tea.mathoverflow.net - Discussion Feed (Deleted question about Rosenberg's proof)

Harald Hanche-Olsen comments on "Deleted question about Rosenberg's proof" (11990)

2010-12-17T15:27:21-08:00

@Sam: The story is getting old (possibly a decade or more), so I think it's a bit late for an academic solution. Or even any solution. If the journal is still afloat, maybe they have mended their ways. In any case, I'd have to ask my colleague for details and for permission to share them. I have forgotten much of it myself. If intensely curious, feel free to email me. My email address is not hard to find.

Sam Nead comments on "Deleted question about Rosenberg's proof" (11989)

2010-12-17T15:10:27-08:00

@Harald - Well, that is incredible. Since legal recourse is probably useless and certainly expensive allow me to suggest that an academic solution is called for. If you give me the details of the story in some way I can verify, then I can promise to never submit to that journal.

Harald Hanche-Olsen comments on "Deleted question about Rosenberg's proof" (11812)

2010-12-13T02:26:04-08:00

@Harald - your story is the first time I've heard of a preprint (as opposed to a paper) being published without the author's permission. Was it published under your colleague's name?

Yes. It is the only such story I have heard, too. It was clearly an attempt by the journal to pad their pages with some good quality material, not an attempt at stealing credit for the result.

Sam Nead comments on "Deleted question about Rosenberg's proof" (11783)

2010-12-12T14:19:45-08:00

@grp - IANAL, nor do I live in the USA. From my reading on-line (haha!) I gather that moral rights are not supported by case law in the USA but are supported by case law in several European countries.

As regards "They may have their own interpretation of the ideas which, among a myriad of possibilities, might do the opposite of delighting you." let me simply reply with a well-known quote of Jefferson:

"If nature has made any one thing less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of every one, and the receiver cannot dispossess himself of it. Its peculiar character, too, is that no one possesses the less, because every other possesses the whole of it. He who receives an idea from me, receives instruction himself without lessening mine; as he who lights his taper at mine, receives light without darkening me."

and a quip of Howard Aiken's:

"Don't worry about people stealing an idea. If it's original, you will have to ram it down their throats."

Sam Nead comments on "Deleted question about Rosenberg's proof" (11776)

2010-12-12T14:04:45-08:00

@bbischof - Plagiarism of published work is common. Here are the two examples that first came to mind:

http://www.eurekajournalwatch.org/index.php/2007_Plagiarism_Ring_Affair

http://en.wikipedia.org/wiki/D%C4%83nu%C5%A3_Marcu

I also have about one student a year (out of 5-10 students) who hands in a rough draft of their second year essay that copies directly from Wikipeda or a textbook. I scold them and silently thank the gods for the day that I thought of requiring rough drafts (so I can modify their behaviour before they turn in the final version).

I'll guess that plagiarism happens all over the place at a fairly low but steady rate. It is the result of people being confused -- they come to believe that academia is about "paper credits". It is not. It is about futhering human understanding, either in the small (say of a single student) or in the large (ie Galileo destroying the heavenly spheres).

grp comments on "Deleted question about Rosenberg's proof" (11774)

2010-12-12T13:51:37-08:00

Sam, thank you for the clarification. I understand the statement better, but still take issue with it, especially upon reading this extract found at the link you kindly provided:

"While the United States became a signatory to the convention in 1988,
it still does not completely recognize moral rights as part of copyright law,
but rather as part of other bodies of law, such as defamation or unfair competition."

I suggest that your choice of words was executed with insufficient care, and that there is still a measure of unintended insult in the phrase. I won't press the matter any further though, and I do not ask for further clarification or rephrasing.

Also, to the earlier point of scruples, anyone who does not respect moral rights might change more than just the name. They may have their own interpretation of the ideas which, among a myriad of possibilities, might do the opposite of delighting you. Either you are assuming extreme consideration on the part of the hypothetical unscrupulous republisher, or I am seeing something you are not in this situation. I still imagine you would be concerned about the scruples of the republisher.

Thanks again for the clarification; I feel my points have been addressed.

Gerhard "Ask Me About System Design" Paseman, 2010.12.12

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11772)

2010-12-12T13:45:37-08:00

You are free to republish my work under my name, because my work is (almost completely) in the public domain.

