Being eviscerated on the battlefield, Russians now have a singular goal of destroying Ukraine to the point that it's as filthy and shitty as Russia - today attempting to blow up the Pechenihy dam in Kharkiv with aerial bombs.
The Trump administration has hit a grim milestone with its 50th strike on a civilian boat in the waters off Latin America. interc.pt/4dIXs5W
Kearny, AZ may run out of water by July 15th. The town council issued an emergency decree in January but water use still went up.
The new order bans “Any…use of water not related to sanitation or use to support life”.
Maybe we should pay attention to climate change after all?
🎉Publication day🎉 for our #CHI2026 extended abstract! I wanted to find out whose voices are commonly reproduced (and whose aren't) in a popular commercial TTS system. Read all about it! w @ninamarkl.bsky.social @lhlew.bsky.social @catlai.bsky.social @chi.acm.org
alice-ross.github.io/publication/...
One day after a temporary ceasefire was brokered between the US and Iran to secure 2 weeks of peace, Israel openly violated the agreement with a terroristic assault on Lebanon killing at least 112 people and wounding 830 more by dropping 100 bombs in ten minutes.
Looks like www.walking4hope.org or #spotthebath isn't on Bluesky yet! It was lovely meeting John Bell and his rubber-duck-covered bathtub yesterday in the centre of Birmingham (though sad as each duck represents a person lost to suicide). Happy to support this excellent charity. #walking4hope
alice-ross.github.io/files/BAAP_k...
Our #BAAP2026 poster, in which we gave some sentences to a popular commercial Text-to-Speech system and asked for voices that were "kind" or "competent" etc. and then analyzed what sociophonetic indexes resulted.
Listen: alias-alice.notion.site/Kind-is-the-...
Those days are gone for me, too! These posts are going to be my only ones, sadly (could only attend today).
Elinor Payne & Angelo Dian discuss intonational rises in Venetan*, which occur in both questions and declaratives, and which confuse the Italians they interact with! Turns out that declaratives can be rises or falls; polar questions also show rise-falls. #BAAP2026
en.wikipedia.org/wiki/Venetia...
Andy Gibson starts off his #BAAP2026 talk by failing to find the lecture hall's resonance 😂
Then he explores the relationship between vowel quality choices and vocal resonance, between more 'speaky' to 'singy' singing styles, at different f0s. He ends by introducing a Praat-based tool for this.
Yanyu Li at #BAAP2026 train English listeners in perceiving lexical tone, using a cue-weighting approach that's previously worked for segmental training. Some listeners were given exaggerated tones and some non-exaggerated tones. Surprisingly, there was no difference, and little learning!
...But this is a little odd because previous results found that listeners were poor at perceiving/processing lexical prominence. So why it is used? (It varies regularly by utterance position, for example.) Consider the role of perception of non-Sg Englishes?
Adam Chong's #BAAP2026 talk looks at "Ethnicity effects in the production of lexical prominence in Singapore English", cos most previous work is on Chinese SgE speakers; there's little data from Malay or Indian SgE speakers. Results: Intonational marking, at a coarse level, does not seem to differ.
Those who use complete bursts are slowest; those who use strategies (like merging the stop with the vowel) are the fastest. There seems to be a speed-accuracy trade-off. Results also show that variability (less accuracy) increases with age (between 5-12), favouring the speed strategy. #BAAP2026
People with Speech Sound Disorders (SSDs) may have accuracy problems with DDK. (It's tough for anyone, really: try it yourself with /pətəkə/!) But what does "accuracy" even mean? What does 'normal' variability look like? Some produce full release bursts, others incomplete bursts; diff. strategies.
Anja Kuschmann et al. (presenter: Jane Stuart-Smith) on "Diadochokinetic* (DDK) rate performance and variability in 5-12-year-old children in Scotland"
*the ability to perform repeated syllables as fast and accurately as possible. #BAAP2026
Data from VariCS: varics.ac.uk
Ianto Gruffydd and Mercedes Durham's #BAAP2026 talk is on pre-/l/ breaking (e.g. "Wales" as "Way-uhls") in South East Wales English. Exploring how to operationalize a 'breaking onset point' as the acoustic measure to differentiate between tokens perceived as having breaking and those without.
Caveat from Liam that this is exploratory and there are a lot of remaining open questions. Questions from the floor are about things that can't be measured by OpenFace, like lower teeth and larynx height.
Analysis of head pose, facial landmarks, and constrictions of facial muscles to explore phonetic correlations. Jaw movement operationalized as chin-nose distance. These results, and results for jaw velocity, correspond to the acoustic differences between the 'straight' and 'queen' styles. #BAAP2026
Liam Ó hír’s talk, “Giving Face and Body: Language, sexuality and bodily hexes in Japanese parodic performance“ is an analysis of Second Street, a gay comedy duo in Tokyo who perform directly comparable gay and straight characters. Love seeing this kind of talk on the big stage at #BAAP2026
Sarah Wang & Claire Nance: /l/-/n/ merger in Chinese. For Chongqing Mandarin L1 speakers, with Standard Mandarin L2, English L3, they didn’t distinguish /l/ and /n/ in their L1, but did in L2 and (even more) in L3. Proficiency modulates how much a speaker extends an L2 contrast to the L3. #BAAP2026
Rasmus Puggaard-Rode gives an interesting articulatory and acoustic analysis of Danish "soft-D" (a vocalized stop). It's articulated like a very front and low vowel, but its acoustics look like a central vowel! (OMG, #Danish is so wild, I love it.) #BAAP2026
...And this difference predicts articulation of English /eɪ/, /aɪ/, and /oɪ/, too! "more dorsal or palatal doming in /i/ associated with earlier formant transitions from the diphthong onglide to the palatal offglide"
Looks like "a form of compensation for variation in vocal tract shape and size."
Two people stand below a projected screen with text on it corresponding to the content of this post.
Starting off Day 2 of #BAAP2026: "Articulatory strategy as a source of variation in acoustic vowel dynamics" by Patrycja Strycharczuk, Justin Lo, & Sam Kirkham.
Individual differences in palate result in individual differences in /i/ production that aren't reflected in acoustics. #phonetics
The richest 0.1 percent of the global population are hiding nearly $3 trillion in offshore accounts to evade taxes — more than the wealth held by the entire bottom half of humanity.
It would take about $93 billion per year to eradicate extreme hunger by 2030.
Taxing the rich is a global necessity.
Cloisonné pin similar to the NASA logo, but reads “NOT FLAT”, with “we checked” underneath l
NASA’s Earth photos from Artemis inspired me to put this on today… 🙂
Getting lots of positive comments on it as I run errands…
📗More book promotion: for the next month, you can download and read Chapter 1 of my & @lhlew.bsky.social 's new book, Sociophonetics: Implications for Phonological and Phonetic Theory, for free! #linguistics academic.oup.com/book/61815/c...
*“The president of the United States would like everyone to know that he is acting with criminal intent, in case there was any ambiguity,” a US law professor said of his social media post with bridge bombing footage*
www.commondreams.org/news/iran-tr...
Dear Christian,
If you’re demanding that the Ten Commandments be put up in public schools, but you’re not holding this president or his administration accountable to a single one of them, please don’t act like you actually take them seriously.
Just out in #earlyview from Amanda Cole, this innovative paper draws on multiple investigation strands to offer methodological advancements in dialect identification tasks, highlighting the importance of social meaning in perceptual processes. Check it out below #openaccess !
