Pr

inci
plesandPr
act
ice

i
n

SecondLanguageAcqui
sit
ion

St
ephenDKr
ashen

Uni
ver
sit
yofSout
her
nCal
i
for
nia
Thisist heoriginalversi
onofPr incipl
esandPr act
ice,aspublishedin1982,wi t
honl yminor
changes. I tisgr at
ify
ingtopoi ntoutt hatmanyoft hepr edi
cti
onsmadei nt hi
sbookwer e
confirmed bysubsequentr esear ch,forexampl e,thesuper ior
ityofcompr ehensibl
e-i
nput
basedmet hodsandshel teredsubj ectmat t
erteaching(Krashen,2003),theineffi
cacyoferror
correcti
on( Truscott
,1996,1999) ,andt he" powerofr eading"(Krashen,2004) .Subsequent
researchhasal so,i
nmyopi nion,conf i
rmedt hati nfootnote5,chapt er3,opt i
on3i st he
correctone,thatweacqui revocabul arybestthroughcompr ehensibl
einput(Krashen,1989;
2003) .

Ihav echangedmyposi ti
onononlyoneissue:AttheendofPr i
nci
plesandPr acti
ce,Isuggest
theuseofaf or
m ofdecepti
on-studentsmayt hi
nkt heyareacquiri
ngv ocabular
yorl ear
ning
subjectmat ter
,butunknown t ot hem,t hey ar
e acqui r
ing because they are get t
ing
compr ehensi
bleinputatt
hesamet i
me.Inowt hi
nkitisveryimportantt
omakeast r
ongef for
t
toinform st
udentsaboutthepr
ocessoflanguageacquisit
ion,sotheycancont i
nuetoimpr ove
ont hei
rown.

Krashen,S.(1989)Weacqui
revocabul
aryandspel
l
ingbyreadi
ng:
Addi
ti
onal
evi
dencef
or
theinputhypothesi
s.Moder
nLanguageJournal
73,440-
464.

Krashen,
S.(2003)Expl
orat
ionsi
nLanguageAcqui
sit
ionandUse:
TheTai
pei
Lect
ures.
Portsmouth,
NH: Hei
nemann.

Kr
ashen,
S.(
2004)ThePowerofReadi
ng.Por
tsmout
h,NH:
Hei
nemann.

Truscot
t,J.(
1996)
.Thecaseagai
nstgr
ammarcor
rect
ioni
nL2wr
it
ingcl
asses.Language
Learni
ng,46(2)
,327-
69.

Truscot
t,J.(
1999)
.What
'swr
ongwi
thor
algr
ammarcor
rect
ion?TheCanadi
anModer
n
LanguageReview,
55(
4),
437-
56.

I
ntr
oduct
iont
otheI
nter
netEdi
ti
on
Copy
right©1982St
ephenKr
ashen

AllRight
sReserv
ed.Thi
spubli
cati
onmaybedownloadedandcopi
edwi t
houtchar
geforall
reasonable,
non-
commerci
aleducati
onal
pur
poses,pr
ovi
dednoalt
erat
ionsinthet
extare
made.

Fi
rstpr
int
ededi
ti
on1982byPer
gamonPr
essI
nc.

Pr
intEdi
ti
onI
SBN0-
08-
028628-
3

Fi
rsti
nter
netedi
ti
onJul
y2009

i
Acknowl
edgment
s
Iam indebt edt omanypeopl ewhohav ehelpedmebot hdirect
lyandi ndir
ectl
y.Icanonl y
ment i
onaf ewoft hem here.Ear lStev
ickandRobi nScarcell
akindlyprovidedmewi thdetail
ed
comment sonnear l
yev eryaspectoft hemanuscr i
pt.Ihav ealsor eceivedagr eatdealof
usefulfeedbackf rom JohnSchumann, JohnOller,Adri
anPalmer ,TracyTer rel
l,
AndrewCohen,
StevenSt ernfel
d,andBat yi
aEl baum.Iam sur et hisbookwoul dbemuchst r
ongeri fIhad
fol
lowedal ltheiradvi
ce.Thet askofwr i
ti
ngthisv ol
umewasmademucheasi erbythesuppor t
andunder standingofmyf amily,mywi feEula,mychi ldr
enDebor ahandDani el,andmy
parentsLeoandJul i
aKr ashen.Iwoul dalsol iketot hankmycol l
eaguesi ntheLi ngui
sti
cs
DepartmentatUSCf ortheirintell
ectualsti
mul ati
onandencour agement .Iespeciall
ythank
LarryHy man,Edwar d Finegan,EugeneBr i
ere,ElaineAnder sen,El
inorOchs,Edwar dPur cel
l,
JohnHawki ns,andBer nardComr i
e.

i
i
Cont
ent
s
I
.Int
roduction:TheRel ati
onshipofTheorytoPr actice 1
A.Thr eeAppr oachestoMet hod 2
1.Theor yofsecondlanguageacqui siti
on 2
2.Appliedli
nguisti
csresearch 3
3.Ideasandi nt
uiti
onsfrom exper
ience 3
B.I
nt eracti
onsAmongAppr oachestoPractice 4
C.Whatt heThr eeApproachesHav etoSayAboutMet hod 6
D.Goal sofThi sBook:tor ei
ntr
oduceteacher stotheor
yand 7
hopef ul
lytogaintheirconfi
denceagain
Notes 8

I
I.SecondLanguageAcqui si ti
onTheor y 9
A.Fi
veHypothesesAboutSecondLanguageAcqui si
tion 10
1.Theacqui sition-learningdi sti
ncti
on 10
2.Thenat ural orderhy pot hesis 12
(
a)Tr ansi t
ional for ms 14
3.TheMoni t
orhy pot hesi s 15
(
a)Indi vidualv ariationinMoni t
oruse 18
4.Thei nputhy pothesi s 20
(
a)St atementoft hehy pothesi
s 20
(
b)Ev idencesuppor tingthehypot
hesi
s 22
5.TheAf fectiveFi l
terhy pot hesi
s 30
B.TheCausati
veVar i
abl einSecondLanguageAcqui sit
ion 32
1.Thecausat i
v ev ari
abl es 32
2.Languaget eachi ng: doesi thel
p? 33
(
a)Whenl anguaget eachinghelps 34
(
b)Whenl anguaget eachingdoesnothelp 35
3.Ex posurev ar i
ables 37
4.Age 43
5.Acculturati
on 45

I
II
.Pr
ovidi
ngInputforAcquisiti
on 57
A.ThePotenti
aloftheSecondLanguageCl assr
oom 58
B.Li
mitati
onsoftheCl assr
oom 59
C.TheRoleofOut put 60
1."
Conversati
on"andlanguageacquisi
ti
on 61
2.Out
putandl earni
ng 61

3
3
3
Cont
ent
scont
.
D.Character
isticsofOptimalI nputforAcqui siti
on 62
1.Optimal i
nputiscompr ehensi bl
e 63
(a)Howt oai dcompr ehensi on 64
2.Optimal i
nputisinteresti
ngand/ orr elevant 66
3.Optimal i
nputisnotgr ammat i
callysequenced 68
(a)Thecaseagai nstt hegr ammat i
cal syll
abus 70
4.Optimal i
nputmustbei nsuf f
ici
entquant ity 71
(a)Quantityrequirement sfori ni
tialreadinesstospeak 71
(b)Quantityrequi r
ement sforhi gherl evelsofprof
ici
ency 72
E.OtherFeaturesthatEncour ageAcqui si
tion 73
1.Thest udentshouldnotbeputont hedef ensi ve 73
2.Providetoolstohelpst udent sobt ainmor ei nput 76
F."
Teaching"Conv ersat
ionalCompet ence 78

I
V.TheRoleofGr ammar ,orPut ti
ngGr ammari ni
tsPlace 83
A.Learni
ngDoesNotBecomeAcqui sit
ion 83
B.ThePl aceofGr ammar 89
1.Gr ammarf orMoni t
oruse: whent heMonitorisused 89
2.WhatcanbeMoni tored 92
(a)Incompet entMoni toruse 94
(b)Rulel earnabi li
ty 96
(c)Someev idence 98
(d)Consequencesoft eaching"hard"rul
es 102
C.TheEf fectsofLear ning:Accur acyofSelf-cor
recti
on 104
1.Fact orsaffectingsel f-corr
ectionaccuracy 105
2.Thedat a 108
D.OtherEf fectsofConsci ousRul es 112
E.PresentationofRul es 113
1.Thededuct ive-induct iv
eissue 113
2.Sequenci ngandl earning 115
F.NotesonEr rorCorrection 116
G.GrammarasSubj ect-mat ter 119

V.Appr
oachestoLanguageTeachi ng 125
A.Present
-dayTeachingMet hods 126
1.Grammar -
translat
ion 127
(
a)Requi rementsforopt
imal
input 128
(
b)Lear ning 129
(
c)Summar y 129
2.Audi
o-l
ingualism 129
(
a)Requi rementsforopt
imal
input 130
(
b)Lear ning 132
(
c)Summar y 132
3.Cogni
ti
v e-
code 132
(
a)Requi rementsforopt
imal
input 133
(
b)Lear ning 134

4
4
Cont
ent
scont
.
(c)Summar y 134
4.Thedi r ectmet hod 135
(a)Requi rement sf oropt imal input 135
(b)Lear ning 137
(c)Summar y 137
5.Thenat ur al appr oach 137
(a)Requi rement sf oropt imal input 138
(b)Lear ning 139
(c)Summar y 140
6.Total phy sical response 140
(a)Requi rement sf oropt imal input 140
(b)Lear ning 142
(c)Summar y 142
7.Suggest opedi a 142
(a)Requi rement sf oropt imal input 144
(b)Lear ning 146
(c)Summar y 146
B.Appli
edLingui st i
csResear ch 146
1.Reviewofmet hodcompar i
sonst udi es 147
(a)Amer icanst udi esofAL, GT, andCC 149
(b)TheGUMEpr oj ect 150
2.Somepr elimi nar yconcl usionst omet hodcompar i
sonstudies 151
3.Mor er ecentmet hodcompar isonst udies 155
(a)TheTPRser i
es 155
(b)Ot heri nputmet hodscompar ed 157
(c)Suggest opedi ar esear ch 158
C.Alt
ernati
vet oMet hods 160
1.Funct i
onoft hecl assr oom 160
2.Thesecondl anguagecl assr oom andr equirement#2 161
3.Theal t er nat ives 162
(a)Conv er sat ion 163
(b)Pl easur er eadi ng 164
(c)Usi ngsubj ectmat terforlanguaget eachi
ng 167
(d)Ev idencef orsubj ectmat tert eaching:theimmersionprogr
ams 170
(e)Ot herpossi bi
liti
esi nsubj ectmat t
erteaching 171
D.Comment sonAchi ev ementTest ing 176
1.Normal consi der ationsi ntestev aluat i
onandsel ecti
on 176
2.I
nstruct ional v alue 177
3.Languager equi r ement s 181
4.Univer sityl ev el ESL 182
E.SomeGapsi nMat erials 182
1.Thel anguagel abor ator y 184
2.Acommentonf ieldt est i
ngofmat erials 185
F.SomePr oblems 186

Bi
bli
ogr
aphy 191

I
ndex 201

5
6
Chapt
erI

I
ntr
oduct
ion:TheRel
ati
onshi
pofTheor
ytoPr
act
ice
Thepur poseoft hisbookist otakeanew l ookatanol dquest i
on:ther elat
ionshipbet ween
secondl anguageteachi
ngpr act
iceandwhati sknownaboutt heprocessofsecondl anguage
acquisiti
on.Theusualwayt odot hisist odi scusssomer esearchr esult
sf i
rst,out l
inea
possi
bl etheory,andthenlistsomei mplicat i
ons.Iwi l
l
,tosomeex tent,foll
ow thispl an.A
si
gnificantport
ionofthi
sbooki s,infact
, dev otedt osummar i
zingthecur rentstateofsecond
l
anguageacqui si
ti
ontheoryasIseei t
.Fol lowingt hi
s,Iwilldraw somegener alconcl usions
aboutappl icat
iontomet hodsandmat erials,andev entuall
ydescr i
bewhatchar acteristi
cs
eff
ectivemat er
ial
sshouldhav e.

Bef oregoi ngt hrought hisex ercise,howev er,itisi mpor t

antt ost atei nadvancet hat
"theor y"and" t
heor eti
calr esear ch"shoul dnotbet heonl yi nputintodeci di
ngonmet hodol ogy
andmat eri
als.Whi lemyemphasi sher ei sont heoryandi t
simpl i
cat ions,i tisdanger oust o
relyonl yon t heory.Ther e ar e atl eastt hree di fferentway s ofar r
ivi
ng atanswer si n
met hodol ogyandmat erials,andwemustconsi deral loft hem.Iwi lldev otet heremai nderof
thisint roduct i
ont oabr iefdescr iptionoft heset hreear eas,andadi scussionofhow t hey
i
nt er
r elatewi theachot her .Wewi l
l t
henseewhateachhast osayaboutmet hodinsecondand
foreignl anguaget eachi ng.Myv i
ew,f ort hosewhol iket hepunchl ineear ly,isthatal lthree
way sar riv
eatpr eciselyt hesameanswer .Thesol utiont oourpr oblemsi nlanguaget eachi ng
l
iesnoti nex pensiv eequi pment ,ex oti
cmet hods,sophi st i
catedlingui sti
canal yses,ornew
l
abor at or i
es,buti nf ullut ilizationofwhatweal readyhav e,speaker soft hel anguagesusi ng
them f orr ealcommuni cat i
on.Iwi l
lalsoconcl udet hatt hebestmet hodsmi ghtalsobet he
mostpl easant ,andt hat ,st r angeasi tseems,l anguageacqui sit
ionoccur swhenl anguagei s
usedf orwhati twasdesi gnedf or,communi cation.

1
A.Thr
eeAppr
oachest
oMet
hod

1.THEORYOFSECONDLANGUAGEACQUI
SITI
ON
2.APPLI
EDLI
NGUI
STI
CSRESEARCH

3.I
DEASANDI
NTUI
TIONSFROM EXPERI
ENCE

B.I
nter
act
ionsAmongAppr
oachest
oPr
act
ice

Befor
edi scussingwhateachappr oachhast osayaboutmet hodsandmat eri
als,I
woul dli
ketomakeamodestpr oposal :t
het hreeappr oachesshouldinfl
uenceandhel peach
other.Itseemsobv i
ous,fi
rstofal l
,thatr esearcherswoul dbei nt
erestedint her esultsof
applied r
esearch,since such experiments can pr ovide pot
enti
alconf i
rming and count er
evidencefortheoriesofsecond l anguage acqui sit
ion.Simil
arl
y,itstandst or eason t hat
appliedli
nguisti
csr esear
chersshoul dpaysomeat tenti
ont ostri
ctl
yt heoreti
calr esearch,
si
nceasuccessf ult heor
ymi ghtgi veresear chersdeeperi nsi
ghtintot her esult
soft heir
studies.

I
talsoseemsr easonabletosuggestt hatresear
chersinbotht heor
eticalandappl
ied
l
i
nguisticswouldbenef i
tbybotht eachi
ngandst udyi
nglanguages,i
nor dert
ogetmor einsi
ght
i
ntot he language acquisi
tion process.Similar
ly,one mightex pectpract i
ti
oner
st o be
i
nter
estedi nt heresultsofr esearch,andonemi ghtalsoex pectresearcherstobev ery
i
nter
estedi ntheopi
nionsofbot hteachersandlanguagestudents.

Fi
g1.1.I
dealrel
ati
onshi
pbetweent
heor
y,appl
i
edl
i
ngui
sti
csr
esear
ch,
ideasandi
ntui
ti
onsand
l
anguageteachi
ngpract
ice.

2
 Fi
g.1.
2.Actualr
elat
ionshi
pbetweentheor
y,appl
i
edl
i
ngui
sti
csr
esear
ch,
ideasand
i
ntui
ti
onsandlanguageteachi
ngpracti
ce.

C.Whatt
heThr
eeAppr
oachesHav
etoSayAboutMet
hod

Thepurposeofthi
sbookistosummar i
zeonecurrenttheoryandstat
etheimplicat
ions
ofthet heoryt
omet hod.Iwil
lbr
ief
lysummar i
zeher ewhatsomeoft heseimpli
cat
ionsar e,
anti
cipati
ngChapterI
II
.Whatcurr
enttheor
yimplies,
quit
esi mply,
isthatl
anguageacquisi
tion,
fi
rstorsecond,occursonlywhencompr ehensi
onofr ealmessagesoccur s,andwhent he
acquireri
snot"onthedefensi
ve"
,touseStevi
ck'
saptphr ase.Languageacquisi
ti
ondoesnot
requi
reex t
ensi
ve

3
useofconsci ousgr ammat icalrules,anddoesnotr equi
retediousdr il
l.Itdoesnotoccur
ov er
night,howev er .Reall anguageacqui sit
iondev el
opsslowly
,andspeaki ngski ll
semer ge
signifi
cantl
ylatert hanlisteningskill
s,evenwhencondi ti
onsareperfect.Thebestmet hodsar e
thereforethose t hatsuppl y" compr ehensi
bleinput"inlow anx i
etysi tuati
ons,cont aining
messagest hatst udentsr eallywantt ohear.Thesemet hodsdonotf orceear l
ypr oduct
ioni n
thesecondl anguage,butal low studentstopr oducewhent heyare" ready",recognizi
ngt hat
i
mpr ovementcomesf rom suppl yingcommuni cativ
eandcompr ehensi bleinput,andnotf rom
forcingandcor rectingpr oduction.

Insever
alplacesint hi
sbookIwi l
latt
emptt omaket hepoi ntthatresearchi nappli
ed
l
ingui
sti
csisveryconsist
entwi ththetheoret
icalr
esearchinsecondl anguageacqui sit
ionand
i
tsimpli
cati
ons.The" bestmet hods"accordingtocompar ativ
er esearchst udi
es( compar i
ng
methodsAandB,asdescr i
bedear l
i
er)appeart obe" i
nputmet hods",thosemet hodsthat
f
ocusonsupplyingcompr ehensibl
einputinpressure-
fr
eesituations.

Wecangetani deaofwhatt he"ideasandi ntuiti

ons"ar eaf eel
si sthe"bestmet hod"by
asur veyofpedagogically-or
ient
edpaper sincurr
entj our nal
sandt het i
tl
esofpr esent ati
onsat
teacherwor kshops.Thet itl
eshavechangedmar kedlyov erthey ears!Adecadeagot eacher-
oriented arti
cl
es and pr esentat
ions focused on gr ammat icaldescr i
pti
on,r eflecti
ng t he
concer n with pr
oduct,and pr ocedures fordril
ling.1*
Currentt it
les more cl earlyr ef
lect
promot ingrealcommuni cati
onint heclassroom,hel pingst udentsunder standspokenand
2
writteninputandparti
cipateinconversati
ons.

Inworkshopsandmi ni
-confer
ences,wenol ongerseepresent
ati
onsonf i
nepoi
ntsof
grammar,orontypesofsubst
ituti
ondril
l
."Ideasthatwork"areideasaboutrol
e-pl
ayi
ng,
using
thenewspaperasat eachi
ngaid,socio-
drama,et c.Moreover
,newermet hodol
ogyhas,asa
pri
marygoal,
theloweri
ngofstudentanxiet
y( seeChapterI
II
).

D.Goal
sofThi
sBook

Thepr
imarygoaloft
hisbooki
st opresentcur
rentt
heor
yandi
tsi
mpl
i
cat
ions.Ther
eis
anot
hergoal
,however
,andt
hatistor
eintr
oduce
*
Super
scr
iptnumber
sref
ert
oNot
esatendofchapt
ers.

7
teacherstot heoryandhopef ul
lytogainthei
rconf i
denceagai n.Thetimehascomet olookto
theoryagain,reali
zingthatthemostcur r
enttheorymayst i
llnotbet hef i
nalwordonsecond
l
anguageacqui si
tion.Iam notaskingpracti
ti
onersormat er
ialsdev el
operstofoll
owallofthe
i
mpl i
cationsoft heorybli
ndly.Myhopei sonlythatourresultswillbeconsideredasanother
sourceofi deasandi nputtomet hodsandmat eri
als,i
npartnershipwithconclusi
onsreached
bypr act
iti
oner sthemselvesf r
om t hei
rownex perienceasl anguaget eacher
sandl anguage
acquirer
s.

Not
es

1 Consider
,forex ampl e,thet ableofcontentsofLanguageLearni
ng,
vol.9,
1959,
whi
ch
i
ncluded:
"
Grammat ical theoryandpr acti
ceinanEnglishgrammarcl
ass"
"
Reachi ngt heFr enchv erb"
;
"
Noun- classesandt hepracti
calteacher
"
"
Mor phemeal ter nantsinSpanishv er
bforms"
"
'
Thechnemes' andt herhythm ofclassact
ivi
ty"
Volume12, 1962, cont ained:
"
Annot atedbi bliographyofgener ati
vegrammar"

2 The1979v olumeoftheTESOLQuar ter

ly,
forexample,cont
ainsarti
clessuchas:"
Usi
ng
r
adiocommer ci
alsassupplementar
ymat er
ial
sinESLl i
steni
ngclasses"
"
Communi cat
ivewrit
ing"
"
Joke-t
ell
i
ngasat ool i
nESL"
ref
lect
ingthecurr
entemphasi soncommuni cat
ioni
ntheclassroom.

(Notes1and2cer tai
nlydonotr epresentawidesampl eofact i
vit
yinthefi
led,
butt
hey
arerepresentat
ive.I
nrecentyear
s,thejournalLanguageLear ni
nghasf ocussedon
theoret
icalandappliedr
esearch,r
atherthanpedagogy .Forthisreason,Iusedthe
TESOLQuar t
erl
y,whichbeganpublicat
ionin1967, f
orcurrenttit
les.
)

8
Chapt
erI
I

SecondLanguageAcqui
si
ti
onTheor
y
Thischapt ersummar i
zescur rentsecondl anguageacqui siti
ont heory. Todot his,i tfir
st
describes some v eryimpor tant hy potheses. The f ir
stt hree, t he acqui sit
ion-learning
disti
ncti
on, t
henat uralorderhy potheses, andt heMoni torhy pot hesis,ar
er eviewedsomewhat
brief
ly,ast heyhav ebeendeal twi thagr eatdeali nsev eralot herbooksandpr ofessional
papers.Enoughdet ailwil
lbepr ovided,howev er,togivetheuni nit
iatedreaderagoodi deaof
thehy pothesesandt hesor tofev idencet hatex i
ststosuppor tthem.Thef ourthhy pot hesis,
the inputhy pothesis,may be t he si ngle mosti mportantconcepti n second l anguage
acquisit
iont heor ytoday.Itisimpor t
antbecausei tat
tempt st oanswert hecr ucialtheor eti
cal
questionofhowweacqui relanguage.I tisal soi mportantbecausei tmayhol dtheanswert o
manyofourev erydayproblemsi nsecondl anguagei nstruct i
onatal llevels.Followi ngt he
discussionoft heinputhy pothesis,wet urnt ot heconceptoft heaf fect
ivefilt
er,ahy pot hesis
ast ohowaf f ectiv
ev ari
ablesrelatet ot hepr ocessofsecondl anguageacqui siti
on.

Thesecondpor t
ionoft hischapterrevi
ewsav ari
etyoffactor
st hathav ebeent hought
to be rel
ated to second l anguage acqui siti
on success,includi
ng i nstruction,diff
erent
measuresofex posur etothesecondl anguage,andt heageoft heacqui rer.Thesef actors,it
wil
lbeclaimed,arenotr eal
lycausat iv
ef act
ors.Whiletheyseem torel
atet osuccessorf ail
ure
toacquiresecondl anguages,t het r
uecausat i
vev ari
ablesinsecondl anguageacqui siti
on
deri
vefrom theinputhy pothesisandt heaf f
ecti
vef i
lt
er--
theamountofcompr ehensi
bleinput
theacquir
erreceivesandunder stands,andthest r
engthoftheaffecti
vefilter,orthedegr eeto
whichtheacquireris"open"tot heinput.

9
A.Fi
veHy
pot
hesesAboutSecondLanguageAcqui
sit
ion

1.THEACQUI
SITI
ON-
LEARNI
NGDI
STI
NCTI
ON

Theacquisit
ion-l
earni
ngdist
inct
ioni
sperhapst
hemost
fundamentalofallthehypothesestobepresent
edher
e.I
tstat
esthat
adult
shav e
twodisti
nctandi ndependentwaysofdevelopi
ngcompet
enceina
secondlanguage.

Thef i
rstwayi slanguageacqui sit
ion, aprocesssi milar
, i
fnoti dent
ical
,
tothewaychi ldrendev elopabil
ityintheirfir
stlanguage.Language
acquisi
ti
oni sasubconsci ousprocess; l
anguageacqui rersarenot
usual
lyawar eoft hef actthatt
heyar eacqui ri
ngl anguage, butareonl y
awareoft hef actt hattheyareusi ngthelanguagef orcommuni cation.
Theresultofl anguageacqui si
ti
on, acquiredcompet ence,isalsosubconsci ous.
Wear egener allynotconsci ouslyawar eoft herulesoft he
l
anguageswehav eacqui red.I
nstead, wehav ea" feel"forcorrectness.
Grammat icalsent ences" sound"right,or"feel"r
ight,ander ror
sf eel
wrong,eveni fwedonotconsci ousl yknowwhatr ul
ewasv iol
ated.

Otherway sofdescri
bingacquisi
ti
oni
ncludei
mpli
citl
earni
ng,
i
nformal lear
ning,andnatur
al l
ear
ning.I
nnon-
techni
cal
language,
acquisi
tion
i
s" pi
cking-up"alanguage.

Thesecondwayt odev elopcompet encei nasecondlanguageisby

anguagel
l earni
ng.Wewi lluset het erm" learni
ng"hencefort
ht oref
er
toconsciousknowl edgeofasecondl anguage, knowingt
her ul
es,bei
ng
awareoft hem, andbeingabl et ot alkaboutt hem.I nnon-
technical
ter
ms, lear
ningis"knowi ngabout "al anguage, knownt omostpeople
as"grammar ",or"r
ules".Somesy nony msi ncludef or
malknowledge
1*
ofalanguage, orexpli
citlearning.

Somesecondl anguaget heori

stshav eassumedt hatchildrenacquire,
whileadultscanonl ylearn.Theacquisi
tion-l
earninghy pothesis
clai
ms, however,thatadul t
salsoacquire,thattheabili
tyto" pi
ck-up"
l
anguagesdoesnotdi sappearatpuberty.Thi sdoesnotmeant hat
adultswill
alway sbeabl etoachievenative-l
ikelevel
sinasecondl anguage.
Itdoesmeant hatadultscanaccesst hesamenat ur
al"language
acquisi
tiondev i
ce"thatchildr
enuse.Asweshal lseelater,acquisi
tion
i
sav er
ypower fulprocessi ntheadul
t.

*Super
scr
iptnumber
sref
ert
oNot
esatendofchapt
ers.

10
10
 Err
orcorrectionhasl ittl
eornoef fectonsubconsci ousacquisiti
on,butisthoughttobe
usefulf orconsci ousl ear
ni ng.Er rorcor r
ectionsupposedl yhelpst helearnert oinduceor
"fi
gur eout"t
her ightf orm ofar ule.If,f
orex ampl e,
astudentofEngl i
shasasecondl anguage
say s" Igoest oschoolev eryday " ,andt het eachercorrectshim orherbyr epeatingthe
utterancecorrectly,thel earneri ssupposedt or eal
izethatthe/s/endi nggoeswi t
ht hethi
rd
personandnott hef irstper son,andal t
erhi sorherconsci ousment alrepresentati
onoft he
rul
e.Thi sappear sr easonabl e,buti tisnotcl earwhetherer r
orcorrecti
onhast hisi mpactin
actual pr
acti
ce( Fansel ow, 1977; Long, 1977).

Ev i
dencef rom childlanguageacqui sit
ion confi
rmst haterrorcorrect
ion doesnot
i
nfluenceacqui si
tiontoanygr eatext
ent.Brownandhi scol l
eagueshav eshownt hatparents
actuall
ycor r
ectonl ya smal lporti
on oft he child'
sl anguage (occasionalpronunci
ation
problems,cer tai
nv erbs,anddirtywords!)
.Theyconcl udef rom theirresearchthatparents
attendf armor etot het r
uthvalueofwhatt hechildissay ingratherthant otheform.For
exampl e,Brown,Cazden, andBell
ugi(1973)reportt
hatasent encesuchas:

Hercur
lmyhai
r

"
wasappr
oved,
becauset
hemot
herwas,
inf
act
,cur
li
ngEv
e'shai
r"(
p.330)
.Ont
heot
herhand,

Wal
tDi
sneycomesononTuesday

was cor r
ected,despit
ei t
s sy nt
acti
c correctness,since Wal tDi sney actual
ly came on
tel
evisi
ononWednesday .Brownetal .concludet hatitseemst obe" t
ruthvalueratherthan
syntact
icwell-f
ormednesst hatchief
lygovernsex pli
citver
balreinforcementbypar ents--
which
rendersmildlyparadox
icalt hefactthattheusualpr oductofsuchat rai
ningscheduleisan
adultwhosespeechi shighlygrammat i
calbutnotnot ablytr
uthful"(p.330).

Theacquisit
ion-
lear
ningdi
sti
nctionmaynotbeuni quetosecondlanguageacquisi
ti
on.
Wecertai
nly"l
earn"smallpart
sofourf i
rstlanguageinschool(e.
g.formostpeopl
e,thewho/
whom dist
incti
on),and simil
ardist
inctionshav e been madei n ot
herdomains( see,for
exampl
e,Reber,1976;Hall
,1959;
andt her evi
ewind'Angl
ejan,
1978).

11
11
2.THENATURALORDERHYPOTHESI
S

Oneoft hemostex ci
tingdiscoveri
esinlanguageacqui si
tionresear
chinr ecentyears
hasbeent hef i
ndi
ngt hattheacquisiti
onofgrammat icalstr
ucturesproceedsinapr edict
able
order.Acquir
ersofagi venlanguaget endtoacquir
ecer t
aingrammat icalst
ruct
uresearly,and
otherslat
er.Theagreementamongi ndivi
dualacqui
rersisnotalway s100%,butthereareclear
,
stati
sti
call
ysigni
fi
cant,simil
arit
ies.

Engl
ishisper hapst hemostst udi edl anguageasf arast henat uralor derhy pot hesisi s
concer ned,andofal lstructur
esofEngl ish,mor phologyi st hemostst udied.Br own( 1973)
reported thatchi l
dr en acqui ri
ng Engl ish as a f irstl anguage t ended t o acqui re cer tai
n
grammat i
calmor phemes,or f uncti
ons wor ds,ear l
ier t han ot hers.For ex ampl e,t he
progressivemar keri ng( asi n"
Hei spl ayingbasebal l
".)andt hepl uralmar ker/ s/( "wodogs"
t )
wereamongt hef i
rstmor phemesacqui red,whi l
et het hirdper sonsi ngul armar ker/ s/( asin
"Helivesi nNew Yor k"
)andt hepossessi ve/s/( "
John' shat ")wer et ypicallyacqui r
edmuch
l
ater,comi ngany wher ef rom sixmont hst ooney earl ater.deVi l
lier
sanddeVi ll
iers( 1973)
confi
rmedBr own'sl ongitudinalresult
scr oss-secti
onal l
y,showi ngt hati t
emst hatBr ownf ound
tobeacqui redear l
iesti nt i
mewer ealsot heonest hatchi ldrent endedt ogetr ightmor eof ten.
Inotherwor ds,forthosemor phemesst udi ed,t
hedi ff
icultyor derwassi milart otheacqui siti
on
order.

Shortlyaf t
erBr own' sresultswer epubl i
shed, DulayandBur t(1974, 1975)reportedthat
childrenacqui ringEngl i
shasasecondl anguageal soshowa" naturalorder "forgrammatical
mor phemes, regardlessoft heirfi
rstlanguage.Thechi ldsecondl anguageor derofacquisi
tion
wasdi ff
erentf r
om t hef i
rstlanguageor der,butdifferentgroupsofsecondl anguageacquirers
showedst r
ikingsi mi l
ari
ties.Dul ayandBur t'
sr esultshav ebeenconf ir
medbyanumberof
i
nv estigator
s( KesslerandI dar,1977;Fabr is,1978;Maki no,1980) .Dul ayandBur tuseda
subsetoft he14mor phemesBr ownor igi
nall
yi nvestigated.Fat hman( 1975)conf i
rmedt he
realit
yoft he nat uralor deri n child second l anguage acqui siti
on wi th hert estofor al
product i
on, t
heSLOPEt est ,whichpr obed20di fferentst r
uct ur
es.

Fol
lowi
ngDulayandBur
t'
swork,Bai
l
ey,Madden,
andKr
ashen(
1974)r
epor
tedanat
ural
or
derf
oradul
tsubj
ect
s,anor
derqui
tesimi
lar

12
12
tothatseeninchi l
dsecondl anguageacqui si
tion.Asweshal lseelat
er,thisnat ur
alor der
appearsonlyundercertai
ncondi t
ions(orrather,itdisappearsonlyundercer t
ainconditions!)
.
Someoft hestudiesconf i
rmingt henaturalor derinadul tsforgrammat i
calmor phemes
i
ncludeAndersen( 1976),whousedcomposi ti
on,Kr ashen,Houck,Giunchi,Bode,Bi rnbaum,
andSt rei(
1977),usingf r
eespeech,andChr istison(1979),alsousingf reespeech.Adul t
resear
chusingt heSLOPEt estalsoconf i
rmst henat uralorderandwi denst hedatabase.
Krashen,Sf
erl
azza,Feldman, andFathman( 1976)f oundanor dersimi
lartoFat hman's(1975)
chil
dsecondlanguageor der,andKayfetz-
Fuller(1978)al soreport
edanat uralorderusingt he
SLOPEt est
.

Asnotedabove,theorderofacqui
sit
ionforsecondlanguageisnott
hesameast he
orderofacquisi
ti
onforf
irstl
anguage,
butt
her ear
esomesi mil
ari
ti
es.Tabl
e2.
1,f
rom Kr
ashen
(1977),pr
esentsanaver
age

1."
TABLE2. Aver
age"or
derofacquisi
ti
onofgrammaticalmorphemesf
orEngl
i
sh
asasecondlanguage(
chil
drenandadult
s)

Not
es:

1.Thi
sorderi
sder
ivedfr
om ananal
ysisofempir
ical
studiesofsecondl
anguageacqui
sit
ion(
Krashen,
1977)
.Moststudi
esshowsigni
fi
cantcor
rel
atonswit
htheaverageorder
.

2.Nocl
aimsar
emadeaboutor
der
ingr
elat
ionsf
ormor
phemesi
nthesamebox
.

3.Manyoft herelat
ionshipsposi
tedherealsoholdforchi
ldfir
stlanguageacquisi
ti
on,
butsomedonot:
I
ngeneral,heboundmor
t phemeshav ethesamer elati
veor
derforfi
rstandsecondlanguageacqui
sit
ion(
ING,
PLURAL,IR.PAST,REG.PAST, IIISI
NGULAR, andPOSSESSI VE)whil
eAUXI LIARYandCOPULAt endtobe
acqui
redrel
ati
vel
ylaterinfir
stlanguageacquisi
ti
ont hani
nsecondlanguageacquisit
ion.

13
13
or
derf
orsecondl
anguage,andshowshowthefi
rstl
anguageor
derdi
ff
ers.Thi
saver
ageorder
i
sther
esul
tofacomparisonofmanyempir
ical
studi
esofgrammati
calmorphemeacqui
sit
ion.

Whi
leEnglishi st
hebestst udi
edlanguage,itisnott heonlyonestudied.Researchi n
orderofacqui
sit
ionf orotherl
anguageisbeginningt oemerge.Asy etunpubli
shedpaper sby
Bruce(1979),deal i
ngwi t
hRussi anasaf orei
gnl anguage,andv anNaer ssen( 1981),for
Spanishasaf or
ei gnlanguage,confi
rmthev ali
dit
yoft henaturalorderhypothesi
sf orother
l
anguages.

Wewi lldealwi
ththepedagogi calimpl i
cati
onsoft henaturalorderhy pothesislater,I
shouldpointouthere,
howev er,t
hatthei mplicati
onoft henaturalorderhypothesisi snott hat
oursyll
abishouldbebasedont heorderf oundi nthestudiesdiscussedher e,thatis,Idonot
recommendt ngi
eachi ngear l
yandt het hir
dper sonsi ngul
ar/ s/late.Wewi l
l,inf act,find
reasontorejectgrammaticalsequencingi nallcaseswher eourgoali slanguageacqui sit
ion.
Wewi ll
deal wit
hthisl
ater
,howev er
,afterwehav efi
nishedlayi
ngt hetheoret
ical groundwor k.

(
a) Tr
ansi
ti
onal
f
orms

Studi
essuppor tingthenaturalorderhy
pothesisshowonl ytheor derinwhi chmat ur
e, or
well
-f
ormedst ructur
esemer ge.Ot herstudi
esr evealt hepat hacqui rerstake enr out
et o
mastery
.(Forar evi
ew, seeDulay,Burt,andKrashen, i
npr ess.Ravem, 1974;Mi l
on, 1974;Gil
lis
andWeber ,
1976; Cancino,Rosansky,andSchumann, 1974; Wode,1978andNel son,
1980aresomesecondl anguagest udiesinthi
sarea. )Ther eissur
prisinguni f
ormi t
yhereas
well
--
acquir
ersmakev erysimi
larerrors,t
ermeddev elopment aler
ror
s, whiletheyar eacquir
ing.
Forexample,inacquiringEnglishnegat i
on,manyf i
rstandsecondl anguageacqui rerspass
thr
oughast ageinwhicht heyplacethenegativemar kerout si
dethesent ence,asi n:

NoMom shar
peni
t. (f
rom Kl imaandBellugi'
s(1966)
study ofchildL1acquisi
ti
on)
and Notl
i
kei
tnow. (f
rom Rav em's(
1974)studyofchil
d
L2acqui
sit
ion)

14
14
At
ypi
cal
lat
erst
agei
stopl
acet
henegat
ivemar
kerbet
weent
hesubj
ectandt
hev
erb,
asi
n:

Inol
i
ket
hisone. (Cancinoetal.(
1975)st
udyofchi
ld
L2acquisit
ion)
and Thi
snohav
ecal
endar
. (fr
om Schumann' s(
1978a)st
udyof
adultL2acquisi
ti
on)

bef
orer
eachi
ngt
hecor
rectf
orm.

Predictablestagesintheacqui
sit
ionofwh-quest
ionsinEngli
shi
ncl
udeanearl
ystage
i
nwhicht hewh-wor dappearsbef
oretherestoft
hesent ence,whi
chi
sother
wisel
efti
nits
nor
mal uninvertedform,asin:

Howhecanbeadoct
or? (Kl
imaandBellugi
,1966,
chi
l
dL1
acquisi
ti
on)
and Whatshei
sdoi
ng? (Ravem,1974,
chil
dL2acqui
sit
ion)

Onlyl
aterdoacqui
rersbegi
ntoi
nver
tthesubj
ectandv
erboft
hesent
ence.(
Adet
ail
edr
evi
ew
canbefoundinDulayetal.
,i
npr
ess.
)

Transiti
onalf
ormshav ebeendescr i
bedf orotherlanguagesandf orot herstructur
es.
Thest agesf oragivent argetlanguageappeart o best r
ikingl
ysimi l
ardespi tethef i
rst
l
anguageoft heacquirer(althoughparti
cularfir
stlanguagesmayi nf
luencet hedur ati
onof
cer
t ai
nst ages;seeSchumann, 1979).Thi
suniformit
yi sthoughttorefl
ecttheoper ati
onoft he
naturall
anguageacqui sit
ionpr ocessthatispartofallofus.(Foradiscussionofsomeoft he
currenti
ssuesandcont r
ov ersi
esconcer ni
ngt henat ur
alor derhypothesis,seeKr ashen,
1981.)

3.THEMONI
TORHYPOTHESI
S

Whiletheacqui si
ti
on-l
earningdisti
ncti
onclaimst hattwoseparateprocessescoexistin
theadult,i
tdoesnotst atehowt heyareusedi nsecondl anguageperformance.TheMoni tor
hypothesispositst hatacquisit
ionandl ear
ningar eusedi nveryspecifi
cway s.Normally
,
acquisit
ion"i
niti
ates"ourutterancesinasecondl anguageandi sresponsibl
ef orourfl
uency.
Learninghasonl yonef uncti
on,andt hatisasaMoni tor,oredit
or.Learni
ngcomesi nt
opl ay
onlytomakechangesi nthef orm ofourut ter
ance,af terishasbeen" produced"byt he
acquiredsystem.Thi scanhappenbef orewespeakorwr i
te,oraf
ter(
self-
correcti
on).Fi
gure
2.1model sthisprocess.

15
15
 Fi
g.2.
1.Acqui
sit
ionandl
ear
ningi
nsecondl
angaugepr
oduct
ion.

Consciousl
earni
ngisavai
lableonl
yasa" Monit
or",whi
chcanal t
ert
heoutputoft
he
acqui
redsystem bef
oreoraftert
heutt
eranceisactual
lyspokenorwri
tt
en.Itist
he
acqui
redsyst
em whichini
ti
atesnormal
,fl
uentspeechutt
erances.

TheMoni torhypothesi
simpl i
est hatf ormalrul
es,orconsci ousl earning,playonl yal i
mited
rolein second language perf
ormance.These l imi t
ati
ons have become ev en cleareras
resear
chhaspr oceededint helastf ewy ear
s.Thisr esearch,r
eviewedi nChapt erIV,str
ongly
suggestst hatsecond language per for
mers can use consci ous r ules only when t hree
condit
ionsar emet.Thesecondi tionsar enecessaryandnotsuf fi
cient,thatis,aper former
maynotf ull
yuti
li
zehisconsciousgr ammarev enwhenal lt
hreecondit i
onsar emet .Ili
stthese
condit
ionshere,withabri
efdescription.Wewi ll
discusst hem i
ngr eaterdetaili
nChapt erIV:

(
i)Time.I
nor dert
othi
nkaboutanduseconsciousrul
eseffect
ivel
y ,
asecondl anguage
perf
ormerneedst ohavesuff
ici
entt
ime.Formostpeople,
normalconversationdoesnotal l
ow
enought i
met othi
nkaboutanduser ul
es.Theov er
-useofrul
esinconv ersati
oncanl eadto
tr
ouble,i.
e.ahesitantsty
leoftal
kingandinat
tenti
ontowhatt heconv ersat
ionalpartneris
sayi
ng.

(
ii)Focusonf or
m.TousetheMonitoref
fect
ivel
y,t
imeisnotenough.Theperf
ormer
mustalsobef ocussedonf
orm,orthi
nki
ngaboutcor r
ect
ness(Dul
ayandBurt,1978)
.Even
whenwehav etime,wemaybesoinv
olvedi
nwhatwear esayi
ngthatwedonotattendt
ohow
wearesayingit
.

(
iii
)Knowt her ul
e.Thi
si saveryfor
midablerequi
rement.Li
nguist
icshastaughtusthat
thestructureofl anguageisex t
remelycompl ex
,andt heyclaimtohav edescribedonlya
fragmentoft hebestknownl anguages.Wecanbesur ethatourstudent
sareex posedonlyto
asmal lpartofthetotalgr
ammaroft helanguage,andweknowt hateventhebeststudent
sdo
notlearneveryrul
et heyareexposedto.

Theevidencefortheproduct
ionschemashowni
nFi
g.2.
1comesor
igi
nal
l
yfr
om t
he
nat
uralor
derstudi
es.(Conf
ir
mingevidencehas

16
beenal sopr oducedf rom ot hersour ces, see,forex ampl e, BialystokandFr ohlich,1977,
1978a, 1978b. )Thesest udi esareconsi stentwi t
ht hisgener alizati
on:weseet henaturalorder
forgrammat i
calmor phemes,t hatis,t hechi l
d's(secondl anguage)di ff
icultyor der(simil
art o
theor derofacqui sit
ion; Krashen, 1977) ,whenwet estsubj ectsi nsituat
ionst hatappeart obe
"Moni tor-free",wher et heyar ef ocusedoncommuni cati
onandnotf or
m.Whenwegi veour
adultsubj ectst est sthatmeett het hreecondi tions, i
.e.apenci l andpaper" gr ammar "-
typetest,
wesee" unnat ural"or ders,or dersunl iket hechi l
dL2or derofacqui sit
ionordi f
fi
cultyorder.
Thei nter pretationoft hisr esultisthatt henat uralor derr eflectst heoper ationoft heacqui red
system al one,wi thoutt hei ntr
usionoft heconsci ousgr ammar ,sinceadul tsecondl anguage
acquisi ti
oni sposi tedt obesi mi l
art ochi ld( second)l anguageacqui si t
ion.Whenweput
peoplei nsi t
uat ionswher ethet hreecondi t
ionsar emet ,whent heyhav et ime,ar efocusedon
form, andknowt her ul
e, theerrorpat ternchanges, reflectingt hecont ri
but i
onoft heconscious
grammar .

Itappearstobet hecasethatunnaturalorder
sarether esul
tofar iseinrankofcertai
n
mor phemes, t
hel ate-acquir
ed,more"l
earnable"it
ems.InEnglishasasecondl anguage,when
performersareputi nsituati
onswheretheycananddoMoni tor,weseear iseinrankofthe
thir
dper sonsi ngularmor phemeandt her egularpast
,bothl ate-
acquir
ed,low ont heli
stin
Table 2.1,and bot hr el
ati
velyst
rai
ghtforward,both syntacti
call
yand semant i
call
y.(See
studiesbyLar sen-Freeman,1975,descri
bedi nChapterIV,Table4.1;andBr own,descri
bedin
2
Note4, ChapterIV.)

Useoft heconsci ousMoni tort hushast heeffectofal l

owingper f
ormerstosuppl y
i
temst hatar enoty etacqui r
ed.Asweshal lseeinChapterIV,however,onl
ycert
ainitemscan
besuppl iedbymostMoni t
oruser s;theMoni tordoesabetterjobwit
hsomepar t
sofgr ammar
thanwi thot hers.Speci f
icall
y,itseemst odobet t
erwithrulesthatcanbechar acteri
zedas
"si
mpl e"i nt wo di f
ferentway s.First
,r ulesthatdo notr equi
reelaborat
emov ement sor
permut ations;rulest hatar esy ntacti
callysimple.Easyr ulesinthissensei ncl
udebound
mor phology ,suchast het hir
dper sonsi ngularinEngl
ish,ort hede+l e=ducont r
actionin
French.Di f
ficul
tr ul
esi nt hissensei ncludetheEnglish wh-quest i
onr ul
e,whichr equir
es
mov ingthequest ionedwor dtot hefrontof

17
17
thesentence,asubj
ect
-auxi
li
aryinversi
on,andi nsomecasest heinser
tonofdoi
i ntheri
ght
place.Rul
escanalsobeeasyanddi ff
icultduetothei
rsemanti
cproperti
es.TheEngl
ishar
ti
cle
system i
seasytodescri
beformally-
-onesi mplyi
nser
tstheoraorsomet imesnothi
ngbefore
thenoun.Butit
ssemanticsar
ev erydiff
icultt
odescri
be(see,
forexample,Hawki
ns,
1978).

Tosummar izet husf ar,Moni toruseresul t

sintheriseinrankofitemst hatare" l
ate-
3
acqui
red"i
nthenat ural order,
itemst hatt
heper f
ormerhaslearnedbuthasnotacquired.
Onlycert
ainit
emscanr i
sei nr ank,howev er
,WhenMoni t
orusei sheavy,thi
sriseinr ankis
enoughtodisturbt henat uralor der.(AsdiscussedinChapt erIV,i
tispossibl
et oseesmal l
changesincertai
nl ate-acquir
edmor phemest hatarenotenought odi
stur
bt henaturalorder;
thi
smaybet ermedl i
ghtMoni toruse.Seeespeci al
lyNote5,ChapterI
V.

Asweshal lseei nChapt erIV,i tisnoteasyt oencour agenot i

ceabl eMoni toruse.
Exper i
ment ationhasshownt hatany thingl essthanar ealgr ammart estwillnotbringoutt he
consciousgr ammari nanyf orce.Key f
et z(1978)f oundnat uralorder sforbot horalandwr i
tten
versi
onsont heSLOPEt est,showi ngt hatsimpl yusi ngt hewr i
ttenmodal i
tyi snotenought o
cause an unnat uralor der.Houck,Rober t
son and Kr ashen ( 1978a)had adul tsubj ects
(universit
yl evelinternat i
onalst udent s)cor rectt heirownwr i
ttenout put,andst il
lf ounda
natural orderf orthecor rectedv ersi
on.Kr ashen, Butler,Birnbaum, andRober tson(1978)f ound
thatev enwhenESLst udentswr itecomposi t
ionswi thpl ent yoftimeandunderi nstr
uctionst o
bev ery" caref ul"
,theef fectofMoni torusewassur prisinglyl i
ght.Thebesthy pothesisnowi s
thatformostpeopl e,evenuni versityst udents,ittakesar ealdiscr ete-
pointgr ammar -
typet est
tomeetal lthr eecondi t
ionsf orMoni toruseandencour agesi gni fi
cantuseoft heconsci ous
grammar .

(
a)I
ndi
vi
dualv
ari
ati
oni
nMoni
toruse

Someoft heindivi
dualv ar
iat
ionweseei nadultsecondlanguageacquisi
ti
onand
performancecanbeaccount edf orint er
msofdi f
ferent
ialuseoftheconsciousMonitor
.
Studiesofcasehi
stor
iessuggestthattheremaybethreebasicty
pesofperf
ormer(Kr
ashen,
1978; St
aff
ordandCovit
t,1978;KouninandKrashen,
1978).

18
18
 (
i)Moni t
orOv er-
users.These ar
e peopl
e who at
temptt o Monitorallt he t
ime,
perf
ormer swhoareconst antl
ychecki
ngthei
rout
putwit
htheirconsciousknowledgeoft he
secondlanguage.Asar esult
,suchper
former
smayspeakhesitant
ly,oft
ensel
f-
correctinthe
middleofutter
ances,andaresoconcernedwit
hcorr
ect
nessthattheycannotspeakwi t
hany
realf
luency.

Ther
emaybet wodi f
fer
entcausesf orov er-
useoft hegrammar.Ov er-usemayf irstof
al
lderivefrom theperformer'
shistoryofex posuretot hesecondlanguage.Manypeopl e,
vi
ctimsofgr ammar-onl
ytypeofinstructi
on,hav esimplynothadt hechancet oacquiremuch
ofthesecondl anguage,andmayhav enochoi cebuttobedependentonl earni
ng.Anot her
ty
pemayber el
atedtopersonali
ty.Theseov erusershavehadachancet oacqui re,andmay
actual
lyhaveacquiredagr eatdealoft hesecondl anguage.Theysimplydonott rustthis
acquir
edcompet enceandonlyfeelsecurewhent heyref
ertotheirMoni
tor"justtobesur e".

(i)Moni
i torunder-
users.Theseareperfor
mer swhohav enotlearned,oriftheyhav e
l
earned,prefernott ousetheirconsci
ousknowledge,evenwhencondi t
ionsallow i
t.Under-
usersaretypicall
yuninf
luencedbyer r
orcor
recti
on,canself
-cor
rectonl
ybyusi nga" feel
"for
corr
ectness(e.g."i
tsoundsright
"),
andrel
ycompl et
elyontheacquir
edsystem.

Staff
ordandCov i
tt(
1978)not
et hatsomeunder-userspay"l
ipservi
ce"tothevalueof
conscious grammar.Theirsubject"I"feltthatpeople need consci
ous rul
es to speak
"cor
rectl
y "
,andthat"
grammari st
hekeyt oeveryl
anguage"."I
"hi
msel f
,however,har
dlyused
consciousrulesatal
l
,inspeechorwri
ti
ng.

(i
i)Theopt
i imalMoni toruser .Ourpedagogi calgoali st
o pr oduce opt i
maluser s,
perf
ormer swhouset heMoni torwheni tisappr opri
ateandwheni tdoesnoti nterfer
ewi th
communi cation.Manyopt imaluser swillnotusegr ammari nordi
nar yconversati
on,wher ei t
mightinterf
ere.(Some v eryski l
l
ed performer s,such assome pr ofessi
onallinguistsand
l
anguaget eachers,mi ghtbeabl etogetawaywi thusingconsider
abl eamount sofconsci ous
knowledgeinconv ersati
on,e.g.River
s,1979, butthi
sisv er
yunusual.Wemi ghtconsi derthese
people"superMoni torusers",afterYori
o,1978. )Inwr i
ti
ng,andinpl annedspeech,howev er,
whent hereistime,optimaluserswi l
lty
picall
ymake

19
19
what
evercorrect
ionst
heycant
orai
set
heaccur
acyoft
hei
rout
put(
see,
forex
ampl
e,Kr
ashen
andPon,1975).

Opti
malMoni t
oruserscant herefor
euset heirl
earnedcompet enceasasupplementto
thei
racqui r
edcompet ence.Someopt i
maluser swhohav enotcompl et
elyacquir
edtheir
secondlanguage,whomakesmal landoccasi onalerrorsinspeech,canuset hei
rconsci
ous
grammarsosuccessf ul
lythatt heycanof tenproducet heil
lusi
onofbei ngnat i
veintheir
writ
ing.(Thisdoesnoti mplythatconsci ouslearni
ngcanent ir
elymakeupf orincomplet
e
acquisi
ti
on.Someunacqui redruleswi llbelear
nableandot her
snot .Theoptimaluserisabl
e
tofi
llpar
toft hegapwithconsciousl earni
ng,butnotallofit
.

4. THEI
NPUTHYPOTHESI
S

Wewi llt
akemuchmor et i
mewi ththishy pot hesi
st hanwedi
dwi t
htheot hersfortwo
reasons.Fi
rst,muchoft hi
smat er
iali
srelativ
el ynew,whi letheot
herhypotheseshavebeen
descri
bedanddi scussedalreadyinseveralpubl ishedbooksandar t
icl
es.Thesecondr eason
i
si tsimportance,boththeoreti
calandpr actical.Thei nputhypot
hesisatt
empt stoanswer
whati sperhapst hemosti mpor t
antquestioni nourf i
eld,andgi
vesananswert hathasa
potenti
ali
mpactonal lar
easofl anguageteachi ng.

Theimpor t
antquesti
onis:Howdoweacqui r
elanguage?I ft
heMoni torhy
pothesi
sis
corr
ect,thatacqui
siti
oniscent
ralandlear
ningmoreperi
pheral
,thenthegoalofourpedagogy
shouldbet oencourageacqui
sit
ion.Thequest
ionofhowweacqui rethenbecomescrucial
.

Thi
ssect i
onisor gani
zedasf oll
ows:Iwil
lfi
rstpresentthei nputhypot
hesi
sbef ore
giv
inganysuppor ti
ngev i
dence.Foll
owingthisi
sadescr i
pti
onoft heev i
dencefr
om research
i
nf ir
standsecondl anguageacqui si
ti
on.Wewi l
lthenbr i
efl
ycov erev i
dencefr
om applied
l
inguist
icsresear
ch,whichisdiscussedinmoredetai
linChapterV.

(
a)St
atementoft
hehy
pot
hesi
s

Letusfir
str est
atet
hequest
ionofhowweacquire:gi
venthecorr
ectnessofthenatur
al
orderhy pot
hesis,howdowemov efr
om onestagetoanother
?Ifanacqui r
erisat"st
age4",
how canhepr ogresst o"
stage5"
?Mor egener
all
y,how dowemov efrom stagei,wherei
represents

20
20
curr
entcompetence,toi+1, t
henextlev
el?Theinputhypothesi
smakesthefoll
owi
ngclai
m:a
necessary(
butnotsuffi
cient
)condit
iontomov efr
om stageitostagei+1isthatt
heacquir
er
underst
andinputthatcont nsi+1,wher
ai e"under
stand"meanst hatt
heacquir
eri
sfocussed
onthemeaningandnott heform oft
hemessage.

Weacqui r
e,inotherwords,onl
ywhenweunder standlanguagethatcontainsstr
ucture
thatis"alit
tlebeyond"wher ewear enow.How ist hispossi bl
e?How canweunder st
and
l
anguaget hatcontai
nsst r
ucturest
hatwehavenoty etacquired?Theanswert ot hi
sapparent
paradoxisthatweusemor ethanourli
ngui
sti
ccompet encetohelpusunder stand.Weal so
usecontext,ourknowledgeoftheworld,
ourextr
a-l
i
ngui sti
cinformati
ontohelpusunder st
and
l
anguagedi r
ectedatus.

Thei nputhy pothesisrunscountertoourusualpedagogicalapproachinsecondand

foreignlanguaget eaching.AsHat ch(1978a)haspoi nt
edout
,ourassumpt ionhasbeenthat
wef i
rstlearnstructures,thenpracti
ceusingthem incommunicati
on,andt hisi
showf l
uency
dev elops.Thei nputhy pothesissaystheopposit
e.Itsaysweacquireby" goingformeani
ng"
fi
rst,andasar esult,weacqui restr
uctur
e!(Fordiscussi
onoffi
rstlanguageacquisi
ti
on,see
MacNamar a,1972.)

Wemayt
husst
atepar
ts(
1)and(
2)oft
hei
nputhy
pot
hesi
sasf
oll
ows:

(
1)Thei
nputhy
pot
hesi
srel
atest
oacqui
si
ti
on,
notl
ear
ning.

(2)We acquire byunder st

andi
ng language t
hatcont
ains str
uct
ureait
beyondourcurr
entlevelofcompetence(i+1).Thi
sisdonewi t
hthehel
pof
contex
torext
ra-l
ingui
sti
cinfor
mati
on.

At hir
dpar toft heinputhypot
hesissay st hatinputmustcont aini+1t obeusef ulf or
l
anguageacqui sit
ion,butitneednotcontainonl yi+1.I tsaysthatiftheacquirerunderstands
theinput ,andt hereisenoughofi ,i+1wi
t l
laut omat i
call
ybepr ov i
ded.Inot herwor ds,if
communi cati
oni ssuccessful,i+1isprov i
ded.Aswewi l
ldiscusslater
,thisi
mpl iesthatt he
bestinputshoul dnotev enat t
empttodeliberatelyaim ati+1.Wear eallfamili
arwithsy l
labi
thattrytodel i
beratel
ycov eri+1.Ther eisa" structureoft heday",andusuallybot hteacher
andst udentfeelthattheai m oft
helessonist ot eachorpr acti
ceaspeci fi
cgrammat i
calitem
orstructure.Oncet hi
sstructurei
s

21
21
"
mastered"
,thesyll
abusproceedst othenex
tone.Thispartofthei
nputhy pot
hesi
simpli
es
t
hatsuchadeliber
ateatt
empttopr ovdei+1i
i snotnecessar
y.Asweshallseelat
er,
ther
eare
r
easonstosuspectthati
tmayev enbeharmful.

Thus,
par
t(3)oft
hei
nputhy
pot
hesi
sis:

(
3)Whencommunicat
ionissuccessf
ul,whentheinputi
sunder
stoodandt
her
e
i
senoughofi
t,i+1wi
llbeprovi
dedautomatical
l
y.

Thef inalpartoft heinputhypot hesi

sst atesthatspeaki ngf l
uencycannotbet aught
4
directly.Rather,i
t"emerges"ov ert
ime, onitsown.Thebestway ,andperhapstheonlyway,to
teachspeaki ng,accordingt ot hi
sv iew,i ssimpl ytopr ovidecompr ehensi
bleinput
.Ear l
y
speechwi llcomewhent heacqui r
erfeels"ready";thi
sstateofr eadinessarri
vesatsomewhat
differentt imes fordi ff
erentpeople,howev er.Ear ly speech,mor eover,i
st ypi
cal
ly not
grammat i
call
yaccurate.Accuracydev elopsov ertimeast heacqui rerhearsandunderstands
mor einput.Part(4)oftheinputhypothesisisthus:

(
4)Pr
oduct
ionabi
l
ityemer
ges.I
tisnott
aughtdi
rect
ly.

(
b)Ev
idencesuppor
ti
ngt
hehy
pot
hesi
s

(
i)Firstl
anguageacqui si
ti
oninchil
dren.Theinputhy pothesisisv eryconsistentwith
whati sknownabout" caret
akerspeech",t
hemodi f
icationst hatparentsandot hersmake
when talking t
oy oung chil
dren.The mosti nter
esting and per hapst he mosti mpor t
ant
charact
eristi
cofcar et
akerspeechf orusi st hatiti snotadel i
berateat t
emptt ot each
l
anguage.Rat her,
asClarkandCl ar
k(1977)pointout,
car etakerspeechi smodi f
iedinor derto
aidcompr ehensi
on.Caretaker
st al
k"si
mpler"inaneffortt omaket hemsel v
esunder stoodby
thechil
d.

Asecondchar acteri
sti
cofi nteresttoushereisthefindingthatcaret
akerspeech,whil
e
i
tissy ntacti
call
ysimpl erthanadul t
-adultspeech,i
s"roughly-
tuned"tothechil
d'scurr
entlevel
oflingui
sticcompet ence,not" f
inely-t
uned" .I
notherwords,car et
akerspeechisnotprecisely
adjustedtot hel ev
elofeachchi d,butt
l endstogetmor ecompl exasthechildprogresses.
Verygoodev idenceforr ough-
tuningcomesf r
om theresearchofCr oss(1977)andNewpor t,
Gleitman,and

22
22
Glei
tman( 1977),whor epor tthatcorr
elati
onsbetweeninputcomplexit
yandmeasur esoft he
chil
d'slinguisti
cmat ur
ity,whi leposit
iveandof t
ensignifi
cant
,arenotusuallyver
ylarge.An
i
nterpretati
onoft hi
sf i
ndi ngi sthatcaretaker
sarenott aki
ngaim exactyati+1.Thei
l nput
theypr ovi
def orchildreni ncludesi+1,butal soincl
udesmanyst ruct
uresthathavealready
beenacqui red,plussomet hathav enot(i+2,i+3,et c.)andthatt
hechi l
dmaynotber eady
foryet.Inotherwor ds, caretakersdonotpr ovi
deagr ammat i
cal
lybasedsyll
abus!(
Foramor e
compl etereviewofr ough- tuni
ng, seeKrashen1980,1981.)

At hi
rdchar acteri
sti
cofcar etakerspeechthatconcer nsusi sknownast he"hereand
now"pr i
nci
ple.Itiswel lest abl
ishedthatcaretakerst al
kmost lyaboutwhatt hechildcan
perceive,whatisint hei mmedi ateenvir
onment.Di scoursewithchildreni
sf armorelikelyto
dealwi t
hwhati sint heroom andhappeni ngnow( "Seet heball
?")thanwhatisnotinther oom
andnotcur r
ent("Whatwi l
lwedoupst air
stomor row?").AsNewpor tetal
.(1977)pointsout,
thi
sisat opi
calconst rai
nt-
-the"hereandnow"pr i
nciplerefl
ect
st hecommoni nt
erestsoft he
caretakerandchild.

Whilet hereisno directev i

denceshowi ng thatcaretakerspeech isi ndeed mor e
effecti
vethanunmodi fi
edinput,theinputhy pot
hesispredi
ctst hatcaretakerspeechwi l
lbe
veryusefulforthechil
d.Fir
st,i
tis,oraimstobe, comprehensible.The"hereandnow"f eature
providesextra-l
ingui
sti
csuppor t(contex
t)t hathelpsthechi l
d understand t
heut ter
ances
contai ngi+1.AsMacNamar
ni a(1972)pointedout,thechi
lddoesnotacqui r
egr ammarf i
rst
andt henusei tinunderst
anding.Thechi ldunderstandsfir
st,andt hishelpshi m acquire
l
anguage.

Asdiscussedearl
i
er,roughl
y-
tunedcaretakerspeechcoversthechil si+1,butdoes
d'
notfocuson i+1ex cl
usivel
y.Part(3)oftheinputhy pot
hesi
sclaimst hatthi
sisopt
imal
.
Rough-t
uni
nghast hef
oll
owingadvantagesinchil
dfir
stlanguageacquisi
ti
on:

(1)I
tensur hati+1i
est scovered,wit
hnoguesswor
kastojustwhati+1i
sfor
eachchi
ld.Ontheot
herhand,
deliber
at m ati
eai +1mightmiss!

(2)Roughly
-tunedi
nputwil
lpr
ovdei+1f
i ormorethanonechi
l
datatime,as
l
ongast heyunderstandwhati
ssai
d.Fi
nel
y-
tunedi
nput
,eveni
faccur
ate(i
.e.
evenifi
t"hi
ts"i+1)
,will
onl
y

23
23
 benef
itt dwhosei+1i
hechi
l sex
act
lyt
hesameaswhati
semphasi
zedi
nthe
i
nput.

(3)Roughly-t
unedinputprovi
desbuil
t-
inrevi
ew.Weneednotbeconcernedwi
th
whet hera chil
d has" master
ed"ast ruct
ure,whet
herthechil
d waspay i
ng
attenti
ontot heinputthatday,orwhetherweprovi
dedenough.Withnatur
al,
roughly-
tunedinput
,i+1wi l
loccurandreoccur
.

In otherwords,ifpart(3)i s cor
rect
,ifiti
sthe case t
hatwith enough nat
ural
communi cati
onandunderst
andingthati+1i sal
wayspr
ovi
ded,t
hecaret
akerneednotwor ry
aboutconsciousl
yprogr
ammi ngstr
ucture.

Thismustbeagoodthi
ng!Addi
ngt
heresponsi
bil
i
tyofgrammaticalsequenci
ngt
o
par
enthoodwoul
dmakeparent
-chi
l
dcommuni
cat
ionmuchl essspont
aneousandf armor
e
di
ff
icul
t.

(
i)Ev
i i
dencefrom secondl anguageacqui si
tion:simplecodes.Thei nputhy pothesi
s
al
sohol dsforsecondl anguageacquisit
ion.First,aspr esent
edear l
i
er ,thesecondl anguage
acquir
er,chil
doradult,isalsoan"acquirer",j
ustl i
ket hechil
dacquiringf i
rstlanguage.Also,
accordi
ngt ohypot
hesis(2),ther
eisanat uralorderofacqui si
ti
onforsecondl anguageaswel l
asfir
stlanguage,sowecant al
kaboutt hesecondl anguageacquirers'i+1aswel l
.Third,
secondlanguageacquirerscanalsorecei
v ethekindofmodi fi
edinputthatchildrenget.

Thi
smodi fiedinputisoft hreesor ts.Foreigner
-tal
kresul t
sf rom themodi fi
cations
nat i
vespeakersmakewi thlessthanf ull
ycompet entspeakersoft hei
rlanguage( see,for
ex ample,Hat
ch, Shapi
ra,andGough, 1978f orsomegoodex ampl es).Teacher-
tal
kisforeigner
-talkinthecl
assr oom,thelanguageofcl assroom managementandex pl
anati
on,wheni tisin
thesecondl anguage.At hi
rdsimplecodei sint
erlanguagetalk,thespeechofot hersecond
l
anguageacqui rers.

Whiletherear esomedif
fer
encesbet weent hesesimplecodesandcar etakerspeech
(Long,1980;Fr eed,1980),t
her
ear ei mport
antsimi l
ari
ti
es.Asi st hecasewi thcar et
aker
speech,modi f
icati
onsmadei nfor
eigner-
tal
kandt eacher-
t k5 ar
al enotmadef orthepur pose
ofl anguaget eaching,butar
emadef orthepurposeofcommuni cat
ion,tohelpt hesecond
l
anguageacqui rerunderst
andwhatisbeingsai
d.Second, t
heav ai
lableresear
chindi cat
esthat
foreigner
-talkandteacher-
tal
k

24
24
areroughly
-tunedtot hel
eveloftheacquir
er,andnotfinel
y-
tuned(Freed,1980;Gaies,1977;
forar ev
iew,seeKr ashen,1980);moreadv ancedsecondlanguageper f
ormerstendt oget
morecompl exinput
, butt
hecorrelat
ionbet
weenpr of
ici
encyandinputcompl ex
ityi
slessthan
perf
ect.
6
Forei
gner -
tal
kandt eacher-
talkmaynotal waysbei nthe"hereandnow" , buthelpful
nati
vespeak er sandteachersfindot herwayst omakei nputcompr ehensi
ble.I
naddi t
ionto
l
inguist
icalter ati
ons,t
heytakeadv antageoftheacqui r
er'
sknowl edgeofthewor l
d, whi
chi s,
of
course,great erthant hatofthechi l
dacqui r
ingaf i
rstlanguage.Teachers,inaddi t
ion,use
pedagogical aids,suchaspicturesandr eal
i
a( seediscussioninChapterII
I).

Theinputhypothesi spr edictsthatt hesesimpli

f i
edcodeswi l
lbev eryusefulforthe
secondl
anguageacqui rer,justascar etakerspeechi sposi t
edtobeusef ulforthechil
d.(For
somepreli
minar
ycasehi stor ydatasupport i
ngthishypothesis,
seeKr ashen, 1980,1981.)The
i
nputhypothesi
salsopr edi ctsthatnat ural,communi cati
ve,roughly-
tuned, comprehensibl
e
i
nputhassomer ealadvant agesov erfinel
y-tunedinputthataimsdirectyati
l
+1,inot
herwords,classroom ex erci
sesthatai mtoteacht hestr
uctureoft heday.

Thecaseagainstthegrammaticalsy
ll
abusi
spr
esent
edinf
ullerdetai
li
nChapterII
I,
but
her
eisabriefsummar y.Theargumentsarever
ysi
mil
artot
hosepresentedagai
nstgiv
ingthe
chi
l
dfinel
y-
tunedinput:

(1)Allst
udent
smaynotbeatt hesamest age.The" st
ruct
ureoftheday"may
notbei+1f ormanyoft
hestudents.Wit
hnat uralcommunicat
ivei
nput
,onthe
ot somei
herhand, +1orotherwil
lbeprovi
dedf orevery
one.

(2)Wi t
hagr ammat icalsyll
abus,eachstr
uctureispresentedonlyonce.Ifa
studentmissesit,i
sabsent ,i
snotpay i
ngattent
ion,ori
ftheresimpl
yhasnot
beenenoughpr acti
ce( i
nput)
,thest
udentmayhav etowaitunti
lnexty
ear,when
allstr
uctur
esar ereviewed!Ont heotherhand,roughl
y-
tunedcomprehensibl
e
i
nputallowsfornaturalrev
iew.

(3)Agrammaticalsy
ll
abusassumesweknowt
heor
derofacquisi
ti
on.Nosuch
assumpti
onisnecessarywhenwer el
yoncompr
ehensi
bleinput
,onr oughl
y-
tunednat
ural
communi cat
ion.

25
25
 (4)Fi
nally
,agrammat i
calsyll
abus,andther esul
ti
nggr ammat i
calfocus,pl
aces
seri
ousconst r
aintsonwhatcanbedi scussed.Tooof t
en,iti
sdi ff
icul
t,i
fnot
i
mpossi bl
e,todiscussorr eadanythingofrealint
erestifourunderlyi
ngmot i
ve
i
st opracticeapar ti
cul
arst r
ucture.Inotherwords,agr ammaticalfocuswill
usuall
ypreventrealcommuni cat
ionusingthesecondl anguage.