+1 for practicing what you preach.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11771)

2010-12-12T13:43:50-08:00

@Sam: Yes, I realize this. My comment was that copyright law is sufficiently strong in the US already. I have no desire to make it any stronger.

bbischof comments on "Deleted question about Rosenberg's proof" (11767)

2010-12-12T13:30:29-08:00

I am sorry, but many of you are speaking in generalities and innuendo, but I don't understand. Could you be more specific? Herald's explanation was useful for instance, but posts like Yemon's and Willy's are unnerving but confusing. What are you two referring to? I feel like this is something that is very useful for young mathematicians to know. Sorry for being so naive, I just have never ran into this before.

Sam Nead comments on "Deleted question about Rosenberg's proof" (11765)

2010-12-12T13:15:13-08:00

@grp, Harry - "Moral rights" is a technical term in the field of copyright. Please see http://en.wikipedia.org/wiki/Moral_rights_%28copyright_law%29

@grp - You are free to republish my work under my name, because my work is (almost completely) in the public domain. If you decide to publish my work under another person's name then I will bring that fact first to your attention (as I will assume some misunderstanding) and then to the attention of your employer. Regardless of what happens after that point I will be somewhat flattered that you chose my theorems and not, say, those of Gowers.

Sam Nead comments on "Deleted question about Rosenberg's proof" (11764)

2010-12-12T13:09:06-08:00

@Yemon - copyright violation is a civil offence while plagiarism is an academic offence. To be honest I would much rather use the tools of academia to police the latter that use the courts to deal with the former. In any case, the free exchange of ideas is far more important that the protection of "academic credits".

grp comments on "Deleted question about Rosenberg's proof" (11763)

2010-12-12T13:08:39-08:00

Sam, I take issue with the phrase "America has no notion of moral rights" . Perhaps a rephrasing or clarification from you would help stave off a flame war.

Also (hypothetically adopting the intent I see behind the phrase), would you take the same delight if I republished your work and associated someone else's name with it rather than your own? I suspect not. I see how scruples come into it, and hope you will see it as well.

Gerhard "Smile When You Say That" Paseman, 2010.12.12

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11762)

2010-12-12T13:04:30-08:00

(Well, in Europe. America has no notion of "moral rights".)

Yeah, but in America, copyright lasts until something stupid, like a hundred years after your death.

Preprints need not be public domain, but published papes should be. Once you've obtained the "journal credit" you no longer should protect your work from being published without your consent. (And I certainly don't think that the inital publisher should control future use/publication of the work!)

This is more in line with what I was thinking. It just came out dumber because I've never spoken with anyone about the process of publishing a paper.

Sam Nead comments on "Deleted question about Rosenberg's proof" (11761)

2010-12-12T13:00:14-08:00

Preprints need not be public domain, but published papes should be. Once you've obtained the "journal credit" you no longer should protect your work from being published without your consent. (And I certainly don't think that the inital publisher should control future use/publication of the work!)

@Harald - your story is the first time I've heard of a preprint (as opposed to a paper) being published without the author's permission. Was it published under your colleague's name? Or was it a straight-up plagiarism? If the latter then the publisher was violating your colleague's "moral rights" even if the preprint was in the public domain. (Well, in Europe. America has no notion of "moral rights".)

@WillieWong - I want my ideas to be spread far and wide -- I would be delighted beyond words if any of my work was ever republished. I don't see how the scruples of the publisher come into it...

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11760)

2010-12-12T12:58:28-08:00

Your faith is touching...

By faith, I guess, you mean naïveté?

that is extremely naive and misguided. Re-publication of your written words can happen outside the academic community, by unscrupulous publishers and such.

Which raises the question, what kind of scrupulous publisher charges obscene amounts of money for access to work done by other people for free?

Yemon Choi comments on "Deleted question about Rosenberg's proof" (11754)

2010-12-12T12:12:15-08:00

A paper being public domain does not mean that anyone can just plagiarize from it, simply because academia deals with plagiarism by other means.

Your faith is touching...

WillieWong comments on "Deleted question about Rosenberg's proof" (11752)

2010-12-12T12:05:08-08:00

@Harry: that is extremely naive and misguided. Re-publication of your written words can happen outside the academic community, by unscrupulous publishers and such.

