23

Running through edit reviews today, I noticed that one user showed up a number of times, adding text to blank tag wikis. Unfortunately, this user was not authoring the wikis, but instead simply copying the text from various sources, such as Wikipedia and/or company web sites (*).

I have two concerns:

  1. As of my review of most of the edits, they had not yet been rejected. One other person had also voted to reject, but another had voted to approve. Less-careful editors are unlikely to pick up on the fact that the content was copied, or may not even realize that doing so is harmful and inappropriate.
  2. The user who is making these edits is apparently targeting tags that they run across and which have no wiki content, and simply doing a search online to find content on other sites to copy to the tag wiki. It seems likely that the person is unaware that this is not useful, not allowed, and potentially harmful.

To elaborate on the latter, and "potentially harmful" in particular: I believe previous discussions in the linked posts (see below) sum it up well, but the short version is: unattributed copying is prohibited on Stack Overflow (so, in and of itself is considered wrong), but much more importantly the copied content is almost never actually useful, instead tending to inhibit others from providing good content, so we wind up with crappy wiki content, from the "displacement effect".


These concerns lead to two questions:

  1. Assuming that a bad edit is made to a tag, is there a way to flag the tag for moderator review? I know that one can just re-edit the tag to undo the edit, but just as editors of regular posts may object and rollback an edit, so too is there a potential for a dispute about tag wiki content. But I didn't see any mechanism for me as a user to get a moderator involved.
  2. Does a user who repeatedly makes edits to tag wikis which are rejected suffer any consequence? Do they receive any sort of notification or feedback that what they are doing fails to meet the standards of the site?


On the second question, I did find this related question: Mass updating of tag wikis with content copied. But while I agree that the proposed changes to the badge requirements could be beneficial, it doesn't really address either of the above.

Other related questions include:
Wiki edit reviewers are not catching plagiarism
The tag wiki suggested edit review mechanism encourages low quality wiki content


(*) (In several instances, the tag itself was of questionable value, having been added recently to just a single post and being a company tag. But there did not appear to me to be a connection between the person adding the copied wiki content and the person who added the company tag, and the fact that most of these tags are just company tags is unrelated to the broader questions above.)

9
  • 1
    You can't flag a tag. You need to find one of his posts and raise a custom flag describing what's wrong and listing the tags.
    – mmking
    Commented Nov 2, 2015 at 3:23
  • Where's Bob?
    – theB
    Commented Nov 2, 2015 at 12:31
  • Related to first concern meta.stackexchange.com/q/238339/213575, the second one is a matter of education.
    – Braiam
    Commented Nov 2, 2015 at 14:35
  • 2
    @pnuts The case for common sense is otherwise: do we really need a tag for everything single thing in the world?
    – Lundin
    Commented Nov 2, 2015 at 14:57
  • @Lundin I presume pnuts was trying to get Hans attention... on another people post(?)
    – Braiam
    Commented Nov 2, 2015 at 15:49
  • 4
    There's a rejection reason for "copied content". SO tags are not supposed to be copies of what otherwise exists. Else, what's the point?
    – Bill Woodger
    Commented Nov 2, 2015 at 16:04
  • 1
    @pnuts: I have nothing against "common sense", if it truly is so. As for elaboration, I believe previous discussions in the linked posts sum it up well, but the short version is: unattributed copying is prohibited on SO (so, in and of itself is considered wrong), and the copied content is almost never actually useful, but tends to inhibit others from providing good content, so we wind up with crappy wiki content, from the "displacement effect".
    – Peter Duniho
    Commented Nov 2, 2015 at 16:12
  • 3
    @pnuts: I think you and Hans are both wrong to be focusing on the question of the word "plagiarism". I never used the word myself, except to tag my question, and I did so primarily because that tag is in fact where people will find related discussion. It is derailing the real point and questions to waste time debating the exact definition and legality of "plagiarism".
    – Peter Duniho
    Commented Nov 2, 2015 at 16:29
  • I must be in the 20% here, but I completely disagree, this is not plagiarism. Our goal on stackoverflow is helping to answer questions. We aren't writing works of English, or reporting sport stories, or writing fiction. We deal with logic, how does one accomplish this? Yes links to original authors / plugin pages are a must. However copy and paste should be completely fine.
    – Leon Gaban
    Commented Feb 20, 2016 at 14:31

3 Answers 3

Reset to default
28

Well, now I'm depressed at the atrocious state of tag wiki understanding, not only among prospective editors (which is only to be expected), nor yet among edit reviewers (which is a sad fact of the SE queue on SO), but even among illustrious users active on Meta! Hans' answer has +18/-9. It completely misses the point.

