Content uploaded by Veronica Benigno
Author content
All content in this area was uploaded by Veronica Benigno on Jul 06, 2017
Content may be subject to copyright.
How long does
it take to learn
a language?
Insights from
research on
language
learning.
May 2017
Veronica Benigno
John de Jong
Alistair Van Moere
Global Scale of English
Research Series
02 How long does it take to learn a language? Insights from research on language learning. May 2017
Contents
Learning as a non-linear process. Which variables to account for? 4
Prociencylevel 5
Motivation 5
Aptitude 6
Learning strategies 6
Learning context 6
Age 6
Learning time estimates: how long does it take to make progress? 7
Defense Language Institute Foreign Language Center 8
NationalCentreonImmigrantIntegrationPolicy 8
Council of Europe 8
Pearson 9
Conclusion 10
References 11
03 How long does it take to learn a language? Insights from research on language learning. May 2017
Executive summary
How long does it take to learn a language? Research
has shown that language is a non-linear process and
that a combination of individual and contextual factors
determines the learning journey and aects the time
each individual needs to make progress. Although
there is no unanimous consent as to how many hours
are needed to gain increasing language prociency,
attempts have been made to produce learning time
estimates - to help educators, institutions, and
ministries set realistic and attainable learning goals as
well as compare dierent programs.
Inthesectionsbelowwerstmentionseveralimportant
factors that impact the time it takes to learn a language, e.g.
motivationandstartingprociencylevel.Then,wepresent
available estimates of learning time and, in light of existing
researchevidence,makearecommendationofhowmany
hoursareapproximatelyneededforanaveragelearnerto
growinprociencyontheGlobalScaleofEnglish(DeJong&
Benigno,2017)andtheCEFR(CouncilofEurope,2001).Such
arecommendationshouldbetakenbythereaderasabasic
guideandadjustedtoreectthespeciccharacteristicsofthe
learning/teaching context.
04 How long does it take to learn a language? Insights from research on language learning. May 2017
Learning as a non-linear process.
Which variables to account for?
Language learning is non-linear, i.e. dynamic, not
uniform or entirely predictable (Larsen-Freeman, 1997).
A U-shaped learning curve, where performance initially
improves, then deteriorates, and eventually recovers is
documented in both rst language research (Lightbown,
1983) and second language research (Bowerman, 1982).
The U-shaped curve is observed in cases where practice
does not lead to improvement in performance as the
result of restructuring processes (McLaughlin, 1990). The
learning journey of each individual is unique, because
learning does not proceed at a regular and continuous
pace but rather goes through peaks and valleys,
improvement and backsliding. According to the typical
learning curve, peaks generally come at the start and at
the end of the learning process, whereas, in between,
learners tend to “get stuck” in what is referred to as a
“learning plateau” (Gass and Selinker, 2001). This means
that although score gains are expected as a result of
learning over time, some learners may experience static
phases or even temporary regression due to a number of
factors, such as lack of exposure, loss of motivation, etc.
Chapters 3, 4 and 5 of the CEFR describe progress in language
prociencyusingbothaquantitative(howmanytasksan
individualcanperform)andaqualitative(howwellanindividual
can perform these tasks) dimension. Chapter 6 points out
thatlearnersmayhaveunevenprolesandachievepartial
competencies(CouncilofEurope,2001,p.133).Dierent
componentsoflanguageabilitydevelopatdierentrates
andfollowdierenttrajectories,e.g.somelearnersmaybe
stronger in some activities, e.g. listening rather than speaking;
ordevelopsomecompetenciesmorequickly,e.g.phonetics
earlierthansyntax.Accordingtotheaction-orientedapproach
outlinedintheCEFR(CouncilofEurope,2001,p.8),learningis
dependent on a number of variables related to the individual
and to their learning experience.
