INFORMATION TO
USERSThismanuseripthasbeenreproducedtromthemicrofilmmaster. UMI
filmsthetext directly fromthe original or copysubmitted. Thus,
sornethesisanddissertation copies are in typewriter face, while
others may be trom any type ofcomputer printer.Thequalityof
thisreproductioni5dependent uponthequalityof thecopy submitled.
Broken or indistinct print,colored or poor quality illustrationsand
photographs, print bleedthrough, substandard margins, and
improperalignment can adversely affect reproduction.lntheunlikely
avent that theauthor did not sendUMI acompletemanuscriptand there
are missing pages, these will be noted. Also, if
unauthorizedcopyright material had to be removed, a note will
indicate the deletion.Oversize materials (e.g., maps, drawings,
chans) are reproduced bysedioning the original, beginning at the
upper left-hand(".omer andcontinuingtrom left to right in equal
sections with small overlaps.Photographs induded in the original
manuscript have been reproducedxerographicaJly in this copy. Higher
quality 6" x 9" black and whitephotographie prints are available
for any photographs or illustrations appearingin this copy for an
additional charge. Contact UMI directly to order.Bell &Howell
Information and Leaming300 North Zeeb Road, Ann Arbor, MI
48106-1346 USA800-521-0600NOTE TO USERSThis reproduction is the
best copy available.DMI-.An Input Enhancement Study with ESL
Children:Effects on the Acquisition of Possessive
DeterminersJoannaL. WhiteA Thesis Submitted to the Faculty
ofGraduate Studies and Research inPanial Fulfillment of the
Requirements of theDegree of Doctor of PhilosophyDepartment of
Second Language EducationMcGill University, MontrealNovember, 1996
Joanna L. White1+1National Libraryof CanadaAcquisitions
andBibliographie Services395 Wellington StreetOttawa ON
K1AON4CanadaBibliothque nationaledu CanadaACOlJisitions etserVices
bibliographiques395. rue WellingtonOttawa ONK1A ON4CanadaYour hie
Votre refrenr:eOur fileNorre rsfsrsnceThe author has granted a
oon-exclusive licence allowing theNational Library of Canada
toreproduce, loan., distribute or sellcopies of this thesis in
microfonn,paper or electronic formats.The author retains ownership
of thecopyright in this thesis. Neither thethesis nor substantial
extracts from itmay be printed or othe1Wisereproduced without the
author' spemusslon.L., auteur a accord une licence nonexclusive
pennettant laBibliothque nationale du Canada dereproduire, prter,
distribuer ouvendre des copies de cette thse sousla forme de
microfiche/film., dereproduction sur papier ou sur
fonnatlectronique.L'auteur conserve la proprit dudroit d'auteur qui
protge cette thse.Ni la thse ni des extraits substantielsde
celle-ci ne doivent tre imprimsou autrement reproduits sans
sonautorisation.0--612-44663-8CanadaThis dissertation is dedicated
to the memory of my tther,LeRoy P. L o n d o n ~who shared his love
of books with everyone he knew.AcknowledgementsMany people have
made important contributions to this doctoral research study.
1would like to begin by thanking Dr. Nina Spada, my supervisor, for
the intellectualguidance and moral support she has offered at every
stage ofthis research. Herrecommendations have strengthened the
study in countless ways, and her high standardsfor conducting and
reporting classroom...based SLA research are reflected in
thisdocument. Equally important, her enthusiasm for the study kept
me going when the dataseemed overwhelming and when other
responsibilities slowed progress. 1am deeplygratefu] for her
help.1would also like to thank the members ofmy committee, Dr.
Janet D o n i n ~ Dr.Mary Maguire; Dr. Lydia White, who shared
their expertise at many points along theway. Their contributions ta
the design ofthe study and the interpretation of the resultshave
been indispensable, and their attention to detail is much
appreciated.The Concordia-McGill Research Group has played a
central raie in supportingthis research. *A very special thanks
goes to Dr. Patsy Lightbown, whose wise counsel,encouragement, and
generosity in ail matters have been greatly appreciated over
theyears. With respect to this study, she shared the resources
ofthe research team during thecollection, coding and transcription
of the data and alwaysfound time to discuss myfindings and my
interpretations of them. 1would like to express particular thanks
toRandall Halter, whose guidance, meticulous work, and kindness
were essential tocanying out the statistical analyses of the data.
1wouJd also like to thank Leila Ranta forsharing her knowledge of
SLA and her experience in classroom-based research at everystep
during the planning, implementation, and interpretation ofthis
study, and for herhumour and emotional support during sorne of the
inevitable difficult times. Thank youto everyone else in the
Research Group forYOUf many suggestions and for your interest. A
number of research assistants and graduate students at McGill and
ConcordiaUniversities have contributed in different ways to helping
me organize the vast amountsof data l collected 1would like to
thank each ofthem individually: Patrick Berger, JudyCook, Pamela
Craven, Tamara Loring, Kay Powell, Martha Trahey,My colleagues at
the TESL Centre at Concordia University have been verysupportive
and cooperative throughout my doctoral studies. l greatly
appreciate theirconfidence in me and their willingness to make
adjustrnents so that l could complete thisresearch. Others in the
university administration have shown flexibility at various
times,and l thank them sincerely for this.l would aiso like to
thank the Grade 6 students, teachers, and principal in the"school
outside ofMontrear' where this study took place. For five months
in1993, theyopened their classrooms to me and shared the excitement
and challenge of teaching andleaming English. The opportunity to
spend this time with themprovided a rich contextin which to
interpret the data. In particular, 1would like to thank the teacher
of GroupE+, \vhose enthusiasm for reading stories to her students
facilitated the implementationof the "book flood" and enriched the
experience immeasurably.My deepest gratitude is reserved for my
husband, Bob, and my daughters, Katbyand Andrea. Without their
unwavering support and encouragement, 1would Dever havehad the
strength to finish this project. Thank you again.*This research was
funded in part through the following grants: McGill Faculty
ofGraduate Studies and Research Special Funds # 3 8 8 5 ~ SSHRC #
410..93-0899 and # 410-96-0520; fCAR # 96ER0153.IfAbstractIn this
c1assroom-basedthe effects of input enhancement on the
acquisitionof a linguistic featureknown to be problematic to
francophone childrenleamingEnglish.were investigated. The research
questions were: 1) Can L2 leamers benefit
tromtypographicallyenhancedinput intheiracquisitionof
thirdpersonsingularpossessivedeterminers? 2) Is typographically
enhanced input more effective than unenhancedinput? 3) [s
typographically enhanced input more effective \vhen combined with a
booktloodH?To investigate these questions. three treatment
conditions were implemented withGrade6ESLlearners. Groups Eand E+
receiveda typographicallyenhanced inputflood. Thisdid notinclude
expl icil referencetothelearners' LI nor was
apedagogicalrulepresented at any time. [n additiontothe
typographically enhanced GroupE+\Vas exposed to extensive reading
and Iistening activities. To ensure that ail groupsin thestudy were
exposed to wrinen input containing the target tatures, Group
Ureadunenhanced versions of the texts read by the other two groups.
