I’m pleased to announce that Mathgen has had its first randomly-generated paper accepted by a ~~reputable~~ journal!

On August 3, 2012, a certain Professor Marcie Rathke of the University of Southern North Dakota at Hoople submitted a very interesting article to Advances in Pure Mathematics, one of the many fine journals put out by Scientific Research Publishing. (Your inbox and/or spam trap very likely contains useful information about their publications at this very moment!) This mathematical tour de force was entitled “Independent, Negative, Canonically Turing Arrows of Equations and Problems in Applied Formal PDE”, and I quote here its intriguing abstract:

Let $\rho = A$. Is it possible to extend isomorphisms? We show that $D’$ is stochastically orthogonal and trivially affine. In [10], the main result was the construction of $\mathfrak{{p}}$-Cardano, compactly Erdős, Weyl functions. This could shed important light on a conjecture of Conway-d’Alembert.

The full text was kindly provided by the author and is available as PDF.

After a remarkable turnaround time of only 10 days, on August 13, 2012, the editors were pleased to inform Professor Rathke that her submission had been accepted for publication. I reproduce here (with Professor Rathke’s kind permission) the notification, which includes the anonymous referee’s report.

Dear Author,

```
```Thank you for your contribution to the Advances in Pure Mathematics (APM). We are pleased to inform you that your manuscript:

ID : 5300285

TITLE : Independent, negative, canonically Turing arrows of equations and problems in applied formal PDE

AUTHORS :Marcie Rathke

has been accepted. Congratulations!

Anyway, the manuscript has some flaws are required to be revised :

(1) For the abstract, I consider that the author can’t introduce the main idea and work of this topic specifically. We can’t catch the main thought from this abstract. So I suggest that the author can reorganize the descriptions and give the keywords of this paper.

(2) In this paper, we may find that there are so many mathematical expressions and notations. But the author doesn’t give any introduction for them. I consider that for these new expressions and notations, the author can indicate the factual meanings of them.

(3) In part 2, the author gives the main results. On theorem 2.4, I consider that the author should give the corresponding proof.

(4) Also, for proposition 3.3 and 3.4, the author has better to show the specific proving processes.

(5) The format of this paper is not very standard. Please follow the format requirements of this journal strictly.

`Please revised your paper and send it to us as soon as possible.`

The author has asked me to include her responses to the referee’s comments:

- The referee’s objection is well taken; indeed, the abstract has not the slightest thing to do with the content of the paper.
- The paper certainly does contain a plethora of mathematical notation, but it is to be hoped that readers with the appropriate background can infer its meaning (or lack thereof) from context.
- It is indeed customary for a mathematical paper to contain a proof of its main result. This omission admittedly represents a slight flaw in the manuscript.
- The author believes the proofs given for the referenced propositions are entirely sufficient [they read, respectively, "This is obvious" and "This is clear"]. However, she respects the referee’s opinion and would consider adding a few additional details.
- On this point the author must strenuously object. The $\LaTeX$ formatting of the manuscript is perfectly standard and in accordance with generally accepted practice. The same cannot be said of APM’s required template, which uses Microsoft Word [!].
Professor Rathke is pleased that the referee nevertheless recommends the paper be accepted, since clearly these minor differences of opinion in no way affect the paper’s overall validity and significance.

However, in spite of this good news, there is a mundane difficulty which will apparently prevent the article’s publication. As an open access journal, APM naturally imposes a “processing charge” on its authors, which for this paper would amount to US$500.00. Unfortunately, due to recent budgetary constraints at the U. of S.N.D. at H., Professor Rathke finds that her research funds are insufficient to meet this expense. It therefore appears that APM’s estimable readership, and the mathematical community at large, will sadly be deprived of seeing the fruits of Professor Rathke’s labor in print.

Bummer.

Congratulations, Nate! Nice work.

Pingback: Randomly generated mathematical research papers | Mathematical Biology

Please note that it is improper to cite the bibliography in the abstract. An abstract must be able to stand alone. The citation code [10] should be replaced by the actual citation. I trust you will inform the author (MathGen, Prof. Rathke, whoever).

