6
$\begingroup$

Let's imagine the following rather common scenario:

  • I find, on another site (not PSE), a fun puzzle and its solution
  • the puzzle and solution are clearly linked, so there's no way to link to one but conceal the other
  • the puzzle is interesting enough that I want to share it on PSE.

There seem to be three possibilities here, none of which is really desirable.

  1. I post the puzzle here on PSE without including a link to the site where I found it. This is unacceptable as it would violate the plagiarism policy.

  2. I post the puzzle here on PSE with a link to the source where I found it. This means anyone trying to solve it here can find the solution online with two simple clicks. Talk of an "honour system" being absolutely unenforceable and therefore so much pie in the sky as far as I'm concerned, this would encourage people to game the system for rep.

  3. I don't post the puzzle on PSE at all. Then PSE loses out on a great puzzle.


I would like to suggest the following fourth possibility:

  1. I post the puzzle here without a link initially. Once it's been solved and the race to answer it and get the pretty green check is no longer on, I edit the question to include attribution.

Would this be acceptable according to site policy on plagiarism?


Note: throughout this discussion, let's assume that copyright violation isn't an issue, e.g. that the site being copied from has the same CC-by-SA licensing as SE. If this were an issue, then there's no point having this discussion at all, because no community-crafted PSE policy can nullify copyright law. So I'm assuming that copying with attribution would be fine, and that the only possible issue at hand is that of plagiarism.

$\endgroup$
  • 4
    $\begingroup$ I realise it's not directly what you're talking about, but it is related, so it bears repeating: simply adding an attribution doesn't magically absolve you - you still can't copy other peoples work if you don't have permission (either directly or via some licence such as CC-BY-SA). If it's a "chestnut", or if you're quoting in order to discuss, or if you're taking a core idea and reworking it into a unique puzzle, that's different, and where attribution comes in. (IANAL etc) $\endgroup$ – Alconja Sep 30 '16 at 1:39
  • 1
    $\begingroup$ You present two alternatives: a link to the source, and no attribution. A possible third, if the author is known: credit the author, but don't cite the source initially. $\endgroup$ – Rosie F Sep 30 '16 at 6:31
  • $\begingroup$ @Alconja Good point, and one worth bearing in mind in all these plagiarism discussions. However, PSE's plagiarism policy says that "posts containing attribution are not plagiarized". I assume this means that they won't be deleted on sight for being plagiarised according to PSE policy, even though they will still have to be deleted if the copyright holder files a DMCA takedown notice. $\endgroup$ – Rand al'Thor Sep 30 '16 at 13:21
  • $\begingroup$ (Personally I never copy other people's puzzles directly, always using just the core idea and rewriting it. But probably not everyone works the same way.) $\endgroup$ – Rand al'Thor Sep 30 '16 at 13:23
  • $\begingroup$ @RosieF That may be possible in some cases, but often crediting the author will be enough to find the source anyway. $\endgroup$ – Rand al'Thor Sep 30 '16 at 13:24
  • $\begingroup$ @randal'thor - of course it's not plagiarism if it has attribution (by definition), but plagiarism and copyright violation are two different things. Also note that the actual plagiarism policy describes things in terms much more in line with "fair use" exceptions to copyright law and not in terms of copying whole puzzles (with or without attribution). $\endgroup$ – Alconja Sep 30 '16 at 13:46
7
$\begingroup$

This should be acceptable ...

As far as I can see, it's the only way to combine the two principles of turning PSE into a great repository of puzzles (both original ones and those copied from elsewhere with suitable attribution) and encouraging people to solve puzzles based on their own ingenuity (rather than simply finding a solution online and using that to write their own).

In the long run, there still won't be any unattributed puzzles lying around: given how good people here are at solving, the puzzle will be solved and the source edited in within a day or two of the original post.

... but all such puzzles should be semi-attributed from the beginning.

I propose a rule that all puzzles posted in this way should be semi-attributed, or 'anonymously' attributed. In lieu of an actual link to the source site, include a short statement to the effect of "this isn't my own work; I found it elsewhere, and will edit in a source when it's been solved". That way, you avoid accusations of passing a puzzle off as your own when it isn't, which (as far as I know) is the main problem that the plagiarism policy exists to address.

Puzzles which are copied from elsewhere but which are posted without a link or an 'anonymous' attribution can be flagged and deleted under the plagiarism rules as per usual.


I can already see some of the possible counterarguments, so let me address them here.

But Rand, what if the puzzle never gets solved?

Unlikely as this is (in my opinion), it has been known for puzzles to go unsolved for months. If the 'anonymous' attribution isn't enough, I propose an amendment to the above rule: if the puzzle goes unsolved for a certain time period (a week?), the link should be edited in. Thus we make 100% sure that no unattributed puzzles stick around on the site, whatever happens. And it will only be a few rare edge cases which need to have the link edited in before being solved.

But Rand, if someone copies the solution from the link, their answer would be liable for deletion under the plagiarism rules!

Not necessarily. While there are many ways of presenting a puzzle, there's often only one way to solve it. It would be perfectly possible for an unscrupulous solver to look at the solution on the other site and get enough information to have an unfair advantage while not copying enough of that solution for their answer to count as plagiarism.

  • For a riddle or lateral-thinking puzzle, they could simply see the solution and then write their own explanation for how it fits each line.

  • For a mathematical or logic puzzle, they could examine the method of solution and write their own implementation of it.

This is why plagiarism of answers is much harder to detect on Puzzling. There's usually a single right answer, and knowledge of what it is is enough to get a massive advantage; the actual writeup of the answer could differ hugely from that in the source it was taken from. That's why the issue of unscrupulous solvers needs to be dealt with by careful formulation of the question rather than by simply deleting their answers, since those answers are completely indistinguishable from answers provided in good faith by people who worked out the solution by themselves.

$\endgroup$
  • 5
    $\begingroup$ If the puzzle is on the Internet, finding the original puzzle would be super easy. That undermines the semi-attributed point since people who only want rep could find the answer. It's nothing a little renaming wouldn't fix but some puzzle are dependent on their wording so it could pose a problem. $\endgroup$ – Areeb Sep 30 '16 at 20:03
  • $\begingroup$ @Areeb makes a good point. If I can copypaste the puzzle into Google search, then attribution is the last of the puzzle asker's worries, however, attribution must always be provided. $\endgroup$ – Buffer Over Read Oct 6 '16 at 22:34
  • $\begingroup$ @TheBitByte The main scenario I've been imagining is where the puzzle is paraphrased, or only the core idea taken, rather than a direct copy-paste. $\endgroup$ – Rand al'Thor Oct 6 '16 at 22:39
  • $\begingroup$ Well, if you know you way around Google search, you might still be able to Google that core idea, but this depends on the specific puzzle in question. $\endgroup$ – Buffer Over Read Oct 6 '16 at 22:52

You must log in to answer this question.

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