Perception of formulaic and novel expressions under acoustic degradation by native and non-native speakers
Time: 01:00pm - 02:00pm
Place: Woodburn Hall 002
In category: Phonetics and phonology
PraatR: An R package for streamlining the phonetic analysis workflow by controlling Praat from within R
Time: 03:30pm - 04:15pm
Place: Ballantine Hall 015
Aaron AlbinPraat (http://www.fon.hum.uva.nl/praat/) is currently the de facto standard software used for acoustic phonetics research. However, since Praat is designed primarily for digital signal processing, its statistical capabilities are limited. As such, most Praat users export the phonetic measurement data to some static representation (e.g. a tab-delimited textfile), then import this into statistical software, such as R (http://cran.us.r-project.org/) or SPSS, and run the quantitative analyses separately. This process of shuttling data back and forth between the two programs unnecessarily complicates the workflow. For the average user, this means many additional clicks to move the data around (e.g. from Praat’s GUI to SPSS); for the advanced user, this means the user must be familiar with multiple languages (e.g. Praat’s scripting language and R). The present talk reports on a R package currently under development to overcome these inefficiencies. The package itself consists of a library of R functions, one for each of the core signal processing routines in Praat (e.g. "To Pitch..." or “To Formant…”). Each of these R functions contains a shell command invoking praatcon.exe (the command-line console version of Praat) with a Praat script that imports a file, applies the requested transformation (with the specified arguments), and saves the output. The end product is that most of Praat’s most powerful capabilities become available within R in the form of OS-level file manipulation functions. Consequently, the package allows the user to conduct the entire analysis without ever having to leave R, speeding up the process and opening many new exciting possibilities by combining the power of these two programs.
In category: Computational linguistics
Individual differences in cognitive ability and L2 speech perception
Time: 05:15pm - 06:15pm
Place: Frangipani Room, Indiana Memorial Union
Joan-Carles Mora (University of Barcelona)Individual differences in cognitive ability (e.g. phonological memory) have been shown to be related to linguistic development in monolingual and bilingual acquisition in vocabulary and oral fluency, but its role in L2 phonological development is largely under-researched. This talk will draw on recent research examining the role of phonological memory, acoustic memory and attention control through novel speech-based tasks to explain inter-learner variability in L2 phonological acquisition.
In category: Second language acquisition
Plasticity in bilingual speech perception and production
Time: 02:30pm - 04:00pm
Place: Ballantine Hall 205
Joan-Carles Mora (University of Barcelona)Previous research on the plasticity of the perceptual system in early and late bilingualism tested through lexical decision tasks has identified asymmetries in the categorization accuracy of contrastive sounds due to extensive exposure to two languages. Such asymmetries are attributable to either properties of the bilingual lexicon or phonetic weakening of contastive sounds, or both. This talk will present evidence from bilingual populations differing in amount of use of their L1 and L2 partly lending support to a phonetic weakening hypothesis. Perception was examined through accuracy and latency measures of degree of categoriality, whereas production was assessed through measures of degree of contrast categoriality (spectral distance scores) in cognate and non-cognate words. These data are discussed with a focus on the special role of cognates in bilingual speech production.
JEvents v3.0.9 Stable
Copyright © 2006-2013
Powered by Joomla!®
Copyright © 2008 The Trustees of Indiana University | Copyright Complaints