You’re right, of course. The abstract uses the same template sentences as the body, which include citations randomly. It’s not clear how to avoid this without leaving the context-free setting, or making a new copy of those templates without citations for the abstract, which would add a lot of duplication to the code. But if you have any thoughts, the code’s on github; feel free to submit a pull request

Trivial addendum to my previous thought: Thank you for clarifying the status of this publisher. I had been uncertain of their degree of reputability. Now I’m certain.

God says…

calculations foundations babe disorders expect preliminary

Idaho motion drunkenness ETEXTS unknown meats everlasting

forgive Cup victorious formlessness desire lift uphold

Mediator alarmed ascertained heightening pages do Know

member descend turn cold physical

I couldn’t decide whether or not this was spam, but I figured it was in the spirit of the project to leave it.

You are completely right, dude!

I love it!

Haha, good decision on leaving the comment online

Reading math articles will never be the same. I will always wonder wheather it was generated or I’m just too stupid to understand it.

It’s a program that does tongues — that thing everybody mocks in the Bible. I thought you might like it.

I wonder if something like it was used by the hybrid in Battlestar Gallactica

Astonishing ! I am definitevely a poor boy, since I have always spent a considerable amount of time trying to figure out what is the right journal to sent my work to…

I get this in Mathgen all the time…

Error

Seed invalid, must be numeric

Try again

That could be a bug in the Javascript. Can you reproduce this? What’s the URL that the page generates for your PDF? And what’s your browser?

Nate, congratulations for your terrific program!

Unfortunately I get the same error with firefox or iexplorer

Hmm. It’s hard for me to fix it if I can’t reproduce it. What I need is the complete URL that you are at when you see the error message. Alternatively, if you remember the name you used on the paper (and it wasn’t the default), then I may be able to find it in my logs that way.

Just clicking the Generate button without changing anything gives that error for me, resulting url:

http://thatsmathematics.com/mathgen/paper.php?nameType%5B1%5D=custom&customName%5B1%5D=A.+Lastname&nameType%5B2%5D=custom&customName%5B2%5D=&nameType%5B3%5D=custom&customName%5B3%5D=&nameType%5B4%5D=custom&customName%5B4%5D=

Ensuring that javascript served from thatsmathematics.com is allowed to run solved that problem for me.

Are you, by any chance, using NoScript?

This should be fixed now.

You should put this on kickstarter.com and raise 500$. I’m sure you’ll get the money.

+1

Hmm. I’m not sure that there’s any point in giving SCIRP $500, even of other people’s money. On the other hand, nothing prevents you from submitting your own paper to APM or one of its many siblings, and starting your own Kickstart/Indiegogo/etc…

Or even better. Generate a paper by Leibnitz and Newton…

Please don’t do this. You’re not playing a joke on them by getting this published — all they want is your money.

A bit late congratulations, this is just awesome! I especially like sentences “It was [...] who first asked [...]” like “It was Euclid who first asked whether ultra-embedded, normal triangles can be studied” in the accepted paper.

Just a brief comment – sometimes references a doubled, like [10,10]. I don’t know whether it’s done on purpose.

Yep, that’s a known bug… but with a context-free grammar, the code has no way to remember that it just used that reference and shouldn’t repeat it.

I’m in for $10 if you get that kickstarter going!

Pingback: В математический журнал приняли к публикации сгенерированный компьютером бред | На пульсе времени

Pingback: Nonsense paper accepted by mathematics journal

Pingback: Mathethread - Seite 7 - inWoW.de - Das deutsche World of Warcraft / Mists of Pandaria Forum

Magnificent.

So many sets are a subset of the empty set… o_O

Consider using Indiegogo, in addition to or instead of the previous commenter’s suggestion of Kickstarter.

This gives a whole new meaning to Lie algebras [\rimshot]

Pro journal doesn’t bother to read the abstract. That’s all it takes to tell that this paper is fake, though the title should have roused suspicions.

Great article! I just had one quibble:

“As an open access journal, APM naturally imposes a “processing charge” on its authors”

Actually, there are plenty of open access math journals that don’t charge fees to publish. The Directory of Open Access Journals includes information on which charge and which don’t. Here’s their directory for math:

http://www.doaj.org/doaj?func=subject&cpid=58&uiLanguage=en

For those that do charge, yes, you have to do some due diligence to determine whether a journal or publisher has real quality standards or is just trying to get as much money from would-be authors as they can manage. Your peers who have had OA publishing experience probably have gotten a good idea of who’s worthwhile and who should be avoided.