Aimmediate anddelayed posttest design was used in this
quasi-experimental study.Results indicatedthat aIl three
instructional treatments improvedthe learners'acquisition of the
target formsand that thosein GroupE+received the greatest
apparentbenefits. At theimmediateleamers inGroupE+outperfonnedthose
intheother two groups on wrinen tasksdesigned tomeasure their
ability torecognize correctinstances of the target faons. Leamersin
Group E+ also outperformed the others on an111oral production task.
On the delayed posttest five\veeks l a t e r ~ however, Groups E
and Uhad caught up with Group E+, and most of the bet\veen-group
differences haddisappeared.The finding that ail leamers
hadsignitcantlyincreasedintheiraccuracyanddevelopment of
possessivedeterminersimmediatelyfollo\\inginstructionsuggests
thatthe target formswere equally salient to theleamersin the three
groups. That ail learnerscontinued to improve but did not reach
advanced developmental stages, however,suggeststhat thesalienceof
thesefeatures intheinput maynot have beensufficientlyexplicit. The
results are discussed in terms ofthe potential need for more
explicitinstructionintheacquisition of thirdpersensingular
possessivedeterminers. Thismaybe particularly important because of
substantial ditIerences in the way inwhich gender ismarked in
English and French.IVRsumCette tude mene en classelangue seconde,
est une enqute sur leseffets d'input mis en vidence sur
racquisition du dtenninant possessif: lmentlinguistique qui s'avre
difficile pour les apprenants francophones. Les questions
derecherche sont les suivantes: 1) Les apprenants peuvent-ils
bnficier d'un input mis envidence par des moyens typographiques
dans l'acquisition du dterminant possessif dela troisime personne
du singulier Chis et her)? 2) L'input typographiquement mis
envidence est-il plus efficace pour l'apprentissage que l'input qui
nel'est pac;? 3) L'inputtypographiquement mis en vidence est-il
plus efficace quand il est associ l'accsaccru aux livres et aux
activits de lecture?Trois conditions de traitement ont t retenues
pour mettre en oeuvre cette tudeauprs d'lves en sixime anne du
primaire. Les groupes E et E+ ont lu des textes danslesquels les
pronoms ont t mis en vidence typographiquement. Dans
cesdeLLxconditions,il n'y a eu aucune rfrence explicite la langue
maternelle ni aux rglesgrammaticales. De plus, le groupe E+ a
particip a un grand nombre d'activits delecture et d'coute. Pour
pouvoir distinguer entre les effets eventuels de la mise envidence
et de la frquence de l'exposition l'iment linguistique vis, le
group U a lules mmes textes lus par les groupes E et E+ dans
lesquelsil n'y avait aucune mise envidence. Le plan de cette tude
quasi-experimentale comprenait trois volets: des pr-tests, des
post-tests, et des post-tests diffrs donns plusieurs semaines aprs
la fin dutraitement.vLes rsultats dmontrent que les lves exposs
au.x trois conditions ont progressdans1- acquisition du dtenninant
possessif. Ce sont les lves du group E+ quisemblaient faire le plus
grand progrs immdiatement aprs le traitement. C e p e n d a n t
~aux post-tests diffrs les groupes E et U ont rattrap le groupe
E+puisque la plupart desdiffrences existant entre les groupes
avaient disparu.Letit que tousles groupes ont fait du progrs
significatif dans leur rendementaux post-tests tout de suite aprsle
traitement laisse supposer que les trois groupes ontperu dufaon
gale rlmentlinguistique en question. Que tousles apprenants
aientcontinu s"amliorer sans toutefois arriver aux tapes les plus
avances fait supposerquelafrquence et la mise en vidence ne sont
pas suffisantes pour amener lesapprenants matriser cet lment
linguistique. Les rsultats de cet tude dbouchent surune proposition
pour un enseignement explicite des fonnes du dterminant possessif
dela troisime personne du singulier. Cela prendune importance
particulire en raison dufait quele genre grammatical est indiqu
diffremment en ce qui a trait au dtenninantspossessifs en anglais
et en franais.VlTable of ContentsAcknowledgementsAbstractRsumTable
of contentsList of tablesList of figuresChapter 1: Input and Second
Language Acquisition1IIIVvuXlXVI1.0 Introduction 11.1 The role
ofthe Iinguistic environment in second language acquisition 61.1.1
Defining input and intake 61. 1.2 The raie of input 71.1.2.1The
innatist perspective 71.1.2.2The cognitive perspective 111.2
Interlanguage knowledge 121.3 Attending to the input 161.3.1 Tomlin
and Villa 171.3.2 Schmidt 181.3.3 Robinson 201.4 Describing input
in the second language c1assroom 211.4. 1 Input from the teacher
211.4.2 Input From other learners 221.4.3 Input from books 241.5
Improving the quality of classroom input empirical studies ofbook
floods 301.6 Problems arising from exclusively meaning-based
instruction 321.7 Chapter summary 35Chapter 2: Form-focused
Instruction and Second Language Acquisition 382.0 introduction2. 1
Theoretical framework2.2 Classroom-based empirical research2.2. 1
Studies with children and adolescents2.2.1.1French immersion
studies2.2. 1.2Intensive ESL2.2.2 Studies with adults2.2.2.1Visual
enhancement2.2.2.2Input processing2.2.2.3Garden path
studies2.2.2.4ImplicitlExplicit feedbackVII3838434343495858636466")
"'"Chapter summary 69 .... -'Chapter 3: Target Features. Research
Questions and Hypotheses 713.0 Introduction 713.1 The acquisition
of pronouns and possessive detenniners 713.1.1 Theoretical issues
713.1.2 Empirical evidence 763.2 Investigating the acquisition of
possessive detenniners throughenhancement 933.2.1 Research
questions and hypotheses 94Chapter 4: t\tlethodol0ID" 1014.0
Introduction 1014.1 Research context lOI4.1.1 Background 1014.1.2
School 1054.2 Participants 1084.3 Design 1104.4 Research schedule