05 How long does it take to learn a language? Insights from research on language learning. May 2017
Sincethesecondhalfofthelastcentury,researchhas
extensivelydiscussedtheimportanceofindividualdierences
(suchasmotivationandlearningstrategies)andcontextual
dierences(suchasexposuretolanguageinputandteaching
methodologies),particularlyintheareaofsecondlanguage
acquisition(Carrol,1963;Gardner,1985;Skehan,1989;
Spolsky,1989;Ellis,1994).Dierenttheoreticalmodelshave
beenproducedtodescribethewayindividualfeaturesinteract
withexternalfeatures.Thesemodels,oftenproducedwithin
dierentareasofresearchsuchaspsychology,linguistics,and
sociology,dierslightlyintheconstructstheypropose(Ellis,
1994)andhavenotprovidedacomprehensiveanddenitive
explanationoftheextenttowhichthedierentvariables
combinetoproduceagivenoutcome.Asanexample,Spolsky
(1989,p.28)presentsanon-hierarchicalmodelinwhichsocial
context,attitude,motivation,age,personality,capabilities,
previous knowledge, and learning opportunities interact -
determining both the linguistic and non-linguistic learner
outcomes. In what follows, we refer to some important factors
aectingalearner’sjourney,andtherefore,learningtime,
without claiming that this list is comprehensive.
Prociency level.TheCEFRisoneofthemostwidelyaccepted
frameworksofreferencetodescribelanguageprociencyand
itsprogress.Theframeworkdivideslanguageprociencyinto
sixmainlevelsfromA1toC2,oftenerroneouslyinterpreted
asallbeingofequalwidth.However,thelogitscaleunderlying
theCEFRscalerevealsthatitslevelsarenotequidistant-with
A2, B1, and B2 being about twice as wide as the A1 level. At the
other end of the framework, C2 has an undetermined width.
Theimplicationofthisobservationisthatlearnerswilltake
much longer to move from A2 to B1, than to move from A1 to
A2.Infact,atamoreadvancedstageoflanguageprociency,
learnersarerequiredtocarryoutawiderrangeofmore
challengingtasksandactivities.Thetimealearnerneedsto
improvewilldependontheirstartingprociencylevel.
Motivation.Studentswhoaredrivenbyaninternaldesireto
learnthelanguage(integrativemotivation)aregenerallymore
successful than students who have more practical reasons
tolearnthelanguage(instrumentalmotivation),e.g.getting
acerticatetogainaccesstoaninstitution(Gardner,1985).
Recentstudies(DörnyeiandSkehan,2005,citedbyEllisand
Larsen-Freeman, 2006) on the relationship between motivation
and language achievement have demonstrated that motivation
correlatesonlymarginallywithachievement-duetothefact
thatlanguagelearningisaverycomplexconstructtowhich
manydierentdimensionscontribute.Forthisreason,itis
arguedthatanadequatemodelofstudents’motivationshould
include the time factor in order to understand how motivation
changesunderthevariousinternalandexternalinuencesthat
individuals are exposed to.
06 How long does it take to learn a language? Insights from research on language learning. May 2017
Aptitude.Thisindividualfeaturecanbeassociatedwith
other learner cognitive features such as intelligence. Carroll
andSapon(1959)developedtheModern Language Aptitude
Test(MLAT),atestwhichmeasuresindividuals’aptitudefor
learningaforeignlanguage.Theyidentiedfourcomponents
oftheconstructofaptitude:(a)phoneticcodingability,(b)
grammaticalsensitivity,(c)rotelearningabilityforforeign
languagematerialsand(d)inductivelearningability.The
underlyingtheoryclaimsthatsomeonewithahigheraptitude
will take less time to achieve a given learning goal under
optimallearningconditions(Carrol,1971;2012).
Learning strategies.Strategiesareusedbystudents,implicitly
andexplicitly,whenapproachingataskandcaninhibitor
facilitatethelearningprocess.Theyvarydependingonthe
learnerprociencylevel,motivation,andlearningstyle.Green
andOxford(1995)classiedstrategiesintosixmaintypes:
metacognitive,e.g.self-monitoring;aective,e.g.anxiety
reduction;social,e.g.askingquestions;memory,e.g.grouping;
cognitive, e.g. summarizing; compensation, e.g. guessing
meanings. It is important that teachers help their students
understand their own learning and develop appropriate
strategies.
