Microsoft Word - MartinezAdrian_7.1.docx


EuroAmerican	
  Journal	
  of	
  Applied	
  Linguistics	
  and	
  Languages	
  
Special	
  Issue,	
  Volume	
  7,	
  Issue	
  1,	
  February	
  2020,	
  59-­‐77	
  
ISSN	
  2376-­‐905X	
  	
  
DOI	
  http://dx.doi.org/10.21283/2376905X.11.191	
  
www.e-­‐journall.org	
  	
  

© Martínez-Adrián 2020. This work is licensed under a Creative Commons Attribution 4.0 International License. 

	
   	
  
	
   The	
  use	
  of	
  previously	
  known	
  languages	
  and	
  target	
  language	
  (English)	
  

during	
  task-­‐based	
  interaction:	
  A	
  pseudolongitudinal	
  study	
  of	
  	
  
primary-­‐school	
  CLIL	
  learners	
  

	
  
	
  

	
  
MARÍA	
  MARTÍNEZ-­‐ADRIÁN	
  

The	
  University	
  of	
  the	
  Basque	
  Country	
  
	
   	
  
	
   Received	
  3	
  May	
  2019;	
  received	
  in	
  revised	
  form	
  8	
  July	
  2019;	
  accepted	
  15	
  August	
  2019	
  

 
ABSTRACT 

EN (Pseudo)longitudinal studies of the use of previously known languages (PKL) and the target language (TL) during interaction 
are still scarce in Content and Language Integrated Learning (CLIL) contexts, particularly those with young learners. This 
paper examines the use of PKLs (Basque-Spanish) and the TL English in various categories (appeals for assistance-
clarification requests-metacomments-discourse markers-private speech) in two groups of CLIL learners (grade 5 and 6) during 
dyadic interaction. A greater use of PKLs in older learners was found, especially in less cooperative and more external to the 
task strategies. As regards TL use, both groups performed similarly, except for metacomments, where younger learners made 
a slightly higher use. When comparing PKLs to TL use, both grade learners resorted to Basque-Spanish more frequently than 
to English, except for metacomments. Learners seem to use their language repertoire for task-management purposes. 
Nonetheless, they still resort to PKLs to keep the flow of speech. 
 

Key words:	
  CLIL, USE OF PREVIOUSLY KNOWN LANGUAGES, TASK-BASED INTERACTION, PRIMARY-SCHOOL LEARNERS, L3 ENGLISH 
 

ES Los estudios pseudolongitudinales sobre el uso de lenguas previamente conocidas (LPCs) y de la lengua meta (LM) en tareas 
de interacción son aún escasos en contextos de Aprendizaje de Contenidos y Lenguas Extranjeras (AICLE), y más aún con 
jóvenes aprendices. Este trabajo examina el uso de LPCs (Euskera-Castellano) y la LM (Inglés) en varias categorías 
(peticiones de ayuda-peticiones de clarificación-metacomentarios- marcadores de discurso-habla privada) en dos grupos de 
aprendices AICLE (5º y 6º curso) durante una interacción diádica. Se encontró un uso mayor de LPCs en el grupo de mayor 
edad, especialmente en estrategias menos cooperativas y más externas a la tarea. Con respecto al uso de la LM, ambos 
grupos tuvieron resultados similares, excepto en los metacomentarios, donde los estudiantes más jóvenes la usaron 
ligeramente más. Al comparar LPCs con LM, los estudiantes de ambos cursos recurrieron a las LPCs con más frecuencia que 
al inglés, excepto en los metacomentarios. Los estudiantes parecen utilizar su repertorio lingüístico para fines de gestión de 
tareas, pero aún recurren a las LPCs para mantener el flujo del habla.  
 

Palabras clave: AICLE, USO DE IDIOMAS CONOCIDOS, TAREAS DE INTERACCIÓN, ESTUDIANTES DE EDUCACIÓN PRIMARIA, INGLÉS COMO L3 
 

IT Gli studi (pseudo)longitudinali sull'uso delle lingue già note e della lingua d’arrivo (TL) in interazione sono ancora scarsi nei 
contesti di apprendimento integrato di lingua e contenuto (CLIL), in particolare quelli con apprendenti di giovane età. Questo 
lavoro esamina l'uso di lingue già note (basco-spagnolo) e dell'inglese come TL in varie categorie (richieste di assistenza e di 
chiarimenti, metacommenti, marcatori discorsivi di discorsi privati) in due gruppi di studenti CLIL (classi di 5a elementare e 1a 
media) durante l'interazione diadica. È emerso un maggiore utilizzo delle lingue già note nei discenti più grandi, soprattutto 
nelle strategie meno collaborative e più esterne al task. In merito all’uso della lingua di arrivo, entrambi i gruppi hanno agito 
similmente, ad eccezione dei metacommenti, che gli studenti più giovani hanno usato leggermente di più. Quando si confronta 
l’uso delle lingue già note con quello della TL entrambi i gruppi fanno ricorso allo spagnolo-basco più frequentemente che 
all'inglese, ad eccezione dei metacommenti. Gli studenti sembrano usare il loro repertorio linguistico per la gestione dei task. 
Tuttavia, ricorrono ancora alle lingue già note per mantenere il flusso del discorso. 
 
Parole chiave: CLIL, USO DI LINGUE GIÀ NOTE, INTERAZIONE BASATA SUI TASK, STUDENTI DELLE SCUOLE ELEMENTARI, INGLESE COME L3 

	
  
	
  
* 	
  	
  María	
  Martínez-­‐Adrián,	
  The	
  University	
  of	
  the	
  Basque	
  Country	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  maria.martineza@ehu.eus	
  
	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 60	
  

1.	
  Introduction	
  
The	
   investigation	
   of	
   the	
   use	
   of	
   previously	
   known	
   languages	
   (henceforth	
   PKL)	
   during	
   task-­‐based	
  

interaction	
  has	
  attracted	
  the	
  attention	
  of	
  researchers	
  in	
  both	
  English	
  as	
  a	
  second	
  language	
  (ESL)	
  and	
  English	
  
as	
   a	
   foreign	
   language	
   (EFL)	
   settings.	
   Very	
   recently,	
   the	
   emergence	
   of	
   Content	
   and	
   Language	
   Integrated	
  
Learning	
   (CLIL)	
   programmes	
  has	
   led	
   to	
   a	
   growing	
  body	
   of	
   research	
   studies	
   analyzing	
   the	
   effect	
   of	
   a	
  more	
  
natural	
   and	
   intense	
   input	
   provided	
   in	
   these	
   settings	
   on	
   the	
   use	
   of	
   previously	
   known	
   language-­‐based	
  
strategies	
   (i.e.,	
  García	
  Mayo	
  &	
  Hidalgo	
  Gordo,	
  2017;	
  García	
  Mayo	
  &	
  Lázaro	
   Ibarrola,	
   2015;	
  Lázaro	
   Ibarrola,	
  
2016;	
  Martínez-­‐Adrián,	
  in	
  press;	
  Martínez-­‐Adrián	
  &	
  Gutiérrez-­‐Mangado,	
  2015a).	
  Most	
  of	
  these	
  investigations	
  
have	
  compared	
  CLIL	
  learners	
  to	
  non-­‐CLIL	
  learners,	
  but	
  (pseudo)longitudinal	
  studies	
  of	
  CLIL	
  learners	
  across	
  
age	
  and	
  proficiency	
  (Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  2018,	
  2019;	
  Lázaro	
  Ibarrola	
  &	
  García	
  Mayo,	
  2012),	
  
which	
  could	
  shed	
  more	
  light	
  on	
  the	
  development	
  of	
  strategy	
  use,	
  are	
  still	
  few	
  in	
  number	
  in	
  CLIL	
  settings,	
  and	
  
even	
  more	
  limited	
  with	
  young	
  leaners	
  (Azkarai	
  &	
  Imaz	
  Agirre,	
  2017).	
  In	
  addition,	
  a	
  more	
  detailed	
  analysis	
  of	
  
the	
  different	
  functions	
  and	
  patterns	
  of	
  both	
  PKL	
  and	
  target	
  language	
  (TL)	
  use	
  is	
  needed.	
  This	
  paper	
  will	
  try	
  to	
  
fill	
   these	
   gaps	
   by	
   examining	
   the	
   use	
   of	
   PKL	
   (Basque	
   and	
   Spanish)	
   and	
   the	
   TL	
   (English)	
   in	
   appeals	
   for	
  
assistance,	
   clarification	
   requests,	
   metacomments,	
   discourse	
   markers	
   and	
   private	
   speech	
   in	
   two	
   different	
  
age/proficiency	
   groups	
  of	
   primary	
   school	
   learners	
   (grade	
  5	
   and	
  6)	
   immersed	
   in	
   a	
   CLIL	
   context	
   during	
   the	
  
performance	
  of	
  a	
  communicative	
  task	
  in	
  dyads.	
  The	
  article	
  is	
  structured	
  as	
  follows:	
  the	
  first	
  section	
  presents	
  
an	
   overview	
   of	
   the	
   literature	
   as	
   regards	
   the	
   use	
   of	
   PKL	
   during	
   oral	
   production.	
   Research	
   questions	
   are	
  
subsequently	
  addressed	
  in	
  Section	
  2,	
  while	
  the	
  methodology	
  is	
  described	
  in	
  Section	
  3.	
  Results	
  are	
  presented	
  
and	
   discussed	
   in	
   Sections	
   4	
   and	
   5,	
   respectively.	
   The	
   last	
   section	
   concludes	
   the	
   paper	
   and	
   offers	
   future	
  
research	
  lines.	
  

	
  
2.	
  Literature	
  review	
  
2.1.	
  The	
  use	
  of	
  previously	
  known	
  languages	
  during	
  task-­‐based	
  interaction	
  

Research	
   on	
   task-­‐based	
   interaction	
   has	
   shown	
   that	
   the	
   use	
   of	
   PKL	
   during	
   the	
   performance	
   of	
  
communicative	
   tasks	
   is	
  more	
   the	
   norm	
   rather	
   than	
   the	
   exception	
   and	
   has	
   been	
   reported	
   to	
   have	
   positive	
  
effects	
  for	
  language	
  learning	
  (Storch	
  &	
  Aldosari,	
  2010).	
  To	
  the	
  present	
  date,	
  a	
  great	
  bulk	
  of	
  studies	
  in	
  ESL	
  and	
  
EFL	
  settings	
  with	
  mainstream	
  learners	
  has	
  been	
  conducted,	
  particularly	
  with	
  adults	
  (i.e.,	
  Alegría	
  de	
  la	
  Colina	
  
&	
   García	
   Mayo,	
   2009;	
   Azkarai,	
   2015;	
   Azkarai	
   &	
   García	
   Mayo,	
   2015;	
   Di	
   Camilla	
   &	
   Antón,	
   2012;	
   Storch	
   &	
  
Aldosari,	
  2010;	
  Swain	
  &	
  Lapkin,	
  2000).	
  These	
  studies	
  have	
  focused	
  on	
  the	
  extent	
  of	
  use	
  and	
  functions	
  of	
  PKL.	
  
In	
   terms	
   of	
   amount,	
   adults	
   have	
   not	
   been	
   found	
   to	
   draw	
   extensively	
   on	
   prior	
   linguistic	
   experience.	
   With	
  
regard	
   to	
   functions,	
   learners	
  mainly	
   employ	
   PKL	
   for	
   a	
  metacognitive	
   function,	
   phatics	
   and	
   for	
   vocabulary	
  
searches	
   (i.e.,	
   Alegría	
   de	
   la	
   Colina	
   &	
   García	
  Mayo,	
   2009;	
   Azkarai	
   &	
   García	
  Mayo,	
   2015;	
   Storch	
   &	
   Aldosari,	
  
2010).	
  The	
  extent	
  to	
  which	
  learners	
  rely	
  on	
  their	
  previous	
  linguistic	
  experience	
  seems	
  to	
  depend	
  on	
  a	
  variety	
  
of	
   factors	
   such	
   as	
   proficiency	
   (Di	
   Camilla	
  &	
  Antón,	
   2012;	
   Storch	
  &	
  Aldosari,	
   2010;	
   Swain	
  &	
  Lapkin,	
   2000),	
  
onset	
  age	
  (Cenoz,	
  2001,	
  2003),	
  gender	
  (Azkarai,	
  2015;	
  Ross-­‐Feldman,	
  2005),	
  task-­‐type	
  (Alegría	
  de	
  la	
  Colina	
  &	
  
García	
  Mayo,	
   2009;	
   Storch	
   &	
   Aldosari,	
   2010;	
   Storch	
   &	
  Wigglesworth,	
   2003)	
   and	
   task-­‐modality	
   (Azkarai	
   &	
  
García	
  Mayo,	
   2015).	
  More	
   limited	
   research	
   is	
   available	
   on	
   younger	
   learners	
   (Pinter,	
   2007;	
   Shintani,	
   2012,	
  
2014;	
  Tognini	
  &	
  Oliver,	
  2012).	
  In	
  particular,	
  studies	
  providing	
  a	
  detailed	
  examination	
  of	
  the	
  functions	
  of	
  PKL,	
  
as	
  documented	
  in	
  research	
  on	
  adult	
  learners,	
  are	
  nearly	
  non-­‐existent	
  for	
  younger	
  learners.	
  	
  
	
  
2.2.	
  Research	
  in	
  CLIL	
  settings	
  

The	
   use	
   of	
   PKL	
   when	
   performing	
   communicative	
   tasks	
   in	
   English	
   has	
   recently	
   been	
   the	
   focus	
   of	
  
several	
  studies	
  conducted	
  in	
  CLIL	
  settings.	
  In	
  CLIL	
  programmes,	
  more	
  intense	
  and	
  natural	
  input	
  is	
  provided	
  
to	
   students	
   as,	
   in	
   addition	
   to	
   English	
   as	
   a	
   school	
   subject,	
   they	
   receive	
   content	
   lessons	
   through	
   the	
   foreign	
  
language	
  (Coyle,	
  2007;	
  Dalton-­‐Puffer,	
  2011).	
  The	
  greater	
  amount	
  of	
  exposure	
  received	
  in	
  these	
  programmes	
  
leads	
   the	
   learner	
   to	
  attain	
  a	
  higher	
   command	
  of	
   the	
  TL	
   (Lasagabaster,	
  2008;	
  Martínez-­‐Adrián	
  &	
  Gutiérrez-­‐
Mangado,	
   2015b;	
   Navés	
   &	
   Victori,	
   2010;	
   Ruiz	
   de	
   Zarobe,	
   2008)	
   and	
   in	
   turn	
   to	
   a	
   lower	
   reliance	
   on	
   PKL	
  
(Gallardo-­‐del-­‐Puerto	
  &	
  Gómez-­‐Lacabex,	
  2013,	
  2017;	
  García	
  Mayo	
  &	
  Lázaro	
   Ibarrola,	
  2015;	
  Lázaro	
   Ibarrola,	
  
2016;	
  Martínez-­‐Adrián,	
  in	
  press;	
  Martínez-­‐Adrián	
  &	
  Gutiérrez-­‐Mangado,	
  2015a).	
  Some	
  of	
  these	
  investigations	
  
on	
   the	
   use	
   of	
   previously	
   known	
   language-­‐based	
   strategies	
   during	
   oral	
   production	
   have	
   compared	
   CLIL	
   to	
  
non-­‐CLIL	
   learners	
   (Gallardo	
   del	
   Puerto,	
   2015;	
   García	
  Mayo	
  &	
  Hidalgo	
   Gordo,	
   2017;	
   García	
  Mayo	
  &	
   Lázaro	
  
Ibarrola,	
   2015;	
   Lázaro	
   Ibarrola,	
   2016;	
   Martínez-­‐Adrián,	
   in	
   press;	
   Martínez-­‐Adrián	
   &	
   Gutiérrez-­‐Mangado,	
  
2015a;	
  Pladevall	
  Ballester	
  &	
  Vraciu,	
  2017)	
  and	
  others,	
  lower	
  in	
  number,	
  offer	
  (pseudo)longitudinal	
  studies	
  of	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 61	
  

CLIL	
  learners	
  (i.e.,	
  Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  2018,	
  2019;	
  Lázaro	
  Ibarrola	
  &	
  García	
  Mayo,	
  2012).	
  
Among	
   these	
   studies,	
   either	
   comparative	
   or	
   (pseudo)longitudinal	
   in	
   nature,	
   some	
   have	
   centered	
   on	
  
secondary-­‐school	
   learners	
   (i.e.,	
   Arratibel-­‐Irazusta	
   &	
   Martínez-­‐Adrián,	
   2018,	
   2019;	
   Lázaro	
   Ibarrola,	
   2016;	
  
Lázaro	
   Ibarrola	
  &	
  García	
  Mayo,	
  2012;	
  Martínez-­‐Adrián	
  &	
  Gutiérrez-­‐Mangado,	
  2015b)	
  while	
  primary-­‐school	
  
learners	
   are	
   currently	
   the	
   focus	
   of	
   several	
   investigations	
   (i.e.,	
   Azkarai	
   &	
   Imaz	
   Agirre,	
   2017;	
   Gallardo	
   del	
  
Puerto,	
  2015;	
  García	
  Mayo	
  &	
  Hidalgo	
  Gordo,	
  2017;	
  García	
  Mayo	
  &	
  Lázaro	
  Ibarrola,	
  2015;	
  Gutiérrez-­‐Mangado,	
  
2015;	
  Martínez-­‐Adrián,	
  in	
  press).	
  In	
  fact,	
  several	
  researchers	
  have	
  called	
  for	
  more	
  research	
  on	
  the	
  acquisition	
  
of	
  foreign	
  languages	
  by	
  primary-­‐school	
  learners	
  in	
  an	
  effort	
  to	
  maximize	
  their	
  learning	
  opportunities	
  (García	
  
Mayo,	
  2018).	
  CLIL	
   studies	
  on	
   the	
  use	
  of	
  PKL	
  during	
   interaction	
  have	
  mainly	
  examined	
  age	
  and	
  proficiency	
  
effects	
   (Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
   2018,	
   2019;	
   Lázaro	
   Ibarrola	
  &	
   García	
  Mayo,	
   2012;	
   Pladevall	
  
Ballester	
   &	
   Vraciu,	
   2017),	
   whereas	
   findings	
   for	
   gender	
   and	
   task	
   type	
   have	
   not	
   been	
   analysed	
   in	
   depth	
  
(Azkarai	
  &	
  García	
  Mayo,	
  2017;	
  Azkarai	
  &	
  Imaz	
  Agirre,	
  2017).	
  	
