REASSIGNMENT OF CONSONANT ALLOPHONES IN RAPID DIALECT ACQUISITION
Nanyang Technological University, Singapore
• Date(s) : 4/6/2013
• Heure / time : 14h30-15h30
• Lieu / location/notice : Salle de conférences, B011, LPL
5 av. Pasteur, Aix-en-Provence
Many recent studies have demonstrated the important role that phonetic
detail plays in the representation of word forms, particularly in
connection with late plasticity. These findings, however, need to be
reconciled with models that assume a prominent role for abstract
phonemic representations. This study (also reported in German, Carlson
& Pierrehumbert, 2013) explores this issue through a dialect imitation
paradigm, and seeks specifically to locate the production system
within a spectrum ranging from exemplar-based models to highly
abstract neo-generative models. In an experiment spanning a week,
American English speakers imitated a Glaswegian (Scottish) English
speaker. The target sounds were allophones of /t/ and /r/, as the
Glaswegian speaker aspirated word-medial /t/ but pronounced /r/ as a
flap initially and medially. This experiment therefore explored (a)
whether speakers could learn to reassign a sound they already produce
(flap) to a different phoneme, and (b) whether they could learn to
reliably produce aspirated /t/ in an unusual phonological context.
Speakers appeared to learn systematically, as they could generalize to
words which they had never heard the Glaswegian speaker pronounce. The
pattern for /t/ was adopted and generalized with high overall
reliability. For flap, there was a mix of categorical learning, with
the allophone simply switching to a different use, and parametric
approximations of the “new” sound. The positional context was clearly
important, as flaps were produced less successfully when word-initial.
And although there was variety in success rates, all speakers learned
to produce a flap for /r/ at least some of the time and retained this
learning over a week’s time. To explain these effects, we propose a
hybrid model of speech production that includes elements of both neo-
generative and exemplar models.
German, J. S., Carlson, K., & Pierrehumbert, J. B. (2013).
Reassignment of consonant allophones in rapid dialect acquisition.
Journal of Phonetics, 41(3), 228-248.
James Sneed German is an Assistant Professor in the Division of
Linguistics & Multilingual Studies at Nanyang Technological University
in Singapore. After receiving his Ph.D. in Linguistics from
Northwestern University in 2008, he worked as a postdoctoral
researcher at LPL within the framework of PRO-GRAM until 2010. His
major areas of interest include the prosody-semantics interface,
rational linguistic behavior (e.g., game-theoretic pragmatics), and
the role of non-linguistic contextual factors in the malleability and
dynamics of the sound-meaning mapping. The primary languages addressed
by this research include American English, Singapore English, French,
and Malay. Recent interdisciplinary collaborations have also led him
to explore the relationship between individual speaker factors and the
phonological system in the variability realized by voice
impersonators, and to explore how an understanding of the cognitive
representation of language can help to speed the development of low-
cost technology. James is currently visiting LPL through the support
of a grant from the Erasmus Mundus MULTI exchange program.
(Un)intelligibility: A phonetic question or is it a matter of attitude
Ying Ying Tan
Division of Linguistics and Multilingual Studies, Nanyang Technological University, Singapore
LPL (Aix-En-Provence): Feb 22nd, 11h, Conference Room
In this seminar, I present the results of a recent study conducted to elicit international responses to Singapore English (SgE), in terms of both intelligibility and attitudes toward the speaker, as compared to American English. One of the biggest concerns of educationists and language policy makers in Singapore is that SgE is not an “internationally acceptable English” (Singapore Ministry of Education, 2001), and is thus not intelligible to the other English speakers. The question however is: is SgE really unintelligible outside of Singapore? If SgE is to be viewed negatively, is it due to unintelligibility or could it be attitudes and perceptions toward this variety of English? There is, at present, very little research to show how well SgE is understood in international contexts, and even less investigating international attitudes towards the variety.It is therefore the aim of this paper to address the following questions, based on the responses of over 200 respondents from over 20 countries as they listen to a set of 15 sound recordings, including read SgE, spontaneous SgE, and read American English:
- How intelligible is SgE internationally and how does SgE compare to AmE in terms of intelligibility?
- What is the attitude toward SgE internationally and how does SgE compare to AmE in terms of attitudes toward these two varieties?
Using these results of this study, I hope to take the research forward in the future by looking at some possible phonetic features that may contribute to unintelligibility.
Ying-Ying TAN is a sociophonetician who looks at how speech and accent, especially of individuals and communities in multilingual environments, encode social information. Outside of phonetics, she also works on language planning and policy, looking at how the State manages ethnolinguistic diversity and engineers multilingual individuals through language, education and media policies. Her articles have appeared in journals such as The International Journal of the Sociology of Language, Journal of Multilingual and Multicultural Development, Journal of Pragmatics, and English World-wide. She has recently been awarded the Fung Global Fellowship by Princeton University, and will spend the 2013/14 academic year at Princeton.