Learning context.Thecontextinwhichthelanguageis
learntplaysacrucialroleindeterminingthesuccessofthe
learningexperience.Itisintuitivelyeasytounderstandthat
learninginanimmersioncontext(asistypicalofsecond
languagelearning)yieldsmoreopportunitiestobeexposed
tothelanguage,speedingupthelearningprocess.Inversely,
research has shown that language learning via instruction
oftendoesnotprovideenoughexposuretoachieveuencyin
the target language. In an instructional context, the choice of
the teaching method is decisive to help students improve as
quicklyaspossible.Anumberofstudieshaveinvestigatedthe
eectofinstructionalmethodsonlanguagelearning,although
considerablecontroversystillexistsabouthowinstructioncan
bestfacilitatelanguagelearning(Ellis,2006).
Age.Ifsecondlanguageacquisitionresearchhasdemonstrated
thatearlylanguagelearningleadstobetterprociencyinthe
longrun(Singleton,1989,p.137),similarndingshavenot
beenfoundinforeignlanguagecontexts.Thatsaid,ithasbeen
shownthatadultlearnersdohaveanadvantageincarrying
outtaskswhicharecognitivelymoredemanding,e.g.tasks
involving metacognitive skills.
07 How long does it take to learn a language? Insights from research on language learning. May 2017
Learning time estimates: how long
does it take to make progress?
How long does it take to learn a language? This is
one of the most troublesome questions for most
practitioners. In view of what has been discussed in
the previous section, it is easy to understand that
too many variables come into play to provide a one-
size-ts-all answer. Below we present the learning
time estimates produced by the US Defense Language
Institute Foreign Language Center, the National
Centre on Immigrant Integration Policy, the Council
of Europe, and Pearson.
Defense Language Institute Foreign Language Center. An
estimateoflearningtimewasproducedbytheDefense
Language Institute Foreign Language Center(DLIFLC).The
DLIFLC, located in California, has the mission to “provide
culturallybasedforeignlanguageeducation,training,
evaluationandsustainmenttoenhancethesecurityofthe
nation” http://www.dlic.edu/about/mission-vision/The
Institutecategorizeslanguagesintofourlevelsofdiculty
forspeakersofEnglishasarstlanguage–agreatdeal
ofresearch(Ellis,2006)havingprovidedevidenceofthe
phenomenonofL1interference,whichplaysamajorrole
onL2acquisition).Generalprociency(correspondingto
level 3 of the Interagency Language Roundtable (ILR) Skill
Level Descriptions; http://www.govtilr.org/) is achieved in
26, 35, 48, 64, and 88 weeks in categories I, II, III, IV, and
Vlanguagesrespectively.CategoryIincludeslanguages
closelyrelatedtoEnglish;category2languagessimilarto
English;category3languageswithlinguisticand/orcultural
dierencesfromEnglish;categoryIVlanguageswith
signicantlinguisticand/orculturaldierencesfromEnglish;
andcategoryVlanguageswhichareexceptionallydicult
for native English speakers. Each week corresponds to
about 30 hours of instruction, accounting for a total of 780,
1050, 1440, 1920, and 2200 to reach what is B2+ on the
CEFR.TheeciencyoftheteachingmethodattheInstitute
isprobablyduetotheschools’teachingmethodologies,
including problem-solving approaches to course materials
andimmersions(Berbeco,2001).Itshouldalsobeadded
thattheDLIFLCmakesuseofhighlysophisticatedteaching
technology,weeklytrainingprogramsandevenisolation
immersionprogramsofupto5dayshttp://liberalarts.
utexas.edu/tlc/_les/prociencyconference/presentations/
DLI/1.pdfFinally,itmustbenotedthatthemotivation
of the learners is high, since their language learning
achievementsarerewardedbyraisingtheirsalary.Since
theDLIFLCestimateappliestothespeciccontextofthe
armyandwasproducedforlanguagelearnerswhoserst
language is English, it should be taken with great caution.
08 How long does it take to learn a language? Insights from research on language learning. May 2017
National Centre on Immigrant Integration Policy. A similar
estimatehasbeenprovidedbyMcHugh,Gelatt,&Fix(2007)
fortheNationalCentreonImmigrantIntegrationPolicy.In
theirreport,theyprovidecensus-basedestimatesonthe
numberofhoursofinstructionneededbymorethan12million
adultimmigrantsintheUSA(lawfulpermanentresidentsor
unauthorized immigrants) to pass the naturalization exam or to
“achievetheEnglishskillsnecessaryforcivicintegration[…]and,
inthecaseofyouthage17to24,theEnglishskillsnecessary
forpostsecondarystudy”(p.3).Accordingtothereport,an
averageof110yearlyhoursofinstructionforsixyears(fora
total of 660 hours on average) are needed to bring learners to
alevelofEnglishprociencyneededtopassthenaturalization
test(forthoseaged25andolder)ortobeginpostsecondary
education(foryouthsaged17to24).Thestudycitesaprevious
studyconductedin2000bytheMassachusettsInstitutefora
New Commonwealth, according to which between 85 and 150
hours are needed for adults without disabilities to move up
onelevelofEnglishprociencyundertheNationalReporting
Systemrange(seehttp://www.cal.org/caelanetwork/pdfs/
NRSFunctioningLevelTable.pdf for further information about
theNationalReportingSystem).
