Wikipedia talk:Copyright problems

I have just added an entry for Amalie von Wallmoden, Countess of Yarmouth after the author of the ODNB article complained on the article's talk page. Our article is heavily based on the ODNB article, which is currently free to view (for the next 5 days) as one of the "Lives of the week". (See http://www.oxforddnb.com/view/article/28579 )

That case is pretty clear-cut, but I have recently noticed a similar problem with some other articles - Falkes de Breauté, William Tresham and Alan Garrett Anderson - and their respective entries in the ODNB: [1] [2] [3] (subscription is required for these).

There is some relevant discussion at User talk:Ironholds and Wikipedia_talk:Did_you_know#Copyvio_and_excessive_quotations, including examples of the text which causes me concern.

Our articles are, I believe, very thinly rewritten versions of the ODNB - not blatant copy & paste, but the same facts, presented in pretty much the same order, often using the same or very similar words. What should be done about these? -- Testing times (talk) 11:48, 20 October 2008 (UTC)Reply

If they are very thinly rewritten, and it certainly looks like they are from the examples you give, they may constitute derivative works or, at best, plagiarism and should be revised. Because copyright law protects creative expression of ideas and not the ideas themselves, many people think that slight revisions eradicate copyright concerns. Of course, as you know, this is not the case. US Federal Courts utilize a "substantial similarity" test. McCarthy's Desk Encyclopedia of Intellectual Property notes about this, "Exact word-for-word or line-for-line infringement does not define the limits of copyright infringement." (Part of this source, including that quote, is available online here). This concept is divided into two: "fragmented literal similarity", which essentially covers lifting chunks and phrases from a source sufficient to outweigh fair use considerations; and "comprehensive nonliteral similarity", where the same language may not be utilized but the derivative work copies the basic form and structure of the original. Substantial similarity is generally tested in courtrooms through two questions: first, does a paraphrase borrow enough from the source to constitute a "substantial taking" from the original (representing a significant portion of that work, for instance, or a key element of it)? If it does, does the use of the material meet the test of fair use? If the answers to these are yes and no, then you've got copyright infringement. Even if the answers are no and yes, however, you may have plagiarism.

It may be helpful to point editors who seem to have trouble with this regards to these university hosted documents: here, here. Also useful is this publication aimed at medical professionals which gives a succinct overview of how to tell when paraphrase crosses the copyright line: this.

As for handling these, the first thing I would do is clarify this for the contributor. This is probably a good faith misunderstanding of copyright & plagiarism; after all, even Wikipedia's user copyright warning suggests that everything is a-okay as long as you have access to a thesaurus. Since this problem has first started to arise here, I've been wondering if we need to make that point more clear in policy somewhere.

The next thing to do is either revise the material or remove it. If the source were openly published, I'd suggest you tag it with the {{copyvio}} to give contributors an opportunity to eliminate substantial similarity concerns. However, when it comes up at the end of the week, I am likely to be the admin reviewing it, and I don't have access to that source. Do you or somebody else involved in these conversations? If so, would it be possible for one or more of you to help revise these articles? Alternatively, earlier versions of them can be restored with a note of explanation at the article's talk page. --Moonriddengirl ^(talk) 15:20, 20 October 2008 (UTC)Reply

Thank you for the guidance. I will direct the author of these articles to this discussion. I have ODNB access, as I suspect does the author here (it is widely available from UK public libraries), but I don't really have time to revise these articles. In the absence of revisions, presumably they should be removed? -- Testing times (talk) 11:37, 22 October 2008 (UTC)Reply

Well, first noting that I am not a lawyer though consideration of US copyright application to text is part of my "day job", I would say yes. Earlier versions should be restored until a complete revision can be done that eradicates "substantial taking". Obviously, we want to be careful not to be overzealous in the application of this, particularly as definition is nebulous and even copyright attorneys cannot always predict where the court will come down. :) But your examples seem to me to closely enough replicate the pattern of the original that they should be regarded as infringement. I can't speak to the rest of the text, but if you remove it with a note of explanation and others disagree, the matter can be opened for wider discussion, hopefully with additional contributors who are able to compare the original. --Moonriddengirl ^(talk) 12:26, 22 October 2008 (UTC)Reply