Ift
heseargumentsar ecorr
ect,t
heymeant hatweshoul dnotat
tempttoteachalong
t
henat
uralor
der,oranyot
herorder,whenourgoalisacqui
sit
ion.(
Thisi
snotnecessar
il
ytrue
whent
hegoalisconsci
ouslear
ning;seeChapt
erI
V.)

(i
ii)Evidencefrom secondlanguageacquisi
ti
on:thesi
lentperi
odandL1infl
uence.The
i
nputhy pothesisisalsoconsistentwithotherfi
ndingsandhy pot
hesesinsecondlanguage
acqui
sit
ion.Oneoft hesecanbet ermedt he"si
lentperi
od",aphenomenont hatismost
noti
ceableinchildsecondlanguageacquisi
tion.

Ithasoftenbeennot edthatchil
drenacquiri
ngasecondl anguagei nanatural
,inf
ormal
l
inguisti
cenv i
ronmentmaysayv er
ylitt
leforseveralmonthsfoll
owingtheirfi
rstexposureto
thesecondl anguage.Whatout putthereisconsistsusuallyofmemor izedlanguage,whole
sentenceslearnedasi ftheywer eonewor d.Hatch( 1972)
,forexample,repor
tedthatPaul,a
fiv
e-year-
oldChi nesespeakeracqui r
ingEnglishasasecondl anguage,didnotr eall
yuse
"creat
ive"language forhi sfi
rstfew mont hsi nt heUni t
ed Stat
es.Hi sonlyout putwas
memor i
zedsentences,suchas

Getoutofher
e.
I
t'
stimetoeatanddr
ink.

Hehadcl earl
yl ear
ned t heseaswhol eut t
eranceswi t
hout ar ealunderstandi
ngoft hei
r
component s(e.g.heprobabl ywouldnotunder st
andt heword"out
"or"time"ifitwereusedin
anothersentence).Suchmemor iz
edsent enceswer eprobablyv
eryusefulforPaul,bot
hinthe
7
cl
assr oom andplayground. When"real
"languagedidstarttoemerge,i
tlookedv er
ymuchl i
ke
fi
rstlanguagedev el
opment ,wit
hshor t,
simplesentencessuchas

Thi
ski
te.Bal
l
no.

26
26
 Theexplanat i
onofthesilentperiodi nter
msoft heinputhypothesi si
sstraight-
forwar d-
-thechildisbuildingupcompet enceint hesecondl anguagev i
al i
stening,byunder standing
thelanguagear oundhim.I naccor dancewi t
hthei nputhypothesis,speakingabi l
ityemer ges
onitsownaf terenoughcompet encehasbeendev el
opedbyl ist
eningandunder standing.We
shouldnotet hatcasehi stor
iesdeal i
ngwi thchi
ldrenacqui r
ingsecondl anguages( seeal so
Hakuta,1974;Er vi
n-Tri
pp,1974)agr ee t hatsev eralmont hs mayel apse unt i
lt heyst art
tal
ki
ng,and t hatt he speech thatemer gesi snoter ror-
fr
ee.Thi sf indi
ng hasi mportant
pedagogicalconsiderat
ions,asweshal l seeinChapt erII
I.

Adults,andchil
dreninformallanguageclasses,ar
eusual l
ynotal l
owedasi l
entperi
od.
Theyareof t
enaskedt opr oduceveryearl
yinasecondl anguage,beforetheyhav eacquired
enoughsyntacti
ccompet encetoex pr
essthei
rideas.Accordingtoahy pothesisfir
stproposed
byNewmar k(1966),perf
or merswhoar easkedt oproducebef oret heyare" ready"wil
lfall
backonfirstlanguagerules,t
hatis,theywil
lusesy nt
acticrul
esoft heirfi
rstlanguagewhi l
e
speaki
ngthesecondl anguage.

Statedmor eformall
y,anacqui r
erwillsubst i
tutesomeL1r ulefori+1,ar uleoft he
secondlanguage, i
ftheacquirerneedsi+1t oex presshi mselfbuthasnoty etacqui redit.The
L1ruleusedmaybequi tesimi l
artotheL2i+1, butmayal sodif
ferincertai
nway s.Whent he
L1 and L2 r ul
es are di
fferent,the r
esult
ing errori sr eferr
ed to often as"i
nt erference".
Butaccor dingt oNewmar k,itisnotint
erf
erenceatal l
;itisnotther esul
toftheL1i nt er
feri
ng
wit
hsecondl anguageperformance,butther esultofi gnorance--
thelackofacqui sitionofan
L2rulethati sneededinper f
ormance.

(
iv)Advant
agesanddisadvant
agesofL2ruleuse.Thesubst
it
uti
onofsomeL1r ul
efor
some i+ 1 has both adv
antages and di
sadv
antages.The advant
ages ar
e shor
tterm,
howev
er,whil
ethedisadv
ant
agesappeartobequiteser
ious.

One obvi
ous advant
age isthatthe use ofan L1 r ul
e al
lows the perf
ormert o
"outperfor
m hiscompetence",t
omeetapr acti
calneedinL2communi cat
ionbeforehehas
acquiredtherel
evanti+1r ul
e.WhentheL1r ul
eusedisidenticaltoarulei
ntheL2( "posi
ti
ve
transfer"
),t
heperfor
merseemst ohavegotsomethi
ngforfree.EveniftheL1rul
eis

27
27
nott
hesameastheL2rul
e,onecouldarguethattheperf
ormerstil
lcomesoutahead,
as,
qui
teof
ten,
hecanst
il
lcommunicatehispointdespi
tetheincor
rectfor
m.

Anotheradvantageisthattheearl
yproduct
ionallowedbytheuseofL1r
ulesalsohel
ps
t
oinvi
teinput--
ital
l
owst heperfor
mert opart
ici
pat
emor ei nconv
ersati
on,
andt
hiscouldmean
morecompr ehensi
bleinputandthusmor esecondlanguageacquisi
ti
on.

Ther earer ealdisadvant agest ofal

li
ngbackont heL1,howev er.First,theL1r ulemay
notbet hesameasanL2r ule,asnot edabov e,anderror
scanr esult
.Theconsci ousMoni t
or
cannot eandr epairtheseer rorsinsomecases,butnotal l
,since,aswehav eseent he
constraintsonMoni torusear esev ere.Thus,useofL1r ul
esr equir
esconst antv igi
l
anceont he
partoft heMoni tor,andisanawkwar dwaytopr oduceformal l
ycorrectsent encesinasecond
l
anguage.( Notet hatMoni torcor recti
onofsucher r
orswi llnot,accor dingt ot het heory,
produceacqui sit
ion,orper manentchange.I twi l
lnoter adicatetheL1r ule,eveni fdone
effect
ivelyov erl ong periodsoft i
me.Realacqui si
ti
oncomesonl yfrom compr ehensi
ble
8
i
nput .
)

Theremaybeanot herseriousdisadvantaget ot heuseofL1r ulesinsecondl anguage

per formance.Ev eni ftheL1r uleissi
mi l
art oanact ualL2r uleort ransiti
onalf or
m,i tisnot
cleart hattheser uleswi l
lhelptheacqui r
erpr ogress-
-theymaynott aket hepl aceof" t
rue"L2
rules int he dev elopment alsequence.I n Kr ashen ( 1982)Idi scuss the hy pothesisthat
acqui sit
ion requires a compar ison between iand i+ 1 ( Clar
k and Ander sen,1980;
Lamendel l
a,1979) .Itmaybet hecaset hatthe" dist
ance"bet weeniandi+1cannotbet oo
great -
-iandi+1canonl ydifferinsmallway s.Transit
ionalforms, Ihypothesize,maybeusef ul
i
nt hattheycant empor ari
lyserveasi,helpingt odecr easetheamountofdi stancebet weeni
andi +1.

I
f,forexample,t
hetargetrul
ei nEngl
ishi sthenegati
ve(i+1,presentedtot hesy
stem
byi nput)
,t he i
nter
mediateform no + v ( provided bythe cr
eati
ve construct
ion sy
stem
i
nternall
y)maybecl osertothemat urenegativeform.Theacquirermayt hususeno+vati ,
rat
herthanamor epri
miti
veform ofthenegativ
e( e.g.no+S).

I
ftransi
ti
onalf
ormscant
emporar
il
yser
veasi,t
henex tquest
ioniswhet
herL1rul
es,
evenwhent heyhappentobesimi
l
artoL2r ul
esortransi
ti
onalforms,canper
for
mt hi
s
funct
ion.Theanswermaybe

28
28
"no".Forexample, Spanishspeaker
sof tenhaveal ongperi
odi nthei
racquisi
ti
onofEnglishin
whicht heyproduceno+vf ortheEngl i
shnegativ
e,astructurethati
ssimi l
artoatransi
ti
onal
formi nEnglishasaf ir
standsecondl anguage(Schumann,1979) .Itmaybet hecaset hat
earlerno+vper
i formanceistheuseoft heL1rule,whi
lelaterno+vper formanceisthetrue
i
ntermedi at
ef orm.I tmaybet he caset hatonlythelattercan helpt hesy st
em "mov e
9
forward".

Tosummar i
ze,useofL1r ulesishy pothesi
zedt obet her esul
toff all
ingbackonf i
rst
l
anguage knowl edge when a second l anguage rulei s needed in product i
on buti s not
avail
able.Itmayt empor ari
lyenhancepr oduct i
on,butmaynotber ealprogressint hesecond
l
anguage.Ther ealcuref or"i
nterference",accordi
ngtoNewmar k,isnotdr il
latthepointsof
contrastbetweent het wol anguages( Newmar kandRei bel
,1973,p.239) .Dr i
llwill
,atbest,
producelearning,and,aswehav eseen, t
hisisonlyashor tt
erm cure.Ther ealcure"issi
mpl y
thecur eforignorance"( Newmar k,1966,p.81) :r
eallanguageacqui si
ti
on.Thi scanhappen
101112
onlywhent heacqui r
erobt ai
nscompr ehensibl
einput.

(v)Appli
edl i
nguisticsresearch.Thei nputhypothesisisal soconsi stentwit
htheresul
ts
ofwhatcanbecal l
ed" met hodcompar i
son"ex peri
ment s.Sev er
alschol arsandgroupsof
scholarshav eattempt edt odet erminedi rectlywhichteachi ngmet hodsar ebestbysimple
compar i
son.Gr oupsofst udentsst udyingsecondandf oreignl anguagesusi ngt
wodi f
fer
ent
met hodsar ecompar ed,bothinl ong-term andshor t-
term studies.Wewi llhaveadetai
ledlook
atthisresearchinChapt erV,butIwi llst
at emyownconcl usi onsinadv ance.Myreadingof
studies compar i
ng t he mor e commonl y used met hods ( audio-l
ingualas compar ed to
grammar -
tr
anslat
ionorcogni tive-code)isasf ol
lows:

(1)"Deductiv
e"methods( rul
efirst
,thenpr act
ice,e.g.grammar-t
ransl
ati
onand
cognit
ive-
code)aresl
ightlymoreef f
ici
entthanaudi o-l
ingualt
eachi
ngforadult
s.
Thedi f
ferencesareoftenst at
isti
call
ysignifi
cant,butar enothuge.Students
clear
lymakesomepr ogr essusinganyoft heseappr oaches.

(
2)Foradol
escent
s,t
her
eisnomeasur
abl
edi
ff
erence.

29
29
 Iinterpretthisfailur
et of i
nd l
argediff
erencesinthi
sway :noneoft hemet hods
comparedi nthesest udiespr ovi
desmuchi nt hewayofcompr ehensi
blei
nput!Theinput
hypot
hesis predict
s,mor eov er,t
hatan approach thatpr
ovi
des subst
anti
alquanti
ti
es of
comprehensibl
einputwi l
ldomuchbet terthananyoftheol
derappr
oaches.

There ar e severalnewermet hods thatdo t his,such as Asher '

s TotalPhy sical
ResponseMet hod( Asher,1966,1969)andTer rel
l'
sNat uralApproach( Terr
ell
,1977).Inthese
methods,classt imei sdev otedtoprov i
dingcompr ehensibleinput,wher et
hef ocusisont he
messageandnott hef orm, andstudentsarenotex pectedt oproducei nthesecondl anguage
unti
ltheythemsel vesdeci detheyare"ready".Report
sconf irmingthesuper i
ori
tyofsuch" i
nput
methods"hav ebeenappear i
nginthepr of
essionalli
ter
atureov erthelasttenyears(e.
g.Asher ,
1972;Gar y,1975;Post ovsky,1974;mor edet ai
lisprov idedinChapt erV).(Thef ocuson
compr ehensibleinputisnott heonlyreasonf orthesuccessoft henewermet hods,howev er;
seediscussionbel owofaf f
ect,andChaptersIIIandV.)

Sincet hebul koft hisbooki sintendedt odealwi thimplicationsofsecondl anguage

acquisi t
iont heory(Chapt ersIII
,IV,andV) ,thissectionshoul dr eall
ybedel ayedunt ill
ater.I
cannotr esi
st,howev er,br i
efl
y stati
ng one i mplication here,si nce,i n my opi nion,t he
i
mpl icationsoft hei nputhy pothesi
sar et rul
yex citi
ngf oral lofusi nt erest
edi nl anguage
acquisi t
ion.Mosti mpor tant,theinputhy pothesispr edictsthatt hecl assroom maybean
excellentpl acef orsecondl anguageacqui sit
ion,atleastupt ot he" intermedi at
e"l evel
.For
beginner s,theclassroom canbemuchbet terthantheout sidewor ld,sincet heout si
deusual l
y
providest hebegi nnerwi thverylit
tl
ecompr ehensiblei nput,especi allyforol deracqui rers
(Wagner -Goughand Hat ch,1975).Int hecl assroom,we canpr ovideanhouradayof
compr ehensibleinput,whi chisprobablymuchbet t
ert hantheout sidecandof orthebegi nner.
Wewi llelaborateont hisabi tmoreafterdiscussionoft heAf f
ectiveFi l
ter.

5.THEAFFECTI
VEFI
LTERHYPOTHESI
S

The Af
fect
ive Fi
lt
erhy pot
hesis st
ates how af
fecti
vef
act
orsr
elat
etot
he second
l
anguageacqui
sit
ionprocess.TheconceptofanAffect
ive

30
30
Fi
lt
erwaspr oposedbyDulayandBurt(
1977),
andi sconsi
stentwit
hthet
heor
eticalworkdone
i
nthear eaofaffect
ivevar
iabl
esandsecondlanguageacquisi
ti
on,aswel
lasthehy pot
heses
pr
eviouslycover
edinthi
schapter
.

Researchoverthelastdecadehasconf i
rmedthatavari
etyofaff
ect
ivevari
abl
esrel
ate
tosuccessinsecondlanguageacquisi
ti
on( r
eviewedi
nKrashen,1981)
.Mostofthosest
udied
canbeplacedintooneofthesethreecategori
es:

1)Mot
( i
vat
ion.Per
for
mer swit
hhighmotiv
ati
ongeneral
lydobet
teri
n second
13
l
anguageacquisi
ti
on(usual
ly
,butnotal
ways,"
int
egr
ativ
e"

2)Sel
( f-
conf
idence.Per
for
mer swithself
-confi
denceandagoodsel
f-
image
t
endtodobett
erinsecondlanguageacquisi
ti
on.

(3)Anxiet
y.Low anxi
etyappear
stobeconduci
vetosecondlanguage
acqui
sit
ion,whet
hermeasur
edasper
sonal
orcl
assr
oom anx
iet
y.

Insev eralplacesIhav ehypothesi

zedt hattheseatti
tudi
nalfactorsrel
atedir
ectl
yto
acqui
sit
ion and notl earni
ng,since theytend to show strongerrelati
onshi
ps t
o second
l
anguageachi evementwhencommuni cat
ive-
typetestsareused,teststhattaptheacquir
ed
rat
herthan t he l
earned sy stem,and when t he student
st aki
ng the testhave used t
he
l
anguagein"acqui si
tion-r
ich"sit
uat
ions,si
tuat
ionswher ecomprehensibl
einputwasplent
if
ul.

TheAf fecti
veFi lt
erhy pothesiscaptur estherelati
onshipbet weenaf f
ect i
vevari
ables
andthepr ocessofsecondl anguageacqui si
tionbyposi ti
ngthatacqui rersvarywi t
hrespectto
thestrengthorl eveloft hei rAf f
ecti
veFi l
ters.Thosewhoseat ti
tudesar enotopt i
malf or
secondlanguageacqui si
tionwi llnotonlytendt oseeklessinput ,
butt heywi l
lalsohaveahi gh
orstrongAf fecti
veFi l
ter-
-ev eni ftheyunder standthemessage,t hei nputwillnotreacht he
partoft hebr ai
nr esponsi blef orlanguageacqui si
ti
on,ort helanguageacqui si
ti
ondev i
ce.
Thosewi that t
it
udesmor econduci v
etosecondl anguageacqui siti
onwi l
lnotonl yseekand
obtai
nmor einput,theywi llalsohav eal owerorweakerf il
ter.Theywi l
lbemor eopentot he
i
nput,andi twil
lstri
ke"deeper "(Stevi
ck,1976) .

TheAf
fect
iveFi
l
terhy
pot
hesi
s,r
epr
esent
edi
nFi
g.2.
2,cl
aimst
hat

31
31
theeff
ectofaff
ecti
s"outsi
de"thelanguageacqui
sit
iondevicepr
oper.Itst
il
lmaint
ainsthat
i
nputisthepr i
marycausati
vevari
ableinsecondlanguageacquisit
ion,af
fect
ivevari
ables
acti
ngtoimpedeorf
acil
i
tatethedel
i
v er
yofinputt
othelanguageacqui
siti
ondevi
ce.

Thefil
terhypothesi
sex pl
ainswhyi
tispossi
bleforanacquir
ertoobtai
nagr eatdealof
comprehensibl
einput,andy etstopshor
t(andsomet i
meswel lshor
t)ofthenativespeaker
l
evel(
or"fossil
i
ze";Seli
nker
,1972).Whenthi
soccurs,i
tisduetotheaff
ecti
vefil
ter
.

Fi
g2.
2.Oper
ati
onoft
he"
aff
ect
ivef
il
ter
".

The" af
fect
ivefil
ter
",posi
tedbyDulayandBur t(1977),actstopr eventinputfrom
beingusedforlanguageacquisi
ti
on.Acquir
erswithoptimal at
ti
tudes( seetext
)ar e
hypothesi
zedt ohave"low"aff
ecti
vefi
lt
ers.Cl
assroomst hatencouragel owfil
ters
arethosethatpr omotelow anxi
etyamongst udents,thatkeepst udents" of
ft he
defensi
ve"(Stevi
ck,1976).

Thi
spicturedoesnotdi mini
sh,inanyway ,theimportanceofaf fecti
vevari
abl
esi n
pedagogy
.TheAf f
ecti
veFil
terhypothesi
simpli
esthatourpedagogi calgoalsshouldnotonly
i
ncludesuppl
yi
ngcompr ehensibl
einput,butal
socreat
ingasituationt hatencouragesalow
fi
lt
er.Asdiscussed i
n ChapterV,sev eralmethodsfocuson j ustt his( e.
g.Counsel
ing-
Learni
ngandSuggest
opedia).

Theinputhypothesi
sand t heconceptoft heAf f
ectiveFil
terdefi
net hel anguage
teacherinanewway .Theef f
ecti
velanguageteacherissomeonewhocanpr ovi
dei nputand
helpmakeitcomprehensibl
einalowanx iet
ysituat
ion.Ofcourse,manyteachershav ef el
tthi
s
14
wayabouttheirt
askforyears,atl
eastunti
ltheyweretoldother
wisebytheex pert
s!

B.TheCausat
iveVar
iabl
einSecondLanguageAcqui
sit
ion

1.THECAUSATI
VEVARI
ABLES

Ourr
evi
ew ofsecondl
anguageacqui
sit
iont
heor
ythusf
arcanbesummar
izedas
f
oll
ows:

1.Acqui
sit
ioni
smor
eimpor
tantt
hanl
ear
ning.

32
32
 2.Inordertoacquire,t
woconditi
onsarenecessary
.Thefi
rstiscomprehensibl
e
(orevenbetter,comprehended)inputcont
ai ngi+1,st
ni ructuresabitbeyond
theacquirer
'scurrentlev
el,andsecond,alow orweakaffecti
vefi
lt
ertoallow
theinput"
in".

Thisi
sequival
enttosayingt hatcomprehensibl
ei nputandthestr
engthofthef i
lt
erare
the t
rue causes ofsecond language acqui si
ti
on.Ot herv ar
iabl
es mayr el
ateto second
l
anguagesuccess,t hatis,wemayseeposi ti
vecorrelati
onsbetweenot hervari
ablesand
measur esofachi
evementinsecondl anguage,butinallcasesinwhichlanguageacquisit
ion
i
sat t
ained,analysi
swillrevealthatt her el
ati
onshipcanbet t
erbeex pl
ainedint ermsof
comprehensibl
einputpl
usf i
l
terlevel.

I
nt hi
ssect ion,wewi l
lperform suchananal ysi
s,l
ookingatseveralf
actorsthathave
beenshownt orelatetosuccessinsecondl anguageacquisi
ti
on.Wewillseethatnotonlycan
theybere-analyzed,butthatthecompr ehensi
bleinput+fi
lt
erexpl
anat
ionhelpstosolvesome
apparentproblemsandcont r
adict
ionsintheresearchli
ter
atur
e.

Wewi l
lbegi
nwi t
htheeffectoflanguaget eachi
ngonsecondl anguageacquisi
tion,t
hen
examinevar
iabl
esrelat
ingt
oex posure(lengthofr esi
denceinthecountr
ywher ethelanguage
i
susedandr epor
teduseoft hesecondl anguage),andt henturntoage.Fi nal
ly,wewi ll
consi
derSchumann'saccul
tur
ationhypot hesi
s,toseewhet heritt
oocanber eanalyz
edi nthi
s
way.

2.LANGUAGETEACHI
NG:
DOESI
THELP?

If acqui
sit
ion is mor e centr
al,and l
earni
ng of l
ess use t o second language
perf
ormance,andi fcomprehensibl
einputandthefi
lt
erar
etheessent i
alcausati
vev ar
iabl
es
forsecondl anguageacqui si
ti
on,theclassr
oom shouldhel
ponl ytotheex tentitsuppli
es
compr ehensibl
einputinanenv i
ronmentconduci
vetoal ow f
il
ter.Thi
smayi ndeedbe,as
mentionedear l
ier
,it
smainf unct
ion.

I
tseemsr easonabl
etohypothesi
zet hatt
hecl assr
oom shouldbeespeci all
yvaluabl
e
f
orbeginners,thosewhocannoteasi l
yuti
lizetheinformalenv
ironmentforinput.Itwi
llbeof
l
essval
uet ot hosewhocan,whohav eothersourcesofcompr ehensi
bleinput,andwhoar e
l
i
ngui
stical
lyadv ancedenoughtotakeadvantageofit.

Thequest
iont
henbecomesnot"
Doesl
anguaget
eachi
nghel
p?"but

33
33
"Whendoesl anguaget eachinghel
p?".A possibleansweri sthis:languaget eachinghelps
wheni tisthemai nsour ceoflow fi
l
tercompr ehensibl
einput,thati s,forbeginnersandf or
for
eignlanguagest udentswhodonothav eachancet ogetinputout sidet heclass.Itwil
lbe
oflesshelpwhenr i
chsour cesofi
nputar eavai
lable.I
ftheresearchl it
eraturesuppor t
sthese
general
izat
ions,itconfi
rmst hegener
alizati
onthatlanguageteachi nghel pssecondl anguage
acquisi
ti
on when i tsupplies comprehensibl
ei nput,which i
st he true cause ofsecond
l
anguageacqui si
tion.

(
a)Whenl
anguaget
eachi
nghel
ps

Br i
ère(1978)invest
igatedt hefactorsthatpredictedsuccessfulacquisit
ionofSpani sh
asasecondl anguageamong920nat i
veMex icanchildr
en,agesf ourthr
ought wel ve.Among
thebestpr edictor
sofSpani shpr ofi
ciencywasat tendanceincl assint hev i
ll
ageschool
(promot ori
a).Thismak essensei ntermsofourhy pothesis,sincethepromot or
iawast he
maj orsourceofcompr ehensiblei nputi
nSpani sh,asoppor t
unit
iestouseSpani shout sidethe
classroom wer enotplenti
ful
.( Thet woot hersigni
fi
cantpredict
or swerethef ather
'sabili
tyto
speakSpani shandt heparents' needforSpanish.)

Someadul tstudiesrepor tfair

lylar
geposit
ivecor r
elati
onsbet weentheamountof
cl
assr oom exposuretot hesecondl anguageandprofi
ciency.Ineachcase,however,itcanbe
arguedt hatt
hecl asswast hepr i
mar ysourceofcompr ehensibl
ei nput
.Krashen,Zelinski,
Jones, andUsprich(1978)testedst udent
sinanextensi
on( eveningandweekend)pr ogram in
English as a second l anguage atQueens Col l
ege i n New Yor k,and report
ed r obust
correlati
onsbetweenr eport
edy earsofformalst
udyandper f
or manceonav ari
etyofESLt ests,
i
.e.:

Test Cor
rel
ati
onwi
thy
ear
soff
ormal
study

Michi
gan(Lado) r= 0.
50
Composit
ion r= 0.
34
Cloze r= 0.
47
SLOPE r= 0.
42(r
epor
tedi
nKr
ashen,
1976)

Al
lcor
rel
ati
onswer
ever
ysi
gni
fi
cant(
p<0.
01orbet
ter
).

Despi
tet
hef
actt
hatt
hesest
udent
swer
eint
heUni
tedSt
atesand

34
34
techni
call
yi nasecondl anguageandnotaf or
eignl anguageenv i
ronment ,itisli
kelythat,in
manycases,t heydi dnothav ear i
chi nputsourceav ail
ablet othem out si
det heclass.First
,
somehadnotbeeni nthecount ryf oral ongtime,t heirpr i
maryex posur etoEngl i
shhav i
ng
beeninaf orei
gnl anguagesi t
uation.Second, si
ncet hesewer eex t
ensionandnotr egularday-
ti
meuni versitystudents,therewast hestrongpossibilit
yt hatmanyoft hem wer enotut i
li
zi
ng
Engli
shv er ymuchi nt heirdailylives,ev enthought heywer eliv
ingi nNew Yor k.Thisi s
confir
medbel ow,whenwenot et hel ackofast rongr elati
onshipf oundf orthesesame
studentsbet weenl engthofr esidencei nt heUnitedSt atesandpr ofi
ciency,andwhenwe
examinet heef fectofinstruct
iononr egularuni
versitystudent swhodohav ear i
chsour ceof
i
nputout sidet heclassroom.( SeeKr ashen,Seliger,andHar t
nett,1974andKr ashenand
Seli
ger,1976, forsimil
arresults.
)

Chi
haraandOl l
er(1978)alsor eportsubst
anti
alcorr
elati
onsbet weenlengthoffor
mal
studyandsecondlanguagepr of
iciency,fi
ndingacorrel
atonofr=0.
i 45forperfor
manceona
cl
oz etestandsimilarresul
tsforot hermeasur es.Theirsubjectswer eJapanesestudyi
ng
Engli
shasaf orei
gnl anguageinJapan,acl earforei
gnl anguagesi t
uat
ioni nwhichthe
cl
assroom wasthemai n,i
fnotonly,sourceofcompr ehensi
bleinput.

(
b)Whenl
anguaget
eachi
ngdoesnothel
p

Notal ltheresearchl i
teratureconcl udest hatlanguaget eachingisgoodf orsecond
l
anguageacqui si
ti
on!( Imustadmi tthatIam ast oundedt oseet hatsof ew studieshave
i
nv estigatedsuchacr ucialissue!Whati spr esentedi nt hi
ssectionispr act i
call
yt heenti
re
l
iteratureont hi
squest i
on.)Ther ear egener ali
zationst hatcanbemadeaboutst udiest
hat
seem t odeci deagainstsecondl anguaget eaching,howev er
.Inallcases,st udentshadar i
ch
sour ceofcompr ehensibl
ei nputout si
det hecl assr oom,andwer ecompet entenoughi nthe
secondl anguagetobeabl etot akeadv antageofi t,i.
e.understand.

Twost udi
esdealwi
thchil
dsecondl anguageacqui
sit
ionandbothwithEngl
ishasa
secondl anguage.Fat
hman(1975)f oundnosi gni
fi
cantdi
ffer
encesi
nEnglishprofi
ciency
betweenchi l
drenwhohadESLinstr
uctionandchil
drenwhodidnot.Al
lchi
l
dreninherst udy
,
however,

35
35
wereenroll
edinEngl i
sh-medium publi
cschoolsinWashington,D.C.andallhadbeeninthe
Uni
ted States f
rom one t ot hr
ee years.I
tcan be hypothesized thatt
heywer e get
ti
ng
comprehensibl
einputfrom t
heschoolandpl ay
ground,
andt heex t
rainputf
rom t
heESLclass
di
dnotmakeadi ff
erence(nordidgrammaranddr i
ll
!
).

HaleandBudar( 1970)st udiedimmi grantadol escentsi nHawai i

anj uniorhi ghschools.
Int heirpaper(tit
led"Ar eTESOLcl assest heonl yanswer ? "
),theynot edt hatt hesubj ects
formedanat uraldivi
sion.Onegr oupwascomposedofst udent swhospokel esscommon
l
anguages.Thesest udentsdi dnothav et he" benef it"ofaf ormalESLpr ogram andwer e
i
sol atedfrom speakersoft heirownl anguage.Thesecondgr oupconsi stedofst udentswho
hadt hechancet oassociat ewi thotherstudent swhospoket heirownf irstlanguage.These
student sdidattendESLcl asses.Hal eandBudarr eportt hatt hefir
stgr oupact ual
lymade
bet t
erprogressinEnglish,af indingthatseemst oquest i
ont hev al
ueofESLcl asses.Thef i
rst
group,howev er
,mayhav e had mor e compr ehensi bl
ei nput ,possiblyt hrough hav i
ng t o
associ at
emor ewi thEnglishspeaker sandwi t hot hernon- nativespeaker susi ngEnglishasa
l
inguaf r
anca.Thisstudyal sof i
tsourgener alizationandconf i
rmst hatthei ssuei snotplusor
mi nusESLorl anguaget eachingbutpl usormi nusl owf i
lt
ercompr ehensibleinput .

Twoadul tstudiesalsoappeart odeci deagai nstthecl assr oom.Upshur( 1968)st udied

threegr oupsoft en forei
gn st udent sst udying law ataspeci alsummersessi on att he
UniversityofMi chigan.Allst udent st ooksemi narsandcl assest hatusedEngl i
shast he
medi um ofi nstr
uction.Inaddi t
ion, t
heyt ookf ormalESLcl asses, eachgr ouptakingadi ffer
ent
amount ,dependi ngonpl acementscor es.Upshur '
sanal ysisoft heirpr ogressinEngl i
shov er
thesummerr evealednosi gnifi
cantef fect
sat t
ri
butablet ot heamountofi nstructi
ont hey
recei
v ed:t hosewi thmor eESLdi dnotnecessar il
yacqui r
emor et hant hosewi t
hl essov erthe
courseoft hesummer .Mason( 1971) ,inast udydoneatt heUni versit
yofHawai i,simply
all
owedasmal lgroupofi nter
medi ate15 l
evelinternati
onalst udentst opost ponear equired
ESLcl assf orasemest er.Theirpr ogressi nESLwascompar edt ost udent swhoact uall
yt ook
thecour se.Masonr eportednosi gnif i
cantdifferencesbet weent het wogr oups.

Thet
woadul
tst
udi
esar
econsi
stentwi
tht
hehy
pot
hesi
s.I
nbot
h

36
36
cases,
student
shadarichsour
ceofcompr
ehensi
bleinputout
sidet
hecl
assr
oom,
andi
nbot
h
casestheywer
eadvancedenought
obeablet
outil
izeit
.

Iconcludef rom thisthatlanguageteachingcer t

ainlycanhelp.Itsprimaryf unct i
oni st o
suppl ycompr ehensibleinputf orthosewhocannotgeti telsewhere,thoseconst rainedby
theirsituati
on( i
.e.foreignlanguagestudentswhodonothav einputsourcesout sidet hecl ass)
orbyt hei rcompet ence( thoseunablet ounderstandthel anguageoft heout si
dewor l
d).Whi l
e
i
ti slessusef ulfort hosewhohav eothersourcesofi nput ,
theresti
llarethingst hecompet ent
classroom cancont ri
butet othei nt
ermediatest udent.Itcansuppl yconsci ousl earningf or
optimalMoni toruse( seeChapt erIV),andgivet oolstohel pt heacquirerutil
i
zet heout side
env i
ronmentmor ef ullyforfurtheracquisi
ti
on( ChapterIII)
.Table2. 2summar izesst udies
16
discussedi nthissect ion.

3.EXPOSUREVARI
ABLES

Severalexposurev ari
ableshav ebeeni nvesti
gatedwit
hr espectt
osecondl anguage
acquisit
ion.Somest udiesshow acl earrelat
ionshipbetweent heamountofex posureand
prof
iciencyandsomedonot .Weshal lseeagaint hatthehypot hesi
sthatcomprehensibl
e
i
nputpl usl ow fi
lterar ethet ruecausat i
vev ari
ablespredi
ctsqui t
ewellwhen exposure
vari
ablesr elateto second language acquisit
ion and when theydo not .Thus,exposure
vari
ablear ealsoi
ndirectandnoti nt
hemsel vescausativ
e.

Severalstudieshaveexamined lengthofr esi

dence (LOR)i nthesecondl anguage
envir
onment .Fort hosestudi
esofchi l
dsecondl anguageacqui si
ti
on,i tcanbepl ausibly
arguedt hatLORmayr efl
ectsi
mplyt heamountofcompr ehensi
bleinputthechildobtains.
(Thisisofcour senotal waysthecasei nchildsecondl anguageacqui sit
ion;al
ltooof t
en
chil
drenl i
vi
nginacount rydonotgetcompr ehensibl
einput,eit
herinoroutofschool .
)We
thus see,i nt hese st
udies,a clearr el
ati
onship between LOR and second l anguage
profi
ciency.

Fathman( 1975)wasdi
scussedabove.I
naddi ti
ont oherfi
ndi
ngonthenon-
eff
ectsof
formali nst
ruct
ion on ESL achi
evement,Fathman al so repor
ted t
hatLOR di
d predict
profi
ciencyforhersampl eofchi
ldren(
ages6- 14,enrol
ledinpubli
cschooli
nWashington,
D.C.)
.

37
37
Tabl
e2.
2.For
mal
inst
ruct
ionandsecondl
anguageacqui
sit
ion

a Al
lsubj
ect
shadbeeni
ntheU.
S.atl
eastoney
ear
.

38
38
Thosewhohadbeeni ntheUnit
edStatesforthr
eeyearsdi
dbett
eront heSLOPEtestthan
thosewhohadbeenintheUni
tedStatesfortwoyear
s,andthi
sgroup,i
ntur
n,out
perfor
med
thosewhohadbeeni
ntheUnit
edStatesforonl
yoneyear
.

Walberg,Hase, andRasher( 1978)st udiedJapanese- speakingchil

dr enwhohadbeeni n
theUnitedStatesar angeofzer ot o12y ears,wi t
hmostr epor t
ingaLORoft hreetofouryears.
Self-
repor
tandr eportoft eacher swer eusedasest i
mat esoft hechildren'spr of
ici
encyin
Engli
sh.Wal berg etal .didf i
ndasi gnif
icantr elati
onshipbet weenLORandpr ofici
encyin
Englishasasecondl anguage,butnot edt hatev enhighercor rel
ati
onswer efoundwi t
ha
"di
mi ni
shi
ngr eturns"model :"Forchil
drenofal lagesint hesampl e,acquisiti
onproceedsata
fastratei
nit
iall
y,buttheamount sofgai ndi minishwi thtime"( p.436).Speci fi
call
y,"
itmaybe
esti
mat edthatequal .
..uni tsar egainedi nt hef irsttwomont hs,thenex tfiv
emont hs,the
foll
owingoney ear,thenex tt woy ear
s,andt henex tei
ghty ears"(p.436).

Ekst
rand (1976)
,howev er
,found no r el
ati
onship between LOR and child second
l
anguagepr ofi
ciencyinhisst udyofimmi grantchil
dreninSweden.Themedi anLORi nhi s
studywasonl y10.5mont hs,
andi tmaybet hecaset hatLORef fect
sarenotseenunl esst he
chil
dren have been i
nt hecount ryforsomemi ni
mum l ength ofti
me( oney ear?).Thi s
conditi
onissatisf
iedintheFat hmanandWal bergetal.st
udies,andmaybeduet othef act
thatapproxi
mat el
yoney earisnecessaryt ooffsettheadvantagetheolderchil
drenhav ei n
rateofacquisit
ioninearlystages(Krashen,Long,andScar cell
a,1979;seealsodiscussi on
belowonage) .

Walbergetal.'
sdiminishingr eturnshypothesispr edi
ctsthatt her
ei salsoamax imum
LOR,bey ondwhi chweseenor elati
onshipbetweenLORandsecondl anguageacqui si
ti
on.
Dataf r
om someot herstudi
esconf i
rmt hi
s.Seli
ger,Kr ashenandLadef oged( 1974)
,Oy ama
(1976,1978),andPat kowski(1980)al lreportnorelati
onshipbet weenLORandpr ofi
ciencyin
Engli
shasasecondl anguageusi ngav ar
ietyofmeasur esforgr oupsofsubj ectswhohad
arri
vedintheUnitedSt at
esatdi fferentages,somear r
ivi
ngaschi ldrenandsomeasadul t
s.In
allcases,however,verylongLORswer einvolv
ed,t hemi nimum bei ngf i
veyearswi t
hmost
beingmuchl onger.

Twoot
herst
udi
est
hatbearont
hei
ssueofLORandchi
l
dsecond

39
39
l
anguageacquisi
ti
onwi
l
lbecov
eredi
nasubsequentsect
ion(
RamseyandWr
ight
,1974;
Minour
a,1979)
.

Adul
tstudi
esofther elat
ionshi
pbetweenLORandsecondl anguageprofici
encyshow,
i
nmyv i
ew,thatLOR" counts"whent herei
sev idencet hati
tref
lect
shi ghinter
acti
onand
ther
eforecomprehensi
bleinput.Becauseofthevariabi
li
tyinfi
l
terl
eveli
nadults,howev er
,we
mightexpectl
owercorrel
ationsforadult
singeneral,ascomparedtochil
dren(
seedi scussi
on
ofagebelow).

Internat ionaluni v ersit

yst udentsf ul l
yi nvolvedi ntheacademi cenv i
ronmentshoul dgi ve
ussi gnifi
cantcor relati
onsbet weenLORandpr oficiencyint hesecondl anguage,pr ovideda
l
ar geenoughr angeofLORi sex amined, sincest udent sshoul dhaveaccesst olargeamount s
ofcompr ehensi blei nput, bothinandoutofcl ass.Twost udiesut i
li
zedi nter
nati
onalst udent s.
Mur akami( 1980)st udied30Japanese- speaki ngst udentsatSout hernI l
li
noisUni v
ersityand
foundasi gnificantcor relati
onbet weenper for manceonadi ctati
ont estofESLandLOR( r=
0.68),andaposi ti
v ebutnotsi gnifi
cantcor relati
onusi ngacl ozetest(r=0. 29).Theser esul t
s
arenear lyident ical tothoseofOl l
er,Per kins, andMur akami (1980),whoex aminedagr oupof
182st udent s,al soatSI U( whichdi dnoti ncludet he30st udentsMur akamist udi
edal one).
Theyal so r epor ta si gni fi
cantcor relation bet ween LOR and di ctati
on (r= 0. 46)butno
signif
icantcor relationf orcloze( correlationnotr epor ted).

LORandpr ofici
encywasalsopr obedinourst udyofex
tensi
onstudent
satQueens
Col
l
ege( ci
tedabove;Krashenetal
.,1978).Corr
elat
ionswer
eoccasi
onal
l
ysignif
icant(
duet
o
t
helar
gesampl esi
z e)butwerequi
temodest .

Cor
rel
ati
on Measur
e

0.
18 Michi
gantest(
Lado)
0.
22 Composit
ion
0.
24 Cl
oze
0.
014 SLOPEtest(Krashen,
1976)

Theseresul
tsarepredi
ctabl
e.Wewoul dexpectex
t ensi
onstudentstohavevar
iabl
e,
andoftenv er
ylow,contactwi
thEngli
shascompar edtofull
-ti
mestudents.Somemayhave
l
ivedintheUnitedStat
esmanyy ear
swithoutmuchcomprehensibl
einput.(Si
mil
arr
esul
tsar
e
repor
ted

40
40
i
nKr
ashen,
Sel
i
ger
,andHar
tnet
t,1974,
andKr
ashenandSel
i
ger
,1976.
)

Twostudi
esex amine"ti
meabr oad"t oforeignl anguagest udents,butdiffersomewhat
i
nenv ironmentandal soinresul
ts.Car roll(1967)r epor t
edt hatforeignlanguagemaj or
sin
Amer i
canuniver
sit
ieswhower eabl et ospendt heirjunioryearsabr oadint hecount r
ywhere
thelanguagewasspokenper f
ormedbet teront heFSIr estofl i
steningcompr ehensi
onthan
thosewhohadonl yspentasummerabr oad.Thesummert r
avelers,i
nt urn,outperf
ormed
thosewhohadnev erbeenabr oad.Thesecl earresul t
spr obablyr efl
ectthef actthatsuch
students,mostli
kely
,soughtouti nteraction,andt huscompr ehensibleinputi nthesecond
l
anguagewhi l
etheywer eabr
oad.

I
nChi haraandOl ler( 1978),studentsofEngl ishasaf oreignlanguagei nJapanwer e
studied.Nor elati
onshipwasf ound,howev er
,betweent heamountoft imespentabr oadand
testsofEngl i
sh( r=0.04f orcloze, wi
thothermeasur esproducingsi milarresul
ts;Chihar
aand
Ollerreporthighercorr
elationsbet weent i
meabr oadandsel f-r
epor tofEngl i
shprofici
ency;r=
0.24f orspeakingandr=0. 23f orli
steni
ngcompr ehension).I
ncont r
astt oCarroll
'
sAmer ican
foreignlanguagemaj orsst udy,inwhi chacqui rerswer ei ndai lycont actwi t
ht hetarget
l
anguage,t i
meabr oad int hi
scaseneed nothav eadi rectr elat
ionship withamountof
compr ehensibleinput
.Tabl e2. 3summar i
zesLORst udi
es.

A secondex posur ev ar i
ablet hathasbeenst udi
edi sreport
eduseoft hesecond
l
anguage.Sev er
alstudies( butsur pr
isingl
yfew)hav easkedwhet hert
hosewhosayt heyuse
thesecondl anguagemor eact uallyacquiremor e.Wewoul dexpectasi gnif
icantrel
ati
onship
between" use"andacqui si
tion,si nceusenear lyal waysentai
lscompr ehensibleinput
.Oft he
thr
eest udi esIknowoft hatex plorethisvari
able,twodoi nfactshowacl earrel
ati
onshipwith
secondl anguagepr oficiency .(Failur
et ofindacl earrel
ati
onshipineverycasemaybeduet o
theunreliabili
tyofself
-report; seeOl l
erandPer kins,1978.)

Al
lstudi
esex aminingreportedusei nvol
veadultsubj ects.JohnsonandKr ug(1980)
st
udi
ed72i nt
ernat
ionalstudentsatSout hernIll
inoi
sUni versityandf oundamodestbut
si
gni
fi
cant0.34 cor
relat
ion between profi
ciencyin Engli
sh ( as measur ed byaccuracyof
gr
ammat i
calmorphemesi nobli
gatoryoccasionsinaninter
viewsi tuat
ion)andsubject
s'

41
Tabl 3.Lengt
e2. hofr
esi
dence(
LOR)andsecondl
anguageacqui
sit
ion(
SLA)

a CI=comprehensi
blei
nput
.
b Aver
ageLOR=3- 4year
s.
c Norel
ati
onshi
pbetweenLORandSLApr
esumabl
ydueher
etor
elat
ivel
yshor
tLOR.

42
42
repor
toft heamountofl ei
sureti
met heyspentspeaki
ngandl i
steningtoEnglish.Ol l
er,
Perki
ns,andMur akami(1980),however
,examini
ngasimilarsampl
e,foundnor elati
onship
betweenar eportof"ti
mespentwi thEngli
shspeaker
s"andsecondlanguagepr
oficiency,as
measuredbydi ct
ationandacloze-t
ypegrammartest.

TheHei delbergproject,ascitedinSchumann( 1978b),examinedf act

orspredict
ing
prof
iciency i
n Ger man as a second l anguage forguest-workers( I
tal
ian and Spanish
speakers)in Ger many .They r epor
ted a corr
elat
ion of0.64 between Ger man syntacti
c
prof
iciencyand" l
eisurecontact"withGermansandoneof0. 53bet weenGer manprofi
ciency
and" workcontact "
.Bothleisureandwor kcontactcanplausibl
ybei nter
pretedasindicat
ing
compr ehensi
bleinput.

4.AGE

Ithasbeenpopul ar
lyassumed thatagei tsel
fi sapr edi
ctorofsecond language
profi
ciency,t hatyoungeracquirer
sar ebett
eratsecondl anguageacquisi
ti
on than older
acquir
er s.I
tcanbear gued,however,t
hatageisnoti nitsel
fapr edi
ctorofsecondlanguage
rate or attainment,and t hatheret oo everyt
hing reduces down t ot he quanti
ty of
compr ehensiblei
nputandt helev
eloftheaff
ecti
vefil
ter.

Krashen,Long,andScarcel
la(1979)rev
iewedt
heavail
ableempir
icalresearchonthe
eff
ectofageandsecondl anguageacquisi
ti
onandconcl
udedthatal
lpubli
shedst udi
eswere
consi
stentwit
hthesethreegener
ali
zati
on:

1.Adul
tsproceedthr
oughtheear
lyst
agesofsecondlanguagedevel
opment
fast
ert
hanchil
drendo(wher
eti
meandexposur
ear
eheldconstant
).

2.Olderchi
l
drenacqui
ref
ast
ert
hany
oungerchi
l
dren,t
imeandex
posur
ehel
d
constant
.

3.Acquir
erswhobeginnat
uralex
posuretosecondlanguagesduri
ngchil
dhood
general
l
yachiev
ehighersecondlanguagepr
ofi
ciencythanthosebegi
nningas
adul
ts.

Thus,iti
snotsi
mpl
ythecaset
hat"
youngeri
sbet
ter
":chi
l
drenar
esuper
iort
oadul
ts
onl
yinthelongrun.

The ex
planat
ions f
ort
hese obser
ved di
ff
erences t
hatseem
most

43
43
plausiblet omei nvolvei nputandt helevelorst rengthoft heaf fecti
vefil
ter.First,letus
considert heol deracqui rersr
' at
esuper iori
ty( generali
zat i
ons(1)and( 2)above).Scarcellaand
Higa( forthcomi ng)f oundt haty oungeracqui rersactual l
yreceived" simpler
"inputi nabl ock
buildi
ngt ask,ar esultthatconf ir
msobser vationsmadebyWagner -GoughandHat ch( 1975) ,
andt hatseemst o predictgr eaterspeedf ory ounger ,and notol deracquirers.Scar cell
a
andHi ganot edt hatt heol deracqui rers(adol escents)wer ebet terablet oregulatebot ht he
quant i
tyandqual ityoft heirinput.Theywer ebet teratencour agingspeechandatget t
ingt he
nativespeakert omodi fyi tforgreatercompr ehensibil
it
y .Theycoul d,forexample,askf orhel p,
change t he t opic,and di rectt he conv ersation better.Theyhad,i n otherwor ds,mor e
"conver sati
onalcompet ence" .Thus, despitethesi mpleri nputdir
ect edatthey oungerchi l
dr en,
i
ti sli
kel ythatol deracqui rersact uall
ygetmor ecompr ehendedi nput,andthismaybeakey
factorint heirfast eri
nitialprogress.

Theremaybeotherreasonsf ort
heolderacqui r
ers'super
ior
it
yinrateofacquisi
ti
on.
Adul
tshavemeansofpr oduci
nglanguageearl
ier,of"beatingtheSil
entPeri
od",meansthat
havenothi
ngtodowi t
hnat ur
allanguageacquisi
tionbutt hatmaynev er
thel
esshelpthem
par
ti
cipat
einconv
ersat
ionandhenceobt ai
ncompr ehensi
bleinput
.

Ihavehy pothesizedi near l

ierpapers( see,forexample,Krashen,1981)thatsigni
ficant
Monitorusei sonl ypossi bleaf t
ert heacquirerhasunder gonef ormaloperati
ons,ast agei n
cognit
ivedev elopmentt hatgener all
yoccursataboutpuber ty(I
nhelderandPiaget,1958).The
avail
abil
it
y oft he consci ous gr ammar,t he Moni tor,al
lows adultst o produce formal l
y
acceptableutter ancesusi ngfirstlanguager ulesrepair
edbyt heMoni tor
,asdiscussedear li
er
i
nt hi
schapt er.Whi letheuseoft hismodedoesnotr equi
recompr ehensibl
einput,i
thelpst he
acquir
ertotal kear l
y,topar ti
cipateinconversat i
ons,andtherebyobtaininput,
atleastsomeof
whichwi l
lbecompr ehensible.

Bothex
planati
onsf
ortheolderacquirer
s'rat
esuperi
ori
tyreducetothegreaterabi
l
ityof
theadul
tandolderchi
ldt
oobtaincompr ehensibl
yinput
.Thus,comprehensi
bleinputagainis
hypot
hesi
zedtobethecausat
ivevari
able,andnotageperse.

Thechi
l
d'ssuperi
ori
tyi
nulti
mateat t
ainmenthasbeenhy
pot
hesi
zedt
obeduet
othe
st
rengt
heni
ngoftheaff
ecti
vef
il
terataboutpuber
ty,

44
44
an ev entt hatmay al so be r el
ated t of ormaloper at
ions (Krashen,1981) .As ar gued
elsewher e,thishy pothesi shassev eraladv antages.Fi rst,i
tclaimst hatchil
d-adultdiff
erences
i
nat tainmentar enotduet oanychangei nt he"languageacqui sit
iondevice"(LAD)butar edue
tot hef i
lt
er,af actort hati s,i
nasense,ex ternalt otheLAD.Second,i tisconsist entwiththe
clai
mt hatadul t
sar est i
ll"acquirer
s",thattheyr etainthenat urall
anguageacqui siti
oncapaci ty
chil
dr en hav e.Ital so al l
owsf orthepossi bil
ityt hatsomeadul tscan achi eveex t
remel y
highl ev el
sofcompet encei nasecondl anguageandmayev enbet akenfornative;itpredi
ct s
thatsuch" GoodLanguageLear ner
s"wi llbe, abov eall,goodacqui r
ers,wit
htheabi lit
ytoobt ain
agr eatdealofcompr ehensi bl
einputwi thal ow af fect
ivef i
lt
er.Inmanycases,t hef i
lter
prev entst headul tonlyf rom goingthel astfewi nches.17

5.ACCULTURATI
ON

Asimilarargumentcanbemadeconcer ni
ngSchumann' sAccultur
ati
on
Hypothesis.Schumann( 1978b)hashy pothesi
zedthatacculturati
on
i
sthe" majorcasual v ar
iableinsecondl anguageacquisit
ion"(p.
29).Schumannmai nt ai
nst hat"Secondlanguageacqui si
ti
onisj ustone
aspectofaccul t
uration,andt hedegreet owhichthelearneraccultur
ates
tothetargetlanguagegr oupwillcontr
ol t
hedegreetowhi chhe
acquir
est hetargetlanguage"( p.34).

Whil
et heAcculturati
onHy pothesi
sseemst oaccountforsecondlanguageacquisit
ion
dat
ai nex
tendedsoj ournsituations,iti
seasilyexpressi
bleintermsofcompr ehensi
bleinput
andlowfi
lterlevel
.Acculturat
ioncanbev iewedasameansofgai ningcomprehensi
bleinput
andlower
ingt hef i
l
ter
.Mor eov er,thecompr ehensi
bleinputhypothesi
saccountsforsecond
l
anguageacqui si
ti
oninsituati
onst hataccult
urati
ondoesnotat tempttodealwit
h.

Schumanndef
inest
wot
ypesofaccul
tur
ati
on:

"I
ntypeoneaccult
urat
ion,t
helearnerissociall
yint
egrat
edwit
ht heTLgr oupand,asa
result
,devel
opssuff
ici
entcont
actwithTLspeaker st
oenablehimtoacquiretheTL.I
nadditi
on,
hei spsychol
ogi
call
yopent otheTLsucht hatinputtowhichheisexposedbecomesi ntake.
Typet woaccult
urati
onhasal lthecharacteri
sticsoftypeone,butint hiscasethelearner
regardstheTLspeaker
sasar efer
ence

45
45
 groupwhosel if
est yl
esandv al
uesheconsci ousl
yorunconsciouslydesir
est oadopt.Both
typesofaccul t
urati
onar esuffi
ci
enttocauseacquisi
tionoft
heTL,butt hedisti
ncti
onismade
tostressthatsocialandpsy chol
ogi
calcontactwit
htheTLgr oupi
st heessenti
alcomponentin
accultur
ati
on( asitrel
atestoSLA)thatthatadopti
onoftheli
festy
leandv al
uesoftheTLgr oup
(char
acteri
sticstradit
ional
lyassoci
atedwiththenotionofaccult
urati
on)isnotnecessaryfor
successfulacquisit
ionoftheTL"(p.
29).

Typeoneaccul t
urat
ioni sthus"weaker"thantypet woi nthatitdoesnoti nvol
ve
adopti
onoft henew li
festyle.Si
nceSchumannhy pothesizest hatty
peonei sal
lthatis
necessaryforsuccessf
ulsecondlanguageacquisi
ti
on,wer estr
ictourdiscussi
ontot
ypeone
accul
turati
on.

Typeoneaccultur
ati
oniseasil
yrestatableintermsoft hef
ramewor kpresentedinthi
s
chapt
er:soci
alint
egrati
onwithresul
ti
ngcont act
sleadst ocompr ehensi
bleinput
,whilethe
openpsychologi
calstateSchumannr eferst oisequi v
alenttoal ow f
il
ter.Theev i
dence
Schumannpresentsi
nsupportoftheAcculturati
onHy pot
hesiscanbesi mi
larl
yint
erpret
ed.

TheHei delber
gpr oj
ect,ment i
onedear l
ier
,studiedv ari
ablescorrelati
ngwi thsuccessf ul
acquisiti
onofGer manbyf orei
gnwor ker s.Report
edamount sofl eisur
econt actwi t
hspeaker s
ofGer mancor rel
atedwithsyntacticper f
or mance(r=0. 64)asdi damountofwor kcontact( r=
0.55).Appar ent l
y,ei
therform ofi nteractionwasef fect i
vei nencouragingsecondl anguage
acquisiti
on.( Schumannnot est hat"amongt hebestspeaker s,thosewhohadl i
ttl
el eisure
contactwi thGer mansal lhadconsi der ablewor kcont act"(p.39);thus,someguest -wor kers
whoacqui r
edGer mandidsowi thoutmuchl ei
surecont act.
)Thisconf i
rmst hatitisint
eract i
on,
andt her esultingcompr ehensibleinput ,that"causes"secondl anguageacqui si
ti
on,av i
ew
consistentwi thboththecompr ehensi bleinputpluslowf i
lt
erv i
ewaswel lastheAccul turation
Hy pot
hesis.

Schumann,inr eport
ingtheHeidel
bergresear
ch,alsonotesthat"l
earner
swhosewor k
requi
redcommuni cati
onwi thco-worker
sdidbet t
erinGer manthanwor kerswhopr ovi
ded
servi
ces( hai
rdr
essers,kit
chenhelp,etc.
)"Al
so,"l
earner
swhowor kedinanenv i
ronmentt
hat
wasnoi syorwhichconst r
ainedmovementwer eatadisadvant
age".

46
46
Theseresul
tsalsosuggestt
hatt
hosewhower
eabletoi
nter
actmoreinthet
argetl
anguage
acqui
redmor eGerman,al
lofwhi
chmeansmor
einputmeeti
ngther
equir
ementsoft
heinput
hypot
hesis.

Schumann dr awsa par all

elbet ween nat ur
al( "free"ori nformal)second l anguage
acquisiti
on and t he pidgini
zation-decr eoli
zation cont i
nuum,suggest i
ng thatear lysecond
l
anguageacqui si
ti
onissi mil
art opi dginizat
ion( secondar yhy bri
dization)andt hatlatesecond
l
anguageacqui si
ti
oni ssimilart ot hemesol ectandacr olectstagesofdecr eolization.18 As
evidence,hedescr ibesthecaseofAl berto,aSpani sh-speaki ngadul tli
vingintheUni tedSt ates
who appear edt obeataconsi derablepsy chol ogicalandsoci aldi st
ance19 f rom Engl i
sh
speaker s,andwhosespeechshowedmar ked si gnsofpi dgini
z ati
on,i .
e.l ackofsev eral
grammat i
calmor phemes,l i
ttl
euseofi nversioni nquest i
ons,anduseofmor epr imiti
ve
tr
ansi t
ionalf ormsi nnegat i
on( Schumann,1978apr ov i
desdet ai l
s).Albertot husf i
tst he
Accul t
urati
on Hy pothesis,since he showed l ittl
e second l anguage acqui sit
ion and l i
ttl
e
acculturati
on,definedast hedegr eeofsoci alandpsy chologicaldi st
ance.Al berti sal soquite
consistent wi tht he theoreticalf ramewor k pr esent ed her e and t he hy pothesi st hat
compr ehensi bl
ei nput and f ilter levelar e pr imar y det erminant s of second l anguage
acquisiti
on.Al berto,i
tcanbecl aimed,r eceiv edl it
tlecompr ehensi bleinputi nEngl i
sh( he
wor kednight s,di
dnotownaTVset ,didnotat tendESLcl asses,andmadel i
ttleef forttogett o
knowEngl i
shspeaker s.,accordingt oSchumann,1978a) ,andpr obabl yhadast rongaf fecti
ve
fi
lt
eraswel l
.20

Stauble'
ssubj ects,r
epor tedi nSt auble( 1978)arealsoconsi deredt obeev i
dencef or
theAccul tur
ationHy pothesi
s.Al lthreewer eSpani shspeaker swhohadbeeni nt heUni ted
States f ormanyy ears,and who had appar entl
y" f
ossil
ized"atdi fferentlev elsi nt heir
developmentofnegat i
on.St aubl eat t
empt edt or el
atetheirpr ogressi nsecondl anguage
acquisition,asr eflectedbyt het ransit
ionalf ormst heyusedf ornegation,andaccul t
uration,
measur ed byan i nformalquest i
onnaire pr obing soci
aland psy chologicaldi stance f rom
speaker sofEngl ish.Thesubj ectXav i
ershowedt heleastpr ogressinEngl i
shnegat ion,but
alsoshowedt hel eastsocialdistance.St auble'squesti
onnairer ev
ealedt hathispsy chological
dist
ance,howev er,was gr eat ert han t hatoft he others,whi ch is consistentwi tht he
Accultur at
ionHy pot hesis.Paz,t hemostadv ancedspeaker,hadt hegreatestsoci aldi stance,
but,along

47
47
wit
h Mar i
a,t he other subject
,had r el
ati
vel
yl ow psychol
ogical di
stance. St
aubl
e's
i
nterpretat
ionoft hesefindi
ngsi st
hatpsy chol
ogi
caldist
ancemaybeamor ei
mportant
determinantofaccult
urat
ion,
andhencelanguageacqui
sit
ion,t
hansocial
dist
ance.

Staubl
e'sdatacanal sobeanal yzedint ermsofourt heoret
icalframewor k.Sinceall
threesubjectshadbeeni ntheUni t
edSt atesformanyy ears,al
lthreehadhadconsi derable
compr ehensi
bleinput( r
ecallourearli
ergeneralizati
ont hatLOR,whenov eral ongper i
odof
ti
me,doesnotpr edictSLA,ahy pothesisconsistentwi thWal getal
ber .'
sdiminishingretur
ns
hypothesis)
,enought oal l
ow a" zerofil
ter"acquirertor eachnat i
ve-speakerlevels.Wecan
thensi mplyhypothesizethatitwasPaz' slowerf il
ter,refl
ectedbyt hel owerpsy chologi
cal
score,thatal
lowedthisacquirert
omakemor epr ogress.21

Fi
nall
y,itcanbecl aimedt hatt hecompr ehensibl
einputpl usf i
lt
erposi t
ionismor e
general.The accul t
urati
on hy pot
hesi s predicts second l anguage acqui si
ti
on onlyi n
i
mmi grati
onandex tendedsoj ournsituations.(I
ndeed,itisunfairtoaski ttoaccountforother
sit
uati
ons,sinceSchumann hasmadei tv er
ycl earinhiswr it
ingst hatt heaccul
turati
on
hypot
hesisisdesignedt oaccountf orsecondl anguageacquisitiononl yinthissit
uati
on.)The
theor
yofSLA pr esentedint hischapt ercannotonl yaccountf orex tendedsojournand
i
mmi grantSLAbutal sopredict ssuccessi nthecl assr
oom,asdet ailedinChapt ersI
II
,IV,and
V,andisclai
medt obeappl icabletoal ll
anguageacqui siti
on.

The Accul turat

ion Hy pothesis has consi derabl e mer i
t.I tmay be t he case t hat
accultur ationisthemostef f
ecti
vewayofl oweringt heaf f
ecti
v efil
terandget ti
nginputf or
i
mmi gr antsandlong- t
ermv i
sit
ors.Figure2.3at tempt st ocapt uretheparall
elbetweensecond
l
anguageacqui siti
onandt heeffectofaccul tur
ati
on." Fr ee"secondl anguageacqui si
ti
onand
thecont inuum aresimilarinthataccul t
urati
onmaybet he" motivati
ngf or
ce"behindboth.22

Creolespeaker sgr aduallyacquirecl oserv ersi

onsoft hest andardast heyar eacculturated
i
nt ot het argetcult
ure.Thi saccul t
ur ati
onbr i
ngst hem i ntocont actwi t
hspeaker soft he
standar d, andmakest hem mor e"open"t othei nput(lower st hefil
ter)
.Also,accult
urat
ionmay
"motiv ate"secondl anguageacqui siti
on.Ast heindivi
dualacqui reraccultur
atesintoacul ture,
heobt ai nsmor einputv i
amor ei
nteract i
on,andi smor e" open"t oit.Thediff
erence

48
48
 Fi
g.2.
3.Accul
tur
ati
on,
pidgi
niz
ati
on-
decr
eol
i
zat
ion,
andsecondl
angaugeacqui
sit
ion.

Secondlangaugeacqui
sit
ionandt hepi
dgini
zat
ion-decr
eol
i
zat
ionconti
nuum ar
esimil
arin
thatbot
hprogressvi
acompr ehensi
blei
nputsuppl
iedi
nalowfi
l
tersi
tuat
ion(ar
eainsi
dethe
box).

a Encul
tur
ati
on=" theprocessbywhichanindi
vi
dualassimil
atest
ohi
sowncultur
eorto
somesegmentofit
",i.
e.thecaseofEuropeaneli
teprofessi
onal
swhoacqui
reEngl
i
shinthei
r
owncount
ri
es,(
Fordiscussi
on,seeSchumann(1978b)
, pp.47-
48).

49
49
i
st hataccult
urati
onist
henecessarymot i
vat
ingforcef
ormovementalongthedecreol
izati
on
continuum,whileiti
sonlyonewayt obri
ngthefilt
erdownandobtai
ncompr ehensi
bleinput.
Inputcanbeobt ai
nedwit
haccult
urati
on,andther
ear emanyt
echni
quesforbri
ngingdownt he
fil
terthathav
enot hi
ngt
odowi t
haccultur
ati
on.

Not
es

1 Theacqui sit
ion-l
earningdist
incti
oni snotnew wi thme.Sev er
alotherschol
ar shave
founditusefultoposi tsimil
arkindsofdi ffer
ences.Bial
ystockandFrohli
ch(1972)distingui
sh
"i
mpl i
cit
"and" expli
cit"lear
ning,andLawl erandSel i
nker(
1971)discussmechanismst hatguide
"aut
omat ic" performance and mechani sms t hat guide "puzzl
e and pr obl
em sol vi
ng
perfor
mance"( p.35) .Also,Corder(1967)andWi ddowson( 1977) suggestsimil
arideas.

2 Thoseofuswhohav est udiedlanguageswi thagr eatdealofi nflect

ionalmor phol ogyi n
school ,usingmet hodst hatf ocusongr ammat icalaccur acy ,oftenhav ef i
rst-handex per i
ence
witht hisphenomenon.Consi derwhathappensj ustbef or eagr ammart esti nal anguagesuch
asGer man;st udentscar eful
lyr eviewt hei nf
lectionalsy stem ( der-das-di
e;den- das-di
e;plust he
l
istofpr eposit
ionst hatt akedi f ferentcases)ont hewayt ot heex am.Assoonast heysi tdown
i
ncl asst otakethet est,theyi mmedi atelyscribblewhatt heycanr ememberoft heinflectional
system ont hesideoft hepage, sot hatwhent heyneedt hecor rectmar ker,theycanf indi tand
usei t.Att heendoft heex am,bef orehandi ngi nt hepaper ,theyer aset heirnot es.The
mor phol ogyont hesi deoft hepagei s,mostl ikel
y,l ate-acquired,andunav ail
ableinr apid
conv ersationformostpeopl e.Thenot esont hesi de,t
hen, actli
keaconsci ousMoni tor,raisi
ng
theaccur acyoftheout putinsi tuationswher et hest udenthast i
me, isfocusedonf orm, andcan
accesst her ul
e;gr ammart est sf illthesecondi ti
onsni cely.Student sthusdomuchbet teri n
ter
msofgr ammat icalaccur acyonsucht estst hant heywoul di nfreeconv ersat
ion,thel ate-
acqui r
ed, ornot-yet-
acqui red, i
t emst hatarelear nablerisingi nrank.

3 Aninter
esti
ngpar al
lelhypothesisist hatwewillseegr eat
ernumbersoft r
ansit
ional
formsi nMoni tor
-f
reecondi t
ions.Thel i
teratureisconsistentwiththi
shy pot
hesi
s,sincethe
transiti
onalformsnot edforadul tacquir
ershav eal
lbeenf oundinsubjectswhoappeart obe
non- users,orunder -users oft he conscious Moni tor
,f orex ample,Schumann's Albert
o
(Schumann,1978) ;Nelson's McGilluniver si
tyjanit
ors( Nelson,1980),and Hanania and
Gr adman' sFatmah( HananiaandGr adman,1977) .Thi
sispr edict
abl
e,sincetr
ansi
ti
onalforms
arehy pothesizedtorefl
ecttheoperationoft heacquir
edsy stem.

4 Tobemor epreci se,speakingski l

lsthatdependonacquiredcompet enceemer geover
ti
measar esul
tofcompr ehensibl
ei nput.Ther eappeartobe,howev er,atl
easttwoway sof
beati
ngt hesystem,atl eastov ertheshor trun.Wecanpr oduceusingmemor iz
edlanguage,or
rout
inesandpat ter
ns( KrashenandScar cel
la,1978),andwecanalsopr oducebyusingthefir
st
l
anguagesur facestructurepl usconsci ousgr ammar( L1plusMonitorMode) .Asweshal lsee
l
ater,bothofthesemet hodsofper formingwi thoutacqui
redcompetencehav edrawbacksand
l
imit
at i
ons.

5 Int
erl
anguagetalk,thespeechofsecondl anguageacquir
erstoeachother,mayormay
notbeusefulforacqui
sit
ion.Thisisanimport
antquest i
onthat,
tomyknowl edge,hasnotbeen
dir
ect
lydealtwithinthepr ofessi
onall
i
terat
ure.Argumentsi nfavorofi
tsutil
it
yf orl
anguage
acqui
sit
ionarethese:i
tsati
sfiesthei
nputhypothesisi
nthatiti
smeant

50
forcommuni cati
onandmightcontaininputatsomeacqui r
ers'i+1.Ontheotherhand t
here
i
st hequest ionofwhethertheungr ammat icali
tyofmuchinterl
anguagetal
kout wei
ghsthese
factor
s.Al so,muchint
erl
anguaget alkinputmi ghtbetoosi
mpl eandmaynotcont ni+1f
ai or
themor eadv ancedacquir
er.SeeKr ashen( 1980,1981)foradi scussi
onofsomeoft he
empi r
icalevidencet
hatmightshedlightont hisissue.

6 I
nar ecentst
udy,
M.Long( 1980)repor
tedt hatf
oreignertal
kdiscoursedidnotcontain
si
gni
ficantl
y mor everbs marked forpresentt ense than nativ
e speaker-nat
ive speaker
di
scourse.I
tisthusnotmoreinthe"now"ofthe"hereandnow" ,toparaphraseLong.

7 Alookatsomeofthememor i
zedsentencesandphr aseschi
ldr
enpickupduringthe
si
lentper i
odconfi
rmsthei
rut i
li
tyinav ar
iet
yofsocialsituati
ons.Quit
eoften,however,t
he
chil
dren donotal ways acquir
et heknowl edgeofex actl
ywhenandhow t ouset hem.A
part
icular
lyviv
idexampl
eist hechi l
d,whohadbeeni ntheUni tedSt
atesapproxi
matelytwo
mont hs,whogreet
edanacquaintanceofminewi t
h"Iki
cky ouass."

8 Consci ousMonitor
ingneednotal waysr esul
tint heful
lrepai
rofanL1i nfl
uenceder r
or.
Iftherepai
rj obappearstobet oocompl exf ortheMoni t
ortodealwi th,theper for
mermay
simplyaborttheentir
esentenceandt rytoex pressthei deainasimplerway .Thi
smaybet he
causeoftheav oi
dancephenomena,f i
rstreportedbySchacht er(1974).I
nSchacht er'
sst
udy,it
wasshownt hatChineseandJapanesespeaker sproducedf ewerrelat
iveclausesi nEngl
ishas
asecondl anguaget handi dFarsiandAr abicspeaker s,butwer emor eaccur ate.Schachter
rel
atesthi
sr esultt
oL1-L2diff
erences:ChineseandJapaneser el
ati
veclausesar econstruct
ed
totheleftoft heheadnoun,whi leFarsiandAr abi
c,l
ikeEngl i
sh,haver el
ativeclausestothe
ri
ghtoftheheadnoun.

Onepossi bleint
erpretat
ionisthattheChi neseandJapanesespeaker sinSchacht er'
s
studyconsciouslyknewt hecor r
ectEnglishrel
ati
vecl auser ul
ebuthadnotacqui redit.Also, i
n
theirproduct
ionofEngl i
sh,theyuti
li
zedtheirL1rule.TheirMoni t
orwast huspresentedwi t
ht he
taskofmov i
ngr elati
veclausesar oundaheadnoun,av erycompl exoperati
on.Inmanycases,
subjectssimplydeci dedt hati twas notwor tht he effort
!Whent heydi
dpr oducer elati
ve
clauses,howev er,theywer eaccurate.Thesewer ethecaseswhent heywentt othetroubl eof
applyingadiffi
cultrule.

