E-J069-084_Banzina_Hewitt_Dillit EuroAmerican  Journal  of  Applied  Linguistics  and  Languages   E-JournALL  Volume  1,  Issue  1,  November  2014,  pages  69-84   ISSN  2376-905X     DOI  http://dx.doi.org/10.21283/2376905X.1.9 www.e-journall.org     Copyright  ©  2014.  This  work  is  licensed  under  the  Creative  Commons  Attribution-­‐NonCommercial-­‐NoDerivs  3.0  Unported  License. 1  Contact:  elina.banzina@sseriga.edu   Using  synchronous  speech  to  facilitate  acquisition  of  English  rhythm:    A  small-­‐scale  study   ELINA  BANZINAa1, LYNNE  E.  HEWITTb, LAURA  C.  DILLEYc aStockholm  School  of  Economics  in  Riga   bBowling  Green  State  University   cMichigan  State  University   Received  24  January  2014;  received  in  revised  form  5  June  2014;  accepted  17  July  2014   ABSTRACT EN While appropriate stress and rhythm is of importance for any speaker’s intelligibility, such properties are critical for international teaching assistants (ITA), who deliver new and complex information to native speaker audiences. Given the limited time available for ITA instruction and the need for a time-efficient rhythm teaching method, this article reports findings of a small-scale feasibility study that tested the effectiveness of a synchronous speech component introduced into conventional rhythm instruction. Synchronous speech involves teacher and learner speaking in unison continuously, which allows the L2 learner to learn rhythm implicitly and uninterruptedly, and provides rich auditory-visual input, ample motor speech practice and real-time feedback, thereby automatizing rhythm patterns. In a 6-week-long pre-post experimental feasibility study, blind listeners evaluated pre- training and post-training recordings of ITA-produced speech. Data revealed a trend towards more improvement in L2 rhythm working with the synchronous speech technique. Results establish feasibility in both instruction and research. Key words:  SPOKEN ENGLISH, PRONUNCIATION TEACHING, ENGLISH RHYTHM, SYNCHRONOUS SPEECH, RHYTHM INSTRUCTION. ES Si un acento y un ritmo apropiados son fundamentales para la comprensión de cada hablante, estas propiedades son indispensables para los asistentes internacionales de la enseñanza (ITA), quienes proporcionan informaciones nuevas y complejas a un público de hablantes nativos. Dado el poco tiempo del que disponen los ITA para su formación y la necesidad de un método de enseñanza eficaz, este artículo transmite los resultados de un estudio realizado a pequeña escala, en el que se ha comprobado la eficacia de la introducción del discurso síncrono como complemento del tradicional método de enseñanza del ritmo. Un discurso síncrono implica que profesores y estudiantes hablen al unísono de forma continua, lo que permite al estudiante de L2 aprender el ritmo de manera implícita e ininterrumpida gracias a un fuerte input auditivo y visual, un incremento de la práctica motora y una retroalimentación inmediata, que conducen a la automatización de los esquemas rítmicos. Para este estudio, unos auditores anónimos evaluaron las grabaciones de discursos de los ITA antes y después del entrenamiento. Los resultados muestran una notable mejora del ritmo en L2 gracias a la técnica del discurso síncrono y su factibilidad en la enseñanza y en la investigación. Palabras clave: INGLÉS HABLADO, ENSEÑANZA Y PRONUNCIACIÓN, RITMO INGLÉS, DISCURSO SÍNCRONO, ENSEÑANZA RITMO. IT Se accento e ritmo appropriati sono fondamentali per la comprensione di ogni parlante, tali proprietà sono indispensabili per gli assistenti internazionali all’insegnamento (ITA), che devono fornire informazioni nuove e complesse a parlanti nativi. Dato il poco tempo a disposizione per la formazione degli ITA e la conseguente necessità di un metodo d’insegnamento efficiente, questo articolo riporta le conclusioni di uno studio su piccola scala con cui si è testata l’efficacia dell’introduzione del parlato sincrono all’interno del tradizionale metodo di insegnamento del ritmo. Il parlato sincrono prevede che insegnanti e studenti parlino all’unisono e senza interruzioni, permettendo agli studenti di L2 di apprendere il ritmo in modo implicito e continuo grazie al ricco input visivo e auditivo, all’incremento della pratica motoria e al feedback immediato, consentendo di automatizzare gli schemi ritmici. Nel corso dello studio, durato sei settimane, ascoltatori esterni hanno valutato registrazioni di parlato degli ITA prima e dopo il training. I risultati mostrano un considerevole miglioramento del ritmo nella L2 grazie alla tecnica del discorso sincrono e dimostrano la sua fattibilità sia nell’insegnamento sia nella ricerca. Parole-chiave:  INGLESE PARLATO, INSEGNAMENTO PRONUNCIA, RITMO INGLESE, DISCORSO SINCRONO, INSEGNAMENTO RITMO. BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 70 1.  Introduction   Appropriate   stress   and   rhythm   patterns   have   been   shown   to   be   critical   for   non-­‐native   speakers’   intelligibility   in  numerous  research  studies  (Anderson-­‐Hsieh,   Johnson,  &  Koehler,  1992;  Derwing,  Munro,  &   Wiebe,   1998;   Tajima,   Port,   &   Dalby,   1997).   This   is   especially   true   in   extended   discourse,   such   as   giving   a   presentation   or   teaching   a   class,   which   is   overall   a   cognitively   demanding   context   (Hahn,   2004).   Use   of   natural  stress  and  rhythm  patterns  is  also  potentially  important  under  degraded  listening  conditions,  such  as   in   large   lecture  halls   and  with  background  noise   in   the   classroom   (Mattys,  Brooks,  &  Cooke,   2009;  Munro,   1998;   Woodford,   Prichard,   &   Jones,   1998).   While   this   applies   to   any   non-­‐native   speaker,   the   need   for   producing  appropriate  speech  patterns  is  especially  critical  for  international  teaching  assistants  (ITAs),  who   often  work  with   relatively   large   American   undergraduate   student   audiences   in   an   acoustically   challenging   environment.  However,   international  graduate  students  often  struggle  with  rhythmic  performance.  There  is   evidence   in   second   language   research   that   non-­‐native   patterns   of   rhythm   and   stress   are   pervasive   in   the   speech  of   second   language   learners   (Anderson-­‐Hsieh  et  al.,  1992;  Flege  &  Bohn,  1989;  MacAllister,  Flege  &   Piske,  2002;  White  &  Mattys,  2007).  These  non-­‐native  rhythmic  patterns  may  result   in  seriously  decreased   intelligibility,  or  at  least  reduce  the  speed  with  which  the  message  is  processed  by  native  listeners  (Anderson-­‐ Hsieh  &  Koehler,   1988;   Braun,   Lemhofer,   &  Mani,   2011;   Field,   2005;   Hahn,   2004).   Unfortunately,   time   for   English   rhythm   training   for   international   teaching   assistants   is   often   limited,   necessitating   a   search   for   an   effective  approach  that  would  lead  to  automaticity  in  a  short  amount  of  time,  maximizing  the  amount  of  both   input  and  active  practice.     The  synchronous  speech  technique,  where  an  instructor  and  a  student  engage  in  reading  of  the  same   text   or   reciting   it   aloud   simultaneously,   was   observed   in   our   practice   to   yield   improvement   in   rhythmic   performance  in  familiar  and  practiced  contexts.  More  importantly,  it  led  to  a  considerable  improvement  also   in   spontaneous   rhythm   use   in   cognitively   demanding   contexts.   The   perceived   benefits   of   the   synchronous   speech   technique   for   second   language   speech   learning   are   the   simultaneously   provided   (i)   rich   auditory-­‐ visual  input,  (ii)  real-­‐time  implicit  feedback,  (iii)  ample  motor  speech  practice,  and  (iv)  uninterrupted  flow.  In   fact,  simultaneity  is  one  of  the  key  factors  that  allows  multiple  goals  to  be  accomplished  in  a  short  amount  of   time,  potentially  leading  to  a  long-­‐lasting  and  consistent  use  of  English  rhythm  and  stress.   Despite   the   observed   benefits   of   this   method,   no   empirical   evidence   exists   regarding   the   effectiveness  of  the  synchronous  speech  technique  for  language  rhythm  acquisition.  We  therefore  attempted   to  establish   its  applicability   to   rhythm   instruction  experimentally,   in  a  6-­‐week-­‐long  small-­‐scale  experiment   where   synchronous   speech  was   incorporated   into  an  existing   conventional  pronunciation   instruction   in  an   oral   proficiency   program   at   a   communication   disorders   department   of   a   Midwestern   U.S.   university.   The   synchronous  speech  technique,  as  it  was  used  in  this  study,  was  geared  towards  the  needs  of  the  particular   audience—international   graduate   students   (ITAs)—who   had   to   prepare   for   giving   oral   presentations   and   lecturing  in  front  of  undergraduate  students.  