  

In	
  the	
  case	
  of	
  CLIL	
  studies	
  with	
  secondary-­‐school	
  learners,	
  Martínez-­‐Adrián	
  and	
  Gutiérrez-­‐Mangado	
  
(2015a)	
   compared	
   the	
   oral	
   production	
   of	
   a	
   group	
   of	
   CLIL	
   learners	
   to	
   a	
   group	
   of	
   non-­‐CLIL	
   learners.	
  
Participants	
  performed	
  an	
  oral	
  narration	
  task	
  in	
  front	
  of	
  a	
  researcher,	
  which	
  was	
  subsequently	
  examined	
  in	
  
terms	
   of	
   the	
   use	
   of	
   the	
   first	
   languages	
   (L1s)	
   (Basque	
   and	
   Spanish)	
   in	
   interactional	
   strategies	
   and	
   transfer	
  
lapses.	
  A	
   lower	
  use	
  of	
   the	
  L1s	
  and	
  a	
  greater	
  use	
  of	
   the	
  TL	
  (English)	
  was	
   found	
   in	
  the	
  case	
  of	
  CLIL	
   learners.	
  
This	
  study	
  also	
  confirmed	
  a	
  negative	
  correlation	
  between	
  general	
  proficiency,	
  lexical	
  richness	
  and	
  use	
  of	
  PKL.	
  
In	
   addition	
   to	
   the	
   use	
   of	
   PKL	
   in	
   interactional	
   strategies,	
   other	
   comparative	
   studies	
  with	
   secondary-­‐school	
  
learners	
  have	
  analysed	
  the	
  production	
  of	
  discourse	
  markers.	
  Lázaro	
  Ibarrola	
  (2016)	
  compared	
  a	
  group	
  of	
  15	
  
CLIL	
  learners	
  to	
  a	
  group	
  of	
  11	
  non-­‐CLIL	
  learners	
  at	
  Time	
  1	
  when	
  they	
  were	
  in	
  their	
  2nd	
  year	
  of	
  compulsory	
  
secondary	
  education	
  and	
  subsequently	
  at	
  Time	
  2	
  when	
  they	
  were	
  in	
  the	
  4th	
  year.	
  All	
  instances	
  of	
  discourse	
  
markers	
   including	
   those	
   in	
   English	
   and	
   those	
   in	
   Basque	
   and	
   Spanish	
   were	
   considered.	
   In	
   addition,	
   the	
  
learners’	
   use	
   of	
   other	
   hesitation	
   phenomena,	
   namely,	
   non-­‐lexical	
   discourse	
   markers	
   was	
   also	
   taken	
   into	
  
account.	
  The	
  use	
  of	
  the	
  L1s	
  (Basque	
  and	
  Spanish)	
  did	
  not	
  decrease	
  with	
  time	
  and	
  no	
  differences	
  were	
  found	
  
between	
  CLIL	
  and	
  non-­‐CLIL	
   (Basque	
  and	
  Spanish)	
   learners.	
  Few	
   instances	
  of	
  discourse	
  markers	
   in	
  English	
  
were	
  produced,	
   the	
  most	
  common	
  of	
   them	
  being	
   ‘well’,	
  which	
   is	
  employed	
   in	
   textbooks	
  and	
  teacher	
   talk.	
  A	
  
difference	
  was	
  found	
  between	
  CLIL	
  and	
  non-­‐CLIL	
  learners	
  in	
  the	
  use	
  of	
  non-­‐lexical	
  discourse	
  markers,	
  with	
  a	
  
higher	
  production	
  on	
  the	
  part	
  of	
  non-­‐CLIL	
  learners.	
  CLIL	
  students	
  were	
  able	
  to	
  use	
  discourse	
  markers	
  in	
  the	
  
L1s	
  while	
  non-­‐CLIL	
  learners	
  were	
  in	
  a	
  previous	
  stage	
  of	
  development	
  and	
  use	
  more	
  hesitation	
  phenomena.	
  	
  

Other	
  investigations	
  with	
  secondary	
  school	
   learners	
  are	
  (pseudo)longitudinal	
  studies	
  of	
  one	
  or	
  two	
  
groups	
   of	
   CLIL	
   learners.	
   Lázaro	
   Ibarrola	
   and	
   García	
   Mayo	
   (2012)	
   tackled	
   the	
   use	
   of	
   the	
   L1	
   in	
   discourse	
  
markers	
   and	
  appeals	
   for	
   assistance	
   as	
  well	
   as	
   the	
  morphosyntactic	
  development	
  of	
   a	
   group	
  of	
   15-­‐year-­‐old	
  
students	
  immersed	
  in	
  a	
  CLIL	
  context	
  in	
  the	
  Basque	
  Country.	
  The	
  use	
  of	
  the	
  L1	
  significantly	
  decreased	
  over	
  the	
  
two-­‐year	
   period	
   investigated,	
   while	
   morphosyntax	
   experienced	
   significant	
   development.	
   However,	
   the	
  
production	
   of	
   English	
   discourse	
   markers	
   was	
   uncommon.	
   Arratibel-­‐Irazusta	
   and	
   Martínez-­‐Adrián	
   (2018)	
  
explored	
  the	
  use	
  of	
  PKL	
  (L1	
  Spanish/second	
  language	
  (L2)	
  Basque)	
  in	
  appeals	
  for	
  assistance,	
  transfer	
  lapses,	
  
codes-­‐switching	
  and	
  discourse	
  markers	
  in	
  an	
  oral	
  narration	
  task	
  in	
  two	
  different	
  age/proficiency	
  CLIL	
  groups	
  
of	
  third	
  language	
  (L3)	
  English	
  learners.	
  Both	
  groups	
  shared	
  the	
  same	
  onset	
  age	
  (Year	
  3)	
  but	
  differed	
  in	
  length	
  
and	
  amount	
  of	
  exposure.	
  The	
  analysis	
  of	
  the	
  results	
  did	
  not	
  yield	
  statistically	
  significant	
  differences	
  between	
  
the	
   groups	
   except	
   for	
   the	
   production	
   of	
   foreignizings,	
   which	
   were	
   more	
   frequent	
   in	
   the	
   younger	
   group.	
  
However,	
  the	
  intragroup	
  analysis	
  revealed	
  how	
  less	
  proficient	
  learners	
  preferred	
  to	
  use	
  more	
  implicit	
  appeals	
  
and	
  employed	
  borrowings	
  and	
  foreignizings	
  in	
  similar	
  proportions,	
  while	
  older	
  learners	
  preferred	
  to	
  use	
  both	
  
explicit	
  and	
  implicit	
  strategies	
  and	
  showed	
  a	
  tendency	
  to	
  use	
  borrowings	
  (a	
  more	
  communicative	
  option).	
  As	
  
for	
   the	
   different	
   categories	
   examined,	
   discourse	
   markers	
   were	
   the	
   most	
   common	
   manifestation	
   of	
   prior	
  
linguistic	
   experience.	
   In	
   a	
   follow-­‐up	
   study	
   involving	
   the	
   same	
   groups	
   of	
   learners,	
   Arratibel-­‐Irazusta	
   and	
  
Martínez-­‐Adrián	
   (2019)	
  delved	
   into	
   the	
  use	
   of	
   previous	
   linguistic	
   knowledge	
   (L1	
   Spanish/L2	
  Basque)	
   as	
   a	
  
communication	
  strategy	
  (appeals	
  for	
  assistance	
  and	
  transfer)	
  together	
  with	
  TL-­‐based	
  strategies	
  (holistic	
  and	
  
analytic	
   strategies).	
   No	
   differences	
   between	
   the	
   two	
   age/proficiency	
   groups	
   were	
   observed.	
   As	
   for	
   the	
  
distribution	
  of	
  categories,	
  holistic	
  strategies	
  were	
  the	
  most	
  employed,	
  a	
  finding	
  attributed	
  to	
  the	
  overriding	
  
effect	
   of	
   CLIL	
   as	
   reported	
   in	
   other	
   investigations	
   with	
   young	
   learners	
   that	
   have	
   analysed	
   self-­‐reported	
  
opinions	
   (Martínez-­‐Adrián,	
   Gallardo-­‐del-­‐Puerto,	
   &	
   Basterrechea,	
   2019).	
   In	
   other	
   words,	
   even	
   if	
   these	
  
participants	
   were	
   considered	
   low-­‐proficient	
   learners,	
   they	
   were	
   exploiting	
   strategies	
   more	
   typical	
   of	
  
advanced	
  learners,	
  a	
  result	
  that	
  might	
  be	
  due	
  to	
  the	
  type	
  of	
  input	
  received	
  in	
  their	
  CLIL	
  lessons,	
  which	
  is	
  full	
  
of	
  reformulations	
  in	
  the	
  TL.	
  In	
  addition,	
  the	
  use	
  of	
  previously	
  known	
  language-­‐based	
  strategies	
  was	
  found	
  to	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 62	
  

correlate	
   negatively	
   with	
   both	
   the	
   level	
   of	
   proficiency	
   and	
   receptive	
   vocabulary.	
   However,	
   no	
   significant	
  
correlations	
  were	
   found	
   in	
  the	
  case	
  of	
  TL-­‐based	
  strategies.	
  Likewise,	
   the	
  analysis	
  did	
  not	
  yield	
  a	
  significant	
  
correlation	
  between	
   the	
  use	
  of	
  PKL	
  and	
   the	
  use	
  of	
  TL-­‐based	
  strategies.	
  Thus,	
   at	
   this	
  point	
  of	
  development,	
  
previously	
  known	
  language-­‐based	
  strategies	
  still	
  coexist	
  with	
  the	
  use	
  of	
  TL-­‐based	
  strategies.	
  

As	
  regards	
  CLIL	
  studies	
  conducted	
  with	
  young	
  learners,	
  most	
  of	
  them	
  have	
  examined	
  the	
  effect	
  of	
  the	
  
instructional	
   context	
   (CLIL	
   vs.	
   Non-­‐CLIL).	
   These	
   studies,	
   with	
   a	
   cross-­‐sectional	
   or	
   (pseudo)longitudinal	
  
design,	
  have	
  compared	
  CLIL	
  to	
  non-­‐CLIL	
  learners	
  (Gallardo-­‐del-­‐Puerto,	
  2015;	
  García	
  Mayo	
  &	
  Lázaro	
  Ibarrola,	
  
2015;	
   García	
   Mayo	
   &	
   Hidalgo	
   Gordo,	
   2017;	
   García	
   Mayo	
   &	
   Imaz	
   Agirre,	
   2017;	
   Gutiérrez-­‐Mangado,	
   2015;	
  
Martínez-­‐Adrián,	
  in	
  press;	
  Pladevall	
  Ballester	
  &	
  Vraciu,	
  2017).	
  	
  

Gallardo-­‐del-­‐Puerto	
   (2015)	
   looked	
   into	
   data	
   from	
   primary	
   school	
   learners	
   performing	
   a	
   peer	
  
interaction	
  task.	
  Two	
  groups	
  of	
  CLIL	
  learners	
  in	
  grades	
  4	
  and	
  6	
  were	
  compared	
  to	
  two	
  age-­‐matched	
  non-­‐CLIL	
  
learners	
   in	
   terms	
   of	
   their	
   production	
   of	
   codeswitching	
   and	
   transfer	
   lapses	
   in	
   a	
   dyadic	
   story-­‐telling	
   task.	
  
Younger	
  and	
  older	
  CLIL	
   learners	
  produced	
   fewer	
   instances	
  of	
   codeswitching	
  and	
   transfer	
   lapses	
   than	
   their	
  
non-­‐CLIL	
  counterparts,	
   these	
  differences	
  reaching	
  significance	
  only	
   in	
   the	
  case	
  of	
   the	
  older	
   learners.	
  As	
   for	
  
the	
   distribution	
   of	
   functions,	
   codeswitching,	
   borrowings	
   and	
   foreignizings	
  were	
   far	
  more	
   frequent	
   in	
   non-­‐
CLIL	
  than	
  in	
  CLIL,	
  while	
  CLIL	
  learners	
  produced	
  more	
  calques	
  than	
  non-­‐CLIL	
  students.	
  This	
  study	
  contrasts	
  
with	
  previous	
   investigations	
   that	
  claimed	
   that	
   foreignizings	
  are	
  characteristic	
  of	
  higher	
  proficiency	
  (=CLIL)	
  
learners	
  (Agustín	
  Llach,	
  2014;	
  Celaya,	
  2008;	
  Celaya	
  &	
  Ruiz	
  de	
  Zarobe,	
  2010).	
  

Martínez-­‐Adrián	
   (in	
   press)	
   explored	
   the	
   use	
   of	
   interactional	
   strategies	
   (appeals	
   for	
   assistance,	
  
clarification	
   requests	
   and	
   metacomments)	
   in	
   the	
   L1	
   (Spanish)	
   and	
   the	
   TL	
   (English)	
   in	
   the	
   same	
   sample	
  
analyzed	
   by	
   Gallardo-­‐del-­‐Puerto	
   (2015).	
   The	
   results	
   indicated	
   that	
   NON-­‐CLIL	
   learners	
   produced	
   more	
  
instances	
  of	
  L1	
  use	
  in	
  interactional	
  strategies	
  and	
  that	
  greater	
  differences	
  emerged	
  as	
  grade	
  level	
  increased.	
  
In	
   addition,	
  metacomments	
  were	
   the	
  most	
   common	
  manifestation	
   of	
   L1	
  use	
   in	
   both	
   years	
   in	
   the	
  non-­‐CLIL	
  
groups,	
  which	
  suggested	
  that	
  non-­‐CLIL	
  learners	
  were	
  not	
  so	
  accustomed	
  to	
  interacting	
  in	
  the	
  TL	
  and	
  not	
  so	
  
well	
   equipped	
   to	
   perform	
   content	
   tasks	
   in	
   the	
  TL.	
   A	
   qualitative	
   inspection	
   of	
   the	
   results	
   also	
   showed	
   that	
  
some	
  differences	
  between	
  groups	
  in	
  their	
  preference	
  for	
  either	
  the	
  L1	
  or	
  the	
  TL	
  existed	
  in	
  the	
  case	
  of	
  appeals	
  
and	
   clarification	
   requests.	
  Nevertheless,	
  metacomments	
  were	
   always	
  produced	
   in	
   the	
   L1	
   in	
   both	
   groups,	
   a	
  
finding	
  in	
  line	
  with	
  previous	
  classroom	
  observation	
  data	
  (Gené	
  Gil,	
  Juan	
  Garau,	
  &	
  Salazar	
  Noguera,	
  2012).	
  

Using	
  the	
  same	
  sample	
  as	
   in	
  Gallardo-­‐del-­‐Puerto	
  (2015)	
  and	
  Martínez-­‐Adrián	
  (in	
  press),	
  Gutiérrez-­‐
Mangado	
  (2015)	
  investigated	
  the	
  pseudolongitudinal	
  development	
  of	
  CLIL	
  and	
  non-­‐CLIL	
  learners	
  as	
  regards	
  
L1	
  use	
  in	
  interactional	
  strategies,	
  transfer	
  lapses	
  and	
  codeswitching.	
  Non-­‐CLIL	
  learners	
  used	
  the	
  L1	
  in	
  these	
  
three	
  categories	
  in	
  the	
  two-­‐year	
  period	
  examined,	
  while	
  a	
  decrease	
  in	
  appeals	
  for	
  assistance	
  and	
  an	
  increase	
  
in	
  borrowings	
  was	
  attested	
  in	
  the	
  CLIL	
  learners.	
  The	
  analysis	
  of	
  the	
  data	
  indicated	
  that	
  older	
  learners	
  seemed	
  
to	
  favour	
  more	
  uncooperative	
  strategies	
  rather	
  than	
  more	
  cooperative	
  ones	
  (i.e.,	
  appeals).	
  

García	
  Mayo	
  and	
  Lázaro	
  Ibarrola	
  (2015)	
  studied	
  the	
  effect	
  of	
  CLIL	
  and	
  age	
  in	
  the	
  oral	
  interaction	
  of	
  
40	
  age-­‐	
  and	
  proficiency-­‐matched	
  dyads	
  in	
  terms	
  of	
  conversational	
  adjustments,	
  repetitions	
  and	
  L1	
  use.	
  CLIL	
  
learners	
  displayed	
   slightly	
  higher	
   linguistic	
   abilities,	
   a	
   finding	
   supporting	
  previous	
   studies	
   comparing	
  CLIL	
  
and	
  non-­‐CLIL	
   learners	
  and	
  were	
   found	
   to	
  negotiate	
  more	
  and	
   to	
   resort	
   to	
   the	
  L1	
   less	
   frequently	
   than	
  non-­‐
CLIL	
  learners.	
  As	
  for	
  the	
  effect	
  of	
  age,	
  older	
  children	
  in	
  both	
  contexts	
  were	
  reported	
  to	
  negotiate	
  less	
  and	
  to	
  
use	
  the	
  L1	
  more	
  frequently	
  than	
  younger	
  children.	
  