Avoi
danceisthuspredi
ctedi
ncaseswherearulehasbeenconsci
ousl
ylear
nedbutnot
acqui
red,andwhent heL1andL2r ul
esarequi
tedif
fer
ent,wher
erepai
rbytheMonitorr
equi
res
dif
fi
cultmentalgy
mnast i
cs.

Avoidancei sal sopr edi

ct edincaseswher etheper formerconsci ouslyknowst her ule
i
mper f ect
ly,notwel lenought omaket henecessar ychancebutwel lenought oseeami smat ch
betweent heL1r ulehehasusedandt hecor recttargetlanguager ule.Sincehecannotr epair
butknowst her ei saner ror
,hecanex ercisehi sopt iontoav oi dt hest ructure.Kleinman' s
avoidancedat a( Kleinman,1977)f it
st hi
sdescr ipti
on.Hi sAr abi c-
speaki ngsubj ectsshowed
evidence ofav oiding the passi vein Engl ish,and hi s Spani sh- and Por tuguese-speaki ng
subject savoidedi nfinit
ivecompl ementsanddi rectobjectpronounsi nsent enceswi t
hinfi
nitive
compl ement s( e.g." Itoldhert oleave")
.I nbot hcases,accor dingt oKl einman,cont rastive
analysispr edictsdi ff i
cult
ies.Thesesubj ects,unl i
keSchacht er
's,wer enotunusual l
yaccur ate
withtheseconst ructionswhent heyproducedt hem.Int hiscase, itispossi blethatt hesubject '
s
knowl edgeoft her ulewasnotcompl eteenought oeffectaper f
ectr epai r
,soav oi
dancewast he
result
.

I
nbothcasesdescri
bedabove,consci
ousrul
esserv
eaf il
teri
ngfuncti
on,t
ell
i
ngt he
per
formerwherehisL1rul
ediff
ersf
rom t
heL2r ul
e.I
nonecase,
repairi
spossibl
ebutdif
fi
cult
,
andintheot
hertheconsci
ousrul
edoesnotpermitr
epai
r.

9 BasedonHy l
tenst
am'sdat
aont heacquisit
ionofnegati
onbyadul
tacqui
rer
sof
Swedi
sh(Hyl
tenst
am,1977),
Hammarber
g(1979)arguest
hatacqui
rer
smaybegi
nat
51
51
di
f f
erentdevel
opmentalst
ages dependi
ng on t
hei
rfir
stlanguage.The nor
malcourse of
developmentintheacqui
sit
ionofnegati
oninSwedishconsi
stsofthefol
lowingt
ransi
ti
onal
stages:

(1a) Acquir
ersplacet
henegat
ivemarkerbef
oreal
lot
herpar
tsoft
heVP,
befor
etheauxi
li
aryandt
hemainverb.

(
1b) Acqui
rer
spl
acet
henegat
ivemar
keraf
tert
heaux
il
iar
ybutbef
oret
he
mai
nver
b.

(
2) Post
-ver
bal
negat
ion.

I
nsubsequentstages,acquir
ersmoveclosertotheSwedishr
uleofpost
-ver
balnegat
ioni
n
mai
nclausesandpre-v
erbalnegat
ioni
nsubordi
nat
eclauses.

Accordi
ng to Hammar ber
g,speakers oflanguages t
hathav e pr e-
verbalnegati
on
(Ser
bo-Croati
an)typi
call
ystartthi
sdevelopmentalsequencesatthebegi nning,atstage1a.
Engli
sh speakers,however,appearto begin at1b.Wedo notsee Engl i
sh speaker
s,in
Hylt
enstam'sdata,whoproduce"neg+aux "st
ruct
ures.Si
nce1b" i
sanEngl i
sh-li
kesoluti
on"(p.
10),onecanhy pot
hesi
sthatEngli
shspeakersski
ppedthe(1a)tr
ansit
ional
st age.

Therearesev
eralpossibi
li
ti
eshere.First
,Hammar berg'
ssuggest
ionmaybet r
ue.I
fso,
i
facquir
erscanskipatransi
ti
onalstaget,
j whenthei
rlanguagehasaruleident
icalt
o t
j+1,t
his
i
mpl i
esthattj wasnotessent i
al-
-i
tdidnothav etoserveasi sdoesnotr
.Thi uleoutthe
possi
bil
i
tythatt
jwoul
dhav ebeenuseful
.

Asecondpossi bi
li
tyist hattj waspr esent, butescapedt heobser ver'
snotice.Indeed, i
t
mayhav ebeenpr esentast jbutnev eruttered.Adul tperf
or merswhohav econsciousl ylearned
the target languager ul
e,orwho hav e ev en lear nedpar tsofit,maybeabl et ouset he
consciousMoni tortodetecttransi t
ionalerrorsandei t
herav oidthem inper f
ormanceorr epai
r
them (see di scussion infoot note7 on av oidance) .Theymay ,howev er,havemor e ofa
tendencyt oacceptsucht r
ansi t
ionalf or
mswhent heycoi ncidewithanL1r ule,ev
eni ftheyar e
error
s( Schachteretal .
,1976).Thi scouldex plainwhyt ransiti
onalformst hatareunlikeL1r ul
es
arelessf requentlyseeni nper formance.I tshoul dbenot ed,however,thattheSer bo-Cr oati
an
speakersinHy l
tenstam'sstudydi dshowcl earsi gnsofst age1b, whichdoesnotcor respondt o
anyruleinSer bo-Croati
an.

Therear ethusatl easttwopossi bili

ti
es--
theEngl i
shspeaker sdidindeedskipast age,
whi chimpliest hattheskippedst agemaynothav ebeencr ucialtofurt
herdev el
opment ,orthe
stagewas" there"butundet ected,duet oi t
sshortdur ati
onand/ oritshavi
ngnotbeenusedi n
theper for
mer '
sout put.Consistentwit
hSchumann' sfindings(Schumann, 1979),t
het r
ansi t
ional
staget hatcoi ncideswitht heL1r ul
ewasqui t
eev ident,bothint hecaseofSer bo-Croati
an
speaker s(stage1a)andEngl ishspeaker s(stage1b).As suggested in the t ext,t hi
s
stagemayhav e,ineachcase,beent wost agesi none,f i
rsttheL1r ul
e,andt hent he" r
eal"
transit
ionalstage, wit
honlythel at
terhelpingrealprogresstocont i
nue.

10 Sever
alscholarshav epoi ntedoutt hatt hi
sv i
ew oftransferistoostronginthatit
predictst heoccur r
enceof" tr
ansf er"er r
orsthati nfactdonotoccur .Thisproblem canbe
resolvedbyposi ti
ngsev eralconstraintsont ransfer,orcondi
ti
onst hatmustbemetbef or
ea
performercansubst i
tut
eaf i
rstlanguager ul
eforsomei +1.
Zobl( 1980a,b,c)notest hattheL1r ul
ei t
selfmustbeapr oduct
iveone.Thisaccounts
forthef actthatFrenchspeakersacqui r
ingEnglishasasecondl anguagedonotmakeer rorsof
thekind:

52
52
 Johncomeshe?
af
tert
heFr
ench:
Jeanvi
ent
-i
l?

TheFr enchrul
e,accordi
ngtoZobl
,isnolongerpr
oducti
vei
nFrench.Ci
ti
ngTerr
y(1970),Zobl
notesthatiti
smai nl
ylimi
tedt
opresentt
ensecontex
ts,ani
ndi
cati
onthatt
herul
eisbecoming
unproducti
ve.

Kellerman( 1978)prov i
desanot hercondi
tionontransf
er :theperformermustper cei
ve
thetransferredruletobepot ential
l
ynon- l
anguagespecifi
c.Kel
lerman' sorigi
nalex peri
mentsin
l
ex i
caltransfershowedt hatforeignlanguagestudentswerelesswi l
l
ingtot ransferfeatur
esof
wordst heyconsi deredtobel ess" core".Forexample,aDutch- speaki
ngst udentofEngl i
sh
wouldbemor elikelytopresumet hathecoul dtr
ansfertheDut chv erb'brechen'(break)inan
Engli
shsent ence:
Hebr okehisleg.
thanin:
Thewav esbrokeont heshore.

Asimi
l
arconstrai
ntexistsinsyntax.Dutchstudent
sofEngl
ish,Kell
ermanr
epor
ts,wer
enot
wi
l
li
ngtoacceptali
ter
al t
ransl
ati
onintoEngli
shoftheDut
chequival
entof:

Thebookr
eadwel
l
.

apparentl
ybecausetheint
ransi
ti
veuseofr ead wasper
cei
vedt
obel
anguage-
speci
fi
cand
i
nfrequent(seeal
soJor
dansandKell
erman,
1978).

Anotherconstr
aintcomesf r
om t hewor kofWode( 1978),andaccountsf orthef i
nding
thatL1i nfl
uencederror
sdonotseem t ooccuratal lst
agesoft heacquirer'
sdev elopment .
Wodest atesthatforanint
erl
ingui
sti
cer rortooccur,t
heL1r uleandt heL2ruleitsubsti
tutesfor
mustmeeta" cruci
alsi
mi l
ari
tymeasure"( p.116).I
notherwor ds,i
fanL1r uleistobeut i
l
ized,it
mustbepr ecededbysomei oftheL2t hatdiff
ersfrom i
tonlyinmi ni
mal ways.Wode'sex ampl e,
from chi
ldsecondl anguageacquisi
tionofEngl ishbyGer manspeaker s,i
ll
ustrat
est hispoint
nicel
y.Wodenot esthaterr
orssuchas:

Johngonott
oschool

occuri n which Ger

man-li
ke post-v
erbalnegati
on is used.These err
ors ar
e notfound i
n
beginningacquir
ers,
butoccur
, accordi
ngtoWode, onl
yaf t
ertheacqui
rerhasreachedt
he"aux-
negation"st
ageandal r
eadypr
oducessent encessuchas:

Johncannotgo.

Theacquir
erthenover
gener
ali
zest
henegat
iver
ulef
rom post
-aux
il
iar
ytopost
-ver
bal
,anduses
thef
ir
stlanguager
ule.

11 Thereisanotherwayinwhi chuseoft heL1mayi ndirect

lyhel psecondl anguage
acqui
sit
ion.Theexist
enceofcognateswil
lhelpt
omakeinputcompr ehensible,
evenifform and
meaningarenotidenti
calacr
osslanguages.Thi
sfact
orwi
lli
ncreasether ateofacquisit
ionbut
notal
tertheorder
.

12 Thehy pothesist hatL1rulescannotcont r

ibut
etor ealprogr
essimpli
esthatfossil
ized
useofaL1r ul
ei sthe" endoft heline"f
oracquisit
ion.Doest hi
smeant hatasingl
eL1er ror,a
singl
epr ol
ongedsubst i
tut onofsomei+1hal
i tsallacqui
siti
on?Itonlyi
mpliesthi
sifweaccept
ast r
ict
lyli
nearviewoft henat ur
alorderhypothesis,t
hatthereisonlyonestream ofprogress
thatacquir
ersfol
lowi nst ri
ctsequence.Clear
ly,t
hisisnott
he

53
case.I fitwere, acquirer swoul dalway sshowusj ustonet r
ansi ti
onaler r
oratat ime!Ofcour se,
i
ndiv i
dual s show us many er rort ypes atonce.Thi si ndicat es thatsev eralst reams of
developmentar et akingpl aceatt hesamet i
me.Thesest reamsappeart obecor r
elated;a
performeratagi venst agei nonest ream wi llusual l
ybeatapr edi ctablestageinanot herst ream.
Schumann( 1980)pr ov idesgoodev idencef ort his,notingthathi ssubj ectswhower eatt heno+
vstagei nnegat ionproducedf ewr elati
vecl ausesorr elati
vecl auseswi t
houtrelati
vepr onouns.
ForL1acqui siti
on, Shipl ey,SmithandGl eitman( 1969)r epor tthatv erbphr aserelat
edi temsar e
correl
at edfairl
yhi ghlyf oror derofacqui si
tion, andnounphr aser elatedi t
emsar ecor rel
at ed, but
agreementacr osst hegr oupsi snothi gh( seeal soKr ashen,MaddenandBai l
ey,1975;and
Ander sen,1978,f orsimi l
arsuggest ions).Ofcour se,itisqui t
epossi blethattransit
ionalf or ms
orrulesf r
om onest r
eam mayhel poutt hosei nanyot herbyser v
ingasi .I
fsayt enpar allel
str
eamsofdev elopmentoccuratanygi vent imei nanacqui rer,itmaybet hecaset hatagi v en
str
eam wi l
linteractwi thsome, butnotal l,oft heot hersinthisway .

13 "
Int
egrati
ve"mot i
vati
onr efer
stot hedesi reto"belike"speakersoft hetargetl
anguage.
Inforei
gnlanguagesituat
ions( e.
g.studyingFr enchinAnglophoneCanada) ,studentswithmore
i
ntegrati
vemot i
vati
onar eusual l
ysuperior,especiallyoverthelongr un(Gar dnerandLamber t,
1972).Insit
uati
onswher ethereissomeur gencyinsecondl anguageacqui siti
onand/orwher e
thereislessdesiret o"int
egrate",thepr esenceofi nt
egrati
vemot ivati
onmaynotr el
ateto
secondlanguageachievement .Rather,"
instrument al
"mot i
vati
on,thedesiretouset helanguage
forpracti
calmeans,maypr edictsuccessbet ter(Lukmani,1972;Gar dnerandLamber t,1972;
Oller
,Baca,andVigi
l,1977).

14 St
evi
ck(
1980)pr
ovi
desapoi
gnantex
ampl
e,ast
oryr
elat
edt
ohi
m byoneofhi
sst
udent
s:

"Foury earsagoIwasl ooki

ngf oranykindofj obIcouldf i
nd.Ihappenedto
getonet eachi ngESLtoacl assofsi xwomenf rom v ari
ouspar t
softhewor l
dwho
spokenoEngl ish.IhadneverheardofESLbef ore.Thesal arywaspoorandIdi dn't
knowi fIwant edtopursueat eachi
ngcareer,
theref
or emyappr oachwasv erycasual
andlowpr essure.Mymet hodusuallyconsi
stedoft hinki
ngupat opi
ctotalkabout,
i
ntroducingit,andencouragingeachstudenttoexpressherf eel
ings.

I
nspi
teofmycasualapproach,t
heteachi
ngjobwasex tr
emelypleasant
.I
hadadeepempathyforany
onewhowasf aci
ngal anguagebarri
erbecauseIhad
j
ustr
etur
nedf
rom atr
ipar
oundtheworldal
oneasamonol ingual
.

Theyall
st ar
tedspeaki
ngEngl
ishfai
rl
ywel laft
erthefi
rsttwoweeksofclass.
Irememberawomanf rom Col
umbi
at ell
i
ngmet hatshehadn' tspokenEngli
sh
befor
ebecauseshewasaf rai
dofmakingmi st
akes.Afterbei
ngi ncl
assf orawhil
e,
she spoke English and made mist
akes and didn'tcare.Idi dn'
tat t
ach much
si
gnif
icancetot hepr ogr
essthatt
hewomenmade.Ihadnoi deahowl ongitt
ook
peopl
et olear
nal anguage.

Graduall
yIbecamequi tecareer
-ori
ented,andmadeaconsci ousdecisi
ont o
tr
yt obeat op-notchESLteacher.Ihadguil
tfeeli
ngsaboutt hecasualwayinwhi chI
hadt aughtthose f i
rstsi
xwomen,and myt eachingevol
v ed i
ntothe tr
adit
ional
authori
tari
anst y
lewiththetext
bookdomi nant.Overtheyears,i
thasgottentowher e
Ifeelfr
ustrat
edi fast
udenttakesclasst
imet orelateapersonalanecdot
e.

Ican l ook back on these f oury ear

s and see agr adualdecl i
ne inthe
performanceofmyst udents.Unti
lrecently
,Ihavebeenassumi ngt hatIneededt obe
mor eattent
ivetotheirmistakesinor dertospeedtheirprogress.Mypr esentstyl
eof
teachingby passesthest udents;feelingsandbasi cneeds,andconcent rat
eson
met hod.Ineverseesuccessesl i
ket hosefir
stsixl
adies."(Fr
om St evick,1980,pp.
4-5).

54
54
15 "
Int
ermediat
e"heremeansknowingenoughEnglishtobeabl etotakeatleastapar
ti
al
academicload,butnotbei
ngableto"passout"oftherequir
eduniversi
tyEngli
shasasecond
l
anguager equi
rement.Thenormalsituat
ionfortheintermediateatt heuniver
sit
yistobe
enrol
ledi
natleastoneESLclassi
nadditi
ontooneormor esubjectmattercour
se.

16 Ther esearchci tedheredeal sex clusivel

ywi t
htheef f
ectofinstructionont heacqui sit
ion
ofsy ntaxandmor phology.Untilrecentl
y ,li
ttleworkhadbeendonet hatex ami nedt heef f
ectof
i
nst r
uct i
onont heacqui sit
ionofpr onunci ati
on.Pur cel
landSut er(1980)r epor tthatacqui sit
ion
ofpr onunciationofEngl i
shasasecondl anguagewaspr edictedbyt hef ollowingf actors:(1)
Theacqui r
er '
sf ir
stl anguage( speakersofAr abicand Far siweresuper iort o speaker sof
JapaneseandThai );(2)Theamountofi nteracti
onwi t
hEnglishspeaker s;(3)Per formanceona
testofphonet i
cabi l
ity;and(4)Thedegr eeofconcer nthespeakerhadabouthi saccent.Fact or
(2)appear stober elatedtocompr ehensi bleinput,whil
e(3)and( 4)mayber elat
edt olear ni
ng.
(1)r ef
lectst heconsequencesoff all
ingbackont hef i
rstlanguage.Theamountoff or mal
classroom traininginESL,howev er,didnotr elat
et opronunciati
onabi l
ity,ev enwhencour ses
wer especifi
cal l
yaimedatt eachingpronunci ati
on.

17 Some st udies seem t o show t hatage ofar r

ival( AOA)pr edi cts second l anguage
attainmentf orchi l
dr en- -t
hati s,t hatt hechi ldwhoar rivesatagesi x ,f orex ampl e,wi llattain
higherl evelsofpr oficiencyt hant hechi l
dwhoar rivesataget en.Whi leAOA doespr edict
ultimat e at tainmentf orchi ldren as a gr oup as compar ed t o adul ts as a gr oup,cl oser
ex ami nat i
onr eveal st hatAOA persei snotaf actorf orchi ldrenconsi der edal one.I ncases
wher eAOAseemst obeaf act or,i
tcanbear guedt hatLOR, andul timat el yCI ,isreal l
ycausat i
ve.
Cummi ns( 1980)hasper formedsuchar eanalysisofRamseyandWr ight 'sdat aon
1,200i mmi grantchi ldreni nCanada( RamseyandWr ight,1974) ,andr eachest hisconcl usion,
not i
ngt hatwhenAOAi scont r
oll
edi nRamseyandWr ight '
sdat a,chi ldr enwi thlongerLOR' s
per form bet teri nav ar i
et yoft asks.Cummi nsal sof oundt hatwhenLORi scont roll
ed,howev er,
chi l
dr enwi thy oungerAOA' sar enotnecessar i
lybet t
er -
-i
nmanycases,t heopposi t
ei st rue.
Mi nour a( 1979)canal sober einterpreted.Shest udied44Japanesechi l
dr enwhohadbeeni n
theUni tedSt at esf orar angeofonet oei ghty ears.Whi leLORpr edi ct
edat tainment( r=0. 79),so
didAOA( r=- 0.75) (
asent encei mi t
ationt estwasused) .Al lthechi ldreni nt hesampl ehad
arrivedi nt heUni tedSt atesataboutt hesamet i
me,howev er
,soLORandAOAwer ehi ghl y
cor related( r=- 0.95) .Itt husmaybear guedt hatLORandt huscompr ehensi bleinput ,wast he
truecausat ivef act or.( Accor dingt omycal culations,t hecor r
elat i
onbet weenAOA andSLA
reducest or=0. 005whent heef f
ectofLORi sremov ed!)TheHei del ber gpr oject,discussedi n
thet ext,al sor epor tsar elationshi pbet weenAOAandSLA,t hist i
meamongadul tst akenasa
group.Thi sal soseemst obeaconf ound,si nceol dersubj ectsseemedt ospendl esst ime
speaki ngGer man( r=- 0.32bet weenAOAandr epor tedl eisuret i
meuseofGer man) .Par tial
cor relationpar ti
all
ingoutt heef fectsofi nteract i
onwi thGer manspeaker sr educest her epor t
ed
cor relationof- 0.57bet weenAOAandSLAt or=- 0.49.Thi scoul d(andshoul d)goev enl ower
wi thamor er eli
abl emeasur eoft heamountofcompr ehensi bl einputsubj ect sact uall
ygot .

18 Pidgini
zat
ion"occur swhenspeakersofdi f
ferentlanguagescomei ntoli
mit
edcontact
andanaux il
iar
yv ehi
cleofcommuni cati
ondev elopst of acil
i
tateinteract
ionamongt hem."
(Schumann,1978b,p.40) .Secondar
yhybridi
zati
oni saf orm ofpidginizati
onthatoccursi
fa
"st
andar
df orm"ofat ar getlanguageisavail
abl
e.Itper si
stsonl yi
fspeaker sremai
natsocial
and psychologi
caldistance from speakers ofthe nor m.( From Whi nnom,1971,cited by
Schumann, 1978b)
.

55
55
Decreoli
zati
onoccurswhenspeaker sofacr eol
e(apidgent hathasbecomeanat i
velanguage
ofagr oup)"gainvar
yingdegr
eesofcont actwitht
hegr oupt hatspeakst hebasel anguageof
thecreole"(Schumann,p.41).I
ti sthepr ocessofmov ingt owardthe" st
andardform"oft he
l
anguage.Cr eoli
stsr
efertoseveralstagesofdecreoli
zation,rangingfrom thecreoleitsel
f,to
thebasolect,whichi
sclosetot hecreole,themesolect,theacrolect,andfinal
l
y,thest andard
for
m.

19 Psy chologi
caldistanceisdet erminedbyf actorssuchasmot ivation,languageand
cult
ureshock, andotheraffecti
vev ari
ables.Socialdist
ancer esult
sf r
om soci alfactors,suchas
therelativedomi nanceofthesoci algroupoft heacqui r
erandspeaker soft het ar
getl anguage,
thecohesi v enessoft hegr oups,similar
ityinculture,et
c.I nSchumann' sv iew factorscausi ng
psychol ogicalandsoci aldist
ance" putthelearnerinasi tuat
ionwher ehei sl argel
ycutof ff
rom
tar
getl anguagei nputand/ordoesnotat tendtoitwheni tisavail
able"(Schumann, 1977, pp.
266-267) .

20 Al
soofi nt
eresti
st hefactt
hatAlbert
o'sgrammati
calmor phemediff
icul
tyor
der(one
cr
oss-sect
ion)corr
elat
essigni
fi
cantl
ywit
ht he"nat
uralor
der
"proposedearl
i r=0.
er( 73,p<
0.
05;analy
sisinKrashen,
1977).Thedatawascol l
ectedf
rom hi
sspontaneousspeech.

21 Thisi
snott heonlyint
erpr
etat
ionoft hi
sresul
t,asEar
lStev
ickhaspoint
edouttome.
Somet
hingelsemayhav ecausedPaz 'superi
orsecondlanguageacqui
sit
ion,andthel
ow
psy
chol
ogicaldi
stancescoremaybear esultofthi
sandnotacause.

22 Ort
he"
remot
ecause"
.Seedi
scussi
oni
nSchumann(
1978b)
,p.48.

56
56
Chapt
erI
II

Pr
ovi
dingI
nputf
orAcqui
sit
ion
Inthischapt er,wet aket hediffi
cultstepfrom t heor
yt opractice.Bef orewedot his,letme
remindt her eaderoft hemainpoi ntofChapt erI:Iconsidertheor ytobeonl yoneofsev eral
possibledet erminantsofmet hodandmat erials.Theseimpl i
cationsneedt obeconf ir
medby
fur
therr esearch(event houghsev erali
mpli
cat ionsdohav eempi ricalconfir
mation)andbyt he
experiencesoft eacher sandst udents.The" i
deal "st
ateisar elationshipwherebyt heoretical
andappl iedresearchersandpr actiti
onersl
earn( andacquire)fr
om eachot her.

We wi l
lcov erone aspectofappli
cati
on inthischapter
:how we can encourage
subconsci
ousacqui si
ti
on.Thisisan import
antquestion,si
ncet hemajorimpli
cati
on of
secondlanguageacquisi
ti
ontheoryi
sthatacquisi
ti
oniscentr
al.I
tther
efor
efoll
owsthatour
majorpedagogical
effor
tsneedtobedevot
edtoencouragingl
anguageacquisi
ti
on.

Thi
spor t
ionbeginswi thabr i
efdiscussionofsomeoft hei mplicat i
onsoft hei nput
hypothesi
swi threspecttother ol
eandpot ent
ialofthesecondlanguagecl assroom, aswel las
i
tslimitat
ions,ascompar edwi t
htheinformalenv ir
onment.Followingt hi s,wedi scusst he
contr
ibuti
ont hatactualoutputcanmake.Asex plai
nedinChapterII,itishy pothesi
zedt hatwe
acqui
r evi
ainput ,whatwer eadandhear ,andnotv i
aout put
,actualtalki
ngandwr i
ti
ng.Out put
doeshav eani ndir
ectrol
etopl ayinencouragi
ngacqui si
ti
on,howev er.

Themaj orpor tionofthi schapterisconcernedwithcharact

eri
zingwhat" goodinput"i
s,
l
ist
ingt hefeatur est hatinputshoul dhav ei fi
tist oencourageacqui si
ti
on.Insubsequent
chapters,wewi lldiscusshowconsci ousl anguagelear
ningfi
tsintothepedagogi calschema,
andint hefi
nalchapt erwewi llexaminesomecommonl anguageteachingmet hodsandsome
aspectsofthei nformalenv ironment,toseet owhatex t
enttheyprovi
det heinputdiscussedi
n
thi
ssect i
onandt hety peoflearningdiscussedinChapterIV.

57
57
 Thegoalofthi
sexerci
seistoprov
ideaf rameworkthathel
psusseewhatmat eri
als
andmethodsactual
lydoforthesecondlanguagestudent
.Thiswil
lhopeful
l
yhelpuncover
gapsandprov
ideuswithwaystosuppl
ementandi mpr
oveexist
ingmater
ial
sandt
echni
ques.

A.ThePot
ent
ialoft
heSecondLanguageCl
assr
oom

Weof t
enheart haty ouhav et o" l
i
vei nthecount ry"inor dert oachi eveanyr eal
profi
ciencyinasecondl anguage,andt hatt hei nformalr ealwor l
denv i
ronmenti sal ways
super i
ortothecl assroom,orf ormalenv i
ronment .Aswesawi nChapt erI I
,therearesev eral
studiesthatappeart osuppor tt
hisasser ti
on.Ot herstudies,howev er,suggest edt hatthe
classroom doeshel paf terall
.Iattemptedt or esolvethisappar entconf l
ictbyhy pothesizi
ng
thatwhatwasr eall
yatissuewascompr ehensibleinput.Thecl assroom isofbenef itwheni ti
s
themaj orsourceofcompr ehensi
bleinput.Whenacqui rer
shav eri
chsour cesofi nputout si
de
theclass,andwhent heyar eprofi
cientenought otakeadv antageofi t(i
.e.understandatleast
someofi t
),t
hecl assr
oom doesnotmakeani mpor tantcontributi
on.

Thus,ther ealadvantageofthei nf
or malenv ir
onmenti st hatitsuppliescompr ehensible
i
nput.If
,howev er,wef il
loursecondl anguagecl assroomswi t
hi nputt hatisopt imalf or
acqui
sit
ion,i
tisqui t
epossiblethatwecanact ual l
ydobet terthant heinformal envi
ronment ,at
l
eastupt otheint er
mediatelevel
.Aswement ionedinChapt erII
,thei nf
ormalenv ir
onmenti s
notal
way swi l
li
ngt osupplycompr ehensibleinputt otheol dersecondl anguagest udent.As
Hatch and hercol l
eagues hav e pointed out ,inputt ot he adul tis mor e compl icated
grammat i
cal
ly,containsawi derrangeofv ocabulary,dealswi thmor ecompl extopics,andi s
gener
all
yhar dertounder stand.Thisissi mplyar efl
ecti
onoft hef actthattheadul twor ldis
morecompl exthant hewor ldofthechi l
d,andourex pectati
onsf oradul tcompr ehensionar e
muchhigher.

I
nt hecaseoftheadultbeginner,t
heclassroom candomuchbet t
erthantheinf
ormal
envir
onment .I
nthesecondlanguageclassroom,wehav ethepotenti
alofsuppl
yi
ngaf ul
l
40-50mi nutesperdayofcompr ehensibl
ei nput,i
nputthatwi l
lencouragelanguage
acquisi
ti
on.Thetruebeginneri
ntheinformalenvir
onment,especi
all
yifheorsheisnot

58
58
adeptatskil
lsofconv ersati
onalmanagementandnegot i
ati
onofmeaning( seediscussi
on
l
aterinthi
schapt er;alsoScar cell
aandHi ga,fort
hcomi ng)
,mayrequi
redaysorev enweeks
beforeheorshecan" pickout"thatmuchcompr ehensiblei
nputf
rom t
hebarrageoflanguage
heard.Thebeginningst udentwi l
lsimplynotunderstandmostoft hel
anguagearoundhi m.It
wil
lbenoise,unusablef oracquisi
tion.

Thev alueofsecondl anguageclasses,t

hen,l
iesnotonlyinthegr
ammari nst
ructi
on,
butinthesi mpler"
teachert al
k",t
hecompr ehensi
bleinput
.Itcanbeaneffi
cientplaceto
achi
eveatl easttheintermediatelev
elsr api
dly
,asl ongast hefocusoft
hecl assi son
prov
idi
nginputforacquisi
ti
on.

B.Li
mit
ati
onsoft
heCl
assr
oom

Despitemyent husiasm fort hesecondlanguagecl assroom,therearesev er

alway sin
whi
ch the out si
de wor ld cl
earlyex cel
s( orsome " modif
icati
on"oft he outside world,a
f
ascinat
ingalternati
v et hatweshal ldi
scusslater),especial
lyfortheint
ermediat
el evelsecond
l
anguagest udent.First,itisveryclearthattheout sidewor l
dcansuppl ymor einput.Livi
ngin
t
hecount r
ywher ethel anguagei sspokencanr esultinanal l
-daysecondlanguagelesson!As
wement ionedear lier,howev er,fort heinfor
malenv ir
onmentt obeofanyuse,t heinput
l
anguagehast obecompr ehensibl
e.Thei nfor
malenv i
ronmentwi l
ltheref
orebeofmor eand
moreuseast heacqui rerprogressesandcanunder standmor eandmor e.

Second,asmanyscholar
shavepoi nt
edout,therangeofdiscour
sethatt
hestudentcan
beexposedtoi nasecondlanguageclassroom isquit
el i
mited,nomatt
erhow "
natural
"we
makei t
.Thereissi
mplynowayt heclassroom canmat cht hevar
iet
yoftheoutsi
dewor ld,
al
thoughwecancertai
nlyex
pandbey ondourcurrentl
imitat
ions.

Theclassr oom wil

lprobablyneverbeablet ocompl et
elyover
comei t
slimit
ations,nor
doesithavet o.Itsgoalisnott osubst
itut
ef ort
heout sidewor l
d,buttobringstudentstot he
pointwheretheycanbegi ntouset heoutsi
dewor ldforf ur
theracqui
sit
ion,towher etheycan
begintounder standt hel
anguageusedont heoutside.,i
tdoest hi
sintwoway s:bysupply i
ng
i
nputso t hatst udentsprogressinlanguage acqui sit
ion,so thattheyunder stand "real"
l
anguaget oatl eastsome

59
59
extent
,andbymaki ngthestudentconver
sati
onal
l
ycompetent,thati
s,bygi v
ingthestudent
toolsto manage conver
sati
ons despi
teal ess t
han per
fectcompet ence i
nt he second
l
anguage.Wer et
urntobothoftheseimport
antpoi
ntsi
nthediscussi
onthatfoll
ows.

C.TheRol
eofOut
put

Asecondpoi ntt hatneedst

obedealtwit
hbefor
edescr
ibi
ngthechar
act
eri
sti
csof
opt
imalinputf
oracqui siti
onistherol
eofout
put,mostcommonl
y,t
heroleofspeech,i
n
l
anguageacqui
sit
ion.1*

TheInputHy pothesismakesacl aimt hatmayseem qui teremar kablet osomepeopl e-

-
weacqui respokenf luencynotbypr act ici
ngt alki
ngbutbyunder standingi nput,byl i
stening
andr eadi
ng.Itis,i
nf act,t
heor eticall
ypossi blet oacquir elanguagewi thoutev ertalki
ng.Thi s
hasbeendemonst ratedforf i
rstl anguageacqui sit
ionbyLenneber g( 1962) ,whodescr i
bedt he
caseofaboywi t
hcongeni taldy sarthria, adi sorderoft heper ipheralspeechor gans, whowas
neverabl etospeak.WhenLenneber gt estedt heboy ,hef oundt hatt hechi l
dwasabl et o
understandspokenEngl i
shper fectl
y .Inot herwor ds,hehadacqui red" compet ence"wi thout
everproducing.Thechi l
dwast estedatageei ght ,andt hereisnowayt otelldir
ectlywhet her
hislackofoutputhadsl oweddownhi sl anguageacqui sition.Itisqui t
epossi blethatifhehad
beenabl etospeak,hewoul dhav eacqui redl anguagesomewhatf aster,duet ot heindi r
ect
contri
butionspeakingcanmaket oacqui siti
on.

Outputhasacontri
buti
ontomaket olanguageacquisi
ti
on,butitisnotadi
rectone:
Si
mply,t
hemor eyoutal
k,themorepeoplewil
ltalktoyou!Act
ualspeaki
ngonthepartoft
he
l
anguageacquir
erwi
ll
thusaff hequant
ectt i
tyofinputpeopl
edir
ectatyou.

I
twi l
lalso affectthe quali
tyoft hei nputdirected att heacquirerConv ersati
onal
partnersof t
entrytohel py ouunderstandbymodi f
yingt heirspeech("f
or ei
gnertalk")
.They
j
udgehow mucht omodi f
ybyseei ngwhet heryouunder standwhatissai d,andal so by
l
isteningt oyout al
k.Asecondl anguagespeakerwhomakesl otsofmi stakes,hasapoor
accent ,andishesit
ant,wil
lmostl i
kelyrecei
ve,ingeneral,
mor emodifi
edinputt hanaspeaker
whoappear scompet entandfluent
.

Engagi
ngi
nconv
ersat
ioni
spr
obabl
ymuchmor
eef
fect
ivet
han

* Super
scr
iptnumber
sref
ert
oNot
esatendofchapt
ers.

60
60
"eavesdropping"forlanguageacqui si
ti
on.Inconversat
ion,thesecondl anguageacquirerhas
somedegr eeofcont rolofthetopic,cansignalt
ot hepartnerthatthereisacompr ehension
problem,et c.Inot herwor ds,hecanmanageandr egulatetheinput ,andmakei tmor e
compr ehensible.Thereisnosuchcont r
olineav esdr
opping!Buti nor dertoparti
cipatein
conversati
on,t her
emustbeatl eastsomet al
k,someout put,from eachpar t
ner
.Hence,t he
i
ndirectcont r
ibuti
onofspeech.

1."
CONVERSATI
ON"ANDLANGUAGEACQUI
SITI
ON

Somescholarshav esuggestedthat"part
icipat
ioni nconver
sati
on"isr esponsibl
ef or
l
anguageacqui si
ti
on.Inthelightoftheabov ediscussion,wecanseet hatthisistrue,i
na
sense."Conver
sation"
,howev er,i
snoti nitsel
ft hecausat i
vevari
ableinsecondl anguage
acquisi
ti
on.I
tisoneway ,andav erygoodway ,t
oobt ai
ni nput
.Iti
stheor
etical
l
yqui tepossible
toacquirewi
thoutpart
ici
pati
ngi nconver
sat
ion,howev er.
2

Fi
gure3.1i
ll
ustrat
est
hei
ndi
rect
,butof
tenconsi
der
abl
e,cont
ri
but
ionout
putcanmake
t
olanguageacqui
sit
ion.

Fi
g.3.
1.Howout
putcont
ri
but
est
olanguageacqui
sit
ioni
ndi
rect
ly.

Compr ehensi ei
bl nputi
sresponsi
bleforprogr
essinlanguageacqui
sit
ion.
Outputispossibl
easar esul
tofacquir
edcompet ence.
Whenper formersspeak,t
heyencourageinput(peopl
espeaktothem).Thi
sis
conversati
on.

2.OUTPUTANDLEARNI
NG

Assuggest edinChapt erII,outputcanpl ayaf air

lydi
rectr ol
einhel pi
nglanguage
l
earni
ng,alt
houghev enher eitisnotnecessar y.Outputai
dslearningbecausei tpr
ovidesa
domainforer r
orcorr
ecti
on.Whenasecondl anguageuserspeaksorwr it
es,heorshemay
makeaner ror.Whenthiser r
oriscor r
ected,thi
ssupposedlyhelpst hel
earnerchangehisor
herconsci
ousment alrepresentat
ionoft heruleoraltert
heenvironmentofr uleappl
icati
on.
(Seedi
scussionofHypothesis(1),ChapterII
.)

Wemayt
huscompar
ean"
out
putappr
oach"t
othei
nputappr
oach

61
61
promotedher
e.Couldwet eachl anguageprimari
lybyencour
agi
ngpr
oduct
ion,
wit
hli
ttl
eorno
i
nput,andcorrecti
ngallerrors?Suchat echni
que,i
nadditi
ontobei
ngmaddening,rel
i
es
enti
rel
yonthestudent
s'abil
it
yt olearngrammar .

Thisi
snottosaythaterrorcorr
ect
ionistot
all
yusel
essandt
hatl
earni
ngisofnoval
ue.
Learni
nghasaroletoplay,ander r
orcor
recti
onmaybeofusei ncert
ainsi
tuat
ions.Wewil
l
ret
urnlat
ert
othequest
ionof" put
ti
nglear
ninginit
splace"
.

Bynowt hereaderhasseenenoughpromisesof"moreonthist
opiclater
",anditi
st i
me
toturntothemajorporti
onofthischapt
er,adescr
ipt
ionofthechar
act
erist
icsofoptimalinput
forsecondlanguageacquisi
ti
on,wherehopeful
lyt
hepromisesthi
sint
roducti
onhasmadewi l
l
bekept.

D. Char act
eri
sti
cs of Opt
imal i
nput f
or
Acqui
sit
ion

Iwillatt
emptinthi
ssect iontopresentasetofr equirementsthatshouldbemetbyany
acti
vityorsetofmat er
ial
sai med atsubconsci ousl anguageacqui si
ti
on.The( test
able)
predi
ctiont hatt
hissetofcharacteri
sti
csmakesi st hatanact i
vi
tythatfit
sthecharacter
ist
ics
ful
lywillencourageacquisi
ti
onatt hef ast
estpossibler at
e.Anact i
vi
tythatfit
snoneoft hem
couldresultinzeroacquisi
ti
on,orv eryli
tt
leacquisit
ion.(Thelatt
er,"
veryli
ttl
e",i
smor el
ikely
.
The" l
anguageacqui si
ti
ondev ice"maybesopower f
ul,evenintheadul t
,thatsomemi nimal
acquisi
tionmayoccurasar esultofanyex posur
et olanguage.)

Thechar acterist
icsdescr i
bedbel ow arenot" weighted"
.Thereisnoat t
emptt ocl ai
m
thatonei smor eimpor tantthananot her,al
thoughsuchcl aimsshouldpossi blybemade.Iwi l
l
l
eav ethist ofut
urer efi
nement s.Also,thereisnoattempther eto"suppor t
"theseconcl usions
byempi ricalevi
dence.Theyder ivef r
om secondl anguageacqui si
ti
ont heory,thehypot heses
presentedi nChapt erII.Iti
st hesehy pothesesthataresuppor t
edbyempi r
icalevi
dence.I n
otherwor ds,wear el ookinghereonl yati mpli
cati
onsoft heory.Thi
sdoesnotmeant hatt he
characteristi
cscannotbet reatedaspr edict
ionsandf urthertest
ed;indeed,t heyshoul dbe
confir
medbybot happl iedli
nguisticsresearchaswel last eacherandst udenti nt
uit
ion,asI
discussedi nChapter1.

Wedi
scusseachchar
act
eri
sti
csepar
atel
y,showi
ngwhatpr
edi
cti
ons

62
62
each charact
eri
sti
c makes withrespectto diff
erentaspects ofmethod,mater
ial
s,and
i
nfor
mali nput
.Foll
owingthesedescr
ipt
ions,
Iwillatt
empttopointoutsomenewpossibi
li
ti
es,
andunderscoretheimpor
tanceofsomeneglectedtradi
ti
onalsourcesofi
nput(
Chapt
erV).

1.OPTI
MALI
NPUTI
SCOMPREHENSI
BLE

Thisisclear
lythemostimportantinputcharact
erist
ic.I
tamount
stothecl ai
mthat
whentheacquirerdoesnotunder
standt hemessage,t herewil
lbenoacqui
sit
ion.Inot
her
words,
incomprehensi
blei
nput
,or"
noise",
wi l
lnothel
p.

Posit
ing compr ehensi
bil
i
ty as a f undamentaland necessary( butnotsuf fici
ent)
requir
ementmakes sev er
alpredicti
ons thatappearto be correct.Itex pl
ains whyi tis
practi
call
yimpossi bleforsomeonet o acqui
reasecond orf or ei
gn languagemer elyby
l
isteni
ng tot he radi
o,unless the acquirerspeaks a v
ery closelyrelated language.A
monol i
ngualEnglishspeaker,f
orex ample,heari
ngPoli
shonther adio,wouldacqui r
enot hi
ng
becausethei nputwouldbeonly"noise".
3

Thisr equir
emental soex plai
nst heappar entfail
ur eofeducat i
onalTV programst o
teachf orei
gnl anguages.Thei nputi
ssi mplynotcompr ehensibl
e.Myownchi l
drenwat ched
programssuchasVi ll
eAllegref ai
thf
ull
yf oryears,andacqui redaboutasmuchasIdi d:They
couldcountf r
om onet oteni nSpanishandr ecogni
zeaf ewwor dssuchascasaandmesa.
Thecompr ehensibili
tyrequirementpr edict
st hatTVwoul d,ingeneral
,besomewhatmor e
successfult hanr adio asal anguaget eacher ,butthatev enTV woul dbeinadequat ein
beginningstages.Er vin-
Tripp( 1973)hasnot edt hathearingchildr
enofdeafpar entsdonot
acquirelanguagef rom TVorr adio,anobser v
at i
onconsistentwiththerequi
rement.
4

This char acteri

stic also ex pl
ains why chil
dren sometimes fai
lto pi
ck up fami ly
l
anguages.Myowncasei s,Ithink,quitety
pical
.Mypar entsspokeYiddi
sharoundthehouse
foryears,occasionallyt oeachot her( totel
lsecret
s),andconstantlytomygr andparents.
Nevert
heless,mysi sterandIf ail
edt oacquireYiddi
sh,withtheexcepti
onofaf ew phrases
androutines.Ont heot herhand,i nmanyf amil
ieschil
drendogr ow upspeakingthefami ly
l
anguageaswel last helanguageoft hecommuni ty.Whatappearstobecruci
aliswhetherthe
famil
ylanguage

63
63
i
sdi r
ectedatt hechild,i
not herwords,whetheranatt
emptismadet omaket helanguage
comprehensibl
e.Whatwehear dv iaeavesdr
oppingwasnotcomprehensible.Itdeal
twi t
h
topi
csthatwer enoteasil
yidenti
fiedandthatwereal
sooft
enbeyondourrangeofex per
ience.
Languagedirectedatusi nYiddishwouldhavebeensimpli
fi
ed,andmorerelevanttous,and
hencemor ecompr ehensi
ble.

Anot herpredict
iont hatthecompr ehensibil
it
yr equi
rementmakesi sthat"j
ustt al
king",
or" f
ree conv ersat
ion",is notl anguage teaching.I n ot
herwords,si mplybeing a nat i
ve
speakerofal anguage doesnoti n ofitselfquali
fy one asa t eacherofthatl anguage.
Consciousandex tensiv
eknowl edgeofgr ammardoesnotmakeoneal anguaget eacher
eit
her.Rat her,thedefi
ningchar acter
isti
cofagoodt eacherissomeonewhocanmakei nput
compr ehensibletoanon- nati
vespeaker ,r
egar dl
essofhi sorherlevelofcompet enceint he
tar
get l anguage. Thi sl eads nat ural
lyt o anot her topi
c,how t eacher
s make i nput
compr ehensible.

(
a)Howt
oai
dcompr
ehensi
on

I
fwear ecorr
ectinpositi
ngcompr ehensi
bil
it
yasacr uci
alr
equir
ementf oropt
imalinput
foracquisi
ti
on,thequest i
onofhow t oai dcompr ehensi
oni saverycentraloneforsecond
l
anguagepedagogy .Indeed,thecompr ehensionrequi
rementsuggeststhatthemainfunction
ofthesecondlanguaget eacheristohelpmakei nputcomprehensi
ble,
todof ortheadul
twhat
the"out
sideworld"cannotorwi l
lnotdo.

Therearebasi call
ytwoway sinwhicht heteachercanaidcompr ehension,li
nguisti
c
andnon-li
ngui
sti
c.Studieshaveshownt hatther
ear emanythingsspeakersdolinguisti
call
yto
maket heirspeechmor ecompr ehensi
bletol esscompet entspeaker
s.Hat ch( 1979)has
summar i
zedthelinguist
icaspectofsimpl
ifi
edinputwhichappeartopromotecompr ehensi
on.
Amongt hesecharacteri
sti
csare:

(1) slowerr
ateandclear
erart
icul
ati
on,
whi chhel
psacqui
rer
stoi
dent
if
ywor
d
boundar
iesmoreeasi
l
y,andall
owsmor eprocessi
ngti
me;

(
2) mor
euseofhi
ghf
requencyv
ocabul
ary
,lesssl
ang,
feweri
dioms;

(
3) sy
ntact
icsi
mpl
i
ficat
ion,
shor
tersent
ences.

64
64
 Suchchar acteri
sti
csandot hersappeartobemor eorl esscommont odiff
erentt ypes
ofsimplecodes, suchascar et
akerspeech, f
orei
gner-talk,
andt eacher-
tal
k(seealsoKrashen,
1980),andcl earlyhelpmakei nputlanguagemor ecompr ehensibl
e.Thereisconsi der abl
e
empiri
calev i
dencet hatthesecodesar esigni
ficantly"simpler"thannativespeaker -nativ
e
speakerlanguage,and,asment ionedinChapt erII,therei sevidenceofsomecor relati
on
betweent helinguisti
cleveloftheacqui r
erandt hecompl exityoftheinputlanguage:mor e
advancedacqui rer
st endtogetmor ecomplexinput.

Doest hismeant hatteachersshoul dconsciousl

ytr ytosimpli
fytheirspeechwhent hey
tal
ktost udents?Shoul dt heyt hi
nkaboutsl owingdown,usi ngmor ecommonv ocabulary,
usi
ng shor tersent ences,l ess compl ex syntax withl ess embeddi ng,etc.? Consciousl
y
ref
err
ingt ot hese" r
ules"mi ghtbehel pfulonoccasion,buti tappearstobet hecaset hatwe
make t hese adj ustment s automat i
cally when we f ocus on t ryi
ng to make our sel
ves
under
st ood.RogerBr own,comment ingonst udiesofcar et
akerspeechi nf i
rstlanguage
acqui
sition,comest oasi mil
arconclusion.Hegi vesthefollowi
ngadv i
cet oparentswant i
ngto
knowhowt o" teach"theirchil
drenlanguagei ntheleastamountoft ime:

Bel
iev
et hatyourchildcanunder
standmor et
hanheorshecansay ,andseek,aboveall
,to
communi cat
e...
.Ther
eisnosetofrulesofhowtot
alkt
oachildthatcanev enapproachwhat
youunconsciouslyknow.Ifyouconcentr
ateoncommunicat
ing,everyt
hingelsewillf
oll
ow.
(Br
own,1977, p.26.
)

AsIhavearguedinseveralplaces(
Krashen,1980,1981)
,thesamesi
tuati
onmayhol
dfort
he
l
anguageteacher.I
fwef ocusoncompr ehensi
onandcommuni cati
on,wewil
lmeetthe
sy
ntact
icr
equirementsforopt i
mali
nput.

Whi
lewef r
eeteachersoftheresponsibi
li
tytoconsciouslycontrolthegrammaroft heir
speech,ot
herr esponsi
bil
it
iesbecomemor eimportant.Onei st omakesur ethatt
hei nputis
i
ndeedcompr ehensi
ble.Ihavenot hi
ngst ar
tli
ngtoaddt ot hel i
ter
atureoncompr ehension
checki
ng,otherthant ounderscoreandemphasi zeitsimport ance.Compr ehensi
onchecki ng
can range f rom si mply asking "Do y ou understand?" occasi onall
y,t o moni tori
ng
comprehensionv i
astudents'
verbalandnon-verbal
responses.

65
65
 Anot hermai nt askoft heteacheri st opr ovi
denon- li
nguist
icmeansofencour aging
comprehension.I n myv i
ew,pr ovidi
ng ex tr
a-li
nguisti
csuppor ti nthef orm ofreali
a and
pi
ctur
esf orbegi nningclassesi snotaf r
il
l,butav er
yimpor tantpartofthetoolst
het eacher
hastoencour agel anguageacqui sit
ion.Theuseofobj ectsandpi cturesinearl
ysecond
l
anguagei nstr
uct i
oncor respondstot hecar etaker'
suseoft he"hereandnow"i nencouraging
fi
rstl
anguageacqui si
tion,inthattheyal lhelpt heacquirerunder st
andmessagescont ai
ning
str
uct
urest hatare" ali
ttl
ebey ond"them.

Goodt eachersalsot akeadv antageoft hestudent'

sknowl edgeoft heworldinhel ping
compr ehensi
onbydi scussi
ngt opicst hatarefamili
artothest udent.Cert
ainly
,discussingor
readi
ngaboutat opicthatist otall
yunknownwi l
lmaket hemessagehar dertounder stand.
Therei sadanger ,howev er,i
nmaki ngt heinputtoo"famili
ar".I
fthemessagei scompl etely
known,i twil
lbeofnoi nt
erest,andt hest udentwil
lprobablynotattend.Wewantt hestudent
tofocusont hemessage, andt heremustbesomemessage, somethingthatthestudentr eall
y
wantst ohearorr eadabout .Thisr equirementisperhapst hehardestonet omeet ,andwe
shallhavemor etosayabouti tbelow, inourdiscussionofcharacteri
sti
cII.
5

As pointed outj usta momentago,compr ehension isa necessar y conditi

on for
l
anguageacqui si
ti
on,buti tisnotsuf
fici
ent.Itisquit
epossi blet ounderstandi nputlanguage,
andy etnotacqui r
e.Thi scanhappeni nsev eraldif
ferentway s:Fi r
st,i
tisqui t
epossi blethat
theinputsi mplydoesnotcont ni+1,t
ai hatitdoesnoti ncludest r
ucturest hatare" abit
beyond"thest udent.Second, i
nmanycaseswedonotut il
izesy ntaxinunder st
anding--wecan
oft
engett hemessagewi thacombi nati
onofv ocabulary,orl exi
calinformation,plusex t
ra-
l
inguist
icinformation.Finall
y,t
he"aff
ecti
vef i
lter
"maybe" up" ,whichcanr esultintheacquirer
understandinginput,eveninputwithi+1,butnotut i
li
zingitforf urt
heracquisiti
on.

2.OPTI
MALI
NPUTI
SINTERESTI
NGAND/
ORRELEVANT

Opti
malinputfocussestheacquireront hemessageandnotonform.Togoast ep
f
urt
her,t
hebestinputi
ssoi nt
erest
ingandrelev
antthatt
heacqui
rermayev
en"f
orget
"thatt
he
messageisencodedinaforei
gnlanguage.

66
66
 Creati
ngmat er
ialsandpr ov
idinginputthatmeett hischaract
eristi
cmayappeart obe
aneasyandobv ioust ask,butmyv iewist hat
,inreal
i
ty,thi
sr equi
rementi snoteasytomeet ,
norhast heprofessionconsi der
editobvious.Itisv ydi
er ff
icultt
opresentanddi scusstopi
cs
ofinteresttoacl assofpeopl ewhosegoal s,i
nter
ests,andbackgr oundsdi f
ferfr
om the
teacher'
sandf rom eachot her'
s.Ial so clai
m thatrelevanceandi nteresthavenotbeen
widelyper cei
ved as r equi
rementsf ori nput
,since so manymat eri
alsf ai
lto meett hi
s
requir
ement .

Iti
sf ai
rl
yeasyt othinkupex ampl esofi nputt hat ,whilecomprehensible,ar
euni versall
y
perceiv
edt obeuni nteresti
ngandi rrelev ant.Amongt hemostobv i
ousex ampl esarepat t
ern
dri
ll
,andmostdi al
oguet y
peex er
cises.Ex per i
ment alev idencesuggest sthatstudent spay
l
itt
leornoat t
ent i
ontomeani ngaf t
ert hef i
rstf ewr epet i
ti
onsi npat
terndril
l(Lee,McCune, and
Patton,1970),andt hesamer esultismostl ikelytruef ordialoguesthatarememor i
zedbyr ote.
Grammat i
calex erci
sesal sofailasinputf oracqui siti
ononsi mil
argrounds.Gr ant
ed,thegoal s
oftheseex ercisesar enot" acquisi
tion" ,andwewi llhav eoccasi ontoex aminewhet herthese
i
nput -
typesf i
llot herneedsi nt hesecondl anguagepr ogram.Nev ert
heless,theyf ailt hi
s
requir
ementdi smal ly
.

Somewhatlessobv i
ousi sthefailureof" meaningfuldril
l"toqual
ifyasoptimalinputfor
acquisi
ti
on."Meaningfuldril
l"isdi sti
nguished f r
om " mechani caldri
ll
",inthatt hef or
mer
requi
resthatrealmeaningbei nvolved( Paul
ston,1972) .Sincemeani ngfuldri
l
lisdesignedt o
provi
depracti
ceonpar ti
culargrammat icalstructures,howev er,i
tisverydif
fi
culttoalsobuild
i
nt heexchangeoftrul
yrelevantorinteresti
ngi nfor
mat i
on,asi n:

Whatti
medoeshegetupinthemor
ning?Whatt
ime
dotheygetupi
nthemor
ning?

Atbest,suchinf
ormati
oni sofonl ymildinter
estt
omember sofalanguageclass.Ibel
i
eve
thati
tisanimpossi
bletaskforteacherstoembedtrul
yint
eresti
ngorrel
evanti
nformati
onint
o
thefor
m ofameaningfuldri
llonadailybasis.

Someot herfair
lywidespreadi nputt y
pest hatf
allshortofthemar koft
ruerel
evance
aret he reading assignmentst hatmostf orei
gn language student
s workt hr
ough in
i
ntroductorycourses.General
ly,thesesel ecti
onsbearv er
yli
ttl
eresemblancet
ot heki
ndof
readi
ngt hestudentswoulddoint heirfi
rstlanguageont hei
rownt i
me.

67
67
 Beforet her eaderfeelsIam bei ngov er l
ycr i
ticalandunf ai
r,Imustsayt hati
tseemst o
met hatt hei nterest-r
elevancer equirementi snear l
yi mpossibl
et osat i
sfyinthest andard
Amer icanf or eignlanguagecour segi veni nthehi ghschoolorcol l
ege, especial
lywhensucha
coursei st akenasar equirement.Itisf areasiert osat isfyi
nESLsi t
uat i
ons,wherethereisa
perceivedneedf orthel anguage.Forex ampl e,incl assescomposedpr imari
lyofimmi grants,
i
nputwi l
lty picallycont ainagreatdealofi nformat i
ont hatisusefultot hestudentforsurviv
al
"on theout side".Uni versit
yESL cour sesf orf oreign studentsof ten incl
udest udyski l
ls,
Englishf oracademi cpur poses,i ntr
oduct i
ont ouni v
er si
tyli
fe,andev enusef ulacademi c
skil
ls.6E SLt eacher sof t
enser veof ficial
lyorunof fi
ciall
yasf riendsand counsel ors,and
thereforepr ov ideagr eatdeal oftrulyrelevantinput.78

3.OPI
MALI
NPUTI
SNOTGRAMMATI
CALLYSEQUENCED

In acqui
sit
ion-or
ient
ed materi
als,we shoul
d notbe consciousl
yconcerned about
i
ncl ngi+ 1i
udi nt heinput. Par
t(3)ofthe InputHy pot
hesi
sclai
mst hatwhen inputis
comprehensibl
e,whenmeani ngissuccessf
ull
ynegoti
ated,i+1wil
lbepresentaut
omatical
l
y,
i
nmostcases. 9

Thi
sr equir
ementcouldbest at
edinaweakerform.( 3)couldber ephr
asedasfol
l
ows:
thereisnoneedt odeli
ber
atel
yincludei+1,sincei
twilloccurnatural
l
y.Thest r
ongf
orm may
becal l
edfori nstead:i
tmaybebet t
ernott
oev enattemptt oi udei+1Thear
ncl guments
againstadel i
berateatt
empttogr ammat i
cal
lysequencewer egivenbriefl
yinChapt
erII
,and
will
beex pandedonher e.

1.I
fwesequence,andeachl esson,orgroupoflessons, f
ocusesononest r
uct ur
e,t
his
assumest hatev eryonei nt heclasshast hesame i+ 1,t hatev eryoneisatt hesame
devel
opment alstageint hesecondl anguage.Becauset hereareindivi
dualdi
ffer
encesi nthe
rat
e of acqui si
tion (due tot he st r
ength of the af f
ecti
vef il
terand t he amountof
comprehensibleinputobtained)
,iti
sex tr
emel yunl
i
kelythatallt
hest udentsi
nanycl assareat
thesamest age.Unsequenced butnat uralinput,itis hypothesized,wil
lcont ainar ich
vari
etyofstructure-
-i
fitiscomprehensible,t
herewilbei
l +1f oreveryoneas

68
68
l
ongast
her
eisenoughi
nput(
wer
etur
nt hequant
ot it
yquest
ionbel
ow)
.

2.Whenweat t
emptt opr esenta" f
inely-tuned"sequence,wegener allypr esenteach
structureorr uleonce.( Therei st he" revi
ew"l essonandt her ear eattempt satr ecy cli
ng,but
rev i
ewdoesnotusual lyworkt hrought heent iresequenceofact ivi
ties--
itsgoali sgener all
yto
"remi nd"and pr ov i
de some addi t
ionalpr actice fora r ulet hati s supposedl y al r
eady
"i
nt ernali
zed".
10)Wh athappenst ot hest udentwhomi ssest her ulethef i
rstt i
mear ound?
Tr aditi
onalreview,meantasar emi nder,wi l
lof tennothel p.Int raditi
onalf oreignl anguage
l
ear ning,asdonei nt heUni t
edSt at es,thest udentmayev enhav etowai tuntilnex ty ear,when
ther uleispr esent edagai n.Unsequencedcommuni cat
ivei nputcont ai
nsbui lt
-inr eview.We
don' thavet owor ryifwemi sst hepr ogressiv etenset oday,i twillbepar toft heinputagai n.
..
andagai n.Compr ehensibleinputt husguar ant eesusnat uralreviewandr ecycling,assumi ng,
asment ionedabov e,thatthereisenoughofi t.

Somer eadersmayf eelt

hatIam sett
ingupandat tackingast r
awman.I tcanbeargued
t
hatsomegr ammat i
cal
ly
-basedcourses,despitealockstepst r
ucturalori
entati
on,doprovide
nputati+1aswel
i l.Whi
letheremaybea" st
ructur
eoft heday ",noteveryutter
ancecontains
t
hetargetstr
ucture.Forexample,i
fthelesson'sfocusist heprogr essi
vetensemar ker
,other
t
enseswillbeusedaswel linbothcl
assroom inputandinther eadings.

Thi
smayappeart obet hecase,butt hereis,never
thel
ess,ar ealpr
oblem wit
hthi
s
approach.Wit
hagr ammat i
calfocus,communi cati
onwi lal
l wayssuffer,ther
ewi l
lal
waysbe
l
essgenui nel
yinterest
ing input.Thet eacher'
smi nd,and themat erial
swr i
ter
'smind,i
s
focusedon"cont
extuali
zing"aparti
cularstr
ucture,
andnotoncommuni cati
ngideas.

Asmycol l
eagueSt evenSt er
nfeldhaspoi ntedoutt ome,whati sproposedher eis
f
undamental
ly diff
erentf rom "contextual
izati
on".Cont ext
uali
zati
on invol
ves invent
ing a
r
eali
sti
ccont
extf orthepresentat
ionofagr ammat icalruleorvocabul
aryit
em.Thegoali nthe
mindofthet
eacheri sthelearni
ngoracqui sit
ionoft her ul
eorwor d.Whatisproposedhereis
t
hatthegoal
,i
nt hemi ndofbot htheteacherandt hest udent,i
stheidea,t
hemessage.

69
69
 Thi
sobj
ect
ioncanbesummar
izedasf
oll
ows:

3.Thev er
yor i
entati
onoft hegrammatically
-basedsy l
labusreducesthequal
ityof
comprehensi
blei
nputanddistor
tst hecommunicat i
vefocus.Teacherswil
lbeconcer
nedwith
how t
heyarespeaki
ng,readi
ngsel ect
ionswil
lbeai medati ncl
udingxnumberofexamplesof
str
uct
ureyalongwi t
hacer tai
nv ocabular
ysampl e,asur eguaranteeofbori
ngandwooden
l
anguage.

4.Stil
lanotherproblem i
sthatthegr ammat icalsequenceat tempt stoguesst heorder
ofacquisit
ion.Severalyearsago,Isuggested( Krashenetal .
,1975)thatanappl i
cationoft he
NaturalOrderHy pothesi
swast heconstructionof" nat uralsy
ll
abi"f
ollowingthenatur alor
der.
Myposi ti
onhaschanged.AsFat hman( 1979)haspoi ntedout,thepracti
cali
mpl i
cationoft he
NaturalOrderHy pothesismayl i
einwhati thast aughtusaboutt heunder ly
ingpr ocessof
l
anguageacqui siti
on.Itseemstomenowt hatweshoul dnotatt
emptt oteachaccor dingtoan
ordersimilartot hatgiveninTabl e2,Chapt erII( oraccor di
ngt oanyot hergr ammat ical
sequence).

Comprehensi
blei
nput
,iti
scl
aimed,wi
l
lautomati
cal
l
yfol
l
owanat
ural
orderi arasi
nsof
+1wi
l
l bepr
ovided(
alongwi
thmanyotherst
ruct
ures)
.

We now summar
ize t
he adv
antages ofnaturalinput(the well
-bal
anced di
etof
wholesomef ood)ov
ert
helockst
epgrammaticalsy
llabus(singl
eorev enmul t
ipl
ev i
tami
n
t
herapy).
11

(
a)Thecaseagai
nstt
hegr
ammat
icalsy
ll
abus
Grammat i
calsy
ll
abus(deli
berat
e Communi
cat
iv nput(i
ei +1i
ncl
udedat
temptt yi
osuppl +1)
natur
all
y,gi
venenoughinput)

1.Allst
udentsmaynotbeatthe 1.i
+1wi
l
lbepr
ovi
dedf
oral
lsamest
age.Thest
ruct
ureof
the studentsevent
ual
ly
daymaynotbet hei+1formany
ofthestudents.

2.Eachstr
uct
urepr
esent
edonl
y 2.Nat
ural
andext
ensi
ver
evi
ew.once.(
Seet
extf
or
di
scussi
onof
"r
eview"
.)

3.Grammat
ical
f ocusmaypr
ev ent 3.Consciousfocusofbot
hstudent
real
andnat
uralcommunicat
ion. andt eacheriscommunicat
ionofi
deas.

4.Assumesweknowor
derof 4.Doesnotassumeweknowor
derofacqui
sit
ion.
acqui
sit
ion.

70
70
4.OPTI
MALI
NPUTMUSTBEI
NSUFFI
CIENTQUANTI
TY

Itisdiffi
cul
tt osayj usthow muchcompr ehensi
ble/l
ow f i
lt
erinputisnecessaryto
achi
eveagi v
enl evelofprofi
ciencyi
nsecondlanguageacqui si
tion,duetoal ackofdata.We
knowenoughnow, however,tobeabletostat
ewi t
hsomeconf idencethattheprofessi
onhas
seri
ousl
yunder esti
matedt heamountofcompr ehensibl
ei nputnecessaryt oachieveeven
moderate,
or"inter
mediate"level
sofprof
ici
encyinsecondlanguageacqui si
tion.

Theor
eti
calar
gumentsforquanti
tyderi
vefrom t
hei
mmedi
atel
yprecedi
ngdi scussi
on.I
hypot
hesi
zedthatnatur
alcommuni cat
iveinputcoul y i+1f
dsuppl orallstudentsiftwo
condi
ti
onsweremet;

(
1) Thei
nputwasnotar
ti
fi
cial
l
yconst
rai
ned(
li
mit
edr
angeofdi
scour
se
t
ypes)

(
2) I
twassuppl
i
edi
nsuf
fi
cientquant
it
y.

Clearl
y,fi
vemi nutesoft
alk,orasi ngl
eparagraphofreading,hasli
tt
lechanceofi ncludi
nga
givenstudentsi+1.Rat
' herthantakeamor ecar ef
ulai
m atthatstudent
'sneeds,ratherthan
"overi
ndivi
dual
izing"i
nst
ructi
on,itisfareasier,Iam suggesti
ng,toincr
easetheamountof
compr ehensi
bleinput
.Again,i
fthereisenough,i+1wi l
lbeprovi
ded,andwil
lbepr ovidedover
andov er.

Asment ionedabov e,wedonothaveenoughdat atostate,

withconfi
dence,howmuch
i
nputisnecessar yt
or eachagivenst
age.Theli
terat
uredoespr ovi
deuswi t
henought ost
ate
somei ni
ti
alhypothesis,howev
er.Bel
ow,webr i
efl
yex ami
newhatt helit
erat
ureimpli
esabout
reachi
ngtheini
tial"
readinesst
ospeak"st
age,andmor eadvancedlevel
s.

(
a)Quant
it
yrequi
rement
sfori
nit
ialr
eadi
nesst
ospeak

Howmuchi nputi
sneededt
oendthe"si
lentperi
od"
?Howmuchi nputi
snecessar
yso
t
hatsecondl
anguageacqui
rer
scanpr
oduceutt
erancesusi
ngacqui
redcompet
ence?

Asher'
sworkonTot
alPhysi
calResponseteachi
ng,amethodthatrequi
resstudent
sto
obeycommandsgi v
eninthesecondlanguage,oft
enwi t
ha" t
otalphysi
calresponse"(
e.g.
standi
ngup),gi
vesussomei deaastohow muchi nputisnecessar
yf ori
niti
alspeaki
ng
readi
ness.

71
Aswewi llseeinChapterV,t hechiefvi
rtueofTotalPhy si
calResponsemaybei tsabil
i
tyto
suppl
yconcentrat
edcompr ehensibl
einput.Asherhasnot edinsev er
alpapers(rev
iewedin
ChapterV)thatTPRst udentsaregener al
lyreadytostartproducti
oninthetargetl
anguage
aft
erabouttenhoursofTotalPhy si
calResponseinput
.12

Informall anguageacqui siti

on research pr esentswhatatf i
rstmayseem t o bea
diff
erentpi ct
ure.The" si
lentper iod"seeni ni nformalchi ldsecondl anguageacqui siti
onmay
l
astasl ongassi xmont hs!Dur ingthist i
me,t hechi l
dmaypr oducev eryl
itt
lei nthesecond
l
anguage,ot hert hanr outi
nesandaf ew pat terns.Thegr eaterlengt hoft he" nat
ur al
"silent
period,ascompar edtoAsher 'sobser vati
ont hatt enhour smaysuf ficemaybeduet ot hefact
thatagr eatdealoft hei nputt hatt hechi l
di nt henat uralenv i
ronmentr eceivesmaybe
i
ncompr ehensible.Asst atedear lieri
nt hi
schapt er,themai nadv antageof" formalinst ruct
ion"
maybei t spotentialforprovidingcompr ehensi bleinputinear lystages,br i
ngingt heacqui rer
tothepoi ntwher eheorshecanbegi ntotakeadv ant ageoft henatur alenvir
onment .Thelong
sil
entper iodininf ormalchi l
dsecondl anguageacqui si
ti
onmaybef urtherevidencet hatt he
i
nfor mal envi
ronmenti sineff
icientinear l
ystages. 13

(
b)Quant
it
yrequi
rement
sforhi
gherl
evel
sofpr
ofi
ciency

Weknowev enl essaboutt heamountofl owf i

l
ter/compr ehensibleinputnecessar yfor
progresst ohi gherl ev elsofcompet ence.Wecangetsomei deaf r
om t heUni tedSt ates
ForeignSer vi
ceI nsti
tut echart
,anest i
mat eoft heamountofcl asst i
menecessar ytoachi evea
FSI2+r atingindi ff
er entforei
gnl anguages( 2+i sdef i
nedas" half
waybet weenmi nimal
professionalprof i
ciencyandwor kingpr ofessionalprofi
ciency",Dill
er,1978,p.100)f oradul t
Englishspeaker s.Accor di
ngtotheFor eignSer vi
ceInstit
uteest i
mat es(reproducedi nDi ll
er,
1978) ,Europeanl anguagessuchasGer man,Fr ench,andI t
ali
anr equireappr oximat ely720
hoursofcl assti
mef ort he"average"st udentt o at
taint he2+ l evel,whi l
emor e" exoti
c"
l
anguages( suchasAr abic,
Korean,andChi nese)requi
re1950hour sofcl assti
me. 14

These fi
gur
es may ,however
,represent an upper bound.They ar
e based on
"cl
assroom hour
s",
which,
iftr
adi
ti
onalmet
hodsar eemployed,
maynotentai
lopti
mali
nput
.In
otherwords,wecandobetter
.

72
72
 "How much i nput?"r emains an empi r
icalquest i
on,one t hatcan pr obabl
y be
adequatelyanswer edbyr esear ch.Tobemor eprecise,wewoul dli
ketoknow:"Howmuchl ow
fi
lt
er/compr ehensibleinputi snecessaryforstudentst oacqui reenoughcompet enceinthe
secondl anguage,sot hatt heycanuset heinformalenv i
ronmentt ocont i
nuei mprov
ing?
"
Despiteourcur rentpauci tyofdat a,whatseemscl eartomenow i sthatwear enotusing
enoughoft heavailableinstructi
ontimeforsuppl y
ingcompr ehensibl
einput
,andt hatwewill
beabl etostimulatemor er apid(andmorecomf ortable)secondl anguageacquisi
ti
onifweput
great
erf ocusoni nput.

