Pronunciation in the
Language Teaching
Curriculum
Pronunciation in Second Language
Learning and Teaching
PSLLT
5th Annual Conference
!
Iowa State University
September 20-21st, 2013
!
!
Welcome to the 5th Annual PSLLT
Conference
Pronunciation in Second Language Learning and Teaching 2013
Pronunciation in the Language Teaching Curriculum
September 20-21, 2013
Iowa State University
Memorial Union, Gold Room and Cardinal Room
Contact Information
pslltconference@gmail.com
Plenary Speaker
Lynda Yates, Macquarie University
Learning how to speak: Pronunciation,
pragmatics and practicalities in the classroom
and beyond
Conference Organizers
John Levis, Shannon McCrocklin
Conference Volunteers
Stephanie Link, Sinem Sonsaat, Ghinwa
Alameen, Monica Richards, Edna Lima, Mandy
Qian
Conference Program Design and Layout
Shannon McCrocklin
Funding
Iowa State LAS Dean’s Office
ISU English Department
Cambridge University Press
Reviewers
John Levis, Ron Thomson, Lucy Pickering, Beth
Zielinski, Tracy Derwing, Murray Munro, Jessica
Sturm, Kimberly Levelle, Shannon McCrocklin,
Stephanie
Link, Sinem Sonsaat, Mandy Qian!
!
Contents
2
Pre-Conference Workshop
4
Detailed Program
8
Plenary Speaker
9
Teaching Tips Round Robin Abstracts
13 Oral Paper Abstracts
26 Poster Abstracts
35 Call for 2013 Proceedings
36 Call for 2014 Papers
37 Getting Online
38 Map & Directions to Conference Dinner
39 Map of ISU Campus
40 City Information
!
!
!
1
Pre-Conference Workshop
Models, metaphors, and the evidence of spontaneous speech: A new
relationship for pronunciation and listening
Richard Cauldwell, Birmingham, UK
This workshop has the goal of improving the teaching of listening, by identifying and
exploiting a new relationship between pronunciation activities and listening goals. New
concepts and techniques (both high- and low-tech) will be illustrated. Participants will leave
the workshop with new ideas to consider, and activities to use immediately in the classroom.
The workshop will begin with thought-provoking theory, and end with the ruthlessly practical:
but throughout there will be a constant reference to the evidence of recordings of
spontaneous speech, and continual opportunities for suggestions and questions from
participants.
Rationale
For pronunciation and speaking, we encourage clear intelligible speech. We present
learners with a model of speech which is built around dictionary pronunciations (citation
forms) and rules of connected speech. We can think of the citation forms as greenhouse
plants – they are isolated forms preceded and followed by a pause, with their component
parts – vowels, consonants, syllables and stresses, all clearly present. The rules of connected
speech – linking, elision, sentence stress, etc - can be thought of as guidelines for
transplanting and arranging greenhouse plants into orderly pleasing arrangements in a
garden. However, the greenhouse forms and the gardening guidelines are not appropriate
for teaching listening. This is because the speech that learners encounter outside the
classroom is more like jungle vegetation than garden or greenhouse plants, much wilder
than the forms they encounter in the classroom. Such speech contains phenomena which
are rarely seen in textbooks and words, like vegetation in the jungle, are blended into their
neighbours in ways which are not predicted by the rules of connected speech. They are
squeezed into bursts of the stream of speech, and it becomes difficult to recognise where
one word begins and another ends, or indeed whether word-endings, syllables, or whole
words have occurred at all. In class, we need to prepare students for their encounters with
jungle listening, while continuing to promote intelligible pronunciation. This workshop will
describe and explore ways of working on these separate but related goals.
2 !
Pre-Conference Location: The Gallery in the Memorial Union
Workshop Timetable
Part 1: Models and metaphors -The goals of listening and pronunciation are different. We
need different models of speech for each goal. We have good models in place for
pronunciation, we have inadequate models for teaching listening. We need to distinguish
between goals and pronunciation activities can serve the goal of listening.
Part 2: Evidence from spontaneous speech -Words have many different soundshapes, of
which the citation form is only one. The soundshapes are formed by interactions between
the language and speaker factors: gender, accent, choices of speed, prominence and
clarity.
Part 3: High-tech solutions: computers, smartphones, tablets, etc. - Recent developments in
technology enable us to examine what happens to words in the stream of speech, to
compare how words sound different as speakers and contexts change. We can manipulate
and play with the sound substance of speech, in ways which promote faster learning of the
listening skill.
Part 4: Low-tech solutions: teachers and learners voices in the classroom -The teacher voice
and student voices can together be used in class to create, savour and handle the sound
substance of the stream of speech. We will look at a number of activities that can be used
and adapted to different teaching contexts.
Pre-Conference Workshop Schedule
9:00-10:30
10:30-11:00
11:00- 12:30
12:30-1:30
1:30-3:00
3:00-3:30
3:30-5:00
!
Part 1: Models and Metaphors (Gallery)
Break
Part 2: Evidence from Spontaneous Speech (Gallery)
Lunch (Provided in the Pioneer Room)
Part 3: High-tech Solutions: Computers, smartphones, tablets, etc (Gallery)
Break
Part 4: Low-tech Solutions: Teachers and learners (Gallery)
3
Friday, September 20th
8:00-8:50am
9:00-9:10
9:10-10:10
10:10-10:30
10:30-10:55
11:00-11:25
11:30-11:55
12:00-1:50
12:30-1:50
Registration (Cardinal Room)
Welcome (Cardinal Room)
Plenary Address by Lynda Yates (Cardinal Room)
Break
Cardinal Room
Gold Room
E. Zimmerman
M. Munro
Teaching the Teachers: How Do
What do you know when you “know”
Pronunciation Textbooks Aid
an L2 vowel?
Inexperienced
Teachers’ Pedagogy?
S. Sonsaat & S. Link
R. Thomson
How do nonnative teachers use
Does vowel learning in one context
pronunciation materials? Implications for generalize to other contexts?
materials development
A. Rocammo
E. Koffi
Learning Pronunciation in Just Ten
Assessment of the Intelligibility of [ ! ] in
Minutes a Day: Adapting Pronunciation
Seven Varieties of L2 Englishes
Training to a Four-Skills German
Classroom
Working Box Lunch (Provided in Pioneer Room)
Posters: Pioneer Room
S. Alexander -Intonation and perceived sincerity in EFL and ESL learner apologies
J. Barcroft & M. Sommers – Better L2 pronunciation is one of the many benefits of
acoustically varied input
C. Barrett – Laying a foundation for rhythm-based pronunciation instruction
L. Cai – An efficient method to build up native sounds in Chinese teaching: Multi-sensory
and multi-cognitive approaches
C. Cárdenas- Scaphoning your language
S. Chibani- Pronunciation teaching in Algeria: From stagnation to progress
M. Delicado Cantero & W. Steed – Fair Dinkum: L2 Spanish in Australia by the book
F. Desmeules-Trudel- VISC effects on the perception of Quebec French nasal vowels by
Brazilian learners
N. Driscoll – Hatsuon Help: a research-based, culturally-sensitive English pronunciation
website for Japanese ELLs
V. Gonzalez Lopez & D. Counselman- The production and perception of Spanish voiceless
stops by novice learners: shedding light on early L2 category formation
S. Halicki – Back door phonetic conditioning: Accent therapy in early French pronunciation
training
Y. Lan – Detecting L2 speech deviations by a communicative experiment procedure:
taking Cantonese speakers’ realizations of English [r] as an example
S. Link, S. Sonsaat, & J. Levis – Confidence in teaching pronunciation: How native and
nonnative teachers negotiate the pronunciation classroom
W. McCartan – Word stress diagnostic procedure shared through a wiki site
C. Nagle - Acquisition of the voicing contrast in L2 Spanish
4 !
D. Olson & H. Offerman- The effects of visual feedback on learner pronunciation: Speech
analysis software in the L2 classroom
L. Pierce – Multi-methodological, cross-disciplinary approaches to pronunciation teaching
S. Shoji – Japanese epenthetic vowels: How Japanese speakers pronounce English words
K. Taylor de Caballero & S. Thompson- Coloring pronunciation across the curriculum with
the Color Vowel Chart
H. Yang – Investigating needs of stakeholders of an oral proficiency test for ITAs to bridge
the gaps between ITAs’ needs and raters’ feedback.
E. Zetterholm – Final stops or not? The importance of final consonants for an intelligible
accent.
E. Zetterholm & M. Tronnier – Different stress patterns meet: Kurdish L1 speakers learn Swedish
2:00- 2:25
Cardinal Room
L. Buss
Beliefs and Practices of Brazilian EFL
Teachers Regarding Pronunciation
2:30-2:55
V. Sardegna
Non-Native Teachers’ Identity Formation
as Qualified Pronunciation Teachers
3:00-3:25
J. Levis, S. Link, S. Sonsaat, T.A.Barriuso
Native and nonnative teachers of
pronunciation: Does language
background make a difference in
learner performance?
Break
3:30-3:55
Gold Room
T. Makino
Pronunciation Characteristics of
Japanese Speakers’ English: A
Preliminary Corpus-Based Study
J.H. Esling
The two-part model of the vocal tract:
a new articulatory basis for phonetics
J. Sturm
Effects of Instruction on Voice Onset
Time in word-initial /p/ for L1 American
English students: A Preliminary Study
Cardinal Room
4:00-4:25
4:30-4:55
6:00
!
Gold Room
J. Koreman, O. Husby, E. Albertsen, P.
Wik, A. Øvregaard, & S. Nefzaoui
L1 variation in foreign language
teaching: challenges and solutions
J. Sereno, L. Lammers, & A. Jongman
P. Watts, A. Huensch, L. Pierce
Perception of foreign-accented speech! Attainable Targets for L2 Learners: How
Proficient L2 Speakers can Bridge the Gap
Conference Dinner at St. Johns (See Map & Directions on Page 37)
S. McCrocklin – Dictation Programs for
Pronunciation Learner Empowerment
5
Saturday, September 21st
8:30-9:00
9:00-10:30
Registration (Cardinal Room)
Teaching Tips Round Robin (Cardinal Room)
C. Keppie - From Mirrors to Mouthwash: Instructional Approaches to Teaching Pronunciation
G.M. Levis & J. Levis – Using introductions to improve initial intelligibility
E. Lima- Using Audacity to help learners improve their suprasegmentals
C. Meyers- Intelligible Accented Speakers as Pronunciation Models
M. Reed - The English syllable: Big news, bad news, and why it's important for intelligibility
M. Richards- Providing individualized homework and accountability for ITAs via Internet and
LMS resources
A. Roccamo - Effective Pronunciation Instruction in Beginner and Intermediate Language
Classrooms
V. Ruellot - Introducing French Nasal Vowels at the Beginner Level
A. Saalfeld - Flipping the phonetics classroom
S. Zhang - Using Tongue Twisters to Supplement CFL students’ Pronunciation and Tone
Practice
10:30-10:55
11:00-11:25
11:30-11:55
12:00-12:30
12:30-2:00
2:00- 2:25
2:30-2:55
3:00-3:25
6 !
Break
Cardinal Room
A. Violin-Wigent
Comparing online vs. face-to-face
classes: A case study of a French
pronunciation class
J. Foote & G. Smith
Is there an App for that? An
investigation of pronunciation
teaching apps
M. Park & S. Huffman
The Potential of ASR for Non-native
English Speakers in Air Traffic Control
Lunch (not provided)
Cardinal Room
Y. Zhuang
Suprasegmentals and second
language teaching: A meta-analysis
O. Kang & F. Chowdhury
Prosodic Features in L2 Accented
Speech: Human versus Machine
S. Staples
Prosodic patterns in nurse-patient
interactions: a comparison of
international and U.S. nurses
Gold Room
T. Isaacs, J. Foote, & P. Trofimovich
Drawing on teachers’ perceptions to
adapt and refine a pedagogicallyoriented comprehensibility scale for use on
university campuses
M. Munro, T. Derwing, R. Thomson, & D. Elliot
Naturalistic L2 Segment Development:
Implications for Pedagogy
B. Zielinski
Demystifying comprehensibility for the
language teaching curriculum
Gold Room
M. Reed
Connecting pronunciation to listening:
Raising learner and instructor awareness
M. Sakai & C. Moorman
Can perceptual training improve
production of L2 phones? A meta-analytic
review
R. Cauldwell
Pronunciation and Listening, the need for
two separate models of speech
3:30-3:55
Break
Cardinal Room
4:00-4:25
G. Alameen
Perception and production of Linking in
Non-Native Speakers of English
4:30-4:55
P. Keyworth
The Acoustic Correlates of Stressshifting Suffixes in Native and
Nonnative English
Closing (Cardinal Room)
5:00-5:30
!
Gold Room
S. McCrocklin & S. Link
What is identity? ESL and Bilinguals' Views
on the Role of Accent
C. Shea & J. Vojtko Rubi
Dialect adaptation and L2 Spanish listeners
7
Plenary Address
Lynda Yates, Macquarie University
Lynda’s research interests centre around
adult language learning and
communication in workplace and further
study contexts, focussing in particular on
speaking skills, pronunciation and
interpersonal pragmatics. In 2009 she was
Acting Director of the AMEP Research
Centre where she had been a senior
researcher since its inception in 2000. After
completing an Honours degree in
languages (Russian and French), Lynda
gained wide experience in adult TESOL in
the U.K., France, Armenia and Egypt, and
has also consulted to industry. She has
been involved in TESOL teacher training
for a number of years at a postgraduate
and professional development level.
(www.ling.mq.edu.au)
Learning how to speak: Pronunciation, pragmatics and
practicalities in the classroom and beyond
It is beyond dispute that learners who want to develop good speaking skills in a language
also need to develop good pronunciation, and yet research continues to report that
pronunciation still has low visibility in the curriculum and is often treated as something of a
poor relation in the classroom. Many teachers are still wary of pronunciation as a specialist
area that is somehow separate from the other skills necessary for successful communication an isolationist tendency that can make its consequent neglect in the curriculum and in
teacher training programs only too easy.
In this plenary I go back to basics and focus on what it is that learners need to do outside the
classroom with the language they are learning. Drawing on studies that have explored the
lives and communicative needs of immigrants and international students, I will illustrate not
only the importance of pronunciation in their lives, but also its close interrelationship with
other spoken skills. I will then consider the implications for how we approach the teaching of
pronunciation proactively as part of developing students’ repertoire of speaking skills in the
classroom and beyond.
8 !
Teaching Tips Round Robin Abstracts
(Listed Alphabetically)
** Please note that the teaching tips will be 10-minute sessions, each repeated multiple
times. You will have a minute to switch tables between sessions. We do not expect that you
will be able to attend all sessions; instead, plan to attend six or seven.