Council of Europe.TheworkcarriedoutbytheCouncilof
Europe to create a unit-credit, transparent, and coherent
systemtoscalelanguageprociencyinEuropestartedasfar
backasthe1970’swiththepublicationoftheThresholdLevel
(1974;1998),theWaystageLevel(1979;1998),andtheVantage
level(2001).Theseverydetaileddocumentsspeciedthe
functions,specicnotions,andgeneralnotionslearnerswould
beexpectedtobeabletoperformatagivenprociencylevel.
IntheThresholdSpecications(1998)theauthorsstatethat
“thereissomeevidencethat,withadequateguidance,absolute
beginners need an average of 375 learning hours - including
independentwork”(p.9)toachievethislevel,whichwilllater
be made to correspond to the B1 level on the CEFR scale. And
withreferencetotheWaystagelevel,theyassumethatthe
learningloadofthislevelwillbe“abouthalfofthatrequired
forThresholdLevel1990.Forbeginninglearnerswhoare
unable or unwilling to commit themselves right from the start
totheexpenditureoftimeandenergyrequiredforthehigher
objective,Waystage1990maybeanacceptablealternative”
(p.9).AshortwhileafterthepublicationoftheThreshold,
Waystage,andVantageSpecications,theCouncilofEurope
made the revised draft of the unpublished Breakthrough
Specicationspubliclyavailablehttps://www.coe.int/t/dg4/
linguistic/Source/FinalBreakthrough%20specication_6Nov01.
rtfOnpage11ofthisdocumentwereadthat“Thediversity
ofthetargetgroups[…]makesitdiculttoassessthelength
ofstudyrequiredtoreachBreakthrough.Adultlearners
with extensive previous language learning experience will
dosomuchmorequicklythanimmigrantsfromapeasant
background with, perhaps, no previous schooling. As a rough
09 How long does it take to learn a language? Insights from research on language learning. May 2017
approximation,thelearningloadmaybeestimatedat,say,some
80-100hoursoftuition.Itwillinanycasebeclearfromthe
specicationitselfthattherewillbeconsiderablevariationinwhat
a language learner who has reached this target will in fact be
capable of doing with what he or she has learnt.”
Pearson.AspartoftheGlobalScaleofEnglish(henceforth:
GSE)researchproject,Pearsonhascarriedoutsomeinitial
investigation into the relationship between learning time and
prociencydevelopment.The GSE is a linear transformation
of the logit scale underlying the descriptors developed
by North (2000) to describe the prociency levels of the
CEFR. The scale, ranging from 10 to 90, was rst used as the
reporting scale for the Pearson Test of English Academic
(Pearson, 2010a) and validated by aligning it to other
international prociency scales such as IELTS and TOEFL
(De Jong, 2009; (De Jong & Benigno, 2017; De Jong & Zheng,
2016; Pearson, 2010a).AstudycarriedoutusingtheVersant
Englishtest(Pearson,2010b)providesevidenceofsignicant
gains in performance between the pre- and post-tests during a
three-weekimmersionprogramme.Followingthisstudy,Pearson
iscurrentlyworkingtocollectadditionalstudentdatatohelp
addressthequestionofhowlongittakestolearnalanguage-in
ordertoacquireagreateramountofevidenceacrossavariety
ofinstructionalcontexts,e.g.youngandadultlearners,dierent
prociencylevels,etc.Table1providesanestimateproduced
byadaptingtheDLIFLCestimateinthelightofexperiencewith
learnersatourEnglishschools.Animportantdierenceismade
between slow learners and fast learners. Fast learners learn in an
idealscenario.Theytakebenetfromanumberofindividualor
context-relatedtraits,forexampletheyarehighlymotivatedand
theirrstlanguageisnottoodistantfromEnglish.The estimate
provided for fast learners claims that learners will take
about 760 hours to enter the B2 CEFR level (at 59 on the
GSE scale).Thisestimateisinlinewiththoseprovidedbythe
DLIFLCguidelinesforcategoryIaswellaswiththoseproduced
bytheNationalCentreonImmigrantIntegrationPolicyandthe
Council of Europe. Note that these are active learning hours,
i.e.,timeexplicitlydevotedtolearningthelanguage,through
instruction and exercises.