Beforeconcl udingthi
ssect i
on,Ishoul
dpointoutt hatwhatIam suggest ingisnotatall
new:alongwi t
hNewmar k(1971),Iam suggesti
ngt hatthe"ex tensive"sideoft heextensi
ve-
i
ntensi
v ereadi
ngdebat eiscorrect,t
hatstudent
spr ofi
tmor ef rom readingf ormeaning,and
readi
nggreatquant it
iesofmat eri
al,thanfr
om whatNewmar kcal l
s"cryptoanalyti
cdecoding"
ofdif
fi
cultparagraphs, andthatstudentsgai
nmor efrom partici
pat i
ngi nconversati
ons,many
conversat
ions,t
hanf rom focusedlist
eningcomprehensionex ercises.

Weturnnow totwoot
herf
eat
urespr
ogr
amsshoul
dcont
aini
ftheyar
etoencour
age
l
anguageacqui
sit
ion.

E.Ot
herFeat
urest
hatEncour
ageAcqui
sit
ion

1.THESTUDENTSHOULDNOTBEPUTONTHEDEFENSI
VE

The phrase "

on the def
ensi
ve"comes f r
om Stevi
ck' lknown book,Memor
s wel y,
Meaning,andMet hod.Whati tmeanstomei sthatmethodsandmateri
alsshoul
dnotbea
testofthestudent'
sabili
ti
esorpri
orexper
iences,shoul
dnotmerel
yrev
ealweaknesses,but
shouldhel
pt hestudentacqui
remore.

Moregener all
y,wear etal
kingaboutkeepingt heaf f
ecti
vefil
ter"low",makingsur ethe
st
udentisopent otheinput.Itmaybethecaset hatifweusepr oceduresthatar e"true"t
othe
i
nputhypothesi
s,andt hatsati
sfyallt
heot herchar acteri
sti
csofopt imali nput
,t hekindof
i
nputthatresul
ts,andt heclassr
oom procedurest hatev ol
ve,wil
lsati
sfyt hisrequirementas
wel
landhel pkeept hef i
lt
erlow.Iwi l
lattempt ,int hissecti
on,toout l
ineaf ew general
pr
oceduresandpr acti
cesthatdothis.

73
73
 Fir
st,andIapologi
zef orharpingont hi
sissuesomuch, i
fweconcentrat
eonsuppl y i
ng
compr ehensibl
einputwheret hefocusi sont hemessageandnotont heform,thiswil
linof
i
tsel
fcont ri
butet oalow filt
er.I
ft het opi cbei
ngdiscussedisatalli
nter
esti
ng,andi fitis
compr ehensibl
e,muchoft he"pressur e"normallyassoci
atedwit
hal anguageclasswi llbe
"of
f",anxiet
ywi llbelowered,andacqui si
ti
on willresult
.Asment i
onedearli
er,Ithinka
desi
rablegoalist hatt
hestudent" f
or get",inasense,t
hatthemessageisactual
lyencodedi n
anotherlanguage.

Second,wewi llbeabl etokeepthef i

lterlow bynoti nsistingont oo-earlypr oduction,
beforet he st udenti s" r
eady "
.Language teacher s( and st udent s)associ at
e pr ogress in
secondl anguageacqui sit
ionwi t
hspeakingfluency( "Doy ouspeakFr ench? "
),andt hel ogical
consequenceoft hi
si sthatwewantourst udentst otalkfrom t hebegi nning.Myper sonalv iew
i
st hatforcingear l
ypr oduct i
on,befor
ethest udenthasbui ltupenoughcompet encet hrough
compr ehensi bleinput,i sper hapsthesinglemostanx iety-provoki ngt hingaboutl anguage
cl
asses!Whi lesomest udentsmaywantt ot al
kassoonaspossi bl
e,ot hersmayf eelless
secureunt i
lt heyhav ebui l
tupmor ecompet ence.I not herwor ds,t hel engthoft hesi l
ent
peri
odi sv ar i
able( seeNot e12) ;Asher
'sten-hourest imatemaybe" average",buti tisnot
carvedinst one.Asaf epr ocedureissimplynott oforcepr oduct i
onandl etthest udentdeci de
whent ostar ttalki
ng.

Closel
yr el
atedt ospeaki ngr eadinessf orpr oduct i
oni st hequest ionofer rorander r
or
cor r
ection.Secondl anguageacqui sit
ionresear cht ell
suscl ear l
yt hater rorsarei nevitabl
e, and
thatt heywi l
lbepl ent i
fuli
near l
yst ages.Togi v ether eaderanest imat e,inanex perimentwe
conduct edatQueensCol lege,wef oundappr oximat elyoneer rorf orev eryf i
vewor dsi n
composi ti
onswr i
ttenbyESLst udent sinapl acementex aminat ionf orourex t
ensioncour se
(Krashenetal .,1978) .Thebet t
erst udent saver agedaboutoneer rorf orev erytenwor ds,and
thel eastproficientaboutoneer rorforev erytwowor ds!Asur emet hodofr aisi
ngt hef i
lteris
attempt ing to cor recter r
or s,especi all
yi n begi nning stages and especi all
yi n spoken
l
anguage!Er rorcor recti
onis, unf ortunatel
y ,
thepr of ession'
sty picalreact iontoer ror,andinmy
viewi thasbeenaser i
ousmi stake.Ther ear esev er alreasonswhyi tisami stake.Wef ocus
her eonwhati spr obabl ythemostser iousf l
awi ner rorcorrect i
on,i tsef f
ectoft heaf f
ect i
ve
fi
lter.

74
74
 Err orcor rectionhast heimmedi at
eef fectofput ti
ngt hestudentont hedef ensive.It
encour ages a st r
ategy i n which the studentwi lltryt o avoi
d mi stakes,av oid di f
fi
cult
construct ions,f ocus l ess on meani ng and mor e on form.I tmay di sruptthe ent ire
communi cativef ocusonanex change.Thiswasbr oughthomet omei nademonst rati
onI
oft
enempl oyt oi l
lustratehow t heI nputHy pothesisappl i
est oclassroom t eaching.I nthe
demonst ration,whi chIhav eborrowedf rom St evenSt ernfel
d,Itelltheaudi encet hatIam
goingt ogi vethem t wol essonsinaf or
eignl anguage( IusuallyuseGer man) .Fort hef i
rst
l
esson,Isi mplyst arttalking,sayi
ngt hi
ngsl i
ke" Iam nowgoi ngtogivey ouaGer manl esson,
butf i
rstofal lletmet ellyousomet hingaboutt heGer manl anguage,et c.etc."Thi
si sdone
entir
elyinGer man,andi snear l
ycompl etel
yi ncompr ehensibletothosei ntheaudi encewho
havenev erbeenex posedt oGerman.Thesecondl essonissomet hi
ngl i
ket his:

Thisismyshoe.( Pointtoshoe)Thisi smy

hand.( Pointt o hand)Thi sis myhead.
(Pointt
ohead)
Thisisahead.( Drawpi ct
ureonboar d)
Herearet woey es.(Dr awey es,
holdupt wof i
ngers)Her
eisamout
h.
(Drawinamout h)
Hereisaci garette.(Drawi nacigar
ette)
Doy ouhav eaci garetteforme?( Walkupt oclassmember,
makeci garettesmoki ngmot i
on,pointtoself
.)

Thepoi ntofl essonnumbert woi st hatwhi l

eitmaynotbev eryinterest i
ng,itisquite
compr ehensible,thankst othesimplel anguage,theex tra-l
inguisticsuppor t
,et c.Therei salso
anat t
emptt obr i
ngdownt hefil
terbydr awingaf unnyheadandaski ngforaci garette.Inthe
discussi
onf ollowingt hi
sbr i
eflesson,Iex plai
ntheset hingsandmaket hecl aimt hatifsuch
i
nputi sprovideov eraper iodoftime, speechwi l
lemer geoni tsown.Whati sofi nt
erestt ous
hereist her eacti
onoft heaudience:i tisoneofr el
ief.Sev eralpeopl ehav ecomeupt ome
afterthel ecture,andsai dsomet hi
ngl i
ke:"Wheny ousai dy ouwer egoi ngt ogi veusa
l
anguagel esson, Igotv erynervous.Iwasaf r
aidyouwoul dcal lonmeandIwoul dhav et osay
somet hi
ng,andIwoul dmakeami stake."Whatt histellsmei sthatl anguagel essonsi nspi
re
fearevenamongpr ofessionall
anguaget eachers,andoneoft her easonsf orthis

75
75
i
sourinsi
stenceonearl
yspeakingandouratt
itudest
owar
dser
ror
s.Whymakest
udent
s
suf
ferf
rom procedur
esthatar
eunpleasantev
ent ous?

Thereismor etosayont het opicofer rorcor r

ecti
on; i
thassomeadv antages,andot her
disadv antages,andwewi l
llookatt hesel ateronwhenwedi scussconsciouslearni
ngi nt he
classroom Whatneedst obesaidher eisonl ythater r
orcorrecti
onisnotthebasi cmechani sm
forimpr ovi
ngsecondl anguageper formance;r ather,weacqui reviacompr ehensibleinput ,
accor dingtothet heor y
.Sinceov eruseofcor recti
onhassuchnegat iv
eef f
ectsf oracquisiti
on,
andsi nceer rorcor recti
oni snotofdi rectbenef ittolanguageacqui sit
ion(seeChapt erI I
,
discussi onofhy pothesisone),asaf epr ocedur eissimplytoel i
minat
eer r
orcorrecti
onent ir
el y
i
ncommuni cati
v e-t
ypeact iv
iti
es,apr ocedureusedwi t
hgr eatsuccessi nTer rel
l'
sNat ur al
Appr oach.Impr ov ementwi l
lcomewi thouter rorcorrect
ion,andmayev encomemor er apidly,
sincet heinputwi ll"
getin",t
hef i
lterwi l
l belower ,andstudentswi l
lbeoffthedefensive.

2.PROVI
DETOOLSTOHELPSTUDENTSOBTAI
NMOREI
NPUT

Ourresponsibi
li
tygoesbey ondt helanguageclassroom. 15 Indeed,asIhav est ated
earl
ier
,ourt askist oprovi
det hestudentswiththetoolst heyneedt ocont i
nuei mproving
wit
houtus.Weneedt oprovi
deenoughi nputsothattheycangai nthel i
nguist
iccompet ence
necessarytobegintot akeadvant
ageoft heinf
ormalenvironment ,theout si
deworld.I
not her
words,theyneedt oknowenoughoft hesecondlanguagesot heycanunder st
andsigni
ficant
port
ions ofnon- cl
assroom language.Buildi
ng thei
rl i
nguistic compet ence tothis point
,
however,isnotenough.

Eveni fwedosucceedi nbr ingingourst udentstothisstage,theywillhav eproblemsi n

usi
ngt hel anguageont heout si
de.Theywi l
lsti
llnotunderstandagr eatdealoft heinputt hey
hear,eveni fi
ti smodi fi
ed.Theywi l
lfindt hemselvesatal ossf orwords,andwi l
lmake
mistakes atal ll evels.Ifwe f ocus onl y on providi
ng t he i
nputf orpur el
y" li
nguisti
c"
compet ence,wewi l
lhav estudent swhoav oidcontactwithnat i
vespeakersf orfeartheywi l
l
notunder standmuchofwhati ssai dt othem, andwhowi l
lhaverealproblemswhent heyar e
engagedi nconv ersati
on,includingpai nfulsi
lenceswhi l
etheysear chforwor ds,confusionand
embar r
assmentduet omi sunder standing,etc.

76
 Thi
si s,Ithink,
thet ypi
calsi tuationinthef or
eignlanguaget eachingscenei nt heUnited
States.Af tert woy earsofi nstruction,thest udentwhoi sevenwi lli
ngt opar ticipateina
conv ersat
ionwi t
haspeakeroft hel anguageheorshehasst udi
edi sr are!Thesol utiont ot
his
problem ist ogi veourst udent st het oolstheyneedt oovercomet hesedi f
fi
culties,t omake
them conv ersat i
onall
ycompet ent.Bygi vi
ngt hem themeansofmanagi ngconv er sati
ons,we
canhel pthem t ocontinuei mpr ov i
ngbyal lowingthem t oparti
cipatei nconversat iondespite
thei
ri nadequaci es.Wecanpr eparet hem forthecer t
aint
ythattheywi l
lnotbeabl et ofindthe
ri
ghtwor d,t hattheywi llnotunder st
andev erythi
ng,andwecanhel pinsuret hatt heywi l
l
continuet oobt aincompr ehensi bl
ei nput.

Whatarethetool
sst udent
sneedtomanageconversat
ionandthusconti
nuetogain
i
nputoutsi
dethecl
assr
oom?I notherwor
d,howcanwehelpourstudent
stoconv
ersedespi
te
l
essthanperf
ectcompetence?

Scarcel
l
a( f
orthcoming)hasst at
edthattherear eatl easttwoway sconv ersat
ional
compet encecanhelpt heacquir
ergai
nmor ecompr ehensi
bleinput:devi
cesthathelpcontrol
the quanti
tyofinput,anddev i
cesthathel
pcont rolthe quali
ty.Thef or
merwi l
lhel pthe
acquirergetmoreinput,t
helatt
erwil
lhel
ptomaket heinputcompr ehensi
ble

Component s ofconv ersati

onalcompet ence incl
uded underthe quantit
y categor
y
i
ncludeway sofst ar
ti
ngconv ersat
ions(
gr eeti
ngs)andway sofkeepingconversati
onsgoing
(e.
g.poli
tenessformulae).Scarcel
la'
ssubjectMi guel
,a21y earoldspeakerofEngl i
shasa
secondlanguage,despit
eonlymodest" l
inguisti
c"competence,wasquit
eadeptatt hesetool
s
ofconversat
ionalcompetence,asthefol
lowingbr i
efexcer
ptshows:

Miguel
: Hi
!Howareya?
NS: Okay .
Miguel
: What'
snew?
NS: Notmuch.Hadat estt
oday.
Miguel
: Ohthat
'st
oobad.Whattest
?

Miguel
,by using a f ew well
-chosen rout
ines,is abl
et oi nit
iat
e and mai ntai
n
conversati
ons.SecondlanguageperformerssuchasMi guelarenott herule,however,as
resear
chhasshownt hatsecondlanguageacquirer
sof t
enhavesurpr
isingandser i
ousgaps
whenitcomest oconversat
ionalcompetence(seeover)
.

77
77
 Therearevari
oustechniquesthatar
eusedt omakeinputmor ecompr ehensibl
e,tocontr
ol
thequali
tyoftheinput.Per
hapst hemostobviousi
ssimplyaski
ngt henativespeakerforhel
p,
"gett
ingthenati
vespeakert oexplai
npart
soft heconv
ersat
ion..
.byusingdi scour
sedev i
ces"
(Scarcel
l
a,p.5).Thesedevicesrangefr
om focussi
ngonasingleproblem wordbyr epeat
ingit
,
asin

NS: SalvadorDali
alsoputoutacookbookbecauseheisagreatexpertoncui
sine.
Miguel
:(l
ooki
ngconfused):
Cookbook?
NS: ( pi
cki
ngupacookbook):Reci
pesfrom Maxime'
s,pl
aces
l
iket
hat.
(
Scarcell
a,p.5)

t
out
ter
ancessuchas"
What
?",
or"
Idon'
tunder
stand.
"

Scarcel
laalsonot
esthatthequali
tyofinputcanbei mpr
ovedbytheuseof" backchannel
cues",cuesthatprovi
dethenativ
espeakerwi thevidencet
hattheconversati
onalpar t
neris
i
ndeedf oll
owingtheconv
ersati
on.Theseincl
udev er
balcuessuchas"Uhuh","Yeah",andnon-
verbalcuessuchasheadnoddingatappropri
ateti
meandey egazebehavior.

Fi
nall
y,thereareconversati
onalst
rategi
esthatavoidincompr ehensi
blei
nput,incl
udi
ng
waysofchangi ngthesubjectt osomethingeasi
ert ounderstandormor efamil
iartothe
acquir
er.Scarcell
a'
ssubj
ectMi guel
isqui
tegoodatthis,ast
hef oll
owingdemonstrates:

NS: . ..Il
i
keclassi
calmusi
ctoo-
-Beet
hov
en,
Schuber
t-
-youknowt
hatki
ndast
uff
.
Miguel
:Youpl aythepi
ano?
Joe: Yeah.
Miguel
:Met oo.

F."
Teachi
ng"Conv
ersat
ionalCompet
ence

Knowledgeoft hecomponent sofconv ersat

ionalcompetenceisonething.Dev
elopi
ng
conv
ersat
ionalcompetenceinstudentsisanother.Thequesti
onthatneedstobeaskedhere
i
swhetherconversat
ionalcompetenceislear
nedoracquired.

Therearegoodarguments,Ithi
nk,agai
nstthehypot
hesi
sthatal
lofconv
ersat
ional
competencei
slear
nabl
e(seeal
sodiscussi
oninScar
cel
l
a,

78
78
f
ort
hcomi
ng)
.Fi
rst
,iti
ssi
mpl
ytoocompl
ex.AsScar
cel
l
apoi
ntsout
:

"Most discour se rules and st r

at egies are v ery complex,char acter
ized by vocabul ary,
pronunci
ation,andpr osodicfeatures,f eatur
esofnon- ver
balcommuni cati
on,and,perhapst oa
l
esserdegree, syntacticfeat
ures.Mor eover,al
lofthesefeatur
esmayv aryaccordi
ngtot hesoci al
context
.Fori nstance,somegr eeti
ngsar eappropri
ateinsomesi tuat
ions,butnotinother s.They
are sometimes shout ed,and somet i
mes spoken qui t
e stif
fl
y .In any given si
tuat i
on,an
appropri
ategr eeti
ngdependsonav arietyoffactor
s.Theseincl
ude:theper sonbeinggreet ed,
the
ti
meofday ,thelocat i
onandt heint eracti
on,otherpeoplepresent,andt hesortofint eracti
on
whichisexpect ed(pleasant,scornful,etc.)
"(p.10).

Second,evenift
hestudentmanagestol
ear
nsomer ul
esofconver
sationalcompet
ence,
they
wil
lnotalwaysbeavail
ablewhentheyar
eneeded:
inMoni
tor-
fr
eesituati
ons.

Mostli
kely,the non-univ
ersalaspects ofconv ersationalcompetence haveto be
acquir
ed.Ani ni
ti
alhy pot
hesisist hattheyare acquired t he samewaygr ammaris,via
comprehensi
blei nput
,and t hatthe requir
ements presented inthis chapt
erneed to be
sati
sfi
edfortheacquisiti
onofconv ersati
onalcompetenceaswel l
,averydiff
icultt
askgiv
en
theti
meanddi scourseconstrai
ntsoftheclassroom.

I
tispossi bl
e, however
,thatasmallsub-
setofconv ersati
onal managementt oolscanbe
dir
ectl
yt aught,
ei t
herasr ul
esorasmemor i
zedrouti
nes,asl ongast heyareeasyt olearn(see
secti
ononl earning,tofoll
ow).Theseincl
uderouti
nesforst arti
ngaconv ersat
ion,somepause
fi
lt
ers,andex pressionsthataskforhelp(Scar
cell
a,p.11) .Also,ahostofi n-
classandout -of
-
cl
ass act i
vit
ies hav e been i nt
roduced i
nr ecentyear s t o encour
age conv ersati
onal
compet ence.

Mymai npoi nti

nt hi
ssectioni st hatconv er
sat
ionalcompet encegivesst udentsthe
toolstheyneedt omanageconv ersation,andi sthusanessentialpartofinstr
uction,sinceit
helps toi nsurethatlanguage acquisiti
on wi lltake pl
ace outside ofclass,and af terthe
i
nst r
uctionalpr ogram ends.Ourr esponsi bil
itydoesnotend wi tht hecompl etion ofthe
semest er:indeed,inmyv i
ew,t
hepur poseofl anguageinstr
ucti
onist oprovidestudentswi t
h
whatt heyneedsot hattheycanprogresswi thoutus.

79
79
Not
es

1 Asdet ail
edi nChapt erII
,speechpr oducti
oncancomef rom anyoft hreedi f
fer
ent
sources.Fi r
st,wecanuseouracqui redcompet enceasi l
lustrat
edi nt heMoni tormodelf or
productioninChapt erII
.Accordingtot heinputhy pothesi
s,thissor tofpr oductiont akessome
ti
met o dev el
op.Anot herwayi svia memor iz
ed pat t
ernsand r outi
nes( seeKr ashen and
Scarcel
la,1978) .At hi
rdwayi sbyex tensi veuseoff i
rstlanguagest r
uctures,asex plai
nedn
ChapterI I.Thel attertwomet hodsofspeechpr oducti
onar eway sof" performi ngwi t
hout
compet ence"( borrowingR.Cl ark'
stermi nology).Asecondl anguageper formercan" l
earnto
speak"v ery quickly usi
ng these met hods,and t hey are ex pli
cit
ly encour aged by some
techni
ques.Theyar esev er
elylimi
tedmodes,howev er.(Seedi scussioni nChapt erI I
,Krashen
andScar cell
a, 1978,andKr ashen1981.)

2 Thi sraisestheinterest
ingquest ionofwhet herparti
cipat i
oni nconv ersati
onisev en
practi
call
ynecessar yfortr
ulysuccessf ulsecondl anguageacqui sit
ion.Itprobablyis.Inaddit
ion
tobeinganef f
ectivemeansofobt ainingcompr ehensibl
einput,conv ersati
onof ferssomeot her
realadvantagest hatwil
lbecomecl eareraswepr oceedinthi
schapt er.Scarcell
a( for
thcoming)
pointsoutt hattherearemanyaspect sof" communi cati
vecompet ence"t hatarepr obabl
ynot
acquirabl
ebyobser vat
ionandi nputalone( seedi scussi
onlateri nthischapt er
).Also,Scarcell
a
pointsoutt hatrealconversati
onent ail
s" ahighdegr eeofper sonali nvol
vement ",whatStev i
ck
(1976)terms" depth"andal oweredaf f
ectivefil
ter.

3 I
nar eviewofthesciencefi
cti
onli
terat
ure,Hatch(1976)poi ntsoutseveralexamplesin
whichaut horsassumet hatitispossibl
et oacqui rehumanl anguagebyl i
steni
ngt oradio
broadcasts.Event heseauthorsseem tounder stand,howev er,thatacquiri
ngl anguageby
l
isteni
ngtoi ncomprehensi
bleinputi
sanabilit
ypossessedonl ybycer tai
nalienswithdiff
erent
,
andappar entl
ysuperi
or"l
anguageacquisi
ti
ondev i
ces".

4 There ar
e anecdotalcasesofpeopl e who hav e picked up second l
anguages v i
a
tel
evi
sion.Larsen-
Freeman( 1979),forexample,cit
edacaseofaGer manspeakerwhoacqui red
Dutchv i
aTV.Thi sisnotatal lstr
ange,asmuchi nputinDut chwoul dbecompr ehensi
bletoa
speakerofsuchacl oselyrelat
edlanguage.Not ethatIam notcl ai
mingthati si
ti mpossibl
et o
acquirelanguagef r
om TV.Iam onl ysay i
ng thatcompr ehensibleinputisnecessaryf or
acquisi
ti
onandt hattel
evisi
onpr ovi
desl i
ttl
ecompr ehensi
blei nputforabeginner
.Int
ermediate
l
evelstudentsmaypr of
itquiteabitfrom tel
evisi
onandev enradi o.

5 Anot herwayt eacher shelpst udentsunder standmessagescont aini

ngst ructur
est hatare
"beyond"t hem i sbyemphasi zi
ngv ocabular
y.Bot hEv elynHat chandIhav estatedthear gument
fori ncreased v ocabul ary worki nr ecenty ears( Hatch,1978a;Kr ashen,1981) ,and our
argument at i
oni s,It hink,simil
ar.Whi leknowledgeofv ocabul arymaynotbesuf fi
cientfor
under st
andi ng al lmessages,t herei sl i
tt
le doubtt hatan i ncreased v ocabularyhel pst he
acquirerunder st
andmor eofwhati shear dorread( seee. g.UlijnandKempen, 1976; Macha,
1979, ont her oleofv ocabul ar
yinr eadi ngcompr ehensi on).Thus, mor ev ocabularyshouldmean
mor ecompr ehensionofi nput,andmor eacqui siti
onofgr ammar .Thi s"new v iew"isqui te
diff
erentf rom ear l
ierposi ti
ons.Languaget eacher shadbeent oldtorest ri
ctint
roductionofnew
vocabul ar
yi nor dertof ocusonsy nt ax
.Nowwear esay i
ngthatv ocabularylearningwi l
lactuall
y
contributet otheacqui sit
ionofsy nt ax.

Thepract
icalimpli
cati
onsofthi
spositi
onarenotcleartome,however
.Shoul dwe
t
eachvocabul
aryinisol
ati
oni nanef
for
ttoboosttheamountofinputt
hati
scomprehensi
ble?
Unf
ort
unatel
y,t
hereisli
tt
leresear
cht
hatspeaksdi
rect
lyt
othequesti
onofhow

80
80
 vocabularyisbestacqui r
ed,and,mosti mportant
,r etained.Thereissomeagr eementamong
teachersthatv ocabul
aryshouldbet aughtincontex t
,r atherthanbyrotememor i
zat
ionofli
st
(seeCelce-MurciaandRosenz weig,1979,forseveralt echniques)
,butitmayevenbet hecase
thatvocabularyshouldnotbedirectl
ytaughtatall!I
tmaybet hecasethati
fwesupplyenough
compr ehensibl
einput,v
ocabularyacquisi
ti
onwilli
nf actt akecareofit
self
.

Letmerest
atethissuggesti
oninthefor
m ofaninfor
malex
peri
ment:Gi
ventenminutes
ofstudyti
me( wai
ti
ngf orabus,et c.
),whi
chactiv
itywouldbemoreusef
ulforthelanguage
acqui
reri
nter
est
edinlong-t
ermretenti
onofvocabul
ary?

(
1) Rot
elear
ningofal
i
st,
usi
ngf
lashcar
dsorsomeequi
val
entt
echni
que.

(
2) Rev
iewi
ngast
oryt
hathas"
newwor
ds"car
eful
l
yincl
uded(
Cont
ext
ual
i
zat
ion)
.

(3) Readingforpleasure,t
ryi
ngonl
ytounderst
andthemessageandl
ookingup
newwordsonl
ywhent heyseem tobeessent
ialt
othemeaningorwhentheacqui
rer
i
scur
iousastothei
rmeani ng.

Method( 3)rel
i
esoncompr ehensi
blei
nputtosuppl ynewv ocabularyi
nenoughf requency,
andt ohelpt heacqui rerdeter
minet hemeaning.Inmet hod( 3)ther
ei snoconsci ousfocuson
vocabular
y,onl yonmeani ng.Thepr edi
cti
on(hope?)ist hatreall
yimpor t
antwor dswil
lr eoccur
natur
allyandt heirmeaningswi l
lbemadei ncr
easingl
yobv iousbythecont ext
.Itdoesnotex clude
thepossibi
li
tyt hattheacquir
ermaybehel pedbyoccasionalglancesatthedicti
onaryoroccasional
defi
nit
ionsbyat eacher.

6 TheAmeri
canLanguageI
nst
it
uteatUSC,f
orexample,
off
ersacour
sei
nty
pingal
ongwi
thi
ts
of
fer
ingsi
nEngl
i
shasaSecondLanguageforfor
eignst
udent
s.

7 Ther
e maybe f urt
heroppor
tuni
ti
es forprovi
ding opt
imali
nputf
orsecond l
anguage
acqui
rer
sattheuni
ver
sit
ylevel
,whi
chwewill
discussi
nChapterV.

8 ForsomeMoni torusersandl i
ngui
sts,ex pli
citgrammat i
calinformat
ionis,oddlyenough,
i
nter
esti
ngandr el
evant,sincesuchacquir sbel
er i
ev eitwil
lhelptheirperfor
mance.Inthiscase,the
medium i
sthemessage.I facourseont hestructureoft hetargetlanguageistaughtinthetarget
l
anguage,andifthestudent sar
e" anal
yti
c"types,thecour sewi l
lbeasuccess.( Seecomment si n
Kr
ashen,1980,andinChapt erV.
)

9 Ther
eareex cept
ions,examplesofcompr ehensibleinputinwhichi+1maynotbepr esent.
Theseincl
udesituati
onsinwhi chthedi scoursei sli
mi ted,andi nmanyi nstr
ument
alusesof
l
anguage i
n which famili
ari
ty wit
haf ew r outines and pat terns may suff
ice f
orsuccessful
communicati
on(
e.g.deali
ngwi t
hgasst at
ionat t
endants, cl
erks,etc.)
.

10 "I
nternal
iz
at i
on",i
nmyi nterpret
ati
on,seemstomeant heacquisi
ti
onofar ul
ethatwasf ir
st
l
earned,wherelearningisassumedt ohav ecausedthesubsequentacquisi
ti
on.Accor di
ngt othe
theor
yofsecondl anguageacquisiti
onpresentedinChapterI
I,thi
sdoesnotoccur.Ihavediscussed
thi
sinseveralt
echnicalpaper
s(Kr ashen,1977)andwil
lrevi
ewt hi
spointi
nalatersect
ion.

11 Anotheranal
ogythatcomest
omi
ndi
s"shot
gun"v
ersussi
ngl
ebul
l
et.Thef
ormerhasa
bet
terchanceofhi
tti
ngthetar
get.

12 Varv
el(1979)descr
ibesasilentper i
odinfor
malinst
ruct
ion(
Sil
entWaymethodology
)that
l
astedconsi
der
ablylonger
,indi
cat
ingt hattheremaybeaf ai
ramountofindi
vi
dualvar
iat
ioninthe
dur
ati
onofthesil
entperi
odforadult
sinl anguagecl
asses:

"Ther
ewasawomanf r
om Tai
wanwhoaf t
erseveralweekswasst i
l
lconspi
cuousl
y
sil
entinclass.Shenevertal
ked,andwhencal leduponwoul donlyanswerina
whisper
,sayi
ngonlywhatwasrequi
red.I
twasclear,howev
er,
thatshewasoneof
81
81
 themostat tent
ivest
udent
sint
heclass,hadaclearunder
standi
ngofwhatwas
beingdone,andseemi ngl
yenj
oyedtheclass.Shealsohadaposi t
iveat
ti
tude
towardswhatandhowshewasl earni
ng.Atnotimewasshecoer cedi
ntoacti
ve
parti
cipat
ion.

"Thenonedayint heni
nthweekofschoolshesatinthefrontr
owandact i
vel
y
parti
cipat
edthr
oughoutthewholehour.Fr
om thatpointon,sheconti
nuedto
parti
cipat
eacti
vel
yinamor eli
mitedwayandatt i
meshel pedother
sandwas
helpedbyother
s..
."(
p.491)

Whiletheremayhav e beenot
herr easonsf orthi
sstudent
'ssil
ence,thi
sexample
suggestst
hatthesil
entper
iodshoul
dberespected,andthatsomestudent
sdev el
opspeaki
ng
readi
nessl
aterthanot
hers.

13 Gi
venthesameamountofcompr ehensibl
einput
,thechil
d'
ssi l
entper i
odinsecond
l
anguageacquisi
ti
onmayt urnouttobelongerthantheaverageadultsi
lentperiodforother
reasons.WhatIam suggest i
ng her
eist hatthe sil
entperi
od in chi
ld second language
acquisi
ti
onwouldnotbeaslongi
fmoreoftheinputt
hechil
dhearsiscompr ehensi
ble.

14 Notethatifweassumet hatanacquirerinthenaturalenvi
ronmentreceivesaboutt wo
hour sperdayofcompr ehensibl
einput,720hourstransl
atesintoaboutoney ear"abroad"
.This
assumest hatclasst
ime=compr ehensibl
einput,whichmaynotbet ruewi t
ht het r
adi
ti
onal
met hodst heFSIchartisbasedon.I ti
s,however,i
naccor dwi t
htheinformallyacceptedidea
thatay earabroadwill
resultinafairdegreeoffl
uencyinthecaseofEur opeanlanguages.

15 Themateri
alcont
ainedinthissecti
onissimplymysummar yofwhatIhavel
earnedf
rom
discussionswit
hStevenSternf
eld,RobinScarcell
a,andBaty
iaEl
baum.Ithankthem notonl
yfor
theinformati
onandintel
l
ectualsti
mulation,butal
sof ort
hei
rpat
ience.

82
82
Chapt
erI
V

TheRoleofGr
ammar
,orPut
ti
ngGr
ammari
nit
s
Place
Asshoul dbeappar entbynow,t heposi
ti
ont akeni nthisbooki sthatsecondl anguage
teachingshoul df ocusonencour agingacqui
siti
on,onpr ov i
dinginputthatstimulatest he
subconsciousl anguageacqui si
ti
onpotenti
alallnormalhumanbei ngshav e.Thisdoesnot
meant osay,howev er,t
hatthereisnoroom atallforconsciouslearni
ng.Consciouslearning
doeshav ear ole,buti ti
snol ongerthel
eadactorintheplay.Thepur poseofthi
ssectionist o
discusswhatt hatr oleis,howwecanputconsci ouslearni
ng, or"gr
ammar "i
ni t
sproperpl ace
i
nt hesecondl anguagepr ogr
am.

A.Lear
ningDoesNotBecomeAcqui
sit
ion

ChapterI
Iattempt edt omakecl earwhatl earningdoesandwhati tdoesnotdointhe
theoreti
calmodelofsecondl anguageperformance.Accor di
ngt ot heMoni t
ormodelfor
performance,consciousl earningactsasanedi tor,asaMoni t
or,"corr
ect
ing"theer
ror
s,or
ratherwhattheperformerper ceiv
estobeer rors,int heoutputoftheacquiredsyst
em.This
canhappenbef orethesent enceisspokenorwr i
tten,orafter
.Consciousknowledgeofrul
es
i
st heref
orenotresponsiblef orourfl
uency,
itdoesnoti ni
ti
ateutt
erances.

Averyimportantpointthatalsoneedstobest atedisthatleani ngdoesnot" tur

ninto"
acquisi
ti
on.Theideat hatwef i
rstl
earnanewr ule,andeventuall
y,thr
oughpr act
ice,
acquireit
,
i
swi despreadandmayseemst osomepeopl etobei nt
uit
ivelyobv ious.Thismodeloft he
acquisi
ti
onprocesswasf ir
stpresentedtomewhenIwasast udentofTESL,andseemedt o
bev erysensi
bleatthet ime.Itwas,Ithought,exactl
ythewayIl earnedlanguagesmy sel
f.I
acceptedaspenetrati
nginsightCarr
oll'
schar
acteri
zati
onofhowl anguagel earni
ngproceeds

83
83
f
rom t
hepoi
ntofv
iewoft
het
hennew"
cogni
ti
ve-
code"school
oft
hought
:

"Oncethestudenthasapr
operdegr
eeofcogni
ti
vecontrolov
erthest
ructureofalanguage,
faci
li
tywil
ldevel
opautomati
cal
l
ywitht
heuseofthelanguageinmeaningfulsi
tuat
ions"(
Carr
oll
,
1966,p.102).

Asment i
onedi
nNote10oftheprevioussect
ion,
thi
spr
ocessofconv
ert
ingl
ear
nedr
ulesi
nto
acqui
redrul
eswascal
led"
int
ernal
izati
on".

Despi
teourfeeli
ngsthati
nternal
izati
ondoesoccur,thet
heorypredict
sthatitdoesnot,
exceptinat r
ivi
alway.Languageacquisiti
on,accordi
ngtothetheor
ypr esent
edinChapt erI
I,
happensi noneway ,whent heacqui r
erunder st
andsinputcontai
ningast ruct
urethatthe
acquir
eris"due"toacquir
e,astr
uctureathisorher"i+1".

Thereisnonecessityforpr
ev i
ousconsciousknowl edgeofar ule.(Thetri
vi
alsensein
whichaconsci ousrul
emi ght"hel
p"languageacquisi
ti
oni siftheper formerusedar uleasa
Monitor,andconsist
entl
yappl i
edittohisownout put.Sinceweunder standourownout put,
partoft hatperf
ormer'
scompr ehensibl
einputwouldi ncl
udeut t
er anceswi ththatstr
ucture.
Whent hedaycamewhent hatperf
ormerwas" ready"toacquirethisal readylear
nedrule,his
ownper formanceofitwoul dqual
ifyascompr ehensi
bleinputat" i+1" .Inot
herwords,self-
sti
mulation!
)

I
naddit
iontothefactthatthetheor
ydoesnotdi
rect
lypredi
ctthatl
earni
ngneedsto
precedeacquisi
ti
on,t
herearev er
ygoodr easonsf
ormaint
aini
ngthisposi
ti
onthatemerge
fr
om observi
ngsecondlanguageperf
ormers.

First,
weof t
enseeacquisit
ionincaseswher
el ear
ningnev eroccur
red.Ther
earemany
perf
or mer swhocanusecompl exstr
uct
uresinasecondl anguagewhodonotknowt her ul
e
consciouslyandnev erdid.Therehavebeenseveralcasehi st
oriesinthesecondlanguage
acquisit
ionlit
erat
urethati
ll
ust
ratethi
sphenomenon,onewhi chIthinki
squitecommon.

Evely
n Hatch'
s student
s,CindyStaff
ord and Gi
ngerCovit
t,i
nterv
iewed one such
second l
anguage perf
ormer,"V"
,an ESL studentatUCLA,who ex hibi
ted consi
derabl
e
competenceinEngli
sh,butwhoadmi t
tedthathehadconsciouscont
rolofveryfew,ifany,
rul
es.The

84
84
fol
lowingex changescomefr
om anint
ervi
ew wi
th"V",whicht
akesplacewhil
eoneofthe
authorsisreviewinghi
scomposi
ti
onerr
ors(f
rom St
affor
dandCov i
tt
,1978;al
soquot
edin
Krashen,1978):

Int
ervi
ewer:(
Wheny ouwr
it
eacomposi
ti
on).
..doyouthi
nkofgr
ammarr
ules?
Doy outhi
nk"Shoul
dIhaveusedt
hepr
esenttenseher
eorwoul
dthepr
esent
conti
nuousbebett
er.
..
"

V:"Idon'
tref
ertothebooksandallt
hat,y
ouknow.Ij ustr
eferi
ttot
his,uh,my
j
udgmentand. .
.sensi
ngifI'
m wri
ti
ngitrightorwrong.BecauseIreal
lydon'
t
know..
.whatwhereexact
lyhow.
..t
hegrammat i
cal
rulesworkout.

Lat
eri
nthei
nter
view,
onei
nvest
igat
orasks:

I
nter
viewer
:Doy
out
hinkgr
ammarr
ulesar
eusef
ul?

V:Usef
ul?Yeah.Wheny
ouwantt
owr
it
etheyar
ever
yver
yusef
ul.

I
nter
viewer
:Buty
oudon'
tuset
hem wheny
ouwr
it
e.

V:Yeah,
Iknow.Idon'
tuset
hem.
..Idon'
tknowhowt
ouset
hem.

Anot
hergoodex
ampl eofan"
under-
user
"oft
heconsci
ousgr
ammari
sHung,st
udi
ed
byCohenandRobbi
ns(1976)
,whost
ated:

"Inevert
aughtanygrammar.IguessIjustneverlearnedt herulesthatwel
l.Iknowthatevery
ti
meIspeaki t'
spret
tycorr
ect
,soInev erthi
nkaboutgr ammar s.Ijustwri
tedownwhat everI
feell
i
keit.Ev
eryt
imeIwri
tesomethingIj
uststopthinking.Idon'tknowwhich(rul
e)t
oapply"(p.
59).

NotonlyiswhatHungsay sreveal
ing,butsoi showhesay sit.Thereare, f
orsur e,err
ors
i
nt hispassage,butt her
eisal socontroloff ai
rl
ycompl exsy ntaxandar ealabi l
it
yf orself
-
expression.(Notallunder -users succeed,of cour se;see,f or ex ampl e,Schumann' s
descripti
onofAlbertoinSchumann( 1978a).)Ifconsciousr ul
eshav etocomef i
r st
,howcan
weex plaincasessuchasV,Hung,andot hers?( Forothercasehi stor
ies,seeKr ashen,1978;
Staff
or dandCov i
tt
,1978;Kouni nandKr ashen1978. )Unl essallcasessuchast hesecanbe
shownt obeinst
ancesoft heuseoft hef i
rstlanguageorr out
inesandpat ternstheex i
stence
ofsuch cases show t hatpr evi
ous consci ous lear
ning i s notnecessar yf orl anguage
acquisiti
on.

85
 Second,weal soseel ear ningt hatneverseemst obecomeacqui sit
ion.Manyf ineESL
perfor mer s, whi let heyhav eacqui redagr eatdealofEngl i
sh,alsoknowmanyconsci ousr ules.
Theynev erthel ess make whatt heyconsi dert o be "careless"er rors on rul
es t hatar e
l
ingui sticallyqui t est r
aightforwar d.Thi soccur swhent heper formerhasl earnedarule,buthas
notacqui redi t.Thi shappenst ypical l
ywi thlate-
acquiredi tems,suchast hethi
rdper son
singul arendi ng on r egularv erbs i n Engli
sh ( "
He goes t o wor k ev er
y day."
).Whati s
particul arl
yi nter estingi sthattheseper formersmayhav eknownt heruleandhav epract i
cedi t
for many y ear s.Ev en af ter t housands of cor rect repet it
ions,and wi that hor ough
under st andingoft her ule,suchper f ormer ssti
llmake" careless"mi stakesoncer t
aini tems.
Whathaspr ev ent edl earningfrom " becomi ng"acquisit
ionint hesecasesi sthefactt hatthe
l
ear nedr uleisst il
l beyondt heacqui rer si
' +1.

Acasehistoryt
hatill
ustratesthissit
uati
onv er
ywel listhatof"P"(KrashenandPon,
1975).Pwasanex cell
entMoni toruser( anoptimaluser,asdescr i
bedinChapt erII
),anadult
wit
haBAi nLinguist
icswit hhonor s,whosewr i
tt
enEngl ishappear ednear l
ynative-l
ike.I
n
casualconver
sation,howev er,Pmadeoccasi onal"careless"error
son" easy"rul
est hatshe
hadknownconsci ouslyfortwentyy ears.Thus,evenwell-
learned,well
-practi
cedrul
esmaynot
tur
nintoacqui
siti
on.

Anex pl
anati
onofP' sprobl
em isthatthei
temsshemi ssedincasualconv
ersationwere
thosethatarelate-
acquir
ed,andheracquisiti
on,
whil
ev eryadv anced,hadsi
mplenotgonet he
fi
nalfew stepsinsy nt
axandmor phology.Shehadlearnedt heruleswell
,however,andwas
abletosupplythem underconditi
onswher eshecoul
dMoni tor.

Athir
dr easonf ordoubti
ngt hatacquisit
ionrequirespr evi
ousl earningist
hef actthat
eventhebestl earnersmasteronl yasmal lsubsetoft her ulesofal anguage.Asdiscussed
earl
ier(Hypothesis3,Chapt erI I
),ev en professi
onall i
nguist sadmi tt hatthei
rconscious
knowledgeofev ent hebeststudiedl anguagesi simperfect,anddi scover i
esofnewrulesar e
report
edwi t
hev eryissueoftechni caljour
nalsinlingui
stictheor y
.Lingui stsof
tensucceedi n
descri
bing,aft
ery earsofanal ysis,whatmanysecondl anguageper formershaveal r
eady
acquir
ed.

My ex pl
anat
ion forthese phenomena i
st hatwhi
lel earni
ng may of
ten pr
ecede
acqui
siti
on,i
tneednot,andi
nf actmaynotevenhel
pdir
ect
ly.Rather,
weacqui
realongafai
rl
y
predi
ctabl
enatur
alorder,
and

86
86
thi
soccurswhenwe recei
ve comprehensi
blei
nput.Occasionall
y,we l earncer
tai
nr ul
es
befor
eweacquir
ethem,andthisgi
vesust heil
l
usionthatthel ear
ningactuall
ycausedthe
acqui
sit
ion.

Professionallanguaget eachers, wi
tht heirfascinati
onforthest ructureofl anguage, and
withthepl easuretheyder ivefrom themast eryanduseofconsci ousr ul
es, areof t
ennotev en
awaret hatacqui si
ti
onwi t
houtpr i
orconsci ousl earningispossible.Thi swasmyunex ami ned
assumpt ionaswel l
.Thepr oceduredescr ibedear l
ierseemedr ightandr easonabl et omeat
onet ime:l anguagel earni
ng,i nt hegener alsense,occur r
ed when onef i
rstconsci ously
graspedar ul
e,thenpr acti
cedi tagai nandagai nunt i
litwas" automat i
c".(Thi sisact uall
y
deductivel ear
ning;ther eisanot herpossi bili
ty,namel y,"i
nducti
v e"learning;seedi scussion
below.)Thegr eatcont ri
butionoflinguisti
cswast odi scoveranddescr iber ul
es, which"appl i
ed
l
inguists"couldtransmi ttolanguaget eacher s,who, intur
n,couldt el
lstudent saboutt hem.

Oneex per
iencet hathelpedt ochangemyt hinki
ngoccur r
edwhenIwast eaching
Engl
ishasasecondl anguagetoan" advanced"adul teducati
onclassatQueensCol l
ege.Asa
memberofat eam,myr esponsibi
l
itywas" st
ructure".SinceIwas,attheti
me,thedi r
ectorof
theEngli
shLanguageI nsti
tut
eatQueens,If eltobl i
gedt opresentanimpressiv
eser i
esof
l
essonsthatdemonst r
atedmycont rolofthesubj ect.Ither
efor
echoset oconcentr
ateont he
ver
bsy st
em,andpresentedacompl etesurveyofal ltenses.

Thef i
rstlessonoft hesessionwasf ocussedont hepresentpr ogressi
vetense.My
obj
ecti
vewast oinfor
m myst udentsthatthepr esentpr ogressi
vehadt hreemeanings:(
1)a
cur
rent
,on-goingact i
onthatwouldsoonbecompl eted,( 2),anactionthatbegansomet i
me
agointhepastandmayormaynotbet aki
ngpaceatt hemoment ,andwoul dendsomet i
me
i
nthefutur
e,and( 3)futur
etense.Ii
l
lustr
atedthisusingt hef amil
iartimef l
owdiagram

andbyshowi
ngt
hatsent
encessuchas

87
87
 Johni
spl
ayi
ngt
hev
iol
i
n.

wer
ethr
eeway
sambi
guous:

(
1)Whati
sthatnoi
sef
rom t
heot
herr
oom?(
Johni
spl
ayi
ngt
hev
iol
i
n.)

(
2)What
'
sJohndoi
ngt
hissummer
?(Hei
spl
ayi
ngt
hev
iol
i
nfort
hel
ocal
symphony
.)

(
3)What
'
sJohndoi
ngt
omor
row?(
He'
spl
ayi
ngt
hev
iol
i
nint
het
alentshow.
)

Noneofmyadv ancedESLstudent
sknewt hi
srul
econsciously
.Infact,
veryfewpeople
do.Ihav epresented t hi
sex ampl
esev eralti
mesatl ectur
est o pract
ici
ng ESLteacher
s,
andIoftenaskt hosewhoconsci ousl
y" know"therul
et hattheprogressi
veisthreeway s
ambi
guoustor ai
set heirhands.Ver
yfewdo, andthoset
hatdocl ai
mt heyknowithaveusual
ly
j
ustf
ini
shedteachingi tinclass.

Whatwasv eryi nterestingwast hatasignif

icantnumberofst udentshada" Eur
eka"
experi
ence.AfterIex plainedt herule,theywouldr emark:"That'
sr i
ght..
.itisthreeway s
ambiguous..
.how aboutt hat !
",orwoul dmakesi mi l
arcomment s.Myi nterpr
etat
ionisthat
thesestudentshadal r eadysubconsci ousl
yacquiredt heprogressiv
et enseandi tsthree
meanings,andwer econf irmingt hatt hei
racquisi
tionwascor r
ect.Ihad,i notherwords,
succeededinprovi
dingl earningwher eacquisi
ti
onwasal readypresent.

Iwoul dli
ket opointoutsev eralthingsaboutt hi
sphenomenon.Fi r
st,myst udentshad
apparentlyacquiredther ulewi t
houthav ingf i
rstlearnedi t.(I
tcoul dbear guedt hattheyknew
i
toncebuthadf or
gotteni t
,andthatt hist empor arylearninghadbeenessent i
al,oratleast
useful,i
nacqui r
ingther ule.Thisispossi ble,butunl i
kely,asal lthr eefuncti
onsar enotusual l
y
taught.Anot herunl i
kel
ypossi bil
i
tyi st ransferf rom thef i
rstl anguage.Mostoft hefir
st
l
anguagesofmyst udentst hatsemest erdi dnothav ethepr ogr essivetense.)Second,t hose
whol earnedwhatt heyhadal r
eadyacqui redt houghtt heywer egai ni
ngagr eatdealfrom the
class.Thissor tofknowl edgeisv erysat i
sfyi
ngt omanypeopl e(incl
udingme) .Itisnot ,
howev er,l
anguaget eaching,ev enthoughi tisofsomev alue.( Wer et
urnt othistopic,whichI
refert
oas" languageappr eciat
ion"
, l
aterint hi
schapt er.
)

Lear
ningsomet
imespr
ecedesacqui
sit
ioni
nreal
time:
Arul
ethati
s

88
88
eventuall
y acquir
ed may hav e been,atone t i
me,lear
ned only
.As Ihav e mai ntai
ned
elsewhere( Kr
ashen,1977),thi
s certainl
ydoes occur,butbyno means est abli
shes the
necessityofpriorl
ear
ningforacquisiti
on.JustbecauseeventAprecededeventBdoesnot
demonst rat
ethatAcausedB.Weseemanycasesofacqui sit
ionwit
houtlear
ning,learni
ng
(even verygood lear
ning t
hati swel lpract
iced)thatdoesnotbecomeacqui si
ti
on,and
acquiredknowledgeofrul
esprecedinglearni
ng.

B.ThePl
aceofGr
ammar

"
Grammar "
,at erm Iwi l
luseasasy nony mf orconsciouslearni
ng,hast wopossible
rol
esint hesecondl anguaget eachingprogram.Fi r
st,itcanbeusedwi t
hsomepr of
itasa
Monitor.Wewi lldiscusst hisuseinmor edetailinthesectionthatfoll
ows.Asecondusef or
grammari sassubj ect-
mat ter
,orfor"l
anguageappr eciat
ion"(sometimescalled"l
ingui
sti
cs")
,
andwewi l
ldiscusst hisrolelat
eron.Neit
herr oleisessenti
al,
neitheri
st hecentr
alpartofthe
pedagogicalprogram, butbot hhavethei
rfunctions.

Sev
erali
ssueswi l
lbedi scussedi
nrel
ati
ontot eachi
nggr
ammarforMoni
toruse:when
r escanbeused,whi
ul chr ulesshouldorcanbelear heef
ned,whatt fect
sofMonitoruseare,
andwhatwecanex pectintermsofMoni t
oreff
ici
ency.

1.GRAMMARFORMONI WHENTHEMONI
TORUSE: TORI
SUSED

Asstat
edi nChapterI
I(Hypothesi
s3),oneofourgoalsinpedagogyi st
oencour age
opti
malMonitoruse.Wewoul dl i
keourst udent
st out
il
izeconsci
ousr ulest
or ai
set hei
r
grammati
calaccuracywhenitdoesnotinter
ferewit
hcommunicati
on.Stateddi
ffer
entl
y,the
opti
malMonioruserknowswhent
t ouseconsciousrul
es.

Asment i
onedearli
er,onenecessar ycondi ti
onforsuccessf ulMoni t
orusei stime.It
takesr ealprocessingtimet or ememberandappl yconsciousr ul
es.Weshoul dnotex pect
mostst udentstosuccessfull
yappl yconsci
ousr ulestothei
rout putduri
ngor alconver
sati
on--
thereis, obv
iously,
li
tt
leti
me.Peopl ewhodoat temptt othi
nkaboutandut i
l
izeconsciousrules
duringconv ersati
onruntwor i
sks.Fi r
st,
theytendt otaketoomucht i
mewheni tisthei
rtur
nt o
speak, andhav eahesitantstyl
et hati
softendiffi
culttol
ist
ent o.Otheroverusersofthe

89
89
Monit
or,i
ntryi
ngtoavoidt
his,pl
anthei
rnextut
ter
ancewhi
let
hei
rconver
sati
onalpar
tneri
s
t
alki
ng.Thei
routputmaybeaccurat
e,buttheyal
ltooof
tendonotpayenoughatt
enti
onto
whattheot
herpersoni
ssayi
ng!

Somepeopl ear ebet t

erthanot her
satMoni t
oruse,andmayact uall
ybeabl eto
successfull
yuseaf airnumberofconsci ousr ules"onl i
ne".Mostpeopler unther iskof
seri
ouslyendangeri
ngt hesuccessoft heconv er
sat i
onwhent heyt
ryt oMonitordur
ingcasual
tal
king.(Successi
nMoni toruseinfreeconv er
sationalsodependsonot herf
actors-
-oneisthe
dif
fi
cultyorcomplexityoft herul
e,whi chwedi scussbel ow.Asecondi st
het opi
c:If indit
mucheasi ertopayattenti
ontothef or
m ofwhatIsayi nasecondl anguagewhenIam t alki
ng
aboutsomet hi
ngIam v eryfamili
arwi thandhav ediscussedbefore,e.
g.secondl anguage
acquisi
ti
on.)

Thepl
acef orMoni toruseiswhent heperformerhast ime,asinwr i
ti
ngandi nprepared
speech.Asstatedearlier,si
mplygivi
ngper for
mer stimedoesnoti nsuret hattheywillusethe
consciousMonitor;hence,condi t
ion2i nChapt erII:Theperformermustbet hinki
ngabout
corr
ectnessorfocussedonf orm.Whengi ventime, andwhenf ocussedonf orm,somepeopl e
can use conscious gr ammart o greatadv antage.I nthe case oft he second language
perfor
merwhohasacqui rednearlyallofthegrammaroft hesecondl anguage,butwhost i
l
l
hassomegaps, theuseoft heconsciousgr ammarcanf i
lli
nmanyoft henon- acquir
editems.
Thiscan,i
nwr i
ti
ngatl east,occasi
onallyr
esulti
nnat ive-
li
keaccuracy.

Ihaveoftenreferredt o"
P",discussedabov e,asaper for
merwhowasabl etodot his.
Despi teheraccent,andoccasi onalmor phol
ogicalerrorsinfreespeech,P' swr i
ti
ng(donei n
class)was near l
yf l
awl ess.Ihav e known manypr ofessi
onal swho al so use conscious
grammart hisway ,coll
eaguesi nli
ngui sti
cswhospeakwi thslightimper f
ecti
onsbutwhose
wr it
ingisnearlyer r
or-f
ree.Somev eryi nter
esti
ngcasesi nvolvespeci ali
stsingrammar ,in
formall i
nguist
ics,scholarswhocer tainlyconsciousl
yknowmanyoft herulestheyvi
olat
ei n
freeconv er
sation.TwocasesIper sonal l
yknowhav e,infact,publi
shedpaper sont hetheor y
ofgr ammart hatrelyheav il
yonEnglish, t
esti
fyi
ngt otheirdeepandt hor oughgraspofEngl i
sh
sy ntax.Yet,inunmoni toredf r
eespeech,t hir
dper sonsi ngularsdr opof f,thepossessive
mar kerisoccasionall
y

90
90
missing,et
c.Bot
hscholarspubl
ishallofthei
rcur
rentworkinEngl
i
shanddonotconsul
t
anyonetorevi
ewthei
rpaper
sforerr
ors,nori
sthi
snecessar
y.

Myownex periencemaybehel pfultor eaders.Iam,att heti

meoft hi
swr i
ti
ng,an
"i
nter mediate"levelspeakerofFr enchasasecondl anguage.( Thismeans,accordi
ngt omy
definiti
on,thatIcanconv ersecomf ort
ablywi t
hamonol ingualspeakerofFrenchasl ongas
(s)hemakessomecompensat i
on.Icannoteav esdropv erywel landhavesomet roublewit
h
radioandf i
lms.Al so,myout putisfair
lyf l
uent,butnoter r
or-
free.)Manypeopleatthislev
el,
i
ncl udingmy sel
f,makeer rorsonrulesthatareeasyt odescr i
be,butthatareapparentl
yfai
rl
y
l
ate- acqui
red.Oner ul
elikethi
sIhav enotedisthesi mpl econtracti
onrul
e:

de+l
e=du.

I
,andmycl assmat esini nter
medi ateconv er
sat
ionalFrenchatUSC,occasi onall
ymi ssthis
onei nf r
eeconv er
sation.Ont heoccasi onswhenIwr i
teFrench,howev er
,Igetitr i
ghtevery
ti
me.( Myaccur acyordi f
ficult
yorderchangeswhenIusemyconsci ousknowl edgeofFrench
grammar .Cor r
ectl
yappl yi
ngt hede+l e=dur ul
er aisesthi
si t
em from al owpositi
oninthe
diffi
cult
yor dert ooneneart het op.Thi sisexactlywhatIat t
empt edtosayi nChapt erI,
Hy pothesis3,indi scussingdi st
ortionsoft henaturalorderinMoni tor
edcondi t
ions.Idi
ffer
from theav eragesubjectint hatIdonotr equi
readiscrete-
pointgrammart esttofocusmeon
form.Mostr eadersoft hi
sbookar epr obablyl
i
kethisaswel l
.)

Thiski ndofbehav i
ori snat ur
alandnor mal .Whati st r
agic,i nmyopi ni
on,ist hat
teacher sex pectperfectperformanceofsuchsi mpl e,yetlate-acquireditemsi nunmoni tored
performance.Ev enqui t
ecompet entsecondl anguageuser s,suchasP, wil
l"miss"suchitems
i
nconv er
sat i
on.Weof tensee,howev er,beginners,studentswhocanbar elyconverseint he
targetl anguage,st ruggli
ng to makecor rectsubj ect
-verb agreementi n whatar et ermed
"communi cative"exerci
ses,fearfulofthet eacher'
sshat teri
ngcor rections.Thecauseoft his
tortureis,fi
rstofall,aconfusionbetweenl i
nguisti
csi mplici
tyandor derofacqui si
ti
on--
iti
snot
atal lthecaset hatthemor el i
nguisti
call
ysimpl eani tem is,theear l
i
eritisacqui r
ed.Some
very" simple"r ul
esmaybeamongt helastt obeacqui red.Second,t hecausei salsoaf ail
ure
todi sti
nguish

91
91
betweenacqui si
ti
onandl ear
ning,afail
uretoreal
izethatconsci ousknowledgeofani t
em
bearsnor el
ati
onshi
ptoaper f
or mer
'sabil
it
ytousei tinunmoni t
oredspeech.Thisabil
i
ty
comesf r
om acquisi
ti
on,andacqui si
ti
oncomef r
om compr ehensibleinput
,notfrom err
or
correct
ion.Theresul
tofsuchtreatmentis,
atbest,ov
eruseoft heMoni t
or.Atworst
,itr
esul
ts
i
nt heestabli
shmentofsuchast rongAff
ecti
veFi
lt
erthatacquisit
ionisimpossibl
e.

2.WHATCANBEMONI
TORED

Condit
iont hr
eeforMonitoruse(
ChapterII
, Hypothesis3)isrel
evanttodiscussingt
his
point
.Inor derf orperfor
merst oMonitorsuccessfull
y,t heymustknow t her ul
et heyare
apply
ing.Toex pandonapoi ntmadei nChapt erII,l
etmeat t
emptt oil
lust
ratejusthow
drast
ical
lythi
sr equir
ementli
mi t
sMonitoruse.Letthiscirclerepr
esental
ltherulesofawel l
-
descri
bedlanguage, suchasEngli
sh:

Letusnow considerallther ulesofEngl i

shthatt
hebestli
ngui
sts"know",orhave
succeededindescri
bing.How manyr ulesdidJespersen(
ever
)know,how muchofEnglish
havescholarssuchasNoam Chomskydescr ibed?Whil
eChomskyof
tensaysthatheandhis
col
leagueshaveonlydescribed"fragment s"ofEngli
sh,wewil
lgi
vetheformalli
ngui
ststhe
benefi
toft hedoubt,andrepresentt heiraccompli
shment
sasapr opersubsetofthefir
st
ci
rcle

Nowletusconsi
dert
herul
esthat"appl
i
edl
i
ngui
sts"know,
wher
eappl
i
edl
i
ngui
stsher
e
r
efer
stotheschol
arwhosetaski
stostudythe

92
92
workofthef ormal,theoreticalli
nguist,andpr esentittothel
anguageteacher
,andper haps
alsot
othel anguagest udent.Lettheaddi ti
onalsmallerci
rcl
einthenex
tfigurer
epresentwhat
theappl
iedl i
nguistknows.Thi swillhav etobeapr opersubsetofwhatt hef
ormall i
ngui
st
knows,sincethef ull
-t
imej oboft het heoreti
cianist oseekoutnew r
ules,whil
etheappl i
ed
l
ingui
stspendsagr eatdealoft i
meex pl
aini
ngt hiswork:

Thenextcir
cler
epresentsal
ltherul
esthatthemostknowledgeabl
elanguageteacher
s
know.Thi
swi l
lbeapropersubsetoftheci
rcl
eint
roducedi
nthelastfi
gure.Teacher
s,aft
eral
l,
haveagreatdealt
odobesidesstudytheworkofappli
edli
ngui
sti
cs:

Sti
llanothercircl
er epresentsthenumberofgrammarr ulesthatthebestt
eacher
s
act
uall
yteach.Thisis,inturn,apropersubsetoft
hesetofci
rcl
est heyknow,si
ncet
eacher
s
wil
lundoubtedl
ypresenttot heirst
udentsonl
yapartoft
hei
rknowledge:

Wedr awnextst
il
lanotherci
rcl
e,whichr
epresent
sal
lther
ulest
hatt
hebestst
udent
s
act
ual
l
ysucceedi
nlear
ning.Weshouldevenputi
n

93
93
onelastci
rcl
e,al
lther
ulest
hatstudent
scancar
ryar
oundi
nthei
rheadsasment
albaggage
andactual
l
yuseinperf
ormance:

Bynow,wear edowntoav er
ysmallcir
cle,ev
engiv
ingev
erygroupdi
scussedthebenef
itof
thedoubt.Ev
enourbestst
udent
swillbeabletol
ear
nanduti
li
zeasmallpartofthegr
ammar
ofalanguageasaconsci
ousMonit
or.

Wecandr awanev ensmal lercir

cleforsomepeopl e.Aswesawi nthedi scussi onof
i
ndi vidualv ari
ationinChapt erII
,someper formersar eei
therlesswi l
li
ngorl essabl et out i
li
ze
consci ousr ul
es.Att heopposi teex t
remef rom thepr of
essionalli
nguistorl anguaget eacher
(see,f orex ampl e,Yor i
o,1978) ,wehav etheMoni t
orunder-user,t
heper for
merwhodoesal l
self-correction by" feel
"and hasno cont rolofconsci ousgr ammar .Per hapsev en mor e
extremei st heincompet entMoni toruser,theperformerwhot hi
nks(s) heknowst her ul
esbut
hast hem ( oratl eastmanyoft hem)wr ong.Thismaymer elybeapr oblem ofnomencl at
ure
(e.g.St affordandCov i
tt'
ssubj ectwho keptr eferr
ingto" deadobj ects"i nst
eadofdi r
ect
object s),butmaybemor eser i
ous.Weshoul dbeawar ethatconf usionsmayex istev enfor
rulest hatappeart oust obev erysimple,andf orr ul
esthatt heperformermayhav eal ready
acqui redandcanut ili
zei nanunmoni tor
edsi t
uation.Suchcasesi l
lustrat
ev ividl
ythecont r
ast
bet weenacqui siti
onandl earni
ng.

(
a)I
ncompet
entMoni
toruse

Sel
i
ger( 1979)r eport
ed a si
mpl e,yetint
eresti
ng experi
mentwhi ch confir
ms t he
exist
enceofincompet entMonit
orusers.Thetaskwasnami ng:Subjectswereshownpi ct
ures
andaskedt osaywhatt heobjectpict
uredwasi nEngl i
sh( e.
g.It'
sapen) .Sel
i
gernot ed
whetherthesubjectsappl i
edthe"a/an"rul
eandwhet hertheycorrectyusedanwhent
l he
foll
owingnounbeganwi t
hav owel
.Thesubjectswerethenasked,aftercompleti
ngthetask,i
f
theyhadnoticedthatsometimesawas

94
94
cal
ledforandsomet mesanwascal
i ledfor.I
ftheysaidthattheydi
dnot i
ce,theywer easked
togiv
et hei
rreasonsf orthedist
incti
on.(I
nallcases,ei
theraoranwasr equired.Therewere
no cases wher et he zero all
omor ph was appropri
ate.)Seli
ger
's subject
si ncluded 29
monolingualEngli
shspeakingchildr
en,ages3to10.8,11" bi
li
ngual
"chi
ldren,ages4t o10,and
15adultESLst udentsatQueensCol l
egeinNewYor k.

I
nmyi nter
pret
at i
on,thisst udycont rast
sacqui sit
ionandl earning.1*T hesubj ect'
sfocus
i
nt hepi ct
urenami ngt askwasonsuppl yi
ngv ocabulary.Theywer enott oldinadv ancet hat
grammat i
calaccuracywas an i ssue and cer tai
nlythe a/ an rule was notpr esented or
discussed i n advance.The t ask,t hen,encour aged use oft he acqui red system;i twas
relatively"unmonitored".Thi si nterpretati
on is consistentwi tht he ev idence rev i
ewed i n
Chapt erI I
,whichconcl udest hatf ormostsubj ects,oneneedst odel i
ber atel
yf ocussubj ects
onf orm usingadev i
cesuchasadi screte-
pointgr ammart estinor dertobr ingoutex tensive
useoft heconsciousgr ammar .Ofcour se,sincet het askwasan" ex periment "
,itisqui te
possi blethatsomesubj ectsmayhav ebeenmor ecarefulthant heynor mal lywoul dbe.The
resul t
soft hedi r
ectquest onaboutaandan,howev
i er
,show t hatitisunl i
kelythatsubj ects
wer eaccessi ngmuchconsci ousknowl edgewhi l
eidenti
fy i
ngpictures.

Seligerrepor
ts"norel
ati
onship"betweenper for
manceonpi ctur
ei denti
fi
cati
onand
whetherthesubject
scouldstat
ear ule!Manysubj ectsdidnot"
dowhatt heysayt heydo".I
f
t
heirresponsest ot
hepost-
taskquest i
onrepresentconsciousl
earni
ng,thisresultconf
ir
ms
j
usthowl i
mi t
edlear
ningi
sforsomepeopl e.Letusex aminether
esult
s.

Aswewoul dex pectfr

om thediscussi
onofagei nChapt erII
,noneoft hebil
ingual
chi
ldr
enproducedcorrectconsci
ousrul
esf ora/an.Thi
sisconsistentwiththeclaimthatpre-
for
maloperat
ionschi
ldrenhavelessext
ensivemet a-
awarenessofgr ammar .Thepotenti
alfor
ext
ensiv
eMoni t
oruseishy pot
hesi
zedtoemer gewithfor
mal operati
ons,ataroundpuberty.

Amongtheadult
s,t
hreeoft hefourwho" knew"ther
ule(couldv er
bal
izei
taftert
het
est
)
"pr
oducednoinst
ancesont hepicturetesttoshowt heyunderstoodhowt herulewastobe
used"(
p.364)
.Thesesubj
ects,i
not herwords,hadlear
nedthea/ andisti
nct
ionbuthadnot

* Super
scr
iptnumber
sref
ert
oNot
esatendofchapt
ers.

95
95
acquir
edi t.Theywer eunabl e,mor eover
,toapplythisconsciousknowledget othepi ct
ure
i
dentif
icati
ont ask,si
ncethenecessar ycondit
ionsforsuccessfulMonit
orusewer enotmet
(condi
tiont hr
ee=knowt her ule,wasmet ,butone=t i
me, andtwo=focusonf orm, werenot).
Thiscasei sex actl
yanal
ogoust ot hede+le=ducasedescr i
bedabove.Thesethreestudents,
Iwouldpr edict,wouldperfor m wellont hisi
tem underdi f
fer
entconditi
ons,i.
e.ifgivena
discr
ete-pointgrammart estthatf ocussedthem onf
orm, contai
ningi
temssuchas:

Final l
y,andwhati sofmosti nteresther e,twochi ldrenandoneadul tper f
or medwel lon
thepi cturei denti
fi
cati
ont estbutproducedi ncorrectr ules( e.g." Youuseanf orsomet hing
that'
sal iv e").Thechi l
dsubj ectsarer emi ni
scentofot hercasesi nt hel it
eratureandmay
simplyr ef lecttheinabi
li
tyofchildr
ent oi nduceorl ear ncor r
ectconsci ousr ules(e.g.achi l
di n
onest udy ,acqui ri
ngFrenchasasecondl anguage,deci dedt hatf emininegenderwasf or
"everythingt hatwasgoodandbeaut iful"(Keny eresandKeny er es,citedinHat ch,1978b) .The
adultwhoper for
medper fectl
yonthet estmaybecl assifiedasan" i
ncompet ent"Moni toruser.
Thissubj ecthadacqui r
edt hea/anrule,buthadnotl earnedi tcor rectl
y.Thef actthathedi d
notappl yhi sconsciousrulet oper
formancewor kedt ohi sbenef i
t!Iwoul dpr edictthatsucha
subjectwoul dperform worseonat estt hatfocussedhi m onf or m.( Idonotwi sht oimpl ythat
somel ear ner sgetallr
uleswr ongwhi l
eot her sgett hem al lr
ight .Clearly
,manyl earnershav e
l
ear nedsomer ul
escorrectlyandsomei ncor rectl
y .
)

Whati sremarkabl
eher ei
sthatthissubj
ecthadf ai
l
edtolear
nwhatmostt eachers
woul
d considerto beanamazi nglysimplerul
e,yethehadappar ent
lyacqui
red it
.This
i
ll
ust
rat
est heindependenceofacqui
sit
ionandlear
ning,aswel
lasjusthow l
i
mit
edl earni
ng
canbeforsomeper for
mers.2

(
b)Rul
elear
nabi
l
ity

Weseef ai
rl
ywi
deindivi
dualvar
iat
ioni
ntheabi
l
ityt
ouset
heconsci
ousMoni
tor
.The
r
angegoesf
rom t
hepr
ofessi
onall
ingui
st,
who

96
96
maybeabl etoconsciouslyl
earnmanyr ulesofgr eatcompl ex
ityandevenappl ythem whi
le
per
for
mingi nasecondl anguage,t
oMoni torunder -
usersandi ncompet
entwr ong-
ruleuser
s.
Despi
tethi
sv ari
ati
on,wecanbegi ntochar acter
izet hekindsofr ul
esthatarelearnabl
efor
mostadult
s,recogni
zingthatevenforsuperMoni torusers(seee.g.Yor
io,1978),t
hissetwil
l
beasmallsubsetofthetotalnumberofrul
esi nalanguage.