Harald Hanche-Olsen comments on "Deleted question about Rosenberg's proof" (11744)

2010-12-12T10:29:22-08:00

A colleague of mine had one of his preprints appear in some obscure journal once to which he had not submitted it. They just took it and published it without asking. This caused him some trouble when he himself submitted the paper to a different journal, being unaware at the time what had happened. After this experience, he always puts a copyright notice on all his preprints. Whether this is a more or less effective way of dealing with the problem than academic sanctions is unknown to me.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11737)

2010-12-12T08:57:42-08:00

Preprints should be public domain, as in, there is no legitimate reason why researchers should care about their copyrights (unless, in the converse case (reverse? inverse? complementary?), when only subscribers to a certain journal can read the paper). A paper being public domain does not mean that anyone can just plagiarize from it, simply because academia deals with plagiarism by other means.

I don't think that copyright law is ever a real cause for concern to people who engage in academic dishonesty or other fraud in general.

Daniel Moskovich comments on "Deleted question about Rosenberg's proof" (11735)

2010-12-12T08:44:58-08:00

You're right- I mis-spoke and said something obviously stupid.
What I meant to say was that, after you publish anything (a paper, a book, an article, a painting, a musical composition, an MO question), then it has a life independent of its authors. Anyone can read it, and anyone can have an opinion on it- and with citation, anyone can use it. It exists as itself, not as an extension of its creators.

Harald Hanche-Olsen comments on "Deleted question about Rosenberg's proof" (11731)

2010-12-12T07:38:11-08:00

Preprints are public domain

What a remarkably odd thing to say. I expect you mean to say anybody can access and read them, which is not at all the same thing.

Daniel Moskovich comments on "Deleted question about Rosenberg's proof" (11689)

2010-12-11T17:08:23-08:00

Preprints are public domain, and I think that their potential errors, as long as they are of wider interest, are a legitimate topic for public discussion. But I think that a wise person phrases such questions diplomatically.
Also, sometimes we (collectively as mathematicians) want to verify a high-profile proof quickly; and sometimes we need help understanding a proof, which the author might not be the best person to give. Not asking because of etiquette might impede our work.
On the high profile side (papers whose veracity might effect an entire subfield), I would discuss it on a blog, rather than on MO, so that people have the chance to discuss the paper and find the error(s) together, if they exist- checking a proof is discussion-y, therefore not really for MO.
On the "not understanding" side, I think this is perfect MO territory as a matter of fact. If, a few years ago, I could have asked on MO about a certain argument I didn't understand (not having the background in that field, tangential to the rest of a paper, but critical in the proof), I would have saved a tremendous amount of time I used, trying to build on a certain result whose proof was wrong. If you don't understand a critical step of an important proof (and have put some effort into understanding it), then for goodness-sakes, ask! Don't refrain from asking for fear of offending the author... and because an author's reply might make insufficient sense, MO becomes an extremely valuable resource.

Martin B. comments on "Deleted question about Rosenberg's proof" (11651)

2010-12-11T02:09:26-08:00

@Ben: Yes, ok. I refered to the reactions to my question, which were unrelated to the specific situation.

Ben Webster comments on "Deleted question about Rosenberg's proof" (11649)

2010-12-11T00:08:26-08:00

Martin-

The issue isn't this particular example and whether you were right or wrong. I think at this point, this thread is clearly about how we deal in general with discussing potential errors in papers on MO. I'm not sure we've hit an consensus yet, but it's turned to a general question about what happens next time around, not the merits of your case.

Martin B. comments on "Deleted question about Rosenberg's proof" (11648)

2010-12-11T00:01:01-08:00

Good points. But be aware of the specific situation.

The author claimed that it is obvious that a union of subsets not containing a given subset $T$ does not contain this subset $T$. This is wrong, I don't have to be precautious about this statement. I have asked MO if we can correct this proof (and all subsequent lemmas using this union) by choosing another object instead of the union, and in fact, there was a canonical choice the author surely had in mind all the time. But it seems to me that nobody wants to think about that and just repeats general advice. So I've closed the question. Besides, I've also had other questions concerning the proof where I was not sure at all if they are actually false, but I'm sure that there are gaps. As you might guess, my patience regarding this proof has gone away.