I'd appreciate it if someone here could take a look at a few representative articles I prepared based on the ODNB when I had access to it. Lord Robert Manners (Royal Navy officer), Thomas Chamberlayne, and Thomas Howard, 1st Earl of Suffolk (long) are representative. For short biographies, where the article is almost a prose list of positions held and dates, it's very difficult to avoid "comprehensive nonliteral similarity", insofar as one has a limited number of facts which are generally best understood in chronological order. User:Choess 22:10, 24 October 2008 (UTC)Reply

Oh, don't I know it. It's one of the banes of my existence. :/ Not having access to the source, I can't weigh in on this, but hopefully somebody will be able to take a look for you and reassure you. :) Perhaps Testing times could take a glance? Again, we don't want to be overzealous in this, but we do want to consider the question. --Moonriddengirl ^(talk) 22:17, 24 October 2008 (UTC)Reply

Well, from a quick look at Lord Robert Manners (Royal Navy officer), the Wikipedia article is clearly based on the ODNB entry[4] - as one may expect when the ODNB is given as the sole reference. Both recite broadly the same facts in broadly the same order, but the specific language used to convey the facts in each article is generally quite different. As people will appreciate, this is very much a question of shades of grey, but my general impression in this case is that there has not been a "substantial taking" in the sense outlined above (unlike the examples that I mentioned further above, where this is a real concern, in my view - but see below about the 1900 DNB).

I have not reviewed Thomas Chamberlayne or Thomas Howard, 1st Earl of Suffolk, but if the language in these articles is similarly distanced from the original, then they should not be a matter for concern either, I guess. -- Testing times (talk) 16:44, 28 October 2008 (UTC)Reply

I have been dumping DNB text directly into Wikipedia as a starting point for several articles. Note: this is direct literal copying, word for word. However, my article source is the 1900 version of the DNB, not the ODNB, and each article is specifically attributed, with a link to the Wikisource version, which in turn is specifically attributed to the page scans and OCR text at archive.org. As far as I can tell, there can be no question that the 1900 DNB published in the US is in the public domain, and that the process by which I am introducing the material into Wikisource and then into Wikipedia is acceptable. (But please tell me if I'm wrong...) I take this approach to ensure that the edit history captures an audit trail of the transformation from the public domain text to the current text. I do not have access to the ODNB. My question is this: If someone updates one of these articles with facts that have been added to the ODNB version, and the update is then substantially identical to the ODNB version, are we in trouble? Note that the ODNB version is likely a derived work, and so is ours, so we run the risk of "convergent evolution" of the articles. -Arch dude (talk) 17:59, 26 October 2008 (UTC)Reply

This is a question I haven't encountered, but at that point we would have the defense of the article's history, which should bear out that the version we duplicated was public domain. If the ODNB sent a take-down to the Wikimedia Foundation, it would be up to our lawyer to decide how to handle it. If all else failed, we could probably revert the updates substantially identical to the ODNB. --Moonriddengirl ^(talk) 23:39, 26 October 2008 (UTC)Reply

Hmm. I think there may be some tricky questions lurking here. Certainly the 1900 DNB is as public domain as the 1911 EB, and can be copied wholesale without copyright concerns (ideally with attribution, of course).

But what if someone essentially copies an ODNB article which is substantially similar (or even identical) to the 1900 DNB? Are my concerns above misplaced, because the ODNB article is public domain anyway? How would we know? Is there an easy way to look up an article in the DNB copies at the Internet Archive (here)? -- Testing times (talk) 16:44, 28 October 2008 (UTC)Reply

