Copyright attribute

[krgomez]

We would like to create a copyright attribute for the UAMN Fine Arts collection to use. We need a place to record the artwork copyright that is separate from media, as the copyright of an artwork pertains to the object, not just an individual image of the object. If there are no objections, we would like this attribute added:

copyright - Statement of copyright ownership.

[Jegelewicz]

Sounds reasonable. Will this be free text or is there a controlled vocabulary?

[krgomez]

This will be free text. We will often be using © artist's name, and so controlled vocabulary won't work.

[AJLinn]

I agree. This is a field we can use in UAM:EH as well.

[dustymc]

Are "who holds copyright in the collection?" or "what copyrights does this artist hold?" the kinds of questions you might ever want to ask of the data?

[AJLinn]

Since copyright is a 1:1 assignment, I think that "who holds copyright in the collection" probably wouldn't be a relevant question. The "what copyrights does this artist hold" question would be answered with a query by the artist and then looking on an individual piece basis. I can see us recording usage/licensing agreement information here as well.

[dustymc]

copyright is a 1:1 assignment

recording usage/licensing agreement information

That may turn out to be an argument for NOT making the model 1:1 - which I don't think will be a problem anyway.

query by the artist

The details of that are what I'm questioning. If finding specimens by the string "© John Smith" is sufficient, then free-text is sufficient. If you want to know it's THAT John Smith (who has changed his name 17 times since you recorded the copyright information, and which you recorded as Jဝhn Smith - the second character in "John" is not a lower-case o), or if you want to see his copyrighted work from his Agent page, or if you want to do anything else involving linked data, then you need a link to the Agent data object.

[AJLinn]

If you want to know it's THAT John Smith (who has changed his name 17 times since you recorded the copyright information, and which you recorded as Jဝhn Smith - the second character in "John" is not a lower-case o), or if you want to see his copyrighted work from his Agent page, or if you want to do anything else involving linked data, then you need a link to the Agent data object.

Ohhhh, that might be interesting. So the copyrights held could be something that shows up in the Agent Activity report then?

What happens if/when the copyright for a particular artist's work are transferred to another individual? E.g., the artist dies and in their will they pass the copyright management to a gallery or a descendent, etc. How might this be indicated?

Maybe it could be something like this? attribute: copyright status value: controlled vocabulary with things like copyright; license; use agreement remarks: whatever - maybe the © John Smith statement or other notations det. date: date the copyright was declared; date license was executed, etc. det. meth.: whatever - artist signature, published in catalog, written agreement on file, etc. determiner: agent who holds the copyright

[krgomez]

For artworks where the copyright holder is the artist, I think it is fine to not have the copyright attribute linked to their agent record directly. It will be obvious. However, for objects where the copyright holder is an heir, agency, etc, I do agree that it would be nice to be able to see the objects associated with an agency on the agent record. For example, copyright for all Ansel Adams works is managed by The Ansel Adams Publishing Rights Trust. It would be nice to be able to go the The Ansel Adams Publishing Rights Trust agent record and see the association with the Adams object records.

I'm not seeing how else this could be linked except for as you are suggesting Angie in making the copyright holder agent the determiner of the copyright status. I think your idea to use a controlled entry with broad copyright status terms and adding the actual copyright statement as a remark could work. However, I think this strategy makes more sense from a data management point of view and maybe wouldn't be quite clear enough for the user. What's important is that the copyright statement is clearly associated with the object and clearly visible to the user. One thing I'm not crazy about with this solution is that it brings the core information that we want to communicate - the copyright statement - to what seems to me like a kind of lesser level as a remark. Another thing is that if a user is viewing an Ansel Adams object record and they are interested in using an image of the work in some way and they see the attribute: copyright status-in copyright; The Ansel Adams Publishing Rights Trust; some sort of date; ©The Ansel Adams Publishing Rights Trust - what they're probably going to do is search online for the Trust. They're probably not going to know to go to the agent record to view contact information, etc. In which case, for the user, it's not important that the Adams records are discoverable through the Trust agent record.

Here are some examples from other art institutions of object records in their online collections:

https://www.guggenheim.org/artwork/1975 https://walkerart.org/collections/artworks/untitled-canis-major-constellation https://www.brooklynmuseum.org/opencollection/objects/1469 - The Brooklyn Museum is a leader in art museums on how to clearly communicate object rights data and copyright policy to the public

I'm not opposed to doing it this way with controlled entry, but it seems imperfect. We definitely want this to work for both the collections though. So if you're thinking it would be better this way we can do that.