The point is almost never "is it legal to copy this text?" Wikipedia has the same license we do, specifically chosen for reusability. And no one is going to complain if you take their ad copy and republish it on an unaffiliated website for free.

Nor is the point usually about the ethics of plagiarism and attribution. Sure, a lot of these lousy copy-paste hackjobs miss that, but it's not like it's difficult to throw in a link to a website you find via search. (Actually digging up the original might be harder, but that's Somebody Else's Problem, right?) Making sure the wiki conforms to the highest academic ideals of generally independent phrasing and such-like is harder, but SO isn't a scholarly journal, so it doesn't exactly matter much.

No. The point is practical: a copy-paste from a Wikipedia article is useless garbage on SO in tag wikis. We don't want it. We wouldn't want it if it was given to us gift-wrapped. Why is that? Because Wikipedia is describing things from a completely different perspective! WP is about explaining things to an audience that doesn't yet know how X works. Tag wikis and usage guidance are about explaining whether this is even about X … and usually, those who tag with X are well aware of the Wikipedia-level explanation. We don't need an explanation of Java's design philosophy and the fields it's commonly used in; we need a quick note that it's not JavaScript.

And, of course, pulling in some marketing department's glowing screeds about how X slices, dices, and Judy-array-fries is even worse. We're not trying to sell someone on what technology to use, we're just letting them know which tag they should enter for what they're already using.

This is the reason for the canned rejection reason that applies here:

This edit copies a significant amount of content from an external source. Generic descriptions such as encyclopedia articles and ad copy do not provide useful guidance; try creating something useful to this community specifically, and be sure to attribute the original author. See: How to reference material written by others.

(That's right, there's a rejection reason coded just for this very scenario!)

So please, stop mixing up these three crucially different reasons for objecting to copy-paste jobs. They are not the same at all.

8
  • 5
    Well, I (obviously) agree 100% with the above. And I very much appreciate the emphatic, detailed, and well-expressed support for the point. But...do you have any answer for either of the questions I actually asked? :)
    – Peter Duniho
    Commented Nov 3, 2015 at 3:27
  • @PeterDuniho: Sadly I do not know of any way to deal with this other than reporting it on Meta or perhaps via a custom flag.
    – Nathan Tuggy
    Commented Nov 3, 2015 at 3:35
  • 1
    Agree with the overall message, but as I understand it, tag wikis (unlike excerpts) aren't meant to be for "explaining whether this is even about X". (Quite what they are for is a little unclear to me; in reality the only ones that aren't just plagiarised content seem to be giant lists of links with no obvious practical purpose.)
    – Mark Amery
    Commented Nov 3, 2015 at 14:40
  • @MarkAmery: I assume full wikis are for explaining how to get from "I am using X" to "I am using X with basic competence", but I'm not sure they do a great job of that.
    – Nathan Tuggy
    Commented Nov 3, 2015 at 17:08
  • 1
    I just reviewed four edits. Bracketing (although unbalanced) an editor who'd decided to past some sample code into someone's answer(!) were three simply-pasted-it wiki edits, so far happily accepted by six different reviewers. stackoverflow.com/review/suggested-edits/10091652 stackoverflow.com/review/suggested-edits/10091113 stackoverflow.com/review/suggested-edits/10091964
    – Bill Woodger
    Commented Nov 4, 2015 at 15:45
  • agree on "tag wiki" and "wikipedia article" have different purposes and the content should reflect that. Could you add a link that explicitly says why tag wikis exist and what they should contain?
    – jfs
    Commented Nov 4, 2015 at 21:55
  • 1
    @J.F.Sebastian: The closest is probably this, which is linked to from the canned rejection reason.
    – Nathan Tuggy
    Commented Nov 4, 2015 at 22:10
  • Let us continue this discussion in chat.
    – Nathan Tuggy
    Commented Nov 5, 2015 at 0:07
3

When reviewing new posts, users sometimes see a warning along the lines of:

Pay attention! This post may contain spam.