There’s also at least one association of OA publishers, OASPA, that requires some basic standards from its members; while folks have had some issues with how well the standards work in practice, they do tend to keep the purely vanity presses out. Their list of membership at

http://oaspa.org/membership/members/

notably does not include this particular journal’s publisher.

With this, the journal has lost the reputation that it never had.

Exactly. The “journal” is a pretty well-known scam. So the situation isn’t even remotely analogous to the humiliation of “Social Text”–a prestigious PoMo journal–and, by extension, PoMo crap as well.

Do some fundraising for the publication charge. I pledge $10.

Pingback: Mathematics Journal Approves Paper Filled With Computer-Generated Gibberish Equations [Science] | Orange Claymore Red Slime

Pingback: What your tuition buys: randomly generated math paper accepted by journal | EduBubble

Awsome code. One criticism; it doesn’t accept names with Apostrophes in it. It wouldn’t take W. O’Shey as a name, had to change it to “W. Oshey”.

Yeah, sorry about that. The web interface has pretty restrictive validation. If you download the code and run it yourself, you can use any author name you want, including arbitrary $\LaTeX$.

Mathgen is nice, but it can’t beat humans for generating truly great nonsense papers.

How about writing one by hand? I suggest including this in the abstract:

“In his seminal paper from 2011, Euclid first asked how many sets are a subset of the empty set.”

I feel like someone just told me, “I trolled wikipedia, and the wikipedians welcomed me to the community even as they reverted my edits for being stupid.”

If you read SciRP’s about us page, it is clear that they are making trade-offs with the goal of maximum accessibility. I think that’s laudable, even though we shouldn’t ignore the costs of this approach.

If you read Newton’s work from a modern perspective, personally I find it impenetrable. My knowledge of calculus and algebra doesn’t help me. He uses arcane-seeming triangle geometry trivia for every single proof. If I have the time, I could read the Descartes and unravel it. But if I didn’t, I would instead write “this doesn’t follow obviously from earlier statements” in red ink all over the paper. Newton’s work is fantastic, but without the time to research his (now arcane) background, it *does* come off as largely unsupported garbage. And he was relatively concrete compared to today’s pure mathematics.

So I find their criticisms are perfectly on point. Since they were providing a free service to you, they did not have the time to research your citations. In service of their goal of accessibility, they didn’t reject your paper merely because of their unfamiliarity with your citations. Instead, all they could tell you is that they would publish your paper if you edited it so that it made sense. You were unable or unwilling to edit it to make sense, so you were not published. Their system worked exactly as intended!

The \$500 is a red herring. Most professional organizations that support peer reviewing charge more than \$100 per year just to join the club. You wouldn’t have even been permitted a seat at the table, no matter your bona fides.

Unlike SciRP, which ultimately rejected your meaningless story because you were unable to meet their terms, slashdot has actually published your meaningless story. Congrats, you successfully trolled someone, but it’s not who you claim to have intended to.

That’s what I thought reading the “acceptance” letter. Based on the grammar, it’s looks like the reviewing has been outsourced…

Outsourced? To me it looks like it’s also autogenerated

One problem with your criticism: you seem to misunderstand how peer-reviewed articles are supposed to work. It is not Wikipedia. A true review, by qualified peers (usually two to three, not just one), would have instantly shot this down. This clearly bypassed that system, and it is that system that we have relied upon, entirely, for maintaining the quality of published scholarly work since (at least) 1665. Anyone can put a PDF on their website, or on another website, but publishing in a peer-reviewed journal is supposed to be a much higher standard to pass (unless of course, like this journal, the publisher happens to be a complete fraud).

In addition, from your comment, “The \$500 is a red herring. Most professional organizations that support peer reviewing charge more than \$100 per year just to join the club. You wouldn’t have even been permitted a seat at the table, no matter your bona fides.” It is clear you have never submitted a paper for review to a journal. Submission is free, and does not require being a member of a professional organization. Sadly, most publishers do charge some publishing fee, and \$500 is not at all an unreasonable price for a reputable journal.

lgstarn is flat-out wrong. Most reputable math journals do not charge a publication fee. A reprint charge is voluntary and is not a fee required for publication. Reputable journals that do charge a publication fee make it voluntary, knowing that some authors can’t afford it, though some can.