The Wikipedia project is Wikipedia:WikiProject Missing encyclopedic articles/DNB. It has a "sister project" at Wikisource, at s:Wikisource:WikiProject DNB. I go to the WP project page to get a pointer to the DNB1900 volume at archive.org, and from there I retrieve the PD article and create the Wikisource version, with strict attribution. After that, I create the WP article by starting with a straignt copy of the Wikisource article, again with strict attribution. This results in a absolutely horrible initial article, but is also causes the article history to become an audit trail, and as long as nobody looks at the article before the first wikification pass, The ugliness doesn't show too much. (Whether the result is a good starting point for a WP article is an aesthetic issue, not a copyright issue.) Now to your question: What if an editor dumps an ODNB article into WP, but the ODNB article is itself a strict copy of the old 1900 DNB article with no copyrightable derivative inclusions? I don't know if any such articles exist, but if one does, and if we are challenged, then we will need to create the equivalent article at Wikisource from the original DNB 1900 material, and then verify that the ODNB version has no copyrightable changes, and then change our references to ref the old DNB version. depending on the later revisions history, it may be easier to re-create WP article. Conclusion: Don't copy from the ODNB even if you think the article is actually DNB1900: it's more trouble than it's worth. -Arch dude (talk) 17:31, 28 October 2008 (UTC)Reply

It could be difficult to tell. Wikipedia:PD#Derived works and restorations of works in the public domain notes that when a new work is built off of a public domain source, only the creative elements in the new work are subject to copyright. If the ODNB article is identical to the 1900 DNB, it's not protected by copyright. If it is identical save for a few new sentences, only those new sentences are protected. I've spent part of this morning addressing your copyright concerns at Amalie von Wallmoden, Countess of Yarmouth. I don't have access to the current ODNB article, but I did have access to the 1909 version, the whole of volume XX being online. :) (And downloadable, it seems, as a 120 mb pdf.) If we're lucky, the article we're comparing will be archived in googlebooks, as that is. If not, given demonstrable duplication, we may have to err on the side of caution. --Moonriddengirl ^(talk) 17:35, 28 October 2008 (UTC)Reply

Thanks for all of that (the Wikiproject's index to the pages at Archive.org could be very useful, and I was not aware of the DNB archive link on the ODNB pages).

As it happens, the DNB article for Amalie von Wallmoden, Countess of Yarmouth (available from here in various formats, or a long way down here in a 2.4MB text file - search for "Wallmoden") is very much shorter than the ODNB version, and quite different in several respects, so congratulations on the comprehensive re-writing of our article. -- Testing times (talk) 18:03, 28 October 2008 (UTC)Reply

I have similar concerns about User:Ironholds's article on Terence Fox, which bears more than a passing resemblance to the ODNB entry.[5] Here are some examples:

This chap died in 1962, and the ODNB entry is almost the same as that first included in the 1981 DNB, so there is no chance of the ODNB entry being public domain. -- Testing times (talk) 12:53, 6 November 2008 (UTC)Reply

I've been vaguely following this debate, and often use ODNB as a source myself, so want to make sure I'm doing it right. When writing a biographical article, surely it's harldy surprising that there is a similar structure, it's natural to write chronologically. ODNB articles fairly frequently cite newspaper obits as references, and if you can track down two or three different obits they will have broadly similar structures. The facts of the life can't be copyrighted, yes? and is chronological treatment actually particularly original? David Underdown (talk) 13:00, 6 November 2008 (UTC)Reply

Exactly the problem I encounter. I've actually rearranged the text I use from the ODNB source in several cases because they sometimes don't do things chronologically. ODNB articles are encyclopedic, and so are ours; there's only so much difference you can get from a source text without resorting to a thesaurus. Ironholds (talk) 13:02, 6 November 2008 (UTC)Reply

(EC) The facts of the life can't be copyrighted, but the creative expression, including the basic structure, can. By that I do not mean "chronological organization." But this example follows very closely on the original, including minor rearrangement of material (in the last sentence) and some chunks of literal duplication ("to study for the mechanical sciences tripos"--changing caps and removing the plural from Science is insufficient). Note, also, "He was the only son born to Corelli Fox, a electrical engineer, and his wife Mabel". (I've emphasized the text copied, in same order, from the source.)