Could we not have something similar for Tag Wikis? Say, when an edit is suggested snippets of its content were tossed at Google, and they were found to exist elsewhere, flag the suggestion for special attention. Then, when presenting it in the suggested edits queue, warn:

Pay attention! This suggestion may contain a significant amount of content from an external source. Generic descriptions such as encyclopedia articles and ad copy do not provide useful guidance, and should be rejected.

What do you know? It's been suggested before.

-1

I personally find accusations of plagiarism to be highly overrated in the specific case of tag wikis. It does require malicious mis-representation first and that's just not there. The tag wiki does not carry the name of the author, you have to dig through the edit history to find the name. Whose only role is accountability, not representation.

The only remaining concern is of copyright violation. Again a weak signal, tag wikis are always excerpts. But a small part of an original Wikipedia article or a company's web site. The kind that are explicitly allowed by the fair-use clause in USA copyright law. If a company is not happy with it then they always have DMCA to fix it. An unlikely outcome, they like to be represented in their own words.

If this kind of copy/pasta personally offends you then just fix it.

8
  • 3
    Yes, it offends us, greatly.
    – Braiam
    Commented Nov 2, 2015 at 14:31
  • 14
    It is not a matter of personal preference. If SO has a rule against plagiarism, then reviewers should follow that rule, period. Reviewers shouldn't be "artists" whose review outcome is determined by opinion, mood or the current location of the planets.
    – Lundin
    Commented Nov 2, 2015 at 14:40
  • 6
    It is not plagiarism. So whatever stone tablet carried down the mountain that says what should be done does not apply.
    – Hans Passant
    Commented Nov 2, 2015 at 15:01
  • 2
    Then again, if SO has a rule against it and the community disagrees with that rule, then perhaps the rule should change, not the community. Reviewers shouldn't be "artists" - I disagree. To take out artistry is to take out the humanity from the reviewers, at which point an algorithm (rather than people) would suffice for reviewing.
    – Reed
    Commented Nov 2, 2015 at 16:08
  • 6
    @HansPassant Well to re-phrase, if SO has rules for what a tag wiki should contain, then reviewers should follow those rules. The question then is, what rules are there and do they make any sense? These are the guidelines for tag excerpts, which are mildly helpful. I think the main issue is the flood of tags which is just a copy/paste of wikipedia with no tag usage guidance. Those tag wikis are pretty useless.
    – Lundin
    Commented Nov 2, 2015 at 16:17
  • 9
    A) I find the assertion that this is not "plagiarism" weak. That attribution is not prominent does not mean it does not exist, and in any case if the SO/SE community chooses to broaden slightly the use of the word to address other behaviors that are disallowed, even if they would not strictly be considered "plagiarism" in some other (e.g. academic) setting, that's its prerogative. B) Whatever you call it, the habit of copying content from outside sources is harmful to the quality of the tag wikis; your answer has failed to actually address either of my actual questions.
    – Peter Duniho
    Commented Nov 2, 2015 at 16:19
  • 1
    "If this kind of copy/pasta personally offends you then just fix it" -- I do and I have. I even followed the bread crumbs, after having seen a few of them, to track down the several others I could find from the same user and fixed those too. That's not the point. It is time-consuming to have to follow behind and clean up someone else's mess, and I would like to know if there is some mechanism by which I can enlist the help of others in some way, and more importantly have some official support in discouraging the user from continuing the harmful practice.
    – Peter Duniho
    Commented Nov 2, 2015 at 16:27
  • 4
    @Jakar Reviewers should not be "artists" because that does not appeal to my aesthetic senses. Furthermore, I am not in the mood to have "artists" doing reviews. Therefore, reviewers should not be "artists", Q.E.D.
    – Lundin
    Commented Nov 2, 2015 at 16:30

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