1104.5 Treatment materials III4.5.1 EnhancedJunenhanced input
materials 1124.5.2 Materials for book flood (intensified exposure
to books) 1154.5.2.1 Selection ofbooks 1164.5.2.2Book tlood
activities 1174.6 Assignment of treatrnent conditions 1224.7
Classroom observations during study 1234.8 Measures 1244.8.1
Baseline tests 1244.8.2 Measures of second language development
1254.8.2.1Grammaticalityjudgement tests 1264.8,2.2Multiple choice
test 1314.8.2.3Oral production task 1334.8.2.4Independent measure
of general abilities in English 1354.8.3 Enhancement activities
questionnaire 135Chapter 5: Analyses and Resulu5.0 Introduction
1395.1 Paper and pencil tests 1405.1.1 Baseline tests 1405.1.2
Grammaticalityjudgement tests 1415.1.2. 1 Passage correction task
141Vill5.1.2.2Truth value task 1545.1.3 Multiple choice test
1545.1.3.1Initial pretest 1545. 1.3.2 Immediate pretes!.. immediate
posnest and delayed posttest scores:third person singular pronouns
and possessive detenniners 1555.1.4 Etfects ofthe treatment:
performance on paper and pencil tasks 1585.2 Oral production
measure 1585.2.1 Quantitative analyses. 1605.2.2 Qualitative
analyses 1785.2.3 Effeet of the treatment: perfonnance on the oral
production task 1855.3 Independent measure ofgeneral abilities in
English 185Chapter 6: Discussion and Conclusions 1886.0
Introduction 1886.1 Results in relation to the hypotheses 1886.2
Salience and explicimess 1896.2.1 Salience 1906.2.2 Explicitnt:ss
1996.3 Developmental sequence 2106.4 Limitations ofthe study 2146.5
Contributions and implications for future research 215References
219Appendices 236Appendix A Classroom observation coding sheet
236Appendix B Sample of treatrnent materials 239Appendix C Criteria
for the selection of books 255Appendix 0 Language measures
258Baldwin-Cartier test de classement 259Passage correction task
261Truth value task 263Multiple choice baseline test 265Multiple
choice pretest/posttest 267Oral production task 269Picture Set A
270Picture Set B 272Picture Set C 274MEQtest 277Enhancement
activities questionnaire 279Appendix E Anova tables 281IXAppendix F
Possessive detenniner stagesCoding sheet used for
individuallearnersFull description of stagesStage development of
individual learnersx302303305307List ofTablesTables in
text3.1reflexive, possessive pronouns, trom Quirk, Greenbaum, Leech
&Svartvik, 1972 72.... .,Accurate use of possessive detenniners
by francophone ESL -'.-adapted trom Lightbown and Spada (1990, p.
442) 89.., ....Acquisition sequence of possessive detenniners by
francophone ESL students .J ..)(Lightbown and Spada., 1990:441,
based on Zobl, 1985) 904.1 Number of intensive ESL groups by grade
and school board during the1992-93 academic year (adapted from
Watts and Snow..1993: 15) 1024.2 Number of boys and girls in each
group 1104.3 PC task: contexts required tor deviant third person
singular pronouns andpossessive determiners and well-formed tokens
supplied 1284.4 Pronouns and possessivt: detenniners tested in
truth value task 1314.5 Third person singular pronoun and
possessive determiner forrns tested in the24-ltem multiple cholce
test 1335.1 Mean scores ln percent and standard deviations by group
on Bald\vin CartierTest de classement 1415.3 Passage correction
task - mean out of24: grammatical corrections of deviantthird
person singular personal pronouns and JX)ssessive detenniners
1445.8 Passage correction task: grammatical correction of deviant
third personslngularmean out of 8 1475.12 Passage correction task,
mean scores out of 8: grammatical corrections ofdeviant possessive
determiners, contexts for his and her 1495.20 Passage correction
task - number of leamers correctly correcting deviantforms in PD
contexts (3 contexts for kin-different and 2 contexts for
bodyparts) 1525.21 Passage correction task - number ofleamers in
each group making one ormore changes to weil fonned and deviant
possessive detenniners 1535.22 Multiple choice test., initial
pretest- group means out of42 1555.24 Multiple choice test; group
scores out of24 1565.29 Mean number ofgrammatical and ungrammatical
uses ofthird personsingular subject and object pronouns and
possessive detenniners usedwith Picture Set A at immediate pretest
and immediate posttest 162Xl5.34 Mean nwnber ofgrammatical and
ungrammatical uses of third per:onsingular subject and object
pronouns and possessive detenniners used\vith Picture Set Bat
inunediate pretest and immediate posttest 1645.39 Accuracy ratios
for personal pronouns in picture description task 1705.42 Accuracy
ratios for Izis and her in picture description task 1725.45
Accuracy ratios for third persan singular possessive detenniners in
fourdomains in picture description task 1755.50 Developmental
sequence in the acquisition of the English agreementrule for
possessive detenniners: adapted from Zobl ( 1984, 1 9 8 5 )
~Lightbown and Spada (1990) 1795.51 Stage development, picture
description task (nurnber oflearners per group) 1815.52 Mean scores
in percentages and standard deviations by group on AIEQTt!st 1856.1
Percentage of leamers who produced two masculine pronominal
fonnsand percentage who producedt\VO feminine pronominal
forrosduring oral production task at end of 5-month intensive
program(Martens, [988, and current study 1956.2 Percentage
oflearners who used both !lis and her correctly at least onceduring
oral production task (Lightbo\\'TI and Spada, 1990, and current
srudy) 1966.3 Comparison of scores in % on possessive detenniner
items in passage 197correction lask (Manens, 1988, and current
study)6.4 Multiple choice delayed posttest - mean score in
percentages for 4 semanticdomains (4 test items per damain) 2006.5
Multiple choice test and passage correction task - mean scores
