Dan Isbell

Dan Isbell

@danielrisbell.bsky.social

Associate Professor of Second Language Studies, University of Hawaiʻi at Mānoa, specializing in language assessment. Associate Editor at Language Learning journal (https://onlinelibrary.wiley.com/journal/14679922). Views own. https://isbell.github.io/

1K followers633 following1.4K posts

It will be interesting to see if the SKA expands meaningfully. While not quite as bad as the situation of the JLPT for international students seeking to study in Japan, the TOPIK exam is limited in testing opportunities - especially abroad and for the options speaking section.

koreajoongangdaily.comAs demand for Korean tests grows, a new exam is broadening access. But will more colleges recognize it?The Sejong Korean Language Assessment offers free testing and an integrated speaking section, but Topik remains more widely recognized by universities and employers.

Maybe this is a bit salty (I'm not on the list!), but with a new Elsevier Top 2% scientists data release, I'm struck by how many people in my field (applied linguistics) making the list are spamming publications related to AI. I browsed one profile with 12+ pubs in 2026 alone with "AI" in the title.

View full thread

not generating a great deal of insight (they point out more gaps/methodological shortcomings than anything). A couple studies ended up in the retracted GenAI for education meta, too. For the last ~8-10 years, the field seems pretty beset by trendy topics: Translanguaging -> COVID-19 -> GenAI

I know trends have always been a thing. Maybe this one feels so big because my field is larger now (not a bad thing), GenAI is something broadly popular (cf. translanguaging, which doesn't have such cross-disciplinary reach), and most worryingly, AI itself might contribute to the number of pubs.

Maybe this is a bit salty (I'm not on the list!), but with a new Elsevier Top 2% scientists data release, I'm struck by how many people in my field (applied linguistics) making the list are spamming publications related to AI. I browsed one profile with 12+ pubs in 2026 alone with "AI" in the title.

No problem or shame in doing work on timely topics, and collaboration is a norm so co-authoring multiple papers a year is typical. But are all these AI publications generating a lot of useful and actionable knowledge? That's where I'm skeptical. Hundreds or thousands of case studies of teachers...

using AI in creating materials, etc. Hundreds/thousands of surveys done about attitudes toward AI. Probably a hundred studies of AI as raters, hundreds on AI feedback for language learners. To my knowledge, very few replication studies, and meta-analyses/systematic reviews seem to be sparse and

Dan Isbell reposted

Over the last decade, there were two hypotheses: H1: the political right genuinely cared about threats to speech from the left. H2: they were using campus speech wars to impose their values and silence dissent. Today, the answer is clear, but I can't find any of the H1 people admitting their error.

Adam Serwer@adamserwer.bsky.social

After years of championing "free speech," Texas lawmakers have enabled a dramatic rise of censorship in Texas universities, banning concepts, canceling courses and closing departments seen as "woke," and even using AI bots as censors. All in the name of freedom. www.theatlantic.com/ideas/2026/1...

The Iraq War is estimated to have costed the US around $2 trillion to date. I think closing schools to save close to 80,000 lives is a better use of my $6,000. Might not be as cost efficient as stuff like vaccines and seatbelt laws, but saving lives is hard to put a price on.

Michael Hobbes@michaelhobbes.bsky.social

lmao this is such a sick way to conclude that COVID school closures were a mistake www.kuow.org/education/20...

Interesting plans to modernize the TOPIK, including AI item generation and scoring as well as home administration (for lower proficiency levels) and improved security. I hope the AI scoring will primarily be feature-based ML models, though - wholly GenAI scoring systems are sub-par.

www-koreatimes-co-kr.cdn.ampproject.orgKorea turns to AI to modernize TOPIK national language exam - The Korea TimesKorea’s national Korean language proficiency exam is preparing for a major digital overhaul, introducing artificial intelligence (AI) to draft test...

and continually kicked down the road. Other 'hot' languages like Korean may face similar issues in testing - outdated admin/delivery makes upgrades a crisis response and opens up markets for competitors (Sejong Korean Assessment, for one). Foreign-made tests are also options...

ACTFL exams, for one, are available much more readily, though considerably more expensive and rely on English knowledge for directions and questions. But I'd imagine some enterprising test developers could build a Japanese equivalent to their flagship English test.

The JLPT is not a 'small' test! It probably has more yearly test takers than Duolingo and comes close to TOEFL. I am not so familiar with the business side of things, but it seems like there has been some complacency - upgrading delivery and administration infrastructure seen as too big of a lift

Without disagreeing, we often also lack the infrastructure/resources to meaningfully bring undergrads into humanities/soc sci research. In many disciplines we lack actual 'labs' with ongoing, large-team projects that trainees can join as well as money to pay them.