CEF
CSE
Hours per level
Total cumulative
Hours required
3 pt GSE-gain
Start Finish Start Finish Fast Slow Fast Slow Fast Slow
<A1 A1 10 22 95 480 95 480 24 120
A1 A2 22 30 95 290 190 770 36 109
A2 B1 30 43 190 616 380 1386 44 142
B1 B2 43 59 380 1109 760 2495 71 208
B2 C1 59 76 760 1996 1520 4491 134 352
*Actual hours will depend on individual factors such as L1,
motivation,intensityofstudy,etc.
TABLE 1
Pearson’sestimateofnumberof
hoursperincreasingprociency
010 How long does it take to learn a language? Insights from research on language learning. May 2017
Conclusion
How long it takes to learn a language is not an easy
question to answer. It is important that practitioners
understand the complexity of factors which aect
learning time in order to design their teaching activity
to meet the specic needs of the learner and the
learning context. There are no shortcuts in learning a
new language but realistic objectives can be achieved
by making sure learners achieve the minimum required
learning hours estimated at each level. Therefore, we
would like to make the following recommendations
concerning the setting of learning goals in relation to
time.
• Keepinmindthatthetimeittakestoachieveprociency
gains depends on both learner-related and external factors
• Somefeaturesaectinglearningtimearemoredicult
to predict than others, for example individual learning
habits.However,stakeholdersplayanimportantrolein
creating optimal conditions for learning, for instance
bymakinginformeddecisionsaboutteachingmaterials,
pedagogical approaches, assessment resources, feedback,
etc.
• Before setting learning goals in relation to time,
itisimportanttoreectonthetypeoflearnerand
their opportunities of exposure to the target language,
thecurriculumrequirementssetbyaspeciccountry,the
alignment between teaching materials and expected
outcomes,andmoregenerally,anypredictablevariable
whichmayhaveanimpactonthelearningresults
• Institutions should maximize the opportunities for active
learning. For example, learners should be helped to develop
critical-thinking,engageintask-basedactivities,reecton
theirownattitudesandmotivations,andideally,spendsome
personal time learning outside the classroom in
more informal settings
• Finally,itisimportantthatstakeholdersinvolvedinsetting
goalshaveanunderstandingofthecomplexityoffactors
aectingthespeedoflearningandarecommittedtosetting
realistic goals.
011 How long does it take to learn a language? Insights from research on language learning. May 2017
References
Berbeco,S.(2016)(eds).Foreign Language Education in America. Perspectives
from K-12, University, Government, and International Learning. Palgrave
Macmillan, London
Bowerman,M.(1982).Starting to talk worse: clues to language acquisition
from children’s late speech errors.InS.StraussandR.Stavy(eds),
U-ShapedBehavioralGrowth,pp.10145.NewYork:AcademicPress
Carroll,J.,B.,andSapon,S.,M.(1959).Modern Language Aptitude Test, Form A.
NewYork:ThePsychologicalCorporation
Carroll,J.,B.(1963).A model of school learning.TeachersCollegeRecord,
64, pp. 723-733
Carrol,J.,B.(1971).Implications of Aptitude Test Research and Psycholinguistic
Theory for Foreign Language Teaching. Paper presented at XVIIth
International Congress, International Association of Applied
Psychology,Liege,Belgium,July27,1971.