From mirrors to mouthwash: Instructional approaches to teaching pronunciation
Christina Keppie - Western Washington University
As a French and linguistics professor who speaks French as a second language, I have had to develop an
innovative curriculum for teaching pronunciation to university undergraduates. While a variety of textbooks on
French phonetics and pronunciation are widely available, none are developed within any particular context
that could be applied to any language-learning classroom setting, for they are primarily concerned with the
scientific end of phonetics rather than possible methods of instruction. Regardless of the language of instruction
or its level, pronunciation should and can be taught within the context of the course’s theme(s). This teaching
tips presentation will focus on a series of interconnected pronunciation methodologies and activities, from using
mirrors to visually see the difference in phoneme pronunciation, to using mouthwash to ensure contact
between the tongue and uvula that is necessary for an accurate pronunciation of the French [R]. These
approaches will be integrated into the context of the story of The Little Prince (Le Petit prince) and popular
French-language comic books to show their applicability to a variety of language teaching courses and texts.
Using introductions to improve initial intelligibility
Greta Muller Levis & John Levis – Iowa State University
Early instruction in speaking any language includes pre-packaged chunks of functionally important language
such as introductions and leave-takings. The presenters will demonstrate how introductions in English can be
used to introduce communicatively-oriented pronunciation from the very beginning. Besides ways to introduce
pronunciation through introductions in English, suggestions for principles that may apply to other languages are
provided.
Using Audacity to help learners improve their suprasegmentals
Edna Lima – Iowa State University
Pronunciation instruction is indispensable in foreign/second language learning as poor phonetic
control and prosody can distract the listener and hinder comprehension of the message (Eskenazi,
1999). Research claims that placing focus on suprasegmentals improves speaker comprehensibility
and greater progress can be attained (Celce-Murcia, Brinton, & Goodwin, 1996). Thus,
suprasegmentals should be the guiding principle for pronunciation teaching (Kang, 2010). Word
stress, rhythm, and intonation are features “suggested by the majority of phonology authorities as
having the greatest implications for intelligibility” (Jenkins, 2000, p. 39) and, by extension, on
comprehensibility.
The activity presented in this session is designed to help adult learners improve their perception and
production of English rhythm. The activity can also be adapted to target other suprasegmental
features and different audiences. Among the advantages of using Audacity to help learners improve
their suprasegmentals are: access to multiple speech models; opportunities for large amount of
practice, which is not possible in a regular classroom; an uninhibiting learning environment; and selfdetermined pace.
!
9
Intelligible accented speakers as pronunciation models
Colleen Meyers - University of Minnesota
A new twist on the “Mirroring” approach is for L2 learners to imitate intelligible accented speakers—
often from their own L1 background—as models. The presenter will show before and after video of a
student whose intonation patterns improved through mirroring a highly proficient intelligible L1 model
and discuss guidelines for choosing which L2 models are best suited to facilitate pronunciation
improvement.
The English syllable: Big news, bad news, and why it's important for intelligibility
Marnie Reed - Boston University
This paper situates the Murphy-Kandil numeric notation system in a metacognitive approach that aims to
raise instructor and learner awareness of the role of the syllable as posing challenges for learner intelligibility
and in establishing the prosody of English. A New Word Checklist is proposed to support speech-intelligibility
training.
The syllable structure of English, allowing complex onsets of up to three consonants and complex codas
of up to four consonants phonetically in monosyllabic words, exceeds the allowable unit for many of the world's
languages. The challenges this poses are met one of two ways: consonant cluster reduction, eliminating one or
two of the cluster consonants, or epenthesis, inserting vowels to restore C-V syllable structure. Both solutions
adversely impact intelligibility.
Lexical stress poses a second challenge: establishing the syllable as the relevant and meaningful unit of timing
for English, the duration of the vowel in the stressed syllable as essential for intelligibility, and the alternation of
stressed and unstressed syllables as the foundation for English prosody. The important news is that stress
alternates in disyllabic words (except compounds) and polysyllabic words. The bad news is that lexical stress
patterns are not predictable, as confirmed by the identification by Murphy and Kandil of 39 stress patterns in
their AWL analysis.
A proposed New Word Checklist and accompanying manipulative materials seek to address instructor
and learner metacognition regarding the role of the syllable at the lexical, phrasal, sentential and discourse
levels. It includes monosyllabic native-language/ target-language syllable-structure comparison. It addresses
the spoken form category of Nation's requirements for knowing a word by adopting the Murphy-Kandil numeric
notation system for lexical stress. It expands the scope of this system to the phrasal and standard sentence
levels. Finally, it establishes a springboard for recognizing contrastive stress and intonation and their role in
conveying speaker intent.
Monica Richards – Iowa State University
Providing individualized homework and accountability for ITAs via Internet and LMS resources
Pronunciation instructors of international teaching assistants (ITAs) frequently provide individual feedback
highlighting the fluency, suprasegmental, and segmental challenges most likely to inhibit a particular ITA’s
successful interaction with undergraduates. Yet providing ITAs with practice actually implementing individual
feedback given, adequate to enabling their development of new, more communicatively effective
pronunciation habits, remains difficult. However, the Internet and learning management systems (LMSes, e.g.
Moodle) contain resources capable of supporting and holding students accountable for focused, self-directed
work on nearly any pronunciation target. This presentation will briefly introduce 1) sample teacher feedback
providing students with individualized 15-minutes-per-day homework options; 2) directions for a Slidesharebased fluency development homework option and a TED Talk-based thought-grouping, phrase stress, and
intonation development homework option); and 3) an LMS-based accountability mechanism for reporting
students’ completion of their individualized 15-minutes-per-day homework.
Effective pronunciation instruction in beginner and intermediate language classrooms
Ashley Roccamo - Penn State University
Foreign language pronunciation skills are a significant component of communicative competence and
should therefore command a place in all foreign language classrooms. A variety of pronunciation instruction
programs have been previously found effective, both in the classroom and the laboratory (Derwing, Munro &
10 !
Wiebe, 1998; Hardison, 2004; Saito & Lyster, 2011). Yet most of these training studies have been conducted in
more advanced foreign language courses that focus on phonetics or oral proficiency. It is uncommon to
encounter a pronunciation training study conducted in beginner or intermediate four-skills classrooms.
The proposed roundtable presentation will address this gap and focus on a module-style pronunciation
unit designed for beginner- and intermediate-level language learners. The roundtable presentation will be
divided into three parts. The first part will explain the motivation for and the significance of a module-style of
pronunciation instruction previously implemented in first- and fourth-semester German classrooms. The second
part of the presentation will provide concrete examples illustrating exactly which types of activities and games
were conducted in the modules. The third part of the presentation will be a hands-on example activity for
roundtable visitors to try themselves. At the conclusion, handouts of sample lessons will be provided to offer
interested instructors ideas for similar interventions in their own classrooms.
The goal of this roundtable presentation is to present interested colleagues with useful information about
a different style of classroom pronunciation training intended for beginner and intermediate learners. It is
hoped that interested parties will use the information as a springboard to apply the techniques to their own
classrooms and teaching strategies.
Viviane Ruellot - Western Michigan University
Introducing French nasal vowels at the beginner level
This presentation proposes tips and recommendations for an introduction to French nasal vowels targeting
adult American learners at the beginner level. It demonstrates an explicit and contrastive approach to the
concept and practice of vowel nasalization. Learners are first made aware of the existence of (partial) vowel
nasalization in American English (Ruhlen, 1973), as illustrated in the words “thing,” “long,” and “man.” They then
are guided in identifying the contrasts between American English and French nasal vowels (e.g., “fan” vs. fin,
“long” vs. lent and “honhonhon!” vs. ont). Finally, learners concentrate on discriminating between French
nasal vowels (e.g., vin, vent and vont). Introduction, explanations, and contextualized practice address both
the perception and the production aspects of French nasal vowel learning and teaching. Although this
presentation primarily targets learners, it will also be of interest to teacher educators.
Flipping the phonetics classroom
Anita Saalfeld - University of Nebraska at Omaha
Although a number of studies have shown gains in student pronunciation following instruction (Derwing, Munro
& Wiebe, 1997; 1998; González-Bueno, 1997; Hahn, 2004; Lord, 2005; 2008; Wang, Spence, Jongman, & Sereno,
1999), the author’s first attempt teaching an articulatory phonetics course did not result in any perceivable
gains in learner pronunciation (as evaluated by native speaker raters). Since one of main goals of the course
was to improve learner pronunciation, these results were disheartening.
In order to address this, a course redesign was implemented in the Fall 2012 semester. The goal of the
redesign was to use existing technology to automate the parts of the course that could be automated so that
class time could be spent working on pronunciation, rather than on lecture. A secondary goal was to improve
student learning of the theoretical material of the course.
In order to meet these objectives, the course was scheduled in a computer classroom. All homework
assignments were digitized, and a series of mini-lectures was produced in order to provide students with
instruction outside of class time (“flipping” the classroom). During class, students used Praat and recordings of
native speakers provided by the instructor and learned to evaluate their vowels and most consonants
compared to those of a native speaker.
The proposed session would provide a description of the considerations involved in flipping the class,
including technology used, problems encountered, and adjustments being implemented in Fall 2013, when the
course will be offered again. It will also present suggestions of how to create differentiated assignments for
heritage speakers of the target language.
Pronunciation data have not yet been analyzed, but comments on anonymous student evaluations indicated
that students felt that the structure of the class was beneficial for learning, and also allowed them to
improvement their pronunciation.
!
11
Using tongue twisters to supplement CFL students’ pronunciation and tone practice
Shenglan Zhang - Iowa State University
Tongue twisters are phrases or sentences that were constructed to put similar but distinct phonemes and tones
together to exercise the jaws, the tongue, and the muscles around the mouth. By purposefully putting the
similar but distinct phonemes and different tones together, tongue twisters sometimes achieve comic effects.
Because of these features, tongue twisters are challenging and engaging as well. The author of this
presentation, based on her teaching experiences, intends to introduce an approach to improving Chinese-asa-Foreign-Language (CFL) learners’ pronunciation and tones by incorporating Chinese tongue twisters as
supplemental materials in the CFL curriculum.
This approach focuses on how to sequence the introduction of different tongue twisters in first year Chinese
language classrooms in an order that is pedagogically appropriate for their Chinese language level so that the
learners’ tones, pronunciation, and prosody can be maximally improved and at the same time their vocabulary
and structure learning can be enhanced. Different ways of introducing tongue twisters are also going to be
discussed. A variety of activities in using tongue twisters are going to be suggested.
12 !
Oral Paper Abstracts
(Listed Alphabetically)
Ghinwa Alameen – Iowa State University
Perception and production of Linking in Non-Native Speakers of English
Linking, as a connected speech phenomenon, is considered a common occurrence in English running
speech and has therefore received some attention in phonology literature. However, little has been done to
investigate pedagogical tools that help learners improve their linking skills.
This pilot study will examine the impact of electronic visual feedback training and audio-only training on
the production of consonant-vowel (C-V) and vowel-vowel (V-V) linking and the extent to which the training
would generalize to novel contexts. In addition to the quantitative data, learners’ responses to questionnaires
about their perceived usefulness of both methods of training and my observation notes during training will
provide qualitative information on the role of computer-assisted speech training in improving L2 pronunciation.
Larissa Buss - Concordia University
Beliefs and Practices of Brazilian EFL Teachers Regarding Pronunciation
Growing interest in second language (L2) pronunciation has led to a proliferation of research and resources in
the field. However, it is still unclear how this increase in knowledge and resources is reflected in the views and
practices of L2 teachers around the world. This study investigated the beliefs and practices of 60 Brazilian
English as a foreign language (EFL) teachers through an online survey. The questionnaire had a total of 64
open- and closed-ended questions on pronunciation teaching practices and beliefs, in addition to
demographic questions. The teachers expressed positive attitudes toward pronunciation instruction and beliefs
that usually reflected current research findings. Overall, teachers thought pronunciation teaching is extremely
important for students at all levels. In terms of teaching goals, more than 70% believed that pronunciation
instruction should aim to make learners “comfortably intelligible” (Abercrombie, 1949) rather than nativelike.
Over 80% wanted more training in teaching pronunciation, similar to teachers in English as a second language
(ESL) contexts (e.g., Baker 2011; Macdonald, 2002). The participants reported different teaching practices from
those mentioned by ESL teachers in Foote, Holtby, and Derwing (2011). Strategies such as students recording
their speech, using mirrors, or viewing mouth diagrams were reported as never or rarely used by most Brazilian
teachers. Brazilian teachers indicated that they used modelling or repetition activities and infrequently
employed metalinguistic explanations, pronunciation rules, or L1/L2 comparisons. Most teachers stated that
they spent more time on segments than on suprasegmentals. Segments, morphophonemic suffixes, and word
stress were identified as the most frequently taught features, while rhythm appears to be the least frequently
taught. The findings will be discussed in light of the different conditions for teaching and learning in second as
compared to foreign language contexts.
Richard Cauldwell - Speech in Action
Pronunciation and Listening, the need for two separate models of speech
!
The model of speech which dominates language teaching is one that is designed for clear intelligible
pronunciation – the careful speech model (CSM). The CSM is something to refer to, to aim at, and to emulate.
Its components include the citation forms of words, the rules for the relationship between clause and speech
unit, and the tunes of statements and questions. It has three main advantages; it is teaching and learning
friendly; it is easy to describe in terms of rules; and it dovetails well with the major components of language
teaching, sentence grammar and vocabulary.
However, to help the learner-listener, we need to recognise that spontaneous speech – indeed any
speech that learners encounter outside of the classroom – is much less orderly than that predicted/advocated
by the CSM. We need to incorporate the unruliness of everyday speech into our teaching, and to achieve this
aim we need also to allow room for another model of speech, the Spontaneous Speech Model (SSM) – a
descriptive model based on the evidence of spontaneous speech. We need the SSM is because the rules of
the CSM have been mistaken for facts about how speech actually is. Even when experts such as Wells (2006:
91!92) warn that their guidelines are not to be taken as evidence of the facts of language use, the warning is
ignored, and the guidelines of the careful speech model are taken as true for all types of speech.
!
13
In my talk I will outline and illustrate the components of a descriptive model of spontaneous speech,
optimised for the teaching of listening.
John H. Esling - University of Victoria
The two-part model of the vocal tract: a new articulatory basis for phonetics
This paper summarizes descriptive phonetic research and reports on experimental studies leading to a new
interpretation of how the vocal tract functions. The paper introduces a completely new view of the vocal tract
– a two-part oral/laryngeal vocal tract model – which will contribute to a clearer understanding by language
teachers of the basic phonetic speech-production mechanism. The new model is extremely relevant to how
pronunciation is represented in language teaching texts as well as to how language teachers hear the sounds
that their students are producing. The presentation introduces the ‘Laryngeal Articulator’ model of the vocal
tract, illustrated with experimental phonetic videos of speech-sound production in a wide range of languages.
It also presents canonical illustrations of the possible ‘states of the larynx’. This revision redefines Catford’s
classification of ‘phonation types’ and ‘states of the glottis’ and Laver’s set of voice quality parameters. From
the point of view of phonetics, this constitutes an innovative approach to teacher education – a professional
development experience which will equip teachers with new auditory and articulatory information to choose
and design new targets for instruction (while perhaps eliminating some older designs). The presentation aims to
provide an alternative for conceiving of ‘how phonetics works’.
