Mark Dingemanse

We live in capitalism. Its power seems inescapable. So did the divine right of kings. Any human power can be resisted and changed by human beings. Resistance and change often begin in art, and very often in our art, the art of words.

Retracted for 'unverified refs'? Strange. This 'flaw' was found only after a Jan 8 Bluesky raid by Gender-biased reviewers + non-academics. Mass Reposts in 1 hour... Does this remind you of something?

it found me too, but I've blocked preemptively because I cannot take the withering heat of a roast identifying my gender-biased bluesky raids in defense of the status quo of arbitrariness

"artifacts" of a "translation" in the LLM era, uh-huh

Some tools for thinking about AI in a time when the companies making them are trying to destroy teaching and learning itself...

1/n 🧵

RETRACTED ARTICLE: A cross-linguistic investigation of /h/ symbolism: the case of H2O Rasheed AL-Jarrah Humanities and Social Sciences Communications volume 13, Article number: 64 (2026) Cite this article 3043 Accesses 47 Altmetric Metrics details This article was retracted on 23 January 2026 09 January 2026 Editor’s Note: The Editorial team is currently investigating concerns raised about the contents of this paper. Editorial action will be taken as appropriate once this process is complete. This article has been updated

aaand it's retracted

"The Editor has retracted this article following concerns about its academic content confirmed in post-publication peer review. Further investigation revealed the presence of references that could not be verified.

The author does not agree with this retraction."

Accountability gap This was not a case of losing random notes or idle chats. Among my discussions with ChatGPT were project folders containing multiple conversations that I had used to develop grant applications, prepare teaching materials, refine publication drafts and design exam analyses. This was intellectual scaffolding that had been built up over a two-year period.

"This was intellectual scaffolding"😢 — Bucher

"A mind propped up by a text generator is a mind populated by premade thoughts and prepared only for the most probable patterns. … In this sense, text generators don’t aid original thinking but are the antithesis to it." — ideophone.org/human-creati...

It's such a shame. I've formally requested they disable this for my papers on the grounds of inaccuracies. No action yet but they do say this is a possibility

Collective Action In Science Diamond Publish one diamond open access article in 5 years

#cognitivescience Don't forget our pledge to publish one article in a scholar-led diamond open access journal in five years is still live. Were half way through! After another 250 sign ups the pledge will activate!

freeourknowledge.org/2024-01-30-c...

"In time, many may come to depend on text generators to streamline prose, co-write computer code, and cover for gaps in their knowledge. The excitement of initial speed gains will obscure that this is at root a process of deskilling (with attendant opportunities for large-scale monetization)."

Header of blog post with image of Emilie du Chatelet and the heading "Human creativity in the age of text generators"

In early 2023 I pitched this piece at a couple of glam journals ideophone.org/human-creati...

I guess it was not AI-credulous enough 🤷
It turned out to be prophetic though.

"We are currently witnessing the first cohort of people tempted into treating text generators as thinking companions."

esp because the sob story in Nature doesn't even begin to touch on the deep data protection and intellectual property violations this guy subjected his colleagues and students to — all he can think of is his chat history, which he calls his "intellectual scaffolding" bsky.app/profile/ding...

bsky.app/profile/ding...

What amazes me most about this case is the brazen admission of incompetence, data protection violations, and research misconduct. This guy just shares articles, grants, email correspondence and student data (!) with OpenAI for training and only regrets not backing up his precious chat history?

The talk of the town in conversational interfaces today is undoubtedly ChatGPT, an instruction-tuned text generator that impresses many because of its fluid prose. Yet striking new capabilities should not detract us from the risks of proprietary systems. Only three months after OpenAI rolled out ChatGPT, it abruptly discontinued API support for its widely used Codex model that had been available as a “free limited beta” since 2021 [ 44 ] — surprising users with only three days’ notice and undercutting at one blow the reproducibility of at least 100 research papers. This is a stark reminder that proprietary systems are designed to offer smooth onboarding and convenience but come at the price of user lock-in and a lack of reliability

As we wrote in 2023: "this is a stark reminder that proprietary systems are designed to offer smooth onboarding and convenience but come at the price of user lock-in and a lack of reliability" dl.acm.org/doi/pdf/10.1...

When two years of academic work vanished with a single click After turning off ChatGPT’s ‘data consent’ option, Marcel Bucher lost the work behind grant applications, teaching materials and publication drafts. Here’s what happened next.

