From: Mary Beckman Subject: today's meeting Date: Fri, 4 Aug 2006 09:53:31 -0400 Just a quick reminder that we are meeting as a group in 29 Derby again today after that interlude of three one-on-one meetings in 21 Oxley. Please be prepared to tell each other a bit about what you accomplished in the separate meetings, and say whether there are articles or such that you want the rest of us to read with you and discuss in group. For example, Jeonghwa has a very interesting article on differences between L1 and L2 processing of English relative clause attachments and David has chosen some articles on vowel normalization that we all may want to read. We can then figure out what our strategy for meetings next week should be. ======================================================== Notes from class meeting on Friday, August 4. ======================================================== 1. We went over some examples of English /twi/ versus /ti/ that Mira recorded (in vs. ), examining the dynamics of the spectral transition from the stop burst through the [w] constriction (voice onset is right at the F2 minimum) into the vowel, and comparing these dynamics with analogous sequences in Korean. The dynamics are *totally* different in a way that makes it easy to see how the burst portion in English /twi/ could be interpreted as the epenthetic vowel. The dynamics of the /kwi/ versus /ki/ examples (in versus ) are similar, but starting at a much lower frequency in the dorsal stop release. Jeonghwa heard the "voiceless vowel" in that case as more equivalent to Korean /u/ than to Korean /i-/, and since /u/ is not typically an epenthetic vowel in loan word assimilation, it would be more difficult to parse this as a syllable. addendum after class ... From: ddurian@ling.ohio-state.edu Subject: Northern California vs. Northern Cities Vowels Date: Fri, 04 Aug 2006 15:33:54 -0400 Hi all, Since this came up today while talking about /twi/ vs. /ti/--here's an interesting comparison of Jocelyn's home dialect vowel system and my (David's) home dialect system. http://www.stanford.edu/~eckert/vowels.html David ======================================================== 2. Jeonghwa's report is in: http://ling.osu.eud/795.10/prosody.txt ======================================================== 3. David designed how he will pick out target vowels in his recordings and also gathered some papers to read about vowel normalization. ======================================================== 4. Jocelyn reported on the hypothesis that she wants to test and the statistical design that she is thinking of trying. Jocelyn's research questions are: 1) How do the intelligibility scores compare across the four listener groups? For example, do Korean L1 listeners have more problems with the Mandarin accented English than Mandarin L1 listeners? 2) How do the reaction times (as a measure of ease of comprehension) compare across the four listener groups? 3) What is the impact of talker consonant "errors" versus talker vowel "errors" on intelligibility? Do the (presumably greater number of) vowel errors in the Mandarin speakers have more of an effect than the consonant errors, as suggested by the Bent, Bradlow, and Smith (2005) conference presentation? The first two questions are relatively easy to do with vanilla-flavored ANOVA. But what to do for question 3? We thought of doing this with a stimulus-by-stimulus logistic regression. ALternatively, it could be block by block, with all of the stimuli from a particular talker that a particular set of listeners heard. ======================================================== Preparation/schedule for classes next week. We will use the same schedule of one-on-one meetings Tuesday through Thursday and then meet as a group on Friday. In preparation for this ... Everyone will send Mary notes from their meeting this week, and a sentence or two describing what they want to accomplish in the one-on-one meeting. If there are articles that we want to read together, in particular, these references should be in the notes, so that we can advertise that to others, and we can discuss the articles on Friday. In the meetings, then, each one will keep notes and give them to Mary to post on the class web site at the end of the session (or the next day). ======================================================== Wrapup on Friday will include ... Learning to apply the R lm function to David's study data, to learn how to do logistic regression in R, so that Jocelyn knows how to do it with her data when she gets them.