If a Set of objects has been enumerated such that all members have URIs, then there is an existing pattern: The set of objects is equivalent to an auction lot, which is used_specific_object from the Provenance Activity, and the parts of the activity are the different Acquisitions and so forth. See: https://linked.art/model/provenance/auctions/#purchase-of-lot
However, for sets where the Set is the only documentation, such as large sets of coins, fragments, archival material, and so on, we don't want to (and indeed cannot) transferred_title_of the Set itself. Two reasons: 1. (semantic) The intellectual ownership of the Set doesn't change. 2. (technical) Sets are conceptual, not physical, and transferred_title_of has a range of Physical Object.
So the proposal is to use members_exemplified_by as our existing escape hatch for these sorts of things. The members of the Set would be exemplified by a Human Made Object, which has the changed_ownership_through relationship, per the simple historical provenance pattern. This would only be used for sets where the items aren't enumerated with their own URIs to avoid multiple patterns for the same thing. If each letter has its own URI and record, then we already have both changed_ownership_through and the full provenance model.
As this doesn't introduce any new relationships or classes, and doesn't change anything, it can come in with 1.1
E.g. a combination of
@ewg118 created this gist in turtle:
https://gist.github.com/ewg118/2c104718a5705f455ed92c2d8c4dfe0d
Or in JSON-LD
{
"type": "Set",
"_label": "1955.190",
"members_exemplified_by": [
{
"type": "HumanMadeObject",
"classified_as": [ {...} ],
"changed_ownership_through": [
{
"type": "Acquisition",
"timespan": {
"type": "TimeSpan",
"begin_of_the_begin": "1955-12-01T00:00:00Z",
"end_of_the_end": "1955-12-01T23:59:59Z"
},
"transferred_title_from": {...},
"transferred_title_to": {...}
}
]
}
]
}
If a Set of objects has been enumerated such that all members have URIs, then there is an existing pattern: The set of objects is equivalent to an auction lot, which is
used_specific_objectfrom the Provenance Activity, and the parts of the activity are the different Acquisitions and so forth. See: https://linked.art/model/provenance/auctions/#purchase-of-lotHowever, for sets where the Set is the only documentation, such as large sets of coins, fragments, archival material, and so on, we don't want to (and indeed cannot)
transferred_title_ofthe Set itself. Two reasons: 1. (semantic) The intellectual ownership of the Set doesn't change. 2. (technical) Sets are conceptual, not physical, andtransferred_title_ofhas a range of Physical Object.So the proposal is to use
members_exemplified_byas our existing escape hatch for these sorts of things. The members of the Set would be exemplified by a Human Made Object, which has thechanged_ownership_throughrelationship, per the simple historical provenance pattern. This would only be used for sets where the items aren't enumerated with their own URIs to avoid multiple patterns for the same thing. If each letter has its own URI and record, then we already have bothchanged_ownership_throughand the full provenance model.As this doesn't introduce any new relationships or classes, and doesn't change anything, it can come in with 1.1
E.g. a combination of
part_of@ewg118 created this gist in turtle:
https://gist.github.com/ewg118/2c104718a5705f455ed92c2d8c4dfe0d
Or in JSON-LD
{ "type": "Set", "_label": "1955.190", "members_exemplified_by": [ { "type": "HumanMadeObject", "classified_as": [ {...} ], "changed_ownership_through": [ { "type": "Acquisition", "timespan": { "type": "TimeSpan", "begin_of_the_begin": "1955-12-01T00:00:00Z", "end_of_the_end": "1955-12-01T23:59:59Z" }, "transferred_title_from": {...}, "transferred_title_to": {...} } ] } ] }