3

This question is motivated by the following three questions:

Question A is the oldest, with a detailed description and a confirmed answer. Question B has a similar description to A, although less detailed, and the same answer has been suggested but not confirmed. Question C has a confirmed answer which is the same as A, but the description is totally different, with different details of the same book being remembered.

Currently, question B has three votes to close it as a duplicate of A, while question C has none. This inspired me to post a meta and start a general (i.e. not just about these particular questions!) discussion about closing identification questions as duplicates.

Should we create a clear policy on when to close ID questions as duplicates?


For the record, so far we have four questions with the tag that are closed as duplicates:

  1. Book about a boy surviving in a forest with animal friends Answer given but not confirmed; dupe target has an accepted answer. Both questions have similar descriptions.

  2. Trilogy of books involving a man's journey through ritual magic Question was inspired by the dupe target; the OP of the duplicate answered both questions after realising they were about the same story, and closed his own question himself.

  3. Book about a boy who leaves his village and goes to a large city. The cover shows two men having a discussion at the foot of Bruegel's Tower of Babel The OP posted the same question on two different SE sites; the SFF version was migrated here and closed as a duplicate of the original Lit version.

  4. Short story about a young boy living in South Africa during Apartheid Answer given but not confirmed; dupe target (which is actually newer) has an accepted answer. Both questions have similar descriptions and the answers seem clearly correct.

5
  • For what it's worth, the sister site SFF has a policy that ID questions are closed as duplicates if and only if both questions are about the same story and this is confirmed by the OP (via acceptance or comment), with similarity of description playing no role whatsoever. (An exception is made for questions like case 3 in this post, where the same OP posts the same question twice - those are always closed.) This is just for context and not saying we should necessarily do the same thing on Lit.
    – Rand al'Thor Mod
    Sep 11 at 12:25
  • Related: literature.meta.stackexchange.com/q/1160/58
    – Mithical Mod
    Sep 11 at 12:44
  • @Mithical Ah, I'd forgotten about that post. But I think this one can stand independently; four years and an actual specific not-100%-clear case to talk about have changed the situation a bit away from "no need to put a rule in place when there's no problem to fix".
    – Rand al'Thor Mod
    Sep 11 at 12:51
  • @Randal'Thor in the event no suggestion gains positive traction, will the moderation team set a policy or will we just float on as we have been?
    – Skooba
    Sep 21 at 15:15
  • 1
    @Skooba I haven't talked about that with the other mods, but I guess it's not a massively burning issue, so we can float on and let close/reopen votes on the main site decide the community's will until this becomes more pressing.
    – Rand al'Thor Mod
    Sep 21 at 17:19

3 Answers 3

2

In an identification question, the asker presents some remembered details about a work of literature, and the answerer proposes an identification for the work, and (ideally) presents evidence that that the work matches the remembered details (allowing for ordinary errors of recall).

This means that an identification question should be closed as a duplicate only if the remembered details are substantially overlapping (again, allowing for errors of recall), so that an answer to one equally answers the other.

In the examples given above, question C, which asks for the source of a famous quotation with a high-sounding tone, possibly containing the words “a fine thing”, is not answered by answer A, which identifies a work in which “a lawyer sacrifices his life for his activist friend by taking his place in prison”. Even though the work is the same, the questions have no overlap.

However, question B, which asks for a work in which “a journalist takes the place of an inmate on death row” does seem to be answered by answer A, as the misremembering of a lawyer as a journalist is well within the usual scope of errors of recall. (Compare this question, in which a prison chaplain was also misremembered as a journalist.) So it is reasonable (but not necessary) to close this as a duplicate without waiting for the OP to return to confirm the identification.

2
  • That this is substantially the same as verbose's answer, except that I've tried to be a bit clearer about what I would like the policy to be. Note that the policy allows but doesn't require questions to be closed as duplicates when the details are substantially overlapping. Sep 14 at 18:06
  • 1
    Upvoted, and I agree 100% with this answer. This is the exact opposite of the policy on SFF, which I personally find absolutely maddening. Sep 26 at 17:54
1

I think that if the details furnished are different enough such that when someone who does not know the sought-after text does not realize upon reading the questions that they in fact are seeking the same work, then it's fine to leave both questions open.

If the details suggest that the question refers to the same work as another question which has an accepted answer, then the question with no accepted answer yet should be closed, with a comment to the OP that if that book isn't the correct one, the OP can edit the question to specify some differences (e.g., "that book has a 40 year old female protagonist, is in Tamil, and is set in France, the one I read had a 15 year old male protagonist, was in Hungarian, and was set in Timor-Leste"). That would cause the question to be reopened when the OP takes action.

That leaves the tricky situation of there being two questions, both open, with substantial similarities in the details of the book. I'd argue that the later question should be closed with a comment to roughly the same effect as above: "is this the same book?" If the later question has some details that are not in the earlier question, then the OP of the earlier question can be asked to verify and incorporate those details, perhaps?

My understanding is that rep gained by votes on closed questions/answers isn't lost; all that happens is that existing answers can't be accepted, and no new answers can be added. Is that correct? If so, closing the duplicates doesn't cause any harm.

2
  • To be clear, your proposed policy is to close based on matching descriptions only? So questions A and C from the OP would be left open, B closed as a dupe of A, and we ignore the presence/absence or confirmation of answers when voting to close? It's a bit subjective, but hopefully workable, especially as this isn't a large-volume issue yet. (And to answer your final paragraph - yes, closure doesn't affect rep, and dupe closure doesn't even count as a "bad" closure for some purposes like migration rejection.)
    – Rand al'Thor Mod
    Sep 12 at 14:26
  • 1
    Btw, I think age shouldn't matter in duplicate closure. If two questions are similar but the later Q&A are better, it's fine to close the earlier one as a duplicate of the later one. That actually happened in case 4 linked in the OP.
    – Rand al'Thor Mod
    Sep 12 at 14:28
-2

Identification questions should only be closed as duplicates once there is confirmation from the asking user that a proper identification has been made. This can come in a couple ways:

  1. Q1 and Q2 have the same accepted answer.
  2. The asking user leaves a comment confirming it is what they were looking for. This can be on a posted answer or as a reply to a comment suggesting the answer has been given elsewhere.
  3. Exception, the same user creates another question looking for the same identification.

I understand that this can leave questions open that appear to be looking for the same work, especially if both identification questions are well written and therefore have many of the same details. However, the only way we can know with 100% certainty is with confirmation from the asker.

As to which question should be left open and which should be closed... Typically the oldest question is the one that remains open as this there is no contention. Sometimes a question or answer with greater detail can usurp an older post but this can be subjective.

2
  • This basically parrots Rand's comment but I agree with that assessment.
    – Skooba
    Sep 12 at 15:53
  • Not parroting my comment, which was only a comment on what another site does, not a proposal for what we should do here :-)
    – Rand al'Thor Mod
    Sep 12 at 18:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .