‘Publication Laundering’

James Heathers
15 min readSep 28, 2022

--

‘Proceedings’ journals are how publishers maintain plausible deniability to let fake science thrive

Depending on who you listen to, 2.5 million to 5 million published academic documents are produced a year.

A lot of them aren’t much good. Maybe most of them.

Or, maybe it would be fair to say: they are not designed to add to the global pool of knowledge. They exist simply because academic jobs measure success via the simple metric of ‘produce words in order and put your name on them’. If that is your remit, why make the words any harder than they have to be.

As a consequence, a bad scientific paper only needs to be minimally cogent. That is, it can’t quite be a shopping list (‘eggs, milk, flour, copy of O Magazine where Oprah will tell you how to gain the strength of twenty longshoreman through rectal cleansing’), but it certainly can be a SciGen paper,

or a Google translated Wikipedia entry,
or the paraphrasing of a different document entirely,
or vaguely-intelligible blather with topic-relevant words salted through,
or a section of the Unabomber’s Manifesto,
or your own work remixed just enough to defeat plagiarism detection.

This is a perfectly adequate amount of face validity to enter a low quality journal.

(There are no rules at all, of course, if a journal is *completely* fake rather than merely bad — the second chapter of War and Peace with random swaps to the nouns and verbs would do, where we get to find out ‘the discotheque at Barry Malone’s was in full custard when Harry, Viscount McGinty’s parrot, turns up to take her crochet needles to another cemetery, and all the fish notice how scrofulous she is’.)

The problem is: having produced smouldering nonsense rather than flaming nonsense, what should we do with it? Improving the text takes time — better ideas, more writing, more time — so that’s possible, but irritating. The trick is finding the soft spots within the publication process such that external credibility can be accrued by said nonsense.

Now! Academics love talking about incentives, and in this case, you could argue that the publisher’s incentive of smouldering nonsense is quite well aligned with the authors of nonsense.

The authors need to put words in order with their name appended. This ensures their employment and promotion. So far, so regular.

But publishers serve their shareholders by [a] publishing increasingly more work over time to cement market dominance, and then some combination of [b] being paid to publish it and [c] bundling it and selling access to it.

So: [1] It appears there is no reputational threat to publishing nonsense, because it can be intermixed with legitimate work and titles, and no-one is paying attention. Likewise, the landscape of global publication is very large, and the publishers largely control the parsing of, access to, and searchability of both legitimate AND nonsense work.

AND, [2] if a problem does arise that cannot be minimized or ignored— and we are being maximally cynical — we can see that the publisher has ensured they are maximally capable of blaming someone else. It is always possible to throw your hands up and say “But, Yer Honour, I Had No Idea That These Papers Were A Problem! I’ve Been Taken Advantage Of, Swear On Me Mum!”

Means [3] there is great scope to justify both higher payment volumes (nonsense-in) and higher bundle prices (nonsense-out).

Combine a pay-to publish service or a papermill, which sells authorship to authors, with a publisher who charges them for services, and you have something uncannily close to money laundering.

Let’s look at an example, which might prove to be an interesting wrinkle on the paper mill / fake paper scam.

AN EXAMPLE!

^ Materials Today: Proceedings publishes not articles but conference proceedings. A submission to the journal is not an individual paper, but rather the sum total of all the papers presented at a conference. The conference applies to publish the proceedings, not the individual authors. They elect a topic and Guest Editors within that topic, and those editors handle the peer review externally to the journal.

The conference representative signs an affidavit that states:

Those of you familiar with research integrity issues might see the risk here: science has a long-standing problem with fake conferences. Every academic has a long list of spam emails inviting them to conferences that don’t exist, or do exist in such a poorly organised state that they don’t count. These became notorious a few years ago when there was a good article written about what one of these was actually like to attend.

And we have reached my assertion: many of those conferences exist not to scam attendance fees from participants, but to provide cover for a ‘proceedings submission’ — which allows the organizers to sell ‘peer reviewed’ papers easily.

This feels particularly straightforward given that we have completely normalized the virtual conference. One ratty-looking Zoom call, and you’ve got a ‘conference’.