Thepr ofessionallit
eraturesupportswhatIt hi
nkisthemostr easonablehypothesi
s:
"Learnabil
it
y"isrelatedt oli
ngui st
icsimpli
cit
y,bot
hf or
malandf unct
ional.Ther ul
eswecan
l
ear nandcar ryar oundi nourheadsf oruseasaMoni t
orarenott hoset hatareearli
est
acquired,noraretheyt hoset hatareimport
antforcommunicat
ion.Rather,theyarethesimple
rules,r
ulesthatareeasi esttodescribeandremember .

Ihavepointedoutelsewherei
notherpubli
cations(f i
rstdi
scussedinKrashenetal.
,
1978)t hatsimplicit
ycanbedef i
nedinatleastt woway s,andbot hdefini
ti
onsarerelevant
here.First
,arulecanbef ormal
lysi
mple.Exampl esofr elati
vel
ysimpler ul
esincl
udeourol d
fri
endst hethir
dper sonsingul
arendi
ngonr egularv erbsinEnglish,andde+l e=du.These
rul
esrequireonlyt headdit
ionofaboundmor pheme( aninfl
ecti
on),orcontr
acti
onoperati
ons.
Simpledeleti
oni salsoprobabl
ynotdi
ffi
cul
tfortheconsci ousMoni tor
.

Othersy ntact i
coper at
ionsappeart obemor ediffi
cul tfortheMoni tor.Permut ations,
andmov ement sofconst i
tuentsf rom onepar tofasent encet oanotherarequi t
edi f
ficultt o
do" i
ny ourhead"whi leinthemi ddleofaconv ersationorev enwhenwr it
ingf orcontent .Itis
probablythecaset hatrul
esr equiringagr eatdealofmov ementandor derchangear eei ther
acquir
edorar enev erdonewel lbymostpeopl e.Thisappl iest orulessuchasf or
mationofwh-
questi
onsi nEngl ish,whichmi ghti nvolvet
hef oll
owi ngsepar at eoperati
ons:(1)placi
ngt hewh
-wor df ir
st;(2)per for
mingsubj ect -
auxi
l
iaryinv ersion,unl esst hereisahel pi
ngv er b;( 3)
perfor
mi ng" do-suppor t
";(4)inflecting"do"cor r
ectlyf ort enseandnumber .Thisisal ott o
remember ,especi allywhent hel earnerhasot hert hingsonhi smi nd,i
ncludingremember ing
otherpar t
sofgr ammar( hemi ghtal sobeMoni toringpr onunci ati
onaswel lassyntax )and
try
ingtokeepupaconv ersati
onwi thanativespeaker .

(Ther
eadermayar
guet hat(
s)hehasnopr
obl
em doi
ngal
ltheset
hingsatt
hesame
t
ime,
andwit
halit
tl
epr
act
iceandgoodteachi
ng

97
97
ever y
oneel secanaswel l.Ifthisiswhati sgoingt hr
oughy ourmi nd,y ouar eprobabl ya
Moni torsuper -user.Thissor tofi nt
erestandabi l
i
tymaybewhatbr oughty oui ntol anguage
sciencei nt hefir
stplace, andgoty ouinterestedinbookssuchast hisone.Youar enott ypical
.
Otherr eader smayar guet hatthewayt omaker ulessuchasquest ionf ormat i
onaut omat i
cis
tolear nanddr il
lthecomponent soneatat imeunt iltheybecomeaut omat i
c.Thi sisex actl
y
the" learningbecomesacqui siti
on"argumentr evi
ewedatt hebegi nni ngoft hissect ion.I
mai ntaint hati n caseswher ethisseemst o wor k,oneoft wo t hingsi shappeni ng:( 1)
acqui si
ti
oni soccurri
ngsepar atel
yandcat chesupt ot hestudent'
slear ningl evel;t
hel earning
thatpr ecededt heacqui siti
ondi dnotpl ayanydi rectrole,moreov er,i nhel pi
ngacqui sit
ion
dev el
op.( 2)Thesuccessf ul l
earnerwasasuperMoni t
oruserandv eryat ypical.)

Simpli
cit
yal soneedst obedef inedi nt ermsofmeani ng.Rul est hatar eformal
lysimple
willnotbeeasil
yl earnableiftheirmeani ngsar esubtleandhar dtoex pl ain.Boththeform and
meani ngofthethirdper sonsi ngularendi ngandde+l e=duar est r
aight forward.Ontheother
hand, whil
ethefor m ofthedef i
nit
eandi ndef i
nit
earti
cleinEngl i
shisv er ysimple,manyoft he
usesofa and t he areenor mousl ycompl ex .Wecer t
ainl
ycannotex pectESLl earnersto
understand,r
emember ,andconsci ouslyappl ydescripti
onssuchast hatcont ai
nedinHawki ns
(1978),a fullvolume dev ot ed tot he ar ticl
ei n Engli
sh.We can al so fi
nd exampl es i
n
punctuati
on.Rulessuchas" capit
ali
zet hef ir
stletterofev erysent ence"ar ef or
mal lyand
functi
onall
yeasy .Someoft heusesoft hecommaandsemi -
colon,howev er,aredif
ficul
tto
describeandprobabl yneedt obeacqui redf oreffect
iveuse.

(
c)Someev
idence

Ther
earenost udiesIknow ofthatdir
ectlyprobewhichst ructuresarelear
nableby
diff
erentst
udentpopul at
ionsandwhicharenot .Severalpapers,howev er,pr
esentevidence
thatisqui
teconsi
stentwiththecl
aimthatonly"easy"r
ulesarelearnablebymostpeopl e.

Onesortofevi
denceisprovi
dedbyst udi
esandcasehi
stor
iesthattel
luswhatsort
sof
"car
eless"err
orssecondlanguagestudent
smake,error
sthatinv
olverulesthatt
hestudent
s
hadformallystudi
edandthat

98
98
theycoul dself
-correct,giventimeandwhenf ocussedonf or m.Inourterms,t hesearer
ules
thathav ebeenl ear
nedbuthav enotbeenacqui r
ed.Theyar e,inallcases,whatappeartobe
l
at e-
acquiredandf or mallysimpl erul
es,inv
olvi
ngmost lyboundmor phology.P,t heopt
imal
Moni toruserwedi scussedear l
ier,mademanyer r
orsonsucheasyi t
emsast hethir
dperson
singularendingonr egularverbs,theuseof" much"and" many "withcountandmassnouns,
andt heirregul
arpast ,amongot hererror
s.Twoopt i
malMoni torusersdescribedinCohenand
Robbi ns(1976)alsomadewhatt heyt
hemselvescall
ed"careless"err
orsonsuchi t
ems.

Bot hUe- LinandEv a,ChinesespeakingESLst udent satUCLA,hadpr obl emswi t

ht he
l
at e-acquiredt hir
dper sonsi ngular/s/
.Ue-Linex pl
ainedt hi
somi ssion" asacar elessmi stake
sincesher eportedknowi ngt her ul
e"(CohenandRobbi ns,1976,p.55) .Si mil
ar ly,"whenEv a
wasshownsent encescont ai
ningsdel eti
on,shewasact uall
yabl et oi dentif
yt heer r
orand
suppl ythesi mmedi ately.Whenaskedt oex plainwhysheomi ttedt hessher eplied:' Probably
j
ustcar eless.'
"(p.58).Ev ahadasi milarexplanat i
onf oromi tt
ingt her egularpast/ ed/ :When
present edwi thoneofherer r
orsont hisform,al soknownt obel ate-acquired( Hy pot hesis3,
Chapt erI I)
,shewasabl et osuppl ythecor rectf orm.Af tercor recti
ngonesent ence." Eva
remar kedt hatshewr ote down t hesent encet hewayshewoul dsayi t:"Foronet hi
ng,
somet i
mesIwoul dwr i
tesomet hi
ngthewayt hatIspeak.Wesayawor dmor eorl essina
car el
essway .Buti fIt
akemyt i
me, sometimesgoov erit
, t
hatwoul dbemucheasi er...
'
."(p.
58) .Myi nterpretati
onist hatgiventime,Evawasabl etoaccessherconsci ousknowl edgeof
Engl i
sh, orMoni tor,aprocedur ethatcanbeef fect i
veforsuchl ate-acqui r
ed, si
mpl er ul
es.

Ev
ahadhadaf airamountofex posuretoEnglishandwasconsi deredt obeadv anced
byCohenandRobbi ns.Shehadl iv
edi nAustr
ali
af ort woandahal fyearsandhadst udied
Engli
shsincegrade5( atthetimeofCohenandRobbi ns'studyshewasaj uni
orincol lege)
.
Shealsoconsideredher sel
fa" goodlanguagel
earner".EvenEv a,however,
hadpr oblemswi th
whatseem t olanguaget eacherstobesi mplerules.Inex plai
ninghererror("Ihavet alkto
Syl
viaalr
eady")sheattri
butedtheer r
ortobeingunclearaboutt herul
e.Sher emarked:"Yeah,I
l
earnedthat.I
t'
sjustsomet hingI'
m notgoodat.It
hinkt hemai n

99
99
problem i
sthatIjustl
earnt
herule-
-oneort
woyear
s,thewhol eti
meIwasgoingt oschool.
..I
t
wasnev erdril
lenoughtome"(CohenandRobbi
ns,p.58).Thisconfi
rmsapointmadeear li
er,
andshowst hatwhil
eMonitorusemaybel i
mit
edtonon- acquir
ed,si
mpl
erules,even"good"
l
earnersmaybeabl etouseandr ecal
lonl
yasmallpartoft herul
eswepresent,eventhose
thatseem tr
ansparentt
ous.3

Ast udybyDuskov a( 1969)al soconf i

rmst hatt hesy ntacticdomai noft heconsci ous
Moni tor,f ormanypeopl e,consi st sofr elativ
elysi mplebutl at e-acquiredi tems.Duskov a
i
nv estigat edwr i
tt
en er rorsi n 50 Czech uni versit
yl evelst udent sst udy i
ng Engl i
sh (EFL) .
Duskov anot edthat"...manyoft her ecurrenter rors..
.ref l
ectnor ealdef icitinknowl edge,since
mostl ear nersknow t heper ti
nentr uleandcanr eadilyappl yit,butt hemechani caloper ation
doesnoty etwor kaut omat i
cally"(p.16) .Thi sgener al
izati
onappl i
esi npar ticul ar,Duskov a
notes,t omor phologi
caler rors.Ex ampl esi ncludet heomi ssionofpl uralsonnouns( rel
ati
v ely
earlyacqui redamonggr ammat i
calmor phemes, Imustadmi t)
.Duskov anot est hatf orplur als
".
..thel earneri sawar eofi twheni tispoi nt
edoutt ohi m andi sabl et ocor recti thimsel f
"
(p.
20) .Anot herex ampl ei ser r
orsi nsubj ect -
ver bagr eement .Agai n,fort hiser ror ,"whent he
l
ear ner'sat tent i
oni sdr awnt othef actt hathehasmadeami stake,hei susual lyablet o
correcti t
"( p.20) .Ot herer rort ypesoft hissor ti ncludeconf usionofpastpar ti
cipleand
i
nfiniti
v e,error sonirregularv er
bs, andadj ect i
ve-nounagr eementi nnumber( e.g.t hiswor ker s).
Inourt er ms, t
heer r
orsr efl
ectaf ailuretoappl yconsci ousr ules,af ailur
et oMoni t
oref f
ecti
v ely.
Thest udent s,Duskov at el
lsus, "
cancer tai
nl yformul atet herule"fort heseer r
ort ypes.

Themor phemest udi

esdescribedinChapt erIIalsocontri
butetothi
spoint.Asy oumay
recal
lf rom Chapt erII
,changesordi st
urbancesi nthe"naturalorder
"wer einterpr
etedas
i
ntrusionsoft heconsciousgrammar .Itisinter
estingtonotejusthowt heorderwasaf fected.
I
nLar sen-Freeman' sst
udy(Larsen-
Freeman, 1975) ,morphemeor derswerepresentedforboth
monitor edandunmoni t
oredconditi
ons( adiscrete-poi
ntpencilandpapergr ammart est,and
theBilingualSy nt
axMeasur e,respectiv
ely).IntheMoni t
or-f
reeconditi
on,Larsen-Freeman
obtai
nedt hef ol
lowingorder(Tabl
e4. 1)whichisqui te"
natural
".

100
100
100
 Tabl
e4.
1

Mor
phemeor
derobt
ainedi
nmoni
tor
-f
reecondi
ti
on(
Lar
sen-
Freeman,
1975)

i
ngcopul a
arti
cle
auxil
iary
shortplural
regularpast
thi
rdper sonsi
ngul
ar
i
rregularpast
l
ongpl ural
possessi v
e

Compar
ethi
stotheunnatur
alorderf
oundint
heMonitoredcondit
ion(
Table4.
2).(Weuset
he
wri
ti
ngt
askasanexample;Lar
sen-Fr
eeman'
s"r
eadi
ng"taskgivessimi
l
arresul
ts.
)

Tabl
e4.
2

Mor
phemeor
derobt
ainedi
nmoni
tor
edcondi
ti
on(
Lar
sen-
Freeman,
1975;
wri
ti
ng)

copula
auxil
iary
thi
rdper sonsi
ngul
ar
i
ng
regularpast
i
rregularpast
arti
cle
l
ongpl ural
shortplural
possessi v
e

Theseordersdifferl
argelyduetot heincreasei nrelat
iverankoftwomor phemes,regul
arpast
andthethirdpersonsingularmar ker
,bot hlate-acquir
ed,orlowinrel
ati
veorderofaccuracyin
the Monit
or -
fr
eenat uralorder.Thisinterpretati
on isconsist
entwit
ht he claim thatwhen
perf
ormersf ocusonf ormt heycani ncreaseaccur acyi nunacqui
redbutl ear
nedpar t
sof
grammar.4

Sti
llmor eevidencecomesf r
om ourcompositi
onstudy(Krashen,Butl
er,Bir
nbaum,and
Robertson,1978) .We asked ESL st udents atUSC to wr i
te compositi
ons undert wo
condit
ions--
"free"(i
nstruct
ionswer etowrit
easmuchaspossi bleinfiv
emi nutes)and"edi
ted"
(i
nstr
uct i
onswer et opaycar efulatt
ent
iontogr ammarandspel l
i
ngandt o" t
akey ourti
me").
Bothcondi t
ionsy i
eldednat ur
alordersforgrammaticalmorphemes,whichwei nt
erpr
etedas
i
ndicatingli
ttl
einterventi
on

101
101
101
oft heconsci
ousMoni t
or.Thiswasduetot
hefact
,we hy pot
hesi
zed,t
hatoursubj
ects
focussedprimari
lyoncommuni cat
ioni
nbot
hcondi
tions,despi
teourinst
ruct
ionst
ot he
contrar
yinthesecondcondi
ti
on.

Closeranal ysi
sofourdat adoesshow somer iseint het hir
dper sonsi ngularint he
edit
ed condi t
ion,howev er(we di d notanal y
ze r egularpastdue t ot oo few obl i
gatory
occasions).Thisrisewasnotenought odist
urbthenat uralorder,butenought osuggestsome
Moni t
oruse.Agai n,weseet hedifferencesinthel ate-
acquired,easyi tem.( Toinjectamor e
theoret
icalpoint,perhapst hecorrectinterpr
etat
ionofmor phemenat uralandunnat uralorders
i
st hatunnaturalor ders,asinLar sen-Freeman( 1975)r efectheav
l yMoni toruse.Increasesi n
certai
nitemswi t
houtchangesi nr ank,asi nourcomposi t
ionst udy,mayr ef
lectli
ghtMoni tor
use.)Table4.3shows t hissmallimpr ovementint hethirdper son singularmor phemei nt he
edit
edcondi ti
on.56

Tabl
e4.3
Accur
acydi
ff
erencesi
nfr
eeandedit
edcondi
ti
onsf
orgr
ammat
icalmor
phemes

Mor
pheme Fr
eeI Edi
ted I Fr
eeI
I Edi
tedI
I

i
ng 0.
87 0.
85 0.
88 0.
82
copula 0.
79 0.
95 0.
86 0.
85
plural 0.
82 0.
82 0.
77 0.
78
arti
cle 0.
86 0.
85 0.
76 0.
83
auxili
ary 0.
82 0.
79 0.
77 0.
76
i
rregularpast 0.
69 0.
81 0.
82 0.
77
thir
dper sonsi
ngul
ar 0.
54 0.
61 0.
32 0.
65

Free:"writ
easmuchasy oucan"i nfi
vemi nut
es.
Edited:"paycaref
ulat
tenti
ont ogrammarandspel l
i
ngandt akeyourti
me".I
:samesubj
ect
s(n=58)
preformedbot hcondi
ti
ons.
I
I:differ
entsubject
sforeachcondi t
ion.
Eachmor phemewasr epresentedbyatleast100obli
gat
oryoccasions.

(
d)Consequencesoft
eachi
ng"
har
d"r
ules

Feli
x(1980)showsuswhathappenswhenst udentsareaskedtolear
nrulesthatare
toodi
ff
icultf
orthem,rul
esthatar
enotonlydi
ff
icul
ttolear
nbutthatareal
sonotyetacquir
ed.
Notonlyweresuchstudent
saskedt ol
ear
ndif
ficul
trul
es,theywereal
soaskedtousethem in
unmonitor
edsit
uati
ons.Fel
ixobserv
edanEFLclassfortenandeleven

102
102
102
yearold studentsin Germany.Among hismanyi nteresting observat
ionswast hisone:
Teacher
st aughtanddemandedcor r
ectuseofel
lipt
icsent ences( asinexchangesoft het
ype:
I
si tadog?Yes,i tis).Despi
tethefactthatthistypeofquest ion-
answerdi al
ogue"was
i
ntensi
vel
ydr i
l
ledeveryday "(
p.8),Fel
ixrepor
tst hatcor rectel l
i
pticsentenceswer eonly
randoml
ysuppl iedf
oraperiodofal
mostthreemonths( i
.e.Itisadog?Yesi ti
sn't
)!

Thi
sresul
tisqui t
epredictable:therulewassi mplytoohar dtolear
nandwasnoty et
acquir
ed.Fel
ixnotesthataccor ding tot he research l
iter
atur
e elli
pti
csentences "
do not
appearunti
lrel
ati
velylat
e"(p.9).Evenwi thinputcontai
ningsentencesofthi
ssor t(
assuming
theinputwascompr ehensi
ble,interesti
ng,etc.
;seeChapt erI
II)
,suchst r
uctureswerefar
beyondthei+1oft hesestudents.

Fel
i
xal sor eport
st hatteachersv
ali
ant
lyt
ri
edtoteach do-
supportandtheEnglish
negati
onr ul
eswi t
hl i
tt
lesuccess.Theseareal
soqui
tedi
ff
icul
t.St
udents,Fel
i
xfound,would
producesentencesliketheseinclass:

(
1)I
t'
snomycow.
(
2)Doesn'
tsheeatappl
es.

Bothoft hesesentencesar eint

erpretabl
easr el
ianceonwhathasbeenacqui r
edwithoutthe
contri
buti
onoft heconsci ousgrammar .Tofull
yappr eci
atet hesigni
fi
canceoftheseerrors,
wef ir
stneedt obr ief
lyreview whati sknow aboutt heacqui si
ti
onofnegat i
onininformal
l
anguageacqui si
ti
on( seealsoChapt erII
).Thefoll
owingst agesarefoundinchil
dL1,chi
ldL2,
andadul tL2acqui sit
ion(whatf ol
l
owsi sasi mpli
fi
cation;seeDul ay,Bur
t,andKrashen,in
press,f
ordetai
ls):

I
.Thenegat
ivemar
kergoesout
sidet
hesent
ence,
asi
n:

nowipefi
nger
wearmit
tenno(exampl
esf
rom Kl
i
maandBel
l
ugi
,1966)

I
I.Thenegat
ivemar
keri
spl
acedbet
weent
hesubj
ectandv
erb,
asi
n:

Henobi
teyou
Henotl
i
ttl
e,hebi
g

103
103
103
 II
I.Postaux
il
iar
ynegat
ioni
sacqui
red;
themar
kernowappear
saf
tert
heaux
il
iar
yver
b,
asi
n:

Thatwasnotme
Idi
dn'
tcaughti
t

Fel
ix'
sex ample( 1)appearst obeast ageIItypetransit
ionalform.This"error"isat ypical
i
ntermediatestageal lacqui
rers(ornearlyall
)got hroughbef orefull
yacquiri
ngt hecor r
ect
for
m.Theappear anceofsuchaner r
ort ypei
sthusconsi stentwit
ht hehypothesisthatthese
chi
ldren,ev
enthought heyar
ei naclassroom,ar eundergoingnormallanguageacqui si
ti
ont o
atleastsomeex tent
,andarer el
yingonacqui redlanguagei nclassroom speech( notethat
Germannegat i
oni salwayspostverbalandpostaux i
li
ary).

Sentence( 2),accordingtoFel i
x,isnotay es/ noquesti
on!Felixmaintai
nst hati tis,
i
nstead,anegat ivedeclaration("
Shedoesn' teatappl es"
).Thus,asisthecasewi thsent ence
(1)
,Felixinter
pretsthiserrorasat ransi
tionalform,t hisonebeinganex ampl
eofst ageIwi th
doesn'tact i
ngasamonomor phemi cnegat ionmar ker.(Iti
squi tecommonf ordon' tt o
perf
ormt hesamef uncti
oni nst
ageI Iinnat uralfir
standsecondl anguageacqui siti
on,e. g.
sentencessuchas" Idon'tcanex plai
n"wher e"don't"actsast henegat i
vemarker ;see,f or
example,Cancino, RosanskyandSchumann, 1974) .Thechild'
sselect
ionofdoesn't( i
nsteadof
no)i sduet ot hepar ti
cularkind ofi nputpr esented int heclassroom,thegr ammat ical
exer
cisesi nwhichdoesn' tappearsinveryhighf requencies.

Such i
nterpret
ati
ons notonly pointtothe real
it
y and strengt
h ofsubconsci
ous
acquisi
ti
on,buttheyalsoconfi
rmthatconsciousl
earni
ngisquiteli
mi t
ed,andthat
,ex
ceptf
or
cert
aincondit
ions,acqui
sit
ioni
sresponsiblef
ormostsecondlanguageperfor
mance.

C.TheEf
fect
sofLear
ning:Accur
acyofSel
f-
cor
rect
ion

Prev
ioussect
ionsofthi
schapterhavedi
scussedwhenperfor
mer sMonit
orandwhich
rul
esareusabl
eforMonitori
ng.Weturnnowt othequest
ionofhowef fect
iveMoni
tor
ingis:
Howmuchcant hesecondlanguageperf
ormeri
mproveaccuracybyconsult
ingt
heconsci
ous
grammar?

104
104
104
 Wecangetsomeappr oximati
onoft heef f
ici
encyoftheconsci ousgr ammarbylooking
athowgoodper f
ormersareatself-
corr
ectionoft hei
rownl i
nguisti
cout put.NoelHouckhas
poi
nted outto me thatself-
corr
ecti
on (as opposed t o"ot her
-correct
ion",orcorr
ecting
someoneelse'
soutput,anacti
vit
ythatincludesdet ect
ingerrorsont est),isthemostv ali
d
obj
ectofstudyininvesti
gat
ingMoni t
orstr engt
h,sincethi
si swhatone' sMoni t
oractuall
y
doesinr
ealper
formance.

Sever
alstudi
eshaveexaminedhowef
fecti
vesel
f-
cor
rect
ioni
s,butbefor
esur
veyi
ngthe
dataanddrawingconclusi
ons,weneedtobr i
efl
ylookatsomeoft hef act
orst
hatcause
accur
acyofself
-cor
rect
iont
ov ar
y.

1.FACTORSAFFECTI
NGSELF-
CORRECTI
ONACCURACY

Fi
rst,
aswecani nferf r
om thediscussioninChapt erI
I,thereisindiv
idualvar
iat
ionwith
respecttoself-correctionef f
ici
ency.Tot heex t
entt
hatsel f-
correcti
oni nv
olvestheconscious
Moni t
or,i
fther eisi ndi vidualv ar
iat
ionindegr eeofandabi li
tyf orMonitoruse,thiswillbe
refl
ectedin self-
cor rection ef f
ici
ency.Wemi ghtexpectmuch bet t
erper for
mancef r
om a
professi
onallinguistwhoi sanopt i
malMoni toruser(e.g."P",from KrashenandPon,1975) ,
thanfrom otherper former s,all
othercondit
ions(seebelow)hel dconstant.

Second, wemi ghtexpectvari

ati
ondependi ngonwhi chaspectsofout
puttheperformer
attempt st ocor r
ect
.Aswedi scussedabov e,t
heMoni t
orappear stoworkbestf orsimple
mor phology ,maybel esseff
ici
entforcompl exsyntax,andmayhav eevenmoretroublewith
otherpar tsoft hegrammar( t
hereis,unfortunat
ely,notevenenoughdat atospeculat
eabout
thel earnabili
tyofmuchoft hegrammar ;seeChapt erIIIforav erybri
efdi
scussionoft he
l
ear nabili
tyofaspectsofconv ersati
onalcompet ence).

Self
-corr
ecti
oneffi
ciencywi llal
sov ar
yaccordingtothecondi ti
onsunderwhichitis
done.Houck,Rober tsonandKr ashen( 1978b)dist
ingui
shedt hefoll
owingconditi
ons.Fi
rst,
thereis"freespeech",ornat ur
alconv ersati
on.(
Inonesense," fr
eewr i
ti
ng"bel
ongsinthis
category,in anot
hersensei tdoesnot ,assomewoul d arguet hatthewr i
tt
enmodal it
y
automaticall
yentai
lsagr eaterfocusonf orm.)I
n"freespeech",self-
corr
ecti
onisupt othe
perfor
mer ,andther
eisnospeci alfocuson

105
105
105
for
m.Rather,t
hef
ocus,i
nmostcases,isoncommunicati
on.Wewouldexpectnatur
al
dif
fi
cul
tyorder
swhenlooki
ngater
ror
singrammat
ical
structur
esi
nthi
scondit
ion.

Asecondcondi ti
on,movinginthedi rect
ionofmor ef ocusonf orm,canbet er
med
"car
eful"speaki
ngorwr i
ti
ng.Thi
sisroughlyequival
entt
ot heeditedcondit
ioninKrashenetal
.
(1978),descri
bedearli
er,andoccurswhenspeaker sorwr i
tersareattempt i
ngtospeakor
writ
e" corr
ectl
y".(
Wemustsubdi v
idethi
scondi t
ionintotwosub- condit
ions,oneforcaref
ul
speakingandonef orcar
efulwri
ti
ng,pr
edicti
ngmor eself
-corr
ectionforwrit
ing.
)

Conditi
ons( 1)and(2)covermostsi t
uationsi
nrealwor l
dinformalcommuni cati
on.We
can,however,specifyothercondi
tionstypical
l
yusedi nlanguagei nst
ruct
ionthatfocust he
perf
ormersti
llmor eonf orm.I
ncondi t
ion(3),t
hestudentisinf
ormedt hatanerrorexist
s,but
doesnotknowwher et heerr
orisorwhatr ulehasbeenbr oken.Thisisroughl
yequi v
alentto
composit
ioncor r
ectioninwhichstudentsar esimpl
ytoldthattherearesomeer r
orsi nthei
r
paperandthattheyshouldbecor r
ected.

Aconditi
onsti
llmor ei
nthedi r
ecti
onoff ocussi
ngonf ormi ndicatestost udent
swhere
theer rori
s,i
nadditi
ont oinfor
mi ngthem thataner r
orex i
sts.Thiscondi ti
on,conditi
on(4)
,
correspondstocompositioncorrect
ioninwhi chtheteacherunder l
inest heerrors.St
il
lmore
focussed,accordi
ng to Houck etal .,is condi
ti
on (5)in whi ch existence,locati
on,and
descripti
onofthevi
olat
edr ul
eareprovided,asinfeedbackofthissor t
:

Themorewemov et owardcondi ti
on(5),themor eeff
ectoft heconsci ousMonitoris
predi
cted,andthegreatertheli
kelihoodof" unnatural
"orderf
orer r
ors.Accordingtoresearch
summar izedinChapterI Ianddi scussedagai ni nthischapter,weseenat uralorder
sf or
condit
ions(1)(freespeech,BSM,f reecomposi t
ion)and( 2)(edit
edwr it
ing),butmightsee
someef fectoftheMoni torincondi t
ion(2)(
i.
e.r i
seinaccuracyi nthir
dper sonsingularin
Krashen etal.
,1978) .Larsen-Fr
eeman' sunnaturalorderwaspr oduced undercondi t
ions
si
mi l
artocondit
ion(4)(butseeNot e4, thi
schapter).

Tabl
e4.
4summar
izest
hef
ivecondi
ti
ons.

106
106
106
 Tabl 4Sel
e4. f-
cor
rect
ioncondi
ti
onsi
nsecondl
anguageper
for
mance

I
nst
ruct
ions: (
1) (2) (
3)Corr
ect (
4)Cor
rect (
5)
None Rewri
te t
heerr
or thi
ser
ror Correctthi
ser
roruse
thi
sr ul
e

I
ncludeserror
:
Exist
ence No No Yes Yes Yes
Locati
on No No No Yes Yes
Rul
ebr oken No No No No Yes

(1)Fr
eespeechorwr i
ti
ng.(
2)Careful
speechorwri
ting
From:Houck,Rober
tsonandKrashen(1978b)
.

107
107
107
 Ther ear esomest udiesav ail
abl ethatgiveusani deaoft heef fi
ci encyofsel f-correct
ion
forsomeoft hecondi t
ionsdescr ibedher e,thatt elluswhatper centageofper for
mer s'err
ors
areact uall
ysel f-
cor r
ected.Theyt hust el
lussomet hingaboutMoni toref fi
ciency,sincet hey
reporttowhatex tentaper former '
ssel f-
correcti
onsi mpr oveout putaccur acy.Inonesense,
howev er,theymaynott r
ulyshow t heef f
ect i
venessoft heMoni tor .They under estimate
Moni t
oruse,si ncet heydonoti ndicatecov ertsel f-
cor r
ection,t hecor rectionthatwenton
beforet heutterancewasspokeorwr it
ten(Recal l,infigureone,Chapt erI I
,thatther earetwo
possiblear r
owsl eadingf r
om t heMoni tortotheout putoft heacqui redsy stem,oneaf fect
ing
outputbef oreandoneaf f
ectingout putaf t
erpr oduct ion.)Ont heot herhand,st udi esthat
reporttheper centageofsuccessf ulsel f
-correct
ional soov erestimat et heamountofact ual
consciousMoni t
oruse,si ncesel f-correcti
oncanal sobedoneusi ngt heacqui redsy st
em
alone,wi t
hone' s" f
eel"forcor r
ectness.Thi siswhatper formersdoi nt heirf i
rstl
anguagewhen
correcti
ngsl i
psoft hetongue.

Sti
llanotherproblem ofinterpr
etat
ionofsuchstudiesi
sthatwedonotr eal
lyknow
whethersubj ect
shadi ndeedhadt hechancetolear
nalltherul
esnecessar
yforsuccessf
ul
sel
f-
cor r
ection.Arewest udy
ingtheeffici
encyofl
earni
ngand/ortheabi
li
tyofper
for
mer sto
appl
ywhatt heyconsciouslyknow?

Sel
f-
corr
ecti
onstudi
esdonotpr ovi
deus,t her
efor
e,withanex actpi ct
ure,butthe
r
esul
tsarequit
eusefult
otheteacherint
erest
edi
nt heoveral
lef
fi
ciencyofself
-corr
ecti
on,and
t
heyprobabl
ygiveusanapproxi
mat i
onoftheeff
ici
encyofconsciouslear
ningandMoni t
ori
ng.

THEDATA

Table4.5andFi g4.1summar i
zet hel i
terat
ureavail
abletomeonsel f
-corr
ecti
on.With
two exceptions,al ldealwith uni
versi
tyl evelESL st udents who,we ex pect,have been
exposedt oaf airamountofformalinstr
uctioni nEngli
shgrammar .Thesubjectinonestudyis
ouroldf r
iend" P",ali
ngui
st.Fathman'ssubj ects(Fathman,1980)ar edescr
ibedas20adul ts
"l
earni
ngEngl ishasasecondl anguageint heUni tedStates,pr
imaril
yinani nfor
malsetti
ng"
and20adul ts"learni
ngEngli
shinaf ormal setti
ng,pri
mar i
l
yinMex ico"(
p.3,manuscript
).

108
108
108
 Tabl 5Accur
e4. acyofsel
f-
cor
rect
ioni
nadul
tper
for
mancei
nEngl
i
shasasecond
l
anguage
St
udy Condi
ti
on1 Er
rort
ypeanal
yzed Resul
ts
(
%ofer
ror
ssel
f-
cor
rect
ed)

1. Schl
ue(
1977) Stream Al
lsyntax, 7.
2%(
99/
1101)
ofspeech morphology
(1)
2. Fathman( 1980) SLOPEt est, Mor
phol
ogy
2 a."Infor
mal"adult
s
oral (seetext
)=20%
interv
iew, (13/65)
Picture b."Formaladul
ts"
description =32%(46/144)
(1)
3. Schlue(1977) Listent o Al
lsyntax, 31%
tapeofown morphology
speech( 2)
4. Houcketal . Inspect Ni
nemor
phemes 17.
5%(
36.
5/236)
(
1978a) t
ranscr
ipt
ion
ofownspeech3
(2)
5. White(1977) Inspect a. Morphology a. 52% (
53/102)
transcriptof b. Syntax b. 27% (
6/22)
responsest o c. "Omissi
ons" c. 53% (
23/43)
BSM4( 4) d. Lexi
cal d. 9% (1/11)

47%(
83/
178)

6.Kr
ashenand I
nspect Mor
phol
ogy
, 95%(
76/
80)Pon(
1975) t
ranscr
ipt
ions5
synt
ax
(
4)

1 Numberi
npar
ent
hesi
sref
erst
ocondi
ti
onsi
nTabl
e4.
4.

2 "
Almostallt
heuncorr
ectederrorswerer
elat
edt
over
bs,
suchas:
omi
ssi
onoft
hecopul
aand
omi
ssi
onofincorr
ectuseofinf
lecti
ons"
.

3 Subj
ect
str
anscr
ibedt
apest
hemsel
ves.

4 Et
ranscr
ibedt
apes(
".
..
(S'
s)wer
epr
esent
edwi
thsomeoft
hei
rer
ror
s")
.

5 E=Subj
ectt
ranscr
ibedt
ape.

109
109
109
 Fi
g.4.
1.Sel
f-
cor
rect
ionaccur
acy
.

1:Schlue(1977).
2a,2b:Fathman( 1980).
3:Schlue(1977).
4:Houcketal .(
1978a).
5:White(1977).
6:KrashenandPon( 1975)
.
(
seeTabl e4.5)

Asfort
hedomai
noflanguageanal
yzed,onest Houcketal
udy( .,
1978a)f
ocussesonl
y
onni
negrammati
calmor
phemes,whil
etheotherscov
ermorphol
ogyandsynt
axi
ngeneral
.

Iti
shardt osaywhetheranyofthestudi
esact uall
ymeett hedescri
ptionofcondi
tion
(
1),si
ncei nallcasesthesubject
sknew t
heywer ebeingtestedandt hatthefocusoft he
i
nvest
igati
onwast hequali
tyandaccur
acyoftheirspeech.Thus,Fat
hman( 1980),andSchlue
(
1977)maybeov erest
imat
esofself-
communi cati
onaccuracyi n"f
reespeech"andmi ght
r
eall
ybelonginconditi
on(2)
.

Figur
e4.1att
empt st
oill
ustr
atehowconditi
ons,anddif
fer
encesi
nsubjects,
aff
ectsel
f-
correct
ionaccuracy
,andgivesusapi ctur
eofwhatwecanex pect
,atl
eastinthedomainof
syntaxandmor phology
.Itsuggests,f
ir
stofall
,thatt
raini
ngandtypeofstudentdomakea
dif
ference:
110
110
110
Fathman' s"
f ormal"studentscor r
ectahigherpercentageoftheirownerrorsthanherinformal
students do,and oursubj ect,"P"
,outperformsev eryone.Itisalso consistentwit hthe
hypothesisthataccur acyincreasesaswef ocusmor eonf orm.Furt
herstudiescouldeasilybe
performedt of il
linthehol esint hefi
gure;theywoul dalsohopeful
lycont r
olforprofici
ency
l
ev el
,sincet heremaybear el
ati
onshi
pbet weent hesheernumberofer r
orscommi ttedand
self-
correct
ionaccur acy.Itwoul dalsobedesi rabletocont r
olforfi
rstlanguageandsome
aspectsofper sonali
ty,duet otheobservedr el
ati
onshipbetweenper sonali
tyandav oidance
behav i
or(Kleinman,1977) .

Possi
blyt
hemosti mpor t
antr esulttoemergefrom thesest
udiesi
st hepointt
hatsel
f-
corr
ectionisneverperfect,neverr eacheswhatsomet eacherswouldconsideraccept
abl
e
perf
ormanceex ceptinthecaseofonev erygoodlearnerwhowaspr esentedwithherown
err
ors!Thefactisthatmanyt eacher sassumesel f
-corr
ecti
onshouldbe100%,t hatst
udent
s
shouldbeabletoapplyalltheyhavel earnedatallt
imes.7

Ifsecondl anguageper f
ormer sdonotf ocusheav il
yonf or mi nself-
cor recti
on, whatdo
theydo?Sev eralst udies,allclassifi
edasf all
ingundercondi t
ion( 2),showt hatr ev isi
onsar e
typi
callyaimedatgr eatercommuni cativeeffectivenessandnotmer elyonform.Hassan( cited
i
nHat ch,1979)not edwhatchangesESLst udent smadeonsecondandt hir
ddr af
tsof
composi ti
ons.Hassanr eport
edt hatt hestudent s" concent r
at edmai nlyonv ocabul arychoi ce,
added mi nor det ails,and made f ewer changes whi ch r esulted in ov erallgr ammar
i
mpr ov ement"( Hatch,1979,p.136) .Schlue( 1977)camet osi mi l
arconclusi ons,not ingt hat
"hersubj ects seemed t o moni tort hei
rspeech qui t
e car efully,butnotf orgr ammat i
cal
correctness.Thei rspeechawar enesswasf ort hemostpar tf ocusedont heirsuccessor
fail
urei nconv eyi
ngt heirmessage.Thus,t heywer ev er
yconcer nedwi thsucht hingsast he
appropr i
atenessoft heirlexi
calchoi ces..
.eveni nt hesel f-
anal ysisact i
vit
y,i
twasof t enhar dt o
maket hesubjectsf ocusonf or mr atherthanonl ex iconandpr onunci ati
on..
."( p.343) .Houck,
Rober t
sonandKr ashen( 1978a)al sonot edt hatmanyoft hecor recti
onsmadebysubj ects
were" obviouslyattempt satimpr ovedi nt
ell
igibil
it
y ,ratherthangr ammat i
calfor m"( p.337) .

Tosummar i
zet husfar:Ourdescr
ipt
ionofwhenwecanMoni
tor
,whatcanhe
Moni
tored,andt
helinguist
iceff
ectofMonitor
ing
al
l

111
111
111
reachsimi l
arconcl usi
ons.Theuseofconsci ousgr
ammari sli
mi ted.Notever
yoneMonitors.
Thosewhodoonl yMoni t
orsomeoft heti
me,anduset heMoni torforonlyasub-
partofthe
grammar .Aswehav ejustseen,theeffect
sofself-
cor
rect
iononaccur acyi
smodest.Second
l
anguageper former scant y
picall
yself-
corr
ectonlyasmal lpercentageofthei
rerror
s,even
whendel iberatel
yfocussedonf orm(condit
ions2to4),andev enwhenweonl yconsi
derthe
easiestaspectsoft hegrammar .

D.Ot
herEf
fect
sofConsci
ousRul
es

Useoft heconsciousgrammar ,wehavemai ntained,islimit

edt oeasi
lylearned,late-
acqui
red r ules,simpl e morphologicaladdit
ions that do not make an ov er
whel ming
contr
ibutiontocommuni cati
ngthespeakerorwr it
er'
smessage.Formostpeopl e,only"local"
rul
escanbel ear
nedandused( Bur tandKiparsky,1972) .Cer t
ainl
y,speaker
sofEngl i
sh
understandsent enceswi t
hmi ssi
ngt hir
dpersonsingularmar kersanddroppedr egularpast
endi
ngsf airl
ywel l
,thankstothepresenceofothermar kersoftenseandpr agmaticknowl edge.

Therei
s,nevert
heless,somerealv
alueinapplyi
ngtheserul
eswhent i
mepermi t
s,when
ruleusedoesnotinterf
erewithcommuni cat
ion.Prov
idi
ngtheselocali
tems,ev
enthought hey
maymakeasmal lcontr
ibuti
ontocommuni cati
on,makeswr i
ti
ngandspeechmor e poli
shed,
i
t adds a cosmet i
c ef f
ect that may be v er
yi mport
ant for many secondlanguage
students.

I
ndeed, i
nt headv ancedsecondl anguageclass,providingsuchpol ishmaybecomet he
main goal,one t hatis quitejustif
ied formanyst udents." Advanced"second l anguage
acqui
rers,especiallythosewhohav ebeeni nthecountrywher ethet argetlanguageisspoken
forafewy ears,mayhav eacquiredagr eatdeal,
butnotall,oft hesecondl anguage,enoughto
meetcommuni cat i
veneed, butsti
l
lshor tofthenati
vespeakerst andar d.Theirchi
efneedmay
beconsciousr ulest ouseasasuppl ementt ot
heiracquiredcompet ence,toenablethem to
appearaseducat edi ntheirsecondlanguageast heyareint heirfi
rst
.

Ido notobj
ecttot hi
ssortofgr ammart eachi
ng.Whati sunf
airi
st o emphasize
accur
acyoncommunicat
ivel
yunessenti
al,l
ateacquir
editemsinbegi
nni
nglanguageclasses,
wit
hstudent
swhoareunabletounder
standthesimplestmessagei
nthesecondlanguage.

112
112
112
E.Pr
esent
ati
onofRul
es

Af ai
ramounthasbeenwr i
ttenabouthow grammarr ulesshoul dbepr esented.One
i
ssueiswhet herrul
esshouldbegi v
en"dir
ectl
y"(deductive)
,orwhet herstudentsshoul dbe
askedtof i
gureoutther ul
esf orthemsel
ves( i
nducti
ve).Anot herissueissequence- -which
rul
esshouldbepresentedfi
rst,
and/oremphasizedmor e.Iwillr
estri
ctmycomment sont hese
i
ssuestotheimpli
cationssecondlanguageacquisi
ti
ont heorymakef orthesequestions.

1.THEDEDUCTI
VE-
INDUCTI
VEI
SSUE

Thi
sissuewasoneofsomeconcer ni nthesecondl anguageacquisit
ionpedagogi
cal
l
it
erat
ureformanyy ears.Formanyschol arsandteachers,deducti
veteachingseemedmuch
morereasonabl
e--whymakest udentsguesst her ul
e?Pr esentaclearexplanati
onandhav e
t
hem practi
ceunt i
ltheruleis"i
nternali
zed".Cognit
ive-
codet eachi
ng,aswel lasgrammar-
t
ransl
ati
on,areexamplesofthe"rul
e-fi
rst
"deductiv
eappr oach.

Proponentsofinducti
veteachi
ngarguedthatthebestwaytoinsurelear
ningwasfor
t
hest
udenttowor koutt herul
ehimself.I
nduct
iveteachi
ngisver
ymuchl i
kerule-
wri
ti
ngin
l
i
ngui
sti
cs.Thelear
nerisgivenacorpusandhastodiscovert
her
egular
it
ies.

Beforeproceedingtosomeoft heresearchbeari
ngont hi
sissue,i
tisimport
antto
clar
ifyonemaj orpoint:bothi nducti
veanddeduct i
velearni
ngarelearni
ng.Neit
herhave
anythi
ngdi rect
lytodowi t
hsubconsci ouslanguageacquisi
ti
on.I
nducti
velear
ningbearsa
superfi
cialr
esemblancet oacquisit
ion,andhasoccasional
lybeenconf
usedwithacqui
sit
ioni
n
theli
teratur
e.AsTable4.6indicates,bothinducti
ve

Tabl
e4.6Acquisi
ti
onandinduct
ivel
ear
ning:
simi
lar
it
iesanddi
ffer
ences
Acqui
sit
ion I
nduct
iveLear
ning

Datafi
rst
,rulef
oll
ows Datafi
rst
,rul
efoll
ows
Rulei
ssubconscious Ruleisconsci
ous
Focusonmeani ng Focusonform
Sl
owpr ogress Mayoccurquickl
y
Requi
reslargeamountsofdata Mayoccuraf terexposuret
o
smallamountofdat
a

113
113
113
l
ear ningandacqui sit
ionshar et hef eaturesofdat a,ori nput,fir
st,wi tht he" r
ule"comi ng
second.Ther ear edeepandf undament aldifferences,howev er.Whent hegoali si nductive
l
ear ning, t
hef ocusisonf orm,andt hel ear
nerat tempt st oanal y
zef ormalaspect soft hedat a
present ed.Whent hegoali sacqui si
tion,theacqui rerat tempt st ounder standt hemessage
cont ainedi nt heinput .Also,t he"rule"dev elopedbyt het wo pr ocessesi sdifferent.An
i
nduct i
vely-
learnedr ul
ei saconsci ousment alrepr esent ati
onofal inguisti
cgener alizati
on--an
acqui redrulei snotconsci ous( wecan,howev er,cer tai
nlyl ear
nl aterwhatwehav eacqui r
ed;
seebel ow),buti smani f
estedbya" feel"forcor rect ness.Al so,inducti
v elearning,sincei tis
consci ouspr oblem-solving,mayoccurv er
yqui ckly--anadeptst udentmay" see"t her egularity
afteronl yaf ewex ampl es.Acqui si
ti
on, howev er,alway stakest i
meandr equiresasubst antial
quant ityofinputdat a.Asdi scussedi nChapt erII,ittakesmor ethanasi ngl
epar agraphanda
fewex erci
sest oacqui rearule.

Thus,from thepointofv i
ew ofsecondl anguageacqui sit
iontheory
,t hedeductiv
e-
i
nductivecontrover
syisnotacent r
alonef orsecondl anguagepedagogy ,sinceitfocusses
onlyonwhi chlear
ningsty
leisbest.Theissuehassomesi gni
ficance,however,andtherehave
beensev eralsuggesti
onsandex perimentalresult
sr el
ati
ngt ot hi
scont r
oversythatareof
i
nterest.

I
thasbeensuggest ed(Hammer l
y,1975)thatcer tai
nst r
uctures"aremostamenabl et o
adeductiveappr oachwhi leothers.
..canbel ear
nedv er
ywel lbyani nducti
veappr oach"(p.
17).Seli
ger(1975)pr esentsdat asuggest i
ngthatretentionov ertimeisbetterwithadeduct ive
approach.Har tnett'
s data suppor tt he hypothesist hatst udents who ar e successfuli n
deducti
vef or
eignl anguagecl assesempl oydif
ferentneur ologicalmechanismst hanlearners
successfulinmor ei nducti
vecl asses,deductivel earnersbei ngmor el eft
-br
ained,anal yt
ic
thi
nkers,andi nductivelearnersbei ngmor er i
ght-brai
ned,anal ogicthinkers(Har t
nett,1974;
Krashen,Seli
gerandHar tnett
,1974) .

Ift
herear eindiv
idualdifferencesi npreferenceofr ul
epresent
ation,ifsomepeople
preferrul
esfir
standot hersprefert of i
gurethingsoutf orthemselves,insi
stenceonthe
"wrong"approachf ort hegr ammarpor t
ionoft hel anguageteachi
ngpr ogram mayr ai
se
anxiet
iesandstrengthent heaff
ectivefil
ter.
8

114
 Thetheoryofsecondl anguageacqui
sit
ionpr esentedi nChapt
erIImakesonl yindi
rect
contr
ibuti
onst othisquestion.Themosti mpor t
antcont ributi
onisit
si nsi
stencethatbot h
deductiv
e and inducti
ve approaches arelearning-or
iented.The "practi
ce"used forr ule
pract
ice(deducti
ve)orrul
e-searchi
ng(i
nduct
ive)wi l
lnotbeopt i
mali
nputforacquisi
ti
on,si
nce
thestudents'
focuswill
bepr imaril
yonfor
mr atherthanont hemessage.

2.SEQUENCI
NGANDLEARNI
NG

Iargued,inChapt erIII
,thatgr ammat i
calsequenci ngwasundesi rabl
ewhent hegoali s
acquisit
ion.Itseemsr easonabl ethatweshoul dpr esentr ulesoneatat i
meinsomeor der
whent hegoali sconsci ousl earning,howev er( "
rulei solation";KrashenandSel iger;1975).
Severalrati
onaleforsequenci nghav ebeensuggest ed.We( Krashen,MaddenandBai l
ey,
1975)oncesuggest edt henat uralor deritself,whichInol ongert hi
nki sthecorr
ectbasisf or
sequencingf oracqui sit
ionorl earni ng.Ot herpr oposal si ncludef requencyofoccur rence,
grammat i
calsi
mpl icity,and" util
ity
"( see,forex ampl e,Lar sen1975) .( Myi mpr
essionist hat
despitetheex i
stenceoft heseopt ions,andt hefai r
lywi despr eaddi scussi
onoft hem int he
professi
onallit
eratur e,thev astmaj or i
tyoft extsut i
li
zesomev ersi
onofl i
ngui
sti
csimpl i
city
,
goingfrom formallyl esscompl extomor ecompl exst ructure.)

Secondl anguageacqui siti

ont heor
y,aspr esentedher e,doesnoty etmakepr edi
cti
ons
ast otheex actlearningsequence.I tdoespredict,howev er,somet hingaboutt hesetofr ul
es
thatcanbel earned.Fi r
st,ifthegoalofgr ammart eachingisi nfactt oprov i
dest udentswitha
Moni t
or,aswedi scussedear l
ierint hi
schapter,simplici
tywi llpl
ayal arger ole.Wecanonl y
teachwhati slearnable,and,r est
ricti
ngthesetev enmor e,whati spor t
able,whatcanbe
carri
edar oundi nt hest udents'heads.( Theset wor equirement sneedt obedi sti
ngui
shed--
l
earningar uledoesnotal way s meanbei ngabl et ousei tinper formance,ev enwhen
conditi
onsar ef avorablef orMoni toruse.)Aswehav eseenabov e,indi scussi ngcasesof
under-usersandi ncompet entMoni torusers,wehav e,asapr ofession,ov erest imatedwhat
mostpeopl ecanl earn, andwhatt heycanr et
ainandusei nper f
ormance.Ev enopt i
malusers,
"goodlanguagel earners"hav elimitsthatarefarbelowmanyt eachers'expect ations.

115
115
115
 Second,unl essourgoali slanguageappr eciation(seebel ow) ,wedon' thav etot each
rul
est hatourst udentshav eal r eadyacqui r
ed.How,t hen,doweknowwhi chi t
emst ot each?
Wecoul d,concei vabl
y, per form adet ail
eder r
oranal ysis oneachst udent,compar et he
result
soft estst hattapl earningandacqui siti
on,anddet erminet hosei temst hathav ebeen
acquired,buthav enotbeenl ear ned,andf ocusonj ustthisset.Thi sispossi ble,butpr obably
unnecessary.The" nat
uralor der "studi escanpr ovideuswi thatl eastsomeoft hei nfor
mat ion
weneed.Whi lesomei ndivi
dualv ar
iationexistsamongsecondl anguageacqui rers,wehav ea
goodi deaofwhati sacqui red" ear l
y"andwhati sacqui r
ed" l
ate"forsomest r
uct ures.Wecan
bef ai
rlycer
tainthatbegi nnersi nESLwi ll
nothav eacqui redthethi r
dper sonsingul ar/s/ort he
possessive/s/,forex ampl e.It hinkt hatav erywor thygoalofappl iedli
ngui st
icsi stoat tempt
todescr i
bet hissetofwhatar et ypicall
ylate-acquired,butlear nabler ul
es,bey ondthef ew
mor phemesandst ructuresweknowaboutnow.

Rul
est
obel
ear
nedshoul
dthusmeett
heset
hreer
equi
rement
s:

1.Lear
nable
2.Port
able
3.Notyetacqui
red

Thesequenci
ngi ssuet henbecomes,orreducesto,deter
mini
ngwhichoft her ul
es
meeti
ngallthr
eeoftheser equirement
sshouldbepresent
edfir
st.Thi
sthussti
llr
emai nsan
i
ssue,butonewehavecont r
ibutedtobyli
mit
ingt
hesetofit
emst hatmustbesequenced.

F.Not
esonEr
rorCor
rect
ion

Anothercontr
oversyrel
atedtoconsci
ouslear
ningistheissueofer
rorcorrect
ion.
Henr
ickson (1978)l
ist
st he"fi
vefundament
alquesti
ons"and r
eviewst
helit
erat
uret hat
addr
essesthem:

1.Shoul
derr
orsbecorrected?
2.I
fso,whenshoulderrorsbecorr
ected?
3.Whichl
earnerer
ror
sshoul dbecorrect
ed?
4.Howshouldlear
nererrorsbecorr
ected?
5.Whoshouldcorr
ectlearnerer
ror
s?

116
116
116
 Secondlanguageacquisit
iont
heor
yhas" answers"t
of ourofthesequest
ions,answers
thatare,asareallotheri
mplicati
onsint
hisbook,themselveshypotheses.I
nthiscase,Iam
predict
ingthati
ferrorcor
rect
ionisdoneaccor
dingtotheprinci
plesdescri
bedbelow,itwi
llbe
effecti
ve.

1.Shoul
der
ror
sbecor
rect
ed?

Accor di
ng tothe second l anguage acqui si
tion theor ypr esented her e,when er ror
correcti
on" works",i
tdoessobyhel pi
ngt hel earnerchangehi sorherconsci ousment al
representat
ionofar ul
e.I
not herwords,i
taf f
ectslearnedcompet encebyi nformi ngt hel
earner
thathisorhercur r
entversionofaconsci ousr ulei
swr ong.Thus, secondl anguageacqui sit
ion
theoryimpliesthatwhent hegoali slear
ning,errorsshoul di ndeedbecor r
ected( butnotatal l
ti
mes; seebel ow;andnotal lrul
es,evenifthegoali slearning).Thet heorymai nt ainshowev er,
thaterrorcor r
ecti
onisnotofusef oracquisiti
on.Acqui si
t i
onoccur s,accordingt otheinput
hypothesis,whenacquirersunder standinputf oritsmeani ng,notwhent heypr oduceout put
andf ocusonf orm.

2.Whenshoul
der
ror
sbecor
rect
ed?

Hendri
ckson,foll
owingBi
rckbi
chl
er(1977)
,suggest
st hati
ngener
aler
rorcorr
ect
ionbe
l
i
mited to " manipulat
ive gr
ammar pr act
ice"
--
more errors may be toler
ated duri
ng
"
communi cat
ivepracti
ce".

Thei mplicationsofsecondl anguageacqui si

ti
ont heoryaresi
mil
ar.Iferrorcorr
ecti
on
aimsatl earni
ng,i tislogicaltosupposet hattheconditionsforerr
orcorrecti
onshoul dbe
i
dent i
calt othecondi t
ionsforutil
izinglearni
ng--weshoul dfocusourstudentsonf orm,and
correcttheirerror
s,onl ywhent heyhav et i
meandwhensuchdi v
ersi
onofat t
entiondoesnot
i
nterferewi t
hcommuni cat
ion.Thisimpl i
esnoer rorcorr
ectioninfr
eeconversat
ion,butall
ows
forer r
orcor r
ection on wr it
ten wor k and grammarex erci
ses.Thisis precisel
yTer r
ell
'
s
procedurei ntheNat uralApproach( descri
bedinChapt erV).

3.Whi
cher
ror
sshoul
dbecor
rect
ed?

Hendr
icksonr
evi
ewst
hreehy
pot
hesesandaccept
sthem al
laspl
ausi
ble.

117
 (1) We should correct" gl
obal"errors,err
orst hatint
erf
ere wit
h
communicat
ionori
mpedet heintell
i
gibi
li
tyofamessage( Bur
tandKipar
sky
,
1972)
.Sucherr
orsdeservetopprior
ityi
ncorrect
ion.

(
2) Er rorsthatar
ethemoststi
gmat
ized,thatcauset
hemostunf
avor
abl
e
r
eact
ions,arethemostimpor
tantt
ocorrect
.

(
3) Er
ror
sthatoccurmostf
requent
lyshoul
dbegi
vent
oppr
ior
it
y.

I
nt he previous section,t he l
ingui
sti
c domai n oft he Moni torwas descr ibed.I
recommendedt hatwer estri
ctt heconsciouslear
ningofr ulesf orMoni toruseaccordi ngto
thesecharact
eri
stics:t
her ulestobel earnedshouldbe( 1)learnable,(
2)por tabl
e,and( 3)not
yetacquir
ed.Thesechar acterist
icsmightalsodescribewhicher rorsshoul dbecorrected,ifi
t
i
si ndeedthecaset hater r
orcor rect
ionaf f
ectsonlytheconsci ousgr ammar .Perhapswe
shouldonlycorrectmi stakest hatrefl
ectrulesthatcanbeusedaspar toftheconsci ous
Monitor.

Thismayappeart obeamodestcont ri
buti
ont ot heissueofwhi cher r
orsaretobe
correct
ed.Manyt eachers,howev er
,trytopoi ntoutorcor rectaller
rors.Thissuggesti
on
reduces the si
ze oft he task considerably.Wi t
hint he smallsetdef ined by t
he three
characteri
sti
csoflearnable,port
able,andnoty etacquired,westil
lhavet omakedeci si
ons,
andher econsider
ationssuchasf r
equency ,contr
ibut
ionst ocommuni cation,andir
ri
tabi
li
ty
mayber el
evant
.Theov erallt
ask,however,isreducedenor mously.

4.Howshoul
der ror
sbecorr
ect
ed?
Hendr
icksonrev
iewssev
eral
met
hodsofer
rorcor
rect
ion,
incl
udi
ngt
het
womostwi
del
y
used:

(
1) pr
ovi
dingt
hecor
rectf
orm(
"di
rect
"cor
rect
ion)
.

(
2) t
hedi
scov
ery(
induct
ive)appr
oach.

Henotesthatlit
tl
er esearchisav ai
lablethatest
abl i
shesthesuperi
ori
tyofonemet hod.
Somer esearchshowst hatdirectcorrecti
oni snotpar ti
cular
lyeff
ecti
ve;st
udentswhohav e
haddirectcorr
ecti
onoft heiroralandwr it
tenout puti
ni nst
ructi
onalpr
ogramsdidnotproduce
fewererror
s(Hendrickson,1976, 1977b,citedinHendrickson,1978;CohenandRobbins,
1976).Thismay,notesHendr ickson,beduet o

118
118
118
t
hel
ackofconsi
stentandsy
stemat
iccor
rect
ion(
All
wri
ght
,1975;
CohenandRobbi
ns,
1976)
.

Secondlanguageacquisi
ti
ont heor
ypredi
ctst
hater
rorcor
rect
ionwi
l
lshow posi
ti
ve
r
esul
tsonlyift
hefoll
owingcondi
tionsaremet
:

(
1) Er
ror
scor
rect
edar
eli
mit
edt
olear
nabl
eandpor
tabl
erul
es.

(2) Err
orsar
ecor
rect
edundercondi
ti
onsthatal
low Moni
toruse.Thi
swi
l
l
giv
ethelear
nert
imet
oreconsi
dert
herul
ethatwasvi
olat
ed.

(3) Measuresevaluati
ngtheeff
icacyoferr
orcorr
ectionareadmi
nist
ered
undercondit
ionsthatall
owMonitoruse,t
oall
owt hel
earnerti
metoref
ertohis
orherconsci
ousknowl edge.

(
4) Subject
sar
e"Moni
tor
-user
s"(
i.
e.t
heyar
enotunder
-user
soft
he
Moni
tor
).

Err
orcorrecti
ont hatisnotdoneundert heseconditi
ons,Ipredi
ct,wil
lnot"wor
k";Iam
al
sonotopt i
misticaboutt heeffi
cacyoferrorcorr
ecti
onev enwhenal ltheaboveconditi
ons
ar
emet .Asisthecasewi t
hcondi t
ionsforMonit
oruse,theyarenecessarybutnotsuf
fici
ent-
-
ev
en undert he" best
"condi ti
ons,correct
ing t
hesi mplestrules,wi
tht hemostl earni
ng-
or
ient
ed students,teachercor recti
ons wi
llnotpr oduce result
sthatwi l
lliv
e up tot he
ex
pectati
onsofmanyi nst r
uctor
s.

B.Gr
ammarasSubj
ectMat
ter

Asment
ionedear
li
er(p.88)
,"gr
ammar"hasanotherpl
acei
nthepedagogi
calprogr
am,
aplacet
hati
snotalwaysclear
lydi
sti
ngui
shedfr
om it
suseasaconsciousMonitor.Thi
sis
gr
ammarassubjectmat
ter.

Manyst udent s(
probabl
yf ewerthanmostofust hink)areinterest
edinthestudyoft he
structureofl anguageperse.Theymayal sobeinter
estedi nlanguagechange,dialects,etc.
Especi all
ysat isf
ying,f
orsomest udents,islear
ningwhathasal readybeenacqui red,the
Eur ekaphenomenondescr ibedearlieri
nt hi
ssecti
on(p.88) .Mystudent swhorecognizedthat
theyhadal readyacquir
edt hethreeusesoft hepresentprogressiv
et enseinEngl
ishwer ev ery
satisfiedandpl easedtohav econsci ousknowl edgecor respondingt othei
rsubconsci ous
knowl edge.Theyal sot
houghtt hatIwasanout st
andinglanguaget eacherforpr
ovidingthem
witht hiskindofinsight
!

119
119
119
 Provi
dinglearni
ngt hatcorr
espondswi thprevi
ousacqui si
ti
onhasi t
sadv antages,butI
donott hinkitislanguaget eachi
ng--i
tisnotinputforacquisiti
on( al
thought helanguageof
cl
assroom di scussi
onmaybe;seebel ow),anditalsodoesnotpr ovi
deusef ull
earningthat
canbeut il
i
zedasasuppl ementt oacqui si
ti
on,asaMoni tor.Itmayser v
eonepur pose,
howev er
:itcandemonst ratetothelanguagestudentthatacquisit
ionisreal,andthatitcanbe
tr
usted.Pointingoutwhathasbeenacqui r
edmayt hussti
mul atemor efait
hi ntheacquisit
ion
process,andl owert
heaffectivefi
l
ter.I
tmayt husbeapar t
ialcuref orover
-useoft heMoni t
or.

Thestudyoft hest
ruct
ur eoflanguage,how itvari
esovert i
me( hi
stori
call
ingui
sti
cs)
andinsociety,hasmanygener aleducat
ionaladvantagesandv aluesthathighschooland
uni
ver
sitylev
ell anguageprogramsmaywantt oincl
udei napr ogram.Itshouldbecl ear,
however,t
hatt eachi
ngcomplexf actsaboutthesecondl anguagei snotlanguageteaching,
butr
atheri
s"languageappreciati
on"orli
nguist
ics.

Teaching grammaras subj ect-

mattercan r esul ti nlanguage acqui si
tion in one
i
nst ance,howev er:whent het argetlanguageisusedasamedi um ofi nstructi
on.Acqui si
ti
on
occur sinthesecl asseswhenst udentsareinterest
edi nt hesubjectmat ter,"
grammar "
.Very
often,whent hi
soccur s,botht eacher sandst udentsar econv i
ncedt hatthest udyoff ormal
grammari sessent i
alforsecondl anguageacqui si
ti
on, andt het
eacheri sskill
edatpr esenti
ng
explanationsi nthet argetlanguagesot hatt hest udent sunder stand.Inot herwor ds,the
teachert al
k,insuchcases,meet stherequir
ement sf ori nputforacqui si
ti
on,aspr esentedin
Chapt erIII
:thei nputiscompr ehensi bl
eandconsi der edt ober elevant.Thef il
terislow in
regardt othel anguageofex planation,asthest udent s'consciousef fort
sar eusuallyont he
subjectmat ter,whatisbeingt alkedabout ,
andnott hemedi um.

Thisisasubt lepoi
nt.Ineffect
,bot
ht eacher
sandst udentsaredeceivi
ngt hemselves.
Theybel i
evethatitisthesubjectmat t
erit
self,t
hestudyofgr ammar ,t
hatisresponsibl
ef or
thestudents'progressinsecondlanguageacquisit
ion,butinreali
tyt
heirpr
ogressi scoming
from themedi um andnott hemessage.Anysubj ectmat t
erthatheldthei
rint
erestwoulddo
j
ustaswel l,sof arassecondl anguageacquisit
ionisconcer ned,aslongasi trequi
red
ex t
ensiv
euseoft het ar
getl
anguage.

Thi
smayunder
li
eandex
plai
nthesuccessofmanygr
ammar
-based

120
120
120
approaches.Theyaretaughtint hetar
getlanguage,andt hispr
ovi
descompr ehensibl
einput
foracquisi
ti
on,inputt
hati srelevantandi nt
eresti
ngasl ongasthestudentbelievesthat
consciousgrammarisgoodf orhim.(Forf
urtherdiscussi
onofsuchaclass,seeKrashen,
1980.)

Not
es

1 Thi
sisnot
,Ishoul
dpoi
ntout
,Sel
i
ger
'si
nter
pret
ati
on.SeeNot
e2.

2 Seli
geri nterprets hisr esult
s as bei ng count ertot he theor yofsecond l anguage
acquisiti
onpr esentedi nChapt erI I.Hisi nt
er pr
etationoft het est,andt het heory,ar
ebot h
diff
erentf rom mi ne.Heconsi der sthetestsi t
uationt obe" f
ormal "
,and" notasampl eof
l
anguagewi thinanat uralcont ext"(p.362) .Ther ei s,It hi
nk,somet r ut
ht ot hisanalysis,as
ment ionedint hetex t
.Subj ects'per f
ormance, howev er ,i
sconsistentwi ththehy pothesist
hatt he
testt appedpr i
mar i
lyacqui si
ti
on- -t
hisissuppor tedbyt hedat aandi sconsi stentwitht he
hypot hesi
st hatMoni toruseoccur sonl ywhensev eral necessar
ycondi t
ionsar emet ,asstatedi n
Chapt erIIandr epeatedi nthischapt er.Mor est rangei shisinterpret
at i
onoft heacqui sit
ion-
l
earni ngdistincti
onandt heMoni torhypothesis:hisr esultsarecount ert o"Moni torTheor y"
,he
clai
ms,si nceMoni torTheor ymai ntainsthat" l
earner sdowhatt heysayt heydo" ,andhisdat a
showst hi
si snotso.Hi sdat adoesi ndeedconf i
rmt hatper f
ormer sdonotal way sdowhatt hey
sayt heydo,but" Moni torTheor y"doesnot ,andnev erhas,madet hecl aimt hatpeopl edowhat
theysayt heydo.

Seligerout l
ineshi sownposi ti
oni nt hesamepaper .Pedagogi calrules, heasser ts,"most
l
ikelyser veasmechani smst ofacil
it
atet hel earner'
sf ocussingont hosecr iterialattri
butesoft he
reallanguageconceptt hatmustbei nduced"( p.368).Theyser v
eas" acqui siti
onf acili
tators"and
"make t he i nduct ive hy pothesistest i
ng pr ocessmor e ef fi
ci
ent "( p.368) .Sel igerpr ov i
des,
unfort unat el
y ,v eryl i
ttlemor ethant hisbywayofdescr ipti
onofhi shy pot hesis,whi chhe
present sas anal t
er nativet o Moni t or Theor y.He al so pr esent snot hingi nt heway of
empi ricalsuppor tforhi sposi t
ion.Ther ei s,mor eover,aser iouspr oblem wi tht hishy pothesis:If
rul
el ear ningi ssoof t enwr ong( apoi ntweagr eeon) ,how cani tbeusef ulasanacqui si
tion
focussi ng dev ice?Al so,aswe hav e seen ear li
eri nthissect ion,acqui siti
on need notbe
precededbyconsci ousl earni
ng.Rat her ,t heav ai
lableevidencesuppor tst hehy pothesist hat
acqui siti
onoccur sonl ywhent heacqui rer '
sat tenti
oni sont hemessage,notont heform oft he
i
nput .Accor ding t ot he InputHy pot hesi s,consci ous r ules do notf acili
tate acqui sit
ion.
Acqui sitionoccur sv iaacompl etelydiffer entr oute.Anal t
er nati
vehy pothesi smustdealwi tht he
evidencesuppor t
ingt heI nputHy pothesi s,andt hear gument sthatacqui si t
iondoesnotr equire
previousl earning.

3 Beforeweconcl udet hatEv aandUe- l

insimplyneedmor edrillandlearni
ng, consi
derthe
possibili
tythattheyareamongt hebet terlearners.CohenandRobbi ns'subjectHung, an"under-
user" of t he Monitor,also made er rors on t he third person si ngular /s/ and / -
ed/,
consistentwi ththehypothesisthatsuchi temsar etypi
call
ylate-
acqui red.IncontracttoUe-li
n
andEv a,Hungcoul dnotsel f
-correctbyr ul
e,howev er.Whenconf ront edwithat hi
rdperson
si
ngul ar/s/deleti
onhehadmade, heremar ked:"IguessIjustneverlearnedt herulethatwell
,so
Ijustwr i
tedownwhat everIfeellikeit.
"(p.52) .Also,"Whenconf rontedwi thasent encehehad
writt
enwher eaneddel eti
oner r
oroccur red( '
Hegotdi scourage'),Hungsuppl eded,buthe
i
comment ed:'Idon'
tsee

121
121
121
why'.
"(p.53).Hungal sonotedthati
twasv eryhardforhimt odet ecter
rorsinhi
sownout put
.
Hungi squot edmanyt i
mesassay i
nghedoesnotpayat tent
iontof orm:"Idon'
tcaret he
grammar( p.50).
..Ij
ustneverl
earnedtherulesthatwell
..
..Ijustwrit
edownwhatIf eell
ikeit(p.
59).
..IgetkindofboredwhenIst udyEnglish"(p.51)
.Agai n,manypeopl e,despi
teexposurein
cl
ass,havepr act
ical
l
ynoi deaofrul
esthatseem strai
ghtforwar dt
ous.

4 Inarecentstudy,
J.Brown( 1980)administ
eredagrammar-
typet
estwhosefor
matwas
qui
tesi
mil
artothatofLarsen-
Freeman.Subject
sonlyhadtosuppl
yonemorpheme,asi
n:

I (
tal
k)t
oJohny
est
erday
.

Thetestwasadmini
ster
edwit
hnoti
melimi
tto66ESLstudent
swithavar
iet
yoffi
rst
l
anguagesatMarymountPal
osVerdesCol
l
ege.Her
eisthedi
ff
icul
tyor
derBrownr
eport
ed:

Auxili
ary 96%cor
rect
Copul a 94.
2%
Regularpast 92%
Pl
ur al 91.
8%
The( defart
) 88.
2%
I
rregularpast 88%
a(indefart) 86.
6%
Ø( art) 85.
8%
Possessive 80.
2%
i
ng 80.
2%
Thirdpersonsing 77%

This order,Brown repor t

s,correl
ates signi
fi
cantl
y wit
h other second language
mor pheme or der
s( rho = 0.
73),compared t o Ander
sen,1978).Itis anal
yzed somewhat
dif
ferentl
yfrom otherst
udiesi
nt hattheal
l
omor phsofarti
clearepr
esentedseparat
ely
; t
heyare
verycloseinrankorder,however
.