onpossessive detenniners in percentages at delayed posttest 2016.6
Picture description task - number of leamers using the with body
parts atdelayedposttest 203XIITables in Appendix E5.2 One-Way ANOVA
comparing scores on Baldwin-Cartier Test deClassement 2825.4
One-Way ANOVA comparing scores al pretes!, immediateand delayed
posttest on Passage Correction task: grammatical correctionsof
deviant third person singular pronouns and possessive jeterrniners
2825.5 ANCOVA comparing scores at immediate and delayed posttests
on passagecorrection task; grammatical corrections ofdeviant third
person singularpronouns and possessive detenniners;immediate
pretest scores as covariate 2835.6 Repeated measures ANOVA showing
effects of group and timeImmediate delayed posttest), passage
correction task 2845.7 Repeated measures ANOVA comparing scores on
grammatical correctionsof deviant third person singular pronouns
and possessive determinersat immediate and delayed posttests,
passage correction task 2845.9 One-Way ANOVA comparing scores al
pretest, immediate posnest, anddelayed posttest on passage
correction task: grammatical corrections of deviantthird persan
singular pronauns 2855.10 ANCOVA comparing scores at immediate and
delayed posttests on passagecorrection task; grammatical
corrections ofdeviant third persan singular immediate pretest
scores as covariate 2865.11 Repeated measures ANOVA comparing
scores on grammatical corrections ofdeviant third person singular
pronouns at immediate and delayed posttests,passage correction task
2865.13 One-VJay ANOVA comparing scores at pretest, immediate
anddelayed posttest on passage correction task: grammatical
correctionsof deviant masculine third person singular masculine
possessivedetenniners (his) 2875.14 One-Way ANOVA comparing scores
at pretest, immediateanddelayed posttest on passage correction
task: grammatical correctionsof deviant masculine third person
singular possessive detenniners (lzer) 2885.15 ANCOVA with
immediate pretest scores as covariate comparing immediateand
delayed posttest scores on passage correction task;
grammaticalcorrections of deviant masculine possessive determiners
(his) 2885.16 ANCOVA with immediate pretest scores as covariate
comparing immediateand delayed posttest scores on passage
correction task; grammaticalcorrections of deviant feminine
possessive detenniners (her) 2895. 17 Repeated measures ANOVA
comparing scores on grammatical correctionsof his at immediate and
delayed posttests, passage correction task
289xiii5.185.195.235.255.265.275.285.305.315.325.335.355.365.375.385.405.41Repeated
measures ANOVAcomparing scores on grammatical correctionsof her at
immediate and delayed posttests, passage correction task
290Statistics for trend analysis, passage correction task
291One-Way ANOVA comparing groups scores on multiple choice
initialpretest 291One-Way ANOVA comparing groups' scores on
multiple choice test at pretest,immediate posttest and delayed
posttest 292ANCOVA with immediate pretest scores as covariate
comparing immediateand delayed posttest scores on multiple choice
test 292Repeated measures ANOVA showing effects of group and time
(pretest,immediate posttest, delayed posttest), multiple choice
test 293Repeated measures ANOVA comparing scores on grammatical
correctionsof /21S al immediate and delayed posttests, passage
correction task 293ANCOYA comparing mean number of grammatical uses
of subject and objectp r o n o u n s ~ Picture Set A, immediate
pretest means as covariate 294ANCOYA comparing mean number of
ungrammatical uses of subject andabject p r o n o u n s ~ Picture
Set A, immediate pretest means as covariate 294ANCOVA comparing
mean number of grammatical uses of possessivedetenniners, Picture
Set A ~ immediate pretest means as covariate 294ANCOVA comparing
mean number of ungrammaticaluses of possessivedeterminers, Picture
Set A, pretest means as covariate, picture descriptiontask
295ANCeyA comparing mean number of grammatical uses of subject
andabject p r o n o u n s ~ Picture Set 8, immediate posttest means
as covariate 295ANceVA comparing mean number of ungrammatical uses
of subject andobject pronouns; Picture Set 8, immediate posttest
means as covariate 295ANCeVA comparing Mean number of grammatical
uses of possessivedetenniners, Picture Set B; immediate posttest
means as covariate 296ANCOVA comparing Mean number of ungrammatical
uses of possessivedetenniners, Picture Set B, immediate posttest
means as covariate 296One-Way ANOVA comparing groups' scores on
subject pronouns, picturedescription task 297One-Way ANQVA
acomparing groups scores on object pronouns, picturedescription
task 297xiv5.43 One-Way ANOVA comparing groups' scores on his
2985.44 One-Way ANOVA comparing groups' scores on her 2985.46
One-Way ANOVA comparing groups' scores on inanimate 2995.47 One-Way
ANOVA comparing groups' scores on kin-same 2995.48 One-Way ANOVA
comparing groups' scores on kin-different 3005.49 One-Way_ANOVA
comparing groups' scores on body parts 3005.53 One-way ANOVA
Comparing Groups' Scores on MEQ Test 301xvList of
Figures4.1Research schedule III5.1 Mean scores for third persan
singular pronouns and possessivepassage correction test 1455.2 Mean
scores for third persan singular passage correctiontask 1485.3 Mean
scores for !lis, passage correction task 1505.4 Mean scores for
/zer,passage correction task 1505.5 Mean scores for multiple choice
test 1575.6 Mean number of grammatical and ungrammatical subject
andobject pronouns, Picture Set A 1625.7 Mean number of grammatical
and ungrammatical possessivedeterminers, Picture Set A 1625.8 Mean
number of grammatical and ungrammatical subject andobject pronouns,