There are tricks in paraphrasing to help avoid this, though sometimes it can be like going around one's back to scratch one's elbow. Helpful here is the use of multiple sources, as incorporating information from various can go a long way to eliminate this concern. Also potentially useful is Wikipedia's own MOS. Some of this information would be included in a WP:LEAD, others in a body. If organized accordingly, we would not be following the basic structure of ODNB anyway. As I've said above, we don't want to be overscrupulous in this respect, but we do need clear separation. --Moonriddengirl ^(talk) 13:14, 6 November 2008 (UTC)Reply

I understand the problem. Honestly, I do. But a thin paraphrase is just not enough, in my view. Here are some more examples from Terence Fox:

Particularly telling, in this context, is the phrase in our article "a series of nervous breakdowns in the 1950s which caused him to resign as shell chair in 1959". ¡¿ resign as shell chair ?! Looking at the ODNB, it is clear where the peculiar turn of phrase "shell chair" in this sentence came from - "a succession of nervous breakdowns in the early 1950s which caused his resignation from the Shell chair in 1959". I could go on. -- Testing times (talk) 18:17, 6 November 2008 (UTC)Reply

"Shell Chair" is not a "telling phrase"; the chair was a Shell endowment, it is perfectly acceptable to refer to it as "shell chair" or "resign as Shell professor". As for the second half; how the hell would you express it differently without being unencyclopedic?Ironholds (talk) 18:49, 6 November 2008 (UTC)Reply

I don't want to get bogged down in the details, but would you really say "to resign as shell chair"? Would you say "Stephen Hawking has resigned as Lucasian chair"? It seems a bit odd to me, but it does directly parallel the phrasing of the original.

You could rephrase the last bit in many ways: how about:

According to his entry in the Oxford Dictionary of National Biography, Fox's natural inclination to intellectual rigour and inability to delegate effectively caused him mental strain which ultimately led to a series of nervous breakdowns in the 1950s. In the end, he resigned as Shell professor in 1959 in favour of Peter Victor Danckwerts. Fox died three years later at the National Hospital for Nervous Diseases in London. He never married.

YMMV, but I think this is substantially better. -- Testing times (talk) 19:13, 6 November 2008 (UTC)Reply

Shell chair seems fairly standard phraseology in British English to me too (though I'd probably cap-up Chair, he is the Shell Professor, holding the Shell Chair, as in the original ODNb, saying "from the" would be better grammatically, but I wouldn't say there's any particular originality in the phrase), same about the earlier comment about the mechanical science tripos, tripos is simply the proper name for a Cambridge degree course. The chagne to Shell Professor is better grammatically if you have "as" in front of it. David Underdown (talk) 09:32, 7 November 2008 (UTC)Reply

I've now tracked down his obit in The Times (Saturday 6 October 1962, p.12 Issue 55515, col A), this has "he was forced by ill-health to resign his chair in 1959", and "he read for the mechanical sciences tripos", so you could argue that the ODNB itself isn't being particularly original in using this phraseology-I've seen similar parallels in other ODNB articles where I've also been able to track down the obits. David Underdown (talk) 09:43, 7 November 2008 (UTC)Reply

Oh, well done in tracking down the original obituary. Using several independent source to write and article almost always helps, as they use phrases to describe the same events.

I agree, "Shell chair" would be a standard phrase to talk about the professorship, but I think you would usually talk about someone holding the chair rather than being the chair.

As I said above, I don't want to get too bogged down in the details of individiaul phrases: the overall impression I get from our article is that it is a thinly rewritten copy of the ODNB entry. The parallels go further than simply repeating the phrases "Shell chair" or "mechanical science[s] tripos". The small extracts from The Times you have given seem to use fairly different terms to report the same facts as the ODNB. By contrast, our article seems to use very similar turns of phrase as the ODNB. -- Testing times (talk) 10:26, 7 November 2008 (UTC)Reply

This article has just appeared on the Main Page in the WP:DYK section. Not great, when we are concurrently discussing whether it is a copyvio. -- Testing times (talk) 11:20, 10 November 2008 (UTC)Reply

I am writing an article for another language edition of Wikipedia on global warming, inspired by the BBC documentary: The Climate Wars. I have used only the format of the program, for example, these themes regarding the history of climate change:

• the program starts with "Letter to Nixon about global cooling in 1971" ->
• "Environmental issues of the 70's" ->
• "Filler" ->
• "Filler" ->
• Ehrlich's "Eco Catastrophe" ->
• Scientists' opinion on new ice age ->
• Cold periods of the 40s to 60s ->
• Heating up during the 70s.