Does this mean that Mathgen has passed the Turing test?

That would imply that a human actually read the paper.

Turing test is meant to be interactive.

Your robot has passed a stripped down (non interactive) version of Turing test.

This however is also yet an advance.

Even if they used the exact word “accepted” in the email, I think it’s premature to claim victory. They demanded revisions, and it ain’t really accepted until they say they are satisfied with the revisions and they don’t require more.

Matt is correct. I’m not sure it’s what they meant, but the letter’s wording is careful enough to leave the matter open

Nate, this is great. I’d chip in on Kickstarter… After reading your post, my colleagues are debating the feasibility of a similar generator for the biomedical literature.

Sadly, I’m also getting a Mathgen error: Seed invalid, must be numeric. Might it have something to do with the antiquated version of IE required on my work laptop?

Here’s the url:

http://thatsmathematics.com/mathgen/paper.php?nameType%5B1%5D=custom&customName%5B1%5D=C+Miller&nameType%5B2%5D=famous&nameType%5B3%5D=famous&nameType%5B4%5D=generic

Thanks for the report. I think this is fixed now, so you could try it again. Generally this is related to broken, disabled or incompatible Javascript, but Mathgen should now work even without Javascript.

Well – that’s really great but there is a very small detail that keeps puzzling me… I wonder how was the list of references generated. This is not pure curiosity; it happens that I am an associate editor to the ONLY existing journal that appears on the list: Journal of Operator Theory. It is a rather respectable journal, but now I begin to ask myself what we did wrong

“Any similarity to real journals, living or dead, is purely coincidental.”

Journal titles are generated from a few different templates, one of which looks like “Journal of AREA”, and among the possibilities for “AREA” is “Operator Theory”. Since real journal names are chosen with only slightly more creativity than my algorithm, there is always the possibility of a collision. I assure you that no slight was intended!

That’s pretty cool. I’ve spent some time in my career writing software that writes software, but not something to write a paper. I wonder, if given sufficient iterations, if some actual advancement in math could be stumbled upon in this fashion? x monkeys * x typewriters * 1,000,000 years = Shakespeare?

Like with the Shakespearean scenario, this leaves the actual work with those finding the wheat amongst the chaff. Like the Maxwell demon, they are the actual source of enthalpy rather than a mere arbiter of it.

Wow you actually did it. I remember we talked about this long time ago at UCSD.

I’m actually surprised at how specific the referee comments were. This in fact raises my opinion of SCIRP. I would have figured they just send a form acceptance email with no substantial comments with just a link where to send the money. This means someone ACTUALLY LOOKED AT IT. This is far more than I would have expected.

Best,

Jiri

Their response generator just needs some work.

Maybe one of us could help them out with that, for a small consideration.

I agree with Jiri. Someone looked at it. The editor may not have read the referee’s comments before or after incorporating them in the letter.

Hoople is definitely in northern North Dakota and not southern North Dakota

Pingback: Math journal publishes computer-generated fake paper « Ted Bunn’s Blog

Pingback: Computer-generated nonsense accepted for publication by a mathematics journal | Comedy Comments

Pingback: Mathematics Journal Approves Paper Filled With Computer-Generated Gibberish | Gizmodo Australia

Pingback: Is math gibberish? « Resource Room Dot Net Blog

I know nothing about maths, but this also happened in IT security a few weeks ago, when the hakin9 magazine published a nonsense generated paper with scigen (i read you forked it): http://seclists.org/nmap-dev/2012/q3/1050

Good job Nate, this made me laugh a lot The best way to find scammers.

Pingback: Math journal accepts computer-generated nonsense paper — Feederoo.com

I am a little surprised that they refereed the paper *before* asking for the loot!

I must admit that I once submitted a paper to an open access journal. I got the request for $500 about two hours later. At that point I retracted the submission, and realized that I had done something rather foolish.