The laryngeal articulator, consisting of the glottal mechanism, the supraglottic tube, the
pharyngeal/epiglottal mechanism, and including three levels of folds (the vocal folds, ventricular folds, and
aryepiglottic folds) is shown to be responsible for the generation of multiple source vibrations and for the
complex modification of the pharyngeal resonating chamber that accounts for a wide range of contrastive
auditory qualities. Many languages of the world exhibit features that can be described in terms of ‘laryngeal
quality’. The acoustic cues of these features illustrate an extensive range of use of the laryngeal constrictor
mechanism, which controls changes from the glottis through the epilaryngeal tube to the aryepiglottic folds,
and which governs changes in the pharyngeal resonator in general. Examples will be shown of the fine control
of laryngeal constriction observed laryngoscopically in over 20 languages. Videofluoroscopy and ultrasound
analyses will also illustrate the parameters of movement available in the laryngeal/pharyngeal space.
Laryngoscopic evidence drawn from e.g. Tibeto-Burman, Semitic, Cushitic, Kwa, and Gur languages
demonstrates the distinctive use of the laryngeal articulator in pharyngeal trilling combined with glottal voicing,
voiceless pharyngeal trilling, and epilaryngeal tube shaping to create opposing vocal register series.
Phonological oppositions that have been described with divergent and confusing terminologies (such as the
[ATR/–ATR] contrast, tonal register in Tibeto-Burman, or pharyngeals in Semitic) can be recharacterized simply
and elegantly using the Laryngeal Articulator Model.
Jennifer Foote and Gabriel Smith - Concordia University
Is there an App for that? An investigation of pronunciation teaching apps.
Technology has long been an integral part of pronunciation instruction. Furthermore, there is research evidence
that computer assisted pronunciation instruction (CAPT) can be effective (Levis, 2007). However, with the rise of
smart phones and tablets, the role of technology in pronunciation instruction has expanded beyond what is
typically envisioned within CAPT. Researchers have offered guidelines for pronunciation materials in ESL
textbooks (Derwing, Diepenbroek, & Foote, 2012) and for CAPT materials (e.g., Pennington, 1999). However, it is
difficult to know what should be expected of pronunciation apps, which vary enormously in both price and
functionality. To understand the potential benefits and challenges of pronunciation apps, we must first have a
sense of what is being marketed and purchased already.
In this study, we conducted a systematic search for pronunciation apps available for Android and
Apple mobile devices. We found over 25 apps for both platforms (excluding the many that only functioned as
speaking dictionaries). The apps were analyzed both for their pronunciation content and the way in which the
content was presented. We found that different apps varied from focusing on one single segment to covering
a wide range segmental and suprasegmental foci. We also looked at the advertising claims of the apps, their
ratings, their prices, and which technological tools were used with each app (e.g., voice recognition, videos,
recording, etc.). We found a wide range of prices and technological sophistication. The potential usefulness of
apps, as well as some of the potential concerns about apps will be discussed, from the perspectives of both
classroom and individual use.
14 !
Talia Isaacs - University of Bristol, Jennifer Foote - Concordia University, Pavel Trofimovich - Concordia University
Drawing on teachers’ perceptions to adapt and refine a pedagogically-oriented comprehensibility scale for
use on university campuses
With an unprecedented number of tertiary students pursuing degrees outside of their home countries, the use
of English as a lingua franca on university campuses is on the rise (OECD, 2012). High-stakes language
assessments which rely on judgments from exam board accredited raters are often used to screen students for
their English language proficiency (e.g., TOEFL, IELTS). However, achieving the requisite proficiency score for
entrance purposes at English-medium higher education institutions by no means guarantees that incoming
students will have sufficient ability to conduct their academic tasks in English, with oral communication often
serving as a major source of difficulty (Spencer-Oatey & Xiong, 2006). University Language Centres could stand
to benefit from the use of a cost-efficient, pedagogically-oriented assessment tool designed to diagnose
sources of difficulty with respect to the aspects of speech most likely to have bearing on the comprehensibility
of their international students’ oral productions. Such a tool could guide teachers in administering feedback by
describing the qualities of comprehensible speech without referring to the native-speaker standard at the high
end of the scale, while simultaneously fostering student awareness-raising.
The goal of this presentation is, therefore, to discuss the development of a user-friendly
comprehensibility scale, intended for formative assessment purposes, for use on Canadian and UK campuses.
The starting point was a “crude” empirically-derived three-level holistic comprehensibility scale, originally
developed from Francophone learners’ picture narratives in the Canadian context (Isaacs & Trofimovich, 2012).
Through a series of focus group sessions involving Canadian and British teacher raters, including crossvalidation
samples (n = 10 raters total; duration: 23 hrs), the scale was adapted and refined for use with learners from
different first language backgrounds on general academic tasks by drawing on teachers’ perceptions. The
discussion will emphasize teacher pronunciation literacy issues and the resulting challenges in elaborating userfriendly scale descriptors.
Okim Kang and Foezur Chowdhury - Northern Arizona University
Prosodic Features in L2 Accented Speech: Human versus Machine
In accented speech, humans use different prosodic cues to convey the communicative values to hearers. In
recent research, suprasegmental features alone were found to account for 50% of the variance in raters’
assessment of oral proficiency (Kang, Rubin, & Pickering, 2010). Thus far in the field of applied linguistics, the
degree of accentedness have been judged either by human impression or by computer-assisted instruments,
which examines some elements of the physical facts of utterances as a supplementary tool. Currently, the
advancement of computing and automatic speech recognition (ASR) technology makes it possible to design
a human-like intelligent machine for oral communication. There have been attempts, albeit still at an
exploratory stage, to include the non-segmental features such as pitch accent contour in the ASR to verify
accented-based speakers. However, the existing ASR largely relies on segmental articulatory-phonetic features
because the ASR model can convert only speech into a stream of phonemes, but cannot model any prosodic
features. Using several algorithms for quantitative analysis, non-segmental features (e.g., speech rate, intensity,
pause duration) have been extracted for the L2 speech assessment (Zechner, Higgins, & Williamson, 2009), but
these algorithms have failed to develop a statistical model for prosodic features. Therefore, the current study
introduces an inventive quantitative model of prosodic features, which can incorporate the intonation tones
and prominence into the system. It will demonstrate a parametric feature extraction technique for tone
choices along with examples of extracted features. In addition, the presenters will discuss the contribution of
suprasegmentals to both human judgments and machine scoring and further explain how a machine
recognizes accented and mispronounced speech produced by language learners. The study will show
potential not only in improving the design of automatic machine scoring systems in L2 accented speech, but
also in advancing a feedback tool in computer-assisted pronunciation teaching.
Paul Keyworth - Saint Cloud State University
The Acoustic Correlates of Stress-shifting Suffixes in Native and Nonnative English
It is widely accepted that certain suffixes in English cause a shift in stress in the root morpheme to the syllable
directly preceding the suffix (e.g., Celce-Murcia, Brinton, and Goodwin, 1996, Kreidler, 2004). These stress-shifting
suffixes are labeled Level 1 [+cyclic] suffixes by generative phonologists (Kisparsky, 1982; Halle and Kenstowicz,
1991). However, these stress-shifts have not yet been scrutinized quantitatively using laboratory phonology
!
15
techniques. Although various studies have been conducted on English lexical stress (ELS) in general, they have
not explored the acoustic properties of the full range of Level 1 [cyclic] suffixes in the lexicon. Also, there is a
lack of consensus on the relative salience of the following acoustic correlates of stress: Fundamental frequency
(F0), duration, and intensity (e.g., Fry, 1955, 1958 vs. Beckman & Edwards, 1994). Furthermore, there is a dearth of
cross-linguistic acoustic data on comparisons of the productions of ELS by native English-speakers (NESs) and
nonnative-English-speakers (NNESs) of different proficiencies and first language (L1) backgrounds. Thus, syllabic
F0, duration, and intensity will be measured using Praat Version 5.3.31 (Boersma & Weenink, 2012) in productions
of seven Level 1 [+cyclic] tokens by NESs and L1-speakers of Mandarin Chinese and Arabic. Relative syllabic
stress ratios will then be calculated. Statistical analyses are expected to provide insights into a possible Acoustic
Correlate Saliency Hierarchy and reveal a main-effect of L1-background in their ordering. L1-speakers of Arabic
– a stress-timed language like English – are anticipated to produce the acoustic correlates of ELS more similarly
to NESs than L1-speakers of Mandarin - a tone language. However, based on the evidence in the literature, the
researcher expects F0 to be over-used as an acoustic correlate of stress by both NNES groups. Finally, the results
may indicate a correlation between L2-input/proficiency and a) native-like placement of stress and b) nativelike ordering of the acoustic correlates of stress in productions by NNESs. These findings would have significant
pedagogical implications for English L2 pronunciation teachers and course-material developers. They may offer
further credence to the idea that Praat can be an effective tool with which to teach suprasegmentals and
provide feedback to learners (Wilson, 2008).
Ettien Koffi - Saint Cloud State University
Assessment of the Intelligibility of [ ! ] in Seven Varieties of L2 Englishes
Munro et al. (1996:328) and Munro and Derwing (2006:493) report the results of perception studies in which they
found that [ ! ] was one of the least well perceived vowels by North American English (NAE) hearers of L2
Englishes. Exploratory acoustic phonetic studies conducted on seven varieties of L2 Englishes support their
findings in part. Indeed, the vowel [ ! ] in these seven varieties of L2 Englishes overlaps acoustically with [ æ ]
and [ " ]. As a result, NAE hearers may have a hard time perceiving [ ! ] accurately. However, confusion data
found in Peterson and Barney (1952) and Hillenbrand et al (1995) also indicate that [ ! ] is among the least well
perceived vowels of NAE. It is perceived accurately 92.2% in Peterson and Barney, and 90.8% in Hillenbrand et
al. The infelicitous perceptions of [ ! ] may be due to the realignment of vowels in the acoustic vowel space
that is going on presently in NAE. As a result, some vowels are encroaching on the acoustic vowel space of [ !
]. Small (2005:79) notes, for instance, that many participants in his acoustic phonetic studies confuse [ ! ] and [
# ]. I contend in this paper that the poor intelligibility of [ ! ] in Munro et al. (1996) and Munro and Derwing
(2006) may have as much to do with the dialect(s) of the intelligibility judges as with the acoustic production of
the L2 talkers. Furthermore, I contend that researchers can gain greater insights into the intelligibility of vowels if
L2 production data is used in tandem with confusion data that is already available in NAE. Doing so can help
us determine whether the intelligibility problem lies with the L2 talkers or with the dialect(s) of the hearers.
Jacques Koreman, Olaf Husby, Egil Albertsen, Preben Wik, Åsta Øvregaard, Sissel Nefzaoui - NTNU, Norway
L1 variation in foreign language teaching: challenges and solutions
To become an active user of a foreign language, pronunciation is a key skill, which lays the foundation for
effective communication. Pronunciation problems typically vary depending on the native language of the
learner. This requires an individualized approach, which often is not feasible in typical classroom situations.
Language labs do offer the possibility of individual instruction. At NTNU, we have developed a ComputerAssisted Listening and Speaking Tutor (CALST) based on a contrastive analysis of segment inventories. The
segment inventories are stored in a database which is based on UPSID and implemented as a wiki. It presently
contains over 500 languages and can easily be extended with new languages. Contrastive analysis results are
visualized as IPA charts in our L1-L2map tool for comparing languages. Segments are color-coded, with red
sound symbols indicating L2 segments which do not occur in the learner’s L1 and may therefore be
challenging. L1-L2map can be used interactively or as a server-client system which returns information to a
computer-assisted pronunciation training (CAPT) system.
Unfamiliar L2 segments are linked to sound contrast exercises in CALST. We shall discuss the selection of
exercises from a large set of relevant contrasts for a given unfamiliar sound, and present a pragmatic
approach to differential substitutions.
CALST has been developed for Norwegian. Because Norwegian does not have an accepted pronunciation
standard, learners must select one dialect as their pronunciation target, while becoming perceptually familiar
16 !
with all other dialects. We shall demonstrate how this is solved in different types of vocabulary training exercises
(Listen&Click, Listen&Speak, Listen&Write) as well as two different types of exercise to train sound contrasts
which are not familiar from the user’s native language (ABX, Minimal pairs/sets).
We shall also explain how the system can be integrated into CAPT systems for other languages.
John Levis, Stephanie Link, Sinem Sonsaat, Taylor Anne Barriuso - Iowa State University
Native and nonnative teachers of pronunciation: Does language background make a difference in learner
performance?
While many teachers feel both that pronunciation is an important skill, they are uncertain that they can teach it
effectively (Bretikreutz, Derwing, & Rossiter, 2001; Burgess & Spencer, 2000; Kawai & Hirose, 2000; MacDonald,
2002). All of these studies examined the beliefs of native speaking teachers. Ambivalent feelings toward
pronunciation teaching are not overcome by the findings that pronunciation instruction leads to pronunciation
improvement in controlled and spontaneous contexts (Saito, 2012); if anything, they may be greatly increased
when the teachers are not native speakers. Unlike the teaching of other language skills, nonnative teachers
may see themselves as inadequate models of pronunciation due to their accents, lack of command of
phonological features, or lack of confidence (Golombek & Rehn Jordan, 2005). However, we do not know if
learners improve their pronunciation skills differently with native or nonnative teachers.
This study reports the pronunciation improvement of learners in two intact classes at an American
university. One was taught by a nonnative and one by a native teacher. The teachers, both PhD students in
applied linguistics, were matched for pronunciation knowledge, teaching skill, and personality. Each taught the
same pronunciation lessons (focused on suprasegmental topics) over seven weeks. The adult students in the
classes were connected with the university, either as students, staff, or those related to the students. All students
took the class as part of the study and not for course credit.
Results show that native listeners’ ratings of the students’ comprehensibility improved similarly with both
teachers, both in controlled and spontaneous speech samples. However, learners showed a preference for
being taught pronunciation by native speaking teachers. The results offer encouragement to nonnative
teachers in teaching pronunciation, suggesting that, like other language skills, pronunciation skills can improve
through knowledgeable teaching practices and that the nativeness of the teachers is not a determining factor.
Takehiko Makino - Chuo University, Tokyo
Pronunciation Characteristics of Japanese Speakers’ English: A Preliminary Corpus-Based Study
The present author is currently developing English Read by Japanese (ERJ) Phonetic Corpus. The Corpus
consists of computer-readable narrow phonetic transcriptions and their corresponding target phonemes of
selected 800 utterances from ERJ speech database, a collection of read-aloud sentences by 200 Japanese
university students. The rationale for the corpus development was the lack of systematic survey of the
characteristics of Japanese speakers’ pronunciation of English. In describing the pronunciation characteristics
of English spoken by Japanese speakers (or speaker of any language), we have been relying on the “rules of
thumb” based on informal observations or theoretical predictions from the L1-L2 phonological differences, such
as L/R confusion or conflation of English vowels into a five-vowel system in the case of Japanese speakers. While
such rules of thumb have had roles to play, corpus-based studies of other areas of linguistic research have
proved that they cannot give us the total picture of what are being studied, and L2 pronunciation should not
be an exception. Indeed, a preliminary survey of the ERJ Phonetic Corpus has revealed some rather
unexpected findings, the most notable of which is the spirantization (fricative realization) of voiceless plosives.