[krgomez]

I had another quick thought. Sorry for my giant essay above. I just realized that of course a determiner can be the copyright holder and link to the agent record even if it is a free text attribute. Now that this obvious fact has occurred to me, I'm not sure what the added benefit would be for a controlled entry.

[Jegelewicz]

Could "copyright holder" be a collector agent type? Then the attribute wouldn't have to include that agent name, it could just have all the other stuff and the attribute determiner could just be whoever entered the copyright info.

[dustymc]

@Jegelewicz yea that was my first thought too, and I still think it's a more-analogous type of data, but its also limited to an agent in a role - I'm not sure it's quite sufficient for this. Someone please tell me I'm wrong....

I'm not really crazy about the attribute determiner pathway, but it's completely recoverable (it's data object keys) - if we get a better idea at some point we can pull agent data from it. That is not true with free-text data. (I have yet to meet a text-based dataset where "It will be obvious" holds up!) Lacking better ideas and assuming that collector really is insufficient, I think attributes sounds reasonable.

Don't confound data and display. We already do weird things...

...with Attributes, more weird things on the specimen detail page (where there's not much data involved so we can afford the computational costs of manipulating it) isn't much of a problem.

I think I sort of like the free-text approach as well - the (controlled) attribute type ("rights holder"? "rights statement"? "license"? "copyright"??) tells you what's going on, the determiner links to an Agent, I think that may be all the structure that's really necessary, "© John Smith" seems like a useful value when accompanied by the linked data.

Displaying linked agent data could be possible (and fun!) as well, or you can just have as many attributes as you need:

• © John Smith (on 2018-07-17; remark: transferred to estate; determiner: John Smith)
• © Estate of John Smith (on 2018-07-18; remark: transferred to other estate; determiner: Estate Of John Smith)
• © Estate of Estate Of John Smith (on 2018-07-19; determiner: Estate of Estate Of John Smith)

Sorta-related, https://github.com/ArctosDB/arctos/issues/1603 seems to be a push to restrict that information, and https://github.com/ArctosDB/arctos/issues/1597 would add administrative data (=stuff most users won't care about) which might display sort of weird.

[Jegelewicz]

Yeah, I like having the agent in the Agent's box, but it does seem to make more sense to have it all together in the attribute....

[krgomez]

Displaying linked agent data could be possible

Can you explain what you mean by this? I'm not sure I follow...

Could "copyright holder" be a collector agent type? Then the attribute wouldn't have to include that agent name, it could just have all the other stuff and the attribute determiner could just be whoever entered the copyright info.

If I understand this correctly you mean that we could potentially just add the copyright holder agents as an additional "collector" agent to the object record? I don't really think this would work because the term collector is too confusing. The copyright holder is not a collector. Even if there were a logical way of putting the copyright holder as an agent along with our other agents associated with the object, it still doesn't link it to the copyright statement which is the important thing for us to be recording.

[dustymc]

linked agent data

We have relationships between agents, we could follow them, or some of them, and do WHATEVER with the data we find there if that's somehow useful. Eg you can click on collectors to get to the agent page, and click on related agents to get to their agent page from there, and etc. Those relationships are capable of doing more than forming links, and maybe the "artist dies and in their will they pass the copyright management to a gallery or a descendent" thing @AJLinn mentioned could be a relationship which does something special (gets displayed on specimen pages or etc.). IDK, clicking links is probably close enough...

collector

Naming things is hard...

Collector is a table which links specimens to agents using roles, one of which is "collector." "Copyright holder" could be a new collector role.

http://arctos.database.museum/info/ctDocumentation.cfm?table=CTCOLLECTOR_ROLE

link it to the copyright statement

Media can do that, and that approach would not require you to change 800 individual attributes when something changes in the copyright statement used by 800 specimens.

[Jegelewicz]

collector

Naming things is hard...

Collector is a table which links specimens to agents using roles, one of which is "collector." "Copyright holder" could be a new collector role.

http://arctos.database.museum/info/ctDocumentation.cfm?table=CTCOLLECTOR_ROLE

See #1492 Maybe those should be "Agents", not collectors and we should present their "role" in the specimen record. We don't have to change the table, just the presentation.