So, here’s how I would scam this Proceedings journal in order to sell paper mill products:

  • Hold a conference, fake or real or somewhere in between. Assert a topic, a location (virtual would make this even easier), prepare a shell website, assign guest editors, etc.
  • Write to a Proceedings journal and request they publish your conference submissions. Shove your tongue firmly in your cheek, and tick all the boxes that promise you’ll follow the process, peer review the submissions, vet the editors etc. as above.
  • Write a series of smouldering nonsense papers.
  • Sell authorship to those papers.
  • Salt the duff papers through the real conference submissions if there are any (and presumably any real/fake admixture is possible).
  • ‘Complete’ the conference, and batch all the submissions up nicely with the right formatting, send to Proceedings journal.
  • Wait for publication.
  • Collect money from ‘authors’ upon publication.

This is a very attractive way to ‘launder’ publications. The items themselves are published closed access, which lowers scrutiny — they can be published OA but that costs money (and that’s giving away profits). As above, the publications get the ‘peer reviewed’ label because the journal trusts you to do yourself (and you do not, or you produce a fake back-and-forth exchange if that is required). Post-publication scrutiny lower as your paper is hidden with many dozens of others turning up all at once. The indexing is good because this is a proper Elsevier journal, and you’ll end up with a ‘real’ paper in Scopus or similar.

This reproduces the typical paper mill scam, but is easier to navigate.

And there is one great benefit if you’re running a papers-on-demand scam: the timing of the publication, very uncertain in traditional peer review unless you also control the reviewers, is now under your control, and you can provide a much better service (“Article acceptance 15–30 days from submission”).

Yes, I have receipts. Let’s get started on them.

1. A lot of content has nothing to do with materials science

Remember, this is a material science journal — ‘ the understanding and application of the properties of matter’ — and the only part of the proceedings that receive any scrutiny appear to be the titles.

As might be expected, then, the titles of the conferences all make perfect sense. An example:

International Conference on Latest Developments in Materials & Manufacturing

Perfectly logical conference name, and fully congruent with the publication condition “only material science papers will be submitted to the journal”.

Unfortunately, the content of these conferences is HUGELY variable. Here are some of my greatest hits from tooling around the search bar, some of them from the conference above.

The design and fulfillment of a Smart Home (SH) material powered by the IoT using the Blynk app

Removal of copper ions from the Contaminated soil by soil washing technique

An innovative hybrid technique for road extraction from noisy satellite images

Computer vision-based surface roughness measurement using artificial neural network

Malware analysis: Reverse engineering tools using santuko linux

Study of antibacterial effect of the hydroalchoholic extract of Teucrium polium against clinical strain of Gardnerella vaginalis in vitro

Phytochemical assay on leaves, bracts, and flowers of Bougainvillea spectabilis and isolation of phenolic materials from bracts

Labor relation in construction industry in India

These have NOTHING to do with material science. Zero. Zilch. Nada.

It is as odd as finding a paper on sustainable fisheries in a psychological journal, or, if you prefer, finding a paper on the psychometric properties of IQ scores in a aquaculture journal.

The search counts are even wilder.

There are ~35K conference papers in Materials Today: Proceedings. Many appear to be on material science topics. But also within those:

There are 550 entries on FISH. From material science conferences.

There are 355 entries on COVID. From material science conferences.

Herbal soup. Whole body vibration. Laser measurement of pumpkins. Detecting financial fraud. Terrorism and its effects on tourism. Noise pollution. Social media sentiment analysis.

What does this tell us? The scrutiny of individual papers is effectively zero. This may have changed over time — perhaps the stipulation in the above contract agreement may be new or recent. Regardless, it is reasonable to expect that a conference on a topic would be… on that topic.

(I’m not cherry-picking. If you just go to the journal’s home page, two out of the three most recently published papers are, respectively, about cryptography and information security!)

2. Many of the papers are obviously nonsense

Quick caveat: I would submit that clever people in non-English speaking countries are quite capable of both ascertaining what is good or bad about a scientific paper and writing one.

However, I would bet you A Big Money that they haven’t written these papers. They are purchasing items here from a full service industry, and that includes writing a paper badly (a.k.a. as fast as possible) for you.

So: in calling work under their name total shit, this is not a reflection on the intelligence or ability of the authors. Their honesty, maybe. If they had any interest in the topic itself, they would be perfectly capable of writing something better.