García	
  Mayo	
  and	
  Hidalgo	
  Gordo	
   (2017)	
   compared	
  a	
  group	
  of	
  CLIL	
   learners	
   to	
  a	
  group	
  of	
   non-­‐CLIL	
  
learners	
  at	
  Time	
  1,	
  when	
  they	
  were	
   in	
  the	
  3rd	
  year	
  of	
  primary	
  education	
  and	
  at	
  Time	
  2,	
  when	
  they	
  were	
   in	
  
their	
  4th	
  year.	
  Both	
  groups	
  completed	
  a	
  jigsaw	
  task	
  that	
  was	
  subsequently	
  codified	
  according	
  to	
  the	
  following	
  
L1	
  functions:	
  metacognitive	
  talk,	
  vocabulary	
  (appeals,	
  borrowings	
  and	
  foreignizings)	
  and	
  discourse	
  markers.	
  
The	
  examination	
  of	
  the	
  results	
  revealed	
  that	
  L1	
  use	
  as	
  a	
  whole	
  was	
  quite	
  limited	
  at	
  both	
  testing	
  times,	
  but	
  a	
  
greater	
  use	
  was	
  observed	
  at	
  Time	
  2	
  in	
  both	
  groups,	
  even	
  though	
  it	
  only	
  reached	
  significance	
  in	
  the	
  non-­‐CLIL	
  
group.	
  Vocabulary	
  was	
  the	
  function	
  most	
  served	
  by	
  the	
  L1	
  in	
  both	
  groups	
  in	
  the	
  two	
  school	
  years,	
  a	
  finding	
  in	
  
line	
  with	
  other	
  investigations	
  exploring	
  young	
  learners	
  in	
  a	
  CLIL	
  setting	
  (Azkarai	
  &	
  García	
  Mayo,	
  2017).	
  As	
  for	
  
the	
   effect	
   of	
   the	
   instructional	
   setting,	
   non-­‐CLIL	
   learners	
   relied	
   on	
   the	
   L1	
   to	
   a	
   greater	
   extent	
   than	
   CLIL	
  
learners,	
   especially	
   in	
   the	
   case	
   of	
   learners’	
   use	
   of	
   the	
   L1	
   for	
   vocabulary	
   purposes.	
   When	
   dealing	
   with	
  
metacognitive	
  talk	
  and	
  discourse	
  markers,	
  the	
  instructional	
  setting	
  did	
  not	
  have	
  a	
  great	
  influence.	
  As	
  regards	
  
the	
  effect	
  of	
  age	
  on	
  functions,	
  only	
  a	
  statistically	
  significant	
  difference	
  was	
  found	
  in	
  the	
  metacognitive	
  use	
  in	
  
favour	
  of	
  the	
  older	
  non-­‐CLIL	
  group.	
  	
  

These	
  results	
  are	
  partially	
  in	
  line	
  with	
  García	
  Mayo	
  and	
  Imaz	
  Agirre	
  (2017)	
  who	
  also	
  considered	
  the	
  
impact	
  of	
   the	
   learning	
  context	
   (CLIL	
  vs.	
  non-­‐CLIL)	
  on	
  L1	
  use	
  adopting	
  a	
   longitudinal	
  perspective.	
  27	
  dyads	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 63	
  

participated	
  in	
  the	
  study.	
  In	
  particular,	
  two	
  groups	
  of	
  CLIL	
  learners	
  were	
  compared	
  to	
  two	
  groups	
  of	
  non-­‐CLIL	
  
learners	
   at	
   two	
   different	
   data	
   collection	
   times.	
   In	
   addition,	
   younger	
   learners	
  were	
   also	
   compared	
   to	
   older	
  
learners	
  in	
  each	
  instructional	
  setting	
  in	
  order	
  to	
  examine	
  the	
  effect	
  of	
  age.	
  At	
  two	
  data	
  collection	
  times,	
  non-­‐
CLIL	
  learners	
  were	
  found	
  to	
  use	
  the	
  L1	
  to	
  a	
  greater	
  extent	
  than	
  their	
  CLIL	
  counterparts.	
  However,	
  while	
  the	
  
use	
  of	
  the	
  L1	
  decreased	
  in	
  the	
  non-­‐CLIL	
  groups	
  over	
  time,	
  in	
  the	
  CLIL	
  groups	
  it	
  remained	
  stable	
  across	
  time	
  in	
  
the	
  younger	
  group	
  of	
  learners,	
  while	
  an	
  increase	
  was	
  observed	
  in	
  the	
  group	
  of	
  older	
  CLIL	
  learners.	
  The	
  nature	
  
of	
  the	
  task,	
  which	
  could	
  have	
  been	
  not	
  so	
  motivating	
  for	
  students,	
  could	
  have	
  yielded	
  these	
  results	
  in	
  the	
  CLIL	
  
group.	
   As	
   regards	
   the	
   effect	
   of	
   age,	
   at	
   Time	
  1,	
   older	
   non-­‐CLIL	
   learners	
  were	
   significantly	
   found	
   to	
  make	
   a	
  
greater	
  use	
  of	
   the	
  L1.	
  At	
  Time	
  2	
  older	
   learners	
   in	
  both	
   instructional	
   settings	
  employed	
   the	
  L1	
   to	
  a	
   greater	
  
extent	
  with	
  significant	
  differences.	
  	
  

Other	
   longitudinal	
  studies	
  comparing	
  CLIL	
  to	
  non-­‐CLIL	
   learners	
  during	
  task-­‐based	
   interaction	
  have	
  
controlled	
  for	
  amount	
  of	
  exposure,	
  a	
  factor	
  that	
  could	
  explain	
  the	
  existence	
  of	
  some	
  contradictory	
  findings	
  in	
  
the	
   literature.	
  Pladevall	
  Ballester	
  and	
  Vraciu	
  (2017)	
  examined	
  L1	
  use	
   in	
   the	
  oral	
  production	
  of	
  5th	
  and	
  6th	
  
grade	
   primary	
   school	
   learners	
   over	
   a	
   period	
   of	
   two	
   academic	
   years.	
   Content	
   and	
   function	
   words,	
  
codeswitching,	
   lexical	
   transfer	
   (i.e.,	
   borrowings,	
   foreignizings)	
   and	
   interactional	
   strategies	
   were	
   codified.	
  
Interactional	
   strategies	
  were	
   further	
   broken	
   down	
   into	
  metacognitive	
   strategies,	
  metatalk	
   strategies,	
   task-­‐
related	
  interactional	
  strategies	
  and	
  private	
  speech.	
  The	
  examination	
  of	
  the	
  results	
  indicated	
  a	
  decrease	
  in	
  L1	
  
use	
   as	
   L2	
   proficiency	
   increased	
   in	
   both	
   CLIL	
   and	
   non-­‐CLIL	
   learners.	
   Primary	
   school	
   learners	
   were	
   also	
  
reported	
  to	
  make	
  use	
  of	
  their	
  L1	
  as	
  a	
  compensatory	
  strategy	
  during	
  L2	
  production	
  irrespective	
  of	
  the	
  type	
  of	
  
instruction	
   (CLIL	
   vs.	
   non-­‐CLIL)	
   received.	
   In	
   terms	
   of	
   interactional	
   strategies,	
   metatalk	
   and	
   private	
   speech	
  
were	
  the	
  most	
  common	
  functions	
  in	
  both	
  groups	
  at	
  the	
  beginning	
  of	
  the	
  study,	
  while,	
  at	
  the	
  very	
  end,	
  the	
  L1	
  
preserved	
   its	
   regulatory	
   function	
   but	
   its	
   scaffolding	
   role	
   disappeared.	
   In	
   the	
   case	
   of	
   lexical	
   transfer,	
  
borrowings	
   and	
   foreignizings	
   decreased	
   with	
   proficiency	
   in	
   both	
   groups	
   and	
   the	
   same	
   applied	
   to	
  
codeswitching.	
  	
  

Apart	
   from	
   these	
   investigations	
   that	
   have	
   compared	
   CLIL	
   to	
   non-­‐CLIL	
   learners,	
   others	
   are	
  
pseudolongitudinal	
   studies	
   of	
   different	
   age/proficiency	
   groups	
   of	
   young	
   CLIL	
   learners	
   (Azkarai	
   &	
   Imaz	
  
Agirre,	
   2017),	
   though	
   limited	
   in	
   number	
   and	
   examining	
   the	
   interaction	
   of	
   age	
   with	
   other	
   variables	
   (i.e.	
  
gender).	
  This	
  line	
  of	
  research	
  is	
  particularly	
  interesting	
  as	
  it	
  could	
  shed	
  more	
  light	
  on	
  the	
  development	
  of	
  the	
  
use	
  of	
  PKL	
  during	
  task-­‐based	
  interaction	
  across	
  age.	
  Azkarai	
  and	
  Imaz	
  Agirre	
  (2017)	
  examined	
  gender	
  effects	
  
on	
   the	
   amount	
  of	
   L1	
  use	
   in	
  Grade	
  3	
   and	
  4	
  primary-­‐school	
   learners	
  while	
  performing	
   a	
   spot-­‐the-­‐difference	
  
task.	
  Boys	
  were	
  found	
  to	
  make	
  use	
  of	
  the	
  L1	
  to	
  a	
  greater	
  extent	
  than	
  girls,	
  and	
  this	
  difference	
  was	
  even	
  more	
  
marked	
  in	
  Grade	
  4	
  learners.	
  Boys	
  seemed	
  to	
  employ	
  the	
  L1	
  to	
  avoid	
  communication	
  breakdowns,	
  while	
  girls	
  
showed	
  a	
  more	
  restricted	
  use	
  of	
  the	
  L1.	
  

The	
  overview	
  of	
  the	
  literature	
  on	
  L1	
  use	
  during	
  task-­‐based	
  interaction	
  has	
  shown	
  a	
  growing	
  body	
  of	
  
research	
  in	
  this	
  area.	
  However,	
  it	
  has	
  uncovered	
  several	
  gaps	
  that	
  still	
  need	
  to	
  be	
  filled.	
  First,	
  studies	
  aimed	
  at	
  
the	
   analysis	
   of	
   the	
   development	
   of	
   use	
   of	
   PKL	
   across	
   age	
   are	
   needed.	
   Second,	
   a	
   detailed	
   analysis	
   of	
   the	
  
different	
  functions	
  and	
  most	
  common	
  manifestations	
  of	
  PKL	
  and	
  TL	
  use	
  is	
  advocated,	
  as	
  to	
  the	
  knowledge	
  of	
  
the	
  author,	
   just	
  one	
   investigation	
   involving	
  younger	
   learners	
  has	
  been	
  conducted	
   in	
   this	
  respect	
  (Martínez-­‐
Adrián,	
  in	
  press).	
  Thus,	
  the	
  present	
  article	
  will	
  contribute	
  to	
  these	
  gaps	
  by	
  examining	
  the	
  use	
  of	
  PKL	
  (Basque	
  
and	
  Spanish)	
  and	
  the	
  TL	
  (English)	
  in	
  appeals	
  for	
  assistance,	
  clarification	
  requests,	
  metacomments,	
  discourse	
  
markers	
  and	
  private	
  speech	
  in	
  two	
  different	
  age/proficiency	
  groups	
  of	
  primary	
  school	
  learners	
  (Grades	
  5	
  and	
  
6)	
   immersed	
   in	
   a	
   CLIL	
   context,	
   while	
   performing	
   a	
   communicative	
   task	
   in	
   pairs.	
   The	
   following	
   research	
  
questions	
  will	
  be	
  addressed:	
  

	
  
-­‐ RQ	
   1.	
   Are	
   there	
   any	
   differences	
   between	
   both	
   age/proficiency	
   groups	
   in	
   terms	
   of	
   the	
   amount	
   of	
  

Basque/Spanish	
  use?	
  
-­‐ RQ2.	
   Are	
   there	
   any	
   differences	
   between	
   both	
   age/proficiency	
   groups	
   in	
   terms	
   of	
   the	
   amount	
   of	
  

English	
  use?	
  
-­‐ RQ3.	
   Do	
   learners	
   in	
   each	
   group	
   show	
   a	
   preference	
   for	
   either	
   Basque/Spanish	
   or	
   English	
   in	
   the	
  

categories	
  analyzed?	
  
-­‐ RQ4.	
   Among	
   the	
   categories	
   examined,	
   which	
   ones	
   are	
   the	
   most	
   commonly	
   served	
   by	
  

Basque/Spanish?	
  	
  
-­‐ RQ5.	
  Among	
  the	
  categories	
  examined,	
  which	
  ones	
  are	
  the	
  most	
  commonly	
  served	
  by	
  English?	
  	
  
	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 64	
  

3.	
  Methodology	
  
3.1.	
  Participants	
  

The	
   study	
   was	
   carried	
   out	
   in	
   a	
   CLIL	
   context	
   in	
   a	
   mid-­‐sized	
   town	
   in	
   the	
   Basque-­‐Autonomous	
  
Community	
  where	
  Basque	
  and	
  Spanish	
  co-­‐exist	
  as	
  official	
  languages.	
  90	
  Basque-­‐Spanish	
  bilingual	
  learners	
  of	
  
L3	
  English	
  from	
  a	
  semi-­‐private	
  school	
  participated	
  in	
  the	
  study.	
  All	
  of	
  them	
  had	
  been	
  exposed	
  to	
  Spanish	
  and	
  
Basque	
   since	
   birth	
   and/or	
   early	
   childhood	
   in	
   the	
   school-­‐context1.	
   They	
   were	
   enrolled	
   in	
   an	
   early	
   partial	
  
immersion	
  programme	
  in	
  which	
  half	
  of	
  the	
  subjects	
  are	
  taught	
  in	
  Spanish	
  and	
  the	
  other	
  half	
  in	
  Basque,	
  except	
  
for	
  the	
  English	
  as	
  a	
  foreign	
  language	
  class	
  and	
  some	
  content	
  subjects	
  that	
  are	
  in	
  English.	
  	
  

All	
  participants	
  started	
  learning	
  English	
  at	
  age	
  4	
  and	
  were	
  enrolled	
  in	
  a	
  CLIL	
  programme	
  at	
  age	
  8.	
  In	
  
fifth-­‐	
  and	
  sixth-­‐years	
  of	
  schooling,	
  learners	
  are	
  exposed	
  to	
  English	
  in	
  the	
  classroom	
  context	
  for	
  5	
  to	
  7	
  weekly	
  
hours	
  in	
  lessons	
  scheduled	
  as	
  English	
  as	
  a	
  school	
  subject	
  (3	
  hours	
  in	
  5th	
  year	
  and	
  4	
  hours	
  in	
  6th	
  year),	
  as	
  well	
  
as	
  in	
  content	
  lessons	
  such	
  as	
  science,	
  arts	
  and	
  crafts	
  or	
  physical	
  education	
  (3	
  hours	
  in	
  5th	
  year	
  and	
  4	
  hours	
  in	
  
6th	
   year).	
   At	
   the	
   time	
   of	
   the	
   data	
   gathering,	
   5th	
   year	
   learners	
   had	
   received	
   714	
   hours	
   of	
   exposure,	
   and	
   6th	
  
graders	
  884	
  hours.	
  As	
  for	
  their	
  English	
  proficiency,	
  the	
  proficiency	
  level	
  test	
  administered	
  at	
  the	
  outset	
  of	
  the	
  
study	
  indicated	
  that	
  5th	
  year	
  learners	
  were	
  in	
  A2-­‐,	
  and	
  6th	
  year	
  ones	
  in	
  the	
  A2	
  level	
  (CEFR;	
  Council	
  of	
  Europe,	
  
2001).	
  Participant	
  characteristics	
  are	
  summarized	
  in	
  Table	
  1.	
  
 

Table 1 
Participant characteristics 

 Learner groups 
CLIL 5 (n = 50) CLIL 6 (n = 40) 

Grade  5 6 
Age (years) 10 11 
Onset age  3.93 4.28 
Exposure (academic years)  6.5 7.5 
Hours of exposure to English  714 884 
Hours of CLIL exposure  162 291 
Proficiency level A2- A2 

 
3.2.	
  Instruments	
  and	
  data	
  analysis	
  

At	
   the	
   beginning	
   of	
   the	
   study,	
   learners	
   completed	
   both	
   a	
   questionnaire	
   on	
   their	
   personal	
   and	
  
linguistic	
  background	
  (languages	
  known,	
  amount	
  of	
  time	
  learning	
  English	
  (at	
  school	
  and/or	
  extramurally)	
  as	
  
well	
   as	
   a	
   proficiency	
   level	
   test,	
   consisting	
   of	
   the	
   reading,	
   listening	
   and	
   writing	
   sections	
   of	
   the	
   Cambridge	
  
English	
   Flyers	
   (see	
   http://www.cambridgeenglish.org/exams/young-­‐learners-­‐english/flyers/testformat/).	
  
