Josef Fruehwald
banner
jofrhwld.bsky.social
Josef Fruehwald
@jofrhwld.bsky.social
i’m a linguist. he/him https://jofrhwld.github.io/
Pinned
Check out the docs & intro tutorials for new-fave! #nwav52
new-fave
forced-alignment-and-vowel-extraction.github.io
fuck it. i’m catholic again
November 22, 2025 at 2:33 AM
imo, this is still one of my greatest theoretical breakthroughs
Josef Fruehwald : Prolegomena to a Grammar of Knitting
YouTube video by Linguistics and English Language at the University of Edinburgh
youtu.be
November 22, 2025 at 2:08 AM
the questions i get about whether or not we’ll have class the monday before thanksgiving presuppose I have a lot of autonomy about when and whether class meets, and am not, say, obliged to show up on regular working day and do the job i’m paid for
November 21, 2025 at 4:33 PM
I’ve hardened against “I don’t have anything against individuals just have reasonable concerns about procedure” after watching it transform into frothing hateful bile in real time.
November 20, 2025 at 8:03 PM
its 2025 and we're attacking AIs with poetry
Looks like LLMs are *very* vulnerable to attack via poetic allusion: "curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90% ..."

https://arxiv.org/html/2511.15304v1
November 20, 2025 at 7:25 PM
for some reason I just scrolled LinkedIn for a minute... and I just can't care about my worksona enough to post there
November 20, 2025 at 5:50 PM
{ggarrow} is such a nice quality of life package for plotting vowel trajectories
Arrows for ggplot2
A ggplot2 extension that adds specialised arrow geometry layers. It offers more arrow options than the standard grid arrows that are built-in many line-based geom layers.
teunbrand.github.io
November 20, 2025 at 3:44 PM
you really hate to see it
November 20, 2025 at 1:42 PM
i feel like it’s very easy to be a normal & imperfect person and not wind up in the epstein files
November 20, 2025 at 1:23 AM
maybe i’m not practiced, but OT feels a *lot* harder to do live at the board
i'm going to tell them about optimality theory today
November 19, 2025 at 9:20 PM
i’m seeing if it lives up the the hype
November 19, 2025 at 7:54 PM
i'm going to tell them about optimality theory today
November 19, 2025 at 3:06 PM
considering embedding this in my CompLing syllabus next semester
Bob Ross - Talent is a pursued interest
YouTube video by Iceberg Studios
www.youtube.com
November 19, 2025 at 11:36 AM
not only did i tell my class about [spʌiɾəɹ] 🕷️ today, i also told them about [tʌigəɹ] 🐅. my belief is that this is an infohazard for creating linguists
November 17, 2025 at 9:17 PM
/oy/ is like the alien DNA in our vowel system
vowels are so good
November 15, 2025 at 11:41 PM
vowels are so good
November 14, 2025 at 10:35 PM
I've gotten some questions along the lines of "When does it make sense to use these DCT parameters instead of just a gam?" So I've written a little blog post: jofrhwld.github.io/blog/posts/2...
How and when to regress over DCTs – Væl Space
jofrhwld.github.io
November 13, 2025 at 7:54 PM
my biggest pet peeve is when an R package does something really cool and makes a plot, but doesn't have a straightforward way to access the data underlying the plot
November 13, 2025 at 3:15 PM
Reposted by Josef Fruehwald
🤏🤏🤏
November 10, 2025 at 9:14 PM
I do try to read widely, but somehow didn't realize that when I published "Generations, lifespans, and the zeitgeist" that psychology calls this type of analysis "Age-Period-Cohort".

But tbh, I like mine better.
November 9, 2025 at 4:30 PM
Reposted by Josef Fruehwald
what star wars really need is a HUNT FOR RED OCTOBER style submarine thriller
November 8, 2025 at 6:48 PM
I think i’ve started seeing “R people”
as maybe an algospeak camouflage for “republican”, and as an Rstats person it’s been confusing!
November 8, 2025 at 3:54 AM
Reposted by Josef Fruehwald
RIP Dick Cheney you would've hated all of this
November 5, 2025 at 2:17 AM
heard an actual [o] at the detroit airport
November 4, 2025 at 9:16 PM
fielding questions about laryngeal realism in class

Q: Is it very uncommon for people to argue [voice] is monovalent?
A: It's not like a fringe theory or anything.

Q: Ok, so people won't, like, argue about it.
A: Oh, they'll *argue* about it!
November 3, 2025 at 10:35 PM