Wikidata:Property proposal/latest start date
latest start date[edit]
Originally proposed at Wikidata:Property proposal/Generic
| Description | (qualifier) latest date on which the statement could have started to be true |
|---|---|
| Data type | Point in time |
| Example 1 | Henry Percy, 3rd Earl of Northumberland (Q2521911)spouse (P26)Eleanor de Poynings, Baroness de Poynings (Q75249624) → latest start date = 25 June 1435 |
| Example 2 | John de Vere, 13th Earl of Oxford (Q1332495)spouse (P26)Elizabeth Scrope (Q75252513) → earliest date (P1319) = 28 November 1508; latest start date = 10 April 1509 |
| Example 3 | Ælfstan (Q3880977)position held (P39)Bishop of London (Q1587771) → earliest date (P1319) = 959; latest start date = 964 |
| Planned use | to represent imprecise marriage dates as presented in eg The Peerage (Q21401824) |
| See also | start time (P580), earliest date (P1319), latest date (P1326), Wikidata:Property proposal/earliest end date |
Motivation[edit]
Suppose a source says that a marriage occurred before some date -- eg [1] "before 25 June 1435", or [2] "between 28 November 1508 and 10 April 1509". How to represent this?
In the past I have occasionally used latest date (P1326) for this, on the basis that 'of course' it must be the wedding day that such a qualifier would describe. But on further thought I feel that's not really right. In the context of a marriage, P1326 logically should indicate the latest date that the couple might still have been married. (Even if that is something that maybe we would be unlikely to often record). So a new qualifier, specifically for this case, would seem to be needed.
The qualifier would also be applicable to other statements that can take start time (P580)/end time (P582) pairs, eg noble title (P97), position held (P39), etc, etc. (sample) Jheald (talk) 21:04, 10 August 2020 (UTC)
Discussion[edit]
- Proposed. Jheald (talk) 21:04, 10 August 2020 (UTC)
Support there are definitely issues with the way we handle dates that could have "uncertainty" ranges right now, I think this is a reasonably approach. ArthurPSmith (talk) 18:16, 11 August 2020 (UTC)- Support Gamaliel (talk) 15:20, 13 August 2020 (UTC)
- Support. This qualifier will be beneficial and provides a good alternative to using the most specific date precision applicable in these situations. For example, the decade 1500s is a viable way to represent "between 28 November 1508 and 10 April 1509" but it isn't really accurate enough to be satisfactory. The Extended Date/Time Format (EDTF) Specification can deal with uncertainty and approximations. Simon Cobb (User:Sic19 ; talk page) 19:03, 18 August 2020 (UTC)