"Filler" refers to sections I didn't cover. I didn't use the narrative as presented in the program, but basically just used the format as used by the program, in other words, discussing issues in the same sequence as they did. Many parts, as presented by the program, differ a lot from my article (since I'm fact-checking everything, lots of sections tend to get rewritten). Am I on safe grounds, as far as copyright is concerned, or can they sue me for the creativity involved in choosing and ordering the information the way they did? Anrie (talk) 11:04, 7 November 2008 (UTC)Reply

Hi. Sorry for the delay in response; obviously, this is not a heavily monitored page. :) This is not a run-of-the-mill question, and my opinion is based on some experience, but not legal certainty. I should also note that my experience relates to US copyright law, which is what governs en-Wikipedia. Copyright law governs creative expression of ideas. If your section heads were one word only, I would think you'd be okay. When you start to run into more complicated phrases, you could hit some trouble. If this were listed as a copyright problem here, I would probably play it safe by renaming the sections. As far as utilizing the division structure, as long as you aren't copying the text or too closely paraphrasing you should be fine. Utilizing multiple sources is precisely the way to avoid that. --Moonriddengirl ^(talk) 18:04, 19 November 2008 (UTC)Reply

Thanks for the reply! Anrie (talk) 16:26, 18 December 2008 (UTC)Reply

CarloscomB added alot of quotes from "Steve Gottlieb", I'm not sure if they're acceptable or not, could someone take a look? [6] 70.55.86.100 (talk) 22:00, 8 November 2008 (UTC)Reply

Hi. Glancing at this one, they seem to lack source information, which makes them very much not okay with WP:NFC. --Moonriddengirl ^(talk) 18:11, 19 November 2008 (UTC)Reply

I have removed all that found, except that I cannot tell where the problem material begins here. --Moonriddengirl ^(talk) 20:11, 19 November 2008 (UTC)Reply

A Image of Wolfgang Paul which I found on the net states: Photo: Humboldt Foundation Reproduction free of charge - specimen copy requested Would a use in Wikipedia be OK? --Stone (talk) 07:21, 14 November 2008 (UTC)Reply

Hi. I don't at a glance see anything indicating what license they are releasing it under or if it is public domain. I'd suggest you pose the question at WP:MCQ, which is specifically for dealing with images. A responder there may be familiar with the site. --Moonriddengirl ^(talk) 18:52, 19 November 2008 (UTC)Reply

I believe there is/was a talk page header template warning users not to re-add excessive fair use images to articles (e.g. "List of" articles). I am however unable to find it. Can someone please point me in the right direction? G.A.S^talk 08:03, 14 November 2008 (UTC)Reply

I'm sorry to say that I don't know where to find it. :) They may be able to point it out at WP:MCQ. --Moonriddengirl ^(talk) 19:07, 19 November 2008 (UTC)Reply

A very experienced editor disagrees with me over the interpretation of the policy. I challenged and removed some article text on grounds of possible copyright violation (diff) (and possible plagiarism). If I understand his comments at Talk:Anglia Regional Co-operative Society#History (and the edit summaries) correctly, he believes that it is ok for him to restore the challenged material (diff), while he works on satisfying the licence requirements. I would prefer the text remain deleted until the issue is resolved.

It would be helpful if a third party assists in editing the article, or joins the discussion at the talk page, to help us resolve this cordially.