When you think two years of running in the chatgpt hamster wheel is "academic work"

(via @mathijsvdsande.bsky.social)
www.nature.com/articles/d41...

Also FWIW my failed ERC StG in 2017 made it to the interview stage but rubbed a formal pragmatics reviewer the wrong way (I didn't cite enough of their friends and so there was 'no theory')

It's a lottery and you can't please everyone

My Vici (comparable to ERC) features both; text of the grant proposal is at markdingemanse.net/futures/ (check Aims and Strands, particularly Strand A 'Conversational structures')

(I'm too lazy to look up the Unicode for that [] in Pān[]ini just as that copyeditor was too lazy to find the proper Unicode for the combining diacritic below n in Pāṇini)

Q13: AUTHOR: Please provide forename for the author Panini

this reminds me of that time when a copyeditor of a piece in which I cited Panini's grammar asked me to "please provide forename for author Pāṇini"

the hegemony of English is fairly recent, so anyone who wants to dig into foundational issues must work multilingually. E.g. these two paragraphs in my 2018 review of the history of ideophone research cite work in German, Sanskrit, English, Latin, Portuguese and French doi.org/10.5334/gjgl.444

Translation offers one way in (this is how I have read and cited work by Vygotsky, Havránek, Pāṇini), collaboration another (e.g. Japanese through my colleague Akita)

but I am in awe at the worlds of scholarship inaccessible to me because they exist in languages I don't read (yet)

2/2

Linguistics: I cite work that I have read and found useful, regardless of the language. Work in my field is in many languages beyond English, and in my papers I have cited work in English (lots), French, German, and (rarely) Dutch and Japanese.

1/2

Prompt: wat is uw enige troost, beide in leven en sterven? Respons: Je verzoek is geblokkeerd door een contentfilter

ja zeg, als zelfs vraag 1 van de Heidelbergse Catechismus het contentfilter niet overleeft dan is 'ie misschien wel iets té goed dichtgetimmerd 😂

Pro tip for scholars: make sure you have a homepage that is not acdemia.edu,researchgate, or google scholar, and has your correct affiliation and contact information. This is doubly true if you are in a less mainstream institution, between jobs, or other unusual setting.

"The first recorded Lithuanian word, reported to have been said on 24 December 1207 from the chronicle of Henry of Latvia, was Ba, an interjection of a Lithuanian raider after he found no loot to pillage in a Livonian church." (Wikipedia)

#FirstWords of a different kind @michaelerard.bsky.social

Welcome to a fake scientific conference, the best way to extort money from scientists: ‘I feel cheated’ - Vox magazine Random topics and poor presentations: copycat conferences are the latest way rogue companies have found of extorting money from scientists. Vox attended one such conference in Budapest, where several ...

Unfortunately it is possible. Science reporter @stanvanpelt.bsky.social did it so we didn't have to www.voxweb.nl/nieuws/welco...

nice, consider citing that too? 😇

lineage of ideas is important, and there's never any harm in generously citing fellow travellers 😊

(and #CiteWomen)

shitty image taken from spamference website showing pictures labelled as BAHMAN JABBARI (a besuited male with red tie and Yale, USA affiliation) and Carmen Villmann (a smiling blonde woman affilated to University Hospital Würzburg, Germany)

come speak on the same stage* as vertically-squished BAHMAN JABBARI and horizontally-squished Carmen Villmann

look Yale affiliation university hospital status signifier look

* more likely same stuffy ballroom with 77 empty chairs and an audience of the 3 other belanyarded speakers who fell for it

blue purplish gradient AI generated bullshit image with BIG ALL CAPS LETTERS with drop shadow shouting out a mashup of keywords likely to tickle the narcissist impulses of people Thinking Big Thoughts just as they have clicked on the link inviting them to be Esteemed Speaker. The words say "exploring the brain: integrating neuroscience, AI, cognition, and clinical innovation"

and when I say shitty branding I mean it

this is a world where gradients, drop shadows, and all caps headings are still all the rage

the GenAI aesthetic was born for this spirit of "idgaf looks okayish to me slap it on the page already"

Timely.

Consider citing work by Guest et al. that precipitated much of this?

e.g., Guest et al. 2025 'Against the uncritical adoption' hcommons.org?get_group_do...

and consider checking out like-minded critiques foreshadowing these arguments, like Birhane 2021 and Weizenbaum 1976?