Why would VTC as duplicate be appropriate for a story identification question? [duplicate]

Question

I keep seeing this over and over and over in the review queue.

Yes, story-idents can have identical answers, but this doesn't make them identical questions. Each person asking remembers different details, and would have little luck looking through older questions to answer their own. If they remember A and C about the story, but the other question describes B and D, this is then not a duplicate question.

Sometimes this even happens when the two answers are not identical (a short story that becomes a television episode or a novel that becomes a movie). This is even more absurd in my opinion.

Furthermore, this is the best resource anywhere on the internet to ask for story identification (at least within the narrow bounds of the genres we support). When we get all antsy and start closing questions in this manner, we're not only upsetting new users, we're making this SE a little worse at one of the things that it really shines at.

We do sometimes close non-identical questions when the answer from one suffices to answer the other... but those are non-story-identification questions and this makes it different. One could be expected to find those answers from the previous if they did some modest research, and so it need not be rehashed in a case like that. But the person who asks for identification just may not remember enough details for that research to be possible. Nor can we feasibly edit the extra details into the original question... to do that to the extreme extent would require every detail of the novel or movie in question.

I contend that if we couldn't merge the details from both versions of the question without leading to absurd scenarios where we replicate the entire work inside the question then this means that they are not duplicate questions at all.

I would implore you all to have a light touch here, but since that just doesn't seem feasible, can we at least come up with some clear policy on the issue?

Answer 1

The underlying concept, I think, is that closing as Duplicate isn't meant to be seen as censure. It just wants to mean "I have an answer for you, and it's over there". It's not a slap on the wrist for a user not doing his homework, because, as you say, with wildly differing questions, the asker will never be able to find the duplicate, even if it fully answers his question.

The problem here is that closing a question as duplicate does come across as censure. It says "your question is bad". Just like the "on hold" status was added to newly closed questions – SE's internal model of "closed" meant "question needs work", but came across as "question is terrible. You're terrible. Jerk". Here, too, the technical solution fails to take into account the people involved.

I doubt any technical solution will be developed, something along the lines of "importing" an answer from another question, linking them into one compound question. The most we can ask for, I think, is better wording.

Answer 2

While I am normally one to encourage the marking/closing of duplicates, I am in agreement that story-identifications should be the exception – and always vote to leave them open when those questions come up in my review queue.

While two story identification questions may seem to be duplicates due to similar details, that doesn't necessarily mean that the two OPs are looking for the same work.

The only time I believe story identifications are justifiably closed as dupes is if, and only if both answers have been accepted, and the works are the same. _{(But I'd still probably vote against it.)}

Answer 3

The rules for closing as duplicate are very simple (and stated explicitly in the actual closing wording posted after the close):

This question already has an answer here:...

In other words, the definition of a duplicate question is "the question's answer is already posted on another question".

It doesn't matter if the question itself is identical, similar, loosely close or wildly different. All that matters is the answer.

Also see:

• Why are questions considered duplicates although they ask different things?
• Changes to "close as duplicate" (part deux)

Answer 4

I don't believe you can have a single rule here for story-ID questions, and looking for one is an exercise in futility. An opinion of "it's black or white, we either accept them or we don't" isn't helpful, irrespective of which side you fall on; we should consider individual questions on a case-by-case basis and on their own merits.

To summarize the hypothetical example given in this question:

• A new question remembers A and C about the story, but...
• ...an already existing and answered question remembers B and D, so the question asserts:
• this is then not a duplicate question.

I dispute that assertion, because:

• The already existing answer may contain information points A, B, C and D in it.

Because of that, you cannot assert straight-up that it's not a duplicate under any circumstances. But you likewise cannot assert straight-up that it is a duplicate either, because the already existing answer may not contain A, B, C and D. you have to form a judgement based on the contents of that answer.

One could extrapolate from here:

• If we have two questions, one remembers A, B and C, the other remembers A, B and D, are they duplicates?
• If we have two questions, one remembers A, B and C, the other remembers A, D and E, are they duplicates?
• Etc?

A blanket exception for story-ID questions is obviously starting to look silly in those two scenarios (particularly the first).

Answer 5

We do sometimes close non-identical questions when the answer from one suffices to answer the other... but those are non-story-identification questions and this makes it different.

No, it doesn't make it different. The network-wide policy of Stack Exchange is that if question X has an answer that also answers question Y, Y is a duplicate of X. The type of question is irrelevant to that definition. The only difference with story identification questions is that we don't definitively know they're duplicates until both have had answers accepted.

One could be expected to find those answers from the previous if they did some modest research, and so it need not be rehashed in a case like that. But the person who asks for identification just may not remember enough details for that research to be possible.

So what? Closing as a duplicate isn't about saying "You didn't do enough research." It's about saying "This information is already available, but it might not be (easily) findable with the tags and keywords you've used, so we're going to use your question as a signpost to the answer to make it easier for future users to find."

The idea is that, regardless of the things you remember or the terminology you use when searching, you're able to get access to the very best information available on the site. To do that we need questions to be marked as duplicates. Having related things without links explicitly stating they're related makes the site worse for future users, not better.

Nor can we feasibly edit the extra details into the original question... to do that to the extreme extent would require every detail of the novel or movie in question.

The entire point of closing questions as duplicates of each other is so we don't have to do that. If question X addresses A and C, and question Y addresses B and D, then by virtue of linking them together we've just covered A, B, C, and D without requiring a single edit.