--Hroðulf (or Hrothulf) (Talk) 10:42, 18 November 2008 (UTC)Reply

As the material clearly predates its placement here, at the article's inception, I have blanked the article pending verification of release. If the contributor does not verify, the material will have to be removed or revised. Thank you for noting your concern. --Moonriddengirl ^(talk) 19:06, 19 November 2008 (UTC)Reply

That statement appears in an edit summary [[7]]. It is in support of a quote of three paragraphs from a book which would be in copyright.

I expect that a brief attributed quote from copyright text can be justified, but where is the line drawn? Are three substantial paragraphs justifiable?

CBHA (talk) 00:27, 2 December 2008 (UTC)Reply

Three substantial paragraphs that do not serve any of the purposes for copyrighted text set out at WP:NFC are a problem. Quotations should be brief and should "illustrate a point, establish a context, or attribute a point of view or idea." I've removed it, as extensive quotations of copyrighted text are prohibited. --Moonriddengirl ^(talk) 01:07, 2 December 2008 (UTC)Reply

The edit summay is quite correct: a long attributed quote is not plagarism. Plagarism is unethical and is a violation of academic norms, and it can get you in trouble at school or as a journalist. Instead, the long quote is a violation of copyright law and it can get you arrested or sued. -Arch dude (talk) 00:54, 16 December 2008 (UTC)Reply

FWIW, three paragraphs, even long paragraphs, from a prose text is legal fair use is all countries that have a fair use provision, except when a short text being copied is copied entirely. (In US copyright law, there is a specific statement that there re no fixed guidelines to fair use. ) It usually however does not meet our Wikipedia standard of what fair use material to include, as it is usually excessive for our purposes in writing an encyclopedia. For purposes of criticism in a school paper or the like, if properly cited, ad provided one doesn't use it for the bulk of the aper, most teachers would or at least in my opinion should accept it. DGG (talk) 18:54, 17 December 2008 (UTC)Reply

Is there a page anywhere that explains how to write things "in your own words"? I'm looking for someplace I can point new users who copy material from other websites in good faith. I'd like to find a general explanation, possibly accompanied by specifics like "short phrases taken from the source are okay but not more than n words." Thanks, FreplySpang 01:28, 3 December 2008 (UTC)Reply

Sadly, I don't know of any such document, although I've been considering that we need one for a while. My best suggestions are from off-wiki: http://owl.english.purdue.edu/owl/resource/619/01/, http://plagiarism.umf.maine.edu/paraphrasing.html and particularly http://www.cinahl.com/library/cinahlnews/Cnews173.pdf. The latter looks at things from a legal perspective, but it helps explain why just replacing words here and there is not sufficient. There is no hard and fast rule about how many words in a row can be taken. You often hear "three words" used as a basic in academic circles, but fewer can be a problem if they are a particularly "apt phrase." Sometimes, more can be okay. We couldn't say "The grotesquely gorgeous John Doe" without quotation marks if our source said "John Doe is grotesquely gorgeous." "Grotesquely gorgeous" is an apt phrase. We'd be okay copying "The 97th Chief Assistant President of the Bumluck Field Squad", though, as there's nothing creative in that title. But this can be very hard for people unfamiliar with the concept to understand. --Moonriddengirl ^(talk) 01:52, 3 December 2008 (UTC)Reply

Oh well, thanks for the external links and explanation! Cheers, FreplySpang 00:21, 5 December 2008 (UTC)Reply

If you see a Wikipedia article that seems to be copied from a website, and revert the Wikipedia article to a version that does not match the text on that website, should you still list the article at WP:CP to bring it to the attention of an administrator (so the copyvio. version can be deleted), or are you done? The instructions are unclear whether reverting is the only step in dealing with such a situation, or whether it is the first step. 69.140.152.55 (talk) 15:29, 15 December 2008 (UTC)Reply

In many cases, you should be done. Quite often, that's what an administrator will do (see Wikipedia:Copyright problems/Advice for admins). It's a good idea to make a note at the article's talk page explaining why you removed the text. I have devised a template of my own I use for the purpose at User:Moonriddengirl/cclean. Should you ever have need of it, please feel free to use it. --Moonriddengirl ^(talk) 01:03, 16 December 2008 (UTC)Reply