Whenever I look up anything on multinomial regression... (1) Uh, we should have used it, but we binarized everything instead (2) Here it is! Run this code. What's a random effect? ...

There's a lawsuit about AI stealing your work. It's the same lawyers taking on Elsevier et al in a separate case. Academics: 1. Check if your work is in LibGen at www.theatlantic.com/technology/a... 2. If so, let the lawyers know at www.lieffcabraser.com/anthropic-au...

Two UB biomedical engineering students from Sri Lanka were taken by ICE in Williamsville today. @ubuffalo.bsky.social

Our short vacation in Rhode Island has been nice, but the food here is really underwhelming. Newport is just expensive and mediocre. It could be good if it weren't so white everywhere.

Nice article by @jrkasstan.bsky.social

The magenta dragon came out well. It's hard to get a good angle of the eyes though.

Husband sees flashlight on Kohl's website (he is addicted to Kohls cash) only to discover that Torchy300 is underwhelmingly smaller than advertised.

So, with the orange enemy-of-the-people firing the director of the bureau of labor statistics, what public sources on the US economy should the engaged public utilize instead? Aren’t there institutes who track these things themselves? Advice @epi.org, @economist.com? #econsky #publiceconomics

While in Oregon for #lsa2025, I basically didn't look at social media. (My therapist also said that would be great.) Coming back after a few weeks, I'm realizing just how many academics have really unhealthy media consumption habits and unhealthy work-life boundaries. We're not good models.

Accepted at Language & Speech. Work on the Triqui corpus with my student Jared Sharp. www.researchgate.net/publication/...

Math professor Johanna Franklin writes home to Waterloo, IL, where a 4th grade teacher sparked her math career. Cutting the NSF, “means not only fewer scientists and less scientific progress in general, but also fewer chances for anyone but the rich to do science.” 🧪 🏠 @johannamath.bsky.social

My husband got me the gayest socks.

Should I cite what formal phonologists say about articulatory timing if they have not actually looked at phonetic variation and production in any quantitative capacity? The field often weighs this work as more meaningful since it’s “clean”, but it’s clean because they don’t look at production.

It's promotion letter writing season for academics in the US. As I'm going through my batch I thought I'd share a few things I did *not* know when I started doing this. (I've served on dept and college's PTR committees, but institutional norms may differ--I'd be interested to hear other takes!)

As a prep for my last lectures on phonetic fieldwork at the LSA summer institute, I created a youtube video guiding users through how I (quickly) mark phonetic variants in recordings. I’m happy to send along the Praat script (or similar versions on my website). www.youtube.com/watch?v=cggx...

#lsa2025 is fun so far. I've met up with several old friends and colleagues. I talk about linguistics all the time! Also, living with Georgia is fun. Our vrbo is amazing.

Join us this summer — 80+ courses, 8 conferences, and linguists surrounded by Oregon’s nature. Don’t miss it! 📅 Sign up now: bit.ly/3CSbbHC 🔗 More info: center.uoregon.edu/LSA/2025 📧 Questions? [email protected] #LSA2025 #Linguistics #SummerInstitute #UOregon

It's way too early but I'm at the airport for my flight to Dallas/Eugene for the LSA. I think I assigned too much reading for my class tomorrow. I am looking forward to seeing many people.

I have published papers w/out ever first haven given a talk on the topic. The next paper may be similar. I wish there were a cheaper venue to more regularly get feedback on phonetics research. The ASA is expensive or inconvenient. The LSA is useless bc no phoneticians go. We need some other venues.

Below is cepstral peak prominence, plotted by time for different human-tagged glottal categories. This reflects about 6,000 Triqui tokens of non-glottalized (coded NA) vs. post-glottalized vowels (/aʔ/). This plot demonstrates a LOT of important findings from my work with a grad student. 1/n

One of the many disappointments of 2025 is that like no famous musician has decided to write music to capture the zeitgeist of the moment. Give us protest songs, mfers. This sh!t writes itself. Only film and television celebrities so far, like the fabulous @janefonda.com or @markruffalo.bsky.social

I'm working on the second phonetics/phonology chapter of my grammar and I'm at the part where one would normally include a vowel chart. I have kind of three options, as I see it, but I'm not sure what to do. One is that I take old recordings and plot vowel midpoints from a few speakers. 1/n

Take my course at the LSA summer institute this summer! (I actually have no idea how many people are going to register.)