Available at http://les.eric.ed.gov/fulltext/ED057071.pdf
Carrol,J.,B.(1989).The Carroll Model: A 25-Year Retrospective and Prospective
View.InEducationalResearcher,Vol.18,No.1(Jan.-Feb.,1989),
pp. 26-31
CouncilofEurope(2001).Common European Framework of Reference for
Languages: Learning, teaching, assessment.CambridgeUniversityPress,
Cambridge
DeJong,J.H.A.L.(2009).Unwarranted claims about CEF alignment of some
international English language tests.PaperpresentedatEALTA
Conference,June2009.RetrievedMay,5,2017,from
www.ealta.eu.org/conference/2009/docs/friday/John_deJong.pdf
DeJong,J.&Benigno,V.(2017).Alignment of the Global Scale of English
to other scales: the concordance between PTE Academic, IELTS, and TOEFL.
Available at https://prodengcom.s3.amazonaws.com/GSE-Alignment
other-scales.pdf
DeJong,J.H.A.L.,&Zheng,Y.(2016).Linking to the CEFR: validation using
a priori and a posteriori evidence.In:Banerjee,J.,&Tsagari,D.(Eds),
ContemporarySecondLanguageAssessment.London:Bloomsbury
Academic, pp. 83-100
Ellis,N.,C.(2006).Selective attention and transfer phenomena in L2 acquisition:
contingency, cue competition, salience, interference, overshadowing,
blocking, and perceptual learning.InAppliedLinguistics,Vol.27(2),
pp. 164-194
Ellis,N.,C.&Larsen-Freeman,D.(2006).Language Emergence: implications for
Applied Linguistics – Introduction to the special issue. In Applied
Linguistics,vol.27(4),pp.558-589
Ellis,R.(1994).The Study of Second Language Acquisition. Oxford:
OxfordUniversityPress.
Ellis,R.(2006).Current Issues in the Teaching of Grammar: An SLA Perspective.
InTESOLQuarterly,vol.40(1),pp.83-107
012 How long does it take to learn a language? Insights from research on language learning. May 2017
Gardner,R.,C.(1985).Social Psychology and Second Language Learning.
The role of attitudes and motivation. London: Edward Arnold
Gass,S.&Selinker,L.(2001).Second Language Acquisition:
An Introductory Course.SecondEdition.LawrenceErlbaumAssociates
Green,J.M.,&Oxford,R.L.(1995).A closer look at learning strategies,
L2prociency,andgender.InTESOLQuarterly,29(2),261-297
Larsen-Freeman,D.(1997).Chaos/Complexity Science and Second Language
Acquisition.InAppliedLinguistics,vol.18(2),pp.141-165
Lightbown,P.(1983).Exploring relationships between developmental and
instructional sequences in L2 acquisition.InH.Seliger&M.H.Long
(Eds.),Classroomorientedresearchinsecondlanguageacquisition,
pp.217–243.Rowley,MA:NewburyHouse
McHugh,M.,Gelatt,J,andFix,M.(2007).Adult English language instruction
in the United States: determining need and investing wisely. National
CentreonImmigrantIntegrationPolicy.Availableat:
le:///C:/Users/verobeni/Downloads/NCIIP_English_Instruction073107.
pdf
McLaughlin,B.(1990).Restructuring.InAppliedLinguistics,vol.11(2),
pp. 113-128
North,B.(2000).The Development of a Common Framework Scale of Language
Prociency.Berlin,NewYork:PeterLang
Pearson(2010a).Aligning PTE Academic Test Scores to the Common European
Framework of Reference for Languages. Available at
http://pearsonpte.com/wp-content/uploads/2014/07/Aligning_PTEA
Scores_CEF.pdf
Pearson(2010b).The Use of the Versant English Test as a Measure of Score
Improvement. Available at https://www.pearsonhighered.com/versant/
assets/VET_Score_Improvement.pdf
Singleton,D.(1989).Language Acquisition. The age factor. Clevedon:
Multilingual Matters
Skehan,P.(1989).IndividualDierencesinSecond-LanguageLearning.
London: Edward Arnold
Spolsky,B.(1989).Conditions for Second Language Learning. Oxford:
OxfordUniversityPress
VanEk,J.,A.&Trim,J.L.M.(1998).Threshold1990.
CambridgeUniversityPress,Cambridge
VanEk,J.,A.&Trim,J.L.M.(1998).Waystage1990.
CambridgeUniversityPress,Cambridge
VanEk,J.,A.&Trim,J.L.M.(2001).Vantage1990.
CambridgeUniversityPress,Cambridge