Wikidata talk:WP EMEW/Manors/Data Modelling
Latest comment: 3 years ago by PKM in topic Ownership
|
Where new questions, issues to think about, options, or issues that particular options might cause can be summarised in a single line, please add that to the main modelling page.
More detailed, in-depth discussions can then proceed here.
Heritage status
edit- Grade II listed park and garden (Q104783251) -- looks like a listing classification, is it appropriate as a value for P31 ?
- I don't think so. Instance of [park, house, stable, whatever], plus heritage designation (P1435), IMHO. See also Wikidata:WikiProject Built heritage/Data structure. - PKM (talk) 22:22, 24 February 2021 (UTC)
Has part/part of?
edit- questionable: location (P276) or part of (P361) for inverse ?
- User:PKM: My first inclination is that a listed park or building should be part of (P361) the estate (and the inverse), as well as link to it via location (P276)
- I agree, unless there is a good reason not to use part of as the inverse of has part?DrThneed (talk) 01:19, 24 February 2021 (UTC)
- (TNA) If eg OpenDomesday settlement ID (P3118) remains attached to settlements/parishes, then queries need consistency to know how to get from manor items to those items
- User:PKM: My first inclination is that a listed park or building should be part of (P361) the estate (and the inverse), as well as link to it via location (P276)
Ownership
edit- owned by (P127) (== "Lordship" ?) -> families? or individuals?
- with individuals, it is a many <-> many relation: one individual may own many manors, a manor may be owned by many individuals with start time (P580), end time (P582)
- (issue: ordering, tiresome if not done by machine; readbility if too many statements - another case for UI collapse box?)(or recommend userscript Wikidata:Tools/Enhance user interface#RearrangeValues)
- In my experience adding ownership items to houses, I was generally the first editor to do so, and I don't know if I ever saw more than around ten such statements on a property. So if this becomes an issue I don't think it is likely to do so very quickly.DrThneed (talk) 01:19, 24 February 2021 (UTC)
- I'm inclined to link ownership to a family, and the family should have has part to the notable members. I don;t think it's practical to expect an editor to make ownership statements on individual heads of the family over a period of 400 or 500 years. (Also, in my experience, many sources will tell us what family owned a property, but interpolating that as the head of the family in every generation might be a step too far.) - PKM (talk) 22:16, 24 February 2021 (UTC)
- Agree, neither practical nor desirable, I think, to have ownership by individual over such a period. I'm just showing my background I think, in that I've worked mostly on data drawn from wills, which is all about individual ownership of individual properties, and my instinct when thinking about how to model something is always to fall back to "but what does the source actually say?". So if the sources say a family owned a manor (as you've pointed out) then that is what I would want to put into WD. I have a suspicion that how we define a family is going to become interesting at some point but again I guess we should be guided by the sources about when a manor changes hands from one family to another.-DrThneed (talk) 00:46, 26 February 2021 (UTC)
- The flip-side is that a particular document will say that a particular person (who we may well have an item for) owned the manor at a particular time. So people documenting that particular person, or that particular document, may well want to record that at that time he was the one who owned the manor, and this is the document that establishes it. It's also the personal owners that we can connect between via genealogical properties
- @DrThneed, PKM: I agree that if somebody is going through eg the VCH entry for a manor, or one of the summary history files TNA has prepared, it would be unbelievably tedious to code in every owner by hand. That's for the far future, if ever. But if information like that happened to already exist in a readily assimilated digital form, then I think it would be something we would definitely add. Jheald (talk) 14:45, 26 February 2021 (UTC)
- @Jheald: Agreed. - PKM (talk) 20:43, 26 February 2021 (UTC)
- Agree, neither practical nor desirable, I think, to have ownership by individual over such a period. I'm just showing my background I think, in that I've worked mostly on data drawn from wills, which is all about individual ownership of individual properties, and my instinct when thinking about how to model something is always to fall back to "but what does the source actually say?". So if the sources say a family owned a manor (as you've pointed out) then that is what I would want to put into WD. I have a suspicion that how we define a family is going to become interesting at some point but again I guess we should be guided by the sources about when a manor changes hands from one family to another.-DrThneed (talk) 00:46, 26 February 2021 (UTC)
- (issue: ordering, tiresome if not done by machine; readbility if too many statements - another case for UI collapse box?)(or recommend userscript Wikidata:Tools/Enhance user interface#RearrangeValues)
- early ownership may be available via http://www.inquisitionspostmortem.ac.uk/
- families -- TNA does have a definition of a family -- perhaps owned by (P127) = <family> at preferred rank, owned by (P127) = <individual> at regular rank? This model might work for external sites like Commons infoboxes.