Picture Set B 1655.9 Mean number of grammatical and ungrammatical
possessivedetenniners, Picture Set B 1655.10 Accuracy ratios forhUi
and 11er, picture description task 1735.11 Accuracy ratios for
third person singular possessive determiners.four description task
1765.12 Stage development ofthird person singular possessive
determinersat three test sessions 1825.13 Developmental paths tor
third persan singular possessive determiners 1846.1 Stage
development of third person singular possessive experimental groups
and comparison group at delayed posttest,oral production task
198XV1Chapter 1Input and Second Language Acquisition1.0
IntroductionFor most of the history of secondlanguage teaching,
instrllctionalapproaches havebeen based on the asswnption that an
explicit focus on language foon is necessary for secondlanguage
(L2) acquisition to take place. In this view of language teaching,
isolated linguisticelements (e.g.sound contrasts, lexical items)
are the organizational units of thesyllabus and the subject matter
of individual lessons. Input is manipulated throughpedagogicaJ
practices which include grammar and vocabulary explanations,
dialoguedisplayerror correction., and fill-in-the-blank
exercises.For the last several decades., however., this
structure-by-structure approach tolanguage teaching bas been
questioned by a number of researchers who see the acquisition
ofL2knowledgeas an essentiallyimplicit processsimilartofirst
language(L1)acquisition.,which proceeds in a systematic and largely
predictable manner through exposure to rich
andvariedlinguisticinput thatis provided inhighJy contextualized
social interactionsforexample., Krashen., 1981., 1982; Dulay and
1973). Themorphemeacquisitionanddevelopmental sequencestudies in
tirst
andsecondlanguageacquisitionresearchwouldseemtoconfirmthehypothesis
that languagedevelopment is under thecontrol ofthelearner's
built-in syllabus and is basically unaffectedby instruction which
focuses explicitlyon targeted grammatical structures (Krashen.,
1977;1981). Although there has beensorne evidence suggesting that
extensive grammar practice and drilling may have an effect
byalteringthenatura! processes,
theseeffectshavebeenshowntobetemporaryinnature(Lightbown, 1983a,b;
and see discussion in Long, 1988).The structure-by-structure
approach has also been questioned by teachers andcurriculum
planners (e.g.Prabhu, 1987), who suggest that to help leamers
attain an accurateand fluent command of the target language, it
makes better sense to provide opportunities forthemto experience
language as a medium of communication than tomakeit anobject
ofstudy. Asa consequence, a number of nstructional
approacheshaveemergedwhichde-emphasizetheimportanceof
explicitteachingof linguisticfonnsandstress,thevalue ofexposure to
comprehensible input (CI) and the use ofthe L2 in
meaningfulinteraction. These include the Natural Approach (Krashen
and Terrell, 1983),comprehension-based second language programs
(e.g.1981; 1992a), andCommunicativeLanguage Teaching (CLT)
(Brurnfit, 1984; LittJewtX>2]PretestPost 1Post 2*includes
1student from Stage 7 at pretestFigure 5.13 Development paths for
third person singular possessive determiners184s..~..
50715.2.3Effect of the treatment: performance on the oral
production taskLike the paper and pencil the oral datadonot reveal
the bet\\"een-groupdifferences that were predicted by the
hypotheses, namely E+ > E >U. In Chapter thesefindings are
discussed.5.J Independeot measure of general abilities in
EnglishTheMEQ test was administered to thethreegroups duringthe
delayedposttestingsession. Mean scores and standard deviations
expressedinpercentages are showll in Table5.52 below. An ANOVA
indicated that the difTerences among the groups were not
significant[F (2,83)=P=.71] (Table Appendix E). \Vhen the scorestor
these three groupswere compared \vith 47 previously tested
intensive classes in Quebec, they were foundto beamong the highest.
Mean scores on this measure for aH50 groups ranged froln a lo\v
of47%to a high of 78%.Table 5.52Mean Scores in Percentages and
Standard Deviations bv Group on AIEQ
TestGroupE+EUN273029l\lean78.1375.4776.97185SD10.9912.5912.76Thequantitativeandqualitative
analyses reportedinthis chapter showthat thepredicted advantage for
typographicalaJone and in combination with a book were not
supported. The next chapter offers Interpretations for these
results.186End Notes for Chapter 51. Two coding conventions should
he noted here. First, the pronoun he was not included in
thesubjectpronoun count whenit was usedasanintroducer, asin "hehave
amother and afather". Second, when the leamer made a falsestart or
self-corrected, ooly the last pronoun\vas counted. In the foUowing
examples from a student in Group U at the delayed posttest,the
pronoun that was cotmted is underlined:1) "and he teU his, his boy,
'show me your brotherNOW!"'; 2) "and the boy teU her mom, his mom
that is not him, and aU kind of stupid thingslike that..".2.
Thebody parts category was kept separatefromtheinanimate category.
Zobl (1985) andMartens (1988) combined these two categories.3. As
noted earlier, in the accuracy ratios caicuIated in this study, the
nwnber of grammatical usesincontext appearsin the numerator. In
contrast,Zobl calculated whathe called"difficultyratios", in which
the nwnber of ungrammatical uses in context appears in the
nwnerator.4. The use of the tenn emergence criteria refers to the
minimwn munber of foons a leamer mustproduce in arder ta he
assigned to each of the eight stages. Emergence criteria say
nothingabout accuracy. The tenn emergence is also used to describe
Stage 3 and Stage 4, when theleamer beginstousegender-markedfaons
but shows noevidenceof usingthe Englishagreement mie.5.