Such a process is not part of standard Japanese phonology and cannot be the case of a negative L2 transfer.
In this paper, the present author is going to point out this or other unexpected findings from the Corpus, and
argue that they should be born in mind in the teaching of English pronunciation to Japanese speakers. He is
also going to talk about the future development of the Corpus, such as the addition of prosodic notations.
Shannon McCrocklin – Iowa State University
Dictation Programs for Pronunciation Learner Empowerment
In language learning, students often recognize a need or desire to improve their pronunciation. Language
programs, however, often neglect pronunciation instruction and students are left to their own devices to
improve. In this situation, many students may feel powerless to work on their pronunciation alone without an
instructor; they do not know how to work autonomously. Additionally, even when pronunciation is addressed in
!
17
more traditional pronunciation classrooms, many pronunciation classroom activities still rely on the teacher to
model “correct” pronunciation and to monitor, evaluate, and give feedback on student production.
Pronunciation teachers also often rely on drills or controlled production activities, giving students little room for
free expression or communicative practice. These types of pronunciation classes seem unlikely to foster student
autonomy because students are not encouraged to develop skills or strategies for monitoring or evaluating
their own pronunciation and are given very little room for free experimentation with the language and
specifically their pronunciation. Students need strategies, skills, or tools that can empower them to experiment
with pronunciation, without relying on the teacher for constant monitoring and feedback, tools that will help
students become more autonomous as pronunciation learners. This presentation will discuss a piloted mixedmethods research study examining whether the use of Windows Speech Recognition (WSR), a dictation
program utilizing Automatic Speech Recognition technology already installed on most PCs as part of Windows,
as part of a hybrid pronunciation class can help foster learner autonomy more than traditional face-to-face
instruction. Survey results indicate that the hybrid course group developed a greater sense of autonomy.
Interview results suggest that the incorporation of WSR gave students a clear strategy for practice outside of
class, expanding their repertoire of available practice strategies.
Shannon McCrocklin and Stephanie Link - Iowa State University
What is identity? ESL and Bilinguals' Views on the Role of Accent
Many researchers and theorists have proposed a connection between accent and identity (Guy, 1988; Ochs,
1993; Setter and Jenkins, 2005; Pierce, 1995; Weedon 1987). Some, however, have gone beyond this, indicating
that students fear obtaining a native speaker accent. “To speak an L2 like a native is to take a drastic step into
the unknown, accompanied by the unconscious fear of no return…” (Daniels, 1995). Yet, this comment may
strike many teachers and researchers as surprising because as Sobkowiak points out, “in my whole teaching
career I have not met a [student] who would not like to sound like a native, or who would fear to step on this
‘road of no return’” (2005). Perhaps the difference in perspective can be explained by a difference in
language learning group, those that have successfully spoken English like a native and those that have not. This
research study examines the perceptions of ESL learners and English speaking bilinguals. Subjects participated
in a semi-structured interview to discuss their experiences interacting with others and their perceptions of
accent and identity. Findings from the interviews suggest that these two groups have very different ideas
about accent and identity. Results from this study provide insights into the possible misconceptions and
assumptions that underscore our work as educators and researchers and can hopefully be used to inform
future teaching in the field of pronunciation.
Murray Munro - Simon Fraser University
What do you know when you “know” an L2 vowel?
From a production standpoint, L2 pronunciation instruction focuses on developing specific kinds of procedural
knowledge to be implemented across a wide range of phonetic contexts and discourse types. Teachers aim to
help learners establish the articulatory routines they need to know in order to produce accurate vowels,
consonants and prosody. But what does it mean to say that a speaker “knows” how to produce a particular
speech sound? This investigation addresses the question by examining productions of pairs of English high
vowels /i $/ and /u #/ by Cantonese speakers. Analyses indicate that vowel intelligibility is highly dependent on
phonetic context and is subject to the effects of task, word frequency, and apparently random variability
within speakers. This outcome suggests that “knowing” a vowel is far from an all-or-nothing phenomenon.
Rather, L2 speakers appear to develop segmental knowledge that varies in its accessibility in ways that are only
partially predictable. Effective pronunciation teaching requires careful attention to the nature of this
production knowledge.
Murray Munro - Simon Fraser University, Tracey Derwing - University of Alberta, Ron Thomson - Brock University,
Debra Elliot - University of Alberta
Naturalistic L2 Segment Development: Implications for Pedagogy
Contemporary views of pronunciation instruction emphasize the development of intelligible speech using
empirically-validated pedagogical principles. Because learners typically have limited time for pronunciation
work, efficient instruction is important. Achieving that goal requires an understanding of adult phonetic
learning. Computer Assisted Pronunciation Instruction (CAPT) can potentially meet this need, if guided by
18 !
evidence about learners’ acquisition processes; however, information on L2 phonological development is
limited. In this longitudinal investigation, we examined consonants and consonant clusters in English learners
from two language backgrounds over two years. We were interested in knowing what general learning
trajectories occur for a selection of English singletons and clusters in initial and final position; which targets
showed excellent performance, either at the outset of the study, or by the end; to what extent speakers from
two different L1 backgrounds share difficulties with the consonant targets; and finally, to what extent the
speakers within each L1 group have similar difficulties. Extensive between- and within-group variability was
observed, with some targets produced well at the outset, and others improving over time. The results argue
against a common curriculum for learners. Instead, individualized pronunciation instruction is essential for
effective use of class time, particularly for consonants. Given current technological capabilities, a CAPT system,
which develops the perceptual skills underlying satisfactory production, in conjunction with production
opportunities in class is best implemented with guidance from trained pronunciation instructors who can assess
learners’ needs and monitor their progress.
Moonyoung Park and Sarah Huffman - Iowa State University
The Potential of ASR for Non-native English Speakers in Air Traffic Control
Air traffic control (ATC), a form of spoken communication between pilots and air traffic controllers, involves a
high-stakes exchange with immediate and potentially severe consequences to human life. In the last decades
of the 20th century, more than 35 major aircraft crashes, and countless more near misses, were found to have
been caused or influenced by miscommunications pertain to non-native English speakers’ English language
ability communication errors. Consequently, International Civil Aviation Organization (ICAO) policy now requires
non-native English speaking pilots and air traffic controllers to demonstrate their aviation English ability to speak
and understand the language used for radiotelephony communications. Among the ICAO’s English language
proficiency requirements, pronunciation has been given high priority for both aviation English listening and
speaking skills. In the area of pronunciation learning, automatic speech recognition (ASR) technologies have
been used in computer-assisted pronunciation teaching for two primary purposes: teaching and assessing the
pronunciation of a target language. In this study we evaluate the potential usefulness of ASR software as a
pronunciation teaching and assessing tool for non-native English speaking air traffic controllers in the context of
army aviation by comparing ASR and human judgment. This study also investigates the problems in non-native
English speaking air traffic controllers’ pronunciation that affect human listeners’ understanding of the ATC
speech. We recorded spoken samples of three Korean army air traffic controllers’, two females’ and one
male’s, aviation English phraseology, as they performed an aviation English training task. For the ASR
evaluation, speech samples were evaluated using the software Dragon Dictation. For the human evaluation,
two native English speakers and two advanced level Korean ESL speakers rated the spoken samples. Findings
from the study point to potential uses for ASR technology in providing corrective feedback to non-native English
speaking air traffic controllers. Results likewise show the value of human raters’ evaluation of specific aspects of
target language pronunciation including accentedness, comprehensibility, and intelligibility.
Marnie Reed - Boston University
Connecting pronunciation to listening: Raising learner and instructor awareness
Learner surveys reveal two listening challenges: word segmentation in connected speech; understanding the
words but not the message. Instruction surveys reveal two deficiencies: listening ability is tested but not taught;
unoperationalized curricular goals offer no assessment or instruction guidelines. A strategy-based
metacognitive approach raises learner and instructor awareness and skills.
As noted by Vandergrift and Goh (2012), listening is a skill which many second language learners report
wanting to learn but for which they often receive misdirected instruction. Despite nominal recognition as one
of the four skills, listening is rarely taught as a language skill. A contributing factor is that curricular guidelines
typical of many ESL programs, such as 'students will be able to comprehend lectures and engage in academic
conversation,' often leave the task of operationalizing these goals to ill-prepared instructors.
For curricular goals to be realized requires identifying the requisite listening comprehension and speech
production skills students need, and providing instructors with the means to teach those skills and to assess
whether students have attained them at the metacognitive and procedural levels. A study undertaken in an
academically-oriented intensive English program pronunciation elective provides encouraging support for a
metacognitive strategy approach proposed by Goh (1997, 2008) and Vandergrift (2004, 2007).
!
19
The study investigated whether training in the production of connected speech features improved
listening for communicated content, and whether training in prosodic features such as contrastive stress and
intonation improved listening for communicative intent. Combined pronunciation and metacognitive training
promoted recognition of known words in the context of rapid, connected speech, and strategies for
production and recognition of the pragmatic functions of intonation promoted ability to infer an utterance's
illocutionary force on the basis of non-standard stress and intonation. Findings indicate that production
accompanied by metacognition improves parsing skills and discourse-level learner awareness and attainment
of speaker intent.
Ashley Roccamo - Penn State University
Learning Pronunciation in Just Ten Minutes a Day: Adapting Pronunciation Training to a Four-Skills German
Classroom
Pronunciation in a second language (L2) is a notoriously difficult skill to acquire. In fact, most second
language (L2) learners do not acquire target-like pronunciation on their own, sometimes to the point of
hindering communication. Yet actual language classroom practice lags behind these research findings.
Oftentimes instructors feel that the time constraints are too tight in a traditional, four-skills classroom, causing
pronunciation instruction to be neglected in favor of grammatical or cultural topics. This paper presents a
solution to this problem, bridging the gap between linguistic research and language teaching by providing
research-based, pronunciation training units for the everyday language instructor that are easy to implement in
a short period of time.
This paper presents the results of a unique eight-week pronunciation training unit designed as a series of
four modules to supplement to both first-semester and fourth-semester German language classrooms. Training
was divided into four two-week modules, each training a specific area of German pronunciation for just 10
minutes a day. Target areas were chosen both for their difficulty for L2 learners as well as their importance for
communicating meaning, and included lexical stress, palatal and velar fricatives ([ç] and [x]), fricative and
vocalized /r/, and the monophthongization of [e] and [o]. Within each module, lessons took the form of warmup activities designed to improve both perception and production skills through a variety of both individual and
partner activities and games. Student recordings from both the experimental and control groups were
collected through a pre-test / posttest design, rated by native speakers for accentedness and
comprehensibility, and compared to measure improvement after training.
The module-style design implemented in this study has a variety of benefits. It allows for easy
adaptation for use in a variety of four-skills foreign language courses while remaining manageable for
instructors and retaining effectiveness for learners. It also avoids taking too much valuable class time away
from practicing the other language skills, particularly vital in beginner-level courses. Breaking pronunciation
training units into modules also provides for an a-la-carte style in which teachers may self-design pronunciation
training based on their students’ most prominent needs. By applying research on L2 pronunciation training
directly to an educational setting, this study provides insights into the development of a pronunciation
instruction unit that provides instructors with useful, tested pedagogical techniques to help improve their
students' pronunciation abilities on a limited time schedule.
Mari Sakai and Colleen Moorman - Georgetown University
Can perceptual training improve production of L2 phones? A meta-analytic review
In 1988, Flege began publishing a series of articles that later developed into the Speech
Learning Model of second language (L2) learning (e.g., Flege, 1988, 1992, 1995, 1999,
2002). Two postulations of this model are that native language phonemic inventories can be altered to include
new L2 sounds, and that accurate perception of L2 sounds must be achieved before accurate production can
occur. In the ensuing 25 years, numerous researchers have attempted to empirically test these hypotheses.
The current investigation employed meta-analytic procedures to summarize findings from experimental and
quasi-experimental studies on the effectiveness of perception training on production outcomes, published from
1988 to the present. Due to the many iterative procedures involved in meta-analysis, a pilot study was
conducted to proceed through all conventional metaanalytic stages (i.e., search and retrieval,
inclusion/exclusion criteria, coding procedures, and effect size calculations). First, 126 potentially relevant
studies were identified for use in the pilot sample. Sixty-two of the studies were retrieved and filtered through
seven inclusion and three exclusion criteria. After this process, a pool of 19 primary studies remained. Using a
20 !
portion of these studies, a coding scheme was developed, piloted, and finalized. Data from five studies were
extracted to calculate 24 unique effect sizes: 12 for perception outcomes and 12 for production outcomes.
The pilot data indicated that all types of perception training led to improvement in both perception and
production modalities with a range of effect sizes. A positive correlation also existed between perception and
production effect sizes. However, given the small sample size of this pilot, more studies need to be analyzed
before results can be generalized. Currently, the full scope of this project is being conducted, and these results
are expected to contribute valuable information to the field of L2 phonological acquisition.
Veronica Sardegna - University of Texas
Non-Native Teachers’ Identity Formation as Qualified Pronunciation Teachers
The expansion of the pronunciation-teaching model to focus on learner comprehensibility and intelligibility
(Levis, 2005) affords non-native teachers the opportunity to assert their legitimacy not only as English
pronunciation teachers, but also as English language speakers (Golombek & Jordan, 2005). Unfortunately, even
after acquiring pedagogical knowledge and skills to teach pronunciation, many non-native teachers still feel
unqualified to teach English pronunciation due to feelings of anxiety and self-consciousness in their identity as
English speakers and teachers. Knowledge is empowering, but anxieties and fears can also be very inhibiting
(Horwitz, 1996). This study investigates non-native pre-service teachers’ journey and identity formation as
qualified English pronunciation teachers while attending a graduate MATESL pronunciation course. This course
empowered the teachers with knowledge and skills to teach pronunciation, but also provided them with
opportunities to use those skills in tutoring sessions within a safe and collaborative environment. The non-native
teachers collaborated with native teachers also taking the course, and their teaching was scaffolded by the
course instructor through lessons, individualized feedback, and observations. This paper focuses on the nonnative teachers’ (N = 13) journey preparing and executing lessons for their tutees (N = 17), and analyses the
effectiveness of the tutoring as well as the participants’ reflections on the experience. Data were collected
from tutees’ recorded pre- and post-read-aloud assessments; tutee and tutor questionnaires; and tutor
portfolios with lesson plans, observation reports, and reflections. The findings revealed that the tutoring was
effective in improving the tutees’ reading of English reduced vowels, contracted words, intonation contours,
primary phrase stress, and linked sounds; (b) the tutees perceived the instruction as both effective and
personally empowering, and (c) the tutors perceived the experience not only to be personally gratifying but
also to be instrumental in building their confidence and knowledge-base to teach pronunciation.