- or: should the statement always be the other way around for individuals? <X> owner of (P1830) <manor> ?
- I have always done ownership statements with inverse statements as well - I think there used to be a constraint flag if you didn't, which has now disappeared.
- with individuals, it is a many <-> many relation: one individual may own many manors, a manor may be owned by many individuals with start time (P580), end time (P582)
- manors sometimes formed a group for ownership, such as an honour (Q1525203) or English feudal barony (Q5378545)
Significant Events
edit- significant event (P793) - what sorts of significant events for manors ?
- This is where I have recorded things like significant renovations, extra wings, redesigns of facades, fires.DrThneed (talk) 01:19, 24 February 2021 (UTC)
- @DrThneed: Thanks! Good thoughts. Important to always remember for this page that items that are manor estate (Q2116450) are not buildings. So we should probably have a separate page (third-level structure!) to look at what are the things most commonly related to manors, and whether there are good style-guides we can put together for those too. Jheald (talk) 11:58, 24 February 2021 (UTC)
- I will try not keep forgetting that! Manors are not manor houses. There is a lot of confusion out there already, looking at Bhaktivedanta Manor (Q2025760) for instance, its label is manor, its description is human settlement, its instance is manor estate (Q2116450), its image is of a building, and it has heritage designation (P1435) of Grade II listed building. Urgh. DrThneed (talk) 21:59, 24 February 2021 (UTC)
- @DrThneed: Thanks! Good thoughts. Important to always remember for this page that items that are manor estate (Q2116450) are not buildings. So we should probably have a separate page (third-level structure!) to look at what are the things most commonly related to manors, and whether there are good style-guides we can put together for those too. Jheald (talk) 11:58, 24 February 2021 (UTC)
- We may also want to remind people that significant place (P7153) exists, which can point towards manors. Once we have the dataset in and start connecting it to things, it will probably be useful to add queries on the properties tab for how things link in to manors, as well as outbound properties; and we may want to write about some of them.
- This is where I have recorded things like significant renovations, extra wings, redesigns of facades, fires.DrThneed (talk) 01:19, 24 February 2021 (UTC)
Coordinates
edit- coordinate location (P625) ?? - according to TNA it may not be possible to place a manor much below the level of a parish. (A limited number straddle more than one parish, but most manors live within a single parish). So should we have P625s on them at all, or instead get users used to having to query for the (much better defined) P625s of some related item - eg the parish/settlement, or the principal house (which conceivably might not exist). Jheald (talk) 11:58, 24 February 2021 (UTC)
Documents and archives
edit- documents: to an archive, a manor is first and foremost an entity that creates documents. See eg the TNA MDR entry for Kittisford, linked below (F253847). So we need to think
- (i) what is/are the properties for the relationship between manors and documents;
- (ii) what documents will we want to create items for? My instinct on the latter is to go slow. Definitely create model items for one or two docs, and consider the classes that docs are likely to fall into. But I think not go out of our way to create items for manorial documents, unless there are particular things we want to say about those documents, or connections we want to record, such as might come out eg if they were looked at by one of the VR study teams (Are there manorial court documents they are planning to look at ? Or is the doc team focus all probate and county sessions?) Jheald (talk) 12:35, 24 February 2021 (UTC)
- maybe here also the place to note the info there is at TNA. First there's the capsule information that's at the top of the MDR entry. As well as that, Dr Jo thinks there may be another couple of fields not shown on the website, but that could be included as columns in a spreadsheet. Finally there's been an ongoing effort to make summary docs of the history of the manors, especially ownership history. These vary in format. Some are 'semi-structured data' Word docs that may have been turned into pdfs and may be available through the TNA site; others may be entirely on paper. (Tech not quite the same in the 1920s). He's going to try to identify a few examples with particularly rich history, from the ones that are online, to give us further ideas as to what information may be valuable to be able to record. Jheald (talk) 12:52, 24 February 2021 (UTC)
- re (i), I would suggest archives at (P485) on the item for the manor, with reference URL and title of the archive doc(s) is the reference. - PKM (talk)