Is hot. Resting. Resting. Sleep. Curl up.

Chatbots — LLMs — do not know facts and are not designed to be able to accurately answer factual questions. They are designed to find and mimic patterns of words, probabilistically. When they’re “right” it’s because correct things are often written down, so those patterns are frequent. That’s all.

Why are so many linguists so bad talking about empirical results? They spend pages perfecting statistical modeling, but basic observations are shoved aside. The effect of this (intended or not) masks observed patterns and effect magnitudes - it’s no better than p-hacking imho.

We went to the New Haven march yesterday instead of the one in Buffalo, but if anyone wants their daily dose of raised /æ/ [eə], watch the channel 4 broadcast and discussion with protesters in Buffalo. #linguistics www.wivb.com/news/local-n...

Exactly. This.

On two separate occasions recently, I have had to contend with the assertion that AI will largely take over many of the tasks I do as a linguist. On both occasions I have been left in a kind of “limbo” when I think about research on endangered languages (most of what I do). 1/n

I do not believe the review process is the place for the author(s) to try to resolve a terminological debate that reviewers are obsessed with - especially if the author(s) are using terms as they are defined in previous literature.

Is anyone interested in investigating how breathy voice and creak influence acoustic measures of nasalization in Triqui? I have controlled recordings looking at this (from 12 speakers) and it’s actually an open question in the acoustics literature. 1/n

The nice thing about only ever looking at mutuals on bsky is that I can be like "that's social media for the day then!" after scrolling past 40 posts.

Is it possible for someone to show up in immigration court remotely via Zoom? I don’t know if such a thing was possible during COVID - or if some safeguard exists like having the person show up at their lawyer’s office for a Zoom call with a judge.

One weird thing about listening to people discussing AI in a lexicography conference is that the tools they propose often assume corpora of 100 million words or more. My reaction on the “advice” they give at the ends of their talks is "so this will never ever be useful to field linguists.”

My talk at the 25th DSNA conference (in Buffalo). This is somewhat more accessible to non-linguists (but those interested in dictionaries). I am happy that I get to complain about FLEx to people who care. I'm hoping I learn about good alternatives. #linguistics www.researchgate.net/publication/...

Cave trolls are feisty. Bring a weapon or run. About 90 minutes painting. The body color was surprisingly hard to mix. #ttrpg

So true! 👇👇👇💯 www.theguardian.com/wellness/ng-...

🚨 A new rule would let career scientists like NSF/NIH program officers be replaced by political appointees Already 14,000+ public comments, deadline is Friday 📣 Comments can be short. Courts consider them—and scientists with NSF/NIH experience are especially impactful Speak up! shorturl.at/WKuBj

What a fun birthday!🎂 www.dworkinsubstack.com/p/mass-prote...

La lingüística sería más interesante y util si las cuestiones normalmente vinieran de una perspectiva de diversidad tipológica. Es fácil enfocarse en cosas más y más específicas en lenguas bien habladas y presumir que sabemos mucho cuando es un saber eurocéntrico. Sigo creyendo que no sabemos mucho.

While I’m not at #tai2025 this year near Munich, here is our talk on (a) the limits of what intonation is and (b) why it seems Triqui completely lacks it. Co-authored with the talented @richardhatcher.bsky.social, who just presented it in Germany. www.acsu.buffalo.edu/~cdicanio/pd...

This is your reminder that this intense week of grading and research is best powered by chocolate and coffee.

The world is destroying me and the things that I used to find so interesting in research and #linguistics. There is a general sense that we’re supposed to be resilient or have a thick skin or what-not. If we don’t, we fail. What if we’re just failing? Am I then responsible for my own destruction?

I don't recommend breaking a finger. I think the worse outcome for this is needing to be on antibiotics for the next 10 days.

I believe I mistakenly assumed that my registration for TAI 2025 was also a hotel reservation because I received an email with “my booking”, though that was completely wrong apparently. If I can find some place else to stay in Herrsching am Ammersee then perhaps I’m still going. Otherwise, 🤷‍♂️.

Springtime haikus christiandicanio.blogspot.com/2025/04/spri...

The program for PHUNY 2 is now up. sites.google.com/view/phuny2/...