source series marked 2019.",
"steps": [
{
"title": "Read the qualification",
"text": "The measurement activity explicitly says its timestamp is the workbook modification time, not an asserted measurement-session time."
},
{
"title": "Preserve the distinctions",
"text": "Keep the source-series label, the workbook timestamp and any independently documented session date as separate information. Do not turn the most precise-looking timestamp into the date the pipe was measured."
},
{
"title": "Ask the right query",
"text": "A query for files modified in 2020 can use this timestamp. A query for instruments measured during 2020 needs evidence of session dates and cannot silently use the same field."
}
],
"takeaway": "Precision of formatting is not precision of historical knowledge.",
"boundary": "No exact acquisition date is inferred from the timestamp in this example.",
"sources": [
{
"label": "Cuntz Positiv · VAO 0.5.0-rc.2",
"href": "https://doi.org/10.5281/zenodo.22151203",
"detail": "Published manifest: identities, measurement observations, realization digests, profiles and rights. Exact examples checked against this release."
},
{
"label": "MODAVIS Ontology Network 0.1.0",
"href": "https://doi.org/10.5281/zenodo.22126086",
"detail": "Identity, state, configuration, evidence, assertions, context and provenance. Teaching examples explain its distinctions without inventing new normative properties."
}
],
"expanded": false,
"caseLabel": "Cuntz",
"evidence": "Released record"
},
{
"type": "heading",
"text": "Validate the model, review the interpretation",
"id": "validation"
},
{
"type": "p",
"text": "OWL distinctions and SHACL constraints help catch incompatible identities and malformed data. They complement, rather than replace, source criticism and organological judgement."
},
{
"type": "heading",
"text": "A source, a fragment and an assertion",
"id": "evidence-chain"
},
{
"type": "table",
"headers": [
"Record",
"Purpose"
],
"rows": [
[
"Source",
"Identifies the publication, collection or other source being consulted."
],
[
"Fixed snapshot",
"Identifies the particular captured or versioned form of that source."
],
[
"Fragment or selector",
"Locates the passage, image region or other relevant part."
],
[
"Assertion",
"States an attributable claim with its scope and status."
],
[
"Decision and projection",
"Records how a reviewed interpretation becomes available in a research view."
]
]
},
{
"type": "p",
"text": "A stable source URL is useful but may not identify the wording that was actually read. A fixed snapshot and a precise selector make a claim more inspectable, while a source citation helps readers assess authorship and context. Fixity establishes content identity; source criticism addresses whether the content warrants the claim."
},
{
"type": "example",
"id": "cuntz-measurement",
"title": "What exactly does “1281” tell us?",
"concept": "An observation connects a property and a result to the thing observed. A unit gives the number a physical interpretation; a source selector identifies its evidential location.",
"scenario": "The published Cuntz manifest contains an observation for rank:ged:pipe:key-036 with property code GesL, value 1281, unit MilliM, and source-cell flag Holzgedackt 8!D9.",
"steps": [
{
"title": "Identify the subject and property",
"text": "The number belongs to the specified pipe entity and the workbook’s GesL property. It is not a dimension of the complete organ or an arbitrary distance on the screen."
},
{
"title": "Retain value and unit together",
"text": "The stored value is 1281 millimetres. Converting it gives 1.281 metres; keep that conversion separate from the original source value and unit."
},
{
"title": "Locate the source",
"text": "The observation links to the exact workbook realization, the measurement activity and the cell selector. A later reader can check the entry and consult the workbook’s definition of GesL."
}
],
"takeaway": "A reusable measurement needs a subject, property, value, unit and evidence location.",
"boundary": "The example reproduces a release observation; it does not independently remeasure the pipe or supply a calibration uncertainty that the record does not provide.",
"sources": [
{
"label": "Cuntz Positiv · VAO 0.5.0-rc.2",
"href": "https://doi.org/10.5281/zenodo.22151203",
"detail": "Published manifest: identities, measurement observations, realization digests, profiles and rights. Exact examples checked against this release."