This is fantastic. Some colleagues and I made up a paper in the humanities and have been trying to get it published (it is, of course, complete nonsense). We have been rejected twice. But if a software program can generate a paper and get it accepted, we have faith. We will press on!

pls surf the website: http://www.scirp.org.

magnificent! I’m busy padding my CV with some of these high-quality publications

Pingback: Incomprehensible Bullshit « azizonomics

Pingback: Math journal accepts computer-generated nonsense paper | It's like, Really?

Hey, Nate, long time no see. This is super-awesome! I’m glad I ran across this. Hope things are going well for you.

Well, MathReviewGen also works fine !

Dear Nate,

This blog post is getting media coverage as “legit math journal accepts fake math paper” (see for instance the post on boing boing).

Which is a bit dishonest; I think Advances in pure math is a fake math journal, borderline spam in fact.

Pingback: Academic scam (on bogus/fake research publications) « Budianto "EonStrife" Tandianus

Pingback: La gran chapuza: “Esto es ortogonal estocásticamente…” « :: ZTFNews.org

I suppose “Advances in Pure Mathematics” uses a similar kind of algorithm to generate the reviews…

First of all, congratulations for your fine work. This blog is really something. But I would like to know if it is possible to create a new version of Mathgen where the user can choose key-words. That would be very helpful for future prank submissions.

Pingback: I Just Wrote a Mathematics Journal Article! – Thinking Christian

Pingback: Incomprehensible Bullshit |

Brilliant!

This suggests that we should test journals’ refereeing procedures – perhaps by submitting papers that prove results that are well-known to anyone reasonably expert in the field. I don’t mean something like the Poincare Conjecture, but some small technical result. Or perhaps better, proving something that is contrary to a well-known fact. Anyway it should be something that any competent referee could reject at once without wasting any time.

I looked up “reputable” to see if it has any additional meanings I wasn’t aware of. There weren’t, but the orthographically similar words “rebuttable” and “refutable” were suggested, so I suppose it was a typo.

Pingback: Why Author Pays Open Access is a Bad Idea | Not Even Wrong

Pingback: "El artículo" de MathGen: la historia de cómo se la han colado a una publicación científica - Gaussianos

I used to attend presentations by Prof. Schickele of USND-H, and one time (late 70s maybe) there was a slide show in which he showed us pictures of the campus. All of the pictures were of the most forlorn-looking ramshackle barn you ever saw, taken from various angles, except for one which also featured an equally forlorn-looking chicken coop (which he described as “the Administration Building”).

Pingback: Computer-generated paper accepted by a journal | Pawel Niewiadomski's WebsitePawel Niewiadomski's Website

Wow! I’m impressed that the journal actually took your paper seriously, and that is a clear sign that the journal is not serious itself. Yet, the paper was not accepted. It’s likely that the reviewer and the editor were just being polite (and greedy…). At least once it’s happened to me that an editor gave me a “half-acceptance” like this, if I carried out some “minor” modifications the reviewer was requiring. The paper got finally accepted, but the “minor” modifications took me more than a month of full-time work. It’s not obvious that you would be able to actually get the paper published.

A great job indeed!

We should try similar experiments with other nonsense journals and expose them.

Pingback: THE MAGNET IS ALWAYS ON -- a cognitive neuroscience blog -- The giant soda ban as epistemological crisis

Pingback: Thank but no thanks: fundraising the publication fee | That's Mathematics!

Pingback: Why Marcie Rathke is not Alan Sokal | That's Mathematics!

Pingback: The Absolute Bottom of the Publishing Barrel | Healthcare

Pingback: Assorted Links: Prorogation, Universities, Online Education, and Publishing | Opinion - Policy Nexus

Pingback: Artigo falso gerado por computador é aceito para publicação em revista científica de matemática « Condição Humana na Modernidade

Awsome work!

I found interesting the fact that also the references contains fake journal title and in some cases cite previous work that does not exists.

In fact it will be interesting to create a software that tryto recconaise this fake papers

“Some”cases???If someone needs a software to recognise this sort of papers, then this person should not be an editor.

Pingback: MUNDO DISCORDIANISTA E FNORDS: Artigo falso gerado por computador é aceito para publicação em revista cientí fica de matemática « O Discordianismo…

Even some “reputable” editors can get caught

The paper

http://www.math.hkbu.edu.hk/~ttang/newspaper/funnyarticle11.pdf

got accepted….

