Open main menu

Wikidata:Property proposal/date of first attestation

date of first attestationEdit

Originally proposed at Wikidata:Property proposal/Lexemes

Descriptiondate of the earliest use of the word, given by an etymological document or, where appropriate, by the oldest document found using it
Data typePoint in time
Exampleordinateur (fr) → 16 April 1955
See alsotime of discovery or invention (P575), Wikidata:Property proposal/place of first attestation


To indicate the etymology of a word. Tubezlob (🙋) 12:13, 24 April 2018 (UTC)


  •   Support David (talk) 15:29, 24 April 2018 (UTC)
  •   Support, good idea ; just not sure if it's for Lexeme or Form. Cdlt, VIGNERON (talk) 17:41, 24 April 2018 (UTC)
  • @VIGNERON: It should be Lexemes (I put the proposal in the Lexemes' section but I forgot to indicate this information in the "domain" field). Tubezlob (🙋) 17:45, 24 April 2018 (UTC)
  •   Oppose there is unlikely a first attestation but some earliest sources so better directly reference these sources instead of their date. -- JakobVoss (talk)
  • @JakobVoss: For example, the word "palooka" in English is used since the 1920s. We don't have any first document, this is the case for most words. We use the first document only if we don't have other information. Tubezlob (🙋) 10:33, 25 April 2018 (UTC)
Who says it is used since the 1920s? The Wiktionary entry mentions explicit sources for when an by whom the term was popularized. The other example "ordinateur" cites that Jacques Perret has proposed the term. It looks the proposed property is too vague to be reliable but something like "proposed or popularized by" could better be filled with sourced statements. -- JakobVoss (talk) 15:50, 26 April 2018 (UTC)
@JakobVoss: Maybe the label of the property is not very accurate. I propose instead "date of appearance".
Yes I agree with you, it is possible to have several properties that are more precise to describe etymology.
  • date of appearance
  • place of appearance
  • proposed or popularized by (should be two properties?) with date and place for qualifiers
  • document of earliest known attestation with date and place for qualifiers
  • derived from (with qualifiers to be more precise)
Are you OK with that? Tubezlob (🙋) 19:44, 26 April 2018 (UTC)
I's hard to say without more examples. I'd rather postpone these properties and start with more obvious properties until we have enough actual Lexemes to work with. There are too many ways to model the information, e.g. something like used by (P1535) with qualifiers could also make sense. -- JakobVoss (talk) 15:18, 27 April 2018 (UTC)
@JakobVoss: What are for you "more obvious properties"? Tubezlob (🙋) 16:08, 30 April 2018 (UTC)
  •   Oppose use inception (P571). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:02, 23 May 2018 (UTC)
  •   Oppose If there actually an document that's the oldest source we should link the document directly. In cases where we don't have such a document inception (P571) should work as proposed by Andy. ChristianKl❫ 08:21, 12 June 2018 (UTC)
  •   Comment Isn't this the same as time of earliest written record (P1249)?
    --- Jura 04:52, 26 July 2018 (UTC)
  • @Tubezlob, ديفيد عادل وهبة خليل 2, VIGNERON: do you think this is still needed? Otherwise, please withdraw.
    --- Jura 12:09, 18 August 2018 (UTC)
    • Not sure, but attested in (P5323) with point in time (P585) (and maybe other qualifiers) seems to be equivalent. @Tubezlob: what do you think? Cdlt, VIGNERON (talk) 13:12, 18 August 2018 (UTC)
      • @Jura1, VIGNERON: What can we do if we don't have a document of first attestation without this property? And attested in (P5323) doesn't say it's the first appearance. Tubezlob (🙋) 07:48, 20 August 2018 (UTC)
        • @Tubezlob: by definition an attestation is in a document, isn't it? Do you have an example where it can be attested outside a document? And true, attested in (P5323) is for all attestations not just the first, but with a qualifier you can say it the first. Cdlt, VIGNERON (talk) 07:57, 20 August 2018 (UTC)
          • @VIGNERON: In dictionaries, they often cite just a century, or a part of a century, or a decade, etc. For example for the word "livre" in the TLFi there is just "Ca 1100". But of course your method can work, for "homosexuel" for example there is a first document: Annales médico-psychol., t. 14, p. 331 Tubezlob (🙋) 08:11, 20 August 2018 (UTC)
            • @Tubezlob: look more carefully, the TLFi indicate the document « Roland, éd. J. Bédier, 610 », it's The Song of Roland (Q185427) edited by Joseph Bédier (Q380842), I just created a specific item for this edition Q56193560   (and we have it on Wikisource so we can find exact citation too). The document is not always easy to find but I think it's worth the effort to find it to be more precise and more reliable. Cdlt, VIGNERON (talk) 08:38, 20 August 2018 (UTC)
              • @VIGNERON: Well done! So OK, let's go with your solution. Anyway, if the use of this property is felt, we will create it in due course. Tubezlob (🙋) 10:08, 20 August 2018 (UTC)

@ديفيد عادل وهبة خليل 2, VIGNERON, JakobVoss, Pigsonthewing, ChristianKl, Jura1:   Withdrawn. Tubezlob (🙋) 10:12, 20 August 2018 (UTC)