Joan Sereno, Lynne Lammers, and Allard Jongman - University of Kansas
Perception of foreign-accented speech
When adults learn to speak a second language, their non-native speech pronunciations often deviate from the
native norm, resulting in speech that sounds accented and may be difficult to understand. These difficulties in
second language pronunciation typically include both segmental and suprasegmental features. Segmental
errors are errors in the production of individual consonants and vowels, while suprasegmental errors include
errors in stress assignment, intonation, timing, phrasing, and rhythm. The goal of the present study is to determine
the extent to which segmental and suprasegmental information, specifically intonation, affect the
accentedness, comprehensibility, and intelligibility of foreign-accented speech. In the present study using
digital signal processing techniques, we acoustically separate segmental features from suprasegmental
features and evaluate the contribution of each to perceived foreign accent. For a set of sentences, we
separated the intonation from the segmental content for both native and non-native speech samples and
superimposed the intonation of one onto the segmental content of the other and then presented these
manipulated sentences to a set of 40 native speakers for their judgment. Results will be presented in terms of
the separate contribution of segments and intonation to accentedness, comprehensibility, and intelligibility.
Christin Shea and Jennifer Vojtko Rubi – University of Iowa
Dialect adaptation and L2 Spanish listeners
While walking past classrooms where college-level Spanish courses are in progress, one quickly notes the wide
variety of dialects found in the speech of instructors – from Spain, Latin America and also non-native Spanish.
Indeed, students often comment ‘it was hard to understand my Spanish teacher at first, but then I got used to
!
21
it’. This is the result of dialect adaptation, or the adjustments carried out by listeners to the
phonological/phonetic/indexical characteristics of a new dialect. In this study we explore the initial stages of
dialect adaption by examining whether different Spanish varieties prime one another in a lexical decision task
and how familiarity with the dialect modulates this.
We used short and long-term repetition priming to examine how effective cross-dialect variants are at
activating lexical items and how each dialect variant is encoded long-term in speaker representations.
Speakers were from Valladolid , Spain (Dialect 1, UNFAMILIAR), which is a variety that exhibits distinción, or the
interdental fricative [%] instead of [s] for the written forms of z, ci and ce and Mexico City (Dialect 2, FAMILIAR), a
variety that does not have distinción and does not elide [s] in codas. Critical items exhibited the [%] – [s]
difference.
We predicted that the familiar dialect would prime the unfamiliar dialect but not vice versa. Participants were
low-intermediate/high intermediate undergraduate Spanish majors and minors. For the repetition priming
paradigm, preliminary results show that the familiar dialect does not prime the unfamiliar dialect, but the
unfamiliar dialect does prime the familiar dialect. In the long-term repetition priming, however, both the familiar
and unfamiliar dialects primed each other. These results suggest that experience and familiarity (i.e., in terms of
similarity of phonological/phonetic forms to the L1) affect lexical recognition in L2 dialects.
Sinem Sonsaat and Stephanie Link - Iowa State University
How do nonnative teachers use pronunciation materials? Implications for materials development
Textbooks and teacher books are crucial components of actual classroom practice both for students and
teachers. They are recognized to be the main sources of enriched input, especially in non-English-speaking
contexts (Tomlinson, 2003). The support of materials can be influential in increasing the self-confidence of
novice teachers and teachers who lack sufficient training (Tomlinson, 2003). There are likely to be several
different ways of using materials and even if the same materials are used by many teachers, the way they are
used may vary depending on the teachers’ experience, expertise or being a native or nonnative speaker of
the target language. Likewise, what teachers expect from materials may show differences as well.
We investigated how two teachers of English (one native and one nonnative speaker) used the same
pronunciation teaching materials (Levis & Muller Levis, 2013). The teachers taught mainly suprasegmental
features to graduate students at a Midwestern US university. The teachers were similar to each other in terms of
their knowledge of language theories, subject matter knowledge, teaching experiences and personal
characteristics. Written and verbal reflections of the teachers’ teaching practices were collected through
teaching notes and weekly focus group meetings with a pronunciation expert after teaching each week. There
were important differences between these two teachers in terms of their use of materials, and expectations
related to the materials. The nonnative teacher prepared more meticulously, but was more likely to distrust
herself, and was more uncomfortable in changing plans. In many ways, the materials were inadequately
informative for her. Specifically, the materials needed to provide greater specificity on answers, more explicit
guidance on linguistic and pedagogical assumptions, and more careful direction on suprasegmental features.
A lack of detailed guidance seems to result in an extra burden and potential frustration for nonnative speaker
teachers during preparation and teaching.
Shelley Staples - Northern Arizona University
Prosodic patterns in nurse-patient interactions: a comparison of international and U.S. nurses
Prosodic features (e.g., prominence, pitch, and intonation) play a major role in effective information
packaging and can also influence perceptions of listeners about speaker affect (Brazil, 1997; Pickering, 2001).
However, prosody is often excluded from oral communication courses for professionals. This is despite the fact
that advanced non-native speakers may still lack mastery of expected prosodic patterns of English (Kang, 2010;
Kang, Rubin, & Pickering, 2010; Pickering, 2001; Wennerstrom, 2001). Previous research has shown that in
professional contexts, such as classroom teaching, deviation from native speaker norms may have a negative
impact on the ability of the speaker to effectively communicate information to listeners and to develop rapport
(Pickering, 2001). Examination of both native and non-native English speakers’ use of prosodic features in
classroom teaching has led to greater prioritizing of pronunciation in courses for international teaching
assistants. Prosodic patterns presumably have a similar impact in other professional contexts, such as nursing. As
in the teaching context, developing rapport and effectively communicating information are essential to
22 !
successful nurse-patient interactions. However, the prosodic patterns used by nurses have not previously been
explored and are not included in language courses for nurses in training. In order to investigate prosodic
differences between native English speaking (U.S.) and non-native English speaking (international) nurses, a
corpus of 102 interactions between nurses and patients was examined (50 U.S. nurse/patient and 52
international nurse/patient). Key speech acts within the interactions (e.g. greetings and expressions of
empathy) were analyzed using instrumental and auditory methods. Differences were found between the two
groups in their use of prominence, tone choice, pitch range, and paratones (use of pitch to mark shifts in
discourse level topics). These findings have implications for language courses for international professionals in a
variety of contexts.
Jessica Sturm - Purdue University
Effects of Instruction on Voice Onset Time in word-initial /p/ for L1 American English students: A Preliminary
Study
Pronunciation is a traditionally neglected aspect of L2 teaching and learning. In French, specifically, this can
be problematic due to the opacity of its orthography. One reason for this neglect is the perception that
pronunciation instruction is ineffective; in other words, it is often believed that learners will never achieve nativelike pronunciation, so why bother? However, non-native-like pronunciation can lead to problems of intelligibility
or perception of foreignness in regards to an “accent.” The present research investigates the efficacy of
phonetics and pronunciation instruction on learners of L2 French, specifically the Voice Onset Time (VOT) of
word-initial /p/, which is longer in American English (where word-initial voiceless stops are aspirated) than in
French (where they are not aspirated). This investigation was accomplished by comparing the progress made
by university students enrolled in a French phonetics and pronunciation course (n= 10) with a control group of
other advanced L2 French students who are not enrolled in phonetics (n = 11) and comparing both groups to
native speakers of French (n = 11). VOT for all three groups was determined using Praat. The researcher
measured, in seconds, the VOT of speakers’ and learners’ word-initial /p/ in recordings of an assigned text.
Learners recorded the text at the beginning and end of the semester and their recordings were compared to a
single set of native speaker recordings. Results indicate that some learners in the experimental group did
reduce their VOT of word-initial /p/ to the native-like short-lag range while the control group learners’ VOT
remained relatively constant.
Ron Thomson - Brock University
Does vowel learning in one context generalize to other contexts?
Implicit in many pronunciation curricula and materials is the belief that training learners how to perceive and
produce speech categories in one context will generalize to novel contexts. In the second language (L2)
speech perception literature, however, many hold the opposite view. Flege’s (1995) Speech Learning Model
has been the catalyst for research demonstrating that knowledge of a sound category in one phonetic
context does not predict knowledge of the same category elsewhere. This view has recently been supported
by Thomson (2012), Thomson, Nearey and Derwing (2009), and Thomson and Isaacs (2009). The present study
extends this line of research by investigating English vowel learning by 15 L1 Spanish speakers in 12 separate
phonetic contexts. Following the High Variability Phonetic Training paradigm (Logan, Lively & Pisoni, 1991),
participants were trained to perceive ten English vowels during 40 computer-based training sessions, over a
period of approximately two months (with nearly 8000 training tokens per learner). While learners were taught
the same vowels during each session, the phonetic context in which vowels were presented varied. Results
indicate that while training improves learners’ vowel perception within specific contexts, generalization of
learning from one context to another does not occur. These findings appear to have major implications for
pronunciation curriculum, indicating, for example, that instructors need to provide learners with vowel
instruction in all possible phonetic contexts. Recommendations for further research will also be discussed.
Anne Violin-Wigent - Michigan State University
Comparing online vs. face-to-face classes: A case study of a French pronunciation class
This presentation intends to compare the outcome of a traditional face-to-face (F2F) class with its online
equivalent. The presenter has been teaching a French phonetics and pronunciation class F2F for a numbers of
years and is scheduled to teach it online during the summer of 2013. Based on comparable assignments,
outcomes will be evaluated by comparing class averages on the last assignment (to see if students have
!
23
reached a similar level) as well as of each student’s progress between the first and last assignments (to
determine if similar improvement is achieved).
It is anticipated that the students in the online class will improve their pronunciation of French more so
than those in the F2F class in spite of the slightly shorter duration (13 vs. 15 weeks). It is hypothesized that such a
difference may stem from the type of feedback received: students in the online class will receive individualized
feedback prior to graded assignments whereas those in the F2F environment received class-wide feedback.
Additionally, accuracy in transcription activities will be compared for a few chapters deemed more
difficult for students, such as liaisons and schwa, among others. For these, it is anticipated that the average
grade for students in the online class will be higher than those in the F2F class because of several factors. First,
due to the semi self-paced format of the online class, students may choose to spend more time re-listening to
“lectures” and re-doing practice activities, hence receiving more input and explanations. Additionally, the
feedback they receive is much more individualized than in a traditional setting.
Finally, students in the online class may not have as many other classes or commitments, since it is
taught during the summer.
Patricia Watts, Amanda Huensch, and Lisa Pierce - University of Illinois
Attainable Targets for L2 Learners: How Proficient L2 Speakers can Bridge the Gap
Recently, the emphasis in pronunciation instruction has shifted from accent reduction to helping
learners achieve intelligibility and comprehensibility (Derwing & Munro, 2005). This shift calls into question the
tendency to rely exclusively on native-speaker (NS) models as targets for learner speech. Munro & Derwing
(1995) have found that even highly accented speech may be both intelligible and comprehensible. This
argument is further strengthened on the basis of affect and motivation. As Cook (2002) noted, teachers and
learners who rely solely on NS models can easily “become frustrated by their attempts to reach what, in effect,
is an impossible target” (p. 331).
Given that most pronunciation textbooks do not include NNS models, teachers must find another source.
Our solution is online speech accent archives, which contain a multiplicity of voice files from native and nonnative speakers of English around the world. Archives—such as the one founded and maintained at George
Mason University—provide rich classroom and self-study resources for learners. In addition to providing access
to models of English pronunciation that are both intelligible and attainable, these archives help learners
become familiar with a variety of accents. With the increasing use of English in global contexts, controlled
exposure to different accents aids students in real-world listening tasks, which often take place between NNSs.
Despite the distinct appeal of speech databases, relatively little work has been done to explore their
pedagogical applications.
This presentation aims to familiarize the audience with ways to utilize accent archives by demonstrating
classroom-tested activities and samples of students’ performances. Featured activities include:
1) Comparing NS and NNS models for suprasegmental features (e.g., stress, intonation, linking)
2) Shadowing attainable target models
3) Identifying common features of NNS speech
Session-attendees will leave with innovative ways to use accent archives as well as tips for creating their own
archive-based materials.
Beth Zielinski - Macquarie University
Demystifying comprehensibility for the language teaching curriculum
Comprehensible speech is essential for successful communication, and thus an important goal for many L2
learners of English. Despite its importance, however, there is a general lack of evidence-based guidelines for
teachers on which pronunciation features they should prioritize in the language teaching curriculum to improve
learners’ comprehensibility.
In this paper I will explore the impact of different pronunciation features on the comprehensibility of 177
utterances produced by three different L2 speakers (L1: Korean, Mandarin and Vietnamese), and judged by a
group of native listeners to be difficult to understand. Each of the utterances was played to and
orthographically transcribed by three native listeners and the relationship between the phonological features in
the speech signal and the listeners’ response to that signal was analysed. This allowed me to explore the extent
to which different features contributed to the compromised comprehensibility of the utterances. The features
investigated were identified in previous studies as possibly contributing to judgements of comprehensibility and
include: word stress errors (Isaacs & Trofimovich, 2012; Zielinski, 2008), sentence stress errors (Hahn, 2004),
24 !
segments in strong and weak syllables (Zielinski, 2008) and segments with high and low functional load (Munro
& Derwing, 2006).
The findings reported here have important implications for establishing pronunciation priorities in the
language curriculum and contribute to much needed evidence-based guidelines to teachers.
Yuan Zhuang - Northern Arizona University
Suprasegmentals and second language teaching: A meta-analysis
As recent research suggesting that the suprasegmental aspects of pronunciation may have the most effect on
oral proficiency and comprehensibility (e.g., Kang, 2010), there is a revival of teaching suprasegmentals over
the last twenty-five years (e.g., Chun, 2002; Saito, 2012). A number of studies have examined the pronunciation
outcomes associated with suprasegmentals instruction. However, it has not been able to reach a satisfactory
level of consensus on the degree of effectiveness of pronunciation teaching or how effective pronunciation
teaching really is. Studies have been conducted in different settings (FL or SL, classroom or lab), using various
teaching and testing tasks with different pronunciation foci. Other factors, including learners’ L1, learners’ age,
learner’s proficiency level, the duration of the pronunciation teaching, the involvement of new technology in
teaching, etc., also vary in studies. Experimental or quasi-experimental studies need more thorough probe into
how English suprasementals are learned and taught. It is necessary to have a more systematic way to
synthesize these results. Meta-analysis, as a statistical process whereby the findings of several studies are pooled
in an effort to draw inferences as to the meaning of a collective body of research (Cooper, 2010), will provide
a clearer understanding of the effectiveness of explicit suprasegmental instruction on the L2 learners’
pronunciation learning. This meta-analysis was conducted on the literature that examined the teaching of
suprasegmentals in ESL and EFL settings. Data from 11 studies (involving 1,369 participants) were extracted and
analyzed following established protocols and procedures for conducting systematic reviews and guidelines for
meta-analysis. Results provided valuable L2 pedagogical implications by showing that ESL/EFL learners’
suprasegmental improvement is reliably associated with explicit teaching. Subgroup mean effect sizes varied
from small to large, depending on the suprasegmental features taught and measured, and were moderated
by methodological features of the primary studies.