[campmlc]

Agree. Change the name of the table to Agents or "Related Agents" or "Specimen Agents" and allow multiple roles. Even with the current limited number of roles, there is a conflict because preparators and makers are not collectors.

On Thu, Jul 19, 2018 at 6:52 PM, Teresa Mayfield notifications@github.com wrote:

collector

Naming things is hard...

Collector is a table which links specimens to agents using roles, one of which is "collector." "Copyright holder" could be a new collector role.

http://arctos.database.museum/info/ctDocumentation.cfm? table=CTCOLLECTOR_ROLE

See #1492 https://github.com/ArctosDB/arctos/issues/1492 Maybe those should be "Agents", not collectors and we should present their "role" in the specimen record. We don't have to change the table, just the presentation.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ArctosDB/arctos/issues/1610#issuecomment-406447979, or mute the thread https://github.com/notifications/unsubscribe-auth/AOH0hCMnxaCfFycpvTMKApzizNJMxSI2ks5uIRvLgaJpZM4VVRK1 .

[dustymc]

just the presentation

Where?

Even with the current limited number of roles, there is a conflict because preparators and makers are not collectors.

Hu??

[Jegelewicz]

presentation:

We are talking about "collectors" here, not "verbatim collectors" although the two conversations are related. Preparators and makers appear on the specimen pages as "Collector(s)", which is misleading. I like the title "Related Agents", then I would prefer to see something like

John Smith, collector Jane Doe, preparator

[dustymc]

Preparators and makers appear on the specimen pages as "Collector(s)",

They do?!?

[Jegelewicz]

Well looky there! I couldn't find a record with them all! Okay then, that means that using a collector type of "copyright holder" would show up pretty nicely, but I am not really the end user. It is up to @krgomez if she prefers this plus an attribute or just putting everything in an attribute. Who else would have a need for this that we make sure has a chance to chime in? @AJLinn ?

[AJLinn]

Collector is a table which links specimens to agents using roles, one of which is "collector." "Copyright holder" could be a new collector role.

I think this will be okay - making an agent role "copyright holder" and a name makes a clear statement to our external users, analogous to the Brooklyn Museum's format for doing this.

link it to the copyright statement

Media can do that, and that approach would not require you to change 800 individual attributes when something changes in the copyright statement used by 800 specimens.

Yeah, this is a good point that @dustymc makes. Could "[UAM] Copyright Status" be a project that gets linked to the object? (But we can only link loans and accessions to projects, is that right?) The copyright statement is in the description of the project, then if we change the statement, we do it in one place and all 800 linkages are seamlessly affected.

[krgomez]

Thanks so much for thinking through this. I was thinking you all were referring to using the collector agent from the collector table. It does seem weird to me to call it the collector table but I get it. I didn't realize it might be possible to add "copyright holder" to the table along with maker, collector and preparator. This seems like it would be a very functional solution to the need to record copyright holders for the collection. I agree with Angie that this would work and could be useful, but I honestly don't know that it's 100% necessary either. I'm not sure.

Even if we were able to have copyright holder agents, I think we still also need a copyright attribute. I think it is important to present a copyright statement that follows the standardized format used by others. The end user is used to seeing this information presented a certain way and it should be easy for them to see it. For any artworks in the public domain or for which the copyright holder hasn't been identified, etc., this isn't something that could be documented as a copyright holder agent. I think we need an attribute to document this.

As for the choice to use a free text field or controlled vocabulary, I think through this discussion I am seeing the possibility of controlled vocabulary. I had been thinking of this one particular way of recording this information, but I see now that there are other options. While I feel strongly that it is important to have a copyright statement and was thinking that it might not seem clear enough as a remark in a copyright status controlled entry attribute, maybe that would be fine. If we were to use controlled vocabulary I think something pretty much like what you suggested @AJLinn would work: copyright status (in copyright, in copyright - rights-holder(s) unlocatable or unidentifiable, public domain, copyright undetermined, copyright not evaluated). I am drawing these statements from www.rightsstatements.org. This is a resource other institutions are using to standardize these statements.

Yeah, this is a good point that @dustymc makes. Could "[UAM] Copyright Status" be a project that gets linked to the object? (But we can only link loans and accessions to projects, is that right?) The copyright statement is in the description of the project, then if we change the statement, we do it in one place and all 800 linkages are seamlessly affected.