In  our  6-­‐week  experiment,  students  in  the  experimental  group   were  not  taught  exclusively  with  the  synchronous  speech  technique  but  rather  a  synchronous  speech  element   was   added   to   the   conventional   instruction   so   that   students   would   receive   explicit   teaching   as   well.   It   is   generally   agreed   in   second   language   research   that   adult   learners   benefit   both   from   implicit   and   explicit   instruction   (Ellis,   2005,   2007;   Hulstijn,   2005).   While   synchronous   speech   allows   rhythm   to   be   learned   implicitly,   having  metalinguistic   knowledge   of   the   English   language   rhythm   is   beneficial   in   order   to   notice   rhythmical  differences  more  easily  and  be  able  to  respond  to  corrective  feedback  more  effectively.  Moreover,   the  experimental  group  was  an  existing  class  and  we  did  not  want  to  employ  a  new  and  previously  untested   instructional   technique  as  our   sole  approach.  We   therefore  added  a   synchronous  speech  component   to   the   traditional  explicit  rhythm  instruction.  This  allowed  for  a  quasi-­‐experimental  design,  examining  whether  the   addition  of  this  component  resulted  in  better  performance  in  that  group.     The   current   study   sought   primarily   to   answer   whether   it   is   experimentally   feasible   to   test   the   effectiveness   of   the   synchronous   speech   technique   applied   in   pronunciation   teaching   and   what   measures   might   be   appropriate   to   establish   its   effectiveness.   The   other   main   goal   of   this   study   was   to   determine   whether   the   use   of   the   synchronous   speech   technique   can   lead   to   improved   rhythmic   performance   in   a   second   language   in   spontaneous   production.   While   the   focus   of   this   study   was   on   rhythm   teaching   and   learning,   we   also   aimed   to   identify   other   areas   of   oral   performance   that   might   benefit   from   synchronous   speech   practice.   Although   the   current   small-­‐scale   study   cannot   establish   definitive   effectiveness   of   the   technique,   we   sought   to   garner   preliminary   data   supporting   its   potential   effectiveness,   both   in   enhancing   performance   in   the   synchronous   contexts,   as   well   as   transferring   improved   performance   to   spontaneous   SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 71 speech.  Thus,  the  present  investigation  is  a  feasibility  study,  working  out  a  protocol  to  test  the  method  and  to   establish  whether  preliminary  results  would  justify  further  testing.     2.  Literature  Review   Languages   differ   from   one   another   in   terms   of   prosody   and   rhythm.   The   rhythm   of   the   English   language   features   alternations   of  more   prominent   and   less   prominent   syllables,   the  more   prominent   ones   demonstrating  greater   intensity,   fundamental   frequency,  and  duration   than   the   less  prominent  ones,  which   often  get  reduced  (Fry,  1955;  Ladefoged,  2005;  Lehiste,  1970).  Production  studies  have  consistently  shown   that   L2   learners’   production   of   these   suprasegmental   features   in   English   deviates   from   native   speakers’   performance.  This  most  often   is   a   reflection  of  non-­‐native   speakers’   experience  with  patterns  and   features   that   are   employed   and   viewed   as   critical   in   their   first   language   phonological   system   and   subsequently,   transferred   to   learners’   L2,   where   they   are   continued   to   be   used   in   L1-­‐specific   ways.   For   example,   prior   research   has   shown   that   L2   learners   tend   to   apply   phonological   knowledge   of   their   native   language  when   perceiving  and  producing  stressed  and  unstressed  syllables  in  English  (Chen,  Robb,  Gilbert,  &  Lerman,  2001;   Lee,  Guion,  &  Harada,  2006;  Yu  &  Andruski,  2010),  duration  cues  (Lehiste  &  Fox,  1992)  and  vowel  reduction   (Braun,  Lemhöfer,  &  Cutler,  2008;  Flege  &  Bohn,  1989;  Zhang,  Nissen,  &  Francis,  2008).     The   use   of   such   non-­‐native   patterns   in   L2   speech   can   have   detrimental   effects   on   speech   intelligibility,  all  the  more  so  in  instructional  speech  that  delivers  novel  and  cognitively  demanding  content.   L2-­‐accented  instructors  are  the  reality  in  many  U.S.  universities  that  are  increasingly  reliant  on  International   Teaching   Assistants   for   filling   the   instructional   needs   of   U.S.   undergraduate   students   (Chiang,   2009).   The   quality   of   instructors’   oral   delivery   should   ensure   access   to   the   academic   content   maximally   quickly   and   effectively,  yet  high  numbers  of  undergraduate  students  experience  difficulties  understanding  or  connecting   with   their   instructors’   speech   (Pickering,   2001;  Rao,   1995).  According   to   Stevens   (1989),   students   cite   the   non-­‐native  use  of  suprasegmentals  as  the  main  reason  for  the  failure  to  decode  their   instructors’  messages.   Suprasegmental  accuracy  has  been  long  viewed  as  more  critical  for  comprehension  than  segmental  accuracy   in  L2  pronunciation  literature  (Anderson-­‐Hsieh  et  al.,  1992;  Derwing,  Munro,  &  Wiebe,  1998;  Morley,  1991).   Some  of   the  most   common  problem  areas  addressed   in   ITA   research   literature  have  been   improper  use  of   stress,  insufficient  vowel  reduction,  inadequate  intonation  patterns,  pausing,  and  speech  rate  (Gorsuch,  2013;   Hahn,   2004;   Kang,   2008;   2010;   Pickering,   2001),   which   have   interfered   with   undergraduate   students’   performance   at   some   level.   The   time   for   speech   training   in   the   ITA   training   curriculum   is   unfortunately   limited   and   often   needs   to   be   shared   with   other   aspects   of   ITA   training   such   as   culture   and   pedagogy   (Anderson-­‐Hsieh,   1990).   This   calls   for   highly   effective   pronunciation   teaching   techniques   that   yield   immediate   improvements   in   the   suprasegmental   realm.   Some   of   the   teaching   approaches   proposed   for   classroom   teaching  of   suprasegmentals  have  been,   for   example,   a   cognitive   approach  with   a  metacognitive   analysis   of   the   suprasegmental   system   and   subsequent   self-­‐study   using   field-­‐specific  materials   (Anderson-­‐ Hsieh,  1990)  and  a  drama-­‐based  approach  (Stevens,  1989).  Often,   the   focus  has  been  entirely  on  self-­‐study   after  pronunciation  instruction,  such  as  an  input-­‐based  approach  in  the  form  of  repeated  reading  along  with   an  audio  recording  (Gorsuch,  2011),  or  covert  practice  with  self-­‐monitoring  and  self-­‐correction  (Dickerson,   2000).  However,  in  order  to  maximize  class  time,  the  search  for  most  effective  teaching  techniques  continues.     The   current   study   was   designed   to   provide   empirical   evidence   for   the   effectiveness   of   the   synchronous  speech  technique,  which  has  been  anecdotally  observed  to  yield  positive  results   in  L2  rhythm   acquisition.   The   exploration   of   this   technique   was   motivated   by   the   empirical   accounts   of   synchronous   speech  being  successfully  used  to  achieve  the  desired  speech  effect  in  the  field  of  communication  disorders,   where  it  is  used  in  stuttering  therapy  for  speech  rate,  prosody  control,  and  fluency  development  (Freeman  &   Armson,   1998;   Ham,   1988;   Kalinowski   &   Saltuklaroglu,   2003;   Kiefte   &   Armson,   2008).   While   L2   speech   acquisition  is  an  entirely  different  discipline  and  of  a  non-­‐clinical  nature,  the  shared  speech  intelligibility  goal   justifies   further   explorations   of   professional   practices   used   in   the   speech   disorder   discipline   as   part   of   an   interdisciplinary  quest  for  most  effective  instructional  tools.     In  the  literature,  the  simultaneous  speech  production  by  two  speakers  has  been  referred  to  in  various   ways:  synchronous  speech  (Cummins,  2003),  choral  speech  or  chorus  repetitions  (Freeman  &  Armson,  1998;   Kjellin,   1999),   and   unison   speech   (Ham,   1988).   It   is   important   to   note   two   things:   first,   the   technique   described  here  does  not  involve  rhythmical,  highly  structured  chanting,  such  as  reciting  poems  and  rhymes,   but   rather   refers   to   any   other   less   structured   text   that   one   would   encounter   in   their   daily   lives.   