Erin Zimmerman - Iowa State University
Teaching the Teachers: How Do Pronunciation Textbooks Aid Inexperienced Teachers’ Pedagogy?
The issues that affect pronunciation instruction include learner variables, teacher capabilities, curriculum, class
size, available time, and available resources. Some of these issues are fixed (teacher capabilities, curriculum,
time), while others might change throughout the term (learner variables, available resources); some of these
issues are known before the term begins (time, curriculum), while others might not be (learner variables, class
size). Any instructor teaching pronunciation must address all of these issues when creating everything from the
term schedule to daily activities, and some of these variables are easier to plan for than others. While teachers
who have been teaching pronunciation for years might not be daunted by these tasks, an inexperienced
teacher likely would be.
So do textbook writers expect instructors to be able to take a textbook and adapt it to a particular
class’s needs? If not, how do writers of pronunciation textbooks and teacher’s manuals attempt to guide
novice pronunciation teachers through the teaching process? What suggestions or supplements do they offer
instructors for adapting the activities in the book for students at different levels of learning or who have different
goals for their language abilities?
This presentation will discuss findings from five of the most popular pronunciation textbooks and
companion teacher’s manuals to determine how well they advise inexperienced instructors on adjusting
materials to an individual learner’s or individual classes’ needs. Using grounded theory, I examine the ways
these textbooks and teacher’s manuals provide information to teachers to uncover what information teachers
are expected to know and how this likely affects inexperienced teachers. This research may be helpful for TESL
programs that invest time and resources to train instructors in teaching pronunciation. It could also shed some
light on what the current textbook writing/design practices are and whether they are meeting the needs of
novice instructors.
!
25
Poster Abstracts
(Listed Alphabetically)
Saowanee Alexander - Ubon Ratchathani University, Thailand
Intonation and Perceived Sincerity in EFL and ESL Learner Apologies
The study explores learner apologies through native speakers’ perception with a focus on relationship between
sincerity and intonation. Thirty Thai learners of English (15 EFL, 15 ESL) and ten native speakers of American
English role-played in apology scenarios in scripted and unscripted task conditions. The apologies were
recorded and played to 45 native speaker judges who determine whether or not the apologies were actual
apologies. The analyses were conducted to examine the extent to which the speakers were successful in
conveying their intent, task effects, and intonational correlates of sincerity or lack of it. The perceptual results
show that while all speaker groups’ sincere apologies were perceived as intended, both groups of learners
failed to convey their ostensible intent. There was also a task effect in which learners’ scripted apologies
tended to be perceived as more sincere than their ostensible counterparts. Regarding intonation, a high pitch
accent and a low boundary tone are associated with sincere apologies while accented first-person pronoun I,
a double pitch accent, and a high boundary tone are associated with ostensible apologies.
Joe Barcroft and Mitchell Sommers – Washington University in St. Louis
Better L2 Pronunciation Is One of the Many Benefits of Acoustically Varied Input
Acoustic variability refers to variations in the speech signal other than those that affect linguistic content. In
second language (L2) instruction, input can be made more acoustically varied by increasing the number of
talkers, speaking styles, or speaking rates used to produce it. This presentation will review research on the
effects of acoustically varied input in two areas of L2 instruction: (1) the perception and pronunciation of the
challenging /r/-/l/ contrast in English for native speakers of languages that do not have this contrast; and (2) L2
vocabulary learning with a variety of different languages (English, Spanish, Russian, etc.).
Studies in the first area have demonstrated that training learners using multiple instead of single talkers
leads to not only better perception of the /r/-/l/ contrast (e.g., Logan, Lively, & Pisoni, 1991) but also better
pronunciation of words containing the contrast (e.g., Yamada, Tokhura, Bradlow & Pisoni, 1995). Studies in the
second area have demonstrated greatly improved L2 vocabulary learning when multiple talkers, speaking
styles, or speaking rates are used to present target words (Barcroft & Sommers, 2005; Sommers & Barcroft, 2007).
Other sources of variability, such as amplitude and fundamental frequency (FO), yield no effect, unless they are
phonetically relevant to learners, such as when speakers of a tonal language (Zapotec) hear target
vocabulary with increased FO variability (Barcroft & Sommers, 2012).
The presentation will emphasize benefits of acoustic variability across different levels of linguistic analysis.
Concrete methods for incorporating acoustically varied input in L2 instruction will also be discussed.
Catrice Barrett – University of Pennsylvania
Laying a Foundation for Rhythm-Based Pronunciation Instruction
Speech rhythm is a suprasegmental feature of natural speech that varies across languages. Linguists have
found that speech rhythm variations are realized according to features such as syllable duration and intensity
(volume) (Dauer, 1981; Setter, 2006). These differences lend a particular "character" to an individual's speech
rhythm, thus creating a rhythmic contrast between a language like Dutch versus Cantonese, for example. For
the proposed poster presentation, I will present data from a pre-dissertation pilot case study. In this study I
perform an acoustic phonetic analysis in order to compare the speech rhythm of an American English speaker
and a (Mandarin) Chinese L2 English speaker. Due to the fact that Mandarin speech rhythm is very distinct
from English speech rhythm, the goal is to determine how much variation is realized between the two speakers
and on which features. Studies in applied linguistics have shown that certain L2 productions of speech rhythm
can lead to intelligibility issues in communication (Anderson-Hsieh, et al., 1992). Through the presentation of this
data, I aim to illustrate two major points to the audience: a.) that the analytical framework of this study is a
basis for further analyzing the development of students’ speech rhythm; and b.) that the features of speech
rhythm which vary significantly for the L2 speaker can be used to guide rhythm-based pedagogy. Rhythm-
26 !
based pedagogy draws on the rhythmic devices of poetry, rap and other forms of music to develop English
language learners' pronunciation.
Ling Cai – Iowa State University
An efficient method to build up native sounds in Chinese teaching: Multi-sensory and multi-cognitive
approaches
Due to the great distance between Chinese and English (or other letter languages), learning Chinese
pronunciation for adult learner is rather challenging. The Chinese language adopted a Romanized spelling
system of phonetic symbols, hanyu pinyin, used it to annotate standard Chinese sounds that are not
represented in Chinese script. Usually adults display considerable cognitive conditioning and bias to the native
language phonetic and phonological rules when they are learning/acquisition FL (Odisho, 2007). There are
some vowels (o, u, ao-diphthong), and consonants (j,q,x,z,c,s) in hanyu pinyin that are pronounced differently
from the Standard English phonology, and some sounds have no equivalent match in English (zh, ch, sh, ü)
which causes difficulties for English speakers.
This presentation intends to share some practical teaching tips to help the learner to identify and recognize the
Chinese sounds, furthermore to build up their native sounds by using multi-sensory and multi-cognitive
approaches. If you like to sing, you will definitely appreciate building a bridge between vocal practicing/lip
configurations in music and learning Chinese pronunciation.
Claudia Cárdenas- Universidad Technológica de Pereira
Scaphoning your Language
Scaphoning your language deals with a project that is being executed in a Language Teacher Education
program which attempts to empower learners with the appropriate strategies to scaffold their pronunciation
skills. The term “scaphoning” has been coined along the completion of this project and resulted from the terms
scaffolding and phonemes which are the core concepts of this research. The former concept has been
considered within this study, since it embeds elements from the constructivist theory of learning that includes
“the elements of the task that are initially beyond the learner’s capacity, thus permitting him to concentrate
upon and complete only those elements that are within his range of competence” (Wood et al, 1976), and the
latter that focuses on the phonemes that refer to the different sounds within a language.
This paper is mainly intended to present a research project concerning the experience of implementing
collaborative study groups (CSG) as a means to enhance and promote metacognitive processes on
Pronunciation courses in a Language Teaching program. Authors such Celce-Murcia, Brinton, & Goodwin,
(2010) argue that the fact of exposing the learners to the appropriate strategies facilitate the development
and achievement of academic goals. Following an action-research model that embedded planning, acting,
observing and reflecting stages, this research study has shed light on the need of incorporating certain tools
that allow language learners or teachers-to-be to monitor their L2 pronunciation learning. Considering this, preservice teachers are instructed on the use and application of four learning pronunciation strategies (LPS) which
in Pawlak's (2010) view are defined as deliberate actions and thoughts that are consciously employed, often in
logical sequence, for learning and gaining greater control over the use of various aspects of pronunciation.
The aforementioned strategies are: Critical listening, transcription, annotations and rehearsing corrections
aloud. Such strategies are carefully completed in the track of the project; that is why, CSGs were embedded as
part of the course considering that Collaborative learning has been deemed as an effective strategy for the
construction of knowledge. With regard to this, Vygostky (1978) suggests that teachers use cooperative learning
exercises where less competent learners benefit from more skillful peers. The current project believe in these
conceptions and makes them the groundwork of this research.
In brief, this presentation will be mainly targeted at sharing the instruments that have been adapted and the
results of this experience that confirm the validity and appropriateness of CSG in pronunciation courses that
promote the use of metacognitive strategies. Results, also, confirm that the inclusion of collaborative learning
tasks encourage students to express their opinions, discuss and negotiate ideas with others (Johnson & Johnson,
1989). Besides, data reveal that phonological awareness is enhanced and monitoring strategies are adopted
throughout the application of the LPS. On the other hand, pedagogical abilities are strengthened in the
process and, finally, the students' proficiency level is related to the error analysis capacity students develop in
the course of the project.
!
27
Safia Chibani
Pronunciation teaching in Algeria : From Stagnation to Progress
Acquiring a native or near native-like accent has always been considered as a redundant skill by both
Algerian teachers and learners due to the narrow space allowed for pronunciation teaching/learning in the
Algerian curricula. Indeed, according to article 49 of the Algerian educational system, communicability is more
important than accentability. This explains why many teachers and learners do not consider the Algerianess of
their English as an obstacle or an inferiority complex as long as they can make themselves understood.
However, with the recent implementation of the competency- based approach at all schooling levels
including university, we can finally say that this misconception of pronunciation redundancy is vanishing
gradually. This paper will be an attempt to assess the contribution of this newly implemented approach to
pronunciation teaching. For this, a systematic comparison between the old and new approach will be
established to detect strengths and weaknesses of both approaches with regard to teaching pronunciation.
Manuel Delicado Cantero – Australian National University and William Steed – James Cook University
Fair Dinkum: L2 Spanish in Australia by the book
While Spanish teaching in the United States is a well-established field, it is still maturing in Australia. Clear
evidence of this is the absence of materials aimed at the Australian English-speaking market. While for the most
part textbooks are fine, problems arise immediately when it comes to addressing L2 Spanish pronunciation in a
pronunciation course, as the models and examples are aimed at an audience with a different language
background. The result is that students report it as difficult to get the most benefit from the materials. Just like
phonetic pedagogy is best achieved using Australian-focused materials (e.g. Fromkin et al. 2012), the same
focus is necessary for effective L2 pronunciation materials.
This study aims to review four well-known primary textbooks for pronunciation courses (Guitart 2004; Stokes 2005;
Piñeros 2008; and Morgan 2010), all primarily and even explicitly written for American students – arguably the
largest market for Spanish pedagogical materials. Our review only measures appropriateness to the Australian
university classroom context not the quality of the texts, which we do not question here. In doing so, we divide
our approach into two main aspects:
1. Phonological and phonetic content (segmental and suprasegmental);
2. Level appropriateness to the linguistic awareness of Australian learners of Spanish.
As for the first aspect, based on our analysis of recordings of Australian Spanish learners and
perspectives from instructors, we have determined that there are specific features that are unique to the
Australian learner, for example r-deletion in codas and the phonetic realization of vowel phonemes –
diphthongs in particular. Given their American focus, it is not surprising that the aforementioned texts do not
cover these features. Furthermore, the examples provided for comparison are from US English, and pose
additional difficulties for typical Australian students.
With regard to the second aspect, according to our survey data, the majority of students studying
Spanish in Australian university classrooms do not have any linguistics training (either from high school or
university study). As a result, they cannot be expected to have metalinguistic awareness, and knowledge of
technical vocabulary – word classes, stress, phonetic and/or phonological transcription, etc. We find that the
base technical level of two of the textbooks is too high for the level at which Australian students will start a
pronunciation course. We have noticed that it is difficult to find a balance between the technical and the
practical aspects of the textbooks in question within the time frame that can be allotted to pronunciation in a
typical Spanish major in Australia.
This study forms part of a larger project analyzing the characteristics of Australian L2 Spanish
pronunciation, based on both surveys of students and teachers and acoustic-phonetic analysis of student data
at different levels of study. The textbook review serves as a benchmark for the ultimate goal of the project,
which is to create materials adequate for use in Australian classrooms.
Félix Desmeules-Trudel- University of Ottawa
VISC effects on the perception of Quebec French nasal vowels by Brazilian learners
Quebec French (QF) and Brazilian Portuguese (BP) both contain contrastive nasal vowels in their phonological
inventories. However, realizations of these segments are different in each language, and phonetic properties of
these realizations might lead a second language (L2) learner of QF to confuse and misidentify certain contrasts
28 !
in the target language. Vowel inherent spectral change (VISC) is one of the acoustic correlates of perceived
diphthongization in QF and BP and is hypothesized to be one of the factors leading to the misidentification of
QF nasal phonemes by Brazilian Portuguese native speakers. The results of an acoustic analysis of VISC in QF (9
participants, 540 productions) and BP (4 participants, 215 productions), and of discrimination and identification
tests (4 participants) showed that QF [oe!] is one of the nasal vowels with the greatest VISC, and also one of the
most misperceived (in terms of both discrimination and identification) vowel productions among the group of
Brazilian participants. QF [&] also displayed important variation with regards to its realizations concerning VISC,
but the positive transfer phenomenon is hypothesized to help BP learners of QF as a L2 to correctly discriminate
and identify this vowel because realizations of BP [&] also display important VISC (similar to QF [&]). Other
potential factors of influence, such as [oe!] lip rounding and the general influence of F3, are taken into account
and discussed with regards to transfer in the perception of QF nasal contrasts by L2 learners.
Nicholas Driscoll – Iowa State University
Hatsuon Help: a research-based, culturally-sensitive English pronunciation website for Japanese ELLs
Hatsuon Help is a new, research-based learning website being developed as a tool to help Japanese learners
with the /l/ and//'/ contrast. The /l/ vs /'/ consonant contrast has long been identified as a difficult one for
Japanese learners of English to both perceive and produce (Goto, 1971, Miyawaki, 1975). The English /l/ and /'/
both differ from the closest Japanese consonant, the apico-alveolar flap (Vance, 1987). Production is easier
than perception for most learners, but perception, when learned, causes a collateral learning of production
(Shimamune & Smith, 1995). To effectively train learners in a consonant contrast, multiple talkers should be
utilized (Lively, Logan & Pisoni, 1993). Videos of mouth movements also assist learners to understand and
replicate the articulations involved in pronunciation (Hardison, 2003). For /l/ and /'/, the prevocalic position is
the most difficult (Sheldon & Strange, 1982), especially before /u/ and /o/ (Hardison, 2003). Thus, the website
focuses on perception through identification and discrimination exercises, multiple speakers, immediate
corrective feedback, videos of mouth movements, and special attention given to prevocalic /l/ and /'/ before
/u/ and /o/ vowels. As a means to be sensitive to the cultural expectations and interest of the Japanese
people, Japan-friendly design aesthetics using cute "working characters" were incorporated into the website,
due to how widely such mascots are used throughout Japan for everything from advertising to instructional
booklets (Alt & Yoda, 2007). Along with the “working character” aesthetic, corrective feedback is delivered via
culture-sensitive sound-effects and visuals, and key concepts will be communicated through educational
comic-style storytelling.