Subsequently,	
  students	
  worked	
  in	
  pairs	
  to	
  narrate	
  a	
  story	
  in	
  English	
  with	
  visual	
  support	
  provided	
  by	
  a	
  series	
  
of	
  wordless	
  pictures.	
  This	
   task	
   type	
  has	
  been	
  widely	
  used	
   for	
  similar	
  research	
  purposes	
  with	
  primary-­‐	
  and	
  
secondary-­‐school	
   children	
   in	
   CLIL	
   and	
   non-­‐CLIL	
   settings	
   (i.e.,	
   Arratibel-­‐Irazusta	
   &	
  Martínez-­‐Adrián,	
   2018;	
  
García	
   Mayo	
   &	
   Lázaro	
   Ibarrola,	
   2015;	
   Lázaro	
   Ibarrola	
   &	
   García	
   Mayo,	
   2012;	
   Martínez-­‐Adrián,	
   in	
   press;	
  
Martínez-­‐Adrián	
  &	
  Gutiérrez-­‐Mangado,	
  2015a,	
  2015b).	
  These	
  vignettes	
  were	
  selected	
  from	
  the	
  Disney	
  movie	
  
Aladdin	
   (1992)2.	
   As	
   for	
   the	
   procedure,	
   12	
   pictures	
   were	
   distributed	
   to	
   the	
   5th	
   and	
   6th	
   graders	
   (6	
   to	
   each	
  
member	
  of	
  the	
  dyad).	
  At	
  the	
  beginning	
  of	
  the	
  task,	
  participants	
  could	
  not	
  see	
  each	
  other.	
  Each	
  dyad	
  member	
  
had	
   to	
   describe	
   the	
   pictures	
   assigned,	
   which	
   had	
   been	
   randomly	
   presented.	
   Then,	
   working	
   together,	
   they	
  
ordered	
   the	
   image	
   and	
   took	
   turns	
   telling	
   the	
   story	
   according	
   to	
   the	
   order	
   that	
   they	
  had	
   agreed	
  upon.	
   The	
  
participants	
  were	
  asked	
  to	
  perform	
  the	
  task	
  naturally,	
  drawing	
  on	
  all	
  of	
  their	
  available	
  resources,	
  without	
  the	
  
assistant’s	
  help	
  for	
  vocabulary	
  queries.	
  Even	
  if	
  this	
  type	
  of	
  task	
  has	
  been	
  found	
  to	
  be	
  achievable	
  and	
  beneficial	
  
for	
   middle	
   school	
   children	
   (see	
   García	
   Mayo,	
   2018	
   for	
   a	
   review	
   of	
   empirical	
   findings	
   related	
   to	
   child	
  
interaction	
  in	
  task-­‐based	
  EFL	
  contexts)	
  and	
  have	
  been	
  used	
  in	
  a	
  wide	
  range	
  of	
  studies	
  with	
  young	
  children	
  in	
  
an	
   EFL	
   context	
   (i.e.,	
   García	
   Mayo	
   &	
   Lázaro	
   Ibarrola,	
   2015;	
   García	
   Mayo	
   &	
   Hidalgo	
   Gordo,	
   2017;	
   Lázaro	
  
Ibarrola	
  &	
  Azpilicueta	
  Martínez,	
  2015;	
  Martínez-­‐Adrián,	
  in	
  press),	
  prior	
  to	
  the	
  administration	
  of	
  the	
  task,	
  the	
  
researchers	
  ensure	
  that	
  the	
  task	
  was	
  appropriate	
  for	
  the	
  learners	
  tested	
  by	
  having	
  several	
  meetings	
  with	
  the	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
1	
  Learners	
  had	
  one	
  of	
  three	
  profiles:	
  L1	
  Spanish/L2	
  Basque,	
  L1	
  Basque/L2	
  Spanish,	
  L1	
  Spanish/L1	
  Basque.	
  
2	
  To	
  avoid	
  one	
  student	
  in	
  the	
  dyad	
  taking	
  on	
  a	
  lead	
  role,	
  pairings	
  were	
  made	
  based	
  on	
  each	
  member’s	
  similar	
  knowledge	
  
about	
  the	
  Aladdin	
  story.	
  	
  	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 65	
  

school	
  teachers.	
  In	
  addition,	
  it	
  was	
  piloted	
  with	
  similarly	
  aged	
  children	
  so	
  as	
  to	
  detect	
  potential	
  problems	
  and	
  
make	
  modifications	
  accordingly.	
  	
  

All	
   the	
   narrations	
   were	
   videotaped	
   and	
   later	
   on	
   transcribed	
   and	
   codified	
   in	
   CHILDES	
   format	
  
(MacWhinney,	
  2000).	
  The	
  oral	
  data	
  was	
  codified	
  in	
  terms	
  of	
  Basque/Spanish	
  use	
  and	
  English	
  use	
  in	
  appeals	
  
for	
   assistance,	
   clarification	
   requests,	
   metacomments	
   and	
   discourse	
   markers,	
   following	
   the	
   categorization	
  
from	
  our	
  previous	
  studies	
  (see	
  Martínez-­‐Adrián,	
  in	
  press,	
  with	
  primary-­‐school	
  children;	
  Arratibel-­‐Irazusta	
  &	
  
Martínez-­‐Adrián,	
  2018,	
  2019,	
  with	
  secondary-­‐school	
  learners).	
  Apart	
  from	
  that,	
  the	
  use	
  of	
  private	
  speech	
  has	
  
been	
   incorporated	
   in	
   the	
   present	
   study	
   in	
   line	
   with	
   other	
   investigations	
   conducted	
   with	
   child	
   learners	
  
(Pladevall	
  Ballester	
  &	
  Vraciu,	
  2017).	
  In	
  what	
  follows,	
  a	
  description	
  of	
  the	
  categories	
  along	
  with	
  illustrations	
  
from	
  our	
  database	
  will	
  be	
  provided.	
  

	
  
3.2.1.	
  Appeals	
  for	
  assistance	
  

Appeals	
   for	
   assistance	
   (also	
  known	
  as	
   interactional	
   strategies,	
   Cenoz,	
   2003)	
   are	
   “direct	
   or	
   indirect	
  
appeals	
  to	
  the	
  interlocutor	
  in	
  order	
  to	
  get	
  help	
  to	
  produce	
  a	
  specific	
  term	
  in	
  English	
  .	
   .	
   .	
  they	
  are	
  considered	
  
intentional	
  and	
  present	
  a	
  marked	
  interrogative	
  intonation	
  pattern”	
  (Cenoz,	
  2003,	
  p.	
  5).	
  Example	
  1	
  illustrates	
  
an	
  appeal	
  for	
  assistance	
  in	
  Spanish	
  and	
  Example	
  2	
  in	
  English:	
  
	
  
1)	
   CH2:	
   ¿Cómo	
  	
   se	
  dice	
  	
  	
  	
  	
  	
  	
  	
  	
  señor?	
  

How	
  	
   do	
  you	
  say	
  man	
  
How	
  	
   do	
  you	
  say	
  a	
  man?	
  
	
  

2)	
   CH2:	
   How	
  do	
  you	
  say	
  a	
  magical	
  thing?	
  [meaning	
  lamp]	
  
	
  
3.2.2.	
  Clarification	
  requests	
  

These	
   are	
   cases	
   in	
   which	
   the	
   learner	
   elicits	
   clarification	
   (Muñoz,	
   2007).	
   Example	
   3	
   provides	
   an	
  
example	
  of	
  clarification	
  in	
  Basque	
  and	
  Example	
  4	
  in	
  English:	
  
 
3)	
   CH2:	
   eh	
  #	
  	
   ba	
  ###	
  	
  hau	
  

	
   then	
  	
  	
  	
  	
  	
  this	
  
	
   so,	
  this?	
  
	
  

4)	
   CH1:	
   All?	
  
	
  
3.2.3.	
  Metacomments	
  

These	
  are	
  instances	
  in	
  which	
  the	
  learner	
  comments	
  on	
  the	
  communicative	
  situation	
  (Muñoz,	
  2007),	
  
as	
  in	
  Example	
  5	
  (use	
  of	
  Spanish)	
  and	
  Example	
  6	
  (use	
  of	
  English):	
  

	
  
5)	
   CH1:	
   no	
  sé	
  esta	
  misma	
  es	
  que	
  yo	
  no	
  me	
  lo	
  he	
  leído.	
  

I	
  don’t	
  know	
  this	
  one	
  is	
  that	
  I	
  have	
  not	
  read	
  it	
  
I	
  don’t	
  know,	
  just	
  this	
  one,	
  because	
  I	
  haven’t	
  read	
  it	
  
	
  

6)	
   CH1:	
   this	
  is	
  the	
  last	
  [one].	
  
	
  

3.2.4.	
  Discourse	
  markers	
  
Also	
  called	
  phatics,	
  discourse	
  markers	
  are	
  “lexical	
  items	
  such	
  as	
  well,	
  so,	
  you	
  know,	
  etc.,	
  which	
  do	
  not	
  

have	
  meaning	
  and	
  whose	
  basic	
   function	
   is	
   to	
   facilitate	
   the	
   flow	
  of	
   speech”	
   (Lázaro	
   Ibarrola	
  &	
  García	
  Mayo,	
  
2012,	
  p.	
  140),	
  as	
  shown	
  in	
  Example	
  7	
  (use	
  of	
  Spanish)	
  and	
  Example	
  8	
  (use	
  of	
  English):	
  

	
  
7)	
   CH1:	
   bueno,	
  princesa,	
  eso	
  es	
  una	
  princesa?	
  

Well,	
  princess,	
  that	
  is	
  a	
  princess	
  
	
  

8)	
   CH2:	
   and	
  this	
  picture,	
  is	
  a,	
  well	
  there	
  is	
  Aladdin	
  and	
  evil	
  and	
  evil	
  is	
  eh	
  eh	
  attacking	
  Aladdin.	
  

	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 66	
  

3.2.5.	
  Private	
  speech	
  
Private	
  speech	
  refers	
  to	
  those	
  segments	
  of	
  the	
  interaction	
  primarily	
  used	
  for	
  self-­‐regulatory	
  purposes	
  

(Pladevall	
   Ballester	
   &	
   Vraciu,	
   2017),	
   as	
   depicted	
   in	
   Example	
   9	
   (use	
   of	
   Spanish)	
   and	
   Example	
   10	
   (use	
   of	
  
English):	
  	
  

	
  
9)	
   CH2:	
   ay	
  the	
  umm	
  the	
  ay	
  ay	
  eh	
  eeh	
  #	
  ay	
  no	
  sé	
  the	
  eeh	
  this	
  (pointing	
  at	
  the	
  character).	
  

ay	
  the	
  umm	
  the	
  ay	
  ay	
  eh	
  eeh	
  	
  ay	
  not	
  know	
  the	
  eeh	
  this	
  
The	
  the	
  I	
  don’t	
  know	
  the	
  this	
  
	
  

10)	
   CH2:	
   to	
  the	
  ###	
  I	
  don’t	
  know.	
  
	
  

To	
   perform	
   statistical	
   analyses,	
   data	
   were	
   entered	
   into	
   SPSS	
   19	
   (IMB	
   Corp.	
   2010),	
   and	
   both	
  
descriptive	
   and	
   inferential	
   analyses	
   were	
   computed.	
   In	
   the	
   case	
   of	
   descriptive	
   analyses,	
   both	
   means	
   and	
  
standard	
   deviations	
   were	
   calculated	
   for	
   the	
   different	
   categories.	
   Kolmogorov-­‐Smirnov	
   tests	
   were	
   run	
   to	
  
verify	
   the	
   normality	
   of	
   distribution	
   of	
   the	
   samples.	
   As	
   data	
   did	
   not	
   follow	
   a	
   normal	
   distribution,	
   Mann-­‐
Whitney	
  U	
  tests	
  were	
  used	
  for	
  inter-­‐group	
  comparisons	
  and	
  Friedman	
  Tests	
  together	
  with	
  Wilcoxon’s	
  Signed	
  
Rank	
  Tests	
  were	
  conducted	
  for	
  intragroup	
  comparisons.	
  Statistical	
  significance	
  was	
  indicated	
  at	
  the	
  p	
  <	
  .01**	
  
and	
  p	
  <	
  .05*	
  levels.	
  Marginally	
  significant	
  differences	
  were	
  indicated	
  at	
  the	
  p	
  <	
  .09#	
  level.	
  

	
  
4.	
  Results	
  

To	
   answer	
   the	
   first	
   two	
   research	
   questions,	
   intergroup	
   comparisons	
   regarding	
   the	
   use	
   of	
  
Basque/Spanish	
  in	
  each	
  category	
  as	
  well	
  as	
  the	
  use	
  of	
  L3	
  English	
  are	
  provided.	
  Table	
  2	
  shows	
  the	
  results	
  for	
  
the	
  total	
  use	
  of	
  PKL	
  (Spanish/Basque).	
  

	
  
Table 2 
Spanish/Basque use during interaction: Descriptive and inferential statistics 

Learner groups Instances of Spanish/Basque use 
Mann-Whitney U test 

Z Significance 
CLIL Year 5 M = 17.00 (SD = 16.02) 

−2.905 **0.004 CLIL Year 6 M = 28.63 (SD = 25.31) 
	
  
As	
   can	
  be	
   observed	
   in	
  Table	
   2,	
   older	
  CLIL	
   learners	
  made	
   a	
   higher	
   use	
   of	
   PKL.	
   This	
   difference	
  was	
  

highly	
  significant	
  from	
  a	
  statistical	
  point	
  of	
  view.	
  The	
  differences	
  were	
  particularly	
  evident	
  in	
  metacomments,	
  
discourse	
   markers	
   and	
   private	
   speech,	
   as	
   depicted	
   in	
   Table	
   3.	
   The	
   L3	
   (English)	
   was	
   also	
   used	
   for	
   these	
  
categories	
   in	
   both	
   groups	
   (see	
   Table	
   4).	
   The	
   analysis	
   of	
   the	
   results	
   indicated	
   that	
   both	
   learner	
   groups	
  
employed	
  L3	
  English	
  with	
  the	
  same	
  frequency,	
  except	
   for	
   the	
  category	
  metacomments	
   in	
  which	
  a	
  statistical	
  
tendency	
   was	
   found	
   in	
   favour	
   of	
   Year	
   5	
   learners.	
   In	
   order	
   to	
   give	
   an	
   answer	
   to	
   the	
   last	
   three	
   questions,	
  
intragroup	
   comparisons	
   were	
   carried	
   out.	
   Tables	
   5-­‐6	
   display	
   the	
   findings	
   concerning	
   the	
   use	
   of	
  
Basque/Spanish	
  compared	
  to	
  English	
  in	
  each	
  category.	
  

	
  
Table 3 
Instances of Basque and Spanish use by categories: Descriptive and inferential statistics 

 Learner groups Mann-Whitney U test 
Categories CLIL Year 5 CLIL Year 6 Z Significance 

Appeals  M = 4.10 (SD = 3.86) M =3.85 (SD = 2.80) −0.229 0.819 
Clarification requests M = 1.20 (SD = 2.00) M = 0.93 (SD = 1.54) −0.404 0.686 
Metacomments M = 7.70 (SD = 9.47) M = 15.48 (SD = 20.42) −2.752 **0.006 
Discourse markers M = 3.02 (SD = 4.11) M = 6.28 (SD = 6.02) −3.319 **0.001 
Private speech M = 0.98 (SD = 1.46) M = 2.10 (SD = 2.82) −2.131 *0.033 
	
  
	
  
	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 67	
  

Table 4 
Instances of L3 English use by categories: Descriptive and inferential statistics 

 Learner groups Mann-Whitney U test 
Categories CLIL Year 5 CLIL Year 6 Z Significance 

Appeals  M =0.20 (SD = 0.40) M =0.10 (SD = 0.30) −1.293 0.196 
Clarification requests M = 0.56 (SD = 0.95) M = 0.30 (SD = 0.56) −0.999 0.318 
Metacomments M = 14.26 (SD = 10.07) M = 11.30 (SD = 11.46) −1.922 #0.055 
Discourse markers M = 0.04 (SD = 0.198) M = 0.65 (SD = 3.48) −1.165 0.244 
Private speech M = 0.96 (SD = 1.52) M = 0.73 (SD = 1.43) −1.216 0.224 

	
  
Table 5 
Comparison of instances of Basque/Spanish use to L3 English use by categories in Year 5: Descriptive and inferential statistics 

 Language use Wilcoxon Rank test 
Categories L1/L2 (Spanish/Basque) L3 (English) Z Significance 

Appeals M = 4.10 (SD = 3.86) M = 0.20 (SD = 0.40) −5.381 **0.001 
Clarification requests M = 1.20 (SD = 2.00) M = 0.56 (SD = 0.95) −1.710 #0.087 
Metacomments M = 7.70 (SD = 9.47) M = 14.26 (SD = 10.07) −3.204 **0.001 
Discourse markers M = 3.02 (SD = 4.11) M = 0.04 (SD = 0.198) −5.106 **0.001 
Private speech M = 0.98 (SD = 1.46) M = 0.96 (SD = 1.52) −0.285 0.775 
	
  
Table 6 
Comparison of instances of Basque/Spanish use to L3 English use by categories in Year 6: Descriptive and inferential statistics 

 Language use Wilcoxon Rank test 
Categories L1/L2 (Spanish/Basque) L3 (English) Z Significance 

Appeals M = 3.85 (SD = 2.80) M = 0.10 (SD = 0.30) −5.188 **0.001 
Clarification requests M = 0.93 (SD = 1.54) M = 0.30 (SD = 0.56) −2.581 *0.010 
Metacomments M = 15.48 (SD = 20.42) M = 11.30 (SD = 11.46) −0.805 0.421 
Discourse markers M = 6.28 (SD = 6.02) M = 0.65 (SD = 3.48) −4.726 **0.001 
Private speech M = 2.10 (SD = 2.82) M = 0.73 (SD = 1.43) −2.840 **0.005 
	
  

In	
  both	
  groups,	
  the	
  tendency	
  was	
  to	
  resort	
  to	
  their	
  PKL	
  more	
  frequently	
  than	
  to	
  English	
  in	
  appeals	
  for	
  
assistance,	
  clarification	
  requests	
  and	
  discourse	
  markers.	
  This	
  preference	
  reached	
  significance	
  in	
  both	
  groups.	
  In	
  
the	
  case	
  of	
  private	
  speech,	
  no	
  differences	
  in	
  Year	
  5	
  students,	
  while	
  a	
  significantly	
  greater	
  use	
  of	
  Basque/Spanish	
  
was	
  observed	
  in	
  Year	
  6	
  learners.	
  As	
  regards	
  metacomments,	
  Year	
  5	
  students	
  preferred	
  to	
  use	
  English	
  in	
  a	
  highly	
  
significant	
  manner,	
  whereas	
  the	
  inexistence	
  of	
  significant	
  differences	
  was	
  obtained	
  in	
  Year	
  6	
  students,	
  a	
  finding	
  
that	
  seems	
  to	
  indicate	
  a	
  more	
  balanced	
  use	
  of	
  both	
  languages	
  in	
  this	
  category.	
  Figure	
  1	
  depicts	
  the	
  descriptive	
  
comparison	
  of	
  the	
  use	
  of	
  Spanish/Basque	
  in	
  the	
  different	
  categories	
  among	
  Year	
  5	
  students.	
  