Second,   synchronous   speech,   as   its   name   suggests,   really   denotes   speech   that   is   created   by   two   people   simultaneously,  without  a  significant  lag  in  production.  This  is  different  from  speech  shadowing,  a  related  yet   BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 72 different   term,   in  which   one   speaker   attempts   to   reproduce   the   other   speaker’s   speech   after   it   is   uttered,   without  a  text  in  front  of  them,  and  with  no  familiarity  with  the  upcoming  utterance  (Cummins,  2003).  While   Cummins  (2003)  points  out  that  there  are  lags  also  in  a  synchronous  speech  condition,  they  are  much  shorter   than   the   shortest   lags   in   shadowed   speech.   Kalinowski   and   Saltuklaroglu   (2003)   describe   the   timing   of   synchronized  speech  as  “one  signal  imitating  the  other  in  a  loosely  defined  temporal  window”  (p.  342).     Kjellin   (1999)   has   provided   an   experience-­‐based   account   of   introducing   synchronous   speech   into   second   language  pronunciation   instruction,  where  relatively  short  phrases  were  used   for   teaching  prosody.   He   notes   that   choral   repetition   is   not   passive   audio-­‐lingual   drilling,   but   rather   efficient   active   work   both   mentally  and  physically  automatizing  a  skill,  the  instant  results  of  which  generates  a  high  level  of  motivation   in  the  student,  so  repetitions  never  get  boring.  Chela-­‐Flores  (1994)  gives  indirect  support  to  the  synchronous   speech  technique  in  second  language  rhythm  acquisition  by  pointing  out  the  need  for  the  student  to  develop  a   feel  for  the  continuous  flow  of  rhythm,  rather  than  working  on  rhythm  on  a  word-­‐by-­‐word  basis  that  does  not   let  the  learner  perceive  it  as  a  whole  and  form  a  general  habit  of  its  use.  In  such  cases,  rhythm  use  is  limited   only  to  the  learned  words  and  phrases  and  does  not  generalize  to  out-­‐of-­‐class  spontaneous  speech  situations.   Similarly,   she   emphasizes   the   difference   between   those   mechanisms   that   lead   to   conscious   knowledge   of   language   rules   but   do   not   normally   develop   oral   fluency,   and   those   mechanisms   that   develop   automatic   language  competency,  claiming  that   in   the  traditional  rhythm  instruction  more  attention   is  usually  given  to   the  former.     “Conventional”   rhythm   teaching   typically   combines   both   explicit   instruction   and   meaningful   and   communicative  practice   to   gradually  move   from   the  declarative   stage  of   skill   acquisition   to   the  procedural   stage   and   finally   to   the   automatic   stage   (DeKeyser,   2007),  which   is   a   relatively   slow  process.   Synchronous   speech  could  hypothetically  accelerate  rhythmic  automaticity  by  focusing  on  the  final  automatization  phase,   where  rhythmic  patterns  would  be  internalized  through  their  smooth  and  fluent  use.  The  intensive  nature  of   synchronous   speech   provides   frequent   perceptual   exposures   to   the   feature   learned   and   a   considerable   amount   of   motoric   practice.   Thus   the   synchronous   speech   technique   could   minimize   reliance   on   the   metalinguistic  rules  of  rhythm  production  and  the  procedural  stage  of  rhythm  learning,  instead  allowing  the   learners  to  fully  immerse  themselves  into  “assisted”  native-­‐like  rhythm  use.   A   neurological   perspective   on   synchronous   speech   could   explain   why   instant   imitation   of   native   speaker’s   speech   gestures   leads   to   automaticity   and   could   potentially   be   a  more   effective   tool   for   rhythm   acquisition  than  more  analytical  approaches.  Kalinowski  and  Saltuklaroglu  (2003)  articulate  the  position  that   all  humans  possess  “mirror  neurons”  that  create  an  immediate  link  between  an  observation  and  an  action  (or   perception   and   production)   and   allow   humans   to   imitate   complex   motor   sequences.   While   the   status   of   mirror   neurons   remains   unclear   and   somewhat   controversial,   recent   reviews   indicate   that   there   is   strong   evidence   to   support   their   existence   and   role   in   motor   imitation   and   hence   motor   learning   (Molenberghs,   Cunnington,  &  Mattingley,  2012).  Thus  synchronous  speech  could  be  conceptualized  as  a  simple  input-­‐output   reflex   mediated   by   mirror   neurons,   during   which   “the   only   cognitive   tasks   required   are   perceiving   and   attending   to   the   external   signal   during   speech   production”   (Kalinowski   &   Saltuklaroglu,   2003,   p.344).   Regardless  of  whether  such  a  strong  claim  can  be  substantiated,  the  mere  presence  of  a  mirror  neuron  system   can  be  seen  as  evidence  for  the  neurological  importance  of  imitation  in  human  motor  learning.     In  sum,  the  strength  of  the  synchronous  speech  technique  lies   in  the  fact  that  rhythm  is   learned  (i)   psycho-­‐motorically   and   implicitly,   (ii)   uninterruptedly   and  holistically,   (iii)   intensively   and   time-­‐efficiently,   while  (vi)  maintaining  consistency  across  speech.  The  flow  of  rhythm  develops  in  time;  interrupting  the  flow   by   the   instructor   and   the   student   taking   turns   would   be   detrimental   to   rhythm   learning.   Therefore,   the   continuity  and  fluency  of  practice  gives  students  an  opportunity  to  use  the  desired  rhythm  in  a  native-­‐like  and   fluent  manner,  forming  a  strong  psycho-­‐motoric  habit  without  much  conscious  awareness.  Additionally,  this   lets   students   hear   themselves   speak   according   to   native   prosodic   norms,   which   might   be   a   powerful   motivator.  Moreover,  temporal  features  of  rhythm  are  hard  to  maintain  without  a  real-­‐time  metronome  that   is  instructor’s  speech,  and  could  be  lost  if  not  paced  and  aligned  by  synchronous  speaking.  The  outside  real-­‐ time  meter   helps   students   set   their   “inner   clock”.   Finally,   automatizing   a   skill   is   the   final   and  most   time-­‐ consuming   stage   of   skill   acquisition,   yet   this   final   stage   of   skill   learning   is   often   cut   short   due   to   time   constraints.   The   synchronous   speech   technique   puts   emphasis   on   the   last   step  —   the   skill   automatization   process  —  from  the  very  first  day,  and  through  fluent  and  intensive  use  could  arguably  turn  the  new  feature   into  a  habit  faster.     SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 73 3.  Methods     3.1.  Research  design   The  small-­‐scale   feasibility   study  used  a  pre-­‐test  post-­‐test   control  group  quasi-­‐experimental  design,   where  the  control  group  (n  =  3)  worked  with  the  regular  rhythm  instruction,  and  the  experimental  group  (n  =   3)  worked  with  a  synchronous  speech  component  added  to  the  regular  instruction.  The  research  participants   were  audio-­‐recorded  at  the  beginning  and  end  of  the  6  weeks  of  rhythm  instruction  period  to  obtain  (i)  initial   baseline  speech  samples  and  (ii)   final  post-­‐instruction  speech  samples.  The  last  ones  assessed  the  potential   transfer  to  novel  contexts/spontaneous  speech  as  students  were  giving  short  unscripted  presentations.     The   baseline   and   final   recordings   were   further   subjected   to   perceptual   measures   to   compare   progress   in   rhythmic   performance   in   each   group.   Recordings   from   all   six   participants  were   first   low-­‐pass   filtered  with  speech  analysis  software  to  retain  only  suprasegmental   information,   in  order   to  eliminate  any   potential   bias   from   context   factors,   and   further   perceptually   rated   by   three   certified,   licensed   speech-­‐ language  pathologists,  all  naïve  to  the  study,  on  (i)  rhythm  pattern  acceptability,  (ii)  speech  rate  acceptability,   and  (iii)  overall  speech  acceptability  for  instructional  purposes.     3.2.  Participants   The  study  was  conducted  within  a  6-­‐week-­‐long  course  of  oral  proficiency  for  international  graduate   teaching   assistants   (ITAs)   in   a   department   of   communication   sciences   and   disorders   at   a   university   in   Midwestern   United   States.   In   this   course,   each   international   student   received   individual   instruction   by   a   speech-­‐language   pathology   graduate   student,   supervised   by   a   licensed   and   certified   speech-­‐language   pathologist.     