You can see the full story at

http://retractionwatch.wordpress.com/2012/04/17/math-paper-retracted-because-it-contains-no-scientific-content/

That seems to be a follow up of the first author’s previous paper

http://www.sciencepub.net/researcher/0105/research0105.pdf#page=62

Pingback: Publishing « genomeek

The author should be ashamed of himself and you guys for encouraging him. The amount of spurious internet content generated by this spoof could exceed that generated by Piltdown man or cold fusion! Conway-d’Alembert’s conjecture alone gives over 1,500 hits just today.

Indeed. Actually, 1501 with this one!

OK perhaps not the Piltdown man hoax but “Cold fusion hoax” at 28,100 could be met. Misinformation is the bane of the internet. In years to come people will come to believe that even Conway-d’Alembert,s conjecture must exist because it has so many internet references

Pingback: On “provisional acceptance” | That's Mathematics!

Pingback: Stephen Huggett on the publication of a randomly generated paper | Open Access

Pingback: Rank zero » Blog Archive » Zweitverwerter (LXVI)

The publisher has already quite a record of misbehaviour. For instance, the article of Yash Paul, Wali Mohammad Shah und Gulshan Singh “Integral mean estimates for polynomials whose zeros are within a circle” [Appl. Math., Irvine 2, No. 1, 141-144 (2011; Zbl 1219.30004)] was found to be largely identical to [

A. AzizandN. A. Rather, J. Math. Anal. Appl. 289, No. 1, 14–29 (2004; Zbl 1040.30002)]. This may, of course, happen also for serious journals; but after notification, there was no retraction or at least a comment linked with the published version, but a lonely, almost invisible Statement of Priority” at the title page of a following issue with the incredible claimPingback: Cứ trả tiền là được nhận đăng, hay là câu chuyện về tạp chí toán học “hề” « ZetaMu

Pingback: Found while foraging (November 3, 2012) « Inspiring Science

Pingback: Quelli che “le Peer Review…” | Critica Scientifica - di Enzo Pennetta

Pingback: All That Glitters Is Not Gold, But Is It Diamond? | SCRIPTed

Pingback: Mein erstes Paper als Wirtschaftsphilosoph | Wirtschaftsphilosoph

Pingback: Mathematical Jargon | Rutgers University Graduate School-New Brunswick Graduate Student Blog

Pingback: Another Mathgen paper accepted | That's Mathematics!

Pingback: Math journals and the fight over open access « mixedmath

Pingback: Desvendar Mistérios – Artigo falso gerado por computador é aceito para publicação em revista científica de matemática

Pingback: Somewhere else, part 17 | Freakonometrics

I would happily chuck in $10 towards publication costs

Thanks, but see this post.

Pingback: hoax paper published in some journal.. hilarious

Pingback: Published: Stupid Manuscripts | Space&Aliens

Pingback: Gaming the system: How to get an astronomical h-index with little scientific impact | The Mad Scientist Confectioner's Club

Pingback: An intriguing invitation from Advances in Pure Mathematics | That's Mathematics!

Pingback: Meer geld voor onderzoek nodig? | The Logic Free Zone

Pingback: Scientific paper easter eggs | The Mad Scientist Confectioner's Club

Pingback: Linkage | An Ergodic Walk

Pingback: Peer Review and Open Access in the Headlines Again | Douglas Whitaker

Pingback: Procedurally Generated Books, Writing, and Language. : Simulate

Pingback: A shocking number of academic journals have accepted studies that are totally fake – Quartz

Pingback: Chcete být profesorem? Stačí zmáčknout tlačítko! | pravy prostor

Pingback: Academic journal recognition, not of a very savoury kind « Robin Kok

Just had a look at the publisher’s website and was surprised that still after this became public the business seems to flourish even in mathematics.

For instance, they claim to have published a paper by C. J. Mozzochi of Princeton

http://dx.doi.org/10.4236/apm.2014.45021

(who has published respected results, e.g.,[H. Iwaniec and C. J. Mozzochi, J. Number Theory 29, No. 1, 60–93 (1988; Zbl 0644.10031)

which however just starts with a misprint in the title (“Condtions”) and continues with a somehow strange introduction going media in res (without much explanation of notation etc.). However, the following content seems to be serious mathematics.

Any idea what’s going on there?