In the movie Angels in the Outfield, Tony Danza plays a washed up pitcher named Mel Clark. In an early scene, one of the kids goes up to him and says, “you used to be Mel Clark!” He responds with, “yeah kid, I used to be.”
Language is beautiful.
It would do some people good to remember that the [saɪ] in cogsci stands for science and not psychology
I’m more sympathetic to (b) as lexical exceptions and abstract representations are weakly equivalent and I have no problem with the latter given the analysis justifies it. In this case, the independent ban on short-long vowel hiatus and short vowels at end of words makes abstraction enticing
The joy of abstraction
Liberman & Prince's explanation of stress being in different locations for "radio" and "Ohio" being due to an underlying length contrast in the penult is so beautiful. Positing two underlying short vowels at the end of "radio" that lengthen based on other phonotactic facts of English...a masterpiece
Berent, Iris. 2026. Three arguments for abstraction in phonology. Glossa: a journal of general linguistics 12(1), 1–24. DOI: doi.org/10.16995/glo...
#linguistics #DiamondOA @openlibhums.org
picture of sign in front of parking garage that says “LOT FULL”
me when my head is full of innate concepts:
Three years ago I pointed out some things that were bad about AI. Basically none of them have changed. Many things have gotten worse. Among other things, the attack on labor has intensified. Here's why to resist dehumanizing your work: www.acsu.buffalo.edu/~cxjacobs/Bl...
I do
Sitting in the coffee shop listening to Mogwai and writing lecture notes.
Sitting in the Champaign twilight and listening to Built to Spill.
Unsurprisingly I had the exact same reactions to reading the abstract/skimming. I should also read the full thing, but initial feelings are a big ¯\_(ツ)_/¯
Isn’t this just generative grammar-informed psycholinguistics, which has always assumed a role for statistical learning, and at least some associationist mechanisms? It’s more common to find gen grammar making use of usage-based insights than the reverse, mind you (in my experience).
I’m hiring an 18-month postdoc to work on physics-informed machine learning for acoustic-articulatory speech inversion at
@phoneticslab.bsky.social
🗓️ Deadline: Friday 10 April.
🔗 More info & applications: hr-jobs.lancs.ac.uk/Vacancy.aspx...
📣 Please share with anyone who might be a good fit!
Phonological rule application is a type of constraint satisfaction.
Hey now, some of us even go as far as mumbling about Lakatos!
I once “joked” that all papers should include essentially this type of statement, but I think it would really help reduce a lot of pointless arguing both in person and in the literature. Though at times it seems like that’s the fun for people ¯\_(ツ)_/¯
Half baked thought: most work on phonological functions in the so called "subregular" space assumes each function is total over an arbitrary input space Sigma*. This is more or less Richness of the Base.
In less grumpy news, I just found out the undergrad whose thesis I'm advising got accepted for a poster presentation and a grad student I'm advising got an oral presentation. Happy for them and 2/3 (with a potential for a fourth) from my research group is a win!
This is the sign to just go to Chicago and have fun instead of going to do more work
Anyway, you win some you lose some, but I was excited to go to CLS now that I'm living in Illinois. Maybe next year.
Not to brag about my native language writing skills, but I promise this was not a problem. The score is whatever and the reviewer said they didn't judge the abstract based on writing, but like when the other two reviewers gave high scores and thoughtful comments...¯\_(ツ)_/¯
CLS reviews are in -- whoops got rejected. Check scores -- strong accept (5); probable accept (4);. ... marginal rejection (2) with additional comment, "please consider recruiting a native speaker to double-check your writing for idiomaticity and legibility in the future." Sick.
every ai company logo
I know Element Theory is mostly an ignored dark art in North America, but if anyone has recommendations, particularly about a specific ET paper on phonetic implementation, I would be happy to have somewhere specific to look.
I’m about 60% done with Phillip Backley’s “An Introduction to Element Theory”. I get the general gist and was hoping there would be a bit more technical/formalization discussion in the book. We’ll see if the back 40% delivers.
After spending the last 4-5 years thinking about Articulatory Phonology and it’s relation to symbolic theories of phonology, I am now setting my sights on Element Theory which is clearly acoustically motivated (necessarily different than being auditorally motivated for them from what I’ve read).
A search function for the Library of Babel is so funny to me. I guess it’s to
show each book has a “real” location. But showing a single result defeats the purpose. Anything you search should have infinite results! It should make me go insane having to wade through gibberish!
in some ways, the thing that is most annoying to me personally about this discourse is that a lot of people seem to be of the view that you can just play with LLMs *instead* of doing philosophy of mind. i'm pretty pessimistic about that approach
100/10