Two  groups  of  participants  were  involved  in  the  study:  first,  six  international  students  (one  female,   five  male),  and  second,  six  native-­‐English  speaking  instructors  (all  female)  who  all  volunteered  to  participate   in  the  experiment.  The  international  students  had  all  been  screened  for  English  language  oral  proficiency  by   ESL  experts  during  the  students’  first  days  on  campus.  If  their  speech  patterns  did  not  meet  the  established   fluency  criteria,   they  were   further  referred  to  the  department  of  communication  sciences  and  disorders   for   placement   in   their   oral   proficiency   program.   The   supervising   SLP   along   with   the   participating   graduate   students   administered   initial   oral   proficiency   screenings   to   each   referred   student   and   identified   individual   pronunciation  goals.  Finally,  those  international  students  who  were  identified  as  having  non-­‐native  prosodic   patterns  were  further  offered  participation  in  the  current  study.     The   six   international   participants   who   volunteered   were   randomly   assigned   to   control   and   experimental   groups,   controlling   only   for   the   native   language   factor:   the   complexity   of   certain   languages   could  affect  the  speed  with  which  rhythm  is  acquired;  therefore,  participants’  native  languages  were  balanced   across   both   groups.   Thus,   there   were   two   speakers   of   Chinese   in   each   group,   while   the   remaining   two   participants  —  one  speaker  of  Nepali  and  one  speaker  of  Sinhala  —  were  randomly  assigned  to  the  control   and  the  experimental  groups,  respectively.     Table 1 Participants’ native language and field of study.   + NATIVE LANGUAGE FIELD OF STUDY Control 1 Chinese Chemistry Control 2 Chinese Chemistry Control 3 Nepali Geology Experimental 1 Chinese Biology Experimental 2 Chinese Math & Statistics Experimental 3 Sinhala Chemistry     BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 74 3.3.  Procedure   Speech  instructors  worked  with  the  ITAs  on  English  rhythmic  patterns  for  a  total  of  six  consecutive   weeks  during  the  semester,  where  approximately  up  to  30  minutes  out  of   the  regular  50-­‐minute  class  each   week   were   devoted   to   teaching   rhythm   and   stress.   Although   each   session   was   conducted   individually,   maximum  measures  were  taken  to  make  sure  that  all  instructors  performed  the  same  way  through  intensive   training   and   constant   monitoring.   Before   meeting   with   their   students,   all   native-­‐speaking   instructors   completed   weekly   joint   training   sessions   where   they   practiced   synchronous   speech   with   each   other   and   shared  their  experiences.   Instructors  were  provided  with  detailed  guidelines  and   initial   texts   to  use   for   the   rhythm  teaching  part  of  each  session.     Each   instructor  met  with   his   or   her   assigned   ITA   in   a   quiet   therapy   room   equipped  with   a   video   camera.  One  of   the   researchers   constantly  monitored   the   instruction  process   via   camera  on   a   computer   in   another  room,  and  gave  instant  feedback  and  made  the  necessary  corrections  right  after  the  session  to  ensure   fidelity  of  treatment.     Both   control   and   experimental   groups   spent   an   equal   amount   of   time   per   session   on   rhythm   instruction.  Within  each  session,  all  students  went  through  controlled,  guided  and  free  speech  activities  (see   Celce-­‐Murcia,  Brinton  &  Goodwin,  2006);  however,  the  proportions  of  these  in  each  group  were  different.  The   experimental  group  received  proportionally  fewer  guided  and  free  speech  activities,  which  were  substituted   with  synchronous  speech  activities  instead.     The  explicit   instruction  on  rhythm  that  both  groups  received   involved  work  on  word  stress  (stress   realization:  duration,  pitch,  loudness,  as  well  as  vowel  reduction),  primary  phrase  stress,  content  and  function   words,   thought   groups,   stress   in   compound   nouns,   and   prominence   in   discourse   context   (stressing   new   information  versus  old,  emphatic  stress,  contrastive  stress).  Each  session  began  with  word-­‐  and  short  phrase-­‐ level   controlled   activities  with   ear   training   (Celce  Murcia   et   al.,   2006)   and  modeling   by   the   instructor   and   imitation   by   the   student.   The   following   guided   activities   involved   more   field-­‐specific   content   and   longer   utterances;  still,  the  attention  of  the  student  was  focused  on  rhythm  production.  Feedback  at  this  stage  was   explicit   and   immediate.   Finally,   the   free   speech   and   communicative   activities   were   aimed   at   developing   automaticity   and   use   of   the   newly   acquired   skills   in   spontaneous   and   also   cognitively   demanding   speech,   which   is   an  everyday   reality   for   teaching  assistants.  These  activities  were  aimed  at   fostering  presentation-­‐ type  of  speech  and  were  made  maximally  meaningful  to  the  ITAs;  for  example,  they  involved  teaching  some   basic  concepts  in  their  field,  describing  research  conducted  in  their  department,  giving  a  motivational  speech   for  prospective  students  in  their  department,  etc.  Feedback  at  this  stage  could  be  instant  or  delayed.     Both   experimental   and   control   groups   received   the   abovementioned   instruction,   but   only   the   experimental  group  received  instruction  using  the  synchronous  speech  technique  in  place  of  some  guided  and   free   speech   activities.   The   experimental   group   followed   specific   detailed   guidelines   for   work   with   synchronous  speech.  The  materials  used  working  with   the  experimental   technique  were   field-­‐specific   texts   and  presentation  materials   that  were   either   selected   for   the   student  or,   later   in   the   semester,   provided  by   students   themselves   (e.g.,   descriptions   of   research   interests,   journal   abstracts,   upcoming   presentations).   Using   contexts   relevant   to   students’   academic   needs   was   an   approach   selected   to   enhance   student   motivation.  The  texts  used  consisted  of  whole  paragraphs,  and  due  to  their  length  could  be  repeated  several   times  only.  Nevertheless,  the  length  of  the  material  also  meant  that  the  flow  of  rhythm  could  be  internalized   more  continuously  and  intensively.   Initially,   each   synchronous   speech  activity  began  with   the   instructor  modeling   for   the   student   and   the  student  silently   lip-­‐reading  along  to  familiarize  him  or  herself  with  the  text.  Next,   the  student   joined  in,   and  both  student  and  instructor  spoke  in  unison.  Depending  on  the  length  of  the  text,  it  was  repeated  three  to   five  times,  with  the  instructor  prompting  the  student  to  gradually  establish  eye  contact  with  the  “audience”,   i.e.,  the  instructor.  This  was  done  in  order  to  minimize  reliance  on  the  script  and  to  create  an  impression  of  a   real  presentation.  Looking  up  frequently  at  the  imaginary  audience  also  helped  maintain  thought  groups  and   appropriate  pausing  in  the  running  speech.  Often,  the  student  and  the  instructor  would  stand  in  the  opposite   corners  of  the  therapy  room  to  create  distance  between  the  speaker  and  the  audience,  creating  the  effect  of  a   real  presentation  where  appropriate  eye  contact  and  appropriate  volume  needs   to  be  maintained;   this  also   tended  to  increase  the  prominence  of  stressed  syllables  in  speech.  Additionally,  to  minimize  the  “crutch”  that   the   instructor’s   voice  might  have   turned   into  during   the   synchronous   speech  activity,   the   instructor  would   significantly   lower  the  volume  of  his/her  voice  at   times  and   let   the  student   take   the   lead  so  students  could   hear   themselves  produce   speech   in   a  near-­‐native  manner  by   themselves.  Afterwards,   the   instructor  would   tune  back  in  and  continue  as  usual.     SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 75 It  was   important   for   the   instructors   to  attend   to   several   things  while  working  with   this   technique:   first,  they  had  to  project  their  voices  loudly  enough  that  the  student  could  speak  in  full  voice  and  still  hear  the   instructor’   voice,  which  was   not   always   easy.   