Christian DiCanio

cdicanio.bsky.social

26 September 2026 03:03

I feel like perhaps I am weird in literally not using any AI tool for anything I do. I have no AI in classes I teach either. I decided a few years ago that I am not going to adjust how I teach or do research to accommodate tools promoting intellectual laziness. Every bit of my content is old school.

Christian DiCanio

cdicanio.bsky.social

07 September 2026 04:25

Yep. In many families (Na-Dene, western Otomanguean, Daghestanian), the contrast is aspirated-plain-ejective/glottalized. Lots of languages also have no contrast (aside from place) in stops. The acoustic-perceptual space around stop production has always been bigger than just VOT.

Christian DiCanio

cdicanio.bsky.social

06 September 2026 03:07

I’ve had a general concern over the past several years that the availability of acoustic analysis suites is having a negative effect on understanding of the acoustics (or of the parameters that are important). Formants are not always accurate. Neither are many other things.

Christian DiCanio

cdicanio.bsky.social

06 September 2026 03:07

I think that there are a lot of important acoustic-phonetic and articulatory-phonetic contributions, but the difficulty is that there are fewer people doing a kind of pure phonetics nowadays. People at Haskins have contributed a lot here, as did earlier phoneticians. But nowadays?

Christian DiCanio

cdicanio.bsky.social

06 September 2026 03:04

Counterpoint - we turn to studies measuring VOffT instead: voice offset time during obstruent production. It’s usually more interesting for processes of lenition.