Thereisnorestrictionon how these fOUT usesmust be distributed over
the differentpicturedescriptions. That is, criterion could be
reached in the description ofonJy one picture althoughthis rarely
occlDTed with these data.6. At the pretest, one leamer in Group E
met the criteria for classification al Stage 7. However,this
indi"idual used only his (Stage 4)al the tirst posttest and
continued to overgeneralize hisat the delayed posttest, with
instances ofher below criterion(Stage 4).187Cbapter 6Discussion and
CODclusions6.0 Introductionln this chapter, an interpretation and
discussion ofthe results are provided. ln Section6.1, the results
are examined in relation to the hypotheses stated in Chapter 3.
Section 6.2presents a discussion of the results within the context
of the theoretical and empirical workon the role of salience and
explicitness in L2 teaching and learmng. Section 6.3 outlines
andexaminestheissuesrelevant tothedevelopmental
frameworkusedintheanalysisof theleamer data.
Section6.4oullinesthelimitations of thestudy.
Section6.5describesthecontributions of thisstudy to
classroomresearch inL2learning, and suggestionsforfutureresearch
are made.6.1 Results in relation to the hypothesesHypotheses 1
and2predictedthat onmeasures ofpossessivedetenniner (PD)development
al the immediate posttest, the group exposed to typographicaHy
enhanced inputin combination with extensive teading and listening
activities (Group E+) would outperfonnthe group exposed to
typographically enhanced input without extensive reading and
listening(Group E), which in turn would outperform the group
exposed to input that wastypographically unenhanced for PDs (Group
U). Hypotheses1and 2 can he summarized asfol1ows: Group E+ >
Group E > GroupU. While mean scores on the wnnen measures
andaccuracy ratios calculated on the oraldata generally
followedthepredicted order, the oolystatistically significant
differences obtained were benveen Group E+ and GroupU. Thus
thefindings do not support the hypotheses ofthis study which
predieted an effect fortypographical
enhancement.188Hypothesis3predicted that thedifferencesthat
wereanticipated at theilnmediateposttest would stillhe significant
at the delayedposttestfiveweekslater. This whichis dependent
onHypotheses1 and2, wasnot supported. Moreover,the
ditferencesbetween Groups E+andUobserved at
theinunediateposttestwerenolonger statisticallysignificant.
Thisfinding does not retlect a declinein performance by leamersin
Group E+betweentheimlnediateanddelayedposttests, but
rathercontinuingilnprovement bytheIearnersin aH three groups
between the iffiJnediate and delayed posttests. The
improvementdemonstrated by Groups E and Uduring this t\ e-week
period was such that leamers in thesetwo groups caught up\\l'ith
Group E+. Inparticular, the strong perfonnance by
GroupUonthedelayedposnest measures was not anticipated. The
tollowingdiscussionexarninesseveral tctors which may have
contributed to equalizing the three groups' chances ofacquiring
PDs.6.2 Salience and explicitnessAs indicatedin Chapters1and 2,
inputis a crucialvariableinand thewaysinwhichlt ismanipuJated,
andenhancedintheleamingenvirorunentcanleadtodifferentresultsintheL2leamer's
acquisitionof thetarget language. A hypothesisin theSLAliterature
which is central to issues concemingthe role ofinput in SLA
inparticular, totheresearch questionsinvestigatedin thisstudyisthe
"noticinghypothesis".This hypothesis states that getting learners
to attend to linguistic fonnsin the inputis a basicprerequisite for
leaming1994;Schmidt and1986). Two questions whichdirectly
arisefromthishypothesisare: a)How caoweget learnersto"-notice"
particularfeatures of the L2 in the input? and 2) Are there more
effective ways for learners to noticethese features which may
enable them to convert the input nto intake?189[ntheSLAliterature.
efforts tocreateinstructional contexts whichdrawthe
L2attentiontofeatures inthe input bymakingthelnmoresalient
orexplicit havevaried. While sorne research has shown that explicit
efforts to get learners to '''notice''mayhe required..
otherresearchershave argued thatless explicit (i.e.
implicit)methodsmaybeequally beneficial. Asindicated in Chapter 2,
Sharwood Smith (1981, 1991) has suggestedthat the expllcit/implicit
distinctionis best viewed as a continuumrather than a
dichotomy.Thus, theinstruetional treatment usedin thepresent
studyis best described asless explicitthanthat
providedintheresearch of Lyster(1994b), SpadaandLightbown(1993),
White(1991)andWhiteetal. (1991), moreexplicit thanthe
"inputinstudiesbyTrahey(1992, 1996), Traheyand\\1ute {1993) andthat
of other booktloodstudies (ElleyandMangubhai,Lightbo\vn. Hafiz and
Tudor, 1989, 1990) and 1ess explicit than thevisual enhancement
conditions inlplementedby Doughty(1991). In thesections whichtllow,
issues related to the conceptualizationandoperationalizationofsuch
notions assalience and explicitness are discussed in relationship
to the findings.6.2.1 Sai ienceAnwnberofSLAresearchers have
pointedout that for input ta becomeintake,learners must attend to
linguistic featuresin the input as weIl as to messages (e.g.
Shaewood1986; VanPattenandCadiemo, 1993). Hulstijn(1989)
proposedthat attentiontafonu at thepoint of input encodingis the
necessary and sufficient condition for learning totake place.
However, the dual requirement of processing input for meaning as
well as fonnrisks imposing excessively large demands on the
leamers' attentional capacity, to the possibledetriment ofeither
ofthe two processes, comprehension or acquisition (V1990).190Third
person singular pronouns and POs were visually enhanced in this
study ta makethemmoresalient andincreasethelikelihoodthat
learnerswouldpay attentiontathemwithoutoverloadingshort-tennmemory.