Brown' sorderi sdif fi

cul
tt ointerpr
etduet othecloselybunchedscor es.Theorder
appearstobesi mi l
artoot herL2or dersi ntheli
teratur
ewi t
ht woexceptons:i
i ngisunusuall
y
ow,andr
l egularpastisunusual l
yhigh( seeChapt erII
,Hypothesis2)
.Thehi ghrankofregul
ar
pastisconsist entwithmyhy pot
hesisthatsuchf ormatsencourageMoni t
oruse,whichresul
ts
i
naj umpi naccur acyoft her ankoflate-acqui
redbuteasyt olearnmorphemes.Ihavenohandy
expl
anationf oring'srelati
vel ypoorshowi ng,norcanIaccountf ort
hethirdpersonsingul
ar'
s
l
owr anki nthisMoni toredt est.Brown'shighcorrelat
ionwithotherstudi
esiscountertosome
ofmycl aims, buttheriseinr egul
arpasti snot.

5 I
tisv eryint
er estingtonot ethataccur
acyf orthethirdpersonsingularinotherMoni tor
-
fr
eest udiesi sv er
ysi mi l
art otheaccuracyfound i nt hecomposi ti
onst udyf orthe" f
ree"
conditi
on.InBai l
ey,Madden, andKrashen(1974)
, usingtheBSM, t
hir
dper sonsingularaccur acy
was0. 41,whileinKr ashenHouck, Giunchi
,Bode,Birnbaum, andSt r
ei(1977),usi
ngf reespeech,
accuracyf orthismor phemewas0. 36.Comparet oTabl e4.3,wher eaccuracyint het wof ree
conditi
onsi s0.54and0. 32,goingupt o0.
61and0. 65i ntheeditedcondition,r
espect i
vely.This
si
mi l
arit
yisconsi stentwi t
ht hehypothesi
sthattheedi t
edcondi ti
oninvolvedli
ghtMoni toruse,
andthatlat e-
acquir
ed, easyitemsaremostaptt obeaf f
ected.

6 Infocussi
ngontheregul
arpastandthi
rdper
sonsingul
ar,Ibynomeanswi sht
oimpl
y
t
hatt
hesearetheonl
ypoint
sofgrammarthatcanbeconsci
ousl
yMoni t
ored.Theyar
e,

122
122
122
rather
,typi
calofwhatcanbeMonitor
ed,andar
econv
eni
entt
ofol
l
owt
hroughsev
eralst
udi
es
sincetheyarementi
onedandanal
yzedsooft
en.

7 Sev eralot herst udiesal so per tai

nt o Moni toring abil
itybutdo notf ocuson sel f-
correcti
on.Asdescr ibedear l
ier,Kr ashen,But ler,Birnbaum andRober tson( 1978)askedESL
studentsatUSCt owr it
e" f
ree"and" edit
ed"composi tionsi nEnglish(conditi
ons1and2) .Inboth
cases,nat uralorderswer efound,wi thasl i
ghti ncr
easei nt hethirdpersonsi ngul armorpheme
i
nt heedi tedcondi t
ion.Ther eal so wasa6% ov eralli ncr
easef ort heedi tedcondi t
ionfor
thesi xmor phemesanal y z
ed, fort hegr oupasawhol ewi t
hsomei ndivi
dualv ar
iationaccording
tofirstl
anguage, Farsispeaker sbei ngthemostef f
icient ,showinga16%gai ni naccur acyinthe
edit
edcondi tion.Thi sisadi fferentmeasur et hant hatdescr i
bedi nt hetex t
,si ncesubjects
wrotecompl etelynewanddi fferentessay s.TuckerandSar of
in(1979)pr esented18" advanced
i
ntermedi ate"Ar abicspeaki ngst udentsatt heAmer icanUni versit
yatCai r
owi th14dev i
ant
sentences.St udent swer easkedt o" drawal i
neunder neat htheerrorandcor rectiti fy
oucan"( p.
32).Thiscor respondst ocondi ti
on( 3).Ther angeoft hepr oporti
onofer rorscorrect edwasf r
om
33%t o83%, dependi ngont heer ror ,wit
her r
orsof" number "bei
ngeasi esttocor rect(asin*SoI
tookt headv i
cesofmypar ent s).

Light
bown,SpadaandWal lace(1980)al socontri
butest oourknowl
edgeofMoni t
or
ef
fi
ciency
.Theygavet hei
rsubj
ects,grade6,8,and10st udentsofEFLinQuebec,at estof
gr
ammat i
cali
tyj
udgmentsinEngl
ish.Subject
swer easkedtocircl
etheer
rorsi
nasentenceand
wri
tethecorr
ectf
orm.Thestudyfocussedont hesestruct
ures:

Plural/s/
Possessive
/s/
Thir
dper sonsingular/s/
Cont r
actabl
ecopula/ s/
Cont r
actabl
eaux i
li
ary
/s/
Be,usedf orex
pressingage( e.
g.Iam si
xy ear
sold.Thi
sisconsi
der
eda
problem str
uct
uref orFrenchspeaker
s.)
Prepositi
onsoflocation(Theyar ngt
egoi oschool
.)

Thetestwasgiv enthreet imes, thef i

rsttwoadmi ni
str
ati
onsbei ngonlytwoweeksapar t,
t
hethirdcomingfi
vemont hsl ater,aftersummerv acati
on.Inbetweenadmi ni
strati
onsIandI I
,
t
herulesusedont het estwer er ev i
ewedi nclass.Lightbownetal .r
epor tsomei mpr ovement
f
rom t
imeIt oti
meI I
--
ther evi
ewi nclassr esult
edinamodestbutnot iceable11%i mpr ovement
f
orgrades8and10,compar edt o3%f orcont rolstudentswhosi mpler et
ookt hetestwi t
hout
r
evi
ew,anda7%i ncreasef ort he6t hgr aders(nocont rol
swer erunf orthe6thgr ade).Inthe
t
hir
dadmi ni
str
ati
on,fi
vemont hsl ater,scor esfel
lbacktoal evel
betweenadmi ni
strati
onIandI I
.

Theresul
tsofLi ghtbownetal .areconsi
stentwiththoser eport
edi nthet ext
,even
thought hetaskisnotsel f
-corr
ecti
onbutiscorr
ecti
onofer r
orspresentedtothest udent
,asin
TuckerandSar af
in.Thest udentswereonl
yabletocorrectapproxi
mat el
y1/ 4to1/ 3ofthese
errors,despi
tetwot osixyearsofformalst
udyanddespitethefactthatthestruct
uresinvol
ved
wer efairl
yst
rai
ghtforward.Thetaskcorr
espondstocondit
ion(3)i
nTabl e4.4.

Review oftherulesinclassalsohadamodestef fect

,muchofwhi chwasl ostaf
ter
summerv acation.IconcurwithLi bownetal
ght .'
sinter
pretat
ionthat"i
mprovementsonthe
secondadmi ni
strat
ionwerebasedont heappl
i
cat i
onofknowl edgetempor
ari
lyret
ainedata
consci
ous l evel,butnotf ul
l
yacqui r
ed".The result
s ofadmi ni
str
ati
on I
IIshow justhow
temporarylearnedknowledgeis.

Li bownetal
ght .al
sopr
ovi
deananal
ysi
sofr
esul
tsf
ori
ndi
vi
dual
str
uct
ures.Theynot
e

123
123
123
thatsubj
ectsmadesi gnifi
cant("dramati
c")improvementf r
om ti
meIt otimeI Iont hebe/ have
rul
eandont hethir
dper sonsingular.Thi
ssuppor t
s,theynote,myhypothesi
st hatsimplerrules
areeasierto consci
ousl ylearn,sincet hedescripti
on ofboth oftheser ul esisrelativel
y
strai
ght
forwar
d.Therewasal soasi gni
fi
cantimprov ementofamuchl esstransparentrule,the
useoflocati
vepreposit
ions,andt heplural,whi
chappearst obe"easy
",didnot ,however,show
l
argegains.

Di
ffi
cult
yor der
sf orthe/ s/mor phemesconf or
medt ot henaturalorderpresentedin
ChapterII
,confir
mi ngbothther el
iabil
i
tyoft henaturalorderit
self
,andt heclai
mt hatittakes
mor ethanconditi
on( 3)todi st
urbt henat uralor
dersignifi
cantl
y(i.
e.conditi
on(3)doesnot
focusonf orm strongl
yenough) .(SeeLi ghtbown,inpress,foradi scussionoftheef fectof
classr
oom inputonmor phemeor ders.)

8 I
nanear l
ierpaper(Krashen,Sel i
gerandHartnett
,1974)
,wesuggestacompr omise:
t
eachr ul
e-fi
rst,whichwillsati
sfyt hedeductivest
udents.Theinductivestudent
scansi mpl e
i
gnoret he rule presentat
ion." Practi
ce"can then serve as practi
ce i
nr ul
e appli
cati
on
(
Monitori
ng)f orthededuct i
vestudent s,andasrul
e-searchi
ngfortheinductiv
est udent
s.The
r
ulecanbe( re)st
atedafterthepr acti
ce,arevi
ew f
ordeduct i
vestudentsandconf ir
mati
onf or
t
heinducti
vest udents'hy
pothesis.

124
124
124
Chapt
erV

Appr
oachest
oLanguageTeachi
ng
Chapter
sI I
IandIV weredev
ot edtoadi scussionofthegener
ali mpl
i
cat
ionsofsecond
l
anguageacquisi
ti
ontheor
y.ChapterII
Idescr i
bedthechar
acter
ist
icsofopti
malinputfor
acqui
sit
ion,hy
pot
hesi
zingt
hatlanguageacquisi
tiont
akespl
acebestwheninputi
sprovided
thati
s:

(
1) Compr
ehensi
ble

(
2) I
nter
est
ingand/
orr
elev
antt
otheacqui
rer
.

(
3) Notgr
ammat
ical
l
ysequenced.

(
4) Pr
ovi
dedi
nsuf
fi
cientquant
it
y.

Thepr esent
ationofthi
sinput,moreover,shoul
dbedonei nawayt hatdoesnotputt he
acqui
rer"ont hedef
ensi
ve";i
tshouldnotr ai
seorstrengt
hentheaf
fecti
vefi
lter
.Inaddi
ti
on,
acqui
rersneedt obeprovi
dedwi t
ht oolstohelpthem obtai
nmoreinputfrom theout
side
worl
d.

ChapterIVt r i
edt o" putlearningini t
splace".Consci ousr ulesshouldbeusedonl ywhen
theydonoti nterferewi thcommuni cati
on.I nadditi
on, onlyasmal lpartofthegr ammari sboth
l
ear nableand" por table"f ormostpeopl e.Rulest hatseem v eryt r
ansparentt opr ofessional
l
inguist sandlanguaget eachersmaybequi teopaqueev ent o" goodl anguagel earners".We
also concl uded t hatt he ef f
ects ofl earning ar e qui te modesti nt erms ofsy ntacti
c
accur acy,butt hatt heappl icati
onofconsci ousr ulesmaypr ovideacosmet i
cef fectthatis
i
mpor tantforsomel anguagest udent s.Chapt erIVal sodi scussedt hefactt hatconsci ous
grammarcanbet aughtassubj ectmat ter;acquisi
tionr esultsifsuchacour seistaughti nthe
targetl anguage.Thi s" languageappr eciati
on"f unction,howev er ,needst obedi sti
nguished
from theMoni torf unct ionf orgrammar .

Thegoal
oft
hef
ir
stpar
toft
hischapt
eri
stoanal
yzesomecur
rentappr
oaches

125
125
125
tolanguaget eachi
ngintermsoft heconclusi
onsofChapt ersIIIandI V.Ther esul
tsofthis
analy
siswi l
lthenbecompar edwiththeresul
tsofwhatappl i
edl i
nguist
icsresearchhasbeen
donei nt
hear eaof" met
hodcompar isons"
.Myi nt
erpret
ati
onwi l
lbet hatinthi
scase,curr
ent
secondl anguageacquisi
ti
ont heoryandappliedlingui
sti
csr esearchcomet ov er
ysimilar
conclusi
ons.

Thenex tsectionexami nessomeal t

ernativest othet r
aditionalclassroom,al ternat
ives
thatseem t ohav etheadv antageofsat i
sfyi
ngi nputr equir
ement#2( the"interesting/ r
elev
ant"
requirement)f arbet terthan anyt radit
ionalcl assroom appr oach.Nex t,ther ei s a br i
ef
discussionoft heimpl i
cati
onst heseideashav ef orsecondl anguaget esting.Iwi llsuggest
thatweneedt ocar eful
lyconsiderwhatef fectsourachi evementt estshav eonst udentand
teacherbehav ior.Ourgoalint esti
ngisthis:whenst udents"studyf orthet est",theyshoul dbe
doingt hingsthatencour ageorcausesecondl anguageacqui sit
ion.Thef inalsect i
onsoft his
chapterdealwi thwhatIper ceivetobesomegapsi nmat erial
s,andsomeoft hepr acti
cal
problemsi nimpl ement i
ngthesuggest i
onsmadeher e.

A.Pr
esent
-dayTeachi
ngMet
hods

Theconcl usionsofChapt ersI I

IandI Vcanbesummar i
zedast hemat ri
xi nFig.5.1.We
cansi mplyask,f oreachappr oacht oclassroom t eaching,t owhatex tenti tsat i
sfi
est he
requi r
ement sf oroptimalinputandt owhatex tentitput slear ningi nitspr operpl ace.Wewi l
l
reviewwhatar esurelythemostwi delyusedmet hods,gr ammar -
translat i
on,audi o-l
ingual
ism,
cogni ti
ve-codet eaching,andonev er sionoft hedi rectmet hod.Wewi llt hencov ersomenew
appr oaches,Asher 's TotalPhy si
calResponse met hod,Ter rell
's Nat uralAppr oach and
Lozanov 's Suggest opedia.(Severalv eryi nt
eresting met hods ar e noti ncluded,such as
Gat tegno'
sSi l
entWayand Cur ren'sCounsel ing-Lear ni
ng met hod.Thi si sduet o several
factors,includingmyownl ackoff ami l
i
aritywi t
ht hesemet hods,t hel ackofempi ri
caldata
compar i
ng t hese met hods to other s,and t he av ail
abili
tyofSt evick's ex cellentanaly si
s
(Stev i
ck,1980) .)Thebr i
efdescr i
pt i
onsuppl i
edatt hebegi nni ngofeachanal ysisisnot
i
ntendedt obeaf ullandadequatedescr i
ptionofeachmet hod, buti sintendedt oser veonlyto
i
nfor mt her eader

126
126
126
 Fi
g.5.
1.Ev
aluat
ionschemaf
ormet
hodsandmat
eri
als

Requi
rement
sforopt
imali
nput

1.Compr ehensible
2.I
nterest i
ng/relevant
3.Notgr ammat i
callysequenced
4.Quant i
ty
5.Fi
lterlevel (
"offthedefensi
ve")
6.Providest oolsforconversati
onal
management

Lear
ning

Restr
ictedt o:
1.Certainrules;thosethatar
e
a.lear nableb.
por table
c.noty etacquired
2.Certainpeople( "Monit
oruser
s")
3.Certainsituati
onsa.t i
me
b.f ocusonf orm

whatmyunderstandi
ngofeachmethodis.I
tshoul
dalsobepoint
edoutthatthisanaly
sis
assumest
hatthemet hodsar
eusedi
nthei
rpur
eforms,
asit
uat
ionwhich,
Iam sure,doesnot
occuri
nev
eryclassr
oom.

1.GRAMMAR-
TRANSLATI
ON

Whil
ether
eissomev
ari
ati
on,gr
ammar
-t
ransl
ati
onusual
l
yconsi
stsoft
hef
oll
owi
ng
act
ivi
ti
es:

(
1) Ex
planat
ionofagr
ammarr
ule,
wit
hex
ampl
esent
ences.

(
2) Vocabul
ary
,pr
esent
edi
nthef
orm ofabi
l
ingual
li
st.

(
3) Areadi
ngselect
ion,
emphasizi
ngtherul
epr
esent
edi
n(1)abov
e
andthevocabul
arypresent
edi
n(2).

(
4) Exer
cisesdesi gnedtoprovi
depract
iceonthegrammarand
vocabularyoft hel
esson.Theseexerci
sesemphasi
zetheconsci
ous
contr
ol ofstructur
e("f
ocuson",i
nthesenseofKrashen
andSeliger,1975)andincludet
ranslat
ioni
nbothdir
ect
ions,
fr
om L1t oL2andL2t oL1.

Mostgrammar-t
ranslati
oncl assesar
edesignedf
orfor
eignlanguagei
nst
ruct
ionand
ar
etaughti
nt hest
udent'
sfirstlanguage.Wetur
nnowt oananaly
sisofgrammar
-tr
ansl
ati
on
i
nter
msoft hematri
xinFig.5.1.

127
127
127
(
a)Requi
rement
sforopt
imali
nput

(i)Comprehensible.Itcanonl ybeclai
medt hatgrammar-tr
anslat
ionprovi
desscrapsof
compr ehensi
blei nput.Themodelsent encesar eusuall
yunderstandable,butthefocusis
entir
elyonf orm,andnotmeani ng.Ther eadi
ngselectionisthepr imarysource,butthe
selecti
onsprovidedar enearlyalway smucht oodif
fi
cult
,oftenr
equiri
ngwhatNewmar k(
1966)
calls"cryt
oanalyt
icdecodi ng".Studentsaref or
cedtor eadwordbywor d,andconsequentl
y
rarelyfocuscompl etel
yon t he message.The sent encesused i nt he exerci
sesmaybe
compr ehensi
ble,buther eagai n,asi nthemodelsent ences,t
heyar edesignedtofocusthe
studentsonf or
m.

(i
i)Interesti
ng/ r
elevant
.Ther ei susuallyanat tempt,especiall
yinr ecenty ear
s,t o
providetopicsofi nterestinther eadingselecti
on,butt heusualtopicsfallfarshortoft he
Forgett
ingPr inci
ple( ChapterI
II)
.Theycl ear
lydonotsei zethestudents'att
entiontosuchan
extentthattheyf or
gett hatiti
swr i
tteninanotherlanguage--
repor
tsofat ri
pt oFrance,evenif
i
tincludestheLouv re,general
lydonotpr ovi
dei nf
ormationthatmosthi ghschoolandcol lege
studentsintheUni tedSt at
esareeagert oobtain.

(i
i)Notgr
i ammat icall
ysequenced.Grammar -tr
anslati
onis,ofcourse,gr
ammatical
l
y
sequenced,themajori
tyoft ext
satt
empt i
ngtopr oceedfrom whattheauthorconsi
der
seasy
rul
est omorecompl exrules.Eachl
essoni nt
roducescertainrul
es,andtheserul
esdominat
e
thelesson.

(i
v)Quanti
ty.Asdiscussedabove,grammar-
tr
ansl
ati
onfai
lstoprovi
deagreatdealof
comprehensi
blei
nput.Thesmal lamountofcompr
ehensi
blei
nputi
nthemodelsentences,
the
readi
ngs,andexerci
sesis,mor eov
er,rarel
ysuppl
emented byteachert
alki
nt he tar
get
l
anguage.

()Af
v fect
ivef
il
terl
evel.I
nChapterII
I,i
twashy
pot
hesi
zedt
hatonewayt
oencour
agea
l
owf
il
terwastobe"tr
ue"totheInputHypot
hesis.

128
128
128
Grammar -
tr
anslat
ion v i
olatesnearl
yev er
ycomponentoft heI nputHy pothesis,and i
tis
therefor
epr edi
ctedt hatt hismethodwi l
lhav etheef fectofput ti
ngt hest udent"onthe
defensive"
.Studentsar eex pect
edtobeablet oproduceimmedi ately
,andareex pectedt
obe
full
yaccurate.Anxiet
yl evel
, i
thasbeenpointedout,i
salsor aisedf orsomest udentswhoare
l
essi ncl
inedtowardgr ammarst udy(under
-users)
,aspointedoutbyRi ver
s,1968.

(
v )Tool
i sforconv ersat
ionalmanagement.Grammar-
tr
ansl
ati
onmakesnoatt
empt
,
ex
pli
cit
lyori
mpli
cit
ly,
tohelpstudentsmanageconver
sati
onswit
hnat
ivespeaker
s.

(
b)Lear
ning

Grammar -
translati
on implici
tl
y assumes t hat consci
ous contr
olof gr ammar is
necessaryf ormast ery.In otherwor ds,learning needs to pr
ecede acqui siti
on.This
assumpt i
onnecessi tatesthatalltargetst r
ucturesbei ntr
oducedandex pl
ained.Therei s,
ther
efore,nolimitat
ionoft hesetofr ulestobel earnedtothosethatarelearnable,port
able,
and noty etacqui r
ed,assuggest ed in ChapterI V.Thereisno attemptt o accountfor
i
ndivi
dualv ari
ati
oni nMoni toruse,nori sthereanyat t
emptt ospeci
fywhenr ul
esar et
obe
used,theimpl i
citassumpt i
onbeingt hatallstudentswi l
lbeabletouseal lther ulesal
lthe
ti
me!

(
c)Summar
y

Grammar-
transl
ati
on,i
ftheaboveanal
ysisiscor
rect
,shoul
dresul
tinver
ylowamount s
ofacquiredcompetence;whatcomprehensi
bleinputi
savail
abl
efacesahighaffect
ivef
il
ter
,
andlearningi
svastl
yoveremphasi
zed.

2.AUDI
O-LI
NGUALI
SM

Herearethecommonf eatur
esofaudi
o-l
inguallanguageteachi
ng.Again,t
heremaybe
subst
anti
alvar
iati
oninpracti
ce.Thel
essontypical
l
ybegi nswithadi al
ogue,whichcontai
ns
thest
ruct
uresandv ocabul
aryofthel
esson.Thest udentisexpectedtomi micthedial
ogue
and

129
129
129
eventual
lymemor i
zeit(ter
med" mim-mem").Oft
en,theclasspract
icesthedial
ogueasa
group,andtheninsmall
ergroups.Thedial
ogueisfol
lowedbypatter
ndr i
l
lonthestruct
ures
i
ntroducedinthedial
ogue.Theaim ofthedri
l
listo"str
engthenhabi
ts"
,tomakethepat t
ern
"aut
omat i
c".

Lado(
1964)notesthataudi
o-l
i
ngualpat
ter
ndri
ll
sfocust
hestudent
s'attent
ionaway
fr
om thenew str
uct
ure.Forexample,thest
udentmaythi
nkheisl
earni
ngv ocabul
aryinan
exer
cisesuchas:

That
'sa .(key,kni
fe,
penci
l
,et
c.)(
cuedbypi
ctur
es,
asi
nLado
andFr
ies,1958)

buti
nreal
i
ty,
accor
dingt
oaudi
o-l
i
ngual
theor
y,t
hest
udenti
smaki
ngt
hepat
ter
naut
omat
ic.

Therear
efourbasi
cdr i
llt
ypes:si
mplerepeti
ti
on,subst
it
uti
on (asint he ex
ample
above),tr
ansf
ormat
ion(e.
g.changi
ngandaff
ir
mativesent
enceintoanegat
ivesent
ence),
and
tr
anslati
on.

Followi
ng dr i
l
l,some audi o-
li
ngual classes provi
de ex planat
ion. According to
proponent sofaudio-
li
ngual
ism,t heexplanati
onisadescript
ionofwhatwaspr acticed,nota
prescri
pt i
on ofwhatt o say .The "r
ules"pr esent
ed ar
et heref
ore nott o be consi der
ed
i
nst r
uctionsonhowt operf
or m.Theex planat
ionsecti
onisconsideredopt i
onal
,since,inour
terms,itis"l
anguageappreciati
on".

(
a)Requi
rement
sforopt
imali
nput

(i)Compr ehensi
ble.I
tcanbemai ntai
nedthataudio-l
ingualmethodol
ogydoespr ovi
de
compr
ehensibleinput.Thedial
oguesandpat ter
npr acti
cear ecert
ainlyunder
standableby
moststudents,alt
houghsomet heori
stshavesaidt hatinear l
ypartsofal essonact ual
compr
ehensionisnotnecessary
,thatpurel
ymechanicaldri
ll
isuseful
.

(i
i
)Interesti
ng/r
elevant
.WhileLado(
1964)advisesthatthedial
oguecont
ain"usef
ul"
l
anguage,thatitbeage-appropr
iat
eandnat
ural
,mostdi al
oguesfal
lfarshor
tofthemarkof
t
rueint
erestandr el
evance.Mostpatt
ernpr
acti
ce,ofcourse,makesnoattemptt
omeett hi
s
r
equir
ement .

130
130
130
 (i
i)Notgr
i ammat i
call
ysequenced.Therei
sacl earsequencei naudi
o-l
i
ngualteachi
ng,
basedusual
lyonl i
ngui
sti
csimpl i
cit
y,butal
soinfl
uencedbyf requencyandpredicti
onsof
di
ffi
cul
tybycont
rasti
veanaly
sis.Asisthecasewit
hgrammar -t
ranslati
on,t
heenti
relessonis
dominat
edbythe"str
uct
ureoftheday ".

(i)Quant
v i
ty.Whi l
eaudio-l
ingualt
eachingi scapabl eoff il
l
inganent i
recl asshourwi th
aural-
orall anguage,i tisquitepossibletoar guet hataudi o-l
ingualism doesnotmeett his
requir
ementaswel lasothermet hods(seebel ow) .Whi l
ethepr esent ati
onofadi alogue,f or
exampl e,mayt akeupaf ul
lperi
od,studentsspendv eryli
ttl
eoft hi
st i
mef ocussingont he
message, whichispr esentedoverandov er.Thegoali st hememor i
zationofthedi alogue,not
thecompr ehensionofamessage.Pat t
ernpracticemayal sobecompr ehensibleintheor y,but
studentspr obablydonotat tendtomeani ngaf terthef ir
stfewr epetiti
ons( Lee,McCuneand
Patton,1970) .Indeed,accordingt osomepr actiti
oners,theideabehi ndpatter npracticeist o
avoidmeani ngal t
oget her
.Forbot hdial
oguesandpat ternpractice,theentirehourmi ghtbe
spentwi thj usta f ew sentences orpat terns,as compar ed t ot he wi de varietyr eal
communi cationgives.

()Af
v fecti
vefil
terlevel
.Audio-l
ingualteachi
ngv i
olat
esseveralaspectsoftheI nput
Hy pot
hesis:producti
onisex pect
edimmedi atel
y,andisexpectedt
obeer ror-
fr
ee.Over
-useof
dril
landr epeti
ti
on,proceduressuchasnotal l
owingstudentsaccesstot hewri
tt
enwor di n
earlyst
agesmayal soaddt oanxi
ety(see,forexample,
SchumannandSchumann( 1978)pp.
5-6).

(v)Tool
i sf
orconversati
onalmanagement .Audi
o-li
ngual
i
sm doesasl ightl
ybetterj
obin
thiscategor
ythandoesgrammar -t
ranslat
ion,
ast hedi
aloguesdocontainmat eri
althatcanbe
used t oinvit
einputand t o contr
oli t
s qualit
y.The appli
cabil
it
y ofdi alogues tof r
ee
conv er
sati
onandt ogenui neconversati
onalmanagementmaybel i
mited,howev er.Most
dialoguesareact
ual
lyscr
ipts,andarenotdesignedtobeusedt onegoti
atemeani ng.

131
(
b)Lear
ning

Theoreti
cal
ly,consci ousl earningi snotanex plicitgoalofaudi o-
li
ngualism.Thegoal ,
rather,i
stohavethest udentov er -
lear nav ari
etyofpat t
er nstobeuseddi rect
lyinperformance.
I
npr act
ice,however,audi o-li
ngualt eachingof tenr esul tsininductivel
earning,thest udent
attemptingtoworkoutaconsci ousr uleont hebasi soft hedialogueandpat ternpracti
ce, with
theex pl
anati
onsect i
onser vi
ngt oconf i
rm ordisconf ir
m hi sguess.Ther eisthusno ex plici
t
attemptt orestri
ctlearningt or ulest hatar el earnabl e,portableandnoty etacquired, nor
i
st her
eanyat temptt oencour ager uleuseonl yi ncer tainsituat
ions.Despitethef actt hat
patternpract
iceattempttof ocusst udentsof frules,ther equir
ementofcompl etecorrectness
probablyencouragesMoni toruseatal ltimes.

(
c)Summar
y

Someacqui sit
ionshoul dr esultfrom t
heuseoft heaudi o-
li
ngualmethod,butnowher e
nearwhati sobtainablewi t
hot hermet hodsthatpr ov
ideal argeramountofcompr ehensibl
e
and i nteresti
ng/relevanti nputwi thal ow fi
lt
er .The dietofdi alogues and patt
er ns wi
ll
occasi onallybeunder stoodandbeati+1,andmayt huscausesomeacqui si
ti
on.Iflear
ned
accor dingt oplan,student swi l
lendupwi thast ockofsentencesandpat ternsthatwi l
lbeof
occasi onalusei nconv ersat
ion,andal soservetheconv er
sationalmanagementgoal , t
osome
extent .Inductivelear ning isi mpli
citl
yencour aged (which mayr ai
se t
he f i
lt
erf orsome
deduct ivelearners;seeChapt erV),butnoat t
empti smadet olimitwhichr ul
esar et obe
l
ear nedorwhent heyar et obeappl i
ed.

3.COGNI
TIVE-
CODE

Cognit
ive-
codebear ssomesimilari
tytogrammar-t
ransl
ati
on,butalsodiff
ersinsome
ways.Whi l
ethegoalofgr ammar-t
ransl
ationisbasi
cal
l
yt ohelpst
udentsreadlit
erat
ureinthe
tar
getlanguage,cogniti
vecodeattempt stohelpthestudentinallf
ourskil
ls,speaki
ngand
l
ist
eninginaddi t
iontoreadingandwr i
ti
ng.Theassumpt i
onsaresimil
ar,however,i
nsofaras
cogni
tivecodepositsthat"compet
encepr ecedes

132
132
132
perf
ormance".Int hi
scase, "
compet ence"isnotthetaci
tknowledgeofthenati
vespeaker,
as
ori
ginal
l
ydefinedbyChomsky( 1965),butisconsciousknowledge.Cogni
ti
vecodeassumes,
asment i
onedi nChapterIII
,that"oncet hestudenthasapr operdegreeofcogni
ti
vecontrol
overthestr
ucturesofal anguage,facil
itywil
ldevel
opautomat i
cal
lywitht
heuseoflanguage
i
nmeani ngf
ul si
tuati
ons"(Carr
oll,
1966, p.102).I
notherwords,l
earni
ngbecomesacquisi
ti
on.

Asi ngrammar -
tr
anslation,t
hel essonbeginswi t
hanex planat i
onofther ule,andt hi
sis
oftendone,i nf orei
gnlanguagesi tuati
ons,inthestudents'fi
rstlanguage.Ex erci
sesf oll
ow,
andt hesear emeantt ohelpt hestudentpr act
icetheruleconsciously.Inotherwor ds,Moni t
or
use is act i
vely promoted.Ex er
cises aref oll
owed by activi
ties label
led "communi cati
ve
compet ence".Thi sterm hasbeenusedi nthel i
ter
atureinsev eralway s;incogni ti
ve-code
l
iteraturei tappearstobesy nonymouswi th"fl
uency".Theseact i
vi
tiesprovidet hepracticein
meani ngf ulsituati
onsr ef
erredt oi nt hequot efrom Carrol
labov e,andi ncl
udedi al
ogues,
games, roleplayingacti
vit
ies,etc.

(
a)Requi
rement
sforopt
imali
nput

(
i) Compr ehensible.The expl anati
on and ex ercise section,as ist he case wi th
grammar -
translat
ion,willprovideveryl i
tt
lecompr ehensibl
ei nput,asthefocus,atallti
mes,i s
on form and notmeani ng.The " communi cati
ve compet ence"section ofcogni t
ive-code
promisestopr ovi
degr eateramount sofcompr ehensibl
ei nput ,
butthispotenti
alisdiminished
i
fact i
vi
ti
esar el i
mitedbyt hedesi r
et ocont ext
ualizethe" ruleoft heday".Thispractice,as
cl
aimedi nChapt erIII
,limitst hestructuresused( whichmaydenyt hestudentt hei+1he
needs),l
imitswhatcanbedi scussed,anddi st
urbsthenat uralnessofthecommuni cat
ion.

(
i)I
i nterest
ing/
rel
evant.Thisdepends,ofcourse,on t
heact i
vi
ti
eschosen f
ort he
communi cat
ive competence secti
on.Regardl
ess ofwhatis chosen,however
,the goal
remainsthelearningofaspeci f
icstr
uctur
e,andbecauseofthisiti
snearl
yimpossibl
et o
sati
sfytheForgett
ingPri
ncipl
e.

133
133
133
 (i) Not gr
i
i ammati
cal
ly sequenced. Cogni
ti
ve-
code,li
ke grammar
-t
ransl
ati
on,i
s
sequenced,
andt
hestr
uct
ureoft
hedaydomi natesal
lpart
softhel
esson.

(i
v)Quant i
ty.Thanks tot he communi cati
ve competence secti
on,ther
ei s greater
quant
it
yofcompr ehensibl
einputincogni t
ive-
code,ascompar edtogr ammar-t
ransl
ation.It
doesnot,however,li
veupt otheidealofaclassfullofcomprehensibl
einputwit
ht ot
alfocus
onthemessage,si ncethecommuni cat
ivecompetencesecti
oni sonlyapartofthepr ogram
andevenhere,t
hef ocusisonf or
m.

(
v)Aff
ecti
vefil
terlev
el.Err
orcorrecti
ononal lout
putispartofmostcogni
tiv
e-code
cl
assr
ooms,student
sar e expect
ed to produce r
ightaway,and itisex
pected t
hatt hi
s
pr
oducti
onwi
llbeaccurat
e.Thispredi
ctsahighfi
lt
erformanystudent
s.

(
vi)Tool
sforconversati
onalmanagement.Ther
eisnoannouncedat
tempttoprov
ide
thi
s,butiti
squit
epossi
blet hatsomeacti
vi
ti
esinthecommuni
cati
vecompet
encesect
ionwil
l
provi
desomeoft heset
ools.

(
b)Lear
ning

Asist hecasewith gr
ammar -
tr
anslat
ion,the assumpti
onofcogniti
ve-code ist
hat
consciouslearni
ngcanbeaccompl i
shedbyev ery
one,thatal
lrul
esarelearnable,andt
hat
consciousknowledgeshouldbeavail
abl
eatal lti
mes.Wecanonl yconcl
udet hatcognit
ive-
codeencour agesover
-use ofthe Monit
or,unlessallrules"
fade away
" assoon ast he
str
ucturesbecomeautomatic.

(
c)Summar
y

Cogniti
ve-code shoul
d provi
de gr
eater quanti
ti
es of comprehensi
bleinputt han
grammar-
translati
ondoes,andhencemor eacquisi
ti
on,butdoesnotcomeneart oful
fi
lli
ng
thepot
enti
aloft hecl
assroom.Asingrammar-
tr
anslati
on,l
earni
ngisov
er emphasi
zed.

134
134
134
4.THEDI
RECTMETHOD

Thet er
m" di
rectmet hod"hasbeenusedt orefert omanydi f
fer
entapproachest
o
secondlanguageteaching.Iwil
luseitheretoref
erspecif
icallyt
odeSauzé'smethodandit
s
presentdayv ersi
ons,namel yPucci
aniandHamel '
smet hodf orFr
ench(see Langueet
Langage),andsimil
arversi
onsforSpanishdevel
opedbyBar cia.

Here aret he char act eri

sti
cs oft he di r
ectmet hod,as Iunder stand i t
.First,all
discussion,allclassroom language,i st hetargetlanguage.Thi si ncludest hel anguageoft he
exercises and t eachert alkused f orcl assroom management .The met hod f ocusseson
i
nduct iveteachingofgr ammar .Thegoaloft heinstr
uct i
oni sf ort hest udent stoguess,or
wor kout ,t
her ulesoft hel anguage.Toai dini nducti
on,t heteacherasksquest i
onst hatare
hopeful l
yinteresti
ngandmeani ngful,andt hest udents'responsei st henusedt opr ovidean
exampl eoft het argetstruct ure.Ifthisi swelldone,itcangi veadi rectmet hodsessi onthe
moodofaconv ersati
oncl ass.Letmer epeatmyf av orit
eex ampl ef r
om anear l
ierpaper
(Krashen,1980) ,onet hatoneofmyt eachersusedi nadi rectmet hodFr enchcl ass.Thegoal
ofthisex erci
sewast oteacht heconj unction"bienque",andt hefactt hatit
spr esencer equir
es
thefollowingv erbtobesubj unct i
ve:

Teacher:Fai
t-
ilbeauauj
our'dhui
?
Student
:Non, I
lnefai
tpasbeauauj our
d'hui.
Teacher:I
rez-
vouscependent&agrave;l
apl agependantleweek-end?Student:Oui
,j'
ir
ai
cependent&agrave;l
aplagependantleweek- end.Teacher
:Ir
ez-v
ous&agr ave;
laplagebi
en
qu'i
lnefassepasbeau?St udent
:Oui,j
'
ir
ai &agrave;l
aplagebienqu'
ilne..
.

Myteacherusedthi
sparti
cul
arexampleonadet
ermi
nedbeach-
goer
,andgener
all
ytr
iedt
o
t
ail
orquesti
onstostudent
s'i
nter
ests.

Thedir
ectmethodinsist
sonaccuracyanderrorsarecorr
ectedinclass.Aft
ersever
al
exchangesofthesortgivenabove,whentheteacherconsi
dersthatenoughex ampleshave
beengiven,
therul
eisdiscussedandexpl
ainedi
nthetargetl
anguage.

(
a)Requi
rement
sforopt
imali
nput

(
i)Compr
ehensi
ble.Thedi
rectmet
hod,
wit
hit
sinsi
stenceont
heuse

135
135
135
ofthetargetlanguageatallti
mes,prov
idesagreatdealofcomprehensi
blei
nput
.Asisthe
casewi t
ht heNat ur
alApproach,t
heentir
eperi
odi sfi
ll
edwithtargetl
anguageusewit
ha
vari
etyoftopi
csandst r
uctur
esutil
i
zed.

(i
i)I
nter
esting/rel
evant.Asment i
onedabov e,t
hereisanatt
emptt omakethelanguage
useintheclassroom ofsomei nter
esttothestudents.Thegoaloft helesson,however,is
grammarteaching,andasdi scussedearl
ier
,thisput sheavyconstraint
sonwhatcanbe
di
scussed.Discussi on is alway s meani
ngful,but isr ar
ely genuinel
y communi cat
ive.
Accor
dingtoPucci aniandHamel 'smanual,
sentencessuchas:

Est
-cequev otr
epantal
onestvi
euxouneuf?(
198)Mangez-
vous
descarott
es?(236)
Quiprepar
eledinerdansvot
refamil
l
e?(
237)
Est
-cequev ousvousrasezt
ouslesmat
ins?(
297)

arer ecommendedt ohel

pt he st
udentinduce vari
ouspoi nt
sofgrammar.Aswi th ot
her
met hodsthatrelyoncont
ext
ual
izat
ion(seee.g.di
scussionoftheSi
lentway)
,therequi
rement
thatalldi
scussionembedagrammarpoi ntmakesthisrequi
rementhardtomeet.

(i
i
i)Notgr
ammati
cal
l
ysequenced.Thedi
rectmet
hod i
sst
ri
ctl
ysequenced,whi
ch
di
stor
tseff
ort
satr
eal
communi
cat
ion.

(
iv)Quanti
ty.Asmenti
onedabov
e,thedi
rectmet
hodmeetsthi
srequi
rementaswel
las
anycl
assroom methodcan,
fil
l
ingt
heenti
rehourwit
hcompr
ehensi
blei
nput.

(v)Aff
ectiv
ef i
lt
erlevel
.Theinsi
stenceongrammati
calaccur
acyatver
yearl
ystages,
theuseoferrorcorr
ecti
on,andthegrammat i
calf
ocusoft
hecoursemaycauseanx
ietyanda
highf
il
terf
orallbutthemostdedicat
edMoni t
oruser
.

(v
i)Toolsf
orconversati
onalmanagement.St
udent
saregi
venthetoolsforint
eract
ion
i
nthecl assr
oom i
nthet argetlanguage-
-t
heyaresoonabl
etoini
ti
atediscussi
onwi ththe
t
eacherandaskquest
ions

136
136
136
aboutgr
ammar .Someofthi
sconv er
sati
onal
,orbet
ter,"
classr
oom competence"wil
lbeuseful
ontheoutsi
de,butsomewillnot
.Thereisnoexpli
citgoalofprov
idi
ngtoolsforconver
sati
on
wit
hamor ecompetentnat
ivespeaker
.

(
b)Lear
ning

The di
rectmet hod presumes thatconscious cont
rolis necessar
yf oracqui
sit
ion,that
consci
ousknowl edgeofgr ammarcan be accessed atal ltimes,and byallst
udents.It
demandsf ul
lcontr
oloflat
e-acqui
redst
ruct
uresinoralpr
oducti
onf r
om thever
ybegi
nning(e.g.
gender
),andmayt husencourageover-
useofthegrammar .

(
c)Summar
y

Thedi r
ectmet hodpr ovi
desgr eateramount sofcompr ehensibleinputt
hanmanyofi ts
compet itors.Itremai ns,howev er,grammar -based,andt hisconstrainsitsabil
i
tytopr ovide
tr
ulyi nteresting messages,and l eads t o over-use oft he Monitor.The directmet hod,
accordingt oi nfor
malr epor t
s,hasbeenv erysuccessf ulwithcer t
ainpopulati
ons,among
studentswhohav eintri
nsicmot i
vationf orlanguagest udyandwhobel ievethatthestudyof
consciousgr ammari sessent ial
.Fort hesest udents,theinductiv
est udyofgr ammari si n
i
tselfinteresti
ng,andpr ov i
desal lthei nterestnecessar y
.Inot herwords,grammari ssubject
mat t
er.Acqui sit
ion,SLAt heorypr edicts,comesf r
om thet eachertalkusedt opresentt he
grammar .(SeeChapt erIVf ordiscussion, andKr ashen,1980.)

5.THENATURALAPPROACH

TheNat uralApproachwasdev el
opedbyTr acyTer r
ellattheUniv
ersityofCali
forniaat
Ir
vineforforeignlanguageinst
ruct
ionattheuniversityandhi ghschooll
evels.Whileori
ginall
y
developedindependentlyof"Monit
orTheor y
",i
tslaterdevelopmentandar t
iculat
ionhavebeen
i
nfluencedbyt hesecondlanguageacquisit
iontheorypr esentedinthi
sv olume.Themet hod
canbedescr ibedbyt hefol
lowi
ngpr i
nci
ples:

137
137
137
 1. Cl
asst
imei
sdev
otedpr
imar
il
ytopr
ovi
dingi
nputf
oracqui
sit
ion.

2. Thet eacherspeaksonlythetargetlanguageinthecl
assroom.
Studentsmayuseei therthefi
rstorsecondlanguage.Ift
hey
chooset orespondinthesecondl anguage,t
heirer
rorsarenot
correct
edunl esscommuni cat
ionisseriousl
yimpair
ed.

3. Homewor
kmayincl
udefor
malgrammarwor
k.Er
rorcor
rect
ion
i
sempl
oyedi
ncorr
ect
inghomework.

4. Thegoal
softhecour seare"semantic"
;acti
vi
ti
esmayinvolvethe
useofacert
ainstructur
e,butthegoalsaretoenabl
estudentsto
tal
kabouti
deas,performtasks,andsolv
eproblems.

(
a)Requi
rement
sforopt
imali
nput

(
i)Compr ehensi
ble.Theent
ir
egoalofclassroom pract
iceintheNatur
alApproachi
sto
pr
ovi
de compr ehensibl
ei nput.Natur
alAppr oach teacher
s ut i
l
ize r
eali
a,pict
ures,and
st
udent
s'prev
iousknowl edgetomakethei
rspeechcompr ehensibl
ef r
om t
hefir
stday.

()I
i
i nt eresti
ng/relevant.Thisisaser iouspr oblem foraforei
gnlanguagecl ass.Nat ural
Approach at tempt st o captur e st
udents'i nt
erestbyusi ng whatTer r
ellterms " Affectiv
e
Acquisit
ionAct i
vi
ties",adapted from Chr istensen,thatencour age di
scussion oft opicsof
personali nterestt ot hest udents( e.
g."Supposey ouar eaf amousper son,andt herei sa
newspaperar ti
cleabouty ou.Tel latleastonet hingabouty our
selfwhi
chi sment ionedi nthe
art
icl
e..
.").Int heear l
yst agesoft heNat uralAppr oach,classroom di
scussionf ocusseson
personalinf ormat i
on, thegoalbei ngtoestabl i
shagr oupfeeli
ng.Lat
er,st
udent sdiscusst heir
pasthistories, andev entuall
yt heyareablet otalkaboutt hei
rhopesandplansf orthef uture.

(
ii
i)Notgrammat i
cal
lysequenced.Thefocusoft
heclassi
snotonthepr
esentat
ionof
grammar .Ther
ei satendencyforcert
ainstr
uctur
estobeusedmoreoft
enincert
ainstages,
butther
ei snodeli
ber
atesequencing.

(
iv)Quant
it
y.Si
ncet
heent
ir
ecl
assper
iodi
sfi
l
ledwi
thcompr
ehensi
ble

138
138
138
i
nput,t
heNatur
alAppr
oachmeet
sthi
srequi
rementaswel
lasanyf
orei
gnl
anguaget
eachi
ng
methodcan.

(
v)Af fect
ivefil
terlev
el.Si
ncetheNaturalAppr
oachat t
emptt oremai n"t
rue"totheInput
Hypothesis,manysour cesofanx ietyarereducedorel i
minated.St udentsdonothav et o
produceint hesecondl anguageunt i
ltheyf
eeltheyareready.Errorcorrect
ionforformisnot
donei nthecl assroom.Al so,anat t
emptismadet odiscusst opicsthatarei nt
erest
ingt o
st
udents.Thi spredict
sl owerfi
lt
erstr
engththanmostothermet hods.

(vi)Toolsforconv er
sat
ionalmanagement.Somet ool
sforconversati
onalmanagement
areprovidedi nthef or
m ofv er
yshortdial
ogues,designedtohelpst udentsconversewith
nati
vespeaker sonpr edi
ctabl
eandf r
equenttopi
cs.Also,st
udentsareintroduced,r
ightf
rom
thebeginning,tophrasesandex pr
essi
onsthatwillhel
pthem cont
roltheteacher'si
nput(e.
g.
"Idon'
tunder st
and",
"Whatdoes mean?",etc.
).

(
b)Lear
ning

TheNat ur alAppr oachi sdesi gnedt obeconsi stentwithwhati sknownofMoni t

or
funct
ioning.Theabsenceofer rorcorr ectionint hecl assroom isarecogniti
ont hatt herear e
constrai
ntsonwhent heconsci ousgr ammari sused:st udentsareex pectedt out il
izet he
Monitoronl yathome,whent heyhav et i
me,whent heyar efocussedonf orm,andwhent hey
know,orar el earning,ar ul
e.Att heuni versi
tyl evel,grammarhomewor ki sassi gnedt o
every
one,buti tisconcei v
ablet hatt heNat uralAppr oachcanbeadapt edf orv ari
ationsi n
Monitoruse, withv aryi
ngamount sofhomewor k,ordi ffer
enttypehomewor kassi gnment sfor
under,oropt i
maluser s.Whi l
el i
ttl
eex per i
ment ationhasbeendonewi t
hchi ldren,SLAt heory
predi
ctst hatyoungerchi ldrenwoul dnotpr ofi
tfr
om gr ammarhomewor k,whileolderchi ldren
andadolescent smi ghtbeabl et ohandl el i
mi t
edamount s.(Formoredet
ail,seeTer rel
l,1977. )

139
139
139
(
c)Summar
y

TheNaturalApproachmakesadel i
berat
eeffor
ttofi
tallr
equi
rementsf
orbothLearning
andAcquisi
ti
on.It
sonl yweakness,accordi
ngt oSLAtheory
,isthati
tremainsaclassroom
method,andforsomest udentst
hisprohibi
tsthecommunicati
onofinter
est
ingandrelevant
topi
cs(seedi
scussionbelow).

6.TOTALPHYSI
CALRESPONSE

Thisuni quemet hodwasdev el

opedbyJamesAsher ,andisdescr i
bedinmanyofhi s
j
our nalpaper sand hi sbook( Asher,1977a) .Tot alPhy si
calResponse,orTPR,consi sts
basicall
y ofobey i
ng commands gi ven byt he instructorthatinvolve an over
tphy sical
response.The i nstructor,forex ample,say s" stand up"and t he class stands up.The
commandsbecomemor ecomplexast heclasspr ogresses,andAshercl ai
mst hatiti
squi t
e
possibletoembedv astamountsofsy ntaxintot hef orm ofacommand.St udentsspeakonl y
whent heyare" r
eady ",whichusual l
yoccur satar ound10hour sofinstr
ucti
on,andconsist sof
studentcommands.I nthetypi
cal TPRcl ass(asdescr i
bedbyAsher ,Kusudo,anddel aTorre,
1974) ,the firstf ew mont hs (45 hour si nt his case)woul d consistof70% l i
stening
compr ehension( obey i
ngcommands) ,20% speaki ng,and10% r eadingandwr it
ing.Asher
(1977b)list
st het hreepr i
nci
plesoft heTPRsy stem:

(
i)Delayspeechf r
om student
sunt
ilunder
standi
ngofspokenl
anguage"
hasbeen
ex
tensi
vel
yinter
nal
ized"(
p.1041).

(
ii
)"Achieveunderst
andi
ngofspokenl
anguaget
hroughut
ter
ancesbyt
hei
nst
ruct
ori
n
t
hei
mperat
ive"(p.1041)
.

(i
ii
)"Expectthat,atsomepointintheunder
standi
ngofspokenl
anguage,st
udent
swi
l
l
i
ndi
cat
ea'readiness't
ot al
k"(
p.1041)
.1*

(
a)Requi
rement
sforopt
imali
nput

()Compr
i ehensi
ble.TPRmeetsthi
srequi
rement
.Thetotalphysi
calresponserequi
red
ofthestudenti
s,ineff
ect,amani
fest
ati
onofhiscompr
ehensi
onoft heteacher'
sutt
erance.I
t
can,infact,bearguedt hataTPRisnotnecessaryf
orcompr ehensi
onorf orprogressin
second

* Super
scr
iptnumber
sref
ert
oNot
esatendofchapt
ers.

140
140
140
l
anguageacqui siti
on,butmer el
yshowst hattheinputhasbeenunder stood.Asher'sown
researchsuppor t
st hev iewthattheuseoft heTPRi snotessent i
al.Aser iesofstudiesusing
chil
dren( Asher,
1966; AsherandPrice,
1967)andadul ts(Kunihi
raandAsher ,1965;Asher,
1965, 1969)showst hatstudentswhomer elyobserveaTPRdoaswel last hosewhoper form
TPR'sont eststhatdemandaTPR.Bot hgroups,thosewhoobser vedTPR' sandt hosewho
perfor
medt hem, outperfor
medst udentswhowr otetheiranswersont ests.Thissuggestst hat
Asher'ssecondpr inciplemaynotbenecessar y,
butmaybesi mplyanef f
ect i
vedevi
cet ofocus
studentsont heinputandt okeepthem act
ivel
yinvolved.

(i
i)Int
eresti
ng/r
elevant.Thenovelt
yandf r
eshnessoftheTPRt echniquepr
obablydoes
agreatdealtomaket hecl assexper
ienceint
eresti
ng.I
tmaybedi f
fi
culttor emai
ninter
est
ing
i
fonehol dstot herequir
ementofpr oducingimperati
ves100%oft het i
me,howev er(
Asher
nowhererecommendst his).

(i
i
i)Notgrammat ical
l
ysequenced.Accor di
ngtoAsher
'sdescripti
on,eachlessondoes
haveagr ammati
calfocusinTPR.I notherwords,commandscontex t
uali
zev ar
iouspoint
sof
grammar.Asdiscussedearli
er,thi
scanhindereffor
tst
omeetr equir
ement2abov e.Ther
eis
nothi
nginher
enti
nt heTPRappr oachthatdemandsagrammat i
calfocus,however.

(
iv)Quanti
ty.TPRcanf
il
lanentir
ecl
assper
iodwit
hcomprehensi
blei
nputi
nthef
orm
ofcommands.Itthushast
hepotent
ial
ofmeet
ingt
hisr
equir
ementf
ull
y.

(v)Affect
ivefi
lterl
evel.TPRmakesonev eryimpor
tantcont
ri
but i
ontoloweri
ngstudent
anxi
ety:student
sar enotaskedt oproduceint hesecondlanguageunt i
ltheythemselves
deci
det heyareready.Theyar e,i
not
herwor ds,al
lowedasil
entperiod.Asherdoesnotstate
expl
i
citl
ywhet hererrorcorrecti
ononearlystudentout
putisrequir
edi nTPR;thismayv ary
fr
om teachertoteacher.I
thasbeenpointedout,

141
141
141
however
,thatt
henecessi
tyofpr
oduci
ngov
ertphy
sical
responsesr
ightawaymaypr
ovoke
anxi
etyi
nsomestudents.

(v
i)Tool
sforconv
ersat
ionalmanagement
.Thi
sisnoex
pli
citment
ionoft
hisi
nAsher
's
paper
s.

(
b)Lear
ning

Theassumpt ionofTPRi st hatgrammarwi llbelearnedi nductively,thatis,student swil

l
wor koutt hecorrectfor m oftheruledur ingtheclassact iv
ity.I
nt ermsoft hetheor ypr esented
i
nt hi
sbook,t hiscanbei nter
pretedascl aimingthatmuchoft hegrammarwi lbeacqui
l red
and/ orinducti
velylear nedi nthet echni calsenseofi nduct i
vel earni
ng.( SeeChapt erIVf or
discussionoft hedi fferencebet weeni nducti
velear ni
ngandacqui si
tion.)Theemphasi son
l
isteningcompr ehensionandt hedel ayofspeechwi l
l,ini tself
,pr eventmuchmi suseof
consciousl earni
ng:st udent swilltendnott omoni tort heiroutputf orf ormi ninappr opri
ate
circumst ancesandt heywi l
lnotuser ulesunsui t
edf orMoni torusei ftherei slessdemandf or
product i
on.

(
c)Summar
y

SecondLanguageAcqui si
ti
ont heorypr edi
ct sthatTPR shoul dr esul
tinsubst ant
ial
l
anguageacquisi
ti
on,andshoul dnotencour ageoveruseoft heconsciousMoni tor.Theuseof
TPRinsurestheacti
veparti
cipati
onofst udents,helpst heteacherknowwhenut terancesare
under
stood,andalsoprovi
descont extstohel pstudent sunderstandthelanguaget heyhear.It
may failto complet
ely sati
sfyt he interesti
ng/relevantr equir
ement,f i
rst
,since i tisa
cl
assroom method,andsecond,becauseofconst raintsimposedbyt hecont i
nuoususeof
i
mperativ
esandt hegr ammat i
calf ocusofl essons.I tshould,howev er,dofarbet terthan
methodssuchasaudio-li
ngualandgr ammar -t
ransl
at i
on.

7.SUGGESTOPEDI
A

From whatIhavereadinthesour
cesavai
labl
etome,t
he"cl
assi
c"Suggest
opediacl
ass,
asconduct
ed in Lozanov'
sInsti
tut
e ofSuggestol
ogyi
n Sof
ia,Bul
gari
a,consist
soft he
fol
l
owi
ng.Coursesaregivento

142
142
142
smallgr
oups,around12student
satati
me,andarei
ntensiv
e, meeti
ngforfourhoursperday
foronemonth.Eachfour
-hourcl
ass,
accor
dingt
oBancroft(1978),
consist
soft hr
eeparts:

1. Revi
ew, donevi
atradi
ti
onal conv
ersati
ons,games,pl
ays,
etc.I
t
mayincludesomeexercisesanderrorcorr
ecti
on,
butdoesnotincl
ude
t
heuseofal anguagelaborpatter
ndr i
ll
.

2. Presentati
onofnewmat er
ial.Newmat eri
ali
sint
roducedinthe
form ofdialoguesbasedonsi tuati
onsfamil
iart
othestudents.
Bancroftnot esthat"newmat eri
ali
spresentedi
nasomewhat
tradi
ti
onal way ,wit
hthenecessarygrammarandt r
anslat
ion"(p.
170).Thedi aloguesareverylong.Accordi
ngtoBushmanand
Madsen( 1976) ,theyrunf
rom 10t o14pages.

3. Thispor tionist he" trulyor i

ginal featur e"ofSuggest opedia
(Bancrof t,
p.170) ,andi si t
sel fdividedi ntot wopar t
s.Inthef i
rst
part,
theact iveseance, thedi aloguei sr eadbyt het eacher ,whi l
e
student sfollowt het ex tandengagei ndeepandr hyt
hmi cYoga
breathing.Theseact ivit
iesar eco- or dinat ed: "I
naccor dancewi th
thestudent s' breathing, thet eacherr eadst hel anguagemat eri
als
i
nt hef oll
owi ngor derandwi t
ht hef ollowi ngt i
mi ng: Bulgarian
(L1)translation( twoseconds) ;foreignl anguagephr ase( four
seconds) ;
pause( twoseconds) .Whi let hef or eignl anguage
phrasei sbei ngr ead, thest udent sr etai nthei rbr eathforf ourseconds,
l
ookatt heappr opriat epartoft het ex t,andment all
yrepeat
tothemsel vest hegi venphr aseorwor d-groupi nt heFL.
Concent rationisgr eatlypromot edbyt her etent ionorsuspensi on
ofbreat h"(Bancr oft ,p.171) .

Thesecondpar t
,label
ledthepassiveorconvertpar
toftheseance,invol
vesmusi c.The
centr
al activ
ityistheteacher'
sr eadi
ngofthedialogue"wit
hanemot i
onalintonat
ion"(Bancroft
,
p.171).Thest udents,"wit
hey esclosed,medi
tateont hetext
"whil
ebaroquemusi cisplayed.
Themusi calsel ecti
onsar especi f
ical
l
ychosent ocontr
ibutetoa" st
ateofr elaxati
onand
meditation..
.t hati s necessaryf orunconscious absorpt
ion ofthe language mat eri
als"
(Bancroft,p.172) .

Indi
scussi
ngadapti
onsoft
heSofiamet
hod,
Bancr
oftnot
est
hat"
threeel
ement
softhe
LozanovMethodareconsi
der
edessent
ialf
ort
hesyst
em t
oworkeff
ect
ivel
y:(
1)anat
tract
ive
cl
assroom (
wit
hsoft

143
143
143
l
ighti
ng)andapleasantcl
assr oom at
mospher
e;(
2)ateacherwi
thadynami
cpersonal
i
t ywho
i
sabl etoactoutt hemat eri
alsandmotivat
ethest
udentstolear
n;(3)ast
ateofr elax
ed
aler
tnessi
nthestudent
s..
."(p.172)
.

InSuggest
opedia,eachmemberoft hecl
assisgivenanewnameandr oletoplay,"
to
overcome inhibi
ti
ons"( p.170)
.Ot herSuggestopedi
at echni
ques and att
itudes wil
lbe
discussed bel
ow,aswewor kthr
ough theanalysisaccordi
ng toinputrequirementsfor
acquisit
ion.

(
a)Requi
rement
sforopt
imali
nput

(i)Compr ehensibl
e.Sev eralSuggest
opediapr
oceduresarespecif
ical
lydesignedtoai d
comprehensibil
it
yofi nput.I
nit
ialdial
oguesarebasedonsit
uationsfamil
iart
ot hestudent
,and
theuseoft hest udents'fir
stlanguageinPar tOneispart
ial
lyjust
ifi
edont hegroundt hatit
hel
pst hest udentconf i
rmt hathehasi ndeedunderst
oodt hetextpresentedint hetarget
l
anguage( Racle,1979,p.100).

(
i)I
i nteresti
ng/relev
ant.Thet opicsoft hedi aloguesar edesignednotonl yt obeof
i
nherentinter
est,butal sotobeofsomepr act
icalv al
ueandr elevanttostudents'needs.Ina
Suggestopediacoursedesi gnedt oteachFr encht oAngl ophonepubl i
cservant
si nCanada,at
thePublicServ i
ceCommi ssi
oni nOt tawa,theai m wast ot akeintoaccountbot hst udent
i
nterestandtheircommuni cat
iveneedsi ntheof fi
cesituati
on( Publ
icServi
ceCommi ssion,
1975).Also,Nov akov,ci t
ed( andt ranslat
edi ntoFr ench)byRacl e,1979,notest hat" Les
sit
uati
onspr ésentéessontt ypiques,réell
es,cont i
ennentunmessageetsontpr ochesde
l
'expéri
encedesél èves,cequi faci
li
teleuractiv
ité"(p.99).

(i
i)Fi
i l
terlevel.Whil
eSuggestopedi
aat temptstomeett heothergoal
sdi scussedboth
aboveandbel ow,itspri
mar yfocusandgr eat
estapparentsuccessishere.Practicall
yev
ery
feat
ureofSuggest opedia isai medatr elax
ingt hestudent
,reducinganxiet
ies,r emov
ing
mentalbl
ocks, andbuil
dingconfi
dence.Herearejustafewmor eexamples:

Thedesignofthecl assr
oom i
smeanttoproduce" apl
easantandwarm envi
ronment
"
(Publ
icSer
viceCommi ssion,1975,p.29)
.St
udentsar eseatedoncomf or
tabl
echairsi
na
cir
cleto"encourage i
nformalcontactand f
ree naturalcommunicat
ion"(Bushman and
Madsen,

144
144
144
1976,p.32).Thetr
adit
ionalclassr
oom,i
tisfel
t,"
cal
lstomindthefr
ust
rat
ion,f
ail
ure,and
art
if
ici
ali
tyofmanypr
eviouslearni
ngef
for
ts"(
BushmanandMadsen,
p.32)
.

Thespecialbreathi
ngex erci
seshaveast hei
rgoalbothincr
easedment alaler
tnessand
reducti
onoft ensi
on.Bancr of
tr eport
sthatAmer i
canadapti
onsofSuggest opediaalsouti
li
ze
physicalex
ercises(str
etchi
ngandbendi ng),and"mind-
calmingexerci
ses"
,inadditi
ontoYogic
breathi
ngtohel pst
udent sachi
ev ethedesi
redstateofrel
axedaler
tness.

Musi
cisalsousedasameansofl oweri
ng anxi
etyanddi mi
nishi
ngtensi
on,and
i
nduci
ngthestat
eofr el
axedal
ert
nessconsi
deredopt
imalforsecondl
anguageacquisi
ti
on
(
seeRacl
e,1980,
pp.73-74)
.

AnotherkeySuggest opedi
ci deaaimedatl oweringt hef i
lteri sthebehav ioroft he
teacher .Suggest opedi
aconsi der
st he"authori
ty"oft het eachert obev er
yi mportant( "
an
i
ntegr alpartofthemet hodandnotj ustadesirabl
echar acteri
sticoft heteacher";St
evick,
1980,p.238) .Thet eacher'sbehaviorismeantt obuil
dt hest udents' confi
dencebot hint heir
ownpot enti
alforsecondl anguageacquisi
ti
onandi nthemet hoditself;theteachershouldbe
confident,butnott yrannical,exerci
se f
ir
m ov er
-al
lcont r
olbutal so encourage student
i
niti
ative(forex cell
entdiscussion,seeStevi
ck,1980,Chapt ers2and18) .

(i
v)Notgr ammat i
cal
lysequenced.Ther ei sadel iberateattemptt oi ncl udeacer t
ain
amountofgr ammardur i
ngthef ir
stonemont hintensiv
ecour se(Racle,1978,p.95l i
ststhe
str
ucturescov eredforFrench).Itdoesnotappeart obet hecase,howev er,thatar i
gid
sequenceisfollowed.All
writ
ersonSuggest opediaIhav er eademphasi zet hatt hefocus,fr
om
theverybeginning,i
soncommuni cat i
on,andt hedi al
oguesdonotseem t of ocusonspeci fi
c
point
s of gr ammar . According to Bushman and Madsen," Dialogues ar er ambling
conversati
onsl oosel
yaggregated ar ound common t hemes,whi ch cov era gr eatdealof
ter
rit
orywithconsiderabl
ebuilt
-inredundancy "(p.33).Inourt erms,Suggest opedi aseemst o
dependont henetofgr ammaticalstructur
espr ovidedbysuccessf ul communi cat i
on.

()Quant
v i
ty.Suggestopediaseemstomeett hi
srequir
ementaswell
.Whil
ethereis
someexpl
anat
ioni nthefi
rstlanguage,t
hel
ongandv ari
eddial
oguedomi
natest
hesessi
on,
bot
haspurei
nputandasabasi sforcommuni
cat
iveuseoftheL2.

145
145
145
 (v)Tool
i sforconv
ersat
ionalmanagement .Thi
sisnotment ionedexpli
cit
ly,butmaybe
covered,since t
he dial
ogues att
emptt o be reali
sti
c.Texts used inthe Publ i
c Servi
ce
Commi ssioncourseinCanadawer eapparentl
ydesignedtoallowandpr omoteconv er
sat
ion
i
nPubl icServi
ceoffi
cesaswellaselsewhere.Thereisnoexpli
citment i
on,howev er
,ofgiv
ing
studentsthetool
stheyneedtoconversewithmor ecompetentspeakers.

(
b)Lear
ning

According t
o Bushman and Madsen ( 1976),"Contentprecedes f
orm.Accur ate
pronunciati
onandgr ammarar etocomei
nduecour se"(p.32)
.Whil
et her
eiserrorcor
rect
ion
and grammarex planati
on in par
tone ofeach lesson,grammaruse i n Suggest
opedia
apparentlydoesnotinter
fer
ewi t
hcommunicat
ion.

(
c)Summar
y

Whi l
eIhav eattemptedtof i
tt heSuggest opediasy st
em,asIunder standi t,intomy
schema,andhav eomi tt
edment i
onofsev eralaspect sofSuggest opediaphi l
osophyt hatits
pract
it
ionerswoul dundoubtedl
yconsi dertobev eryimpor t
ant,i
tappear sthatSuggest opedi a
comesv erycloset ocompletel
ymat chingther equirementsf oroptimalinput.Element st hat
Lozanovmi ghtconsidertoi
nvokehy per-orsuper -memor y,orthat"desuggest"li
mi t
ations,are,
i
n ourt erms,condi t
ionsthatlowert heaf fecti
v ef i
l
terand t hatallow thesubconsci ous
l
anguage acqui sit
ion syst
em t o oper at
e atf ull,ornearf ul
lcapaci t
y and ef fi
ciency.
Suggestopediaalsoseemst oputgrammari ni t
spr operplace.

B.Appl
iedLi
ngui
sti
csResear
ch

Wet urnnow toat

temptstoempir
ical
l
yt estt
eachi
ngmethodsvi
a"appl
i
edli
ngui
sti
cs
research"
.Idefinedappl
i
edlingui
sti
csr
esearchinChapterIasr
esear
chaimedatsol
vi
nga
practi
calprobl
em, wi
thor

146
146
146
wit
houtr ef
erenceto an underl
yi
ng t
heory.A port
ion ofappli
ed li
ngui
sti
csr esear
ch has
consist
edofempi r
icalcompari
sonsoflanguageteachi
ngmet hods,andthepurposeoft hi
s
secti
onistorevi
ewt hemajorf
indi
ngsofthesest
udiesinordert
omaket hefol
lowingpoint
s:

1. Whenoldermet hodssuchasgr ammar-transl

ati
on,cognit
ive-
code,
andaudio-l
ingual arecompar edwitheachot her,
wesee
smalldi
fferences, ornodiff
erencesinter msofeffi
cacy.Cogniti
ve-
code,
insomest udies,showsav er
ysl i
ghtsuper i
ori
tyf
or
adultst
udentswhencompar edt oaudio-li
ngual
,andnodi f
ferences
areseenwhenadol escent
sar ecompar ed.

2. Dat aarenotaspl ent

ifulaswewoul dli
kethem t
obef ornewer
met hods,buttheresult
swedohav earequit
econsistentwithour
theoreti
calanaly
sisoft heprev
ioussecti
on.Newerappr oaches,
suchasTot alPhy si
calResponse,producesi
gnif
icant
lybet t
er
result
sthanolderappr oaches.

Wewil
lconcludethatweseel i
tt
ledif
fer
encebetweenoldermethodssincet heyallf
ail
manyoftherequirementsf oropti
mali nputand ov er
emphasizeconsciouslearning.The
newermet
hodsputt othemet hodcompar i
sontestsati
sfyt
her equi
rementsbetter,andare
al
soshowntooutperf
ormt heirri
val
s.

1.REVI
EW OFMETHODCOMPARI
SONSTUDI
ES

Beforepr oceedingdirect
lytothestudies,itshoul dbepoi ntedoutt hatclassroom
research,whileitoftenproducesval
uableandi nter
est i
ngdat a,doesnotpr oduce"defi
niti
ve"
data.Thisisbecauseoft hemany"confoundi
ngv ar
iables"thatpreventus,innear
lyallcases,
from concludingthatitmusthavebeenapar t
iculartreatmentormet hodthatwasresponsible
fortheresultsobtained.I
tmaybeuseful t
oli
stsomeoft hesepotentialconf
oundshere.

I
fst
udentsinapproachAdi dbet t
eri
nachievementt eststhanstudent
sinappr oachB,
assumingtheywer eequaltostartwit
h,itmaybet heteacherr at
herthantheactualmet hod
thatwasresponsibleforthediff
erence.Evenifthesamet eachertaughtbothclasses,the
teachermayhavepr ef
err
edoneappr oachtotheother
,ormayhav eevenli
kedthestudentsin
oneclassbett
er!ClassAmi ghthavebeentaughtearl
yint hemor ning,andcl
assBr ightaft
er
l
unch.Thus,st
udent si
nclassA

147
147
147
 Tabl 1.Amer
e5. icanst
udi
escompar
ingf
orei
gnl
anguaget
eachi
ngmet
hods

St
udy Met
hods TL n Test
s:Speaki
ngLCReadi
ngWr
it
eAt
ti
tudet
owar
dmet
hod

Scherer& GT, AL Ger

man 130, 150 Year 1: AL AL GT GT AL
Wertheimer Year 2: nd AL nd GT AL

Chast
ain& CC,AL Spani
sh 51,48 Year 1: AL nd CC CC
Woerdehof
f1 35,
31 Year 2: AL nd nd nd

Muel
l
er CC,
ALFr
ench not CC CC CC CC
gi
ven (
fewer
dr
opout
s)

AL=audi
o-l
ingual
GT=grammar -
transl
ati
on
CC=Cognit
ivecode
nd=nodif
ference

1 I
ncl
udesbot
hChast
ain&Woer
dehof
f(1968)andChast
ain(
1970)
.

148
148
148
mi ghthav ebeenmor ealert.ClassBmighthavebeenl ocatedneart heat
hleti
cfiel
d,causing
mor edi stracti
ons.Textsmayhav emadeadiff
erence.Theremi ghtal
sohavebeenasel ection
biasont hepar tofthestudents;somemayhav edeliberat
elyenrol
ledincl
assAbecauset hey
knew i twas" special
".Thet eachercancer
tai
nlyaddot herpotenti
alconfounds.(Formor e
technical discussi
on,seeCampbel l
andSt
anley,
1963. )

Manyoft hemet hodologicalproblemscanber educed.IfapproachAconsistentl

ydoes
betteri
nmanyst udi
es,involvi
ngdi f
ferentclassroomswi t
hmanyst udentsindif
ferentschool
s,
thatis,i
fAisrel
iablysuper i
orinav ari
etyofcondi ti
onsusingav ar
ietyofmeasures,theresul
ts
ofclassroom ex
per i
ment sareatl eastwor thconsider
ing,especi
all
yiftheyareconsist
entwith
conclusi
onsderivedfrom othersour ces,e.g.secondlanguageacquisit
iontheor
y.

