Wikidata:Property proposal/exonerated of

From Wikidata
Jump to navigation Jump to search

exonerated of[edit]

Originally proposed at Wikidata:Property proposal/Person

   Done: exonerated of (P7781) (Talk and documentation)
Descriptioncrime of which the subject was found not to be guilty after having previously been found guilty by a court for that crime.
Representsexoneration (Q5420623)
Data typeItem
Domainhuman (Q5)
Allowed valuescrime (Q83267)
Example 1Clarence Elkins (Q5126508)murder (Q132821)
Example 2Clarence Elkins (Q5126508)rape (Q47092)
Example 3Kenny Richey (Q6391188)murder (Q132821)
Sourcehttps://en.wikipedia.org/wiki/List_of_wrongful_convictions_in_the_United_States
Planned useAdd to those who have been convicted of a crime and then exonerated; citation required
Number of IDs in source20+
Expected completenesseventually complete (Q21873974)
Robot and gadget jobsbots could identify candidates, but I don't think they should add the data

Motivation[edit]

I've been doing some research on people who have been exonerated through the efforts of the Innocence Project, such as Clarence Elkins. His entry lists him as having been "convicted of" murder, but there is no structured data to show that he was exonerated of the crime. That lack of information is both incomplete and compounds the significant injustice of his experience. (The information about his exoneration is represented textually in the description, but is not reflected in a machine-readable way.) He and the many other people whose convictions have been overturned deserve to have their innocence reflected in our data. Kenirwin (talk) 03:00, 4 August 2019 (UTC)[reply]

Discussion[edit]

I'm proposing this as a property to be applied to a human item, but I could also imagine this data being represented differently: as a qualifier to a "convicted of" crime, example: convicted of (P1399) -> sea piracy (Q45393) + qualifier: "exonerated" -> point in time (Q186408). If we have to choose between the two data models, I prefer the main proposal because it allows for multiple attributes on the exoneration and to have its own citations. But I wonder if it would also make sense to have an "exonerated" or "conviction overturned" qualifier too. What do you think? -Kenirwin (talk) 03:08, 4 August 2019 (UTC)[reply]

  •  Support David (talk) 06:41, 4 August 2019 (UTC)[reply]
  •  Comment I prefer the second model, because it makes clear which murder conviction was overturned -- someone may have more than one conviction of the same kind, so it's good to tie the exoneration to the particular offence. But, as you note, it raises a difficulty of not being able to put qualifiers on a qualifier -- eg which court considered the appeal. Should it also be possible to record appeals that were unsuccessful? We should also have a way of recording pardons. It would be good if we could come up with a more generic way to record all sorts of revisions post-conviction. (including eg sentences appealed and increased). Jheald (talk) 20:10, 5 August 2019 (UTC)[reply]
  •  Comment Should this be two different properties, one intended as a qualifier and one intended as an item property? As a qualifier I think it would just have a date; as an item property, I imagine there are a lot of properties it might take. I agree with Jheald that it would be good to do a more thorough examination of how we handle all sorts of outcomes. How does the Wikidata community go about such things? -Kenirwin (talk) 21:04, 5 August 2019 (UTC)[reply]
  •  Comment In any form I support this property's existence. I think it makes sense to model out other rulings as well. I haven't found a data model for this yet. It should also carry a citation required constraint. For now, I favor it having parallel usage with "convicted of"...i.e. it would be its own property where editors could specify what the person was exonerated of as a qualifier. Wskent (talk) 22:36, 5 August 2019 (UTC)[reply]
  •  Support creation of exonerated of. In response to questions about time, etc., we can do Kenny Richey (Q6391188)<exonerated of>murder (Q132821)statement is subject of (P805)<exoneration item>, and the exoneration item has instance of (P31)exoneration (Q5420623) and then can have its own point in time (P585) for when, etc. Its better to store the associated data in a separate item, in my opinion. --DannyS712 (talk) 05:35, 7 August 2019 (UTC)[reply]
  • I'm not sure how exoneration works but is a person who's exonerated still convicted of the crime? I would expect that they aren't and thus it would be proper in our data-model to deprecitate the claim. ChristianKl12:03, 7 August 2019 (UTC)[reply]
*@ChristianKl: - that's a really good point.
* Support In theory, once someone has been exonerated, the conviction should be expunged from the person's record. BUT, I don't think an exoneration is needed for a conviction to be expunged--so I support having a property to make that distinction. To me the deprecation of a conviction claim just reflects expungement, not necessarily exoneration. Gtsulab (talk) 21:56, 15 October 2019 (UTC)[reply]