	
  

  
Figure 1. Categories served by Basque/Spanish in Year 5 

0 
1 
2 
3 
4 
5 
6 
7 
8 
9 

Metacomments  Appeals  Discourse markers Clarification requests Private speech 



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 68	
  

As	
   illustrated	
   by	
   Figure	
   1,	
  metacomments,	
   followed	
  by	
   appeals	
   and	
  discourse	
  markers,	
  were	
  most	
  
commonly	
   served	
   by	
   the	
   use	
   of	
   Basque/Spanish.	
   Clarification	
   requests	
   and	
   private	
   speech	
   were	
   the	
  
categories	
  with	
  the	
  least	
  use	
  of	
  these	
  languages.	
  In	
  order	
  to	
  see	
  whether	
  these	
  differences	
  were	
  statistically	
  
significant,	
   a	
   Friedman’s	
   Test	
   was	
   conducted.	
   Statistically	
   significant	
   differences	
   emerged	
   (Chi-­‐
Square=324.708;	
  p-­‐value=0.001),	
  and	
  therefore	
  post-­‐hoc	
  analyses	
  were	
  conducted.	
  A	
  Wilcoxon’s	
  Signed	
  Rank	
  
test	
   showed	
   that	
   metacomments	
   significantly	
   differed	
   from	
   the	
   other	
   categories.	
   In	
   addition,	
   statistically	
  
significant	
  differences	
  were	
  also	
  observed	
  among	
  appeals,	
  discourse	
  markers	
  and	
  clarification	
  requests,	
  while	
  
no	
  differences	
  emerged	
  between	
  clarification	
  requests	
  and	
  private	
  speech.	
  Table	
  7	
  presents	
  the	
  summary	
  of	
  
these	
  results.	
  The	
  same	
  descriptive	
  analysis	
  was	
  carried	
  out	
  for	
  year	
  6	
  (Figure	
  2).	
  
	
  

Table 7 
Use of Basque/Spanish: Wilcoxon’s Signed Rank Test for Year 5 

Categories Z Significance 
Clarification requests-Appeals −4.842 **0.001 
Metacomments-Appeals −2.432 *0.015 
Discourse markers-Appeals −2.124 *0.034 
Private speech-Appeals −4.753 **0.001 
Metacomments-Clarification requests −5.032 **0.001 
Discourse markers-Clarification requests −3.605 **0.001 
Private speech-Clarification requests −0.621 0.534 
Discourse markers-metacomments −3.158 **0.002 
Private speech-Metacomments −4.890 **0.001 
Private speech-Discourse markers −4.101 **0.001 

 

Figure 2. Categories served by Basque/Spanish in Year 6 
	
  	
  

As	
  in	
  Year	
  5,	
  metacomments	
  were	
  the	
  most	
  common	
  manifestation	
  of	
  Basque/Spanish	
  use,	
  followed	
  
by	
   discourse	
  markers	
   and	
   appeals	
   for	
   assistance.	
   The	
   use	
   of	
   these	
   languages	
  was	
   less	
   common	
   in	
   private	
  
speech,	
   and	
   even	
   less	
   in	
   clarification	
   requests.	
   The	
   Friedman	
   test	
   indicated	
   the	
   existence	
   of	
   significant	
  
differences	
   across	
   these	
   categories	
   (Chi-­‐Square=	
  292.418;	
  p-­‐value=0.001).	
   Consequently,	
   post-­‐hoc	
   analyses	
  
were	
  conducted.	
  The	
  Wilcoxon’s	
  Signed	
  Rank	
  test	
  (Table	
  8)	
  confirmed	
  the	
  existence	
  of	
  significant	
  differences	
  
between	
  metacomments	
  and	
  the	
  rest	
  of	
  the	
  categories.	
  Similarly,	
  the	
  differences	
  between	
  appeals,	
  discourse	
  
markers,	
  clarification	
  requests	
  and	
  private	
  speech	
  were	
  significant	
  or	
  nearly	
  significant.	
  Figure	
  3	
  shows	
  the	
  
descriptive	
  analysis	
  for	
  the	
  use	
  of	
  English	
  in	
  the	
  different	
  categories	
  explored.	
  

	
  
	
  
	
  
	
  

0 
2 
4 
6 
8 

10 
12 
14 
16 
18 

Metacomments  Appeals  Discourse markers Clarification requests Private speech 



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 69	
  

Table 8 
Use of Basque/Spanish: Wilcoxon’s Signed Rank Test for Year 6 

Categories Z Significance 
Clarification requests-Appeals −4.443 **0.001 
Metacomments-Appeals −5.144 **0.001 
Discourse markers-Appeals −1.853 #0.064 
Private speech-Appeals −2.876 **0.004 
Metacomments-Clarification requests −5.376 **0.001 
Discourse markers-Clarification requests −4.784 **0.001 
Private speech-Clarification requests −2.465 *0.014 
Discourse markers-metacomments −3.033 **0.002 
Private speech-Metacomments −5.306 **0.001 
Private speech-Discourse markers −3.904 **0.001 

	
  
	
  

 
Figure 3. Categories served by L3 English in Year 5 
 

The	
   use	
   of	
   English	
   was	
   prevalent	
   in	
   metacomments,	
   while	
   the	
   use	
   of	
   this	
   language	
   in	
   the	
   other	
  
categories	
   was	
   scarce.	
   Particularly,	
   the	
   use	
   of	
   English	
   discourse	
   markers	
   was	
   nearly	
   non-­‐existent.	
   The	
  
Friedman	
  test	
  revealed	
  the	
  existence	
  of	
  significant	
  differences	
  across	
  these	
  categories	
  (Chi-­‐Square	
  =	
  324.708;	
  
p	
  =	
  0.001).	
  A	
  Wilcoxon’s	
  Signed	
  Rank	
  test	
  indicated	
  that	
  metacomments	
  significantly	
  differed	
  from	
  the	
  rest	
  of	
  
the	
   categories.	
   Statistically	
   significant	
   differences	
   emerged	
   among	
   the	
   other	
   categories,	
   except	
   for	
   the	
  
contrast	
  between	
  private	
  speech	
  and	
  clarification	
  requests	
  (see	
  Table	
  9).	
  	
  

	
  
Table 9 
Use of English: Wilcoxon’s Signed Rank Test for Year 5 

Categories Z Significance 
Clarification requests-Appeals −2.640 0.008* 
Metacomments-Appeals −6.156 0.001** 
Discourse markers-Appeals −2.530 0.011* 
Private speech-Appeals −3.571 0.001** 
Metacomments-Clarification requests −6.155 0.001** 
Discourse markers-Clarification requests −3.342 0.001** 
Private speech-Clarification requests −1.335 0.182 
Discourse markers-metacomments −6.156 0.001** 
Private speech-Metacomments −6.141 0.001** 
Private speech-Discourse markers −4.073 0.001** 

	
  

0 
2 
4 
6 
8 

10 
12 
14 
16 

Metacomments  Appeals  Discourse markers Clarification requests Private speech 



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 70	
  

Figure	
   4	
   shows	
   the	
   results	
   for	
   the	
   use	
   of	
   English	
   by	
   Year	
   6	
   learners.	
   Year	
   6	
   students	
   also	
   made	
  
substantial	
   use	
   of	
   English	
   in	
   metacomments	
   compared	
   to	
   the	
   other	
   categories,	
   in	
   which	
   the	
   use	
   of	
   this	
  
language	
  was	
   less	
   frequent.	
   In	
   particular,	
   appeals	
  were	
   the	
   least	
   frequently	
   used	
   category.	
   The	
   interaction	
  
among	
   these	
   categories	
  was	
   significant	
   (Chi-­‐Square	
   =	
   292.418;	
  p	
  =	
   0.001).	
   Post-­‐hoc	
   analyses	
   showed	
   that	
  
metacomments	
  significantly	
  differed	
  from	
  the	
  rest	
  of	
  the	
  categories.	
  Statistically	
  significant	
  differences	
  were	
  
obtained	
  between	
  the	
  rest	
  of	
  the	
  categories,	
  except	
  for	
  the	
  contrast	
  between	
  discourse	
  markers	
  and	
  appeals	
  
for	
  assistance,	
  as	
  well	
  as	
  discourse	
  markers	
  and	
  clarification	
  requests	
  (Table	
  10).	
  	
  

	
  
	
  

 
Figure 4. Categories served by L3 English in Year 6 

 
Table 10 
Use of English: Wilcoxon’s Signed Rank test for Year 6 

Categories Z Significance 
Clarification requests-Appeals −1.890 0.059# 
Metacomments-Appeals −5.092 0.001** 
Discourse markers-Appeals −0.647 0.518 
Private speech-Appeals −2.724 0.006* 
Metacomments-Clarification requests −5.219 0.001** 
Discourse markers-Clarification requests −1.016 0.310 
Private speech-Clarification requests −2.041 0.041* 
Discourse markers-metacomments −5.089 0.001** 
Private speech-Metacomments −5.164 0.001** 
Private speech-Discourse markers −1.771 0.077# 

	
  
5.	
  Discussion	
  	
  

In	
  this	
  section,	
  the	
  five	
  research	
  questions	
  of	
  the	
  study	
  will	
  be	
  answered.	
  As	
  regards	
  the	
  first	
  research	
  
question	
   (Are	
   there	
   any	
   differences	
   between	
   both	
   age/proficiency	
   groups	
   in	
   terms	
   of	
   the	
   amount	
   of	
  
Basque/Spanish	
  use?),	
  the	
  intergroup	
  analysis	
  showed	
  a	
  greater	
  use	
  of	
  PKL	
  in	
  Year	
  6	
  learners,	
  a	
  result	
  in	
  line	
  
with	
   findings	
   from	
  other	
  studies	
  carried	
  out	
   in	
  a	
  CLIL	
  context	
   (Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  2018;	
  
García	
  Mayo	
  &	
  Hidalgo	
  Gordo,	
  2017;	
  García	
  Mayo	
  &	
  Imaz-­‐Agirre,	
  2017;	
  García	
  Mayo	
  &	
  Lázaro	
  Ibarrola,	
  2015).	
  
This	
  greater	
  use	
  of	
  Basque/Spanish	
  is	
  especially	
  reflected	
  in	
  metacomments,	
  discourse	
  markers	
  and	
  private	
  
speech.	
   Due	
   to	
   older	
   learners’	
   cognitive	
   maturity	
   and	
   on	
   economy	
   grounds,	
   they	
   seem	
   to	
   make	
   use	
   of	
  
strategies	
   that	
   are	
   less	
   cooperative	
   and	
   more	
   external	
   to	
   the	
   task.	
   In	
   fact,	
   studies	
   conducted	
   with	
   CLIL	
  
learners	
   (both	
   primary-­‐school	
   learners	
   and	
   adolescents)	
   have	
   shown	
   the	
   use	
   of	
   PKL	
   to	
   be	
   prevalent	
   in	
  
metacomments	
  and	
  discourse	
  markers,	
  and	
  this	
  use	
  did	
  not	
  decrease	
  as	
  age	
  increased	
  (Arratibel-­‐Irazusta	
  &	
  
Martínez-­‐Adrián,	
   2018;	
   Gutiérrez-­‐Mangado,	
   2015;	
   Lázaro	
   Ibarrola,	
   2016)3.	
   As	
   regards	
   more	
   cooperative	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
3	
   Also	
   note	
   that	
   private	
   speech,	
   an	
   underresearched	
   category	
   in	
   CLIL	
   studies,	
   was	
   also	
   maintained	
   over	
   the	
   period	
  
examined	
  in	
  the	
  study	
  by	
  Pladevall	
  and	
  Vraciu	
  (2017).	
  

0 

2 

4 

6 

8 

10 

12 

Metacomments  Appeals  Discourse markers Clarification requests Private speech 



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 71	
  

strategies	
  (appeals	
  for	
  assistance	
  and	
  clarification	
  requests),	
  the	
  intergroup	
  analysis	
  did	
  not	
  yield	
  statistically	
  
significant	
   differences.	
   But	
   despite	
   the	
   inexistence	
   of	
   statistically	
   significant	
   differences,	
   the	
   descriptive	
  
means	
   point	
   to	
   a	
   greater	
   use	
   by	
   Year	
   5	
   learners.	
   In	
   previous	
   investigations	
   with	
   young	
   CLIL	
   learners	
  
(Gutierrez-­‐Mangado,	
  2015),	
  similar	
  tendencies	
  were	
  observed	
  for	
  these	
  two	
  categories.	
  These	
  strategies	
  are	
  
more	
   internal	
   to	
   the	
  task,	
  and	
  we	
  could	
  claim	
  that	
  use	
  decreases	
  as	
  age	
  and	
  proficiency	
   increase.	
  A	
  greater	
  
gap	
   in	
  proficiency	
  could	
   likely	
  result	
   in	
  statistically	
  significant	
  differences	
   in	
  these	
  categories.	
  Nevertheless,	
  
the	
  results	
  from	
  the	
  present	
  study	
  contrast	
  with	
  Lázaro	
  Ibarrola	
  and	
  García	
  Mayo	
  (2012).	
  In	
  their	
  longitudinal	
  
study	
   with	
   secondary-­‐school	
   learners	
   in	
   which	
   a	
   wider	
   gap	
   in	
   proficiency	
   existed,	
   L1	
   use	
   in	
   appeals	
   for	
  
assistance	
  and	
  discourse	
  markers	
  decreased	
  as	
  proficiency	
  increased.	
  

Even	
   if	
   a	
   higher	
   use	
   of	
   Basque/Spanish	
   was	
   observed	
   in	
   the	
   older	
   group	
   of	
   learners,	
   as	
   in	
   other	
  
investigations	
  with	
   young	
  CLIL	
   learners	
   (García	
  Mayo	
  &	
  Hidalgo	
  Gordo,	
   2017;	
  García	
  Mayo	
  &	
   Imaz	
  Agirre,	
  
2017;	
   García	
  Mayo	
   &	
   Lázaro	
   Ibarrola,	
   2015),	
   the	
   underlying	
   reasons	
   cited	
   in	
   these	
   investigations	
   for	
   this	
  
higher	
  use	
  are	
  not	
  applicable	
  to	
  our	
  sample.	
  Unlike	
  the	
  older	
  learners	
  in	
  these	
  investigations,	
  for	
  whom	
  the	
  
task	
  performed	
  may	
  not	
  have	
  been	
  very	
  appealing	
  or	
   in	
  other	
   cases	
   (e.g.	
  when	
   task	
   repetition	
  was	
  carried	
  
out)	
  not	
  so	
  motivating	
  the	
  second	
  time	
  they	
  performed	
  the	
  task,	
  the	
  older	
  learners	
  in	
  the	
  present	
  study	
  were	
  
highly	
  engaged	
  while	
  carrying	
  out	
  the	
  task.	
  Their	
  attitude	
  towards	
  performing	
  a	
  different	
  task	
  from	
  the	
  one	
  
used	
   in	
  the	
  aforementioned	
  studies	
  suggested	
  that	
  the	
  task	
  was	
   indeed	
  attractive	
  to	
  them.	
  Thus,	
  we	
  cannot	
  
really	
  argue	
  that	
  low	
  motivation	
  explains	
  the	
  higher	
  use	
  of	
  PKL	
  on	
  the	
  part	
  of	
  older	
  learners.	
  The	
  higher	
  use	
  
of	
  PKL	
  by	
  older	
  learners	
  seems	
  to	
  be	
  a	
  trend	
  and	
  deserves	
  further	
  investigation	
  in	
  future	
  studies.	
  