I have yet to use projects and so I don't really understand how this would work but am curious!

[dustymc]

agent role "copyright holder"

Sweet! Is

Agent holding copyright to physical material.

a suitable definition?

project

The core function of Projects is to show usage. There are a few "proper" ways you could link specimens to Projects:

• Link the accession to a project. (Accessions are administrative, so there's no "real" reason you couldn't reaccession stuff if that doesn't quite work, but I realize things generally aren't actually quite that arbitrary.)
• Link a loan to a project. Loans are intended to (among other things) document collection usage, and you're "using" the object to figure out copyright. A bit of a stretch maybe...
• File an issue about projects and borrows - if you don't own the copyright maybe you don't own the object and it makes some sort of sense to "loan" it to yourself (borrows are "incoming loans" - sorta). I'm out of my depth here, I have no idea if that makes actual sense...

If none of that works, Projects have URLs and Media is "things with URLs." You could use Media to attach specimens to projects (from either direction - specimens have URLs too). I don't see what that could DO that a simple PDF (HTML doc, whatever) couldn't, but if there's some other reason to use Projects and avoid project-transactions it's available.

also need a copyright attribute

I think I like it. Collector (table, not role) provides the link to Agents, the attribute provides the text (controlled or whatever ya'll want to type), the attribute determiner is just what it sounds like - the person (probably associated with the collection) making the assertion of copyright/use restrictions. I don't really like the dependent data (one components of this won't make complete sense without the others), but there's no stretching of data either - everything is in the normal place doing the normal things.

Yay??

[krgomez]

Great! That will be wonderful to have a copyright holder agent role to use. The definition looks good to me. Thanks! Also thanks for describing how Projects might be used for this. There seems to be lots of possibilities.

We have decided that what would be most useful for both EH and ART is to have a controlled vocabulary attribute. We would like the attribute to be called "copyright status" and the values to be:

In Copyright: Item is protected by copyright and/or related rights. In Copyright - Rights-Holder(s) Unlocatable or Unidentifiable: Item is protected by copyright and/or related rights. However, for this Item, either (a) no rights-holder(s) have been identified or (b) one or more rights-holder(s) have been identified but none have been located. See also: http://rightsstatements.org/vocab/InC-RUU/1.0/ Copyright Undetermined: The copyright and related rights status of this Item has been reviewed by the organization that has made the Item available, but the organization was unable to make a conclusive determination as to the copyright status of the Item. See also: http://rightsstatements.org/vocab/UND/1.0/ Copyright Not Evaluated: The copyright and related rights status of this Item has not been evaluated. See also: http://rightsstatements.org/vocab/CNE/1.0/ Public Domain: Item is in the Public Domain. License Agreement: Item is in copyright and a license agreement is secured with copyright holder.

We can record the copyright statement specific to the object in the remarks.

Does this work? Does it look okay @AJLinn and @marecaguthrie ? Would anyone else be using this attribute that would find some other value useful?

[Jegelewicz]

This thread just made my day! :-)

[dustymc]

copyright holder agent

Done.

The Attribute looks fine to me, but I'll wait as long as possible in hopes of more feedback. Let me know when/if it's holding up progress...

[AJLinn]

Looks perfect to me. Thanks to all who have contributed!

[marecaguthrie]

Woooo! This is fantastic. Thank you all!

On Fri, Jul 20, 2018 at 2:03 PM Angela Linn notifications@github.com wrote:

Looks perfect to me. Thanks to all who have contributed!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ArctosDB/arctos/issues/1610#issuecomment-406738345, or mute the thread https://github.com/notifications/unsubscribe-auth/AlUox1qmyh9on4IPlqP7sF63ZXshBKszks5uIlPFgaJpZM4VVRK1 .

-- Sent from Gmail Mobile

[krgomez]

@dustymc I added an agent using the copyright holder role to one of our records just to see. It saves properly but it's not displaying on the record. Maybe it just needs some time or something but I thought I'd let you know. This is the record I added it to if that helps somehow: https://arctos.database.museum/guid/UAM:Art:UAP1983-004-001.

[Jegelewicz]

The Attribute looks fine to me, but I'll wait as long as possible in hopes of more feedback. Let me know when/if it's holding up progress...

Created a new issue for the attribute to get it out of the weeds. #1634