},
{
"label": "VAO Standard 0.5.0",
"href": "https://doi.org/10.5281/zenodo.22214248",
"detail": "The standard contract is separate from the Cuntz dataset content version 0.5.0-rc.2."
}
],
"expanded": false,
"caseLabel": "Cuntz",
"evidence": "Released record"
},
{
"type": "heading",
"text": "Retain conflict as information",
"id": "retain-conflict"
},
{
"type": "p",
"text": "If two sources disagree about a maker or a date, preserve both attributed assertions and record the interpretation separately. An accepted convenience view can show a selected result while still exposing alternatives and reasons. Rejecting an interpretation should not make the original source vanish, and superseded decisions should remain distinguishable from currently accepted ones."
},
{
"type": "example",
"id": "cuntz-assertions",
"title": "How can two accounts disagree without corrupting the database?",
"concept": "An assertion is an attributable claim. A review decision records how that claim is interpreted; a projection is a selected view of the reviewed knowledge.",
"scenario": "The Cuntz thesis description and the digital inventory present different register counts. Treating the latest imported count as an unconditional replacement would erase the question the discrepancy raises.",
"steps": [
{
"title": "Store both claims with their scope",
"text": "Retain the four-register physical description and the five-label digital configuration, including their sources and intended subjects."
},
{
"title": "Review the relationship",
"text": "Ask whether the accounts concern comparable states and counting units. The missing Regal 8′ mapping remains an explicit review issue rather than an invented explanation."
},
{
"title": "Publish an understandable view",
"text": "A catalogue summary can show a selected statement and its qualification, while an evidence view exposes the other assertion and the reason the relationship remains unresolved."
}
],
"takeaway": "A clean interface can present an editorial decision without deleting the evidence behind competing interpretations.",
"boundary": "This is a teaching application of the assertion model, not a claim that a new authoritative review of the historical register mapping has been completed.",
"sources": [
{
"label": "Cuntz Positiv · VAO 0.5.0-rc.2",
"href": "https://doi.org/10.5281/zenodo.22151203",
"detail": "Published manifest: identities, measurement observations, realization digests, profiles and rights. Exact examples checked against this release."
},
{
"label": "MODAVIS Ontology Network 0.1.0",
"href": "https://doi.org/10.5281/zenodo.22126086",
"detail": "Identity, state, configuration, evidence, assertions, context and provenance. Teaching examples explain its distinctions without inventing new normative properties."
},
{
"label": "Dominik Ukolov · Musikinstrumente im virtuellen Raum (2026)",
"detail": "Submitted dissertation, 11 September 2026; §§8.2.2.1 and 8.4.1, printed pp. 264–265 and 275–276. The historical description and digital inventory require a resolved mapping; the manuscript is not redistributed here."
}
],
"expanded": false,
"caseLabel": "Cuntz",
"evidence": "Unresolved interpretation"
},
{
"type": "heading",
"text": "Context is different from provenance",
"id": "context-provenance"
},
{
"type": "p",
"text": "Provenance explains how a statement entered the research record. Context explains where or when the statement applies. Knowing who extracted a stop assignment does not establish which historical configuration it describes. Conversely, knowing the intended configuration does not identify the source and processing steps that support it. Both relationships matter, and an unspecified context should remain unspecified rather than becoming a universal claim."
},
{
"type": "note",
"title": "Review follows the claim",
"text": "A reviewer may accept one interpretation from a source while leaving another unresolved. A blanket “trusted source” label is too coarse to express the different grounds of individual claims."
},
{
"type": "links",
"items": [
{
"label": "MODAVIS Ontology Network 0.1.0",
"href": "https://doi.org/10.5281/zenodo.22126086",
"detail": "The published semantic model."
}
]
}
],
"sources": [
{
"label": "MODAVIS Ontology Network 0.1.0",
"href": "https://doi.org/10.5281/zenodo.22126086",
"detail": "The published semantic model."
}
]
},
"figures": {
"evidence": ""
}
}
}