Veronica Gonzalez-Lopez- Denison University and David Counselman- Ohio Wesleyan University
The production and perception of Spanish voiceless stops by novice learners: shedding light on early L2
category formation
This study offers evidence supporting the SLM (Flege 1995) postulate regarding perceptual distance among the
L1 and L2 phones. Many studies focus on the end state of L2 pronunciation in a variety of language pairings
(Baker et al. 2008; de Leeuw, Schmid and Mennen 2010; Flege 1995, 2002; Flege and Eefting 1987, 1988), but
more research on the emergence of L2 phonetic categories is needed (but se Arteaga 2000; González López
2012; Hurtado and Estrada 2010; Zampini 1998; Zampini and Green 2001). This experiment examines the /ptk/
production and perception of two groups of L1 English first-semester Spanish students to test two hypotheses: 1)
phonetics instruction positively affects L2 production accuracy, and 2) category formation will only be
observed for L2 phones that are perceptually different from L1 counterparts. The target group (N=18) received
explicit training in articulatory phonetics while the control group (N=16) did not. Production and perception
tasks were administered at the beginning and end of a 14- week semester. Each participant produced 30
counterbalanced /ptk/ tokens in English and Spanish and listened to 60 instances of /ptk/ (with varying degrees
of VOT) followed by a neutral vowel. Findings show significantly shorter VOTs for L2 Spanish voiceless stops than
L1 English. Both groups improved significantly their production of /k/ on the posttest. Contrary to the first
hypothesis, there were no significant differences between the two groups. However, linear regressions revealed
a significant correlation between the perception and production of Spanish /k/, confirming the second
hypothesis and offering evidence supporting the SLM. Overall, these findings suggest that category formation
may begin with initial exposure as long as L2 sounds are perceived as different from the L1 phones, but the
effects of instruction are not noticeable in the first semester.
!
29
Shannon Halicki – West Liberty University
Back Door Phonetic Conditioning: Accent Therapy in Early French Pronunciation Training
This paper reports on an instructed language study in that compares the outcomes of two distinctive methods
of phonetics training in elementary French classes. For American adults learning French, it is typically
challenging to pronounce French /R/ (a uvular liquid) and initial voiceless stops (/p/, /t/ & /k/). In the case of
French /R/, learners must activate a new, uvular place of articulation, and French initial voiceless stops pose a
challenge because of reduced VOT compared to their English counterparts. While the substitution of English /r/
and stops may not seriously impede comprehensibility, it does contribute to robustly accented speech. Twentysix L1 English learners of French participated in the current study and were randomly assigned to the
experimental group or the control group. Participants in the control group received fifteen minutes of
repetition-based pronunciation instruction at the end of each Elementary French class for two weeks (a total of
eight sessions). Participants in the experimental group received “accent therapy,” a training technique in
which they were instructed to enunciate carefully written English sentences with a French accent. While the
experimental group also received fifteen minutes of pronunciation training per session, only six minutes of that
time was spent performing repetition, with the other eight minutes being spent practicing the accent therapy
technique. Subjects performed a recorded reading task one week after the completion of the sessions. Three
professors of French who were naïve to the experiment rated the overall accent for each recording as well as
the “correctness” of the production of French /R/ and voiceless stops. Results suggest that the accent therapy
technique had a measurable and significant effect on overall accent as well as specific production of French
/R/ and voiceless stops.
Yizhou Lan – City University of Hong Kong
Detecting L2 speech deviations by a communicative experiment procedure: taking Cantonese speakers’
realizations of English [r] as an example
It is argued that the purpose of marrying L2 speech research and L2 teaching and learning is to help improving
learners and listeners’ intelligibility and facilitate smoother communication (Munro and Dewing, 2003). However,
the reliability and validity of experiments of L1 and L2 speech production has long been accredited to the
convenience of controlled careful reading procedure (e.g., Flege et al 1999; Pepperkamp and Bouchon,
2011). In a communicative context, such a paradigm faces a challenge. In controlling other elements of
linguistic processing, we might have been isolating pronunciation as an independent modular in language
processing. Based on a parallel-distributed processing model of language (McClelland and Elman, 1986), it is
questionable whether the reading paradigm can effectively examine how L2 sounds are acquired in
communication. This study aims to explore other experiment procedures, especially in a stage of finding out
real error patterns in communication.
In order to compare reading and self-generated speech in their potential to expose L2 speech deviations, we
examined whether native English speakers’ ratings of ESL learners’ production were affected by experiment
procedure. Participants are 6 advanced Cantonese speakers of English with 20 years of exposure to English.
Stimuli are real r- or -r- words, each in five vowel contexts (e.g., root, rob, cry) with equal number of fillers. In
experiment 1, participants read the stimuli carefully in a carrier sentence; in experiment 2, same participants
were told to make up a story out of the same stimuli used in experiment 1. Results showed that in experiment 2,
more numbers and types of errors were found by native English speakers than in experiment 1. Results imply that
though both experiments uses the same controlled stimuli, gearing the procedure to a more communicative
one will help find more types of errors and helps learners greater in later training.
Stephanie Link, Sinem Sonsaat, and John Levis – Iowa State University
Confidence in teaching pronunciation: How native and nonnative teachers negotiate the pronunciation
classroom
Teachers’ actual classroom practices and decision-making processes are affected by various factors such as
their knowledge of language theories, subject matter knowledge, beliefs, experiences, practices (Richards,
1998), assumptions about language (Woods, 1996), cognition (Baker & Murphy, 2011), confidence levels
(Murdoch, 1994), and identities (Kumaravadivelu, 2012). When it comes to teaching pronunciation, which is
thought to be more challenging to teach than the other skills (Macdonald, 2002), some of these factors, such
as native-speaker status, identity and confidence, may especially shape the confidence and classroom
30 !
practices of teachers based on their reaction to uncertainty (Campbell, 2007). Murdoch (1994) points out that
“a teacher’s confidence is most dependent on his or her own degree of language competence” (p. 258), or
with pronunciation, a teacher’s perception of language competence. Identity, especially represented by
status as native or nonnative English-speaking teachers, might also affect teachers’ confidence in different
ways. In turn, these factors, may lead to different classroom behaviors even if the teachers are using the same
materials.
Any teacher can face uncertainty when teaching pronunciation, but only by addressing the sources of their
uncertainty can teachers be confident. In this paper, we focus on the differences of classroom practices,
decision-making processes and the use of pronunciation teaching materials by two teachers of English, one a
native speaker and one a nonnative speaker of English. The teachers taught pronunciation skills, mainly through
suprasegmental features, to two intact classes of graduate students at a Midwestern University. The teachers,
both PhD students in Applied Linguistics, were similar to each other in their knowledge of language theories,
subject matter knowledge, teaching experiences and personal characteristics. Written and verbal reflections of
the teachers’ teaching practices showed how the teachers prepared to teach, reacted to errors in materials,
and adjusted their classroom practices in different ways. Suggestions are made for the kinds of teacher
preparation necessary for successful NS and NNS teacher preparation in teaching pronunciation.
William McCartan – Seton Hall University
Word stress diagnostic procedure shared through a wiki site
The complex Standard American English (SAE) vowel system is rendered more complex by a complicated
system of word stress patterns. Word stress influences vowel length, vowel reduction, and the rhythm or music
of SAE; consequently, word-stress errors can render words or phrases difficult or impossible to understand.
Since advanced learners sometimes know thousands of words, correcting individual word stress errors as they
occur is an endless task; therefore, a diagnostic procedure is needed for identifying categories of errors. I
believe the most efficient approach is to categorize errors according to underlying stress patterns which
requires an inventory of stress patterns.
To meet this need I built a 7,500 record FileMaker Pro (FMP) database. Each record has a field for a word,
number of syllables, stress pattern, part of speech, and level of difficulty. The database enabled me to
formulate a comprehensive inventory of stress patterns, and generate word lists representative of specific stress
patterns. The diagnostic materials consist of short lists of words representative of stress patterns and stress
assignment rules. By reading the lists of words students reveal the stress patterns they use correctly and
incorrectly. Information acquired from the diagnostic procedure guides the development of instructional
materials to meet each student’s needs.
To share the approach, I established a pbwiki site which makes available the stress pattern inventory,
diagnostic and instructional materials, including word lists, and the evaluation procedure.
The presentation will consist of a review of relevant background information, brief demonstration of materials,
procedures and audio clips, and a description of the pbwikisite. Background information, materials, and
opportunities for further discussion will be available through the pbwiki site.
Charles Nagle - Georgetown University
Acquisition of the voicing contrast in L2 Spanish
Since Lisker and Abramson’s (1964) study on voice onset time, research has shown that learners’ gradually
acquire the phonetic properties of target language stops (e.g. Gass, 1984; Kissling, 2012; Zampini, 1998). Yet, we
still know relatively little about individual variation in rate of acquisition and production precision.
The current study focuses on acquisition of the voicing contrast in L2 Spanish using the bilabial stop consonants
/p/ and /b/ as a test case. Spanish presents an added layer of complexity given that voiced stops weaken to
fricatives or approximants in all environments except after a homorganic consonant. Research has shown this
to be a challenging alternation even for advanced learners (Face & Menke, 2009; Shively, 2008; Zampini, 1994)
and one that may prove resistant to focused instruction (Kissling, 2012). By examining the stop consonants, we
may monitor the acquisition of both phonemic and allophonic contrasts in a foreign language and how
internal phonetic factors such as stress or position within the word and external factors such as motivation and
relationship to the L1 bear on phonological SLA. 26 English-speaking participants who were beginning learners
of Spanish completed both perception and production tasks as well as background and motivation
questionnaires. The researcher created a set of fictitious characters to feature as target items by crossing lexical
!
31
stress and position within the word. Both tasks were highly contextualized and required learners to process
sentential elements beyond the minimal pair characters thereby simulating more realistic conditions. Data was
collected in two waves. Analyses indicate that learners are highly accurate at discriminating the stops even at
Time 1. On the other hand, their production shows considerable variation in voice onset time and lenition of the
voiced stops in required contexts. This study also explores the role motivation plays in phonological SLA.
Daniel Olson and Heather Offerman- Purdue University
The Effects of Visual Feedback on Learner Pronunciation: Speech analysis software in the L2 classroom
There has recently been renewed interest in the previously “marginalized” area of SLA pronunciation (Derwing
& Munro, 2005) and the concept of Electronic Visual Feedback (e.g. de Bot, 1983), driven by advances in
speech technology. Although speech analysis software is common in phonetics courses (Lord, 2005), practical
L2 classroom implementation is lacking (Levis, 2007). Moreover, some have claimed that speech software may
be too complicated for successful classroom implementation (e.g. Setter & Jenkins, 2005). Building on work
demonstrating positive student response to visual analysis (Olson, forthcoming), this project seeks to quantify
pronunciation gains in a visual feedback paradigm in the lower-level L2 classroom. Specifically two questions
are addressed: (1) Does guided visual analysis improve student pronunciation? and (2) Are benefits
generalizable to similar phonetic contexts?
A guided-inductive approach (Mayer, 2004) was used to incorporate PRAAT speech analysis software
(Boersma & Weenink, 2011) into a 3rd semester Spanish-language course. L2 learners of Spanish (n=24)
completed three pronunciation activities during the semester. Each activity focused on an intervocalic voiced
stop (V_V context: English [b, d, g]; Spanish [(, !, )]). Students recorded lexical items and utterances, visually
analyzed and compared their target phoneme productions with native speaker productions, and re-recorded.
Although each activity focused on one phoneme, all three phonemes were recorded during each activity to
allow for analysis of local (i.e. each lesson’s target phonemes) and global (i.e. non-target phonemes)
performance. Analysis focused on the intensity ratio (preceding vowel to target consonant) and closure
duration (Colantoni & Marinescu, 2010; Hualde, 2010).
Preliminary analysis reveals mixed results. Students demonstrated no local improvement for /d/ (activity 1), but
significant improvement for /g/ (activity 2) (Figure 1). Additionally a trend towards improvement of non-target
phonemes emerged, specifically during the second activity. Discussion focuses on the utility of speech analysis
technology and individual differences.
Lisa Pierce – University of Illinois
Multi-methodological, cross-disciplinary approaches to pronunciation teaching
Every theory of Second Language Acquisition (SLA) asserts the importance of input (language) to provide
evidence for the syntactic, phonological and morphological structures of the target language. Educators in
language instruction assert the necessity of - not just input - but comprehensible input in meaningful contexts
with some concomitant amount of attention paid. However, there is a surfeit of evidence that input has little
impact on the acquisition of the adult learners’ L2 phonology which is reflected in the persistence of accentedspeech. Indeed, there is a burgeoning area of research that suggests that the L1 phonological system is not
“passively failing” to identify novel forms but is so robust that “perceptual illusions” result from active repairs to illformed input. In addition, there is evidence to suggest that cognitive processing of language for meaning is
different from processing for acoustic details] – so the notion that interacting with language in a meaningful
context is sufficient for phonological or pronunciation learning is dubious.
When best practices fail to produce measurable learning, it is time to reassess what informs our teaching
practices. This work considers ways in which experimental methods and findings from Laboratory Phonology,
Neuroscience, Psychology, and Speech and Hearing Science can guide new approaches to pronunciation
training while providing concrete examples of methods that can be repurposed for the classroom.].
Shinichi Shoji – University of South Carolina
Japanese epenthetic vowels: How Japanese speakers pronounce English words
The primary objective of this study is to examine which epenthetic vowels Japanese speakers utilize in
pronouncing English words and how they use them in the framework of optimality theory (OT). While most of the
early studies are limited to explaining partial phenomena, this study aims to offer a unified explanation with a
32 !
single ranking of constraints. Analysis of Japanese loanwords from English and two experiments with native
Japanese-speaking participants were conducted to support the proposed ranking.
Early studies argue that Japanese speakers epenthesize [*] as the default vowel due to its
unmarkedness ([-round, +back, +high]) and its low perceptual salience. Also, [o] and [i] are epenthesized
following certain consonants. The current study proposes that these epentheses are explained with the
constraints in the ranking of: COMPLEX, CODACOND >> MAX-IO >> PALATAL-FRONT, SYLLABLEINVENTORYSTRUCTURE (SIS),
IDENT-IO >> *ROUND, *LOW >> *FRONT >> HIGH >> BACK >> DEP-IO. In the first experiment, 94.7% of participants'
responses followed the constraints above especially in the adaptation of coda-consonants. Also, regressive
vowel-harmony was detected in the cluster-condition.