One of the most unfortunate consequences for this is that doing research becomes a norm for undergrads in STEM but an exception for undergrads in the humanities. By claiming to “help” students by putting “pedagogy” before “research” we actually do them a disservice by drawing an artificial boundary.

My uni has a program and modest funding for undergrad research/scholarship/creative activity, but it seems to work best when a student can be brought into a physical/bio science lab. Folks in many humanities/soc sci disciplines would have to *come up with new/tailored projects* for mentees.

That makes the whole endeavor a massive investment of time and effort. Most of my projects involve small teams of colleagues/grad students, and on top of that I supervise loads of grad student research. Maybe it's a me problem, but coming up with and supervising even more projects is beyond me.

I don't want this to come across as victim blaming, but this story also highlights why it just isn't worth being on X. The reach isn't worth the Nazi trolling, because you're mostly reaching Nazi trolls.

Very good news. The proposed rule has had universities, from the top all the way down to individual departments, scrambling to figure out how to continue supporting international students. Good to be relieved of all that... for now.

Josh Gerstein@joshgerstein.bsky.social

BREAKING: Judge Dennis Saylor (GWBush/MA) temporarily blocks Trump admin drive to impose a 4-year limit on student visas now issued for length of study. Calls the rationale for the new rule "exceptionally weak." Doc: www.documentcloud.org/documents/28...

Dan Isbell reposted

A recurrent theme in this excellent story about China’s progress is that people there regard Trump as essential to their success, including his ability to drive scientific talent from the US www.newyorker.com/magazine/202...

Michael Clemens@mclem.org

The US President's new proposal to tax international students has been approved by the White House for release this week. It's an unprecedented $100,000 tax on int'l grads who stay to work after graduation on Optional Practical Training (most grads do), a crucial bridge to remaining by any channel.

I grudgingly use LinkedIn because it is useful for connecting with people in my field and keeping up to date on relevant industry news. But man, sometimes the slop and spam on that website...

Dan Isbell reposted

I can’t wait for the 18 part series in the Atlantic about “The Crisis at Business Schools” which will detail of every instance of academic fraud at the nation’s top business schools, the coverups, lack of accountability, and their implications for business ethics and academic standards at B schools.

View full thread

What matters more, for communication and for digesting input to support more implicit learning of syntax, is vocabulary and decoding. Rote memorization and deliberate practice works for establishing those initial form-meaning connections!

And apps (and textbooks, and classroom instruction) can give lots of give lots of useful opportunities for learning and practice. They won't be enough in the end, but can definitely be a useful start and continued supplement. Just be sure to add in other activities and resources.

View full thread

Some purists in language teaching like to point out that you can never really develop mental representations of complex grammatical phenomena (read: syntax) through explicit instruction and deliberate practice, whether it comes from a textbook, teacher, or app. They're probably right! But...

some of those things don't matter so much from a communicative standpoint. I communicate, daily and successfully, with people whose L2 English grammar doesn't match my L1 grammar on things like plurals or articles. And my continued flubbing of Korean particles also doesn't matter much.

I'm glad about the subtitle for this article - and I'm someone who readily acknowledges that yes, you can learn a language (to some extent) from an app. But to really reach higher levels, you'll need more. And interacting with other humans in a new language is going to be more personally rewarding.

Part of why apps work is, in my view, because they deliver structured opportunity to get familiar with the forms of a new language. Even if they can get a bit rote at times, or turn into mindless games, but they still give you something that leads to usable knowledge of the language.

Sad news for #linguistics. Lots of post-grad earnings pressure from (and soon to be on, from federal govt) universities is influencing programs and curricula, with language and linguistics getting pinched at many institutions. (ling/lang refugees welcome here at the University of Hawaiʻi)

ntdaily.com‘My program’s going to sh-t’: Linguistics students move forward knowing major is being ‘phased out’A small program at the university, the linguistics department has taken pride in documenting endangered languages and preparing the next generation to research language.

Dan Isbell reposted

One of the first "outrageous" claims I saw as Jason Arday's media hounding unfolded was that he ran 30 marathons in 35 days! Wow! What a thing to claim! And the newspapers all suggested there was no evidence he actually did it. Well, they clearly didn't look very hard 🙄 🧵 1/26

Dan Isbell reposted

Me, a US-based academic: I don’t care what any scholar believes, we all need to stand up publicly whenever someone’s academic freedom is threatened Me, continuing: also, the first time this has happened, motivating me to sign my first ever open letter, is in late 2026, to a race scientist, in Ghent