It wasexpectedthat additional saliencewouldresult from an increase
in the frequency with which learners encountered the foonsas
theycompleted a set of leaming tasks. However, there is evidence
from the oral data suggestingthat POs May have been equally salient
in the input available to leamers in all three groupsalthough this
was not intended.Thefirst
evidencecornesfromthenumerousself-correctionsandfaIse
startsthatoccurred as leamers struggled to describe each ofthe
pictures. The following example showshowa leamer eventually arrived
at the correct PD form after a number of unsuccessful tries:Leamer
Uh theboyshaveuh ailthe band-aid And the herlegituh hand.And
00..lnterviewer Where are the band-aids did you say?Leamer ln the,
in his leg. And uh her, uh his hand.Of particularinterest was
whether there was anincreaseinthe number of self-corrections
involvingpronouns andPDs immediatelyfollowing the
two-weektreatmentperiod since such anincreasemight berelatedtathe
salience of pronominal formsintheinput treatmenl Toinvestigate
this,thedataforPietureSet Awereusedto compare thenwnber of
self-corrections made by learners in each group immediately before
andimmediately after the two-week instructionaJ period. Learners in
ail three groups showed asimilar increase of about 10%.
Furthennore, the percentageofpronoun andPDself-corrections out of
the total nwnber of self corrections was simi1ar in ail groups at
bath thepretest(14-19%) andthe posttest (25...29%).
Althoughnofurther analysesof theself-correction data were carriedo
~ these findings suggest that the enhanced andunenhanced191input
had a sunilar effect on promoting self-corrections ofthe target
fonns..Additional evidence that salience was similar across groups
cornes from the findingthat leamers in Groups E and Uwere more
accurate on her than on his al the irnmediateposttest, a finding
which is contradietory to claims made by other researchers that
learners aremore accuratein their use of masculine pronominal
foonsin carly stages of development(Zobl, 1985; Martens, 1988). One
interpretation of this finding is that the higher frequencyof her
as compared to hisin the enhanced and unenhanced input treatment
materials madethe feminine forms more salient than the masculine
fonns. Recall that Alanen (1995) foundthat
leamersexposedtoenhancedinput overgeneralizedthemost frequent
fonns. Thefindingthat accuracywas sili1ilaronhis andher for
learners inGroup E+ would notconstitute counter evidence since this
group' s exposure to stories containing large amounts ofinput that
was moregender-balancedmight beexpectedtoreduce the effects
oftheimbalance. Thusitappearsthatwhilethetreatment conditions were
designedtoprovidethreedifferent typesof i n p u ~ otherfactors
mayhavebeenoperatingwhichreducedtheimpact ofthese
differences.Another factor involved the nature of the written tests
and the frequency with whichtheywereadministered Inparticular, the
multiplechoice test createdcontexts
whichcontrastedhisandherandrequiredthelearnerstochooseamongseveral
fonns. It isplausible that the process of deliberating over the
forms at the pretest and immediate posttestdrew the leamers'
attention to the gaps in their knowledge and increased the salience
of thefOnDS that they encountered in the enhanced and unenhanced
treatment materials, as weil asin the regular c1assroominput.
Leamers who found the fonns puzzlingwould havefonnulated hypotheses
about the English rule, and they would have had Many
opportunitiesto test their hypotheses. Thus the testing procedure
itselfmay have enhanced the target fonns192similarly for in ail
three of the treatInent groups.Thee\'idence presentedabove lends
support tothe c1aitn that different types ofenhancement may have
contributed equally to the leamers'acquisition of PDs in this
study.thequestionas towhythe most salient typeof enhancement
provided(i.e.typographica1 enhancement)didnot benefit leamees
morestill remains. OneexplanationInay he related to the learners'
tanlilianty \vith the target tonns. lt is certain that learners
hadalreadyencounteredthetarget fonns intheir regular intensive
ESLprogram. FollowingCook' s (1991. 1993)they mayhave been
abletodecode (understand)messagescontainingPDs eventhoughmost
hadnot yetbrokenthecode(workedout theunderlyingrule). [notherwords,
thetonns maynot have beennovel enoughtoattract the
leamers'attention to the extent predieted (see Harley,tor a similar
interpretation).Another tctor maybe thenunlber of tormsthatwere
visuallymanipulated. A totalof sixpronominal
tormsweretypographicallyenhanced: thesubjeetpronounsIzeandshe:the
abjectpronouns lum and lier:thepossessive detenninershi.'Iandht!r.
This\vas doneinordertoncrease thesalenceof the gendercontrast andto
implicitlydrawthe learners'attentiontothefaerthat
filsandIzerarepart of thethirdpersonsingularpronounsystemAlthough
PDs were always enlarged more than subject and objectit is possible
thatlearnersdidnot perceivethisdifference. Recall frolnChapter
4that learnershadnohelpfromtheir teachersinmaking the distinction
and were simplyinfonned that sorne wordsinthe texts were
highlighted because "these are words you have trouble and \ve want
youto notice how they areused". Thus sornelearners mayhavefound
thepages clutteredan...., 'C
.cfic:: .. t: ..&Jfi CI} CIMC)fa.0M CI}lUU.... i!
zCl)0 J: W (&J
l'l'
1.0CI),
f-l
0()0,f&1Hra.H0!Teacher Feedback on ErrorDate:
Teacher:School:-/for grammar; x for vocabularYi Pfor
pronunclatlonExp11 c 1tl correct lon1Act./BD Imgllclt
correct1Qn,Hetallnqulstlc otherlCali onlIodel Other correct Model
parallnaulstlcanother st. no signal -+ stress f+lnt. +lnt.
