Unfortunately, I was not able to present this little study at this year's #ESSV conference. If you missed me in Eichstätt, you can check out the paper now! 🤓 We show that #LLM based ASR or phonetic transcription struggles with novel phonetic forms. #CRC1646 #LINCC

Nach den jüngsten Diskussionen um Arbeit in der Wissenschaft war meine erste Interpretation des Titels “Jetzt aussteigen oder für immer alt aussehen” ein Aufruf an alle älteren “Nachwuchs”wissenschaftler*innen den akademischen Zirkus zu verlassen 😳 #IchBinHanna

EXTENDED DEADLINE FOR REGISTRATION: May 31st!

Very cool study! Sad to see this bias against diverse speaker groups, again and again. We found a similar disadvantage for non-standard speech with multi-ethnolectal speakers in German.

The Call for Special Sessions is now open, until June 1st 2026! Please head to the ICPhS 2027 website for more details! ➡️ icphs2027.ca/special-sess...

Don't forget to acknowledge your invisible colleagues!

You've seen the meme that Gen-Z is like Christopher Columbus? Unpopular opinion: It’s the same with a large part of current #LLM “research” – always discovering something which has always been around and known to people for ages, selling it as new … and misinterpreting the discoveries. You agree?

Mitmachen! Forschung unterstützen! ...und vielleicht sogar Spaß dabei haben! 🤓

It’s a very interesting programme. My course on sociophonetics will deal with multiethnolectal German, German rap and #Creativity. You definitely need to check this one out! #Kiezdeutsch #Deutschrap #phonetics

Very fascinating findings! Does this suggest that – within exemplar-theoretic models with a perception-production-feedback-loop – once activated, ego-exemplars do no longer compete on higher levels of abstraction during self-monitoring and correction?

Nice! What one considers a “pretty crazy theory” is of course totally subjective. Dependency parsing was introduced into (pre-LLM) NLP as a hack, because computational linguists struggled too much with constituency grammars. No one else really cared so much about dependencies. Pretty crazy, right?

Congratulations to Dr. Simeon Junker, from the CLAUSE group at Bielefeld University, who successfully defended his doctoral dissertation on the topic “Visual Context and Communicative Goals in Referring Expression Generation”! 🥳 @clausebielefeld.bsky.social

My quote of the day: »be the Madonna in the crowd not the phone zombie …dance like nobody's filming it« 🥳

»Wer was echtes will, der kommt zu mir Gütesiegel Straßendeutsch, es gibt kein'n echteren hier« ― Twin. Familienangelegenheit.

»Das aktuelle Wissenschaftssystem schätzt Selbstausbeutung bis zum Zusammenbruch, einen gesunden Umgang mit Arbeit hingegen überhaupt nicht.«

Too slay for syntax? Join my course on #sociophonetics and discuss #Deutschrap , #Kiezdeutsch and #Creativity at the #LINCC #DGfS Summer School in Bielefeld! 😎 Registration is possible until April 30th, 2026! www.uni-bielefeld.de/sfb/sfb1646/...

Sampling ist ein wesentliches Element im HipHop. Schon vor einigen Jahren habe ich in einem Seminar über diesen besonderen Rechtsstreit #Pelham vs #Kraftwerk aus der Frühzeit des #Deutschrap gesprochen. War damals schon alt. Wie es aussieht, nimmt diese unendliche Geschichte kein Ende. 😯

I'm very excited. I am teaching again! My first seminar since the pandemic. My first seminar since ChatGPT. I'm curious, how much the world actually changed. 🤓

Familien mit kleinen Kindern in Berlin gesucht. Melden Sie sich gerne.

Interessiert an einem Stipendium für die #DGfS Sommerschule 2026 in Bielefeld? Die Bewerbungsfrist für Nachwuchs aus Deutschland ohne Förderung der Heimatuniversität oder anderer Geldgeber endet am 15. April 2026. www.uni-bielefeld.de/sfb/sfb1646/...

😂

Freue mich schon auf die P&P an der Friedrich-Schiller-Universität Jena 😎 www.gw.uni-jena.de/98626/2026-p-p

3-year full time position in phonetics and speech technology at Bielefeld University (Germany), suitable for PhD candidates: jobs.uni-bielefeld.de/job/view/487...

Only 499 days to go! 🤓 @icphs2027.bsky.social icphs2027.ca

Before we forget: of course, CLAUSE Bielefeld is also at @eaclmeeting.bsky.social 2026 in Rabat/Marocco! Our group is involved in a main conference papers:

Thanks to the amazing dedicated efforts of community-minded colleagues over many years, the International Speech Communications Association (ISCA) maintains the largest open archive of research papers in speech communications www.isca-archive.org

Are you interested in how people #cheer for long distance runners? There are some specific #prosodic patterns, nicely explained by Marzena Zygis from #ZAS also for laymen. run.outsideonline.com/road/road-ra... If you want to read the paper, it is here: www.degruyterbrill.com/document/doi...