Here are some examples:

Overview of risk factors in the use of antihypertensive biomaterials drugs in medical prescriptions

A drug utilisation search is also packed with information about the effectiveness of drug use, as well as information about the drug itself. Prioritization is effective when it is used to facilitate the allocation of resources in a sensible manner, resources to assist you in maintaining or improving your fitness. Men and women always had a risk of hypertension. The current inquiry’s main purpose is to present an overview about the antihypertensive drug exploitation sample, threat part drug detection evidence in the manufacturing and the identification of the threat. It provides richer insights into individualised but premium pharmacological care of hypertension by examining prescription trends and use of antihypertensive medicines.

Put to one side the fact that this, again, has nothing to do with material science whatsoever, and look at it on its merits.

It is, in a word, shit. A wordy and circuitous description of a process, not an outcome. It has some of the hallmarks of Tortured Text (‘drug exploitation’? Is the senior author “Tarantino, Q”?) It is the smouldering nonsense of before.

Robots and its types for industrial applications

Robots are playing a very important role in industries. The main task in the robotics, is to reach the end effector at desired position. Robots are used in every sector like hospitals, mining, domestic purpose, industries etc. The inverse kinematic robotics problem has proved to be of great significance because the solutions found provide control over the position and orientation of the robot hand. In this paper, a brief review of robots and its types has been discussed. The materials used in the robots is discussed in this paper.

Same again. Again, no material science content (the last line not withstanding, which amounts to ‘robots have materials in them’). Again, just empty verbiage.

One more for luck.

Vitamin D and fertility

Vitamin D is one of the steroid hormones. Approximately 80% of vitamin D comes from skin production under the effect of sunlight. A small portion of vitamin D also comes from food and/or dietary supplements. Vitamin D is well known for its role in phosphocalcic metabolism and bone mineralization. Several studies suggest that vitamin D has reproductive effects in women and men. Indeed, vitamin D receptors (VDRs) and enzymes that control vitamin D metabolism are expressed in both female and male reproductive organs, which may suggest that vitamin D may have effects on sperm quality, fertility, and conception rate. This article presents a description of the function of vitamin D in female and male fertility.

This is more cogent, but somehow even more irrelevant.

And I reiterate:

Material. science. journal.
Material. science. conference.

3. Papers being sold on the open market end up in the journal

This is straightforward to understand.

Here is a Facebook marketplace selling paper slots.

Now, here is the paper itself.

These title-to-later-paper pairs are often hard to match up, because it is best practice for the scammers to only give the research area a later paper will appear in, or list a title and then change it — otherwise, you can do what I did, and simply search title listed in the For Sale ad and find the eventual publication.

However, here they slipped up and did not change the title enough, just fiddled a few words. Or were just never expecting any scrutiny.

You can also see the classic hallmarks of an auctioned author list, which is where all the authors are from different institutions across multiple countries. Also, we should respect the fact that this paper mill provides great service! They are producing the goods, and keeping closely to their promised timeframe.

Here’s another one from the same screengrab.

(This one fulfils all three of my criteria: It is not material science, it is nonsense, and the title is listed on an open market marked as a submission to the journal pre-publication.)

Conclusion

This is just one example from one journal. I don’t have the time or the energy or the support to kick all the beams in this house and find out which are rotten with termites. I haven’t done any citation analysis to see if these papers come with ‘citation agreements’ (later papers which cite them as part of the ‘package’). I don’t even know how many other journals are which have similar policies.

But: I think we need to appreciate just how well a journal which publishes conference proceedings wholesale and without scrutiny allows a perfect sweet spot to launder fake papers. If other proceedings journals operate under the ‘no editorial staff, no copy editing staff, outsource peer review, a signed affidavit is good enough’ model, then we have a gigantic problem.

Out of all the money laundering methods, I see the strongest parallel to the ‘chain of cash-only businesses’ technique, mixing the bad money in with the good. Scams are rarely new, they are usually new wine in old bottles, and people who run scams are clever. If a scam makes instant intuitive sense to you, because you see hallmarks of it elsewhere, you can bet they’ve already thought of it.

And this makes all the sense in the world. You have control over the timeframe and submission parameters, a ‘real’ outlet, minimal scrutiny but the appearance of peer review, and very high volume as 50, 100, etc. items can be batched up and published all at once — this is good for either hiding garbage amongst the real submissions, or having a tremendous sales volume available for authors that wish to buy publication credentials.