(
a)Amer
icanst
udi
esofAL,
GT,
andCC

Tabl
e5. 1summar i
zessever
alcomparisonsofteachingmethodsallofwhichlast
ed
oneortwoy ear
s.Thesestudi
eswereallconcernedwithforei
gnlanguageteachi
nginthe
Unit
edStates,compari
ngtheaudio-
li
ngualsystem wi
theithergrammar-t
ransl
ati
on(GT)or
cogni
ti
ve-
code( CC)
.

Scher erandWer thei

mer(1964)f
oundsomedi f
fer
encesbet weenaudi o-li
ngual
( AL)and
grammar-translat
ion(GT)af t
eryearone,differencesthatappear edtor eflectthemodal it
y
eachmet hodemphasi zed.Thesediff
erenceswer eattenuatedaf t
erthesecondy ear,and
combinedscor esforsub-test
sshowednosi gnifi
cantdif
ferencesbetweent het womet hods.It
wasconcl udedt hatstudentstendtodowel linthosear easemphasi zedi nt het eachi
ng
method.

Chastai
n and Woer dehof
f( 1968)and Chast ai
n( 1970)f ound si
mi l
arresults af
ter
comparingaudi
o-l
ingualandcognit
ive-codet eaching,fi
ndingsomedi ff
erencesafteryearone,
dif
fer
encesthatcouldplausi
blybetracedt ot hoseskill
semphasi zedinthemet hodused,but
nodiff
erencesaf
teryeartwo.Chastain( 1970)al sonotedt hatmalestendedtodobet terwith
AL,whil
efemalesdidbetteri
nCCsect ions.Wer etur
nt otheseint
eresti
ngfindi
ngslater.

Muel
l
er(
1971)l
i
mit
edhi
sst
udyt
ooney
ear
,compar
ingALandCC

149
149
149
teachi
ng.Forthoseski
l
lstested,CCwassuperior
,whi
leALcl
assesscor
edatnat
ional(
MLA)
norms.Ther esul
tsofprevi
ousst udiesf
orceustoaskwhet
herthi
sadvant
agewouldhave
beenmaintai
nedinthesecondy ear.

Tabl
e5.2givesussomeideaastot hedegreeofsuper
ior
it
yshownbyonemet hodover
another.Whatisobv i
ousist
hatbot
hmet hodsresulti
nsomepr ogr
ess;st
udentsdobett
erat
theendoft hecour sethanatthebegi
nning.Whi l
edif
fer
encesareoccasional
l
ysigni
fi
cant
,
theyarecert
ainl
ynothuge.

Tabl
e5.2
Degr
eeofsuper
ior
it
yshowni
ncompar
ati
vemethodstudi
es(
Amer
icanser
ies)

Li
steni
ng
MLACooper
ati
veTest
s:Readi
ng1 Wr
it
ing1 comp. Speaki
ng2

AL 26 59 25 51
CC 30 64 26 49

1: Si
gni
fi
cantdi
ff
erencei
nfav
orofCC.

2: Si
gni
fi
cantdi
ff
erencei
nfav
orofAL.

From:ChastainandWoerdehof
f(1968)
Testsadminister
edaf
teroneyearofuni
ver
sit
ylev
elst
udyofSpani
sh.

(
b)TheGUMEpr
oject

Thefirstgr
oupofst udieslookedatlanguaget eachi
ngefficacyoveroneortwoyears,
usi
ngprofi
ciencytest
s.Anothergroupofstudiesfocussedratheronspecifi
cstr
uctur
esovera
shor
terti
mespan.Thesest udiesarether
esultoft heGUMEpr oject,whi
chdealtwit
hEngli
sh
asaforei
gnlanguageinSweden.TheGUMEpr ojectstudi
esaresummar i
zedinTabl
e5.3.

TheGUMEpr ojectaimedt ocompar eALt ypeteachi

ngwi th" cogniti
ve"met hods,the
l
atterbeingquit
esimilartothecogniti
ve-codesy stem.Iwi l
lnotpresentt heirr
esultsstudy-by-
study,butwi
llat
tempti nst
eadtosummar izet heoveral
lresul
ts;theinterestedreadercanr efer
toTable5.3fordetai
ls,ortothestudi
est hemsel ves(seeespeciall
yv onElekandOskar sson,
1975,foracompl et
er evi
ewoft headultstudies).

150
150
150
Statedv er
ysimply
,theGUMEpr ojectfoundnoov eral
ldif
ferencesbetweenwhatt
heytermed
"i
mpl i
cit"methods (si
mil
art o AL)and " ex pl
i
cit"met hods (simil
arto CC f
oradolescent
subjects.Foradultsubjects,explici
tmet hods wer ef ound to be somewhatbet
ter.The
dif
ferenceforadul
tswasst ati
sti
call
ysignif
icant,butnotverylarge.

Togoi ntosli
ghtlymor edet ai
lfortheadol escents,despit
et heov eral
lfindi
ngsofno
diff
erences,somesub- groupsdi dbett
erusingex pli
citmet hods:(
1)an" accelerated"classi
n
Lev i
n'
sst udy,(2)femal es,invonEl ekandOskar sson,af indi
ngsi mil
art othatofChast ai
n
(1970).Onecl assinv onEl ekandOskar sson'
sstudy ,describedasbeinglowert hant henorm
i
n" verbalint
elli
gence"(v onElekandOskar sson,1975,p.29)wasr eport
edt ohav ehadmor e
troublethanotherclasseswi ththeexpli
citmethod.

I
naddi ti
ont o si
mpl
ecompar i
sonsofex pl
icitand implici
tmethods,v onElekand
Oskarssonalsocompar edvar
iouscombinationsofthesemet hods.Theyf
oundt hatIMEX(see
Tabl
e5. 3)wassuper i
ortoI
M alone;i
not herwords,addingsomegr ammati
calex pl
anat
ionto
amet hodbasedt ot
all
yonpatterndr
il
lswashel pf
ul(seefootnotetwoonTable
5.
3).Howev er,EXIM wasnotsuperi
ortoEX:addi ngpatter
ndr il
l
st oacogni
tiv
eappr oachdid
nothelp.

Tabl
e5. 4isi
ncl
udedtogivethereaderanideaoft hedegr
eeofsuper
ior
itytheex
pl i
cit
methodsshowedwi thadultsubj
ectsintheGUMEst udi
es.Asisthecasewi thAmerican
st
udiesdescri
bedintheprevi
oussecti
on,thedif
ferencesarenotlar
ge.Cl
earl
y,bothgr
oups
makeprogress.

2.SOMEPRELI
MINARYCONCLUSI
ONSTOMETHODCOMPARI
SON
STUDI
ES

Takenasawhole,Ameri
canandSwedi shst udiesshowonlysmalldiff
erences,i
fatall
,
betweent hemethodstheyi
nvesti
gated.Student
smakeatl eastsomepr ogressnomat t
er
whatmet hodisused,
aresul
tthathadtwodi f
fer
entki ndsofr
eactionsi
nthef i
eldoflanguage
teachi
ng.Stevi
ck(1976)not
edt heimpli
citcontradicti
on,whichhestatedint heform ofa
ri
ddle:

"
Int
hefiel
doflanguaget
eachi
ng,Met
hodAi st
helogi
calcont
radi
cti
onofMethodB:if
t
heassumpti
onsfrom whi
chAclai
mst obeder
ivedar
ecorr
ect,t
henBcannotwork,
andvice-
v
ersa.y
etone

151
 Tabl 3.GUMEpr
e5. ojectr
esear
chcompar
ingt
eachi
ngmet
hods

St
udy Met
hod St
udent
s Mat
eri
als Resul
ts

Ol
sson,
1969 I
mpl
i
cit
1 Age14 Onestr
uctureNodi
ff
erences
EXSwedish (passiv
e)
EXEngl
ish

Lev
in,
1972 I
mpl
i
cit Ages14-15 Noover
alldi
ff
erence;
EXEngl
ish "
advanced"groupexcel
sin
EXSwedish EXSwedish

Lev
in,
1972 I
mpli
cit Age 13 Nooveral
ldi
ff
erence;
EXSwedish more"abl
e"st
udentsdowel
l
EXEngl
ish wit
hEXSwedi sh,
butl
essabledowor se

VonElek& IM2 Adul

ts Tenl
essons EXPLI
CITsi
gni
fi
cant
lybet
ter
Oskar
sson, EX n=125
1975

VonElek& IM Adul
ts Asabov
e EXPLI
CITsi
gni
fi
cant
lybet
ter
Oskar
sson, ex N= 91
1975

VonElek& I
M Age 12 As abov
e Nodif
fer
ence,
duetol
ow
Oskar
sson, EX per
for
manceofoneEXPLI
CIT
1975 cl
ass

152
152
152
Tabl
e5.
3.-
-Cont
inued

St
udy Met
hod St
udent
s Mat
eri
als Resul
ts

VonElek& EX,
IM,EXI
M, Adul
tsn 4lessonson EXsuper i
ortoIM;IMEX
Oskar
sson, I
MEX3 =277 2str
uctur
es bettert
hanIM,
1975 butnotsi
gnif
icant
;
EXsuperi
ortoEXIM
(notpr
edi
cted)

VonElek& EX,
IM, Age12 4l
essonson Girl
stendtoconfor
m
Oskar
sson, EXI
M,I
XEM n=335 2st
ruct
ures t
ot headultpat
ter
n
1975 (seeabove)butboys
donot

1: I M ="
impli
cit
"(patter
ndril
l
sonly).EX="expl
i
cit
"(pat
ter
ndr
il
lsi
ncombi
nat
ionwi
th
ex
planat
ion)
.EXSwedi sh=ex pl
anati
oninSwedish.
EXEngl i
sh=ex pl
anati
oninEngli
sh.

2: IM ="structuredandgr adedpat terndril

l
s, perf
ormedont hebasi sofsituati
onalpi
ctures
projectedonascr eeni nfrontoft hecl ass..
.noex pli
citex planati
ons,compar i
sonswi ththe
sourcel anguage,ort r
anslati
on exercises"( vonEl ekand Oskar sson,1975,p.16) .EX =
"studentswer egi venex pl
i
citinformationaboutt hesy ntacti
cchar acteristi
csoft hestr
uctures
beingpr acti
ced. .
.compar isonswer emadewi tht hecor respondi ngst ruct
uresi nSwedi sh..
.
grammarwast aughtdeduct ivel
y..
.explanationsanddi rectionswer egi v
enbef oremainpr actice
witht hestructureunderst udy.
..exerciseswer emost lyoft hef i
l
l-i
nt ypeort ransl
ati
on...no
patterndri
ll
swer eper f
ormed"( v
onEl ekandOskar sson, 1975, p.16-17).

3: I
MEX=i denti
caltoIM wi
tht
headdi
ti
onofex
planat
ion.EXI
M =i
dent
icalt
oEXwi
th
addi
ti
onofor
alpat
terndri
ll
s.

153
153
153
 TABLE5.4
Thedegreeofsuperi
ori
tyshowni
ncompar
ati
ve
methodstudi
es(GUMEproject
)

Gr
oup Test Pr
e-t
est SD Post
-t
est SD Pr
ogr
ess

I
M A 26.
94 8.
61 33.
00 9.
31 6.
06
B 23.
75 7.
64 27.
53 7.
79 3.
78
C 5.
86 3.
40 9.
40 4.
16 3.
54
EX A 25.
71 6.
61 36.
59 9.
17 10.
88
B 21.
82 5.
19 29.
18 8.
14 7.
36
C 5.
65 3.
28 11.
84 4.
39 6.
19

Descr
iptionoft ests:
A:60items: Studentslistent
oor alconversati
on.Theychooseoneoft hree
alternati
vest ofi
llinmissi
ngpar toflastexchange( nomor ethantwowords)
.
B:50items: describedasan" ordinar
ymul ti
plechoicetest"
(p.66)
C:20items: writt
enpr oducti
ontest;students"fi
l
linthe
crucialelementi nincompleteEnglishsentences.Themeani ngofeach
sent encewascl arifi
edeit
herbyt hecompl eteSwedi shequiv
alent
,orbya
cur eword"( p.66).

Fr
om:
VonEl
ekandOskar
sson(
1972)
.

col
l
eagueisget
ti
ngexcell
entr
esul
tswi
thAandanot
heri
sget
ti
ngex
cel
l
entr
esul
tswi
thB.
Howisthi
spossi
ble?
"(p.104)
.

Toapplythi
sriddletotheresult
sofcompar ati
vemet hodresearch,wecanaskhowt he
cogni
ti
ve appr
oach,whi ch assumest hatlearni
ng becomesacqui sit
ion,can gi
veresul
ts
comparabl
et o audi
o-li
ngualteaching,which is based on the hypothesi
sthatwe learn
l
anguagebyconditi
oningandhabit-st
rengt
hening.

Beforet urningtoapossi bleanswer ,itshouldbenot edt hatt herewasadi ff

erent
reacti
on:notev eryonesawt hecont r
adict
ionStevi
cksaw.Manymet hodologistsandt eachers
simplyassumedt hatthesoluti
onwassi mplytobeecl ecti
c, t
ochoosepar tsofeachsy stem in
thebeliefthattheanswermustbesomewher einthemi ddle.Asfairmi ndedast hissounds, i
t
oftenresulted,inmyopi ni
on,inteacherschoosingthewor stfr
om each,t hepar tsleastli
kely
to encouragel anguageacqui sit
ion:patter
n dril
lfrom t heaudi o-l
ingualsy stem,and r ule
explanati
onf rom thecogni
ti
v eandgr ammar-tr
anslat
ionappr oach!

Inmyi nterpret
at i
on,ther esult
sofmet hodcompar isonstudiesusingaudi o-
li
ngual,
grammar -
transl
ati
on,andcogni ti
vecodear equit
econsistentwit
ht hetheoret
icalanal
ysisof
thesemet hodspr esentedint hepr evi
oussecti
on:accordi
ngt othi
sanal ysi
s,noneoft hese
met hodsdoesapar ti
cularl
yeffecti
v ej
obinencouragi
ngsubconsciousl anguageacquisi
tion,
alt
hougheachwi l
lprovideatleastsome, andthecogni
ti
ve

154
154
154
methodswi l
lallowsomewhatmor elearning.Thispredi
ctstheclosesimi
lar
lyinef
fecti
veness,
andt hesli
ghtsuper iori
tycognit
ivesy stemsshow f orol
dersubj ect
sandt hemor e"ver
bal"
adol
escents.Italsopr edi
ctsthatothermet hodsshoulddomuch,muchbet t
er.Unfort
unatel
y,
wedonothav edet ail
edmet hodcompar i
sondat aonal lthenewermet hods,butsomei s
avai
labl
e,andt heresultsarequit
econsi stentwiththi
spredicti
on.

3.MORERECENTMETHODCOMPARI
SONSTUDI
ES

Wet urnnow t ost udiesthatinvol

vethenewermet hods,met hodswi thbet t
erreport
cards,accordingt osecondl anguageacquisi
t i
ont heory,thangr ammar -
transl
ation,audio-
l
ingual,orcognit
ive-
code.Wedonothav edetai
ledr epor
tsonev erymet hodcompar edt oevery
other,andsomeoft henewappr oacheshav enev erbeent est
ed.Thest udiesthathav ebeen
performed,howev er
,indicatethatthosemethodst hatprovidemor eoftheinputnecessar yfor
acquisit
ion,
andt hat"putgrammari ni
tspl
ace",aresuperiortoolderapproaches.

(
a)TheTPRser
ies

Asherhasdoneat hor
oughjobinputti
nghi
smet hodtot heempi
ricaltest
.Hehas
compared TPR t
o othermet
hodsusing f
orei
gn l
anguage cl
assesand second l
anguage
cl
asses,usi
ngchi
ldr
enandusi
ngadult
s.Her
eisabri
efsurvey
.

TheTPRser ieswithadul t
sbeginswi t
hAsher( 1972),whichcompar edst udent sina
TPRGer mancour sewi t
hcont rolsina" standard"coll
egecour se.Asherrepor tedt hataf t
er
only32hour sofTPRinstr
uction, TPRstudent soutperf
ormedcont rol
s,whohadhad150hour s
ofclassti
me, i
natestofli
steningcompr ehension,andequaledcont r
olsi
ntest sofr eadingand
wri
ting.Asher '
sstudentsprogr essednear lyfivetimesfaster!Thisisincont rastt othev ery
smal ldi
fferencesseeninol dercompar ativemet hodex per
imentscompar ingaudi o-lingual
,
cogniti
ve-code,andgrammar -translat
ion.

Asher
,Kusudoanddel aTor
re(1974)compar edTPRst udentsst
udyingSpanishatthe
fi
rsty earuniv
ersi
tylev
elwith AL contr
ols.After45 hoursofTPR i nstruct
ion,st
udents
outperfor
medcont r
olswho hadhad150hour sinli
steni
ngcompr ehension,andequaled
controls'
per
formanceonareadi
ngtest(PimsleurSpani
shProfi
ciencyTest
).

155
155
155
Aft
er45addit
ionalhoursofi
nstr
ucti
on,
TPRst udent
sperf
ormedbeyondthe50thper
cent
il
e
onaSpanishprofi
ciencyt
estdesi
gnedforst
udentswit
h150hoursonall
skil
ls

Asher(1977a)compared30ESLst udentsusingTPRtocont r
olsusi
ngaudi
o-l
i
ngual
i
nst
ructi
on,andr eport
edthatTPRstudent
sout per
formedcontr
olswhohadhadt hesame
amountoftr
aini
ng( 120hour
s)butwhohadstart
edatahigherl
evelcl
ass.

TPRst udieshav eal sobeendoneusi ngchi l

drenassubj ect
s.Asher( 1977b)i sr eall
y
threeex peri
ment sinone,al li
nvolvi
ngSpanishasaf orei
gnlanguagei ngradesf i
vet hrough
nine.I nEx peri
mentI ,TPRcl assesingradesix,andacl assconsist
ingofsev ent handei ghth
grader swer ecompar edtogr adeninecontrol
s.Thecont r
olscoveredsimil
armat er i
al,
butt heir
i
nst ructionfocussedonr epet i
ti
on,andformalinst
ructi
oninreadingandwr it
ing, "emphasi zi
ng
Spani shgr ammar "
.Allgr
oupshadat otalof40hour sofclasstime.AllTPRcl asses( seven
differentclassesi nall
)exceededt hecontrol
sonat estofwrit
tenpr oduct
ion( subjectswer e
asked t o wr i
te a shortst orybased on a car t
oon,and wer e graded on the numberof
meani ngfulexpressi
onspr oduced).

I
nEx perimentI I
,nineel ementaryschoolTPRcl asses,fr
om gr adesfivethrougheight,
andanadul teducationTPRcl ass,werecompar edwi t
ht wocontrolclassesfrom gradesseven
throughnine.Thi stime,TPRcl asseshadonl y20hour sofi nstr
uctionwhi l
econtrolshad200
hour sofinstructi
onsi milartot hatdescr
ibedinEx peri
mentI .Thetestusedwast he" Spani
sh
PictureTestf orListeni
ng" ,whichaskedst udentstojudgewhet heragi vensentencewast r
ue
orf al
seinrelationtoapi cture.AllTPRclasses,wi t
htheex cepti
onofgr adefive,
out perf
ormed
cont r
olsafter100hour s,andt headultcl
ass,afteronly20hour s,outperfor
medcont rol
safter
200hour s.Simi l
arresultswer eobt ai
nedusingar eadi
ngt est.

InEx per
imentIII
,fi
ft
handsixt
hgradeTPRandcontr
olcl
asseswerecomparedontests
thatmeasuredfluency(e.
g."Wri
teasmanySpanishor
der
sorsentencesasyoucanrecal
l
...
")
.
Bothgroupshadequalcl asst
ime.TPRstudent
sdidsi
gni
fi
cant
lybett
erthancont
rol
sonbot h
fl
uencytasks.

TheTPRr esul
tsarecl
earandconsistent
,andthemagni
tudeofsuperi
ori
tyofTPRi s
quit
est r
iki
ng.Event heonesub-groupthatdidnottur
nouttobesuper i
or(gradefiv
ei n
Experi
mentIIabove)canbeexpl
ained:t
heTPRadv ant
agewasoutwei
ghedbythefactt
hatthe
contr
ols

156
156
156
wereolder
,and,
asmentionedinChapt
erI
I,
olderchi
l
drenar
eknownt
obef
ast
eracqui
rer
s
(
Krashen,LongandScar
cell
a,1979)
.

(
b)Ot
heri
nputmet
hodscompar
ed

Av ari
etyofst udieshavebeendoneex amini
ngt heeff
icacyofmethodst hat
,li
keTPR,
focusonpr ovi
dingcompr ehensibl
einputanddonotf orceearl
yproducti
on.Noneoft hese
met hodshasbeenanal yzedinthepr ev
ioussecti
on,sincetheyarenot"standard"orwi del
y
used,butt heystrengthenbot hthecasef orTPRandt hehypothesi
sthatmethodsal l
owinga
sil
entperiodwi l
ldobet tert
hanmet hodst hatdonot,evenwhen" speaki
ngskil
ls"aretested
dir
ectly.

Gary(
1975)examinedchildrenst udyi
ngSpanishasaf orei
gnlanguageoveraperiodof
fi
vemont hs.Herexperi
mentalgroupdi dnotspeakatal lf
orthef i
rst14weeksbut ,i
nstead,
hadt oproduce"acti
veresponses"t hatdemonst ratedcomprehension.Also,t
heywer enot
forcedtospeakformuchoft henex tsev enweeks.Theex peri
mentalgroupwasshownt obe
superiort
othecontrolgroupinl i
steningcompr ehensi
onandequali nspeaking,despi
tethe
factthatt
hecontr
olshadmor e"practice"inspeaki
ng.

Postovsky(1974)usedstudentsattheDef enceLanguageI nstitute,studyingRussianin

ani ntensi
ve12weekcour se,sixhoursperday ,inaf ai
rl
ystandar daudi o- l
ingualcourse.The
"experiment al
"groupdidnotspeakforthefir
stfourweeks, butwrot ethei rresponses.Thet wo
groupswer ecombi nedafterfourweeks.Atmi d-t
erms,theex periment algr oupex cell
ed in
reading,wr iti
ng,and speaki
ng tests(especial
lywi threspectto" cont rolofgr ammar "and
"readingaloud"),andaft
er12weeks, t
heywer esignifi
cant
lybett
eri nlisteningcompr ehension.

Swaf f
erandWoodr uff(1978)ex ami nedt heef f
ectsofaf ir
sty earcol legeGer man
courset aughtatt heUni versi
tyofTex as.Asi st hecasewi t
ht hestudi esjustci t
ed,thei
r
approachwasnotex actl
yanyoft hest andar donesdescr i
bedint hefi
rstpar toft hi
schapter,
butitfitstherequir
ementf orprovi
dingopt i
mali nputforacquisit
ionandforput t
inglearni
ngin
i
tspl acev erywell.Thef i
rstfourweeksoft hecour sewer eTPRbased,wi t
ht heemphasi s
swit
chi ngtoreading"forglobalmeani ng"( p.28).Studentswer enotrequiredt ospeakatal li
n
Germanf orthef ir
sttwoweeksoft hecl ass,and" t
hereaf
terstudentswer eencour agedto
speakonav oluntarybasi
s"(p.28).Also, "overtcorrecti
onsof

157
157
157
beginningst udents'productioner r
ors(was)keptatami ni
mum"( p.28).Low filt
erstrength
wasf ur t
herencour agedbyt heuseofr elaxat
ionex er
ci sesandyogabr eat
hing.Also,"except
forabr ief(f
ive-
mi nute)questionandanswerper i
odatt hecloseofeachhour ,Ger
manwast he
exclusivelanguageofi nstruction"(
p.28).Nodr il
lwasused, andtheonlygrammart aughtwas
thosef eatur
es" consideredessent i
alf
orlist
eningandr eadingcomprehension"(p.30).Swaffer
and Woodr uff'
smet hod thusappear st o supplycompr ehensi
bleinputin quant i
ty,using
techniquest hatencour ageal oweraff
ectivefi
lter
,anddoesnotencour agetheov er-
useoft he
Moni t
or .

TheSwaf f
erandWoodr uffprogr am wasev aluat
edinsev er
alway s,andallindi
cated
cl
ear l
ythatt hemet hodwasahugei mpr ovementov erotherappr oaches.First
,ascomparedt o
previ
ousy ears,mor est udent scont i
nuedont osecondsemest erGer man.Second,German
coursest aughtt henew wayr eceivedmuchbet terev al
uat i
onsf r
om t hestudents.Third,
studentscompl etingt hecour seper f ormedwel labov ethenat i
onalnor msont heModer n
Language Associ at i
on r eading and l isteni
ng tests( 70th and 69t h percenti
l
es),and last,
studentsel f
-reportoft heirownabi l
iti
eswas,i nmyopi ni
on,amaz i
ng:78% oft hestudents
fi
nishingt hef i
rsty ear" expressedconf i
dencethatt heycoul dr eadGer manandgr aspmai n
i
deasatl eastmostoft het i
me"( p.32) ;48%sai dt heycouldunder standspokenGer manat
l
eastmostoft het ime.Ido notknow ofcont roldat af ort hislastquestion,butfrom
experience,theser esponsesar equiteunusual .
2

(
c)Suggest
opedi
aresear
ch

Whi l
et here have been reports ofst udent sl earni
ng 1000 wor ds perday using
Suggestopedia,inthenor malonemont hintensiv ecour se,studentscov erabout2000wor ds.
Lozanovi squot edassay ingthat"aft
ercompl eti
ngt hecour se,thest udentscanex pr
ess
themselvesf r
eelywithi
nt heframeworkoft heirlex i
calcapacity,andcanr eadnewspapersand
books."(I
ntervi
ewpubl i
shedi nPravda,r
epr i
ntedi nOst randerandSchr oeder,1976,p.
74)Thesear eex cel
lentresult
s,butarenotsuper human- -
themont h'
scour se,asnotedear
li
er,
i
squiteintensive,meetingfourhoursperday ,sixday saweek, f
orat otalofnearly
100hour s.Intermsofcl asshoursalone,thisisequi valenttomor et hanoney earofst
udyat
theuniversi
tylevel
.Ifstudentscanindeed

158
158
158
"getby"i
nconver
sati
oni nthetar
getlanguageandr eadmanythingsinitaswel
l
,Lozanov'
s
approachmaybejustaboutassuccessfulasother"
inputmet
hods",suchast
hemethodused
bySwaf f
erandWoodruff
,whoreportsi
mi l
arresul
ts.

BushmanandMadsen( 1976)putSuggest opedi

at ot heex peri
ment altestinasmal l
scale st udy done atBr igham Young Uni v ersity.(Lozanov has car ri
ed outex t
ensive
exper i
ment ati
onwhi chr epor t
edlydemonst rat
et hesuper i
ori
tyofSuggest opediaov erAL-type
met hods.Det ail
soft hesest udiesarenotav ai
lablet ome.Forav erycr i
ti
calrevi
ew, seeScov el
,
1979. )Sixdiffer
entclassesatBYU, teachingFi nnishasaf or
eignlanguage, withanav erageof
sevenst udentsi neachcl asswer eused.Twocont rolclasseswer et aughtusi ngt he"full
"
Suggest opediat reat
mentandt wowi thamodi fiedt r
eatment .Themodi fi
edSuggest opedia
classesf ol
l
owedal laspect sofSuggest opediabutl ackedmusi c,theeasychai r
s,andt he
"l
ivi
ngr oom env ironment".Theywer eheldi nsteadi nordinaryclassrooms.Eachcl assr eceived
10hour sofi nstruct
ionandcov eredsimi l
arlingui st
icmat eri
al.Tocont rolforteacheref fect
,
twoi nstructor
st aughtallthreet r
eatment s.

Suggest opedi
a student si n bothf ul
land modi fied classes clearl
y outperformed
contr
olsinav ocabularytestandwer e"vastl
ysuperior"i
nat estof"communi cat
ion".(Inthi
s
test
,student
swer eratedont heirsuccessinconv eyi
ngamessaget oanat i
vespeaker .
)There
werenosi gnif
icantdif
ferencesbet weenSuggest opediaclassesandcont r
olsonagr ammar
testorinapr onunciati
ont est;thi
sr esul
tsuppor t
st hehy pothesi
st hatSuggestopediawas
superi
ortothecont rolgroup,sincecont r
olclasseshadf armor eworkonpr onunciati
onand
grammari ntheform ofpatterndril
lsandrepetit
ionexerci
ses.

Bushman and Madsen al so pr obed st udent'

s personalr eacti
ons t ot he di ff
erent
tr
eatment s,and reported no dif
ferences bet ween gr oups:t here was no di fference in
measur edaffectbetweenSuggestopedi candcont r
olgroups.Thi sconflicts,tosomeex tent
,
withreportsfrom t
heCanadi anPubli
cSer viceCommi ssion,i
nt hei
rreportofaf ullone-mont h
Frenchcour se.Theyreportedchangedat ti
tudest owardl anguagel ear
ni ng( "
learning"used
hereint hegeneralsense),andeven" ar ealandt otalchangei ntheper sonhi msel f"(p.33).
JustasLozanovmai nt
ainshappensinSuggest opedia,Canadianresearcher sreport

159
159
159
"thestudentdi
scoverednewcapabili
tiesi
nhimsel
f,becameawareofwhathewasabl etodo,
reali
zedtheextentofhiscreati
vi
tyandhispotent
ial
;he'f
oundhimsel
f',whichgavehim more
self-
confi
denceandsel f
-assur
ance"(p.33)
.Inourter
ms,theybecameawar eofther
eali
tyof
theirownsecondl anguageacquisi
ti
oncapacit
yandt hefactt
hatitr
emai nsverypowerf
ulin
theadult.

C.Al
ter
nat
ivest
oMet
hods

Thepr ev i
oussect i
onat tempt edt oshow sev er
althings.Fi r
st,thatwecananal yzeall
commonl y used appr oaches t o classr oom second language t eaching interms oft he
requi
rement sforopt i
mali nputpr esent edinChapterII
Iandt hecr i
teriaforteachi
ngconscious
grammarr ules,aspr esented in Chapt erIV.Second,i twasseen t hatcertai
n methods
sati
sfi
edt heser equi
rement sandcr i
teri
abet t
erthanot hers.Third,i twascl ai
medt hatt he
avail
ableappl ied li
nguisti
csr esear ch revealsthatthose met hodst hatareshown t o be
superi
orinmet hodcompar i
sonr esear chcomecl osertosatisf
yingt hecr it
eri
athatder
ivefrom
secondlanguageacqui sit
iontheor y.

Whatwecanconcl udefr
om thi
ssurveyandreviewisthatther
eisnoonewayt oteach,
noonemet hodthatisclearl
ythebest
.Somemet hodsareclear
lymor eeff
ect
ivet
hanothers,
however,andtheclaim madeher eisthatthesameunder l
yi
ngpr i
ncipl
eswil
lholdforany
successf
ulsecondl
anguaget eachi
ngprogr
am, t
hepr i
nci
plesoutl
inedinChapt
ersI
IIandIV.

Thepurposeoft hissect i
oni stoex pl
oreway s ofhel pingpeopl
eacqui r
esecond
l
anguagest hatgobey ondcl assr
oom met hods.Inthesect
ionst hatf
oll
ow,Iwil
lrevi
ewwhatI
considertheessentialf
unct ionofthecl assroom tobe,andsomel imit
ati
onsinherentinall
cl
assroom met hods.Foll
owi ngthis,
Iwilldiscusssomepossi bil
iti
esi
nlanguageteachingthat
areconsistentwithmyv iewsont herol eoft heclassr
oom,andwhi ch,atthesamet i
me,
bypassorav oi
dsomepr oblemst hatari
sewi thclassr
oom met hods.

1.FUNCTI
ONOFTHECLASSROOM

Quit
esimpl
y,t
herol
eofthesecondorfor
eignl
anguagecl
assr
oom i
stobr
ingast
udent
t
oapointwher
ehecanbegi
ntousetheoutsi
de

160
160
160
worldforfurthersecondl anguageacquisit
ion.Asex pressedi
nChapt erIIt
hismeanswehav e
to provide student
s with enough compr ehensi
blei nputto bri
ng theirsecond language
compet ence tot he pointwher et heycan begi nt o under
stand language hear
d" on the
outsi
de",read,andpar ti
cipateinconv er
sations.Si
ncet heywil
lbel essthanfull
ycompet ent
,
wealsoneedt oprovi
det hem withtoolsforencouragingandregulat
inginput.

Inotherwords,al
lsecondlanguageclassesar etr
ansit
ional
,andnosecondl anguage
cl
asscanbeex pectedtodot heentir
ejob.Asdi scussedi nChapt erI
II
,secondlanguage
cl
assesarebestthoughtofasplacestogaincompr ehensibl
einputinearl
ystages,whenthe
acqui
rerdoesnotyethavet
hecompet encet
ounder standtheinputprovi
dedontheoutside.

2.THESECONDLANGUAGECLASSROOM ANDREQUI
REMENT#2

As we hav e seen,manyoft he newermet hods make v ali

antat t
empt st o meet
requir
ement#2ofChapt erI I
I,topr ovi
deinputthati
sgenuinelyinterestingandr elev
ant.The
NaturalApproachattempt stodot hisbyfocussi
ngonper sonaltopics,Communi tyLanguage
Learningbyhav i
ngthest udent sgeneratethei
rowni nput
.Thel i
teraturecontainsmanyot her
usefulandi nt
erest
ingsuggest i
onsast ohow t osolvetheproblem ofwhatt ot al
kabout:
Stevi
ck (1980)hasanex cellentdiscussion ofthe possibi
li
tyofusi ngpoet r
yf orsecond
l
anguagest udents,
andWi nn-
Bel lOlsen(1977)hasnumerousothersuggest i
ons.

Therear etwof undament alproblemswi t

hanyat t
emptt omeetr equi rement#2i nthe
secondlanguagecl assr oom, howev er.Thef irsti
sthatwhati sofi nterestt osomepeopl emay
notbeofi nteresttoother s.St evi
cknot est hisinrel
ationt ohi spoet r
yex per i
ment:onet eacher
whot ri
edpoet rynotedt hatf orherst udents, "
poetryj
usti sn'ttheirthi
ng; theypr ef
erpoli
tics"(p.
225).Stevickal sonotest hatsomest udentsmayobj ecttohumani sti
cappr oaches,suchas
Communi tyLanguageLear ning,thatpr omot epersonalgr owt hal ongwi thsecondl anguage
acquisi
ti
on: "Some( st
udent s)..
.will
eager lyaccepta' humani sti
c'languagecour seasanar ena,
orasamedi um,i nwhi cht of indnewadv ent uresindiscov eringthemsel v esandot herpeopl e,
andinwhi cht heycangoont obecomemor ethantheyhadbeenbef ore.Ot hersof

161
161
161
them,however,maydecidethatt hel
anguageclassisnotaplacewher
et heychooset o
confr
onttheissuesofali
enati
on,orofpersonalval
ues.
..t
heymayjustwanttobet aught
well
..
..Wemustrespectt
hisdeci
sion"(
p.293)
.

Whatisperhapsamor efundament alpr oblem isthatthesecondl anguageclassroom is

seen,bymanyst udentsandt eacher s,asanar tif
ici
allingui
sticenvir
onmentr egar dl
essof
att
empt stopromot e"real"communi cat i
on.Thef actthatitisaclassroom,andt hef actthat
theclassiscall
ed" Spanish"or"Engli
sh" ,ofitselfmaysubv ertanyefforttomeetrequi r
ement
two,andmaypr eventstudentsf r
om f ull
yfocussi ngont hemeani ngofwhati ssai d.Inother
words,t hefi
l
termayal waysbe" up",tosomeex t
ent,andmanyst udentswi l
lnev ergetso
i
nterestedinwhatisbeingsai dthattheyf orgetitisi nasecondl anguage.

Thereareot herli
mi tati
onsoft heclassr
oom t hatarenotr elat
edtoi t
sf ailuretof ully
sati
sfyrequir
ementt wo.Asment ionedinChapt erI
II,t
hereisreallynowayt hecl assroom can
provi
det hev ar
ietyofsecondl anguageusenecessar yf orrealcompet encei nasecond
l
anguage, nomat terhowv ar i
edt hepresentati
on,nomat terhowmanydi ff
erentsi tuati
onsar e
usedinr ol
eplayingactivi
ties.Ther eisalsonowayt heclassroom canprovidet hequant i
tyof
i
nputr equir
edf ort rul
yadv ancedcompet encei nasecondl anguage.Thesear enotr eal
probl
ems, whenweconsi derwhatt heclassroom isfor.I
fthest udentcanmaket het ransit
ion
tother ealworl
d,ifthest udentcanbegi nt ouset heoutsideforcompr ehensiblei nput,both
quanti
tyandv ar
ietywi l
lbepr ovided.

3.THEALTERNATI
VES

Wewi llex amineanumberofpossi bl

eal t
ernati
vesandsuppl ement si nthesect i
ons
thatfoll
ow, and, aswedi dwit
hlanguageteachingmet hodsear l
i
er,anal
y zethem accordingto
thepr edicti
onsmadebysecond l anguageacqui si
ti
ont heoryand sur veywhatempi r
ical
evidencet her
ei st hatconfi
rmsthattheseapproachesar eeffecti
ve.Wewi llfi
rstl
ookatsome
veryobv ioussuppl ementar
yactivi
ti
estot hesecondl anguagecl assr
oom,conv ersati
on( t
he
realthing,withnat i
vespeakersoft hetar
getl anguage)andpl easurereading,thenmov eto
somei deasandpr ogramsthathavemetwi threalsuccessinsomecont extsandt hatcouldbe
extendedt oot hercont ex
ts.

162
162
162
(
a)Conv
ersat
ion

Beforemaki ngthe predict

ion t
hat"conversati
on wit
h native speaker
s"isgood f or
secondlanguageacqui siti
on,Ineedtodefi
neconv ersati
oninali
mi tedway:conver
sati
onher e
refer
sonlyt ointeracti
onwi thanat i
vespeakerwhoi smot i
vat
edt otr
yt ohelpthesecond
l
anguageacqui rerunderstand,andwhoi sgenuinelyint
erest
edint heacquir
erasaper son.In
otherwords,conv ersat
ionwi t
h" f
orei
gnert
alk",i
fi ti
snecessary,andwi t
har ealorpotenti
al
fri
end,businessassociate,et
c.

Ther ewil
lbenoat t
empttodemonst rat
ethatthissortofconv ersati
onhasanyef f
ecton
consciousl ear
ning.Itonl ycouldiftheconver
sati
onalpar t
nerwereal anguageteacherand/ or
theacqui rerwer eanex tr
emelygi ft
edinducti
ve learner.Itdoesappeart obet hecase,
howev er,thatconv ersationdefinedinthiswayhast hemax imum i mpactonsubconsci ous
acquisit
ion.Thef ollowinganal ysi
sdemonst r
ateswhatt omostl aypeoplei scompl etely
obvious,thatconv ersationwithsomeonewhoi sint
erestedinint
eract ingwit
hyou, andwhoi s
try
ingtohel pyouunder standwhathei ssayi
ng,i
sgoodf orsecondl anguageacqui sit
ion.

Requi
rement
sforopt
imali
nput

()Compr
i ehensi
ble.Ifmeaningissuccessf
ull
ynegoti
ated,
ift
heconver
sationalpart
ner
i
sabletoadjusthisspeech,useex tr
a-l
i
nguist
ici
nformat
ionandcontex
t,andift heacquir
er
hasenoughli
nguist
iccompet enceandcanr egul
atethequali
tyofi
nput
,theinputprovidein
conv
ersati
onwil
lbecompr ehensibl
e.

()I
i
i nt
erest
ing/relev
ant
.Conversat
ionhasthebestchanceofmeeti
ngthisrequi
rement
ofal
lthemethodswehav econsi
dered.I
thasthebestchanceofachi
evi
ngtotalf
ocusont he
messageandofbr ingingtheacqui
rertothepointoffor
getti
ngthatt
heinputisinanother
l
anguage.

(
ii
i)Notgr
ammat
ical
l
ysequenced.Thi
srequi
rementi
scl
ear
lysat
isf
ied.

(i
v)Quant
it
y.Conver
sati
oncert
ainl
yhasthepotenti
alf
orsat
isf
yingt
hisr
equi
rement
,
dependi
ngontheper
sonalci
rcumst
ancesoft
heacqui
rer.

163
163
163
 (v)Fi
lt
erst r
ength.I
nf reeconv ersati
onwi thasy mpathet i
cnati
v espeaker,fi
l
terstrength
shouldbel ow.Ingener al
,thereislittl
eornoer rorcorrecti
onf orfor
m andmostpeopl edonot
demandper fectaccuracyorcompl et
eut t
erances,asdol anguaget eachers.Thet opicof
conversati
oni sofcour seunpr edictablebutisgener all
yoff argreaterinter
estthanany t
hing
thatgoesoni naclassroom, andt hisalsowi l
lcontri
butet oal owerlevelofanxietyandal ow
fi
lter
.Apossi bletension-
raiserinf reeconv er
sationist hechanceoft heacquirer"getti
ngi n
overhishead" ,andnotunder standingwhati ssaidt ohim.Ifhei sprepared,ifhehast oolsfor
conversati
onal managementandi swi l
l
ingtouset hem, thi
sislessofapr obl
em.

(
vi)Tool
sforconv
ersat
ionalmanagement
.Conver
sati
onwillgi
vet
heacquir
erachance
topracti
cethetool
shehaslearnedandgivehi
m per
hapsthebestoppor
tuni
tyt
oacqui
renew
ones.

(
b)Pl
easur
ereadi
ng

Aswas t hecase wi th conv ersation,cov ered int hepr evioussect ion,Iwi lldef i
ne
"reading"inaspeci alway .Idonotmeani ntensiver eading,anal ysisofwr i
ttenpr ose,reading
andt henanswer ingquest ionsofcont ent,orr eadingaspr epar ationf ordiscussionorwr i
ting
assignment s.Thesor tofr eadingt obeanal yzedher eisex t
ensiv e,andconcer nssubj ect
mat terthatthestudentwoul dr eadinhi sfir
stlanguagef orpl easur e.Itiscompl etelyv ol
untary.
Indoi ngpleasurereading,reader shavet heopt i
onofski ppingwhol esect ionst heyf i
ndeither
toodi ff
icul
torlessi nt
eresting( e.g.detail
eddescr i
ptionsi nfiction).Theyev enhav et heoption
ofput ti
ngt hebookorst orydownandsel ect i
nganot herafterr eadingaf ewpages.Theycan
skipwor dstheydonotunder stand,iftheyt hinkt heyar ef ollowingt hemai npoi nt,andt hey
hav etheopt i
on,ofcourse, oflooki ngupev erywor d,ifthatist heirsty l
e.

I
notherwor ds,weareconsi deri
ngpurepleasurereading.Whatisreaddependsont he
studentandwhati sav ai
l
ablet ohim.Forsomepeopl e,itmaybemy st
erynovels,f
orothers,
sciencef i
cti
on,andf orother
s,comi cbooks.Theonlyr equir
ementi st
hatthest or
yormai n
i
deabecompr ehensibl
e andt hatthetopi
cbe somet hing thestudenti
sgenuinelyi
nter
ested
i
n, thathewouldr eadinhisfi
rstlanguage.

164
164
164
 Iwil
lnottrytoshowthatpleasurereadi
nghasanyeffectonl
earni
ng.Theanal
ysis,as
donewithconversati
onint
hepr evioussecti
on,wi
llberest
rict
edtotheeffect
sofpleasure
readi
ngonacquisit
ion.

Requi
rement
sforopt
imali
nput

() Compr
i ehensible. We hav e defined pleasure readi
ng as r eadi
ng t hat is
comprehensi
ble,sother eisnopr oblem here.Iwouldl i
ketonot e,howev er
,thatpleasure
readi
ngismadecompr ehensiblebyt hereader'
sownsel ect
ionofpassagesandt ext
s,andby
therej
ecti
onofr eadi
ngmat eri
althati stoodiffi
cul
t.Thesuccessofpl easur
er eadi
ngt hus
dependsont hereader'
swi ll
ingnesst ofindmat eri
alathislevelandr ejectmateri
althatis
beyondhim.

()I
i
i nteresti
ng/r
elevant
.Readi
ng as defi
ned her
eis by def
ini
ti
on i
nter
esti
ng and
r
elev
ant,si
ncet hestudenthastheopt
ion(whichmustbeexerci
sed)ofonl
yreadingthi
ngs
t
hatareofpersonali
nterest
.

(i
i
i)Notgr
ammat icall
ysequenced.Thi
srequi
rementi
smet
,unl
esst
hest
udenti
nsi
sts
onr
eadi
ngspeci
all
y-
preparedpedagogical
mater
ial
s.

(i
v)Quant
it
y.Readingcert
ainl
yhast hepot
entialforsati
sfy
ingthi
sr equi
rement
.The
onl
ypr
oblemsar
epract
ical:
theavail
abi
l
ityofmat
eri
als,thei
rcost
,andthestudent
s'ti
me.

(
v)Fil
terlev
el.Ifthestudentisabletofindmater
ial
st hatarecomprehensibl
eandthat
ar
ei nt
eresti
ng,t hi
sr equi
rement is easily met
.Ther ei s no fr
ustr
ation caused by
i
ncomprehensi
ble messages,no ear l
ydemands f oroutput ,no demands forpr emat
ure
gr
ammat i
calaccuracy.Thepleasurereadershoul
dbecompl etelyof
fthedefensi
ve.

(
vi)Toolsforconver
sati
onalmanagement.Pl
easur
er eadi
ng mi
ghtev en made a
cont
ri
buti
ontowar
dmeetingthi
srequi
rement
,ift
het
ext
sreadincl
udesomedi
alogue.

165
165
165
 Att hispointIwoul dl i
ket oi ncl udeaper sonalobser v ati
onaboutpl easur er eading.I
hav ebeenat t
empt i
ng,ov erthel astf ew y ear s,t oi mpr ov emyFr ench,largel yv i
apl easure
readi ng,anat temptt hathasbeensuccessf ul.Most l
yt hroughi nput ,Ihav ei ncreasedmy
compet encefr
om " advancedbegi nner "t o"highi ntermedi ate".Idef inet heinter mediat elev eli
n
thef ollowingway :Requi r
ingonl ysome" downshi ft
ing"ont hepar tofanat ivespeakert obe
ablet oconv erseeasi l
y,andbei ngabl et or eadmostt extswi thoutadi ctionary,wi t
hout
necessar il
yknowi ngev erywor d.Icannow r eadagr eatdealofFr enchwi t
houtadi ctionary,
andev ender i
ver ealpleasur ef rom i t.Bei ngaMoni toruserandsomeonewi thani ntrinsi
c
i
nter esti nthestructureofl anguage, Ioccasi onal lylookatgr ammarbooks( theonest hatgav e
mesomucht roubleinhi ghschool )
.Ihav enot iced, tomysur pri
se, thatther eadi ngpassages
att heendoft heel ement arygr ammarbookst i
llgivemet r
oubl e!If i
ndt hem mor edi ffi
cult
than" raw",uneditedFr ench,Fr enchwr i
ttenf ornat i
v espeaker s.Ther eason" pedagogi cal"
passagesar emor edi ff
icultfort hei nt ermedi atei st hatt heyar epackedf ul
lofsubj unctives,
condi ti
onnelpassé,f uturant erieur ,and al lmannerofi nfrequentv ocabul ary!inr eading
throught hesepassages,If oundt hem di ffi
cul tt ounder stand,andex t
remel yf rustrati
ng:t he
topicswer enotev enofmi ldinter est,andIf eltmyaf fectivef i
lt
ergoi ngup,asIencount ered
wor daf terwor dIdi dnotknow.Myf rustrationwasf ur t
heraggr avatedbyt hef actt hatI
realizedt hatIwashav i
ngt roublewi that extdesi gnedf orsecondy earst udent s!

Whatthisexperi
encesuggestsisthatourint
ermediat
estudent
smayf indr ealt
exts,
readforint
erestandpleasur
e,easi
erthanourpedagogicalmater
ial
s.Mor
eover,iftheabove
analy
sisiscorrect
,itmaybet hatf
reepleasur
ereadingwillr
esul
tinmoreacquisit
ionofthe
l
anguage.

Fort hosewhoobj ectont hegr oundst hatreadinginl anguagecour sesshoul dbe

restri
ctedt otheclassics,t
oser i
ousl it
erature,Icanonl ysaythattheabi l
it
ytoread" l
iter
ature"
willbef aci
lit
atedbyt hedev el
opmentofahi ghlev elofcompet enceinthesecondl anguage.I
personal yagr
l eewiththosewhof eelthatamaj orgoalofl anguageinstructi
onintheuni versi
ty
i
st hest udyofl it
erat
urewr i
ttenint hesecondl anguage.Idonott hink,
howev er,t
hatweneed
tost ar
toutwi t
hseriouslit
eraturei mmedi atel
y.Dev oti
ngsev eralmont hstofreer eadingof
easiermat erialmightbethef astestwayt obr i
ngst udentstothepoint

166
166
166
wher
etheycanr
eadgr
eatl
i
ter
atur
einasecondl
anguagewi
thoutaser
iousl
anguagebar
ri
er.

Summar
y

Bothconversat
ionandpleasurereadi
nghav et hepotent i
alofmeet i
ngtherequir
ements
foropt
imali
nputforacquisi
ti
onv erywell
.Wehav er eachedt heconclusionthataninter
esti
ng
conver
sat
ion in a second l
anguage,and r eadi
ng somet hing forpleasure,are excel
l
ent
l
anguagelessons.Thiscomesasnosur pr
iset othemi ll
ionsofpeopl ewhohav eacquired
l
anguageusingonlythese"methods",
andhav eacquiredthem v erywell
.

(
c)Usi
ngsubj
ectmat
terf
orl
anguaget
eachi
ng

Anot herclassofal t
ernat
ivestoclassroom teachingi
nvolvestheuseofsubj ectmatt
er
i
nt hesecondl anguagecl assr
oom,usingt hesecondl anguageasav ehicl
e,asalanguageof
presentationandex planati
on.Idonotmeanbysubj ectmatterteachingwhati sknownas
submer si
on,mi xi
ngsecondl anguagest udentsinwi t
hnat i
vespeaker s.Idomeanspeci al
classesf orsecondl anguagestudents,classesinwhi chnonat i
vespeaker spart
ici
pateas
students,inwhi chteacher smakesomel i
nguist
icandcul t
uraladjustmentsinordertohelp
theirst
udent sunderst and.

Inthissect i
on,wewi llfi
rstputsubj ectmatterteachingthr ought hefami l
iaranal
ysis
accordingt ot hepredi
ct i
onsmadebysecondl anguageacqui sit
iont heory.Aswast hecase
withconv ersationandpl easurer eading,therewil
lbenoat t
emptt ocl aimt hatsubjectmatter
teachi
ng hel ps conscious l earning in any way .We wi l
lt hen t ur
nt o sev eralconcrete
mani f
estati
onsofsubj ectmat tert eaching:thesuccessfulimmer sionpr ogramsunder wayi n
Canadaandt heUnitedSt at
es,andsomeunt ri
edpossibil
it
iesfort heuseofsubj ectmatterin
secondl anguagesi t
uations.

Requi
rement
sofopt
imali
nput

(
i)Comprehensi
ble.Subjectmat
tert
eachingwil
lbeofusef oracqui
sit
iononl
ytothe
ext
enti
tiscomprehensi
ble.Whatthi
smeansisthatdi
ff
erentsubj
ectsmaybeofmor euseto
st
udent
satdif
ferent

167
167
167
l
ev el
s.Caz den( 1979)poi ntsoutt hatonecoul dmakeacaset hatmat hematicsisi dealfor
teachi
ng in an i mperfectlyacqui r
ed L2.Ther eisal imi t
ed vocabul ary
,lessi nteracti
onal
demand t han i n some ot hersubj ects,and consi derable extr
a-li
nguisti
c suppor tt o aid
compr ehension.( She poi nts out
,howev er
,thatcompl ex "storypr oblems"mi ghtbe an
obstacleforbeginninglev elstudents.)Studentswithmor esecondl anguagepr ofi
ciencycoul d
handlesubj ectmat terthati smor edi spl
acedi nti
meandspace,t hatsuppl yf ewerconcr ete
refer
ents,such ashi storyand l it
er atur
e.Thepoi ntisnotsi mpl yt hatsecond l anguage
studentscansur viveinsubj ectmat terclasses,butalsot hattheywi l
lreceivecompr ehensible
i
nputt hatwillhel
pt hem impr ovemor einthesecondl anguage.

Thecompr ehensibil
i
tyrequi
rementarguesagai
nstsubmer si
on,agai
nstmixingsecond
l
anguageacquir
ersi nwithnati
vespeaker
sbef oret
hesecondlanguageacquir
ersreachhi
gher
l
evel
sofprofici
ency :t
hepr esenceofnativ
espeakersinsur
est hatagoodpr oport
ionofthe
l
anguageheardbyt heinter
mediateacqui
rerwill
notbecomprehensibl
e.

(
i)I
i nt
eresti
ng/relev
ant
.Subjectmat termaynotal waysbeinterest
ing,buti
tisrel
evant
.
Whenst udentsarefocussedont hesubj ectmat ter
,thereisaverygoodchancet heywi l
lbe
focussedofft
hef orm ofthel
anguagei ti
spr esentedin.Subj
ectmatteraf
fordsagoodchance
ofmeet i
ngthe"forget
tingpri
nci
ple",ofthest udentbeingsofocussedonwhati ssaidthathe
i
snotawar eofhowi tissaid.

(i
i)Notgr
i ammat ical
l
ysequenced.Thi
srequi
rementi
sal
soclearl
ymet.Infact,itis
har
dtoimagesubjectmattert
eachi
ngnotmeet i
ngit.Thi
swoul
dr equi
recont
extual
izing
bey
ondourwi
l
destdreams.

(i
v)Quant
it
y.Cl
ear
ly,t
her
eisthepotentialofsuppl
yinggreatquant
it
iesofi
nputthi
sway.
Subj
ectmattert
eachi
nginthesecondlanguageaut omatical
l
yr eachest
hepedagogicali
deal
offi
l
li
ngtheenti
recl
asshourwi
thcompr ehensibl
einput.

(
v)Fi
l
terst
rengt
h.Subj
ectmat
tert
eachi
ngmayi
nvol
ve,
andi
nfact

168
168
168
requi
re,somemi ni
mum amountofanx i
ety.Thisanx i
ety,however
,isnotdir
ectedatthe
l
anguagei ti
spr esent
edin,i
fthemessageiscompr ehensibl
e.Subj
ectmat
terteacher
scan
keepthelanguageporti
onrel
ati
vel
yanx
iet
y-f
reeandthefil
terdownby:

(
1) i
nsur
ingcompr
ehensi
bil
i
tyoft
hemessage;

(
2) notdemandi
ngpr
emat
urepr
oduct
ion;

(
3) notdemandi
ngf
ull
grammat
ical
accur
acyf
rom st
udent
s.

Subjectmattersecondl anguageteachersmi ghtconsi

dertesti
ngpr oceduresthatrequi
reless
l
inguist
icproducti
on( shortanswersinst
eadofl ongessays),andclassdiscussionprocedures
thattakestudents'li
nguisti
ccapaciti
esi nt
oconsi der
ati
on( notcorrecti
nger r
orsonf orm or
evenal l
owinguseoft heL1wher epr acti
cal,asi ntheNat ur
alAppr oach)
.Thepoi ntto
rememberi sthatf ur
therlanguageacquisit
ioncomeswi t
hmor ecompr ehensibl
einput,fr
om
teachertal
kandr eading,andnotfrom demandsf orproduct
ion.

(v)Tool
i sforconversati
onalmanagement .Subjectmat t
erteachingmaynotpr ovidethe
toolsnecessarytomai ntainconversationsont heout side,butitcanleadt othelearni
ngand
acquisit
ionofacademi ccommuni cativ
ecompet enceinanot hercultur
e.Inacl asscomposed
entir
elyofimmi grant
sandf or
eignst udents,t
eacher scanbeawar eofcul tur
aldiff
erencesin
academi cbehaviorandt eachclassroom behav ior,ei
therv i
alearni
ng,f orobvi
ousaspect sof
classroom behavior(
standingornotst andi
ngwhent het eacherenterstheroom;whatsor tof
papert ohandinhomewor kon,etc.)oracquisi
ti
on, formor esubtleaspects.

Summar
y

Subjectmat tert eaching has,t hus,t he fullpot

enti
alf orencouragi
ng l anguage
acquisi
tion.Thismaybeagoodpl acetopoi ntoutthatbysubjectmattert
eaching,Idonot
mean" Engl ishforSpeci alPur poses"orf or"AcademicPurposes".ESPandEAPar e,tomy
understanding,standardl anguaget eachingcl
asseswhosesy l
labiarebasedonananal y
sisof
thetaskst udentswillf
aceandt helanguagetheywillneed(
see,forexample,
Robinson,
1980).Subj ectmatterteachingappear s

169
169
169
tomet obef undament all
ydif f
erent ,
althoughitmaymeetmanyoft hegoal sESPi sdesi gned
for.Whi l
eESP r equiresadet ail
edanal ysisofthesynt ax,vocabular
y,anddi scourseofa
subfi
eld,tobedev elopedi ntoasy l
labusandpr esent
edbi tbybi t
,subj ectmat terteaching
focussesonl yont het opi
c,t hei nformationorskil
ltobel earned,t
heassumpt ionbei ngthat
muchoft hesy nt ax,vocabular y,anddi scoursesty
lewillbeacqui r
edal ongwi tht hesubject
mat t
er.(Thisideai snotent irelyf orei
gnt oESP;sever
alESPcour sesemphasi ze" authenti
c
acti
vit
ies".See,forex ampl e,Robi nson,p.39;Widdowson,citedinRobinson, p.23.)

(d) Evi
dence f
or subj
ect mat
ter t
eachi
ng: t
he i
mmer
sion
progr
ams

I
mmer si
onbili
ngualpr ogramshav edemonst rat
edwhati spossibl
einsecondl anguage
acquisi
ti
on using subjectmat t
er.Ini mmersion programs,i ni
ti
all
y monol
ingualmaj ori
ty
chi
ldrenareschooledi nami nori
tylanguage(FrenchinAngl ophoneCanada;Spanishi nthe
Unit
edSt at
es).Theyar et aughtt hei
racademicsubj ect
st ot
allyinthesecondl anguage.In
whatisknownas" totalearlyi
mmer si
on",i
nputinthesecondl anguagebegi
nsinkindergarten.
Lateimmer si
onpr ogramsmaybegi nl at
er,af
tert hechil
drenhav ehadatl eastay earof
i
nstruct
ioni
nt hesecondl anguage.

Theimmersi
onpr ogramsappeart obesuccessf uli nmanyway s.Themanyr epor t
sthat
havebeenpubl i
shedconf i
rm ov erandov erthatimmer sionst
udent sacqui
rehighl evelsof
compet encyi
nt he second language ( whi
let heymaynotr each nati
ve-
li
ke levels,they
outperfor
m peers who hav e had standardforeign language classes)
,theymake nor mal
progressinschool,doingaswel linsubject-
mat terasmonol inguals,andtheydonotf all
behindpeersi
nfir
stlanguagedev elopment(forrevi
ews, seeLamber tandTucker,
1972; Swain,
1974).

CohenandSwai n( 1976)di
scusst hesesuccessesi nl i
ghtofthel ackofsuccessof
manyot hertypesofbili
ngualprograms.Amongt hedif
ferencesbetweenimmer si
onandot her
progr
ams,t hesechar acteri
sti
csofimmer si
onmayhel pt oexplai
nitssuccess.Cohenand
Swain pointoutt hati n earl
yimmer si
on "allkinder
garten pupil
sare unili
ngualin L1.I n
essence,thesuccessf ulprogram starsoutasasegr
t egatedonel i
nguisti
call
y"(p.47).As
mentionedabov e,
thisraisesthest
udent s'
chancesof

170
170
170
gett
ingcomprehensi
blei
nput
,sinceteacher
scannotgauget
hei
rspeechonl
ytonat
ive
speaker
s,l
eav
ingsecondl
anguageacqui
rersbehi
nd.

CohenandSwai npoi ntoutseveralot herfactorst hat,inourterms,leadtoal ower

aff
ecti
v efil
terinimmer sionpr ograms.Thel i
nguisticsegr egati
on"eli
mi nat
esthekindof
ri
dicul
et hatstudent
sex ertonl essprofi
cientper former s"(p.47),teachershavepositi
ve
expectat
ions,andtheprogram isv ol
unt
ary.Also,"i
nki ndergarten,t
hechil
drenarepermi
ttedto
speakintheL1unt i
ltheyarer eadytospeakint heL2"( p.48).Thus,aSil
entPeriodi
sall
owed.

Thei mmer sionex peri

ence, itneedst obeemphasi zed,doesnotbr ingt hesestudent sto
nati
vespeak erlevels,andimmer sionst udents' secondlanguagecompet encemayhav egaps,
especi
al l
ywheni tcomest ointeracti
onabi li
ti
esi ncasualconv ersation.(SeeConner s,Menar d
andSingh, 1978, whor eportproblemsi mmer sionstudentshav eint hisar ea; ontheot herhand,
seeBruck,Lamber tandTucker ,1974,f orar epor tonwhati mmer sionchi ldrencandoi nthis
area.
)Iti sthoughtt hatthesegapsex i
stonl ybecauset hesecondl anguagei nputdoesnot
i
ncludei nputfrom peer s.Immer sionchildrenheart hesecondl anguageonl yf r
om thet eacher
andonl yinclass.Consi deri
ngt hi
sl imit
ation,theirachi
evement sar er emar kable.

Theimmer sionprogramsshow uswhati spossibl

eli
nguist
ical
lyfrom subj
ectmatt
er
teachi
ng,whensoci alandpsychol
ogicalproblemsar eel
imi
natedorr educed.Theyprov
ide
strongempiri
calev i
dencethatsubjectmatt
ert eachi
ngcannotonlyteachsubjectmatt
erbut
thelanguagei
tist aughtinaswell
,aslongast heinputi
smadecompr ehensibl
e.

(
e)Ot
herpossi
bil
i
tiesi
nsubj
ectmat
tert
eachi
ng

Thereisnor easonthatsubjectmat terteachingcannotbeex t

endedtoot hersecond
l
anguageacqui sit
iondomai ns,and ut i
l
ized to atl eastsupplementt hesecond language
cl
assroom andpr ovidesomehel pinthediffi
culttr
ansi t
ionfrom l
anguagecl
asst orealworld.
Onesuchdomai nist heuni
ver
sity.Iwil
ldiscussher ethesituati
onintheAmericanuniversi
ty,
butthepri
ncipl
escanbegener al
izedtoanyhi ghereducat i
onsituat
ioninwhi
chlargenumber s
ofsecondlanguagespeakersar eenrol
led.

171
171
171
 Practi
cal
lyev erylar
geAmer icanuni ver
sit
yhasanESLpr ogr am.Theyr angeinqual it
y,
ofcour se,from ex cell
enttosub-standar d,butregar
dlessofqual i
t y
,iti
smyf eeli
ngt hatforei
gn
student sregardt hem asanobst acl e.ESLi s,moreover,perceivedasi rrel
evantatj ustthose
l
ev elsthatbotht heor yandappliedr esear chconcludeiti
si r
relevant:atthe"intermediate"level
.
Manyf oreignstudent snol ongerf eelt heyneedESLwhent heyar eablet osur vi
veinr egular
classes,y etwell-meani ngadmi nistrator sfeelthatforthef oreignst udents'ownpr otecti
on,
theirlevelofEnglishcompet enceshoul dbehi gher.

Appli
edr esear
chconfir
mst hatint
ermediat
eESLi snotproducti
v e.Thestudiesof
Upshur(1968)andMason( 1971),r
evi
ewedi nChapterI
I,whichshowedt hatextr
aESLdoes
nothel
pwhenst udentsar
eenroll
edinr
egularcl
asses,i
ncludedonl
ystudentsatthi
slevel
.

Forsome" goodlanguagelearners"(acquir
ers),theanswert ot hi
spr obl
em mayv ery
wellbet heeli
minationofthe"ESL"requirementorpl aci
ngt helevelorrequir
edpr ofi
ciencyin
Engli
shl ower.Forot her
s,however,t
hiswoul dnotbet hebestsol ut
ion.Thef eeli
ngsofESL
administrat
orsthatsomest udentsneed "mor e"isqui ter ealand just
ifi
ed.Allt oo oft
en,
studentsareableonl ytosurviv
ei nclasseswher ethel anguagedemandi sv er
ylow,and/ or
theyendupr elyi
ngheav i
lyonnativ
elanguagehel p,i
nt heform oftext
sorclassmat es.

Subjectmattert
eachingmaybepar toftheanswert othis" t
ransit
ion"pr oblem.WhatI
proposei sthattheuniversit
yconsidercl assesfori
nternati
onalst udentsinsubj ectmat t
er,
classesinwhi chinter
nationalstudentsar einfact"segregated",tobeof feredi nal lar
eas
forei
gnst udentsarelikel
yt oenr ol
l,andt obemadeav ai
lableonav oluntarybasi s.Such
courses woul d gi
vef ullacademi c creditand coverr egularsubj ectmat ter.The mai n
diff
erenceswoul dbethef actthatthest udentsmaybeunf ami l
i
arwi thAmer icanacademi c
practi
ces.3

Theabsenceofnat i
vespeakersint heclasswoul dhel ptoi nsur
et hatthei nputis
comprehensi
bleforthesamer easoni mmer si
onpr ovidesmor ecompr ehensibl
ei nputthan
submersi
on.Boththelevelofcomplexi
tyoft heclassroom presentat
ionandt heamountand
complexi
tyofoutsidereadi
ngwouldber egulat
edt ot heli
nguisti
clevelofthecl ass.Other
modif
icati
onsthatwoul
dhelpcompr ehensi
bili
tyarealsopossible:

172
172
172
Wewouldexpectl
owerdemandsonst
udentoutput
,incl
udi
ngatol
eranceforer
rors(manyof
whichwil
lbeel
imi
natedbymorecomprehensi
blei
nputovert
ime)andt est
srequir
ingshor
t
answer
sinli
euofl
ongessay
s.

Int
er nat
ionalcl assescan notonl ytakei nto accountand hel p eli
minatelinguist
ic
defi
ciencies,theycanal sohelpf i
l
lsev eralothergapsi ni nt ernat
ionalstudents'knowledge.
Theycanpr ovideananx i
ety-
fr
ee,oratl eastanx i
ety
-l
ow, init
ialex posuretotheAmer i
canst y
le
ofeducat ion.St udentswi l
lbeabl et oacquiret hesubt let
iesofAmer i
cancl assroom st y
le
behavior,andl earnmanyoft heobv i
ousdi ff
erencest hatex istbetweenwhati sacceptable
behaviorinacl assroom intheircountryandwhati sex pect edi ntheAmer icanuniversit
y.In
otherwor ds,internati
onalclassescangi vest udentssomeoft het ool
sf orcommuni cati
ve
compet encei nthecont extoftheclassroom.45

(
i)TheNewEngl andproblem.I nternat
ionalstudent scanal sof i
llf
oreignstudentsinon
cult
urali nformati
on t hati s presupposed i n courses f ornat i
ve speaker s and Amer i
can
student s.Thomas Jabl onskioft he Hi st
oryDepar t
mentatUSC has been t eaching an
Amer icanhi st
orycourseex cl
usivelyforinter
nat i
onalstudent sforthel astthreeyears,andhe
haspoi ntedoutt o met hatmanyi nternati
onalst udent slacki nformationt hatAmer i
can
professorst akeforgrant ed.Acl earexampl eishi sfi
ndingt hatmanyofhi sstudentsdi dnot
haveacl earideaofwher eNew Englandwas,apoi ntofi nf
ormat i
ont hatwasessent i
altoa
parti
cularpr esentati
on.I nfor
mat iongapssuchast hisonear enotobv i
ous,andpr obably
abound.Theyhav eabet terchanceofbei ngfil
ledininternati
onalcl asses,wherest udentsare
encour agedt oaskquest ions,andwher einstructorspresupposel ess.

()Ther
i
i ol
eofESLi nsubjectmat t
erteachi
ng.Theestabl
i
shmentofi
nter
national
cl
assesdoesnotsignaltheendofESL,al thoughitmayresul
tinsomemodif
icat
ion,and
hopef
ull
yimpr
ovement,ofourESLoffer
ings.

Fir
st,whil
ewecan i magi
nepushing subj
ectmatt
ercl
asses"down"t othel owest
l
inguisti
cpr of
ici
encylev
elpossi
ble,
wemayal wayshav
eaneedf
orthesecondlanguageclass
atthebegi nni
nglevel.I
tisanempiri
calquest
ionj
usthowmuchcompetenceandi nst
ruct
ion
(i
.e.comprehensible

173
173
173
i
nput
)isnecessar
ybef
orest
udent
scanbeginspecialsubjectmat
tercl
asses,
butt
her
ewi
l
l,i
n
mostcases,
beaneedforagener
alcl
assatthebeginning.6

Second,asdi scussedinChapt
erIVther
earemanyaspect soflanguaget hatare
consci
ouslylear
nabl
e,bot hi
n"gr
ammar "(
mostl
ymorphol
ogyforthemajori
tyofstudents)
and di
scourse(consciousrul
esforthefi
nepoint
sofwr i
ti
ng,incl
udi
ng punctuat
ionand
organi
zat
ion).

Al
so,alar
gepercentageoff orei
gnstudentsmaydesi remor eEngl
ishthantheycanget
i
nt he classr
oom situati
on in or dertof acil
it
ate part
ici
pati
on in Amer i
can socialli
fe.
I
ntermediat
elevelcl
assest hatfocusonpr ovidi
ngthet oolsforcommuni cat
ivecompetence
andconv er
sati
onalmanagementwoul dbev eryhelpfulforstudentswi t
hmor eintegr
ati
ve
ori
entati
onand/orwhoplant oremainintheUni t
edStatesforextendedperi
odsoftime.

Inaddit
ion,ESLteacher
smightserv
et heusefulfuncti
onofassi
sti
ngandconsul
ti
ng
wi
tht
hesubjectmattert
eacher
swhoteachint
ernat
ionalsecti
ons.