-+stres.li Quest-ron aule Gest. Volee l'ace: 1...1j1---
.1,1Appendix B: Sampie of treatment materialsActivity 2 - The Frog
PrincePart of the text andrelated task'lQEnbanced
treatment240Activity 2Patrt 1'l'be . ~ r O l ....inceOnce upon a
lime there wu a king. Ht had a beautiful, young daughter. For
htrbirthday, the king gave hera golden bail thal she played wilh
every day.The king and his daughler Iived neac a dark foresl There
was a deep weil near thecastle. Sometimcs, the princess would sil
by the weil and play with herball. One day.the princess threw
hergolden baU in the air but il did nol fall into htrhands. It fell
intothe weIl. Splash! Th weil wu deep and the princess was sure she
would never sec herball again. So sh, cried and cried and could not
stop.-What is the matter'?- said a voice behind htr. The girl
looked around, and sh, saw afrag. Ilewa:s in the weil, his head
sticking out of the Waler.Oh, ifs you" said the girl. "My baU feU
into the weU:-1 can help,- said the frog. "1 an get your baU. What
will you give Ole if 1do']""Whatevcr you want,- said ahe princesse
l'lI give you my beautiful gald ring. l'IIgive you tlowcn from my
larden..( do not wanl your bcautiful gold ring or f10wers from your
garden, said the frag._Uut 1would Iike to live with you and be your
friend.ACt\'ity 2l'art 2....ioct5s, King or .'rogWho dues lhe
underliMd wonJ reJer 10'1 Wrile l'in the blank. if;t reJers10 the
princess.wrile Kin Ihe bl"nk if;, reJn 10 the king, and write F in
the b/anJc if;,reJeTS 10lhefrog.Ilnecessary, loole /Jacle Dllhe
siory. l'he fini OM is dOMJor )lU",1. For htr birthday, IK had
givenh ~ r a Bolden bail.2. The princess lived with lJJ.m. near~
dark foresl.3. She played with !JH. golden bail.4. She dropped
!lHgolden bail in the weil.S. !!.l. was in the weil, sticking hls
head out of the water6. She offered ta give !J.im. h ~ , fine
golden ring and flowen from her garden.1. The frog said he wanlcd
to be !JHfriend.241Enhanced AlIinstances of he, she, l1im,11er,
lus.11er were enhanced in the texts and activity sheets.Theteacher
corrected the answersraeachtaskwiththe c1assbeforebeginning
thenextone.Activity 1: Encvclooedia BrownWorkingindividually,
studentsreadashort mystery story thattookplace in
thesummeraboutaboydetective, EncyclopediaBro\vn,
andthentriedtasolvethemysterywithapartner. They were encouraged ta
reread relevant parts of the story.Studentsread a similar story
that tookplacein the winter, toldl'romthepoint of view ofa girl
detective. They\Vereencouragedtatinddiffrencesbetweenthestaries
andthenta solve the mystery asinPart 1.Studentsread 35 sentences
and,withoutlooking back at the texts, identified which of
thetwostarieseachsentencecamefrom. Cluesconsist of
thirdpersonsingularpronounsandpossessivedeterminers. Students
correctedtheirpartners' papersbytindingeachsentence in the
appropriate story.Activity 2: The Frog PrinceAfter eliciting
information about afamiliar fairytale,the Frog Prince, the teacher
askedstudents taidentifycharacteristicsofthis andother fairy tales.
Workingindividually,students read a traditional version of the Frog
Prince.Students worked \Vith apartner to answer 20 questions about
the story; the task requiredthemtowritetheinitial of
thepersontowhomthe underlinedpronounor possessivedeterminer
referred (K for the King; P for the Princess; F for the
frog).242Parts 3 and 4Afterlooking ata cartoonwiththecaptton .... [
washappier when1 wasastudentsdiscussed why hislife as ahuman might
havebeen more difficult thanhis Iit as afrog.Theteacherreadthe
illustratedstorybook TheFrogPrinceContinuedwhile studentsfollowed
along withtheir own textsinwhich pronounsand possessive
determinerswereenhanced.Studentsreadaletterfromtheprincesstoherfather
inwhichsorneof thecontent waswrong.
Errorsprimarilyinvolvedpronounsandpossessivebut other errorsoffaet
were included. Students circledthe errors inthe letter and\Vfote
the correctinfonnationabovetheerror.
Studentsexchangedtheirpapersandreferredbacktothestory to correct
them.Activitv J: HelenKellerAlter discussing what it
wouldbeliketobe deaf andblind, students looked at apictureof Helen
Keller and discussed whattheykne\v about her. Working indi\ldually,
they reada story about her discovery of the meaning of the word
waler.Working in groups of six, students took tums answering
questions about thewater story.Pronouns were used instead ofproper
nouns whenever possible. The task requiredstudents to go back to
the text tofindtheinformation.Studentsread another
textaboutHelenKeller and askedeachother questionsabout herlife.
Questions contained pronouns instead of proper nouns.The teacher
asked the questions with third person singular pronouns and
possessivedeterminers about Helen Keller and her teacher, Annie
Sullivan. Students \Vererequiredto understand which woman the
pronouns referred to.Students answeredadditional factual questions
about Helenandher teacher, Annie.They had to reread portions of the
tex! to answer them. Answers were corrected with
theclass..243Students\Vere told that Helen \vas a difficult chi Id
who oftenfought \vith her teacher and\Vth her father. They read
questions withmasculine andtminine pronounsandhad topredicttowhich
of n,o stories each
questionreferred"FightingwithAnnieorFiQhting\vith Father. They
corrected their own answers by reading the relevant texts.Activity
4: PoemsActivin lStudents followedalongastheteacher
readaloudtenhumorous poems inwhichthegender of
thecharacterscouIdonlybedetenninedbyunderstandingthepronounsandpossessivedetenniners.
As theyreadalong" theyselected atitlefor each poemfromalist
pro\ided. Workingwitha partner. students
confinnedtheirchoiceoftitles andanswered questionsaboutthegenders
of thcharacters (e.g. lsthespeakera boyor agirl? \}/hat makes you
think so?).Working in pairs, students grouped the poems according
to themes.Each student chose a poem to practice: several students
pertonned or read them aloud tothe class.Students read a new poem
and created a title \vith a partner.Activio 5Omitted.Activitv 6:
Brothers and SistersStudentsread one of nvo complementary stories
about a brother and sister \vho didnotget a l o n ~ One storywas
written fromthe sister' s perspective, the other fromthebrother' s
perspective.244Working in groups ofthree, students decided whether
six statements containing pronounsandpossessive
determinersweretrueorfaIse, based onthestory theyhadread.
Theyunderlined passagesin the textsupporting their responses.
Students compared ans\versfor thet\VO different texts andfoundthat
they \vere different. They then read theotherstory. The teacher
pointed out that the perspective of the speaker can change the
story.Students read a story about triplets, separatedinearly
childhoo