With	
  respect	
  to	
  the	
  second	
  research	
  question—Are	
  there	
  any	
  differences	
  between	
  both	
  age/proficiency	
  
groups	
  in	
  terms	
  of	
  the	
  amount	
  of	
  English	
  use?—both	
  groups	
  were	
  able	
  to	
  perform	
  the	
  categories	
  examined	
  in	
  
English,	
  a	
  result	
   in	
   line	
  with	
  the	
  bilingual	
  nature	
  of	
   the	
  educational	
  programme	
  in	
  which	
  the	
   learners	
  were	
  
immersed.	
  The	
  use	
  of	
  both	
  Basque/Spanish	
  and	
  English	
  was	
  evident	
  in	
  the	
  different	
  categories.	
  These	
  results	
  
contrast	
   with	
   the	
   findings	
   obtained	
   in	
   previous	
   investigations	
   with	
   younger	
   learners	
   (Martínez-­‐Adrián,	
   in	
  
press)	
  in	
  which	
  similar	
  categories	
  were	
  examined.	
  In	
  particular,	
  metacomments	
  were	
  always	
  produced	
  in	
  L1	
  
Spanish	
  in	
  Martínez-­‐Adrián	
  (in	
  press),	
  while	
  both	
  Basque/Spanish	
  and	
  English	
  were	
  employed	
  in	
  the	
  present	
  
study.	
  The	
  differences	
  existing	
  in	
  the	
  CLIL	
  programmes	
  in	
  which	
  learners	
  were	
  enrolled	
  in	
  these	
  studies	
  could	
  
account	
   for	
   these	
   results.	
   In	
   the	
   present	
   study,	
   learners	
   received	
   instruction	
   in	
   content	
   subjects	
   such	
   as	
  
science,	
  arts	
  and	
  crafts	
  and	
  physical	
  education	
  exclusively	
  in	
  English.	
  In	
  Martínez-­‐Adrián	
  (in	
  press),	
  Years	
  4	
  
and	
  6	
  learners	
  were	
  receiving	
  half	
  of	
  their	
  lessons	
  in	
  English	
  and	
  half	
  in	
  Spanish4.	
  	
  

In	
   terms	
   of	
   quantity,	
   the	
   intergroup	
   analysis	
   revealed	
   that	
   English	
   was	
   employed	
   in	
   similar	
  
proportions	
  in	
  Years	
  5	
  and	
  6	
  as	
  in	
  previous	
  investigations	
  in	
  which	
  a	
  small	
  gap	
  in	
  age	
  and	
  proficiency	
  was	
  not	
  
so	
  big	
  as	
  to	
  result	
  in	
  statistically	
  significant	
  differences	
  (see	
  Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  2019,	
  with	
  
secondary-­‐school	
   learners	
   and	
   oral	
   production;	
   Martínez-­‐Adrián	
   et	
   al.,	
   2019,	
   with	
   self-­‐reported	
   opinions	
  
gathered	
  from	
  the	
  same	
  sample	
  tested	
  in	
  the	
  present	
  study),	
  except	
  for	
  the	
  metacomments	
  category	
  in	
  which	
  
a	
  statistical	
  tendency	
  was	
  found	
  in	
  favour	
  of	
  Year	
  5	
  learners.	
  This	
  result	
  is	
  related	
  to	
  the	
  answer	
  to	
  the	
  next	
  
research	
  question,	
  which	
  centered	
  on	
  language	
  preference	
  in	
  each	
  category.	
  In	
  the	
  case	
  of	
  this	
  category,	
  Year	
  
5	
  preferred	
  the	
  use	
  of	
  English	
  over	
  Basque/Spanish.	
  As	
  neither	
  interviews	
  with	
  the	
  teachers	
  of	
  this	
  group	
  of	
  
learners	
  nor	
  the	
  observation	
  of	
  the	
  lessons	
  were	
  carried	
  out,	
  the	
  advocacy	
  for	
  a	
  more	
  English-­‐only	
  policy	
  in	
  
the	
  Year	
  5	
  group,	
  which	
  could	
  explain	
  this	
  result,	
  remains	
  tentative.	
  	
  

As	
   for	
   the	
   third	
   research	
   question—Do	
   learners	
   in	
   each	
   group	
   show	
   a	
   preference	
   for	
   either	
  
Basque/Spanish	
  or	
  English	
  in	
  the	
  categories	
  analyzed?—	
  in	
  Year	
  5	
  a	
  significantly	
  higher	
  use	
  of	
  Basque/Spanish	
  
was	
   observed	
   in	
   nearly	
   all	
   the	
   categories	
   except	
   for	
   metacomments,	
   in	
   which	
   statistically	
   significant	
  
differences	
   emerged	
   in	
   favour	
   of	
   English	
  use.	
  A	
   similar	
   pattern	
  was	
   attested	
   in	
  Year	
  6,	
   even	
   though	
   in	
   the	
  
category	
   metacomments,	
   learners	
   used	
   Basque/Spanish	
   and	
   English	
   in	
   similar	
   proportions	
   given	
   the	
  
inexistence	
   of	
   statistically	
   significant	
   differences	
   in	
   this	
   respect.	
   Nevertheless,	
   even	
   if	
   the	
   analysis	
   did	
   not	
  
yield	
  statistical	
  significance,	
  the	
  descriptive	
  means	
  point	
  to	
  a	
  tendency	
  to	
  increase	
  the	
  use	
  of	
  Basque/Spanish	
  
over	
  English	
  in	
  this	
  category.	
  	
  

In	
   general,	
   the	
   trend	
   is	
   for	
   learners	
   to	
   use	
  Basque/Spanish	
   to	
   a	
   greater	
   extent	
   than	
  English	
   in	
   the	
  
categories	
  examined	
  when	
   they	
  are	
   interacting,	
  as	
   it	
   is	
   the	
   language	
  shared	
  by	
  both	
  members	
  of	
   the	
  dyads	
  
(see	
  Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  2018	
  and	
  Lázaro	
  Ibarrola,	
  2016	
  for	
  similar	
  results	
  in	
  the	
  category	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
4	
  Also	
  note	
  that	
  the	
  present	
  study	
  and	
  Martínez-­‐Adrián	
  (in	
  press)	
  differ	
  in	
  design,	
  as	
  Martínez-­‐Adrián	
  (in	
  press)	
  compared	
  
CLIL	
  to	
  non-­‐CLIL	
  learners	
  in	
  Years	
  4	
  and	
  6,	
  no	
  inferential	
  analysis	
  was	
  carried	
  out	
  when	
  contrasting	
  the	
  use	
  of	
  English	
  and	
  
Spanish,	
  and	
  the	
  task	
  performed,	
  although	
  similar	
  in	
  procedure,	
  differed	
  in	
  terms	
  of	
  the	
  visual	
  stimuli	
  employed.	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 72	
  

discourse	
  markers).	
  However,	
  the	
  category	
  metacomments	
  deserves	
  further	
  examination	
  as	
  a	
  preference	
  for	
  
English	
   existed	
   among	
   Year-­‐5	
   students	
   and	
   a	
   more	
   balanced	
   use	
   of	
   Basque/Spanish	
   and	
   English	
   was	
  
observed	
   in	
   Year-­‐6	
   learners.	
   What	
   these	
   results	
   reveal	
   is	
   that	
   English	
   is	
   special	
   for	
   the	
   performance	
   of	
  
metacomments.	
  This	
  category	
   is	
  specifically	
   linked	
  to	
   the	
  task	
  procedure.	
  Learners	
  use	
  English	
  to	
  a	
  greater	
  
extent	
   or	
   as	
   much	
   as	
   they	
   use	
   Basque/Spanish	
   when	
   they	
   have	
   to,	
   among	
   other	
   things,	
   move	
   the	
   task	
  
forward,when	
  they	
  have	
  to	
  classifiy	
  the	
  pictures	
  or	
  when	
  they	
  have	
  to	
  order	
  them.	
  In	
  this	
  respect,	
  the	
  more	
  
functional	
   use	
   of	
   English	
   is	
   observed.	
   We	
   should	
   not	
   overlook	
   the	
   fact	
   that	
   English	
   is	
   a	
   language	
   of	
  
communication	
  rather	
  than	
  an	
  object	
  of	
  study	
  in	
  CLIL	
  programmes	
  (Martínez-­‐Adrián	
  &	
  Gutiérrez-­‐Mangado,	
  
2015a),	
   which	
   could	
   explain	
   the	
   use	
   of	
   English	
   for	
   metacognitive	
   purposes.	
   This	
   result	
   also	
   aligns	
   with	
  
Martínez-­‐Adrián	
   et	
   al.’s	
   study	
   (2019)	
   in	
   which	
   the	
   preference	
   for	
   the	
   use	
   of	
   English	
   in	
   paraphrases	
   was	
  
evident	
   in	
   the	
   same	
   learners’	
   self-­‐reported	
   opinions.	
   In	
   that	
   study,	
   this	
  was	
   explained	
   in	
   light	
   of	
   the	
   input	
  
provided	
  to	
  CLIL	
  learners	
  and	
  the	
  fact	
  that	
  learners	
  use	
  the	
  foreign	
  language	
  as	
  a	
  means	
  of	
  communication,	
  
which	
  might	
  lead	
  them	
  to	
  manipulate	
  the	
  TL	
  to	
  a	
  higher	
  rate	
  compared	
  to	
  mainstream	
  learners.	
  Even	
  if	
  the	
  
use	
  of	
  English	
  in	
  paraphrases	
  or	
  descriptions	
  is	
  slightly	
  different	
  from	
  its	
  use	
  in	
  metacomments,	
  what	
  seems	
  
to	
  be	
  clear	
  is	
  that	
  the	
  use	
  of	
  English	
  as	
  a	
  language	
  of	
  communication	
  in	
  CLIL	
  lessons	
  may	
  be	
  the	
  underlying	
  
reason	
  for	
  both	
  episodes.	
  	
  

If	
  we	
  tackle	
  the	
  differences	
  between	
  Year	
  5	
  and	
  Year	
  6	
  learners	
  in	
  terms	
  of	
  metacomments,	
  even	
  if	
  we	
  
observed	
  a	
  similar	
  use	
  of	
  Basque/Spanish	
  and	
  English	
  in	
  Year	
  6	
  learners,	
  the	
  tendency	
  seems	
  to	
  augment	
  the	
  
use	
   of	
   their	
   PKL.	
   It	
   seems	
   as	
   if	
   these	
   Year	
   6	
   learners	
   were	
   more	
   practical	
   and	
   were	
   guided	
   by	
   economy	
  
principles	
  to	
  solve	
  the	
  task	
  in	
  a	
  faster	
  way	
  (Lin,	
  2015).	
  This	
  is	
  not	
  surprising	
  as	
  the	
  same	
  learners	
  in	
  Year	
  6	
  
were	
   found	
   to	
   show	
   a	
   preference	
   for	
   more	
   communicative	
   options,	
   such	
   as	
   borrowings,	
   to	
   avoid	
  
communication	
  breakdowns	
  in	
  the	
  self-­‐report	
  questionnaire	
  administered	
  to	
  them	
  in	
  Martínez-­‐Adrián	
  et	
  al.	
  
(2019).	
   In	
   fact,	
   other	
   studies	
   with	
   primary-­‐school	
   learners	
   have	
   reported	
   similar	
   tendencies	
   (Gutiérrez-­‐
Mangado,	
   2015)	
   and	
   the	
   same	
   applies	
   to	
   secondary-­‐school	
   learners	
   (Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  
2018).	
  On	
  the	
  other	
  hand,	
  Year	
  5	
  learners	
  seem	
  to	
  be	
  more	
  conservative	
  by	
  using	
  the	
  language	
  of	
  instruction	
  
to	
  a	
  greater	
  extent	
  while	
  making	
  comments	
  related	
  to	
  the	
  task.	
  However,	
  this	
  argument	
  remains	
  tentative	
  due	
  
to	
  the	
  lack	
  of	
  questionnaires	
  or	
  interviews	
  carried	
  out	
  with	
  these	
  learners	
  and	
  their	
  teachers	
  about	
  the	
  use	
  of	
  
Basque/Spanish	
   and	
   English	
   in	
   their	
   classes.	
   The	
   use	
   of	
   more	
   qualitative	
   data	
   on	
   the	
   language	
   policy	
  
implemented	
  in	
  each	
  class	
  would	
  undoubtedly	
  illuminate	
  the	
  discussion	
  of	
  the	
  results.	
  	
  

All	
  in	
  all,	
  what	
  these	
  results	
  seem	
  to	
  evince	
  is	
  that,	
  even	
  if	
  English	
  is	
  widely	
  used	
  in	
  metacomments,	
  
Basque/Spanish	
   use	
   overrides	
   the	
   use	
   of	
   English	
   in	
   the	
   rest	
   of	
   the	
   categories.	
   This	
   aligns	
   with	
   previous	
  
research	
   studies	
   in	
   which	
   previously	
   known	
   language-­‐based	
   strategies	
   coexist	
   with	
   the	
   use	
   of	
   TL-­‐based	
  
strategies	
   (Arratibel-­‐Irazusta	
   &	
   Martínez-­‐Adrián,	
   2019;	
   Martínez-­‐Adrián,	
   in	
   press;	
   Martínez-­‐Adrián	
   et	
   al.,	
  
2019;	
   Pladevall	
   Ballester	
   &	
   Vraciu,	
   2017).	
   This	
   suggests	
   that	
   these	
   learners	
   are	
   functioning	
   in	
   a	
   bilingual	
  
mode,	
  rather	
  than	
  in	
  a	
  monolingual	
  mode	
  (Martínez-­‐Adrián,	
  in	
  press).	
  

In	
  response	
  to	
  the	
  fourth	
  research	
  question—Among	
  the	
  categories	
  examined,	
  which	
  ones	
  are	
  the	
  most	
  
commonly	
  served	
  by	
  Basque/Spanish?—metacomments	
  are	
  the	
  most	
  common	
  manifestation	
  of	
  the	
  use	
  of	
  PKL,	
  
followed	
  by	
  appeals	
   for	
   assistance	
   and	
  discourse	
  markers	
   in	
  Year	
  5.	
  A	
   similar	
   result	
   is	
   obtained	
   in	
  Year	
  6.	
  
These	
  are	
  the	
  functions	
  characterized	
  by	
  a	
  greater	
  use	
  of	
  the	
  L1	
  in	
  previous	
  investigations	
  with	
  both	
  children	
  
(Azkarai	
  &	
  García	
  Mayo,	
   2017;	
   García	
  Mayo	
  &	
  Hidalgo	
  Gordo,	
   2017;	
  Martínez-­‐Adrián,	
   in	
   press)	
   and	
   adults	
  
(Alegría	
   de	
   la	
   Colina	
   &	
   García	
   Mayo,	
   2009;	
   Arratibel-­‐Irazusta	
   &	
   Martínez-­‐Adrián,	
   2018;	
   Azkarai	
   &	
   García	
  
Mayo,	
  2015;	
  Storch	
  &	
  Aldosari,	
  2010).	
  The	
  use	
  of	
  Basque/Spanish	
  in	
  metacomments	
  allowed	
  the	
  learners	
  to	
  
progress	
   in	
   their	
   completion	
   of	
   the	
   task	
   and	
   comply	
   with	
   task	
   requirements.	
   Similarly,	
   the	
   use	
   of	
   these	
  
languages	
  in	
  appeals	
  for	
  assistance	
  regarding	
  deliberations	
  over	
  vocabulary	
  and	
  in	
  discourse	
  markers	
  to	
  keep	
  
the	
  flow	
  of	
  speech	
  prevented	
  them	
  from	
  communication	
  breakdowns.	
  In	
  fact,	
  the	
  use	
  of	
  appeals	
  for	
  assistance	
  
was	
  also	
  one	
  of	
  the	
  strategies	
  that	
  ranked	
  higher	
  in	
  frequency	
  in	
  the	
  self-­‐report	
  questionnaire	
  administered	
  
to	
  the	
  same	
  learners	
  in	
  Martínez-­‐Adrián	
  et	
  al.	
  (2019).	
  	
  

The	
   least	
   used	
   functions	
   were	
   clarification	
   requests	
   and	
   private	
   speech.	
   The	
   lower	
   use	
   of	
   L1s	
   in	
  
clarification	
   requests	
   has	
   also	
   been	
   attested	
   in	
   research	
   with	
   younger	
   learners	
   (Azkarai	
   &	
   García	
   Mayo,	
  
2017).	
  Private	
  speech	
   is	
   similarly	
  uncommon	
   in	
  other	
   investigations	
  with	
  children	
   (García	
  Mayo	
  &	
  Hidalgo	
  
Gordo,	
  2017)	
  as	
  well	
  as	
  in	
  studies	
  with	
  adults	
  conducted	
  in	
  an	
  EFL	
  setting	
  (Alegría	
  de	
  la	
  Colina	
  &	
  García	
  Mayo,	
  
2009).	
  	
  

Finally,	
  in	
  response	
  to	
  the	
  last	
  research	
  question—Among	
  the	
  categories	
  examined,	
  which	
  ones	
  are	
  the	
  
most	
   commonly	
   served	
   by	
   English?—metacomments	
   were	
   the	
   most	
   productive	
   category	
   in	
   both	
   years.	
  
