Join Ed King for his tutorial on sound manipulations as part of the Phonetics & Phonology Workshop. It will take place in the Greenberg Room from noon to 1:15. Regarding the tutorial, Ed writes:
“This tutorial will go over some sound manipulations that are commonly used in linguistic experiments. Topics will include basic manipulations of pitch, intensity, and duration, and move on to continuum creation for categorical perception work (especially fricatives, vowel formants, and nasality). Some familiarity with phonetics and audio software will be assumed. I’ll be using Praat, but will be stressing higher-level concepts that should be applicable to your software of choice (and I’ll give some Praat background just in case).”
See you there!
Uriel Cohen Priva will be presenting at this afternoon’s Colloquium. Come to the Greenberg Room at 3:30 to hear his talk, “Preferential treatment: Explaining why languages play favorites with their sounds.” You can read the abstract below:
Investigations of language change are invaluable for understanding the physiological and cognitive forces that shape language. But there are many unexplained mysteries about language change, such as the repeated and independent reduction and deletion of some sounds in certain languages. For instance, in American English /t/ can be deleted word- finally (as in the casual pronunciation of “jus’” for “just”) and several varieties of English phonetically reduce the sound /t/ to a tap (the /t/ in American English “butter”) or a glottal stop.
I present a new model, MULE (Most information Utility, Least Effort), which shows that language-specific phonetic reduction of sounds follows from the balance between information utility and effort avoidance. Read the rest of this entry »
Beginning tomorrow morning and continuing until Sunday evening, the Berkeley Linguistics Society will be holding its 38th annual meeting. You can find a program and any other information you’re looking for at the Berkeley Linguistics Society website.
Come to the Humanities Center, Room 153, on Monday at 12:30 for lunch and Tom Griffith’s (Berkeley) talk entitled “Monte Carlo as Mechanism: Sampling and Human Cognition.” The abstract is below:
Human behavior is consistent with the predictions made by Bayesian models of cognition across a wide range of problems. This raises an interesting question: How are people solving these problems, given the computational challenges posed by Bayesian inference? When we look at the distribution of people’s responses on specific tasks, we often see that this distribution is similar to the posterior distribution produced by applying Bayes’ rule. This phenomenon – which we call “probability matching to the posterior” – suggests a possible answer to our question: That people are approximating Bayesian inference by producing a small number of samples from the posterior distribution. Recent work in computer science and statistics has resulted in a number of sophisticated Monte Carlo methods for approximating Bayesian inference, such as importance sampling and particle filters, which we have explored as candidate explanations for human behavior. Considering the unique constraints of human cognition has also led us to develop novel algorithms that have surprising properties, such as a sequential Monte Carlo scheme based on the “win-stay, lose-shift” principle. In this talk, I will present the results of experiments with both adults and children that explore the potential of the Monte Carlo principle as an account of human cognition.
Christine Sheil writes:
“Our first meeting of the semester is tomorrow from 3-4:30 in 1229 Dwinelle. Adrian Brasoveanu (UCSC) will be presenting his research on the Pragmatics of quantifier scope, based on corpus study.”
You can find the rest of the UCB Syntax Circle’s schedule at their website.
On Monday the 13th, Tanya Nikitina will visit the Spoken Syntax Lab meeting (2:15 in Margaret Jacks 110) and give a talk based on her fieldwork: “Clause-internal correlatives and other unique constructions of Mande: searching for an explanation of typological rara.”
Join Scott Grimm for his talk “Grammatical number and individuation.” Come to the Greenberg Room at 1pm on Tuesday to hear all about it.
But stay tuned for announcements about the next meeting, which will be on Feb 28.
From Tyler Schnoebelen:
Here’s what’s been happening on the corpus linguistics blog since the beginning of the year:
An introduction to using the (super-awesome) Corpus of Contemporary American English. It’s online, easy, big, and you should be using it. The intro looks at “What a __!” phenomena:
Getting started in CHILDES, with examples of how moms, dads, boys, and girls differ in their uses of “big” and “little”:
Resources on Arabic, African languages, and Native American (especially Californian) languages:
You can also find popular posts from last year: topic-modeling erotica, the Sarah Palin of the Canterbury Tales, sentiment corpora, cultural keywords, Tgrep2 basics/alternatives, and Robin’s guest post on dealing with knockouts in Goldvarb.
Speaking of guest posts–have a corpus, a tool, or a technique to share? Or have corpus questions? Email TylerS at Stanford.
HELSINKI — A former finance minister won Finland’s presidential election Sunday and will become the country’s first conservative head of state in five decades.
Sauli Niinisto took 63 percent of the votes, compared to 37 percent for his rival, Greens candidate Pekka Haavisto, official results showed with 100 percent of ballots counted.
Both Niinisto and Haavisto entered Parliament in 1987. They come from affluent backgrounds, share a gentlemanly manner and, in true Finnish fashion, were not provoked into confrontation during debates.
For more details, see the Washington Post article.
(Thanks to Arto Anttila, fearless Finnish correspondent)