I'm very happy with Gabber's proof. We don't need local properties of various types of spectra at all. I will write it up for my diploma thesis and you can e-mail me if you want to read it.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11637)

2010-12-10T17:26:47-08:00

Dear Matt,

Ah, okay. I thought that was just a given. However, I didn't read Martin's question, so I don't know how he asked it. I have asked such questions on MO before, and I always presume that it is either my misunderstanding or a typo (the book I'm referencing has very terse proofs as well as a veritable multitude of typos (and by typo, I mean something more like "small (albeit confusing) error that is inconsequential to the conclusions of the proof")).

(The comment above has quadruply nested delimiters =X)

Emerton comments on "Deleted question about Rosenberg's proof" (11634)

2010-12-10T17:07:51-08:00

Dear Harry,

Diplomacy is part of it, but I think is more than this; it is a question of mind-set. I can tell you from being on the other side that having someone suggest that you have made a mistake, when in fact they are just misunderstanding, can be very frustrating. If you (by which I mean "one", not you in particular) think you have found a mistake in someone's work, it is just as likely (and probably more so, in my experience) that you are simply misunderstanding.

So I think such questions should be asked with a presumption that one is misunderstanding things, rather than that there is a mistake.

Regards,

Matt

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11633)

2010-12-10T16:53:49-08:00

It seems like the only thing at issue here is how diplomatically the question is worded. Is that a fair assessment?

bbischof comments on "Deleted question about Rosenberg's proof" (11632)

2010-12-10T16:22:01-08:00

I agree with the recent posts of Andy and Emerton 100%. My mistake above was not putting this in perspective.

Thanks for the good advice from those above, I am glad I will never have to make this mistake myself.

Emerton comments on "Deleted question about Rosenberg's proof" (11628)

2010-12-10T14:28:39-08:00

I strongly agree with Andy. The issue is not (just) getting bad questions. Rather, it is that one should always be very polite and cautious in pointing out possible mistakes in other peoples work, and a widely read professional forum does not lend itself to such politeness and caution.

An important point to remember is that claiming a new result, especially a significant one, involves putting oneself right out on the edge, and there is always the danger of coming a cropper. One has to be respectful of this. On the one hand, if you made a mistake in a paper, you would presumably rather not have people trumpeting this all over MO, but would prefer to get the chance to quitely withdraw the claim; it's just professional courtesy to offer the same opportunity to others. On the other hand, if you are incorrect in your claim of a mistake, then you look somewhat foolish, and the author accused of making a mistake may well get annoyed. It's better to prevent such possibilities by not initiating them in the first place.

There is a reason that experienced mathematicians begin questions about another person's work (even if they have their doubts) with the expression "I don't understand [such-and-such in your argument]", putting the fault on themselves, not the person whose work it is. If there really is a mistake, it will come out with patient questioning, and if not, one avoids having egg on one's face.

P.S. I think Noah's suggest question, "I don't understand step X", is fine. This is the kind of question people ask all the time. It is very different to an accusation of a mistake.

Andy Putman comments on "Deleted question about Rosenberg's proof" (11627)

2010-12-10T14:11:50-08:00

Accusing someone of being wrong in a paper of theirs is very serious. There is a strong tradition in mathematics that you don't do this in too public a manner; for instance, it is generally considered bad form to cast aspersions on someone else's proof in a lecture or paper without clearing it with them beforehand. In the rare situations in which this has to be done, MO is not the appropriate forum.

I also strongly worry about young people confusing small (easily correctable by experts) mistakes with genuine errors (as appears from bbischof's comment above to be the case with regards to the OP's question). This creates a bad situation both for the questioner and the author of the paper.

This is not to say that one can't ask questions about papers on MO. However, they need to be framed in such a way that no-one would read them as accusing the author of being wrong.

Noah Snyder comments on "Deleted question about Rosenberg's proof" (11625)

2010-12-10T14:06:57-08:00

This is why "as long as it doesn't happen too often" was an important part of my conditions.

bbischof comments on "Deleted question about Rosenberg's proof" (11624)

2010-12-10T14:03:07-08:00

@Andy After reading your comment, I realized the somewhat obvious issue with this. Am I correct in assuming that your worry is that every time someone doesn't understand a result in a paper, they post it as a purported mistake, and this will lead to an enormous amount of garbage questions on MO.