I also examine some Japanese loanwords exhibiting consonant-deletion instead of vowel-epenthesis,
which violates the ranking of MAX-IO >> DEP-IO. I attribute this to perceptional phenomenon; Japanese
speakers delete consonants when they do not perceive them, especially when the input is aural. This analysis
claims that different patterns of perception filter the input for tableaux. The second experiment confirmed that,
although most participants epenthesized vowels to adapt consonants following the constraints above, they
deleted more aurally input consonants (9.1%) than orthographically input ones (2.1%).
Overall, the dominant adaptive strategy of Japanese speakers in pronouncing English words is vowelepenthesis, and the vowel is selected based on the constraints above. Meanwhile, deletion of coda or
consonants in clusters is another prominent strategy when the consonants involved lack perceptibility among
Japanese speakers.
Karen Taylor de Caballero- Santa Fe Community College and Shirley Thompson- English Language Training
Solutions LLC
Coloring Pronunciation Across the Curriculum with The Color Vowel Chart
The teaching of pronunciation is often marginalized in language instruction. Reasons stem from a number of
misperceptions among teachers: that there is no time to “fit” pronunciation into the curriculum, that it is
culturally insensitive to change learners’ accents, that pronunciation instruction is futile, that pronunciation is
too hard to teach and too boring to learn. These beliefs point to a lack of effective teacher education: if
teachers themselves are challenged by phonology, how can we expect them to teach it? More importantly,
unless we model pronunciation across the teacher education curriculum, how can we expect teachers to
integrate pronunciation into their own teaching?
The presenters have developed a visual-kinesthetic approach to teaching pronunciation known as The Color
Vowel Chart. As a visual aid, the Chart features a unique color and key word phrase for each of the English
vowel sounds. As an instructional approach, the Chart enables teachers and learners at all levels and across
skill areas to identify word- and phrase-level stress, thus teaching intelligibility from the first class.
This presentation will feature three case studies of how The Color Vowel Chart is being used to integrate
pronunciation into language program curricula. The first will highlight a university TESOL program in Washington
DC that uses the Chart in multiple courses to help pre-service teachers form a pedagogical bridge between
phonology and the teaching of pronunciation. The second case study, of a university-based intensive English
program in Oregon, will demonstrate how The Color Vowel Chart has served as a curricular thread across all
skills-based courses. The third will demonstrate how a volunteer-taught Adult ESL program in Washington DC
uses the Chart to train volunteers and teach low-literacy English learners to focus on what we characterize as
the “deal breaker” of spoken English: word and phrase stress.
Hyejin Yang – Iowa State University
Investigating needs of stakeholders of an oral proficiency test for ITAs to bridge the gaps between ITAs’ needs
and raters’ feedback.
Oral English Certification Test (OECT) at Iowa State University is a test that evaluates international graduate
students (ITAs)’ speaking proficiency in order to determine whether they are capable of performing
instructional duties as teaching assistants. Students who fail in the test are not able to receive teaching
assistantships and subsequently enroll in English 180 courses for improving their pronunciation and speaking skill.
Despite its’ huge influence of the test on ITAs (Saif, 2006), however, less attention has relatively been drawn to
examine nature of raters’ feedback on ITA’s performance and whether recipients of feedback, such as ITAs,
180 instructors, and departments, could benefit from feedback.
!
33
This study aims to investigate gaps between raters’ feedback and needs of diverse stakeholders,
including ITAs, the OECT raters, and English 180 instructors, of the test by exploring the OECT raters’ actual
feedback on pronunciation and speaking skill and the other stakeholders’ needs. Instruments employed in this
mixed-methods study included (a) two sets of focus group interviews with five OECT raters, (b) one focus group
interviews with two instructors of speaking classes for ITAs, and (c) one questionnaire with twenty two
international graduate students who were in preparation for the test. The quantitative data was analyzed using
descriptive and inferential statistics. Participants’ responses on the interviews and the questionnaires were
transcribed and analyzed thoroughly.
Result of the study revealed that the stakeholders agreed on huge impact of raters’ feedback to
enhancement of students’ speaking proficiency and even to English 180 class instruction. As for the
stakeholders’ views on quality of raters’ feedback, it was found that feedback itself did not clearly deliver
strengths and weaknesses of ITAs’ speaking performance on the test.
This poster presentation will report on more detailed- findings of this study and suggestions for future
improvement of raters’ feedback in order for ITAs to receive more individualized feedback to succeed in the
OECT.
Elisabeth Zetterholm – Linnaeus University Sweden
Final stops or not? The importance of final consonants for an intelligible accent.
Many of the Karen people of Myanmar live in refugee camps in Thailand. The UN helps some of them move to
other countries and families are spread all over the world. Despite difficulties, a chance at integration into the
new society depends upon learning the society’s language. After eight years in … (the name of the country)
most of them still are not able to pronounce words with an intelligible accent for a native speaker. One reason
is the phonetic and phonological differences between the two languages. Karen is a SinoTibetan language
and the two main dialects are Pwo and Sgaw Karen. They are described as having a three--‐way contrast for
plosives, namely unvoiced aspirated plosives [p+ t+ k+], unvoiced unaspirated plosives [p t k] and voiced
unaspirated plosives [b d] in word--‐initial and word--‐medial positions. As other SinoTibetan languages Karen
has a monosyllabic structure and final stops are not pronounced. This causes problems when learning e.g. a
Germanic language where final stops sometimes are of importance for intelligible speech. A training program is
constructed with sentences and minimal pairs of words ending with final stops. The learners will listen to native
speakers; imitate and record themselves, in order to be aware of their own pronunciation, and to get closer to
the native speakers’ pronunciation. Recordings of the learners, before and after six weeks of training will be
closely analyzed to get an opinion about the utility of this method.
Elisabeth Zetterholm – Linnaeus University Sweden and Mechtild Tronnier – Lund University
Different stress patterns meet: Kurdish L1‐speakers learn Swedish
The Kurdish language is an Indo--‐European language spoken in Turkey, Iran, Iraq and Syria. Based on
migration, many speakers of Kurdish now live in other countries all over the world and it is among the ten most
frequent languages in Swedish as a second language courses, as of today. Kurmanju, or the Northern Kurdish,
and Sorani, or the Central Kurdish, are the two main dialects. As an L2--‐learner it is important to be aware of
differences in phonology and prosody between L1 and L2, as problems in direct communication may occur
otherwise. In Swedish, a stress--‐timed language, a stressed syllable in a word is perceived by the means of
intensity, pitch and length. In Kurdish, a syllable--‐timed language, stress is not correlated with length variation.
Most of the words have their stress on the last syllable, but nominal compounds usually have only one stressed
syllable on the last element in the compound. In Swedish, on the other hand, the primary stress falls mainly on
the first element in a compound and a secondary stress occurs on the second element. Our investigation of
Kurdish L1 speakers learning Swedish indicates potential problems with stress placement and this is not only the
case for compounds, but also non--‐compounded words, no matter which Kurdish dialect the speakers
represent. This causes problems in communication although the meaning of the word does not change. But L2-‐speakers are sometimes hard to understand as the stressed syllable serves as a perceptual anchor. Thus, stress
misplacement may cause a delay in the communication flow. Recordings of six Kurdish speakers have been
analyzed when reading sentences in L2. It is obvious that the stress placement and pattern based on the
adequate combination of pitch and length is hard to master.
34 !
Call for 2013 Proceedings
Dear Presenters from the 2013 PSLLT conference,
We'd like to encourage you to consider writing up your paper or poster for the 2013 PSLLT
proceedings. Your paper will be due December 1, 2013.
The proceedings are an important part of the conference and are essential in making your
work known to a wider audience. There is nothing like the proceedings anywhere else in
pronunciation teaching. The publication of high-quality, interesting papers helps the field of
second language pronunciation to move forward.
Guidelines for proceedings submissions for 2013
All presenters of papers and posters are invited to submit a written version of their
paper for consideration in the electronic conference proceedings. All
submissions will be reviewed by at least two readers who will make suggestions
and recommendations to the authors and the editors.
Manuscripts should:
•
•
•
•
•
•
•
follow the Publication Manual of the American Psychological
Association, 6th ed. (2001) especially for tables, charts, and graphs
be doubled spaced and include no more than 3000 words
(excluding references, tables, notes, appendices etc.)
use Times New Roman, 12 point font
include an abstract of no more than 200 words
include a biographical statement of the author(s) not to exceed
120 words per author
include the contact information of the author(s): name,
affiliation, address, telephone number, and email address
All proceedings contributions are due by December 1, 2013. Any
questions should be directed to John Levis at
pslltconference@gmail.com.
!
!
35
CALL FOR PAPERS
Sixth Annual Conference
Pronunciation in Second Language Learning and Teaching
University of California, Santa Barbara
Santa Barbara, California USA
Looking at L2 Pronunciation Research from Varying Perspectives
September 5-6, 2014
Plenary Speaker
Alene Moyer, University of Maryland
Pronunciation instruction is increasingly popular in language classrooms around the world, in
second language and foreign language contexts. Issues of intelligibility (Munro & Derwing, 1995)
vs. nativeness (Levis, 2005), functional load (Brown, 1991; Munro & Derwing, 2006), effective
instructional techniques for overcoming learning plateaus (Acton, 1986; Hardison, 2004;
Goodwin, 2006), fluency (Derwing et al., 2008) and the relative roles of suprasegmentals and
segmentals in instruction (Hahn, 2004) have all been examined in multiple studies. However, a
large majority of important research into pronunciation has been carried out with English as the
target language, despite the importance of L2 pronunciation in other languages, such as
Japanese (e.g., Hirata, 2004), Spanish (e.g., Lord, 2008), French (Ruellot, 2006), German (Moyer,
1999), Chinese (Liu et al, 2000), and Dutch (Bongaerts, Mennen & Slik, 2000), among others.
Research from a wide variety of L2 learning contexts is essential to filling out the current Englishcentric research agenda.
The 6th Pronunciation in Second Language Learning and Teaching Conference invites proposals
for papers and posters on all topics related to naturalistic and classroom pronunciation
acquisition and learning. We especially welcome proposals for papers on pronunciation in a
wide variety of L2s other than English. Possible paper topics include descriptive and experimental
studies, re-examinations of key research findings (e.g., intelligibility, comprehensibility and
accentedness) in new languages, technology in the teaching of pronunciation, and innovative
approaches to teacher education.
In addition to papers related to the place of pronunciation in L2s other than English, the
conference invites proposals for papers or posters on any aspect of pronunciation research,
teaching and learning. Papers will be given in English.
Submission of paper abstracts will be possible beginning October 1, 2013 at
http://linguistlist.org/easyabs/PSLLT2014. Submissions are due by April 12th, 2014. Notification of
acceptance or rejection will be sent out by May 15th, 2014.
For further information about the conference, contact Dorothy Chun, Conference Organizer at
pslltconference@gmail.com.
36 !
Getting Online
Wireless Internet Access on Campus
Iowa State University has a service that allows campus visitors access to the Internet via the
campus network. This service allows visitors to browse the web and check email.
When a visitor opens a web browser (Internet Explorer, Firefox, etc.) a website will appear
asking the visitor to indicate their affiliation with Iowa State University (if any), their name,
email address, and phone number. When the visitor has completed the registration process
and re-booted their computer they will be able to access the Internet.
Initial guest access is either 7 days or the length of the event (if selected from the pull down
menu). After expiration, visitors will not be able to register for guest access for 20 days.
To set up guest access, follow these steps:
1. Open a web browser.
2. If the ISU NetReg page does not automatically appear, enter “netreg.iastate.edu” into
the address bar and press Enter.
3. Select the specific event you are attending if it is listed and select “Register Here” or
select “Guest Registration Here”.
4. Enter your local contact information and select “Next”
5. Read the terms and select “I Agree”
6. Reboot your computer when instructed to do so (approximately a one-minute wait) to
complete your registration process.
Questions regarding registrations for guest access may be directed to the IT Services Solution
Center at 515-294-4000.
!
37
Map and Directions to Conference Dinner
St. Johns by the Campus is close to the Memorial Union, easily within walking distance (0.2
miles). From the Memorial Union, head West on Lincoln Way, you will see St. John’s on your
left.
38 !
Map of ISU Campus
!
39
Ames City Information
Transportation
There are two local taxi companies in Ames:
Ames Taxi: (515) 232-1343
Nighthawk Cab: (515) 203-0423
The city bus system for Ames is Cyride (www.cyride.com). The regular fare for an adult is
$1.25.
What to do in Ames
Octagon Art Festival
At the Octagon Art Festival more than 100 artists from all over the Midwest sell high-quality
original artwork. The festival is held outdoors on Main Street (in the Cultural District) rain or
shine. This year the art festival will be held on Sunday, September 22nd from 10am-5pm. For
more information visit http://www.octagonarts.org/en/octagon_art_festival/
ISU Carillon Festival and Competition
The festival includes a carillon concert by Geert D’hollander from Lake Wales, a carillon
competition for young composers, a master class, a lecture on writing carillon music, and
campanile tours. The festival will be held at the ISU Carillon on Saturday, September 21st from
10am-4pm. For more information visit http://www.music.iastate.edu/carillon
Reiman Gardens and Butterfly House
The campus gardens are located south of central campus on University Boulevard. There is
also a Butterfly Garden that is a big attraction for many visitors. For more information, go to
http://www.reimangardens.iastate.edu
Ada Hayden Heritage Park
A longer, pleasant walk can be had at Ada Hayden Heritage Park (no charge), about 5
miles north of the Memorial Union. Directions and information can be found at
http://cityofames.org/parksrecweb/AdaHaydenHeritagePark.htm
40 !
Ames Restaurants
Close to Campus
Legends American Grill - 19 Stanton Ave Ste 701
www.legendsamericangrill.com
Thai Kitchen – 2410 Chamberlain
www.thaikitchenames.com
Battle’s Barbecue – 218 Welch Ave
Es Tas (Mexican) – 120 Welch Ave
Pita Pit – 114 Welch Ave
The Fighting Burrito – 117 Welch Ave
Café Baudelaire (Brazilian Drinks and Food) 2504 Lincoln Way
Stomping Grounds (Coffee and Food) – 303 Welch Ave, Suite 101
The Scallion (Korean) – 118 Hayward Ave
India Palace - 120 Hayward Ave
Off Campus
Olde Main Brewing Company - 316 Main St
www.oldemainbrewing.com
The Spice (Thai) - 402 Main St
www.thespiceames.com
Aunt Maude’s – 547 Main St.
www.auntmaudesames.com
Hickory Park (a true Ames experience) - 1404 S Duff Ave
www.hickoryparkames.com
The Café - 2616 Northridge Pkwy
http://thecafeames.com
The Mandarin (Chinese) - 415 Lincoln Way
www.mandarinofames.com
Café Diem (coffee house) – 323 Main Street
www.Café-diem.com
London Underground (Pub-Large selection of beers) – 212 Main St.
!
41