Second,   the   instructors  were   trained   to   use   expressive   body   language  (e.g.,  raising  eyebrows,  looking  up,  leaning  forward)  to  amplify  rhythmic  patterns.  Body  language  is   of  benefit  in  such  instruction  since  it  allows  speech  synchronization  through  peripheral  vision  while  reading  a   text   (Cummins,   2003),   alerts   the   student   that   a   new   thought   group   is   coming,   and,   overall,   serves   as   a   reminder   of   the   physiological   aspect   of   stress   production.   In   this   way,   body   language   acts   as   a   real-­‐time   feedback  mechanism,  since  it  can  convey  the  right  timing  and  prominence  without  using  voice  or  any  other   signal.  Overall,   corrective   feedback  was   implicit,  delivered  real-­‐time  via  amplified  auditory  and  visual   cues.   For   example,   if   the   student   rushed   ahead,   the   instructor   could   slow   down   her/his   speech   even   more   or   prolong  the  stressed  syllables  if  they  were  of  insufficient  duration  in  the  student’s  production.       3.4.  Data  Collection   To  establish  the  efficacy  of  the  synchronous  speech  technique,  pre-­‐test  and  post-­‐test  recordings  were   obtained   from   each   group.   Recordings   were   made   with   Panasonic   RR-­‐US470   digital   audio   recorders   and   SHURE  C606  professional  microphones  attached  to  them,  which  were  distributed  to  all   instructors.  Overall,   students  were  recorded  six  or  more  times  during  the  semester,  but  only  the  first  and  the  last  recordings  were   used  in  this  study  as  we  were  interested  in  the  maximum  result  each  technique  could  achieve  in  the  limited   amount  of  time  allocated  for  rhythm  instruction.  Students  were  unaware  of  which  recordings  would  be  used   for   research   purposes,   and   recordings  were   regularly   used   in   class   for  work  with   segmentals,   so   students   were  relatively  comfortable  with  being  recorded.     For  baseline  recordings,  all  participants  read  a  scripted   text   (see  Appendix  A).  They  were  asked   to   first  familiarize  themselves  with  the  text  and  read  it  silently,  and  then  read  it  and  say  each  line  aloud,  looking   up,   as   if   presenting   it   to   their   class.   For   the   final   recordings   that   assessed   retention   and   transfer   to   spontaneous  production,  students  were  recorded  as  they  produced  a  novel  and  spontaneous  speech  that  they   had   a   few   minutes   to   prepare.   Prior   to   recordings,   students   were   told   to   remember   the   segmental   and   suprasegmental   features   they   had   worked   on   during   the   semester   and   use   them   accurately   in   their   final   speech;   the  very   final   session  was  devoted  entirely   to  assessment  and   involved  no  rhythm   instruction.  The   speech   had   to   be   on   some   concept   from   their   field   they   could   explain   and   then   students   were   asked   to   “present”   it   while   they   were   being   recorded.   The   strong   focus   on   the   content   made   attending   to   speech   patterns  cognitively  demanding,  as  students  had  to  construct  a  novel  text  in  their  minds  and  simultaneously   maintain  appropriate  rhythm.  Since  most  of   the  cognitive  resources  were  spent  on   the  challenging  context,   leaving   little   for  attending   to   the  rhythmic  performance,   the  ability   to  maintain  rhythm   in   the  spontaneous   and   content-­‐wise  difficult   speech  was   taken  as   an   indicator   that   a   certain  degree  of   automaticity  had  been   achieved.       3.5.  Data  Analysis   Data  were   analyzed   using   descriptive   statistics   to   assess   the   outcomes   of   the   synchronous   speech   component  in  the  experimental  group  and  establish  research  feasibility  of  this  method  for  future  studies.  The   small   sample   size   of   this   study   did   not   allow   the   use   of   inferential   statistical   methods   or   the   ability   to   generalize   results   to   larger   populations.   Selected   recorded   speech   samples   from   all   six   participants   were   randomized  and  presented  to  three  naïve  speech-­‐language  pathologists  who  assigned  impressionistic  ratings   to  the  speech  in  terms  of  (i)  rhythm  pattern  accuracy,  (ii)  speech  rate  acceptability,  and  (iii)  overall  speech   acceptability  as   instructional   speech.  Regarding   the   latter,  we  hypothesized   that   there  might  potentially  be   additional  factors  besides  the  first  two  that  could  influence  the  perceived  acceptability  of  speech.     All  speech  samples  were  approximately  30  seconds  long,  and  contained  not  less  than  five  sentences.   Such  sample  length  was  deemed  to  be  sufficient  to  appropriately  reflect  the  target  patterns  and  decrease  the   burden  on  raters.  Since  the  baseline  recordings  were  all  done  on  the  same  text  for  all  speakers,  the  30-­‐second   portion  was  randomly  spliced  out  of  various  time  points  so  that  the  content  would  not  repeat  itself  and  raters   would   not   develop   familiarity   effects  with   speech   that  would  make   them   rate   samples   later   in   time  more   favorably.  The  final  recordings,  on  the  other  hand,  were  recorded  on  spontaneous  speech  that  was  different   for  each  individual  speaker;  therefore,  the  30-­‐second  speech  samples  were  spliced  out  from  the  middle  of  the   recording  for  all  participants.     To   ensure   that   raters   only   focused   on   the   suprasegmental   aspect   of   speech   and   interference   of   segmental   information   was   eliminated,   all   speech   samples   were   low-­‐pass   filtered   at   1000   Hz   using   Praat   BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 76 speech  analysis  software  (Boersma  &  Weenik,  2011).  Low-­‐pass  filtering  passes  lower  frequencies  and  rejects   higher  frequencies,  making  segmental  information  relatively  unavailable  to  the  listener  as  words  are  almost   impossible  to  make  out.  Suprasegmental  information,  on  the  other  hand,  was  all  preserved  and  thus  the  only   quality  information  available  to  raters.   Speech   samples   were   presented   to   raters   on   a   computer   via   headphones.   A   100   mm   long   visual   rating  scale  was  used,  where  raters  selected  a  point  on  the  continuous  blank  scale,  marked  as,   for  example,   “non-­‐native  rhythm”  at  one  end,  and  “native-­‐rhythm”  at  the  other  end  (see  Appendix  B).  Raters  were  told  to   make   use   of   the  whole   scale.   Care  was   taken   to  make   sure   that   all   raters   defined   the   notion   of   “rhythm”   identically   and   were   able   to   apply   it   to   the   ratings;   therefore,   they   received   a   short   introduction   to   the   rhythmical   system   of   the   English   language,   followed   by   a   short   training   session   on   random   and   study-­‐ unrelated  low-­‐pass-­‐filtered  native  and  non-­‐native  speech  samples.     Raters   could   listen   to   each   speech   sample   twice   and   they   then  marked   the   scale   according   to   the   perceived   performance   in   English.   The   inter-­‐rater   reliability   was   high,   r   (34)   =.71,   p   <   .01.   Intra-­‐rater   reliability  on  the  same  speech  samples  that  raters  had  to  rate  twice  was  even  higher,  r   (16)  =   .93,  p  <   .01).   Each  individual  ITA’s  progress  during  the  semester  was  expressed  as  a  difference  in  scaled  units  (0  -­‐  100  on   the   visual   rating   scale)   between   the  baseline   recording   and   the   final   recording   (i.e.,   final   perceptual   rating   minus  initial  perceptual  rating),  averaged  across  all  three  raters.     4.  Results   The  two  groups’  performances  were  rated  on  a  scale  of  1  to  100  for  rhythm  accuracy  in  spontaneous   oral   production,   as   well   as   acceptability   of   speech   rate   and   overall   speech   patterns   in   spontaneous   production.  An  individual  differences  analysis  revealed  that  one  of  the  participants  in  the  Experimental  group   exhibited  a  profile  that  matched  neither  the  control  nor  experimental  group:  her  performance  at  the  end  of   the  semester  was  well  below  the  average  in  any  of  the  groups;  therefore,  this  participant  was  excluded  from   further  analyses.     Figure   1   presents   results   of   the   perceptual   improvement   in   rhythmic   pattern   use   in   spontaneous   speech,  which  was  the  focus  of  this  study.  