Students	
   employed	
   the	
   language	
   of	
   instruction,	
   which	
   together	
   with	
   Spanish,	
   was	
   also	
   used	
   for	
   task	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 73	
  

management.	
   This	
   finding	
   could	
   be	
   related	
   to	
   the	
   use	
   of	
   the	
   TL	
   in	
   CLIL	
   programmes	
   with	
   a	
   meaningful	
  
purpose.	
  However,	
   the	
  use	
  of	
  English	
   is	
  uncommon	
   in	
  discourse	
  markers	
  and	
  appeals	
   for	
  assistance.	
  Other	
  
studies	
   have	
   also	
   reported	
   the	
   inexistence	
   of	
   English	
   discourse	
   markers	
   in	
   CLIL	
   learners’	
   discourse	
  
(Arratibel-­‐Irazusta	
  &	
  Martínez-­‐Adrián,	
  2018;	
  Lázaro	
  Ibarrola,	
  2016).	
  Likewise,	
  when	
  learners	
  noticed	
  a	
  gap	
  in	
  
knowledge	
   while	
   interacting,	
   they	
   preferred	
   to	
   use	
   Basque/Spanish	
   rather	
   than	
   English,	
   so	
   the	
  
communicative	
   flow	
  of	
   speech	
  would	
  not	
  be	
   interrupted	
  and	
  doubts	
   concerning	
  vocabulary	
   items	
  could	
  be	
  
solved	
  much	
  more	
  effectively	
  and	
  efficiently.	
  In	
  this	
  fashion,	
  the	
  message	
  can	
  be	
  conveyed.	
  	
  

	
  
6.	
  Conclusion	
  

The	
  goal	
  of	
  this	
  study	
  has	
  been	
  to	
  examine	
  the	
  use	
  of	
  PKL	
  (Basque	
  and	
  Spanish)	
  and	
  the	
  TL	
  (English)	
  
in	
  appeals	
  for	
  assistance,	
  clarification	
  requests,	
  metacomments,	
  discourse	
  markers	
  and	
  private	
  speech	
  in	
  two	
  
different	
   age/proficiency	
   groups	
   of	
   primary	
   school	
   learners	
   (grades	
   5	
   and	
   6)	
   immersed	
   in	
   a	
   CLIL	
   context,	
  
while	
   performing	
   a	
   communicative	
   task	
   in	
   pairs.	
   In	
   terms	
   of	
   amount,	
   there	
  were	
   differences	
   between	
   age	
  
groups,	
  with	
  greater	
  use	
  of	
  PKL	
  among	
  the	
  older	
  learners,	
  especially	
  in	
  less	
  cooperative	
  strategies	
  and	
  those	
  
more	
  external	
   to	
   the	
   task	
  (metacomments,	
  discourse	
  markers	
  and	
  private	
  speech).	
  As	
   for	
   the	
  use	
  of	
   the	
  TL	
  
(English),	
   the	
   analysis	
   indicated	
   that	
   both	
   age	
   groups	
   behaved	
   in	
   a	
   similar	
   way,	
   except	
   with	
   regard	
   to	
  
metacomments,	
  which	
  younger	
  learners	
  used	
  slightly	
  more	
  frequently.	
  The	
  intragroup	
  analysis	
  revealed	
  that	
  
both	
  age	
  groups	
  resorted	
  to	
  Basque/Spanish	
  more	
  frequently	
  than	
  to	
  English,	
  except	
   for	
  metacomments,	
   in	
  
which	
   the	
   younger	
   group	
   preferred	
   to	
   use	
   English,	
   while	
   older	
   learners	
  made	
   use	
   of	
   Basque/Spanish	
   and	
  
English	
   in	
   similar	
   proportions.	
   The	
   fact	
   that	
   English	
   has	
   a	
   more	
   functional	
   use	
   in	
   CLIL	
   settings,	
   as	
   the	
  
language	
   of	
   communication	
   in	
   these	
   meaning-­‐oriented	
   settings,	
   could	
   account	
   for	
   the	
   use	
   of	
   English	
   for	
  
metacognitive	
  purposes	
  by	
  these	
  learners	
  during	
  dyadic	
  interaction.	
  In	
  terms	
  of	
  types,	
  a	
  similar	
  pattern	
  was	
  
observed	
  as	
  regards	
  the	
  most	
  common	
  manifestation	
  of	
  Basque/Spanish	
  use	
  and	
  English	
  use.	
  Metacomments,	
  
followed	
  by	
  discourse	
  markers	
  and	
  appeals	
  for	
  assistance	
  were	
  the	
  categories	
  most	
  commonly	
  served	
  by	
  PKL.	
  
In	
  the	
  case	
  of	
  TL	
  use,	
  metacomments	
  were	
  widely	
  employed.	
  What	
  this	
  analysis	
  indicates	
  is	
  that	
  learners	
  use	
  
their	
   language	
  repertoire	
  for	
  task-­‐management	
  purposes,	
  a	
   finding	
  which	
  could	
  be	
  ascribed	
  to	
  the	
  bilingual	
  
nature	
  of	
  the	
  CLIL	
  programme	
  in	
  which	
  they	
  were	
  immersed.	
  	
  

In	
  the	
  light	
  of	
  the	
  results,	
  two	
  main	
  pedagogical	
  implications	
  may	
  be	
  drawn.	
  The	
  examination	
  of	
  the	
  
findings	
   has	
   shown	
   the	
   beneficial	
   effect	
   of	
   CLIL	
   on	
   the	
   learners’	
   use	
   of	
   their	
   language	
   repertoire	
   for	
   task-­‐
management	
   purposes.	
   However,	
   the	
   less	
   balanced	
   use	
   of	
   their	
   PKL	
   and	
   the	
   TL	
   attested	
   in	
   appeals	
   for	
  
assistance	
   indicates	
   the	
   need	
   for	
   more	
   TL-­‐based	
   strategies	
   (i.e.,	
   use	
   of	
   synonyms,	
   descriptions,	
  
reformulations)	
  among	
  students	
  or	
  the	
  enhancement	
  of	
  paraphrasing	
  during	
  teachers’	
  discourse,	
  which	
  could	
  
provide	
  students	
  with	
  rich	
  and	
  repeated	
  modeling	
  (Dalton-­‐Puffer,	
  2016,	
  as	
  cited	
  in	
  Pavón	
  &	
  Ramos,	
  2019).	
  It	
  
is	
  not	
  a	
  matter	
  of	
  banning	
  the	
  use	
  of	
  their	
  PKL,	
  but	
  of	
  making	
  students	
  aware	
  of	
  other	
  strategies	
  that	
  could	
  
help	
   them	
   during	
   interaction	
   when	
   they	
   might	
   feel	
   stuck.	
   Similarly,	
   more	
   systematic	
   activities	
   aimed	
   at	
  
practicing	
   TL	
   lexical	
   discourse	
   markers	
   could	
   promote	
   the	
   development	
   of	
   more	
   natural	
   speech.	
   Taken	
  
together,	
  these	
  two	
  measures	
  could	
  boost	
  learners’	
  flow	
  of	
  speech	
  in	
  the	
  TL.	
  By	
  integrating	
  the	
  development	
  
of	
  strategy	
  use	
  in	
  the	
  curriculum,	
  we	
  would	
  be	
  contributing	
  to	
  a	
  more	
  balanced	
  use	
  of	
  learners’	
  PKL	
  and	
  the	
  
TL,	
  as	
  advocated	
  by	
  recent	
  research	
  (see	
  Martínez-­‐Adrián,	
  Gutiérrez-­‐Mangado,	
  &	
  Gallardo-­‐del-­‐Puerto,	
  2019).	
  
In	
   other	
  words,	
   gradually	
   bridging	
   PKL	
   and	
   the	
   TL	
   in	
   the	
   everyday/academic	
   register	
   could	
   lead	
   to	
  more	
  
optimal	
   communication	
   and	
   maximize	
   the	
   effectiveness	
   of	
   CLIL	
   in	
   terms	
   of	
   both	
   content	
   and	
   language	
  
learning	
  (Lo	
  &	
  Lin,	
  2019).	
  

For	
   future	
  research,	
  as	
   indicated	
   in	
  Martínez-­‐Adrián	
  et	
  al.	
   (2019),	
   statistical	
  correlations	
  should	
  be	
  
conducted	
   between	
   learners’	
   self-­‐reported	
   opinions	
   and	
   their	
   performance	
   during	
   oral	
   interaction.	
   In	
  
general,	
  a	
  call	
   for	
  more	
  research	
  along	
  these	
   lines	
  has	
  been	
  made	
   in	
   the	
   literature	
  (Gao,	
  2007).	
  This	
   line	
  of	
  
research	
  is	
  especially	
  inexistent	
  in	
  studies	
  on	
  young	
  learners.	
  Similarly,	
  longitudinal	
  investigations	
  will	
  shed	
  
more	
  light	
  on	
  how	
  the	
  use	
  of	
  PKL	
  and	
  the	
  TL	
  evolve	
  over	
  time.	
  Likewise,	
  more	
  qualitative	
  data	
  (observations,	
  
interviews	
   with	
   both	
   learners	
   and	
   teachers	
   in	
   different	
   subjects)	
   would	
   elucidate	
   some	
   of	
   the	
   arguments	
  
provided	
  in	
  the	
  discussion	
  of	
  the	
  results.	
  The	
  impact	
  of	
  different	
  types	
  of	
  pairings	
  and	
  task-­‐modality	
  effects	
  
are	
  also	
  worth	
  investigating	
  so	
  as	
  to	
  explore	
  the	
  best	
  learning	
  conditions	
  for	
  young	
  learners.	
  Finally,	
  a	
  follow-­‐
up	
   study	
   comparing	
   the	
   performance	
   of	
   CLIL	
   to	
   non-­‐CLIL	
   leaners	
   will	
   allow	
   us	
   to	
   reach	
   more	
   robust	
  
conclusions	
  concerning	
  the	
  effect	
  of	
  more	
  intense	
  and	
  meaningful	
  input	
  during	
  CLIL	
  lessons.	
  	
  

	
  
	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 74	
  

Acknowledgements	
  
The	
   author	
   would	
   like	
   to	
   acknowledge	
   the	
   grants	
   awarded	
   by	
   the	
   Spanish	
   Ministry	
   of	
   Economy	
   and	
  
Competitiveness	
   (FFI2012-­‐32212)	
   and	
   (FFI2016-­‐74950-­‐P)(AEI/FEDER/UE),	
   the	
   University	
   of	
   the	
   Basque	
  
Country	
  (UPV/EHU)	
  (UFI	
  11/06)	
  and	
  the	
  Basque	
  Government	
  (IT904-­‐16).	
  

	
  
	
  
	
  
References	
  
Agustín	
  Llach	
  ,	
  María	
  del	
  Pilar	
  (2014).	
  L1	
  use	
  in	
  children	
  EFL	
  learners	
  in	
  traditional	
  versus	
  CLIL	
  instruction.	
  

Paper	
  presented	
  at	
  the	
  International	
  Conference	
  on	
  Child	
  Foreign	
  Language	
  Acquisition.	
  Universidad	
  
del	
  País	
  Vasco,	
  16-­‐17	
  October.	
  

Alegría	
  de	
  la	
  Colina,	
  Ana,	
  &	
  García	
  Mayo,	
  María	
  del	
  Pilar	
  (2009).	
  Oral	
  interaction	
  in	
  task-­‐based	
  EFL	
  learning:	
  
The	
  use	
  of	
   the	
  L1	
  as	
  a	
  cognitive	
   tool.	
   International	
  Review	
  of	
  Applied	
  Linguistics	
   (IRAL),	
  47(3),	
  325-­‐
345.	
  

Arratibel-­‐Irazusta,	
  Izaskun,	
  &	
  Martínez-­‐Adrián,	
  María	
  (2018).	
  The	
  use	
  of	
  previously	
  known	
  languages	
  in	
  the	
  
oral	
   production	
   of	
   L3	
   English	
   learners:	
   A	
   pseudolongitudinal	
   study.	
   European	
   Journal	
   of	
   Applied	
  
Linguistics,	
  6(2),	
  229-­‐254.	
  

Arratibel-­‐Irazusta,	
   Izaskun,	
  &	
  Martínez-­‐Adrián,	
  María	
   (2019).	
  Communication	
   strategies	
   in	
  L3	
  English	
  CLIL	
  
learners:	
  A	
  study	
  of	
  crosslinguistic	
  influence	
  and	
  target	
  language-­‐based	
  strategies	
  in	
  oral	
  production.	
  
In	
   M.	
   Juncal	
   Gutiérrez-­‐Mangado,	
   María	
   Martínez-­‐Adrián,	
   &	
   Francisco	
   Gallardo-­‐del-­‐Puerto	
   (Eds.),	
  
Crosslinguistic	
  influence:	
  From	
  empirical	
  evidence	
  to	
  classroom	
  practice	
  (pp.	
  169-­‐189).	
  Dordrecht,	
  The	
  
Netherlands:	
  Springer.	
  

Azkarai,	
   Agurtzane	
   (2015).	
   L1	
   use	
   in	
   EFL	
   task-­‐based	
   interaction:	
   A	
  matter	
   of	
   gender?	
  European	
   Journal	
   of	
  
Applied	
  Linguistics,	
  3(2),	
  159-­‐179.	
  

Azkarai,	
  Agurtzane,	
  &	
  García	
  Mayo,	
  María	
  del	
  Pilar	
  (2015).	
  Task	
  modality	
  and	
  L1	
  use	
  in	
  EFL	
  oral	
  interaction.	
  
Language	
  Teaching	
  Research,	
  19,	
  550-­‐571.	
  

Azkarai,	
  Agurtzane,	
  &	
  García	
  Mayo,	
  María	
  del	
  Pilar	
  (2017).	
  Task	
  repetition	
  effects	
  on	
  L1	
  use	
  in	
  EFL	
  child	
  task-­‐
based	
  interaction.	
  Language	
  Teaching	
  Research,	
  21,	
  480-­‐495.	
  

Azkarai,	
  Agurtzane,	
  &	
  Imaz	
  Agirre,	
  Ainara	
  (2017).	
  Gender	
  and	
  age	
  in	
  child	
  interaction	
  in	
  an	
  EFL	
  CLIL	
  context:	
  
An	
   exploratory	
   study.	
   In	
  María	
   del	
   Pilar	
   García	
  Mayo	
   (Ed.),	
  Learning	
   foreign	
   languages	
   in	
   primary	
  
school:	
  Research	
  insights	
  (pp.	
  103-­‐123).	
  Clevedon,	
  United	
  Kingdom:	
  Multilingual	
  Matters.	
  

Celaya,	
  María	
   Luz	
   (2008).	
   I	
   study	
   natus	
   in	
   English:	
   Lexical	
   transfer	
   in	
   CLIL	
   and	
   regular	
   learners.	
   In	
  Rafael	
  
Monroy	
  &	
  Aquilino	
  Sánchez	
  (Eds.),	
  25	
  Años	
  de	
   lingüística	
  aplicada	
  en	
  España:	
  Hitos	
  y	
  retos	
   (pp.	
  43–
49).	
  Murcia,	
  España:	
  Editum	
  (Ediciones	
  de	
  la	
  Universidad	
  de	
  Murcia).	
  

Celaya,	
  María	
  Luz,	
  &	
  Ruiz	
  de	
  Zarobe,	
  Yolanda	
  (2010).	
  First	
  language	
  and	
  age	
  in	
  CLIL	
  and	
  Non-­‐CLIL	
  contexts.	
  
International	
  CLIL	
  Research	
  Journal,	
  1(3),	
  60-­‐66.	
  

Cenoz,	
  Jasone	
  (2001).	
  The	
  effect	
  of	
  linguistic	
  distance,	
  L2	
  status	
  and	
  age	
  on	
  crosslinguistic	
  influence	
  in	
  Third	
  
Language	
   Acquisition.	
   In	
   Jasone	
   Cenoz,	
   Britta	
   Hufeisen	
   &	
   Ulrike	
   Jessner	
   (Eds.),	
   Cross-­‐linguistic	
  
influence	
   in	
   Third	
   Language	
   cquisition:	
   Psycholinguistic	
   perspectives	
   (pp.	
   8-­‐20).	
   Clevedon,	
   United	
  
Kingdom:	
  Multilingual	
  Matters.	
  

Cenoz,	
   Jasone	
   (2003).	
   Cross-­‐linguistic	
   influence	
   in	
   third	
   language	
   acquisition:	
   Implications	
   for	
   the	
  
organization	
   of	
   the	
   multilingual	
   mental	
   lexicon.	
   Bulletin	
   VALS-­‐ASLA	
   (Vereinigung	
   für	
   angewandte	
  
Linguistik	
  in	
  der	
  Schweiz),	
  78,	
  1-­‐11.	
  

Coyle,	
  Do	
   (2007).	
  Content	
  and	
   language	
   integrated	
   learning:	
  Toward	
  a	
  connected	
  research	
  agenda	
   for	
  CLIL	
  
pedagogies.	
  The	
  International	
  Journal	
  of	
  Bilingual	
  Education	
  and	
  Bilingualism,	
  10,	
  543-­‐562.	
  

Dalton-­‐Puffer,	
   Christiane	
   (2011).	
   Content	
   and	
   language	
   integrated	
   learning:	
   From	
   practice	
   to	
   principles.	
  
Annual	
  Review	
  of	
  Applied	
  Linguistics,	
  31,	
  182-­‐204.	
  	
  

DiCamilla,	
   Fred,	
   &	
   Antón,	
   Marta	
   (2012).	
   Functions	
   of	
   L1	
   in	
   the	
   collaborative	
   interaction	
   of	
   beginning	
   and	
  
advanced	
  second	
  language	
  learners.	
  International	
  Journal	
  of	
  Applied	
  Linguistics,	
  22,	
  160-­‐188.	
  