Wie zum Beispiel diesen Kurs hier: Sociophonetics: The Creative Use of the Multiethnolect in German Rap. 😎

There's nothing new under the sun

There's nothing new under the sun

»Not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain―that is, not only write it but know that it had written it. No mechanism could feel […] pleasure at its successes […]«

@lingucat.bsky.social #Tristaben share.google/XL41WYsPXtiQ...

Is the bouba-kiki effect evolutionary so much older than language? New research with just-hatched chicks “…makes the question of bouba-kiki being a solution to language evolution less interesting because it is prelanguage,” comments Aleksandra Ćwiek. 🤯 @acwiek.bsky.social

“…invoking linguistic determinism to explain LLM performance isn’t profound. It’s outdated. If we want serious conversations about AI and reasoning, we need to start from what linguistics, neuroscience, and cognitive science have already clarified over the last half century.” I totally agree!

My word of the day: »fašnik« — /ˈfáʃník/ m noun, short rising accent, Croatian, Kajkavian, from German „Fasching“, meaning carnival. At first glance, this is an example of relatively simple metathesis. However, -nik is a common masculine noun suffix. The word thus does not feel like a loan.

Do you have a favourite letter? Behold: Latin Capital Letter Sharp S: ẞ This is my favourite German letter (used in German only ― though not in Switzerland). Its codification fixed some orthographical problems with names, for example. Cute footnote: (lower case) ß is called ringel-s in Dutch.

»fair play must be given to the machine […] if a machine is expected to be infallible, it cannot also be intelligent.« — Alan M. Turing, 1947. Turing asked why we should allow for occasional “blunders” from human experts but not from machines. (cited by A. Hodges, The Enigma, 2012:361)

🚨 Reminder 🚨 The call for games for #DGKL2026 is still open until February 25th! Do you have a game you developed for e.g. teaching linguistics, participant engagement in experiments, or else? Let people try your game in Bielefeld! 👉 CfG here: t1p.de/opy89 #linguistics #langsky #BLinguistik

“If you knew how something ended — would you still begin?” D. W. Moore's beautiful summary of Ted Chiang’s “Stories of your life” Just re-read this short story of dread, despair and memories of joy… and linguistics. Highly recommend!

Habt euch lieb! Nicht nur heute.

Don’t forget to enter the Koala Competition for your chance to win a free registration for #Interspeech2026! Competition entries close 26th Feb 2026, with the winner to be announced on 17th March interspeech2026.org/en-AU/pages/...

The #Interspeech2026 deadlines are approaching, with the call for papers closing on 25 Feb 2026. Log in to the submission portal early, and check out the welcome message and instructions in advance! interspeech2026.org/en-AU/pages/...

Very proud and happy about all my collaborations so far ― past and present! I learned so much from all of you. #CRC1646 #LINCC

Very cool way to learn more about spectrograms by having fun with squiggles and blobs. 🤓 Another fascinating video. As always!

#Throwback2016 LabPhon15 at Cornell University. Fond memories of collaborations past. Presented joint work with J. Roux, A. Coetzee, J. Bruni and my late Doktorvater G. Dogil on devoicing in Tswana. With A. Schweitzer and N. Lewandowski we showed that personality affects convergence. What a time!

More upcoming: At this year’s #ESSV conference, I will present a little study where we test some ASR systems with co-authors Laurens Winkler and Sina Zarrieß from the #CRC1646 #LINCC discussing “How well can LLMs handle novel phonetic forms?” ― www.essv.de/essv2026 @clausebielefeld.bsky.social

Upcoming: At this year’s #ESSV conference, I will present joint work by Robert Fromont, Jennifer Hay, Daniel Duran, Allie Osborne, Melanie Weirich, Miriam Oschkinat & Stefanie Jannedy: “Evaluating Full Automation of Computational Sociophonetics for the #Plapper Corpus” ― www.essv.de/essv2026

Fritz Seebauer from the #Bielefeld University #Phonetics group successfully defended his doctoral dissertation “Synthetic Speech Evaluation - Quality and Context”. Congratulations Dr. Seebauer! 🥳👨‍🎓

#Throwback2016 Buenos Aires, Argentina: International Congress on Acoustics. Amongst hardcore acoustics with numerical simulations — no LLMs, not even deep learning — me with a humble 3D computer game promoting #gamification in #phonetics research. I dare say, I was a professional game designer. 😎

Ever wanted to find out whether Bielefeld really exist? 🕵‍♀️ this is your chance! Two days left to submit an abstract for #DGKL2026 👇 conferences.uni-bielefeld.de/frontend/ind... 🐦🐦 #BLinguistik #Langsky