The  amount  of   improvement  in  the  control  group  (n  =  3)  was  an   average  of  18.4  units,  while  in  the  experimental  group  (n  =  2)  the  improvement  was  an  average  of  30.7  units.   This   indicates   that,   perceptually,   the   experimental   group   experienced   greater   improvement   in   second   language  rhythm  use  in  spontaneous  production  than  the  control  group,  potentially  due  to  the  addition  of  the   synchronous  speech  technique.         Figure 1. Difference in speech rhythm acceptability ratings pre- and post-instruction on a scale from 0 to 100, indicating perceived improvement in speech rhythm in control and experimental groups. SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 77 Figure  2  illustrates  perceptual  ratings  for  improvement  in  the  overall  speech  patterns.  A  slight  trend   toward  greater  overall  speech  improvement   in  the  experimental  group  (M  =  33.6),  compared  to  the  control   group  (M  =  19.3),  can  be  observed.         Figure 2. Difference in overall speech pattern acceptability ratings pre- and post-instruction on a scale from 0 to 100, indicating perceived improvement in overall speech patterns in control and experimental groups.   Finally,  while  both  groups  were  rated  as  having  made  comparable  progress  in  improving  speech  rate,   improvement  was  more  pronounced  in  the  experimental  group  (M  =  37.5),  compared  to  the  control  group  (M   =  20.5),  again  showing  a  trend  towards  greater  benefits  of  adding  the  synchronous  speech  technique.       Figure 3. Difference in speech rate acceptability ratings pre- and post-instruction on a scale from 0 to 100.       BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 78 5.  Discussion  and  pedagogical  implications   The   number   of   international   graduate   students   teaching   at   U.S.   universities   is   large   and   growing.   Interactions   between   ITAs   and   American   undergraduates   have   often   resulted   in   communication   failures,   which   is   hardly   surprising   given   that   the   amount   of   time   allotted   to   ITA   preparation,   speech   training   in   particular,   typically   is   short.   Various   approaches   for   teaching  English   speech  patterns  have  been  proposed   and   tested   empirically,   but   the   synchronous   speech   technique   so   far   has   not   been   addressed   in   research   literature.   This   study   attempted,   first,   to   establish   the   feasibility   of   testing   the   effectiveness   of   the   synchronous   speech   technique  with  experimental  methods  and  develop  a  protocol   to   test   the  method,   and   second,   to   provide   preliminary   experimental   evidence   for   the   effectiveness   of   the   synchronous   speech   technique   in  L2   rhythm  teaching.  The  hypothesized  benefits  of  using  synchronous  speech   in  pronunciation   instruction   were   the   simultaneously   available   rich   input   and   practice,   as   well   as   implicit,   fluent   and   continuous   rhythm   learning   that   could   potentially   allow   for   automatizing   rhythm   production   faster   than   other  rhythm  teaching  methods  and  techniques.   Out   of   the   six   volunteers  who   agreed   to   participate   in   our   small-­‐scale   study   conducted  within   an   existing  pronunciation  course,  one  participant  was  eventually  eliminated  due  to  exhibiting  performance  that   was  well  below  the  average  for  both  the  experimental  and  control  groups,  apparently  failing  to  benefit  from   any   type   of   instruction.   There   could   be   a   range   of   possible   explanations   here,   including   the   fact   that   this   course  was  mandatory   for   all   six   potential   teaching   assistants   and   certain  motivation-­‐related   issues  might   have  played  a  role.     The  study  findings  with  the  five  remaining  participants  partially  confirmed  our  casual  observations   and   intuitions   regarding   the   effectiveness   of   the   synchronous   speech   technique:   observable   differences   between   the   experimental   and   control   groups   were   found,   such   that   there   was   a   trend   toward   more   pronounced  improvement  in  speech  rhythm  acquisition  when  the  synchronous  speech  technique  was  added   to  instruction.  Considering  the  small  sample  size  (six  participants  total,  out  of  which  one  was  eliminated  as  an   outlier)  and  the  limited  time  for  instruction  (a  total  of  six  weeks,  not  more  than  half  an  hour  per  week),  it  can   be  considered  an  encouraging  preliminary  finding.  An  added  synchronous  speech  component  thus  has  shown   potential   to   accelerate   learning   of   second   language   rhythm   patterns   and   aid   their   transfer   to   new   and   untrained   contexts.   However,   most   surprising   and   encouraging   might   be   the   fact   that   these   rhythmic   improvements  were   shown   in   spontaneous   speech  production,  which   is   the  ultimate  goal  of  pronunciation   training.   What   is   more,   improved   rhythmic   performance   was   demonstrated   in   cognitively   demanding   contexts,  where  most   of   the   cognitive   resources  were   devoted   to   the   content   of   the   speech.   These   results   indicate   that   automaticity   in   second   language   rhythm  use  may   be   achieved   faster  with   the   addition   of   the   synchronous  speech  technique  than  with  conventional  rhythm  teaching  techniques  alone.   Initial  screenings  indicated  that  speech  rate  normalization,  specifically  its  reduction,  was  among  the   many  semester  goals  set  for  these  students.  Although  trending  toward  greater  improvement,  the  application   of   the   synchronous   speech   technique   did   not   appear   to   help  with   speech   rate   normalization   considerably   more  than  the  “conventional”  rhythm  teaching  techniques.  Neither  baseline  nor  final  ratings  as  assigned  by   the  three  raters  reached  the  maximum  score;  therefore,  the  possibility  of  a  ceiling  effect  was  eliminated.  One   possible   explanation   for   the   clear   lack   of   difference   between   groups   could   be   that   the   time   spent   on   the   presentation   type   of   speech   for   both   groups,   irrespective   of   the   type   of   rhythmic   training   received,   led   to   increased  awareness  of  the  needs  of  the  audience  and  consequently  speech  rate  normalization.     Similarly,   the   experimental   group   was   not   perceived   by   raters   as   having   convincingly   more   improvement  in  the  overall  speech  patterns  (intonation,  pausing  and  other  factors  included)  than  the  control   group,  although  a  trend  toward  more  improvement  was  again  observed.  Possibly,  the  amount  of  time  devoted   to  working  with  the  synchronous  speech  was  insufficient  to  reveal   its   full  potential   in  this  realm.  The  small   sample  size  and  unequal  number  of  participants  limit  any  conclusions  in  this  regard.     From  our  preliminary  results,  we  can  tentatively  suggest  that  one  aspect  of  second  language  speech   learning   where   synchronous   speech   could   have   superiority   over   traditional   techniques   is   rhythm   pattern   acquisition,   which  was   the   focus   of   this   study.   More   importantly,   the   findings   for   the   experimental   group   indicate   that   30   minutes   of   instruction   over   six   weeks   might   be   sufficient   to   see   the   benefits   of   the   synchronous  speech  technique  for  automating  second  language  rhythm  patterns.  The  presence  of  the  outlier   who   decreased   her   rhythmic   accuracy   and   speech   acceptability   following   six   weeks   of   instruction   is   a   particular   complicating   factor   in   interpreting   the   present   results.   Thus   a   larger-­‐scale   study  with   a   greater   sample  size  and  a  longer  instruction  period  is  needed.     SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 79 Overall,   our   results   supported   the   feasibility   of   testing   the   technique   experimentally   by   using   the   selected  measures  of  improvement  and  the  efficacy  of  this  method  for  rhythm  training  in  general.  The  use  of   the  low-­‐pass  filtered  speech  and  impressionistic  ratings  as  a  measure  proved  to  be  sensitive  enough  to  detect   differences   in   the   rhythmic   improvement   after   a   limited   number   of   training   sessions.   The   fact   that   the   experiment   was   conducted  within   an   existing   class   dictated   that   synchronous   speech   in   the   experimental   group  not  be  used  in  isolation  but  in  addition  to  more  explicit  rhythm  teaching  instruction  to  minimize  risks   for  the  students;  as  a  result,  the  ability  for  synchronous  speech  to  achieve  the  current  result  independently  is   yet  unknown.  