Professor of Linguistics, research at the intersection of meaning, modalities, and the mind (she/her) ☀️
Joined November 2008
- Tweets501
- Following581
- Followers848
- Likes509
Kate Davidson retweeted
Here's work where we try to take analogies quite seriously, plus our takes about LLM experiments serving to raise the bar for what we are willing to take as behavioral evidence for things like composition 💇♀️
New paper with @najoungkim and @TeaAnd_OrCoffee!
We previously found that humans and LLMs generalize to novel adjective-noun inferences (e.g. "Is a homemade currency still a currency?")
Turns out analogy isn't enough to generalize, so it's likely evidence of composition! 🧵
Kate Davidson retweeted
so proud of @HayleyRossLing for getting a best paper award at @GenBench this year!! 🎉🪅🎉 I'm sure @TeaAnd_OrCoffee would be too :) check out our paper and share if you think homemade cats are cats!
New paper with @najoungkim and @TeaAnd_OrCoffee testing if LLMs can draw adjective-noun inferences like humans! Turns out they often can, and even generalize to unseen combinations. But they're more optimistic about "artificial intelligence" than humans.
arxiv.org/abs/2410.17482
It's been a true honor to advise Dr. @YuhanZhang_ , such a well deserved celebration and exciting work at the psycholinguistics/experimental ling intersection!
🙌🏻 I defended my PhD thesis “The Rational Processing of Language Illusions” yesterday!🎓 I am extremely grateful for my amazing advisors @LanguageMIT and @TeaAnd_OrCoffee and dear committee members @roger_p_levy and Kevin Ryan!
Kate Davidson retweeted
It was really fun to work with Hayley and Kate testing human generalization patterns on adjective-noun compositions! Hayley will be presenting this at @ELM_Conference. Also, follow her for updates on future work including cool comp ling stuff!
New preprint with @najoungkim & @TeaAnd_OrCoffee on fake reefs and other cases of (novel) adjective-noun composition: lingbuzz.net/lingbuzz/008012
Whether a fake N is an N or not depends on noun + context, but people handle novel AN pairs just fine 🙂
Stay tuned for results on LLMs!
Middle of the night #TISLR14 poster presentations "in Japan" is surprisingly fun (although not sure I can do it two nights in a row, and I expect to pay for it later today)!
Thanks to everybody who came to ours (work on locus optionality with Annemarie Kocab and @ahn_dorothy : projects.iq.harvard.edu/file…), and the other poster presenters that I got to learn from 😊
I think this is one of the coolest projects in recent linguistics, combining historical change and sign languages and semantics and just all kinds of insight (invokes pronouns, reflexives, copulas…). And they’re doing a webinar on it, tomorrow!
Dr. Rachel Mayberry and I will be presenting our work about the grammaticalization of SELF across 150 years of ASL in a webinar by @LingSocAm on July 21 at 9 am PST/12 pm EST! Registration link is here! Everyone welcome! register.gotowebinar.com/reg…
Had been looking forward this whole year to give a keynote at @E_SPPMilan2022, to represent Linguistics and learn from cognitive science-oriented Philosophers and Psychologists. Instead, like so many I ended up with COVID this week 😢 (first time, whole family...). That said,
I'm also taking the moment to reflect on conferences. Even though I'll miss this one, I really do value bringing humans together to see each other as people and ask the stupid questions to other fields, especially folks who don't social media (it shouldn't be a requirement...)
Kate Davidson retweeted
The name change is official: I successfully defended my dissertation - “Compositional routes to (non)intersectivity” - yesterday!
Reviewers (aka Gennaro) call it “deeply unsettling” and “1/3 right”, which is all the praise I could ever ask for.
My experience at #hsp2022 today has been that I kept thinking that each talk I was watching was my favorite, and then the next talk changed my mind; currently really enjoying this session on a(h)naphora by @ahnaphor, @kirbyconrod, and @VArvindam
Kate Davidson retweeted
Hi all! Registration for #HSP2022 is now live! 🎉 You can register (and see more about the conference) on our website:
hsp2022.ucsc.edu
Quantification ([e.g. some/all/none/...][X][Y]) is a known source of linguistic/cognitive complexity, and it is present from the earliest generations of Nicaraguan Sign Language: new paper 📝 with Annemarie Kocab and Jesse Snedeker in Cognitive Science onlinelibrary.wiley.com/doi/…
2022's first paper 📝 in Semantics & Pragmatics (semprag.org/index.php/sp) is the result of one of my very favorite collaborations, with Deanna Gagne at Gallaudet. Learn about the expression of quantificational domain information in ASL (More is up!). Open access.