Mathieu Morey (Ex-Post-Doc Erasmus Mundus at Nanyang Technological University, currently at LIF)
10:30-11:00 Neologism Detection of Modern Chinese in Taiwan: A corpus-based study of PTT
Tsun-Jui Liu (Master Erasmus Mundus, National Taiwan University)
Pierre Magistry (Post-Doc Erasmus Mundus Candidate, Alpage)12:00-12:30 Verbs of interpersonal manipulation and causation in Mayrinax Atayal
Yu-Chih Lin (Master Erasmus Mundus, National Taiwan University)
Helen KaiYun Chen (Aix-Marseille Université & The Hong Kong Polytechnic University)
Location: Laboratoire Parole et Langage
Date: April 20th,
The objective of the present research is to investigate the ways in which Mandarin interlocutors utilize prosodic resources as part of the means to achieve interactional actions and sequential-organization in natural conversation. By approaches of interactional prosody and conversational phonetics toward the analysis of sound realization of Mandarin repair in natural conversation, this study provides a preliminary exploration on how the combination of detailed prosodic features forms various sound patterns in reflecting the interactional aspect and sequential organization of Mandarin conversation.
The current study examines the particular example of recycling repair, defined as “a repeat of part of a conversational turn,” among various methods used in accomplishing Mandarin same-turn self-repair. The data corpus consists of 260 cases of recycling repair culled from a collection of both video- and audio-taped, naturally occurred Mandarin interaction. For each recycling, acoustic measurements, including duration, pitch, silence, cut-off and lengthening were carried out by using Praat. Additional judgments would be made based on the analyst’s impressionistic interpretation of these acoustic cues. After taking measurements, the results were compared and the sound patterns that emerged from recyclings with the same sound manifestation were identified.
In this talk I will present 3 of the sound patterns identified, based on selected examples of Mandarin recyclings. The findings from the current talk highlight the interaction-specific, sequence-specific, and function-specific examples of recyclings in relation to the use of the prosodic patterns. The result from the current exploratory work on the correlation between sound patterns in Mandarin recyclings and interaction contributes significantly to an interdisciplinary study that focuses on broadening the interactional linguistic theory by paying close attention to the sound patterns in natural conversation.
Seminar on Chinese prosody at the LPL prosody group
Date: March 20th
Helen Chen (The Hong Kong Polytechnic University / LPL, Aix-Marseille University): The conversational phonology/ interactional prosody approach to sound (prosodic) pattern(s) in Mandarin recycling repair
Yu-Ying Chuang (National Taiwan University / LPL, Aix-Marseille University): An Introduction to Pan-Mandarin ToBI
Sheng-Fu Wang (National Taiwan University / LPL, Aix-Marseille University): Tone Sandhi Group in Taiwanese
NARRATIVE STORY-TELLING OF MANDARIN-SPEAKING PATIENTS WITH SCHIZOPHRENIA
Hsu-Ning, Graduate Institute of Linguistics, National Taiwan University
(Erasmus Mundus Master currently at Università di Pisa)
• Date(s) : 21/2/2012
• Heure / time : 14h00
• Organisateur / organiser : LPL
The present study investigates the narrative story-telling of
Mandarin-speaking patients with schizophrenia. Schizophrenia is a
mental disease characterized by abnormalities in perception and
expression, with disorganized speech being one of its representative
characteristics. This study aims to provide a comprehensive
characterization of schizophrenic speech in Mandarin. Narrative story-
telling samples were collected from two groups of participants: the
normal control group (20 participants) and the schizophrenic group
(22 participants). Participants were asked to narrate three picture
stories: (1) the Couple Story, (2) the Old Man and the Dog Story, and
(3) the Frog Story. Narrative samples were analyzed using a Mandarin
version of Narrative Assessment Profile (NAP), a profile analysis
that was designed to evaluate discourse coherence in both typical and
atypical speakers (McCabe & Bliss, 2003). Eight dimensions of the
narrative performance were examined. They are topic maintenance,
event sequencing, reasonableness, referential skill, background
information, evaluation, conjunctive cohesion, and the narrative
pattern. In addition to the discourse level analysis, lexical choices
of specific connectives (causal connectives and contrastive
connectives) and verbal predicates (perceptual verbs, psych verbs,
and mental verbs) were computed. The study also includes an analysis
of character representation within narratives.
Schizophrenic narrators in this study had a significantly weaker
narrative performance in both discourse level and word level. Their
performance in six narrative dimensions (i.e. topic maintenance,
reasonableness, background information, evaluation, conjunctive
cohesion, and the narrative pattern) was significantly weaker.
However, they seemed to preserve an intact ability to identify basic
elements in narratives, performing comparably well in dimensions of
event sequencing and referential skill. Lexical analysis showed that
the schizophrenic group used less contrastive connectives and mental
verbs when telling their stories, echoing the findings of NAP.
Patients also demonstrated a more superficial character
representation in narratives, from which the diminishments of
language ability were related with the decline of Theory of Mind
Talk given at the Laboratoire Parole et Langage on Dec 2nd, 2011.
TAN Ying Ying (Nanyang Technological University)
In linguistic ecologies where multiple languages co-exist and language
contact creates language variation and language change, the process of data
collection is complicated by a set of factors that will need careful
manipulation and consideration. This talk aims to highlight some linguistic
complexities that need to be considered in a multilingual context such as
Singapore, where its linguistic ecology consists of over 20 different languages
and its population almost entirely multilingual. Besides presenting an overview
of the linguistic situation in Singapore, I will also present the findings of a
few of my studies on prosody, ethnic accent perception and a sociophonetic
study of rhotics in Singapore English, and argue that there are phonetic
variants, both in production and perception, that are closely correlated to
ethnicity, age, education and socioeconomic status. Given such complexities,
one could ask what then is the possibility and feasibility of building a speech
recognition system for Singapore. And in the age where there is an increasing
need to profile speakers by the way they speak, one could also ask how these
phonetic variants will complicate matters in a multilingual context such as
Singapore. These are some of the questions that I would like to pose as future