Theoretically,  the  explicit  component  of  rhythm  teaching  that  is  needed  to  help  students  notice   second   language   features   could   be   substituted   with   real-­‐time   amplification   of   rhythmic   cues   during   synchronous  speech,  thus  making  rhythmic  acquisition  possible  without  referring  to  metalinguistic  means  of   instruction.     This   study   also   demonstrated   that   synchronizing   speech   for   native   and   non-­‐native   speakers   in   an   instructional  setting  was  feasible,  dispelling  our  initial  concerns.  Although  Cummins  (2002,  2003)  observed   that  when  two  people  recite  a  common  text  in  unison,  they  rapidly  attune  their  speech  to  each  other  and  are   able   to   successfully   synchronize   their   speaking  with   relative   ease,   these   observations  were   limited   to   two   native  speakers  who  share  the  same  language  and  expectations  regarding  its  prosody.  This  study  proved  that   such  observations  can  be  extended  also  to  native—non-­‐native  speaker  dyads;  however,  it  does  not  allow  for   generalizations   regarding   the   effectiveness   and   implementation   of   the   technique   in   larger   groups,   which   would  require  various  adjustments  and  more  creativity  on  the  part  of  the  instructor.     Although  the  population  that  was  targeted  with  the  synchronous  speech  technique  in  this  study  was   international   teaching   assistants,   its   application   is   not   limited   to   ITAs   or   academic   settings   but   can   be   successfully  used  in  a  variety  of  second  language  teaching  contexts  where  a  more  holistic  approach  to  rhythm   teaching  is  preferred  and  time  for  speech  instruction  is  limited.         References     Anderson-­‐Hsieh,   Janet   (1990).   Teaching   suprasegmentals   to   international   teaching   assistants   using   field-­‐ specific  materials.  English  for  Specific  Purposes,  9,  195-­‐214.   Anderson-­‐Hsieh,  Janet,  Johnson,  Robert,  &  Koehler,  Kenneth  (1992).  The  relationship  between  native  speaker   judgments  of  nonnative  pronunciation  and  deviance   in  segmentals,  prosody,  and  syllable  structure.   Language  Learning,  42(4),  529-­‐555.     Anderson-­‐Hsieh,   Janet,  &  Koehler,  Kenneth  (1988).  The  effect  of   foreign  accent  and  speaking  rate  on  native   speaker  comprehension.  Language  Learning,  38  (4),  561–613.   Braun,   Bettina,   Lemhöfer,   Kristin,   &   Cutler,   Anne   (2008).   English  word   stress   as   produced   by   English   and   Dutch   speakers:   The   role   of   segmental   and   suprasegmental   differences.   Proceedings   from   INTERSPEECH  2008:  9th  Annual  Conference  of   the   International  Speech  Communication  Association.   Brisbane,  Australia,  p.1953.   Braun,  Bettina,  Lemhöfer,  Kristin,  &  Mani,  Nivedita  (2011).  Perceiving  unstressed  vowels  in  foreign-­‐accented   English.  Journal  of  the  Acoustical  Society  of  America,  129  (1),  376–387.   Boersma,  Paul,  &  Weenink,  David  (2011).  Praat:  doing  phonetics  by  computer  [Computer  program].  Version   5.2.16,  retrieved  20  February  2011  from  http://www.praat.org.   Celce-­‐Murcia,   Marianne,   Brinton,   Donna,   &   Goodwin,   Janet   (1996).   Teaching   pronunciation:   Reference   for   teachers  of  English  to  speakers  of  other  languages.  Cambridge,  United  Kingdom:  Cambridge  University   Press.   Chela-­‐Flores,   Bertha  (1994).     On   the   acquisition   of   English   rhythm:   Theoretical   and   practical   issues.    International  Review  of  Applied  Linguistics,  32(2),  232-­‐242.     Chen,   Yang,   Robb,   Michael   P.,   Gilbert,   Harvey   R.,   &   Lerman,   Jay   W.   (2001).   A   study   of   sentence   stress   production   in  Mandarin  speakers  of  American  English.     Journal  of  the  Acoustical  Society  of  America,   109(4),  1681–1690.   BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 80 Chiang,   Shiao-­‐Yun   (2009).  Dealing  with   communication  problems   in   the   instructional   interactions  between   international   teaching   assistants   and   American   college   students.  Language   and   Education,   23(5),   461-­‐478.   Cummins,  Fred  (2002).  On  synchronous  speech.    Acoustic  Research  Letters  Online,  3(1),  7-­‐11.   Cummins,  Fred  (2003).  Practice  and  performance  in  speech  produced  synchronously.  Journal  of  Phonetics,  31,   139–148.   DeKeyser,  Robert  (2007).    Skill  acquisition  theory.  In  Jessica  Williams  and  Bill  Van  Patten  (Eds.),  Theories  in   second  language  acquisition:  An  introduction.  Mahwah,  NJ:  Erlbaum,  97-­‐113.   Derwing,Tracey,   Munro,   Murray,   &   Wiebe,   Grace   (1998).   Evidence   in   favor   of   a   broad   framework  for  pronunciation  instruction.  Language  Learning,  48(3),  393–410.   Dickerson,   Wayne   (2000).   Covert   rehearsal   as   a   bridge   to   accurate   fluency.   Presented   at   the   34th   Annual   Meeting  of  the  Teachers  of  English  to  Speakers  of  Other  Languages,  Vancouver,  BC.   Ellis,   Nick   C.   (2005).   At   the   interface:   Dynamic   interactions   of   explicit   and   implicit   language   knowledge.   Studies  in  Second  Language  Acquisition,  27,  305  –  352.   Ellis,  Nick  C.  (2007).  Implicit  and  explicit  knowledge  about  language.  In  Jasone  Cenoz  &  Nancy  H.  Hornberger   (Eds.),   Encyclopedia   of   language   and   education   (Vol.   6,   pp.   119–132).   New   York   City,   New   York:   Springer.   Field,  John  (2005).  Intelligibility  and  the  listener:  The  role  of  lexical  stress.  TESOL  Quarterly,  39,  399-­‐423.   Flege,  James  E.,  &  Bohn,  Ocke-­‐Schwen  (1989).  An  instrumental  study  of  vowel  reduction  and  stress  placement   in  Spanish-­‐accented  English.  Studies  in  Second  Language  Acquisition,  11,  35-­‐62.   Freeman,  Kathryn,  &  Armson,  Joy  (1998).  Extent  and  stability  of  stuttering  reduction  during  choral  reading.   Journal  of  Speech-­‐Language  Pathology  and  Audiology,  22,  188–202.   Fry,   Dennis   B.   (1955).   Duration   and   intensity   as   physical   correlates   of   linguistic   stress.   Journal   of   the   Acoustical  Society  of  America,  27,  765–768.     Gorsuch,  Greta  J.  (2011).  Improving  speaking  fluency  for  international  teaching  assistants  by  increasing  input.   TESL-­‐EJ,  14(4),  1-­‐25.     Gorsuch,  Greta  J.  (2013).  Helping  international  teaching  assistants  acquire  discourse  intonation:  Explicit  and   implicit  L2  knowledge.  Journal  of  Teaching  English  for  Specific  and  Academic  Purposes,  1(2),  67-­‐92.   Hahn,  Laura  (2004).  Primary  stress  and  intelligibility:  Research  to  motivate  the  teaching  of  suprasegmentals.   TESOL  Quarterly,  38(2),  201-­‐223.   Ham,  Richard  E.  (1988).  Unison  speech  and  rate  control  therapy.  Journal  of  Fluency  Disorders,  13(2),  115-­‐126.   Hulstijn,  Jan  H.  (2005).  Theoretical  and  empirical  issues  in  the  study  of  implicit  and  explicit  second  language   learning.  Studies  in  Second  Language  Acquisition,  27,  129-­‐140.   Kalinowski,   Joseph,  &  Saltuklaroglu,  Tim  (2003).  Choral  speech:   the  amelioration  of  stuttering  via   imitation   and  the  mirror  neuronal  system.  Neuroscience  and  Biobehavioral  Reviews,  27(4),  339-­‐347.   Kang,  Okim  (2010).  Relative  salience  of  suprasegmental  features  on  judgments  of  L2  comprehensibility  and   accentedness.  System,  38,  301-­‐315.   Kang,   Okim   (2008).   The   effect   of   rater   background   characteristics   on   the   rating   of   International   Teaching   Assistants  speaking  proficiency.  Spaan  Fellow  Working  Papers,  6,  181-­‐205.   Kiefte,   Michael,   &   Armson,   Joy   (2008).   Dissecting   choral   speech:   Properties   of   the   accompanist   critical   to   stuttering  reduction.  Journal  of  Communication  Disorders,  41,  33–48.   Kjellin,  Olle  (1999).  Accent  Addition:  Prosody  and  Perception  Facilitates  Second  Language  Learning.  In  Osamu   Fujimura,   Brian   D.   Joseph,   &   Bohumil   Palek   (Eds.),   Proceedings   of   LP'98   at   Ohio   State   University,   Columbus,  Ohio,  September  1998.  Prague,  Czech  Republic:  The  Karolinum  Press.   Ladefoged,  Peter  (2005).  A  course  in  phonetics  (5th  ed.).  Boston,  Massachusetts:  Heinle.   Lee,   Borim,   Guion,   Susan   G.,   &   Harada,   Tetsuo   (2006).   Acoustic   analysis   of   the   production   of   unstressed   English   vowels   by   early   and   late   Korean   and   Japanese   bilinguals.   