One curiosity that remains is if paper mills are always involved. There are examples of this being a full-service industry where you buy paper slots (see the FB links below!) but it might be easier to simply be a conference organiser — hold a ‘conference’, register it with a journal, charge quietly for submissions written by the authors themselves, reject unpaid submissions (or maybe accept them, for bulk/legitimacy/cover), then publish wholesale.

Either way, it’s a problem, so let’s talk solutions. A suggested remediation:

  • Price a proceedings submission with an upfront cost.
  • Actually look at the article titles and keywords to make sure they relate to the ostensive focus of the journals. This could be automated. Or it could be very, very fast. Ten minutes per conference, tops.
  • Peer review a random subset of the submissions even briefly. Peer review has many problems, but ‘being unable to tell nonsense about diabetes from real research about material science’ isn’t one of them.
  • Return some or all of the upfront fee upon acceptance so legitimate small, regional and LDC conferences are not priced out.
  • Keep the money if more than X% of the papers are out of scope or obvious crap.

Or, you know, you could read what you publish.

ADDENDA:

[A] Those of you interested in fake paper marketplaces that aren’t the regular commercial websites can browse these Facebook groups at your leisure. I’m sorry the links look ass, but Medium won’t let me change them. Cheekily, these cats often use the phrase ‘Call For Authors’, reminiscent of ‘Call For Papers’, and sometimes list the fees upfront.

If you find someone doing a pay-to-play scam, and you click their profile link, you should be able to see all their posts within the group isolated. These can be … revealing.

[B] Big thanks to Nick ‘Colonel Mustard’ Brown and Gideon ‘Haunted Egg’ M-K for readthroughs and suggestions.

[C] Analysing the content above for ‘tortured phrases’ is an exercise left for the reader.

[D] Another fun way of looking at this problem, if you’re in a hurry, is to simply search the journal articles on PubMed. Now, PubMed is pretty much Science Google for us cats in the biological / life sciences, but for materials science it should show precious few relevant entries.

^ However, at time of writing, this shows 152 entries — which is far more ‘biomedical’ research than you’d ever realistically expect from the proceedings of material science conferences. Probably because these entries are in PMC database or something. Either way, pick another material science journal and you’ll see far fewer entries.

[E] Recently, Nick Wise (who I don’t know — hi Nick!) was at the center of an investigation provoking a colossal 500 retractions. Turns out what I’ve done here is not particularly novel, but more an independent replication of some of those findings.

That is to say: I found out about this after finishing writing the above. I’m not going to go back and editorialize, as that would be deeply bullshit. Instead, I’ll just summarise the findings here given in the RetractionWatch article.

(Frankly, if I’d read this beforehand, it could have saved me a lot of trouble!)

(1) Three journals were involved — Journal of Physics: Conference Series, IOP Conference Series: Materials Science & Engineering, IOP Conference Series: Earth & Environmental Science. Yes, all conference proceedings journals.

Those of you who actually retain information that you read, unlike yours truly, might recognise two of those journals from previous mass retractions stemming from work with the Problematic Paper screener.

(2) The MS&E journal, at least, seems to have exactly the same ‘promise you’ll do peer review, swear on a bible’ trick as the journal here. It is outlined in full here and the conditions are essentially identical to the MS:P journal.

(3) As far as I can tell, exactly the same problems are afoot. Let’s look at two quick examples from “Materials Science & Engineering”, because why break a habit.

https://iopscience.iop.org/article/10.1088/1757-899X/1020/1/012017

Link: https://iopscience.iop.org/article/10.1088/1757-899X/928/6/062030

Yup. Again, we see the same hallmarks. Not Materials Science or Engineering. More like public health. And it is, of course, absolute boiled nonsense. ‘Piece-wise linear regression’ as a novel approach? Anything I could say would be garbled screaming.

But as Facebook search is almost unusable — you cannot text-grab even a little — I cannot find these articles in any paper marketplaces. Maybe someone with better Google-fu will have more luck.

Finally, at the risk of being boring, I reiterate my previous:

“…I think we need to appreciate just how well a journal which publishes conference proceedings wholesale and without scrutiny allows a perfect sweet spot to launder fake papers. If other proceedings journals operate under the ‘no editorial staff, no copy editing staff, outsource peer review, a signed affidavit is good enough’ model, then we have a gigantic problem.”

Gigantic problem confirmed. Now what?

--

--