I overlooked this in my last post because I was thinking of the situation at hand, which is quite far from this situation.

So now I revise my position to I don't know.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11623)

2010-12-10T13:52:40-08:00

@Andy: I'm inclined to agree with Noah and Bryan on this. Could you explain your reasoning?

Noah Snyder comments on "Deleted question about Rosenberg's proof" (11622)

2010-12-10T13:39:44-08:00

I think if the poster has done their due dillegence, and if the question is phrased as "I don't understand step X" rather than "I think step X is wrong" and it doesn't happen too often, then I don't have a big problem with such a question.

Andy Putman comments on "Deleted question about Rosenberg's proof" (11621)

2010-12-10T13:32:43-08:00

@Harry : Yes, absolutely. Allowing that kind of thing would be damaging for MO.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11620)

2010-12-10T13:22:26-08:00

@Andy: Even in the case that the author has been e-mailed?

José Figueroa comments on "Deleted question about Rosenberg's proof" (11619)

2010-12-10T12:48:09-08:00

I tend to agree with Andy; although I have in the past asked a question about an erroneous theorem in a book I was studying and one of the answers saved me potentially a lot of time. So I think that we should perhaps exercise caution and treat each case on its individual merits. I have no opinion on this particular question, though.

Andy Putman comments on "Deleted question about Rosenberg's proof" (11618)

2010-12-10T12:26:07-08:00

I disagree strongly with Harry and bbischof. I don't think that MO is a good place for a discussion of possible errors in people's papers. That's a rabbit hole we just shouldn't go down.

bbischof comments on "Deleted question about Rosenberg's proof" (11617)

2010-12-10T12:17:04-08:00

I agree with Harry that the question should be undeleted. I think questioning and being critical is extremely important to mathematics, and that this forum IS a good place for reports of possible errors.

In response to the tone of the post. The first version was probably ok. The later iterations were a bit incendiary. I think it is better to be a bit less intense when suggesting there are issues. I think a good way of posting this sort of thing, is saying that you think there might be an error, rather than asserting that the result is incorrect. However, I am far from the politeness police. You can choose to be as aggressive as you wish, but I think the comments reflected a preference for politeness.

Regarding you emailing Rosenberg. I have spoken with him about this. Yes, he is busy right now and of the opinion that you can resolve your difficulties on your own with your advisor. He also told me that one of the issues is that you want a simpler proof in a specific setting. Since he has not thought about this simpler proof, it is not surprising that he not create one for you. As I told you in personal email, working through his machinery might be advantageous to you, even though it can appear daunting.

Please reopen the question for at least the reason that you can answer it after you understand Ofer Gabber's and Zoran Skoda's correspondence with you . Others(at least I) would be interested to hear these things.

DISCLAIMER: I do NOT speak for Rosenberg. I do NOT speak for the MO community. This is just my interpretation of the situation. I cannot answer the questions that you have asked, and unfortunately don't have time right now to attempt to.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11611)

2010-12-10T10:31:27-08:00

Oh, then I have no problems with Martin posting it on MO, and further, I think he should undelete his question.

Scott Carnahan comments on "Deleted question about Rosenberg's proof" (11610)

2010-12-10T08:52:40-08:00

@Harry: In the comments to the question, Martin mentioned that he had emailed the author, but that the author was too busy.

Harry Gindi comments on "Deleted question about Rosenberg's proof" (11609)

2010-12-10T08:45:13-08:00

You should always always always e-mail the author first.

Martin B. comments on "Deleted question about Rosenberg's proof" (11608)

2010-12-10T08:33:27-08:00

Since the question has gained some attention, I would like to mention here why I have deleted my question titled "Supposed errors in Rosenberg’s proof of the Reconstruction Theorem".

a) I continue the discussion with Zoran Skoda via E-Mail.
b) Ofer Gabber has sent me his proof and it is far more comprehensive, transparent and short than Rosenberg's.
c) It seems to me that this community is not a good place to report errors in preprints.