Studies   in   Second   Language   Acquisition,  28,  487–513.   Lehiste,  Ilse  (1970).  Suprasegmentals.  Cambridge,  Massachussets:  MIT  Press.     SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 81 Lehiste,  Ilse,  &  Fox,  Robert  A.  (1992).  Perception  of  prominence  by  Estonian  and  English  listeners.  Language   and  Speech,  35(4),  419–434.   Mattys,   Sven   L.,   Brooks,   Joanna,   &   Cooke,   Martin   (2009).   Recognizing   speech   under   a   processing   load:   Dissociating  energetic  from  informational  factors.  Cognitive  Psychology,  59,  203-­‐243.   McAllister,   Robert,   Flege,   James   E.,   &   Piske,   Thorsten   (2002).   The   influence   of   L1   on   the   acquisition   of   Swedish  quantity  by  native  speakers  of  Spanish,  English  and  Estonian.  Journal  of  Phonetics,  30,  229– 258.     Molenberghs,  Pascal,  Cunnington,  Ross,  &  Mattingley,  Jason  B.  (2012).  Brain  regions  with  mirror  properties:  A   meta-­‐analysis  of  125  human  fMRI  studies.  Neuroscience  &  Biobehavioral  Reviews,  36(1),  341-­‐349.   Morley,  Joan  (1991).The  pronunciation  component  in  teaching  English  to  speakers  of  other  languages.  TESOL   Quarterly,  25  (3),  481-­‐520.   Munro,   Murray   J.   (1998).   The   effects   of   noise   on   the   intelligibility   of   foreign-­‐accented   speech.   Studies   in   Second  Language  Acquisition,  20,  139–154.   Pickering,   Lucy   (2001).   The   role   of   tone   choice   in   improving   ITA   communication   in   the   classroom.  TESOL   Quarterly,  35(2),  233-­‐255.   Rao,  Nagesh  (1995).  The  “oh  no!”  syndrome:  A  language  expectation  model  of  undergraduate  negative  reactions   toward  foreign  teaching  assistants.  Paper  presented  at  the  79th  Annual  Meeting  of  the  International   Communication  Association,  Albuquerque,  New  Mexico.     Stevens,   Scott   G.   (1989).   A   dramatic   approach   to   improving   the   intelligibility   of   ITAs.   English   for   Specific   Purposes,  8,  181–194.   Tajima,   Keiichi,   Port,   Robert,   &   Dalby,   Jonathan   (1997).   Effects   of   temporal   correction   on   intelligibility   of   foreign  accented  English.  Journal  of  Phonetics,  25,  1–24.   White,  Laurence,  &  Mattys,  Sven  L.   (2007)  Calibrating  rhythm:  First   language  and  second   language  studies.   Journal  of  Phonetics,  35(4),  501–522.   Woodford,   Charles,   Prichard,   Cheryl,   &   Jones,   Robin   (1998).   Listening   conditions   in   higher   education   classrooms:  One  method  of  improving  them.  Education,  119(1),  129-­‐134.   Yu,  Vickie,  &  Andruski,  Jean  (2010).  A  cross-­‐language  study  of  perception  of  lexical  stress  in  English.  Journal  of   Psycholinguistic  Research,  39(4),  323–344.   Zhang,  Yanhong,  Nissen,  Shawn  L.,  &  Francis,  Alexander  L.  (2008).  Acoustic  characteristics  of  English  lexical   stress   produced   by   native   Mandarin   speakers.   Journal   of   the   Acoustical   Society   of   America   123(6),   4498–4513.           BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 82 Appendix  A     Scripted  text  for  baseline  recordings     “When  world-­‐class  people  are  invited  to  campus,  and  they  are  important  in  your  discipline,  be  sure  to  attend,   if  at  all  possible.  Attend  their  lecture  even  if  they  are  from  a  different  discipline  than  yours.  Such  people  come   to   give   seminars   and   to   spend   a   day   or   two   with   faculty   and   students.   Often,   they   become   events   to   remember.  You  might  want  to  attend  just  to  see  who  those  people  are.  Go  even  if  you  are  not  interested  in  the   topic.   What   you   will   hear,   if   you   listen   carefully,   is   how   they   think   about   the   world,   how   they   approach   problems,   and   how   they   work   on   them.   They   might   change   your   perspectives   about   your   own   areas   of   teaching  and  research.      If   you  want   to   survive   in   a   research-­‐oriented   institution,   publish   as  much   as   possible.  Write  most   of   your   articles  for  refereed  journals.  Papers  presented  at  meetings  get  you  funds  to  travel.  However,  even  if  refereed,   these  papers  don’t  really  count  for  tenure,  salary  raise,  or  promotion.   Your  first  papers  might  be  rejected.  But  remember  –  every  paper  has  a  market,  unless  they  are  real   stinkers.  If  the  work  is  sound,  it  will  get  published.  Submit  your  papers  first  to  the  best  journals.  Work  your   way  down  the  list  if  a  paper  is  rejected.  In  general,  publish  early  and  often.  Begin  writing  for  publication  while   still  in  graduate  school.  Data  show  that  people  who  publish  while  still  in  graduate  school  usually  continue  to   publish  at  a  fast  rate  after  they  graduate  –  unlike  those  who  did  not  publish  while  still  a  student.”     (Adapted  from  Gray,  P.,  &  Drew,  D.  E.  (2008).  What  they  didn't  teach  you  in  graduate  school:  199  helpful  hints   for  success  in  your  academic  career.  Stylus  Publishing:  Sterling,  Virginia.)       SYNCHRONOUS  SPEECH       E-­‐JournALL  1  (2014),  pp.  69-­‐84 83 Appendix  B     Visual  Rating  Scale                                                     BANZINA,  HEWITT,  AND  DILLIT       E-­‐JournALL  1  (2014),  69-­‐84 84   Elina Banzina, Department of Languages, Stockholm School of Economics in Riga – Riga, Latvia elina.banzina@sseriga.edu EN Elina Banzina, Ph.D., is an Assistant Professor at the Department of Languages of Stockholm School of Economics in Riga. Her research interests involve English phonetics, perception and production of foreign-accented speech, intelligibility, and pronunciation teaching techniques, with a special focus on language rhythm ES Elina Banzina. Ph.D., es Profesora en el Departamento de Lenguas de la Escuela de Economía de Estocolmo, en Riga. Sus líneas de investigación incluyen la fonética inglesa, la percepción y la producción de la lengua hablada con acento extranjero, inteligibilidad y técnicas para la enseñanza de la pronunciación, con especial hincapié en el ritmo de la lengua.   IT Elina Banzina, Ph.D., è Ricercatrice confermata presso Dipartimento di Lingue della Scuola di Economia di Stoccolma nella città di Riga. I suoi interessi di ricerca includono la fonetica inglese, la percezione e la produzione di parlato con accento straniero, intellegibilità e tecniche per l’insegnamento della pronuncia, con particolare enfasi sul ritmo della lingua.   Lynne E. Hewitt, Department of Communication Sciences and Disorders, Bowling Green State University lhewitt@bgsu.edu EN Lynne Hewitt, Ph.D., CCC-SLP, is an Associate Professor and Chair of the Department of Communication Sciences and Disorders, Bowling Green State University, Bowling Green, OH. Her undergraduate training was in English and linguistics. She is a certified and licensed speech-language pathologist with clinical and research interests in language assessment and intervention. ES Lynne Hewitt, PhD., CCC-SLP (Certificado de competencia clínica en patologías del lenguaje), es Profesora y Directora del Departamento de Ciencias y Trastornos de la Comunicación en la Bowling Green State University en Bowling Green, en Ohio. Es licenciada en Inglés y Lingüística y posee certificación y licencia de patóloga del lenguaje. Sus intereses de investigación abarcan también la evaluación y la intervención lingüística.   IT Lynne Hewitt, Ph.D., CCC-SLP (Certificato di competenza clinica in patologie del linguaggio), è Professore Associato e capo del Dipartimento  di Scienze e Disordini della Comunicazione, alla Bowling Green State University sita a Bowling Green, in Ohio. Si è laurata in Inglese e Linguistica ed ha la certificazione e la licenza come patologa del linguaggio. I suoi interessi di ricerca riguardano anche la valutazione e l’intervento linguistici.   Laura C. Dilley, Department of Communicative Sciences and Disorders, Michigan State University ldilley@msu.edu EN Laura Dilley, Ph.D, is an Assistant Professor in the Department of Communicative Sciences and Disorders, Michigan State University. Her research is in the area of phonetics, psycholinguistics, and phonology, particularly with respect to speech prosody. ES Laura Dilley  Ph.D., es Profesora en el Departamento de Ciencias y Trastornos de la Comunicación en la Michigan State University. Su investigación se centra en el area de la fonética, psicolingüística y fonología, con especial interés para la prosodia de la lengua hablada.     IT Laura Dilley, Ph.D, è Ricercatrice confermata presso il Dipartimento di Scienze e Disordini della Comunicazione, alla Michigan State University. I suoi interessi di ricerca sono nell’area della fonetica, della psicolinguistica e della fonologia, con particolare enfasi sulla prosodia del parlato.