Gallardo-­‐del-­‐Puerto,	
  Francisco	
  (2015).	
  L1	
  influence	
  in	
  CLIL	
  vs.	
  EFL	
  schoolchildren:	
  A	
  study	
  of	
  codeswitching	
  
and	
   transfer	
   lapses.	
   Paper	
   presented	
   at	
   the	
   33th	
   AESLA	
   International	
   Conference.	
   Universidad	
  
Politécnica	
  de	
  Madrid,	
  16-­‐18	
  April.	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 75	
  

Gallardo-­‐del-­‐Puerto,	
  Francisco,	
  &	
  Gómez	
  Lacabex,	
  Esther	
   (2013)	
  The	
   impact	
  of	
  additional	
  CLIL	
  exposure	
  on	
  
oral	
  English	
  production.	
  Journal	
  of	
  English	
  Studies,	
  11,	
  113-­‐131.	
  

Gallardo-­‐del-­‐Puerto,	
  Francisco,	
   &	
   Gómez	
   Lacabex,	
   Esther	
   (2017).	
  Oral	
   production	
   outcomes	
   in	
   CLIL:	
   An	
  
attempt	
  to	
  manage	
  amount	
  of	
  exposure.	
  European	
  Journal	
  of	
  Applied	
  Linguistics,	
  5(1),	
  31-­‐54.	
  

Gao,	
   Xuesong	
   (2007).	
  Has	
   language	
   learning	
   strategy	
   research	
   come	
   to	
   an	
   end?	
  A	
   response	
   to	
   Tseng	
   et	
   al.	
  
2006.	
  Applied	
  Linguistics,	
  28,	
  615-­‐620.	
  

García	
  Mayo,	
  María	
  del	
  Pilar	
   (2018).	
  Child	
   task-­‐based	
   interaction	
   in	
  EFL	
   settings:	
  Research	
  and	
  challenges.	
  
International	
  Journal	
  of	
  English	
  Studies,	
  18(2),	
  119-­‐143.	
  

García	
   Mayo,	
   María	
   del	
   Pilar,	
   &	
   Hidalgo	
   Gordo,	
   María	
   de	
   los	
   Ángeles	
   (2017).	
   L1	
   use	
   among	
   young	
   EFL	
  
mainstream	
  and	
  CLIL	
  learners	
  in	
  task-­‐supported	
  interaction.	
  System,	
  67,	
  132-­‐145.	
  

García	
  Mayo,	
  María	
  del	
  Pilar,	
  &	
   Imaz	
  Agirre,	
  Ainara	
   (2017).	
  Child	
  EFL	
   interaction;	
  age,	
   instructional	
   setting	
  
and	
   development.	
   In	
   Janet	
   Enever,	
   &	
   Eva	
   Lindgren	
   (Eds.),	
   Researching	
   the	
   complexity	
   of	
   early	
  
language	
  learning	
  in	
  instructed	
  contexts	
  (pp.	
  249-­‐268).	
  Bristol,	
  United	
  Kingdom:	
  Multilingual	
  Matters.	
  

García	
  Mayo,	
  María	
  del	
  Piar,	
  &	
  Lázaro	
   Ibarrola,	
  Amparo	
   (2015).	
  Do	
  children	
  negotiate	
   for	
  meaning	
   in	
   task-­‐
based	
  interaction?	
  Evidence	
  from	
  CLIL	
  and	
  EFL	
  settings.	
  System,	
  54,	
  40-­‐54.	
  

Gené	
  Gil,	
  María,	
   Juan	
  Garau,	
  María,	
  &	
   Salazar	
  Noguera,	
   Joana	
   (2012).	
   A	
   case	
   study	
   exploring	
   oral	
   language	
  
choice	
  between	
  the	
   target	
   language	
  and	
   the	
  L1s	
   in	
  mainstream	
  CLIL	
  and	
  EFL	
  secondary	
  education.	
  
Revista	
  de	
  Lingüística	
  y	
  Lenguas	
  Aplicadas,	
  7,	
  133-­‐146.	
  

Gutiérrez-­‐Mangado,	
  M.	
   Juncal	
   (2015).	
  The	
  development	
  of	
   L1	
  use	
  by	
   schoolchildren	
   in	
  CLIL	
   and	
  NON-­‐CLIL	
  
contexts.	
   Paper	
   presented	
   at	
   the	
   33th	
  AESLA	
   International	
   Conference.	
   Universidad	
   Politécnica	
   de	
  
Madrid,	
  16-­‐18	
  April.	
  

Lasagabaster,	
  David	
   (2008).	
  Foreign	
   language	
  competence	
   in	
   content	
  and	
   language	
   integrated	
  courses.	
  The	
  
Open	
  Applied	
  Linguistics	
  Journal,	
  1,	
  30-­‐41.	
  	
  

Lázaro	
  Ibarrola,	
  Amparo	
  (2016).	
  Are	
  CLIL	
   learners	
  simply	
   faster	
  or	
  also	
  different?	
  Evidence	
   from	
  L1	
  use	
   in	
  
the	
  repair	
  sequences	
  and	
  discourse	
  markers	
  of	
  CLIL	
  and	
  EFL	
  learners.	
  Vigo	
  International	
   Journal	
  of	
  
Applied	
  Linguistics	
  (VIAL),	
  13,	
  127-­‐145.	
  

Lázaro	
   Ibarrola,	
  Amparo,	
  &	
  Azpilicueta	
  Martínez,	
  Raúl	
   (2015).	
   Investigating	
  negotiation	
  of	
  meaning	
   in	
  EFL	
  
children	
  with	
  very	
  low	
  levels	
  of	
  proficiency.	
  International	
  Journal	
  of	
  English	
  Studies,	
  15,	
  1-­‐21.	
  

Lázaro	
  Ibarrola,	
  Amparo,	
  &	
  García	
  Mayo,	
  María	
  del	
  Pilar	
  (2012).	
  L1	
  use	
  and	
  morphosyntactic	
  development	
  in	
  
the	
  oral	
  production	
  of	
  EFL	
  learners	
  in	
  a	
  CLIL	
  context.	
  International	
  Review	
  of	
  Applied	
  Linguistics,	
  50,	
  
135-­‐160.	
  

Lin,	
   Angel	
   M.	
   (2015).	
   Conceptualising	
   the	
   potential	
   role	
   of	
   L1	
   in	
   CLIL.	
   Language,	
   Culture	
   and	
   Curriculum,	
  
28(1),	
  74-­‐89.	
  

Lo,	
   Yuen	
   Yi,	
   &	
   Lin,	
   Angel	
   M.	
   Y.	
   (2019).	
   Curriculum	
   genres	
   and	
   task	
   structure	
   as	
   frameworks	
   to	
   analyse	
  
teachers’	
  use	
  of	
  L1	
   in	
  CBI	
  classrooms.	
   International	
   Journal	
  of	
  Bilingual	
  Education	
  and	
  Bilingualism,	
  
22(1),	
  78-­‐90.	
  

MacWhinney,	
   Brian	
   (2000).	
  The	
   CHILDES	
   project:	
   Tools	
   for	
   analyzing	
   talk	
   (3rd	
   ed.).	
   Mahwah,	
   New	
   Jersey:	
  
Lawrence	
  Erlbaum.	
  

Martínez-­‐Adrián,	
  María	
   (In	
   press).	
   ¿Los	
   juntamos?	
  A	
   study	
   of	
   L1	
   use	
   in	
   interactional	
   strategies	
   in	
   CLIL	
   vs.	
  
non-­‐CLIL	
  primary	
  school	
   learners.	
   International	
  Review	
  of	
  Applied	
  Linguistics	
   in	
  Language	
  Teaching	
  
(IRAL).	
  

Martínez-­‐Adrián,	
   María,	
   &	
   Gutiérrez-­‐Mangado,	
   M.	
   Juncal	
   (2015a).	
   L1	
   use,	
   lexical	
   richness,	
   accuracy	
   and	
  
complexity	
   in	
   CLIL	
   and	
   non-­‐CLIL	
   learners.	
   Atlantis,	
   Journal	
   of	
   the	
   Spanish	
   Association	
   for	
   Anglo-­‐
American	
  Studies,	
  37,	
  175-­‐197.	
  

Martínez-­‐Adrián,	
  María,	
  &	
  Gutiérrez-­‐Mangado,	
  M.	
  Juncal	
  (2015b).	
  Is	
  CLIL	
  instruction	
  beneficial	
  both	
  in	
  terms	
  
of	
   general	
   proficiency	
   and	
   specific	
   areas	
   of	
   grammar?	
   Journal	
   of	
   Immersion	
   and	
   Content-­‐Based	
  
Language	
  Education,	
  3(1),	
  51-­‐76.	
  

Martínez-­‐Adrián,	
  María,	
  Gallardo-­‐del-­‐Puerto,	
  Francisco,	
  &	
  Basterrechea,	
  María	
  (2019).	
  On	
  self-­‐reported	
  use	
  of	
  
communication	
  strategies	
  by	
  CLIL	
  learners	
  in	
  primary	
  education.	
  Language	
  Teaching	
  Research,	
  23(1),	
  
39-­‐57.	
  



MARTÍNEZ-­‐ADRIÁN	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 76	
  

Martínez-­‐Adrián,	
  María,	
  Gutiérrez-­‐Mangado,	
  M.	
  Juncal,	
  &	
  Gallardo-­‐del-­‐Puerto,	
  Francisco	
  (2019).	
  Introduction.	
  
Special	
   issue	
   on	
   L1	
   use	
   in	
   content-­‐based	
   and	
   CLIL	
   settings.	
   International	
   Journal	
   of	
   Bilingual	
  
Education	
  and	
  Bilingualism,	
  22(1),	
  1-­‐4.	
  

Muñoz,	
   Carmen	
   (2007).	
   Cross-­‐linguistic	
   influence	
   and	
   language	
   switches	
   in	
   L4	
   oral	
   production.	
   Vigo	
  
International	
  Journal	
  of	
  Applied	
  Linguistics,	
  4,	
  73-­‐94.	
  

Navés,	
   Teresa,	
   &	
   Victori,	
   Mia	
   (2010).	
   CLIL	
   in	
   Catalonia:	
   An	
   overview	
   of	
   research	
   studies.	
   In	
   David	
  
Lasagabaster,	
   &	
   Yolanda	
   Ruiz	
   De	
   Zarobe	
   (Eds.),	
  CLIL	
   in	
   Spain:	
   Implementation,	
   results	
   and	
   teacher	
  
training	
  (pp.	
  30-­‐54).	
  Newcastle	
  upon	
  Tyne,	
  United	
  Kingdom:	
  Cambridge	
  Scholars	
  Publishing.	
  

Pavón,	
  Víctor,	
  &	
  Ramos,	
  María	
  del	
  Carmen	
  (2019).	
  The	
  controversial	
  relationship	
  between	
  the	
  L1	
  and	
  CLIL:	
  
Analysing	
   students’	
   L1	
   communication	
   strategies	
   in	
   classroom	
   interaction.	
   International	
   Journal	
   of	
  
Bilingual	
  Education	
  and	
  Bilingualism,	
  22(1),	
  35-­‐48.	
  

Pinter,	
   Anna	
   M.	
   (2007).	
   Some	
   benefits	
   of	
   peer-­‐peer	
   interaction:	
   10-­‐year-­‐old	
   children	
   practicing	
   with	
   a	
  
communication	
  task.	
  Language	
  Teaching	
  Research,	
  11,	
  189-­‐207.	
  

Pladevall	
   Ballester,	
   Elisabet,	
  &	
  Vraciu,	
   Alexandra	
   (2017).	
   Exploring	
   early	
   EFL:	
   L1	
   use	
   in	
   oral	
   narratives	
   by	
  
CLIL	
   and	
  non-­‐CLIL	
  primary	
   school	
   learners.	
   In	
  María	
   del	
   Pilar	
  García	
  Mayo	
   (Ed.),	
  Learning	
   foreign	
  
languages	
   in	
  primary	
   school:	
  Research	
   insights	
   (pp.	
  124-­‐148).	
  Bristol,	
  United	
  Kingdom:	
  Multilingual	
  
Matters.	
  

Ross-­‐Feldman,	
  Lauren	
  (2005).	
  Task-­‐based	
  interactions	
  between	
  second	
  language	
  learners:	
  Exploring	
  the	
  role	
  of	
  
gender.	
  Unpublished	
  doctoral	
  dissertation.	
  Georgetown	
  University.	
  	
  

Ruiz	
   de	
   Zarobe,	
   Yolanda	
   (2008).	
   CLIL	
   and	
   foreign	
   language	
   learning:	
   A	
   longitudinal	
   study	
   in	
   the	
   Basque	
  
Country.	
  International	
  CLIL	
  Research	
  Journal,	
  1(1),	
  60-­‐73.	
  

Shintani,	
  Natsuko	
  (2012).	
  Repeating	
  input-­‐based	
  tasks	
  with	
  young	
  beginner	
  learners:	
  The	
  role	
  of	
  the	
  teacher.	
  
RELC	
  Journal,	
  43,	
  39-­‐51.	
  

Shintani,	
  Natsuko	
   (2014).	
  Using	
   tasks	
  with	
   young	
  beginner	
   learners:	
  The	
   role	
   of	
   the	
   teacher.	
   Innovation	
   in	
  
Language	
  Learning	
  and	
  Teaching,	
  8,	
  279-­‐294.	
  

Storch,	
  Neomi,	
  &	
  Aldosari,	
  Ali	
   (2010).	
  Learners’	
  use	
  of	
   first	
   language	
   (Arabic)	
   in	
  pair	
  work	
   in	
  an	
  EFL	
  class.	
  
Language	
  Teaching	
  Research,	
  14,	
  355-­‐375.	
  

Storch,	
  Neomi,	
  &	
  Wigglesworth,	
  Gillian	
  (2003).	
   Is	
   there	
  a	
  role	
   for	
   the	
  use	
  of	
   the	
  L1	
   in	
  an	
  L2	
  setting?	
  TESOL	
  
Quaterly,	
  37(4),	
  760-­‐770.	
  

Swain,	
  Merril,	
  &	
  Lapkin,	
  Sharon	
  (2000).	
  Task-­‐based	
  second	
  language	
  learning:	
  The	
  uses	
  of	
  the	
  first	
  language.	
  
Language	
  Teaching	
  Research,	
  4,	
  251-­‐274.	
  

Togini,	
  Rita,	
  &	
  Oliver,	
  Rhonda	
  (2012).	
  L1	
  use	
  in	
  primary	
  and	
  secondary	
  foreign	
  language	
  classrooms	
  and	
  its	
  
contribution	
   to	
   learning.	
   In	
   Eva	
   Alcón	
   Soler,	
   &	
   María	
   del	
   Pilar	
   Safont	
   Jordá	
   (Eds.),	
  Discourse	
   and	
  
learning	
  across	
  L2	
  instructional	
  contexts	
  (pp.	
  53-­‐78).	
  Amterdam,	
  The	
  Netherlands:	
  Rodopi.	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  



THE	
  USE	
  OF	
  PREVIOUSLY	
  KNOWN	
  LANGUAGES	
  

E-­‐JournALL	
  7(1)	
  (2020),	
  pp.	
  59-­‐77 77	
  

María Martínez-Adrián, The University of the Basque Country 
maria.martineza@ehu.eus  

  

EN María Martínez-Adrián is an Associate Professor at the UPV/EHU and a member of the Laslab group 
(www.laslab.org), highlighted by the Basque Government for excellence in research. She has co-edited the volume 
Contemporary Approaches to Second Language Acquisition published by John Benjamins (Spanish Association of 
Applied Linguistics award 2014), a special issue on L1 use in CLIL/CBI settings published by the International Journal 
of Bilingual Education and Bilingualism, and the monograph Crosslinguistic influence: From empirical evidence to 
classroom practice published by Springer. Her work has appeared in books published by Multilingual Matters and 
Springer, as well as in journals such as RESLA, IJES, ITL, JICB, LTR, EUJAL, IRAL and INJAL, among others. 
 

ES María Martínez-Adrián ejerce como Profesora Titular en la UPV/EHU y es miembro del Grupo Consolidado de 
Excelencia Laslab (www.laslab.org),. Ha coeditado el volumen Contemporary Approaches to Second Language 
Acquisition [Acercamientos contemporáneos a la Adquisición de Segundas Lenguas] publicado por John Benjamins 
(Premio de la Asociación Española de Lingüística Aplicada, 2014), una edición especial sobre el uso de la L1 en entornos 
AICLE/IBC publicada por el International Journal of Bilingual Education and Bilingualism [Revista internacional de 
educación bilingüe y bilingüismo], y la monografía Crosslinguistic influence: From empirical evidence to classroom practice 
[Influencia interlingüística: De la evidencia empírica a la práctica en el aula] publicada por Springer. Asimismo, cuenta con 
capítulos de libro publicados por Multiligual Matters y Springer, al igual que artículos en revistas académicas como 
RESLA, IJES, ITL, JICB, LTR, EUJAL, IRAL y INJAL, entre otras.  
 

IT María Martínez-Adrián è professoressa associata all'UPV/EHU e membro del gruppo Laslab (www.laslab.org), lodato dal 
governo basco per eccellenza nella ricerca. Ha co-curato il volume Contemporary Approaches to Second Language 
Acquisition pubblicato da John Benjamins (premio dell'Associazione Spagnola di Linguistica applicata 2014), un fascicolo 
speciale sull'uso di L1 in contesti CLIL/CBI pubblicato dall'International Journal of Bilingual Education and Bilingualism, e la 
monografia Crosslinguistic influence: From empirical evidence to classroom practice pubblicata da Springer. Il suo lavoro è 
apparso in libri pubblicati da Multilingual Matters e Springer, oltre che, tra l’altro, su riviste come RESLA, IJES, ITL, JICB, 
LTR, EUJAL, IRAL e INJAL.