Fi
gure5.
2present
saschemaoft
hepossi
blei
nter
act
ionbet
weenanESLcomponent
andanacademi
ccomponent
.

Fi
g.5.
2.ESLandacademiccomponentsofi
nter
nat
ional
student
s'pr
ogr
am att
he
uni
ver
sit
ylev
el

Lev
el ESLcomponent Academi
ccomponent

Begi
nni
ng Classroom l anguaget eaching, Nonef ocussingon
topicsofgener al
i
nt erest;intr
oduct i
ontoUni ver
sity
l
ife
I
ntermedi
ateOpt
ional cour sewor kon I
nternati
onalsectionsof
1.Engl ishgr ammar( Moni t
or) subjectmat tercour
ses
2.St yl
istics(lear
nable) (
opt i
onal
)
3.Conv ersation(seetext)
Advanced None Regularsecti
onsofsubj ect
mattercour
ses

SeeNot
e6f
orsuggest
ionsf
orat
ransi
ti
onbet
weent
hebegi
nni
ngandi
nter
medi
atel
evel
s.

(
ii
i)Theneedf orappl
iedlinguist
icsr
esear
ch.I
fIwer
esi
mpl
ytoasser
tthatt
he
I
nter
nat
ionalSt
udent
sprogram asoutl
inedabovewas

174
174
174
"
theanswer"t
oourf or
eignst
udentpr
oblemsintheAmer i
canuniv
ersi
ty,Iwouldberepeat
ing
t
hesinsofthepast,
clai
mingthatweneedonlyconsulttheoryi
nordert
ocomet othecorr
ect
f
orm ofpract
ice.Toretur
ntot hemessageofChapt erI,thi
sisnotsuf f
ici
ent
.Atleastthe
f
oll
owingquesti
onsneedtobeansweredwit
hempiricaldata:

1. Dost
udent
sini
nter
nat
ional
classesacqui
remor
eEngl
i
sh?

2. Dotheyl
earnasmuchsubjectmat
terasthosewhoelectt
otake
st
andar
dcourses?
3. Dotheyhav
emor esuccessi
nthei
rstudi
esovert
helongrun?

4. Dot
heyf
eel
mor
ecomf
ort
abl
eint
heacademi
cenv
ironment
?

Cl
ear
ly,
theanswer
stot
hesequest
ionswi
l
lbeofbot
htheor
eti
cal
andpr
act
ical
int
erest
.

(
iv)Adul
tESLandsubj ectmat ter.TheESLprofessioni nt
heUni t
edStateshasalready
beenex peri
mentingwithaf orm ofsubj ectmat t
erteachingatthe" adul
teducati
on"lev
el,in
coursesdesignedf oradultimmi grantst ot heUni
tedSt ates.(Thi
si sanotherexampleof
teachersandadmi ni
str
atorsnotwai ti
ngf ortheor
yandr esearch,
butdiscover
ing"whatworks"
ont hei
rown;seeChapt erIfordiscussion.)S.Brown(1979)descr i
besoneex per
imentofthis
sortinLosAngeles.

Whil
epar tofESLi nstr
uctioni nBr own' sschooli s" t
hemor et r
adit
ionalgrammar -
oriented"sty
leclass,studentsalsopar ti
cipateinunitscov eri
ng"li
fesit
uati
ons"topi
csthatlast
fr
om t wot ofourweeks.Ex amplesi ncl
udet heuseofcommuni t
yservi
ces(post-of
fi
ce,li
brary,
etc.),consumereducat ion,empl oy ment( cov eri
ng classif
ied ads,employmentagenci es,
unions, et
c.)
,familyli
fe(e.g.weddi nginvi
tations,bir
thdaypar ti
es,et
c.)
,ci
ti
zenship(e.
g.traffi
c
andpar ki
ngt i
ckets,voti
ng,t axes,etc.)
,andot her"li
fesi t
uati
ons".Teacherscanuseguest
speaker s,fi
l
ms,f iel
dt ri
ps,andcommer cialmat eri
alsi nhelpingstudentsunderstandt he
"mechani csofli
fe"inanewcount r
y .

Agai
n,asisusuallyt
hecase, noevi
denceisyetavai
labl
econf i
rmingtheuti
li
tyofsucha
program.Twooft hethreesourcesofinspir
ati
onforprogr
amspr esentedinChapterI,second
l
anguage acquisit
ion theory and teacherinsi
ght/i
ntui
ti
on,pr edi
ct,howev er,thatsuch
programswillbeofgr eatuseforlanguageacquisi
ti
on,inaddit
iont othei
robv i
ouspr act
ical
value,asl
ongastheinputiscompr ehensi
bl 7
e.

175
D.Comment
sonAchi
evementTest
ing

Inthi
ssect i
on,wewi l
lconsiderthei mplicati
onsofsecondlanguageacqui siti
ont heory
ont esting.Iwi l
lbeginwithav eryshortr evi
ewofwhatwenor mall
yconsi dertober elev
anti n
selecti
ngt estsf orsecondlanguageachi evement ,andthekindsoftestoptionswechosef rom.
Aswast hecasei ndescri
binglanguaget eachingsy st
emsear l
ier
,thisisnotdonei nanef f
ort
tosuppl ynew i nformati
on,butt oestabl i
shacommonsetofassumpt ions;Iwi llassume,
therefore,somef amili
ari
tywiththestandardlit
er at
ureinsecondlanguaget esti
ng( e.g.Harris,
1969; Valette,1977; Ol
ler
,1979).Iwillt
henf ocusononemaj orconsider
ation,whatOl l
er(1979)
termst he"instructi
onalvalue"ofat est,andsuggestt hatifwet aket hisproper tyoft ests
seriously,secondl anguageacqui si
ti
ont heorysev erel
yli
mitsouropt i
onsi nachiev ementt est
selection.

1.NORMALCONSI
DERATI
ONSI
NTESTEVALUATI
ONANDSELECTI
ON

Thest andar dl i
ter
at ureont estsandmeasur ement stell
sust hatagoodt estneedst o
meetcertai
nst andards.Itmustber eliable,thatis,itmustconsi stentlygiv
ethesamer esult
s
underdif
ferentcondi t
ions.I tmustal sobev al
id,thatis,itshouldr eall
ymeasur ewhati tis
supposedtomeasur e.Test i
ngex pertsal soadv iseust omakesur eat estispracti
cal,thatiti
s
economical
, easyt oscor e, andeasyt oi nterpret(Har r
is,1969,pp.21- 22).Har
risalsosuggest
thatweconsi dert hef acev al
idi
tyofat est," t
hewayt het estlooks--totheex aminees,test
admini
strat
or s,educat ors,andt helik e"(p.21) ,notingthati fatestdoesnotappeart obea
vali
dmeasur e,whet heritisornoti nr eality,
student sandt eacherswillnottakeitseriously
.

Teachersandadmi nist
ratorsi
nsecondl anguagepr ogramsnowhav eawi dev
ariet
yof
teststochoosef rom.Test sareusuall
ycl assif
iedaccordingt othemodalitytheyuse(r
eading,
writi
ng,speaking,li
steni
ng)andt hei
rpl aceamongt hedi scretepoi
nt/
integrati
veconti
nuum.
Discret
e-pointtestsareteststhatattempt" tof ocusattentionononepoi ntofgr ammarata
ti
me"( Oll
er,1979,p.37).Anex t
remedi scretepointtestrequiresaminimum ofknowl edgeof
contextoutsidet hesentencecont ai
ningt heitem tested.Her eisanex ampl eofadi scr
ete-
pointit
em:

176
176
176
 Mar
y i
nNewYor
ksi
nce1960.

a.isl
iv
ingb.has
l
ivedc.li
ves

I
nt egrati
vet
ests,ont heotherhand, makenoat t
empttof ocusononeaspectofl anguageata
ti
me;accor ding to Ol l
er(1979)," Whereas discr
eteitems at t
emptt ot estknowl edge of
l
anguageonebi tatat i
me,int
egr ati
vetestsat t
emptt oassessal earner'
scapaci t
yt ouse
manybi t
sallatthesamet i
me,andpossi blywhileexerci
singseveralpresumedcomponent s
ofagr ammaticalsystem,andper hapsmor ethanoneoft het r
adit
ional
ly-r
ecognizedskill
sor
aspect sofskill
s"(p.37) .Examplesoft estthatareusual l
yconsi der
edi nt
egrativ
ei ncl
ude
readingcompr ehension,cl
ozetests,dict
ation,
composi t
ions,andtestsoforalcommuni cat
ion.

2.I
NSTRUCTI
ONALVALUE

Iwouldli
ketofocusher eononlyoneaspectofoneki ndoftesti
ng,thei
nstructi
onal
valueofachi
evementtests,andmakeonlyonepoint.Test
shaveahugei mpactonclassroom
behavi
or,andneedt obeselectedtoencour
agestudentstoengageinacti
vi
ti
esthatwi l
lhel
p
them acqui
remor elanguage.Itmaybet hattheinstr
ucti
onalval
uecrit
eri
onispossi bl
yof
mor ei
mpor t
ancethanthecrit
erial
i
stedabove.

Statedsi mpl
y,thesortoftestsel ect
edhasahugei mpactont heclass.Ifstudent s
knowi nadv ancewhatsor toft
estwillbeusedt omeasuret hei
rachiev
ementinacour se, t
hey
will
,naturall
y,tendtostudyforthetest,andteacherswillf
eelpressur
etoteacht othet est.I
suggestwehar nesst hi
snaturaltendencyand sel ectteststhatwi l
lencouragest udent
preparati
ont hatini
tsel
fcausesmoresecondl anguageacquisi
ti
on.8

Jones( 1979)givesagoodex ampleoft her

esultsofharnessingt hi
stendency,whi
chhe
call
st he'backwash"ef fect.Inteachinganel ementaryGermancour seatt heuniver
sit
ylevel,
hedeci dedtogi veanor almi dter
m,ashor t(
fiveminute)conversat
iondoneonaone-on- one
basis.Jonesnot edont histestthatf ewofhi sst
udent swerepr of
icientintheareaofsoci al
communi cat
ion:" WhenIgr eetedthem, askedhowt heywere,orsaidgood- bye,themajori
tyof
them hadnor esponsebut

177
177
177
awkwar
dlaught
er,
event
hought
heyhadpr
act
icedt
hesepr
otocol
sint
hecl
assr
oom"(
p.56)
.

The ef fectoft hi
s mi dt
erm ex peri
ence on the cl
ass was st r
iki
ng:" The teachi
ng
assistant stoldmeshor t
lyafterthef i
rstoralt
estthatstudentswer ebeggingf ormor eoral
practicei nt hecl assr
oom.Thesi tuati
onwasmuchdi f
fer
entont hesecondt est .Theywere
waiting f orme.I twasobv i
oust hatt heyhad madeagr eateffortto develop speaki
ng
profi
ci encyi nav eryshorttime.Thet estnotonlygavemev i
talinf
ormationaboutt hei
rabil
ity
tospeakt hel anguage,butitalsoser vedasamot ivati
nginfluenceforthem tospendmor e
ti
medev elopingthisimportantskil
l"(pp.56-
57).

Whati fJones'oraltesthadf ai
l
edt heusualstandardsforr eli
abili
ty?Whatif,for
exampl e,
t herat
inghadbeenmadebysev er
al j
udgesandthei
rinter
rat
err el
iabil
i
tyhadnotmet
therequiredlevel?Thepowerfulbackwasheffect
,Iam suggesti
ng,may ,incertai
nsit
uat
ions,
moret hanmakeupf ort
hisproblem.

Thebasi cpr oblem Iam speaki ngofher eist hef actthatpract i

ceincer taintypesof
testsdoesnotnecessar il
yl eadt omor eacqui siti
onoft hesecondl anguage.Thi sf actor
eli
mi nat essomet estswithver yf
inet r
ackr ecordswhenj udgedont hebasisofr eliabil
i
tyand
vali
dity.Ther ei snoev i
dence,f orex ampl e,thatpr acti
cingcl ozet estsincl asshel pst he
studentacqui r
emor eofthel anguage,ori mprov esper f
ormanceoncl ozetests.Ther eisv ery
goodev idence,ont heotherhand, thatparti
cipatinginconv ersati
on,andr eadingforcont entor
pleasur e, do hel p t he student acqui re l anguage. Conv ersat
ional pr acti
ce pr ov i
des
compr ehensi bl
ei nputandhel psthest udentacqui rethet oolsneededf orconv ersati
onwi th
nativespeaker s,whi chint ur
nr esultsinmor ei nputandmor elanguageacqui siti
on.Readi ng
forcont enti
sal soanef fecti
v ewayofget t
ingi nputthatmeet sther equir
ement sf oropt i
mal
i
nputf oracqui siti
on,aswesawear li
erinthischapt er.

Achievementt ests,Iam suggesti

ng,shouldmeetthisrequir
ement:preparat
ionf ort
he
test,orstudyingfort het est
,shouldobviousl
yencouraget hestudenttodot hi
ngst hatwill
providemor ecompr ehensi bl
einputandthet oolstogai
nev enmor einputwhent heclassis
over.Thisdr asti
call
yr educesouropt i
ons,butal so,i
nar ealsense,simplif
iesthet askof
achievementt esti
ng.Letusf ir
stexaminewhatt heconsequencesofthisphi
losophymi ghtbe
i
nt heareaoff orei
gnl anguagetesti
ng.

178
178
178
 Achiev
ementt est
ingi
nforei
gnlanguageclassesat
temptstoassesswhetherast
udent
hasmett her equi
rementsofagi vencourse,andsomet i
meswhet herhehassat i
sfi
eda
l
anguagerequirementatani
nsti
tut
ion.Iwi
lldeal
witheachofthesesi
tuat
ionsi
nturn.

Fort hef orei

gnl anguagecl ass,Iseeonl yt woopt i
ons.Oneoft hem i sfairl
ytradit
ional:
reading compr ehension.I fst udent sknow i n adv ance thatt heywi llbe gi ven a reading
compr ehensiont est,at esti nwhi cht heyar easkedt oreadseveral shor tpassagesandanswer
gener alquest ionsaboutt hecont entofwhatt heyhav eread,theywi llbeencour agedt oread.
Theywi llbeencour agedt ost udyf ort het esti
nt hesimplestandmostobv i
ousway ,andwi ll
seekoutr eadingoppor tuniti
esi nt hesecondl anguage.Asl ongast heyknow t heywi llbe
present edwi thav ari
etyofpassages( ondi f
ferenttopics)andasl ongast hequest i
onsf ocus
ont he" gist"oft hepassageanddonotr el
yononespeci f
icwor dorst r ucture,itcertai
nlywill
bet hecaset hatgener alr eadi ngf orpl easureandi nt
erestwillpr eparet hem f orsuchat est.
Teacher swi l
lbeencour agedt opr ov idecompr ehensiblereadingmat er i
als,andst udentswi ll
beencour aged t o go out sidet heboundsoft hecl assroom i nsear chofsuppl ement ar
y
mat erials.Mosti mpor tant ,iftheyr ead, theywi l
l acquiremor eoft hetar getl anguage.

The reading comprehension t

estis especi al
l
yusef ul,si nce thereis generall
yno
probl
em inpurchasingorconstr
ucti
ngt est
st hatmeett hestati
sticalrequir
ement sment i
oned
above.Thestandardlit
erat
urehasmanysuggest i
onsonconst r
uct i
ngr eadi
ngt est
s( see,for
example,Harris,1969,chapter6),and reli
abili
tymeasur esand v ar
ioustypesofv ali
dity
measurescaneasi l
ybeobtained.Readi
ngt estscanbeconst ructedt hatarepracti
calandt hat
haveobviousfacevali
dit
y.

Asecondki ndoftestismor ecompl i

cat
ed,but,atthemoment ,Iseenoot herv ali
d
opt i
ons.Whati sneededisat estthatwil
lencour
agestudent
stoengagei nconv ersat
ions,that
requiresuseoft het ool
sofcommuni cati
vecompetence.Manyst andar
dor altestsfailt
odo
this.At estinwhicht hestudentanswersquesti
onsdoesnotr equir
einter
actionalabili
ty,nor
doesat estinwhi chastudentsimplytalksorevenasksquestions.Whatisneededi sat rue
testofconv ersat
ionalmanagement .

Iwil
latt
empttogiv
ear oughdescr
ipt
ionofwhatat
estofconv
ersat
ionalmanagement
woul
dl ookli
ke:I
deall
y,i
twoul dinv
olvebothtest
erandstudenti
naconv er
sat
ionabout
somet
hingreal
,aprobl
em that

179
179
179
hast obesol ved,at opicthatneedst obedi scussed.Second,t hestudentwoul dber atedon
hisabili
tyt omanaget heconv ersationandcommuni cate,notongr ammat icalaccuracy.If,f
or
exampl e,thest udenthadwor d-findi
ngdi f
ficult
iest hatresultedonl yinanembar rassedsilence
onhi spar t,hewoul dbegr adeddown.I ft hest udentwer eablet o" cover"thepr obl
em wi th
appropriatef i
ll
ers( j
ustamoment ..
.whatIwantt osayi s.
..howdoy ousay ..
.?)
,hewoul dnot
onlynotbepenal izedbutwoul dbegr adedupf orhav ingt heabi l
itytokeept heconversation
going and notl ose the fl
oor!St udents woul d al so be gi ven credi tforpol it
eness and
appropriateness,si nceami nimum amountoft hisknowl edgei sabsol utel
ynecessar yfor
successf ulconv ersat
ion.Mosti mpor t
ant ,they woul d be gi ven creditf orsuccessf ul
communi cation,f orsuccessfullycompl etingt hecommuni cativ
eex change.St udentswho
wer eablet ogett heex aminertohel pthem woul dal sobegr adedmor ehighly,theassumpt ion
beingt hatt hosewhocanel icitneededv ocabul aryandhel pt henat ivespeakergi vet hem
compr ehensi bl
ei nputwill
havemor esuccessi nsecondl anguageacqui si
ti
oni nthelongr un.9

Therearepredictableobj ectionsonecanmaket osuchat estingplan.Mostobv i

ously,it
can be ar gued thatsuch t ests,especi al
lyt he second one,wi l
ldo not hi
ng f ort he
developmentofgr ammat i
calaccur acy, andwillonl
yencour agesl oppyspeech, alaissez-
faire,
"anythi
nggoes"at t
it
udet owar dl anguage,andt heestablishmentofper manentbadhabi ts.
Secondl anguageacquisitiontheor y,howev er,makesqui t
edi f
ferentpredicti
ons:i
ftestsofthi s
sortencour agestudent st opar t
icipatei nconv er
sati
onanddev elopt heskil
lst omanage
conversati
ons,theywi l
lcont r
ibut eagr eatdealt othedevel opmentofgr ammat i
calaccuracy .
Indeed,theywi l
ldevel
op,per haps,mor egrammat i
calaccur acyint helongr unt
hananyot her
kindofmeasur e!Theywi llgivet hest udentthet ool
sheneedst oobt ai
ncompr ehensiblei
nput ,
andt hi
si nt ur
nwi l
lresultinsubsequentl anguageacqui siti
on,i mprovementaf t
ert heterm
ends.

Theconv er
sati
onalmanagementt estpromi sestobeverydif
fi
cultt
ogr adereli
abl
y,and
thankstothisunrel
iabi
li
ty,
itmayf ai
ltomeetaccept abl
estandardsofvali
dit
y.Itwil
lbehardto
trai
nr at
ers and hardt oi nventtopi
cs t o discuss.Neverthel
ess,ithas t he pr
omise of
stimul
atingstudentstodevelopconversati
onalskillst
hatwil
lenablethem t
o

180
180
180
usethel
anguagedespi
tethei
rlessthanper
fectpr
ofi
ciency
,thushel
pingt
oinsur
econt
inued
progr
essi
nsecondlanguageacquisi
ti
onaft
ertheter
m hasended.

Onecoul dalsoar guet hatatleastsomegr ammart esti

ngshouldbei ncluded.As
emphasi
zedinChapt erII
,wehav enotr
eject
edtheteachingofformalgrammar .I
thasi t
suse
asaMonitor,whenusingt heMoni t
ordoesnotint
erfer
ewi thcommuni cat
ion.I
tistheref
orea
por
ti
onofthei nstr
ucti
onalprogram.Shouldn'
twethereforetestgrammaraswel l
,intheform
oft
esti
ngourst udent
s'abili
ti
estoMoni t
orthei
rout
putundercondi ti
onsconducivetotheuse
oft
heMoni t
or?

Thear gumentappear edplausibletomeatonet ime.TracyTerrel

lpresentedmewi tha
counter-
argumentt ot estinggrammar ,andIthinkheisri
ght:ifweallow grammart esti
ng,it
willgrow andsoondomi natethetest i
ngprogram,andhencet hecur r
icul
um.Whi l
elimited
grammart estingisconsi stentwiththelimit
edroleofthegrammar,thereisar ealdangerthat
teachersandadmi nistratorswi l
lrev erttotheirol
dway sandgr aduall
yr et
urnt otesting
grammarex clusi
vel
y!

3.LANGUAGEREQUI
REMENTS

Manyuni versit
iesandsomehi ghschool sst illhavel anguager equirement s.Thisi s
usuallyexpressedast henecessi tyofst udy i
ngaf oreignlanguagef oragi venper i
odoft i
me,
twot ofoursemest ers.If,however ,thegoaloft heclassr oom ist obr i
ngst udent stot hepoint
wheret heycancont i
nuet oacqui rethel anguagebyusi ngt heout sidewor ld,orr esour
ces
outsidethecl assr oom, thissuggest sthatweshoul dconsi dert est
ingst udentst oseewhet her
theyhav ereachedt hi
sl evel
:cant heycont i
nuet oobt ai ncompr ehensibleinput ?Thet eststhat
probet hi
scoul dbet hei denti
calonespr oposedt obeofmax imum educat ionalv alueint he
precedingsect i
on:Readi ngcompr ehensionandconv ersati
onalmanagementar enotonl yt he
mostappr opr iatef orachi ev
ementt estsgi venatt heendoft hesemest er,butmayal sobet he
mostappr opriatel eavingex ams.Thet est saskonl yt hesequest ions:cant hestudentr ead
wellenoughi nt hesecondl anguagesot hathecanr eadtextswi thouthav ingt oconsul ta
dict
ionaryex cessi velyandwi t
houtunduepai n,i
.e.wit houtwhatNewmar kcalls"crytoanal
y ti
c
decoding".Isheabl etocommuni cat eeffect i
velywithanat i
vespeakerwhoi swi l
li
ngt ohelp?

181
181
181
 Ofcour se,Ihaveleftmanyser iousquest
ionsunsett
led,suchast herangeoft
opicsto
bereadanddi scussed,thepr oblem ofrequi
ri
ngequallevelsofpr ofi
ciencyi
ncognate(and
hencemor ecompr ehensible)languagesandmor eexoti
cl anguages,andhow thepassing
l
evelis deter
mi ned.Some appl i
ed resear
ch mayev entuall
yhel pt o sol
vethem;att he
moment,howev er,t
heformatseemscl ear.

4.UNI
VERSI
TYLEVELESL

Secondlanguageacqui sit
iontheory,aspresent edinthi
sv ol
ume, givesnomagi caland
obviousanswert othediff
icultquesti
onofESLt esti
ngatt heuniver
sityl
ev el
.Thegoalofsuch
testi
ngistodet erminewhet herstudentsknowenoughEngl i
sht ostudyinEnglish.Asiswel l
known,t he"backwash" ef fecthasbeenapr oblem i nthi
sar ea fory ears: manyf orei
gn
studentsstudyf ortheTOEFLex aminati
onexcl usi
vely,andarehel pedt odosobyspeci al
coursesdesignedt odojustthis(seediscussi
oni nWi ggon,1979).

Applyingthesamear gumentsherethatweusedear l
ier,emphasizi
ngt heinstr
uctional
val
ueoft ests,wecomet otheconclusi
ont hatsubjectmat tertesti
ngwoul dbeofbenef itat
thi
sl ev
elaswel l
.Thisismor eeasi
lysaidt handone;i twoul dbepr ohi
biti
velyexpensivet o
design standardized subjectmat t
ertestsi n al
ldi scipl
ines fori nt
ernati
onalst udents.
Int
ernati
onalcour ses,asoutli
nedintheprev i
oussect i
ons,maybeast epinthisdirecti
on,as
l
ongast heyusesubj ectmat tert
est
sasf i
nal s;astudents'r
eleasefrom theESLr equirement
couldbeatl eastparti
all
ydependentonhisabi li
tyt
opassi nternati
onalcourses.

E.SomeGapsi
nMat
eri
als

I
ftheconcl usionswehav ereachedint hisvolumear ecorr
ect,itimpli
esthatwehav e
somef airl
yser i
ousgapsi nourmat er
ial
s.Bef orelist
ingwher eIthinkthesegapsar e,letme
fi
rstofallnotethatmat eri
alsneedt omeett hesamer equirementsthatmet hodsdo,asl i
sted
i
nChapt ersI I
IandIV.Ifmat er
ial
sar esupposedt ohelpstudentsinlanguageacquisit
ion,they
shouldeithersupplyinputthatiscompr ehensible,i
nterest
ing/rel
evant,andnotgrammat i
call
y
sequencedt hemselv es,ortheyshoul dprovidest udentswi ththemeansofobt aini
ngsuch
i
nput.Ifmat eri
alsaremeantt ohelplanguagel earni
ng, t
hey

182
182
182
shouldf ocusonr ul
est hatarel ear
nabl
e,portable,et
c.Whil
el ear
ning materi
alscanbe
cri
ti
cized,itismyi mpressi
ont hatther
ei snol ackofmateri
alsforthi
spurpose,andthat
curr
entt extscanbeusef ulforthelanguagelearningcomponentofsecondlanguageand
for
eignl anguagecourses.Wewi l
lther
eforef
ocusonwhatsor tsofmat er
ial
sneedt obe
developedtoencourageacquisi
ti
on.

Thenewmat eri
alswi l
lbedesi gned,Ihope,tofi
llabasi cneed,helpingbeginningand
i
ntermediatestudentsobtaincompr ehensiblei
nputoutsidethecl assr
oom.Thi sisanobv ious
problem forforei
gnl anguagest udent s,andisespeci al
lycr ucialf
orstudentsof" exoti
c"
l
anguages.I tisalsoamaj orprobl em forstudent
sofcommonl yspokenl anguagesand
secondl anguagest udents;studentsatbegi nninglevelsdonothav ethecompet encet o
engagenat i
vespeaker sinconv er
sat i
on,andcannotunder standr adi
oandTVorr eadeasi ly
.
Weneedmat eri
als,i
naddi ti
ontot heinputprovi
dedbyt heclassroom, tobr
ingstudentstot he
pointwheretheycanut i
li
zet heoutsidewor l
d.

Oneobv iousandconv eni

entsourceofcompr ehensibl
einputi sreading.Asdi scussed
earl
ierint hischapter,pleasurer eadi
ng meetst her equir
ement st o quali
fyasi nputf or
acquisit
ion verywell.The pr obl
em wehav et odayist hatr eadersdesigned f orsecond
l
anguagest udentsdonotmeett heser equi
rements.Whati scur r
entlyavail
ableisof tennot
compr ehensibl
e;asment ionedear li
er,the onl
yr eadi
ngmanyf orei
gnl anguage st udents
encounterar eparagr
aphst hatareloadedupwi t
hcompl exvocabularyandsy ntax.Itisnear l
y
al
way sgr ammat i
call
ysequenced; writ
ersarecaref
ulonlytoincludesy nt
axt hatthestudenti s
supposedt ohav estudiedori scur r
entl
ylearni
ng.Also,thereissi mplynotenoughr eading
avai
lable.

Thesecondlanguagestudentneedsmassiv
eamountsofcomprehensi
ble,i
nter
est
ing
readi
ngmateri
al,enoughsot hathecanr eadforpl
easur
eand/ori
nterestf
oranhouran
eveni
ng,i
fhewantsto,forsev
eralmonths.

Somecur
rentt
ext
sar
eint
her
ightdi
rect
ion,
butt
heyar
efl
awedi
nsev
eral
way
s.

(
1)Theuseofexerci
ses,quest
ionsthattestst
udent
soncontentanddri
llt
hem ont he
grammarandv ocabul
aryused.Teachersare,ofcourse,f
reet
oignoretheseexerci
ses,but
theyoftent
akeupmostoft hepagesoft hereader.Whil
eitcanbear guedt
hatex er
cises
provi
de

183
183
183
l
earning,whilethet
extprovi
desacquisi
ti
on,It
hinkitisdangerouspract
icetotr
yt ocombi ne
thetwoi nthisway.Fi
rst
,thenecessi
tyofansweri
ngcont entquest
ionscanrui
nt hepleasure
ofreading.Second,
theyencour
agereadingmoreforform andlessf
orcontent
.

Theassumpt i
onunder l
yi
ngmanyoft heexercisesfoundinr eadersseemst obet hat
student sneed" r
eview"and" pr
acti
ce"onnewv ocabularyandgrammar ,other
wisetheywi l
lnot
retainit.Thisis,itseemst ome,asel f
-f
ulf
il
li
ngpr ophecy.Withfewerex erci
ses,students
mi ghtread mor e,and hav e a bet
terchance ofencount er
ing t
hese i t
ems int ext
s.Wi th
excessi veex er
cises,wemaybedest royi
ngourst udents'desi
ret or eadforpleasureand
i
nt erestint hesecondl anguage,thusinsuri
ngthatmanywi l
lindeednev erseet henew
structuresandwor dsagain.

(
2)Cur r
entreaderssimplydonotpr ovi
deenough.Par toft
heproblem i
st hei
ncl
usion
ofex er
cises,whichtakeupv aluablespace.Ther eaderoft hefut
urewi l
lbet hi
ck,f
ullof
readi
ng,andonv ari
edt opi
cs.Student swillbeabl
etopi ckandchoosetheirt
opics.Todothis,
theyneedal ottochoosef r
om.Si mplyincludi
ngonest oryaboutt
heWi neCountryofFr
ance,
anotheraboutsports,
andonemy ster
ystor yi
snotenough.

(3)Final
ly,writ
ersofsuchr eadersneedtoridt
hemselvesofthei l
lusi
ont hateachli
ne,
eachparagraph,mustcount ,and i ntr
oduce some new st r
ucture orv ocabular
yitem.As
emphasizedmanyt imesint hisbook,suchgrammaticalemphasiswi l
lseriousl
ydistor
tany
att
emptt owrit
eany thi
ngofinterest
.Weneednotwor ryabouteachline.Ifwepr ovi
deenough
comprehensibl
ei nput,
ever
ythingthestudentneedswi
llbether
e.

1.THELANGUAGELABORATORY

Asmanyr eadersknow,t her

ehasbeenagr eatdealofdiscussi
onanddebat
einthe
appli
edl ingui
sti
csliter
atur
eov ert hevir
tuesoft hel anguagelab.Inmyv i
ew,iti
snota
questionofwhet herthelabis" good"or"bad",butsimplywhetheritcanbeusedtosuppl
y
i
nputt hatisusefulforacqui
sit
ion,andtherebysupplementwhatwecanpr ovi
deincl
assand
i
nr eading.

Notonl
ycant
hel
abbeusedi
nthi
sway
,buti
tappear
stobet
hecase

184
184
184
thati ti sf areasier ,technological l
yspeaki ng,t o use t he lab asa meansofsuppl yi
ng
compr ehensibleinputt hanforot herpur poses.Thet raditi
onaluseoft helanguagelabput sa
tremendoust echnologi calandpedagogi calbur denont het eacher:theteacherisexpect edto
moni torst udentoutput ,andcor rectt hei
rer rors.Usingt hel abasasour ceofcompr ehensi bl
e
i
nputi seasi er.Herear esomepossi bil
it
ies:tapedst or i
es,wi t
hpi ct
ur estoaidcompr ehension
andaddt oenj oyment ,class-t
y pel ectures,suppl ement edwi t
hlect urenotes(onrealt opics,
designedt osuppl ementi nternati
onalcl asses,notsampl el ecturesonr andom aspect sof
chemi st r
yort hehistoryofapr etendki ngdom) ,radiopr ograms,commer ci
als,et
c.Inot her
wor ds, compr ehensibleinput,withsi mpleai dst ocompr ehension.

I
nmyv iew,thelabshouldbear esource,aplacestudent
scangotogeti
nputona
v
ari
etyofsubj
ectmatterswheneveritisconv enientf
orthem.Theoldv
iewoft
hel
ab,wi
tht
he
v
igi
l
antdri
llmaster
,doesnotall
owt his.
10

2.ACOMMENTONFI
ELDTESTI
NGOFMATERI
ALS

Thi
ssl ightl
ynewappr oacht omat eri
alsmi ghtalsonecessi t
ateasl i
ghtl
ynewappr oach
tofieldtesting.It hinkIcanbesti llustrat ethisbyr elat
ingaconv er
sati
onIhadsev eraly ear s
agowi th ar epresent ati
v e ofa publ i
shi nghouse t hati sacti
vei nbot h ESL andf or ei
gn
l
anguagemat eri
als.Hehadcomet oseemebecauseofourwor kont heor derofacqui siti
on
ofgr ammat icalst r
uct ures( e.g.Bai l
ey,MaddenandKr ashen,1974;Kr ashen etal .,1978;
Krashenetal .,1976;Houck,Rober sonandKr ashen,1978a)f eel
ingthatourwor k,andsi mi lar
worddonebyot herr esearcher s,mightgi vehiswr it
ersabet t
ersequencet obaset heirreader s
on.Heaccept editasagi venf actthatr eadersdesi gnedf orstudentsneededt obecont roll
ed
forstructur
es, andthatournat uralor derst udieswoul dpr ovi
deasuper iorbasisforthis.Book
one,f orexampl e,woul dcont ainonl yt hosest r
ucturesf oundt obeear lyacquir
ed,bookt wo
woul daddt hosest r
uct uresslightlyf art
herdownont henat ur
alorder,et
c.

Ihavearguedagainstthi
sphi
losophysev
eralti
mesinthi
svolume.AsStevi
ck(1980)
not
es,i
tleadstoast y
le"whichi
sli
nguist
ical
l
yanti
septi
candemoti
onall
yst
eri
l
e"(p.203;see
al
sohisexcel
lentdi
scussi
on,

185
185
185
pp.203- 204) .Ipr esented my argument s agai nst thi
s appr oach tot his publi
sher'
s
representati
ve,andaskedhim whatform offieldtestinghisr eadersunderwent.Hisresponse
wast hatt heli
nguist
icanaly
siswasdeemedsuf f
ici
ent :hi
spubl i
shi
nghousepr ovi
deswrit
ers
withagui de,indi
cati
ngwhatst r
uct
uresaret obei ncludedfordi f
fer
entlevel
s.Iftheproposed
text
sdoi ndeedonl ycontai
nthesest
ructures,theyareconsi der edworthyandhav epassedthe
test
.Hi spurposeinseeingmewast orevisethisguideaccor dingtothenaturalorder
.

Her eisanal ter

nat i
veappr oacht odev elopingandf iel
dt est i
ngr eader
s,onet hatis
consistentwi t
ht hephilosophysetf ort
hi nt hi
sbook.Thef irststepi stousewr it
erswhoar e
genuinelyinterestedint ell
i
ng,orr e-tell
ingast ory ,andwhoar eint erest
edi nandsy mpat heti
c
withtheaudi ence.Theysi mpl ywr it
e,f ocussingont hestory ,usingwhatt heyintui
tiv
elyf eel
theyneedt otelli
tandmakei tcompr ehensi bl
e( r ecal
lBrown' sadv icetopar ent
s,repeatedon
page65) .
Thef i
eldtestisnotasy nt acticanal ysis.Itisdonei nor dert oanswert hequest ions:
domember soft heintendedaudi ence under stand it? Do t heyenj oyi t?Dot heyfind i t
i
nteresting?Woul dt heyr eaditont heirown( notasanassi gnment )?Iftheanswer stot hese
questionsar eint heaffirmati
ve,secondl anguageacqui sit
iont heor ytell
sust hati+1wi l
lbe
ther
e,t hatther eadingisl i
nguist
ical l
yappr opriateandi twillhelpt hereaderacqui r
emor eof
thetargetlanguage.

We mayappl ysimil
arcriteri
at o otherkinds ofmat er
ial
s,i
.e.the l
ab materi
als
recommendedear l
ier,andmateri
alsdesignedt ohelpteachsubj
ectmat
ter(seeNot
e10) .Are
theycompr ehensi
ble?Ar etheyinterest
ing/
relevant
?et c.Onlythest
udentsandlanguage
acquir
erscananswert hesequest
ions.

Letusalsonotforgettheobviousquesti
onthatneedst obeaskedaboutal
lmat eri
als:
dotheyactual
lyresul
tinmor eprof
iciencyi
nthetargetlanguage?Thetheor
ypredict
st hatif
materi
alssati
sfyourrequir
ements,thiswil
lhappen,but,asemphasizedinChapterI,t
hisis
notenough.Appli
edli
nguist
icsr
esearchneedstoconfir
mi t.

F.SomePr
obl
ems

Evenifthet
heorypresent
edhereistot
all
ycor
rect
,andmysuggest
ionsf
orappl
i
cat
ion
ar
einf
acttheappr
opri
ateones,ther
earesome

186
186
186
seri
ousprobl
emst hatneedtobement ionedbeforeconcludi
ng.Thesehav etodowi t
ht he
accept
ance, by t eachers and students, of language acquisi
tion as primary, and
comprehensi
bleinputasthemeansofencour agi
nglanguageacquisi
ti
on.Theseprobl
emsar e
causedbyhefactt hatacqui
sit
iondi
ffersfrom l
earni
ngintwomaj orways:acqui
sit
i ssl
oni ow
andsubtl
e,whil
elearningisf
astand,forsomepeopl e,obv
ious.

Acquisiti
ont akest ime;ittakesfarmor ethanf ivehour sperweekov erninemont hsto
acqui r
et hesubj uncti
ve.I tmay ,iffact,t
akey ear s.Goodl ingui
sts,ont heot herhand,can
consci ousl
yl earnagr eatdeali nav eryshor
tt ime.Al so,whenwehav eacqui redsomet hing,
wear ehardlyawar eofi t.I
nasense,i tfeel
sasi fitwasal waysthere,andsomet hinganyone
cando.Lear ningi sdifferent.Somepeopl ederivegr eatpleasurefr
om thel earni
nganduseof
consci ousr ules,andIam oneoft hem!"Mast eri
ng"t hesubjuncti
vei nFr enchwasv ery
satisfyi
ngf orme, andIr ekindl
et hi
ssenseofv ictoryeverytimeIplanandsaysent encessuch
as" Ilfautquej 'ail
le"
.Iti ssomet imeshardforpeopl elikeust ounderstandt hatt hi
ssortof
pleasur abl
eact ivit
yisnotr eallanguageacquisiti
on.

Thi
sl eadst oonemaj orproblem.Languagecur r
iculum andt ext
sar edesignedby
peopleli
keus,peopl ewhol earnquicklyandwhoderivesatisfact
ionf r
om it(St
evick'
s"group
G",p.253;Stevick,1980).Thevastmaj or
it
yofourstudent
s,howev er,arenotasinter
estedin
thestr
uctureofl anguageaswear e,andgetthei
rpl
easureselsewher e!

Butwhataboutt hose st udentswho bel ieved us,and wi llonl yacceptconsci ous

grammaranddr il
last hecor eofal anguagecl ass,andwhoex pectal loft hei
rer r
orst obe
corrected( seee. g.Cat hcartandOl sen,1976) ?Icanonl yrecommendt wosor tsofsol ution,
onel ongt erm andoneshor tterm.I ftheessent i
alsoft hisbookar ecor rect,inthelongt erm,
thesest udentsandt heirteacher swi l
lbeeducat ed.Ideaschangesl owly,howev er
,andsome
short-t
erm solutionsar eneeded.Oneoft hese, suggestedbyTonyPf annkuche, i
stopr esenta
shortcour seonl anguageacqui siti
onaspar tofthelanguaget eachingpr ogram, orjustpriorto
i
t.Ithinkt hi
sisj usti
fied, especiall
yifweconcei veoft helanguagerequi rementi nhighschool s
andcol legesasi ncludingski l
lsandi nformationabouthowt oacqui reanyl anguage,notj ust
theonepr esentedi ncl ass.Anot herappr oach,andonet hatIam per sonallynotabov eusingi n
mycl asses, i
sdecept ion.Wecant eachv ocabularyor

187
187
187
grammar ,and,aslongasi tisdoneinthetargetlanguage,
agr eatdealofacquisi
ti
onwil
ltake
place,themedi um beingt hemessage.Wecant eachsi
tuationall
y,gi
vingstudent
suseful
,
shortdial
oguest hatsati
sfythecrav
ingforl
earningandmemor izedlanguage,
but,att
hesame
ti
me,pr esentcompr ehensi
bleinput
.Final
ly
,thesubj ectmatterinter
nati
onalcl
asseswil
lalso
providecompr ehensibl
einputforastudent,whetherhebelievesinsubconsciousacqui
sit
ion
ornot .

It
hinkthatIhavepr esent
edaconser vati
veviewofl anguageacqui si
tiont heoryandits
appli
cati
ons,conserv
ativeinthesenset hatitatt
empt st
obeconsi stentwithal lempiri
caldata
thatareknownt ome.Itisconsistentwiththewayt housandsofpeopl ehaveacqui r
edsecond
l
anguaget hr
oughouthistory,andi nmanycasesacqui redt hem v erywel l.Theyacqui red
second languagewhi l
et heywer ef ocussed onsomet hing el
se,whi let heywer egaining
i
nteresti
ngorneededinformat i
on,orint
eractingwit
hpeopl etheylikedtobewi th.

Not
es

1 Thesepri
nci
plesderi
vef
rom whatAsherconsi
der
stobet
het
hreecr
it
icalel
ement
sof
chi
l
dlanguageacqui
sit
ion:

(
1) l
ist
eni
ngi
nadvanceofspeaking:"
Itmaybethatl
ist
eningcompr
ehensi
onmaps
t
hebl
uepri
ntf
ort
hef
utureacqui
sit
ionofspeaki
ng"(
p.1041).

(
2) "
..
.t
he underst
anding ofspoken l
anguage may be acquir
ed when adul
ts
mani
pul
atethephysi
calbehavi
oroft
heinf
antthr
oughcommands. .
."

(3) "
..
.l
ist
eni
ng ski
l
lmaypr oduce a '
readiness'f
ort
he chi
l
dto speak.
..As
under
standi
ngdev el
ops,ther
ei sapointofr eadinesst
ospeakinwhi
cht hechil
d
spont
aneouslybegi
nstoproduceut
ter
ances"(
p.1041) .

2 Whilenotst r
ictlyamet hodcompar i
sonex peri
ment ,Newmar k'sMi ni malLanguage
TeachingPr ogram forf oreignl anguaget eachingatt heUni versitylev el,reportedi nNewmar k
(1971),isofgr eatinterest.Newmar k'
sstudent sspentt heirinstructionalweekasf ollows:three
hoursi nconv ersati
onsect ionswi thnativ
espeaker s;twohour sofex tensi ver eadi ng( "
inorder
toencour agescanni ngandr apidreading,assignment sar epur posel ylonger( 10- 20pages)t han
studentscanst udycr ytoanal y
ticall
y,andex ami nationsonr eadi ngspur poselyencour agerapid
sketchyr eading",p.16) ;threehour sint helabf orwor kondi alogues;andf ourhour swi th
"l
earning"ty peact i
vit
ies( studyofaconv enti
onalgr ammar ,readi nganddi scussi oni ngener al
l
inguisti
cs).Cl early,the f i
rstt hreepor t
ionsf ocuson acqui siti
on,wi tht heconv ersat
ional
secti
onsandex tensivereadi ngassi gnment sprovidingcompr ehensi bleinput .Newmar kreports
thathisstudent sconsi stentlyreacht heMLAnor msf ortwoy earsi nreadi ngaf teronl yoney ear
i
nhi sprogr am.

3 Insomecases, inter
nati
onalclassesarei
mpr acti
calori
mpossible.Oneex ampleisthe
l
argelectureclassinel
ement ar
ysciences.Apossibi
lit
yistheint
ernati
onaldiscussionsect
ion
and/or"pre-l
ectur
e"secti
on,inwhichdi ffi
cul
tvocabularyisexpl
ained,andt het opi
coft he
l
ecturediscussedinadvance.

188
188
188
4 I
nsomecases,i nternati
onalclassesarei
mpr acti
calori
mpossible.Oneex ampleisthe
l
argelectureclassinel
ement arysciences.Apossibi
li
tyistheint
ernati
onaldiscussionsect
ion
and/or"pre-l
ectur
e"secti
on,i nwhichdi ff
icul
tvocabular
yi sexpl
ained,andt het opi
coft he
l
ecturediscussedinadvance.

1. Abi
l
ityt
otakenot
esi
nlect
ures.

2. Abi
l
ityt
otakenot
esonwr
it
ten,
tex
tual
mat
eri
als.

3. Abil
it
ytoorganizeessayty
peex
ami
nat
ionquest
ionsandwr
it
e
accur
atel
yundert
hepressur
eofti
me.

4. Abi
li
tyt
or ecogni
zeandunderst
andt
het
hinki
ngst
rat
egi
esi
mpl
i
citi
n
obj
ect
ivet
ypetestquesti
ons(p.79)
.

Thei nternati
onalcl
ass,i
tcanbear gued,prov i
desanaturalsyll
abusf ortheacqui sitionofst udy
skil
l
s; needssuchast hoselist
edabov ewi l
lbemet,aswel l
asot hersnotpr edictedbyt heneeds
survey( seef oot
not
efiv
ef oranex ample) .Second,i
nter
nationalstudentsmaynotr egar d"st udy
skil
l
s"cl assesasessentialtot heirneedsandascont r
ibuti
ngdi rectl
yt ot heireducat ional
program ( alt
houghSchwabepoi ntsoutt hatherstudent
satt heUni versi
tyofCal i
forniaatDav is
appeart obemor emoti
vatedforandi nterestedinherESLcl i
nicthanregularESLcl asses) .They
maysi mpl ybeanotherobstacl
et ogetthr oughbeforest
udent scanpur suet heirmaj orinterest.

Thisargumentati
onandspecul
ati
onneedstobesupplementedwit
hr esear
chont he
appl
i
edlevel,
todeter
mi ne,whet
hert
hebestappr
oachi
sthecl
i
nicalone,t
heint
ernati
onalcl
ass
al
one,orsomecombi nati
on.

5 To gi vea concr eteex ampl eofan easi l

y-l
earnableaspectofcl assroom/ academic
behav i
or,Glor i
aHel l
erhaspoi nt edoutt omet hatseveralofherESLst udent swoul dhandi n
homewor kassi gnment sont hr eer i
ngnot ebookpaperwi t
ht her i
ngsont hewr ong( right)si
de
(onwhatweconsi dertobet hebackoft hepaper ).Thistriv
ialer r
ormi ghtbei nterpretedasa
signofsl oppinessi nar egularcl assandmi ghtnotbecor r
ect ed.Itwouldbeant ici
pat edorat
l
eastcor rectedi naninternati
onalst udents'class,andisagoodex ampl eofasi mple,learnabl
e
rulethatmakesar ealdif
ference.Usi ngthecor rectsideofthepapermaynotmakeast udenta
betterstudentori mpr ovehisgr aspofsubj ectmat t
er,butitwi l
laffecthisimagei ntheey esof
thet eacher." Lear
ning"smal laspect sofcl assroom andacademi cbehav iormayt hushav e
similarfunctionsasl earningl ate-acquiredaspect sofl anguage( ChapterI V):theymaynotbe
essentialforcommuni cati
on, butadd" poli
sh",givi
nganof t
eni mportantcosmet i
cef f
ect .

6 Her
ei sapossibl
esummeri ntensi
veprogram, meantfortheinternat
ionalst
udentwit
ha
few y
earsoff or
malEngli
shinstr
uctioninhisowncount r
ywhoi snoty etreadyforacademic
workinEngli
sh.Thegoaloft heprogram i
st oprovidesubjectmatterinstr
ucti
oninareasthat
are,atthe same ti
me,v eryrelevanttot he student
s'needs and i nter
ests,and thatare
l
ingui
sti
cal
l
ycompr ehensi
ble

(1)Cour
sework,taughtbysubjectmat t
erteacher
s.Thest
udentsel
ect
s
cour
sesfr
om al
istconsi
sti
ngofcoursessuchasthese:

(
i)Mat
hemat
icsr
evi
ew,
from al
gebr
athr
oughcal
cul
us.

(
ii
)Comput
eroper
ati
on(
notpr
ogr
ammi
ng)
.

(i
ii
)Ameri
canconsumereconomi
cs(
"Sy
lvi
aPor
ter
"),
incl
udi
ngcr
edi
t,banki
ng,
shoppi
ngstr
ategi
es,
etc.

189
189
189
 (
iv)Engl
i
shgr
ammar(
"l
anguageappr
eci
ati
on"
,orl
i
ngui
sti
cs)
.

(
v)Engl
i
shgr
ammarf
orMoni
toruse.

(2)Once a degree offluency is achi

eved,discussi
on gr
oups wit
h both more
exper
iencedforei
gnstudent
s( i
nEnglish)andwi t
hnativ
espeaker swhoar
ei nt
erest
edinthe
samear eaofst
udycansupplementtheformalcourseof
feri
ngs.

Mypr edi
cti
onist
hatsuchaprogr
am wouldresulti
nfarmoreacquisit
ionofEngli
sht han
the standardint
ensiv
e progr
am,would be per
cei
v ed ofasmor er elev
antbyi nternati
onal
students,andwouldbeofconsider
abl
evaluei
nfurt
heringthest
udents'educati
onalprogressin
hisspecialt
y.

7 Inareaswher ethereareenoughstudentst osupportsuchclasses,otherfor

msofsubject
matt
erteachingshoul dalsoworkinadul tESL,incl
udingjobr elat
edclassesf ori
mmigrantsor
non-
nati
vespeaker sofEngl i
sh,andtopicsofi nter
est,e.
g.introducti
ontoAmer i
canli
ter
atur
e,
Ameri
canspor ts,cooking,etc.Thepointi sthatanyt opicwi llworkasl ongast heinputis
comprehensi
bleandt hestudentsaregenuinel
yinterest
edinthesubj ectmatter.

8 AsCar r
oll(1980)notes:"Itisonlynat uralforstudentstoshapet hei
rlearningeffor
tsso
ast obemax i
mal l
ysuccessf ulont ests,andi ft hetestsmeasureobj ecti
vesthatar einsome
way sdifferentfr
om t hoseoft hei nstr
uction,st udentswillworktowardst hoseobj ecti
vesand
paylessat tenti
ont oachi ev
ingot herobjectiv
es.Thenat ureofexternalexaminationswi l
loft
en
shapet hebehav ioroft het eachersthemsel v
es.Wesomet i
mescompl ainthatt eacher
sdo
nothingbut
'
teachf orthetests'"(
p.528).

9 Ihav enot otallysatisfactor

yt opi
cst osuggestt hatar e"real"andt hatpresentr eal
problemst obesol ved.I naconsul t
ingsessi onwi thKarlSchev i
ll
e's" PEFL"pr ojectatt he
Universi
tyofCal i
fornia atBer keley(Depar t
mentofEducat i
on),If eelwe came cl ose t o
developingsome.Her eisoneex ample.Exami nerandstudentar egivent hef ol
lowingsi tuati
on:
theyar esi bli
ngs,andl ivei nasmal lapar tmentwi thal argef amily.Al lthechi l
dr enshar e
bedrooms.Theol destbr otherhasdeci dedt hathewant sar oom ofhi sown.Af amilymeet i
ng
needst ot akepl acet odeci dewhatt odo,becausei ft
hebr othergetshi sway ,t
her ewi l
lbe
i
ntolerablespaceconst r
aintsont herestoft hef amil
y.Theex aminerandst udentdi scusst he
sit
uation,withthegoal ofrecommendi ngt othef amil
ywhatt hepossiblesol ut
ionsar e.Thet opic
i
snot" r
eal",si
ncei tisacont r
ivedsi t
uati
on,buti nourrehearsals,wef oundt hatitwaspossi ble
tostimulatesomei nteresti
ngbackandf ort
hdi scussi
on.

10 Thei nt
ernati
onalclassesIproposedearl
ier
,speci
alsectionsofsubjectmatt
erclasses
forint
ernat
ionalstudents,mightalsoprof
itf
rom speci
almateri
als.Thesemightincl
udetextsin
areaswheresl i
ghtlyeasierreadingi
snotav ai
l
able,suppl
ementst oexist
ingtext
s,and,asjust
ment i
oned,tapedlectur
essuppl ementedwit
hnotes.

190
190
190
Bi
bli
ogr
aphy
ALLWRI
GHT, R.(
1975)Probl
emsinthestudyofthelanguaget
eacher
'str
eatmentoferr
or.I
nM.BurtandH.Dulay(
Eds.
)
NewDir
ect
ionsinSecondLanguageLear
ning,Teachi
ng,andBil
i
ngualEducat
ion.Washi
ngton,
D.C.
:TESOL.pp.96-
109.

ANDERSEN, R.(
1976)Afunct
ionalacqui
sit
ionhi
erar
chyst
udyi
nPuer
toRi
co.Paperpr
esent
edatt
he10t
hannual
TESOL
conf
erence,
NewYor k,
NewYor k,March,1976

ANDERSEN,
R.(
1978)Ani
mpl
i
cat
ional
model
forsecondl
anguager ch.LanguageLear
esear ning28:221-
282.

ASHER,J.(
1965)Thestr
ategyoft
het
otal
phy
sical
response:
anappl
i
cat
iont
olear
ni an.I
ngRussi nter
nat
ionalRev
iewof
Appl
iedLi
ngui
sti
cs3:291-300.

ASHER,
J.(
1966)Thel
ear
ningst
rat
egyoft
het
otal
phy
sical
response:
arev
iew.Moder
nLanguageJour
nal50:79-
84.ASHER,

J.(
1969)Thet
otal
phy
sical
responseappr
oacht
osecondl
anguagel
ear
ning.Moder
nLanguageJour
nal53:3-
17.ASHER,
J.

(1972)Chi
ldr
en'
sfi
rstl
anguageasamodel
forsecondl
anguagel
ear
ning.Moder
nLanguageJour
nal56:
133-139.

ASHER,
J.(1977a)Lear
ningAnot
herLanguageThr
oughAct
ions:TheCompl
eteTeacher
'sGui
debook.LosGat
os,
Cal
i
f:Sky
OaksPr
oducti
ons.

ASHER,
J.(
1977b)Chi
l
drenl
ear
ninganot
herl
anguage:
adev
elopment
alhy
pot s.Chi
hesi l
dDev
elopment48:1040-
1048.

ASHER,J.andPr i
ce,B.(1967)Thel
ear
ningst
rat
egyoft
het
otalphy
sicalr
esponse:someagedi
ff
erences.Chi
l
d
Devel
opment38:1219-
1227.

ASHER,J.
,KUSUDO, J.andDELATORRE,
R.(
1974)Lear
ningasecondl
anguaget
hroughcommands:
thesecondf
iel
dtest
.
ModernLanguageJournal58:24-
32.

BAILEY,N.
,MADDEN, C.andKRASHEN,
S.(
1974)I
sther
ea"
nat
ural
sequence"i
nadul
tsecondl
anguagel
ear
ning?
LanguageLear
ning21:235-
243.

BANCROFT, J.(1978)TheLozanovmet
hodandi
tsAmer
icanadapt
atons.Moder
i n
LanguageJournal62:167-
174.

BI
ALYSTOCK,E.andFROHLI
CH,
M.(
1977)Aspect
sofsecondl
anguagel
ear
ningi
ncl
assr
oom set
tngs.Wor
i kingPaper
son
Bi
l
ingual
i
sm 13:1-
26.

BI
ALYSTOCK,E.andFROHLI
CH,
M.(
1978a)Theaur
algr
ammart
est
:descr
ipt
ionandi
mpl
i
cat
ions.Wor
kingPaper
son
Bi
l
ingual
i
sm 15:15-
35.

BIALYSTOCK,E.andFROHLI CH,
M.(
1978b)Var
iabl
esofcl
assr
oom achi
evementi
nsecondl
anguagel
ear
ning.Moder
n
LanguageJournal62:327-
335.

BIRKBI
CHLER,D.(1977)Communicati
onandbey ond.I
nJ.Phi
l
li
ps( )TheLanguageConnect
Ed. ion:Fr
om t
heCl
assr
oom t
o
theWorl
d.Skoki
e,Il
l.
:Nati
onal
Textbook.pp.53-
94.

BRI
ERE,
E.(1978)Var
iabl
esaf
fect
ingnat
iveMexi
canchi
l
dren'
slear
ningSpani anguage.LanguageLear
shasasecondl ning
28:
159-
174.

BROWN, J.(
1980)Anexplanati
onofmor
pheme-groupi
nter
act
ions.Paperpr
esent
edatt
heLosAngel
esSecondLanguage
Acqui
sit
ionResear
chForum, UCLA,
Febr
uar
y,1980.

BROWN, 1973)AFi
R.( rstLanguage.Cambr
idge:
Har
var
dPr
ess.

191
191
191
BROWN, R.(
1977)I
ntr
oduct
ion.I
nC.SnowandC.Fer
guson( )Tal
Eds. ki
ngt
oChi
l
dren.NewYor
k:Cambr
idgeUni
ver
sit
y
Press.pp.1-
27.

BROWN,R.
,CAZDEN,C.andBELLUGI
,U.(1973)Thechi
ld'
sgrammarfr
om ItoII
I.I
nC.FergusonandD.Sl
obi
n(Eds.
)
St
udi
esofChi
ldLanguageDev
elopment
.NewYor k:Hol
tRinehar
tandWinst
on.pp.295-
333.

BROWN,S.(1979)Li
fesi
tuat
ions:
incor
por
ati
ngcommuni
tyr
esour
cesi
ntot
headul
tESLcur
ri um.CATESOLOccasi
cul onal
Paper
s5:48-65.

BRUCE,L.(
1979)Theacquisi
ti
onofgr
ammat
ical
mor
phemesbyadul
tst
udent
sofRussi
anasaf
orei
gnl
anguage.MAPaper
,
Depar
tmentofLi
nguist
ics,
USC.

BRUCK,M.,LAMBERT,W.andTUCKER, G.R.(1974)Bil
ingual
school
i
ngt
hrought
heel
ement
arygr
ades:
theSt
.Lamber
t
Proj
ectandgr en.LanguageLear
adesev ning24:183-204.

BURT,
M.andKI
PARSKY, 1972)TheGoof
C.( icon:ARepai
rManualf
orEngl
i
sh.Rowl
ey,
Ma:
Newbur
yHouse.

BUSHMAN,R.andMADSEN,H.(1976)Adescri
pti
onandev
aluat
ionofSuggest
opedi
a--
anewt
eachi
ngmet
hodol
ogy
.InJ.
Fansel
owandR.Cr
ymes( )OnTESOL'
Eds. 76.Washi
ngt
on:TESOL.pp.29-
38.

CAMPBELL, D.andSTANLEY, 1963)Exper

J.( iment
alandQuasi
-Exper
iment
alDesi
gns
forResear
ch.NewYork:RandMcNall
y.

CANCINO,H.
,ROSANSKY,E,andSCHUMANN,
J.(
1975)Theacqui
sit
ionoft
heEngl
i
shauxi
l
iar
ybynat
iveSpani
shspeaker
s.
TESOLQuart
erl
y9:421-
430.

CARROLL,
J.(1966)Thecont
ribut
ionsofpsy
chol
ogi
calt
heoryandeducati
onalresear
cht
oteachi
ngoff
orei
gnl
anguages.I
n
A.Val
dman(Ed.)Tr
endsinLanguageTeachi
ng.NewYork:
McGr aw-
Hil
l.pp.93-
106.

CARROLL,J.(1967)Forei
gnlanguagepr
ofi
ciencyl
evel
sat
tai
nedbyl
anguagemaj
orsneargr
aduat
ionf
r ege.For
om col
l eign
LanguageAnnals1:131-151.

CARROLL, J.(
1980)For
eignl
anguagetesti
ng:per
sist
entpr
obl
ems.I
nK.Cr
oft( )Readi
Ed. ngsonEngl
i
shasaSecond
Language.Cambri
dge,Ma:Wint
hrop.pp.518-
530.

CATHCART,R.andOLSEN,
J.(1976)Teacher
s'andst
udents'
prefer
enceforcor
rect
ionofcl
assr
oom conv
ersat
ioner
ror
s.I
n
J.Fansel
owandR.Cry
mes(Eds.)OnTESOL'76.Washi
ngton;TESOL.pp.41-
53.

CAZDEN,C.(1979)Cur
ri
cul
um/languagecont
ext
sforbili
ngual
educat
ion.I
nE.Bri
ere(
Ed.)LanguageDevel
opmentina
Bi
li
ngualSet
ti
ng.Pomona,Cal
i
fornia:
Nati
onalMult
il
ingualMul
ti
cul
tur
alMater
ial
sDevel
opmentCent er
.pp.129-
138.

CELCE-MURCIA,M.andROSENZWEIG,
F.(1979)Teachingvocabul
aryi
ntheESLcl
assr
oom.I
nM.Cel ce-
Murci
aandL.
McInt
osh(Eds.)Teachi
ngEngl
i
shasaSecondorForeignLanguage.Rowl
ey,Ma.
:Newbur
yHouse.pp.241-
257.

CHASTAIN,K.(1970)Amethodol
ogicalstudycompari
ngt
heaudi
o-l
i
ngual
habi
ttheor
yandt
hecogni
ti
vecodel
ear
ningt
heor
y:
acont
inuat
ion.ModernLanguageJournal54:257-
266.

CHASTAIN,K.andWOERDEHOFF,F.(
1968)Amethodological
studycompar
ingt
heaudi
o-l
i
ngual
habi
ttheor
yandt
he
cogni
ti
vecode-l
ear
ningt
heor
y.Moder
nLanguageJournal52:268-279.

CHIHARA,T.andOLLER,J.(1978)At
ti
tudesandat
tai
nedpr
ofi
ciencyi
nEFL:
asoci
oli
ngui
sti
cst
udyofadul
tJapanese
s.LanguageLear
speaker ning28:55-
68.

CHOMSKY, 1965)Aspect
N.( soft
heTheor
yofSy
ntax.Cambr
idge:
MITPr
ess.

CHRISTI
SON,M.(1979)Nat
ural
sequenci
ngi
nadul
tsecondl
anguageacqui
sit
ion.
TESOLQuart
erl
y13:122.

192
CLARK,E.andANDERSEN,E.(
1980)Spont
aneousr
epai
rs:
awar
enessi
nthepr
ocessofacqui
ri anguage.Paper
ngl sand
Repor
tsinChil
dLanguageDev
.16:1-
12.

CLARK,
H.andCLARK, 1977)Psy
E.( chol
ogyandLanguage.NewYor
k:Har
cour
tBr
aceJov
anov
ich.

COHEN,A.andROBBINS,M.(
1976)Towardsassessingint
erl
anguageper
formance:
ther
elat
ionshi
pbet
weensel
ect
inger
ror
s,
l
ear
ner'
scharact
eri
sti
cs,
andl
earner
'sexpl
anat
ions.LanguageLearni
ng26:45-
66.

COHEN,A.andSWAI
N,M.(
1976)Bi
l
ingual
Educat
ion:
the"
Immer
sion"model
int
heNor
thAmer
icancont .TESOLQuar
ext ter
ly
10:
45-53.

CORDER,
S.P.(
1967)Thesi
gni
fi
canceofl
ear
ner
'ser
r s.I
or nter
nat
ionalRev
iewofAppl
i
edLi
ngui
sti
cs5:161-
170.

CROSS,T.(1977)Mother
'sspeechadj
ustment
s:thecont
ri
buti
onsofsel
ectedchi
ldl
ist
enerv
ari
abl
es.I
nC.SnowandC.
Fer
guson,Talki
ngtoChil
dren.NewYork:
Cambr i
dgeUniv
ersi
tyPr
ess.pp.151-
188.

d'
ANGLEJAN, A.(
1978)Languagel
earni
nginandoutofcl
assr
ooms.InJ.Ri
char
ds( )Under
Ed. standi
ngSecondand
For
eignLanguageLear
ning.Rowl
ey,
Ma: Newbur
yHouse.pp.218-
236.

DEVILLERS,P.andDEVILLERS,J.(
1973)ACr
oss-secti
onal
studyoft
heacqui
sit
ionofgr
ammat
ical
mor
phemesi
nchi
l
d
speech.Jour
nalofPsy
choli
ngui
sti
cResear
ch2:267-278.

DI
LLER, 1978)TheLanguageTeachi
K.( ngCont
rov
ersy
.Rowl
ey,
Ma.
:Newbur
yHouse.

DULAY,
H.andBURT,
M.(
1974)Nat
ural
sequencesi
nchi
l
dsecondl
anguageacqui
sit
ion.LanguageLear
ning24:37-
53.

DULAY,H.andBURT,M.(1975)Anewappr oachtodi
scover
inguni
versalst
rategi
esofchi
l
dsecondlanguageacqui
sit
ion.I
n
D.Dato( )Dev
Eds. el
opmentalPsychol
i
nguisti
cs:Theor
yandAppl icat
ions.Georget
ownUniver
sit
yRoundTabl eon
LanguagesandLi
ngui
sti
cs.Washi
ngton:Georget
ownUniver
sityPr
ess.pp.209-233.

DULAY,H.andBURT,
M.(1977)Remar
ksoncreat
ivi
tyi
nlanguageacquisi
ti
on.InM.Bur
t,H.Dul
ayandM.Fi
nnochi
aro
( )Vi
Eds. ewpoint
sonEngl
i
shasaSecondLanguage.NewYork:Regents.pp.95-
126.

DULAY,H.andBURT,M.(
1978)Somegui
del
i
nesf
ort
heassessmentofor
all
anguagepr
ofi
ci nance.TESOL
encyanddomi
Quar
ter
ly12:177-
192.

DULAY,
H.,
BURT, S.LanguageTwo.NewYor
M.andKRASHEN, k:Ox
for
d.I
npr
ess.

DUSKOVA,
L.(
1969)Onsour
cesofer
rori
nfor
eignl
anguagel
ear
ning.I
nter
nat
ionalRev
iewofAppl
i
edLi
ngui
sti
cs4:11-
36.

EKSTRAND, L.(
1976)Ageandlengt
hofr esi
denceasv ari
ablesrel
atedt
ot headj
ustmentofmigrantchi
ldr
en,wi
thspeci
al
refer
encetosecondl
anguagel
earning.I
nG.Nickel(Ed.)Proceedi
ngsoftheFourt
hInter
nati
onalCongressofAppl
ied
Lingui
sti
cs.Vol
.3.St
utt
gart
:HochschulVerl
ag,pp.179-197.

ERVI
N-TRI
PP,S.( 1973)Somest r
ategiesforthefir
standsecondy ear
s.I
nA.Di
l( )LanguageAcqui
Ed. sit
ionand
Communicat
iveChoice.St
anf
ord:
StanfordUni
versi
tyPress.pp.204-
238.

ERVI
N-TRI
PP,
S.(
1974)I
ssecondl
anguagel
ear
ningl
i
ket
hef
ir
st?TESOLQuar
ter
ly8:111-
127.

FABRIS,
M.(
1978)Theacqui
sit
ionofEngl
i
shgr
ammat
ical
funct
orsbychi
l
dsecondl
anguagel
ear s.TESOLQuar
ner ter
ly12:
482.

FANSELOW,
J.(
1977)Thet
reat
mentofer
rori
nor
al k.For
wor eignLanguageAnnal
s10:583-
593.

193
193
193
FATHMAN,A.(
1975)Ther
elat
ionshi
pbet
weenageandsecondl
anguagepr
oduct
iveabi
l
it.LanguageLear
y ning 25:
245-
266.

FATHMAN, A.(
1979)Thev
alueofmor
phemeor
derst
udi
esf
orsecondl
anguagel
ear
ning.Wor
kingPaper
sonBi
l
ingual
i
sm
18:179-
199.

FATHMAN, A.(1980)I
nfl
uencesofageandset
ti
ngonsecondl
anguageor
alpr
ofi
ciency
.Paperpr
esent
edatLosAngel
es
SecondLanguageResearchForum,UCLA,
Febr
uary
,1980.

FELIX,S.(
1980)Theeff
ectoffor
mal i
nstr
ucti
ononsecondl
anguagel
ear
ning.Paperpr
esent
edatLosAngel
esSecond
LanguageResearchFor
um, UCLA,Febr
uary,
1980.

FREED,B.(1980)Tal
kingtofor
eigner
sversustal
ki
ngtochil
dren:si
mil
ari
ti
esanddiffer
ences.I
nR.Scar
cel
l
aandS.Kr
ashen
( )Resear
Eds. chinSecondLanguageAcquisi
ti
on.Rowl
ey,Ma.:NewburyHouse.pp.19-27.

GAIES,S.(1977)Thenatur
eofl i
ngui
sti
ci nputi
nf or
mall anguagelear
ning:l
ingui
sti
candcommunicat
ivestr
ategi
esi
nESL
teacher
s'classr
oom l
anguage.InH.D.Br own,C.Yor i
oandR.Cr y
mes( )Teachi
Eds. ngandLear
ningEngli
shasaSecond
Language:Trendsi
nResearchandPract
ice.Washington:TESOL.pp.204-212.

GARDNER,
R.andLAMBERT, 1972)At
W.( ti
tudesandMot
ivat
ioni
nSecond-
LanguageLear
ning.Rowl
ey,
Ma.
:Newbur
y
House.

GARY,J.O.(1975)Del
ayedoral
pract
icei
nini
ti
alst
agesofsecondlanguagelear
ning.I
nM.BurtandM.Dulay( )On
Eds.
TESOL'75:NewDi r
ect
ionsi
nSecondLanguageLearni
ng,
TeachingandBili
ngualEducat
ion.Washi
ngt
on:
TESOL.pp.
89-
95.

GILLI
S,M.andWEBER,R.(
1976)Theemer
genceofsent
encemodal
i
tiesi
ntheEngl
i
shofJapanese-
speaki
ngchi
l
dren.
LanguageLear
ning26:
77-
94.

HAKUTA,K.(
1974)Aprel
iminar
yrepor
tofthedevel
opmentofgr
ammat
ical
mor
phemesi
naJapanesegi
rll
ear
ningEngl
i
shas
anguage.Wor
asecondl kingPapersonBi
li
ngual
ism 3:18-
43.

HALE,
T.andBUDAR,
E.(
1970)Ar
eTESOLcl
assest
heonl ?Moder
yanswer nLanguageJour
nal54:487-
492.

HALL, 1959)TheSi
E.( l
entLanguage.Gr
eenwi
ch,
Conn.
:Fawcet
t.

HAMMARBERG,B.(1979)Onintr
ali
ngual,
inter
li
ngualanddev
elopmental
sol
uti
onsi
nint
erl
anguage.Paperpr
esent
edatt
he
Fi
ft
hScandi
nav
ianConferenceofLi
nguist
ics,Fr
ostval
len,
Apr
il
,1979.

HAMMERLY,H.(
1975)Thededuct
ion/
induct
ioncont
rov
er .Moder
sy nLanguageJour
nal
LIX:
15-
18.

HANANI
A,E.andGRADMAN, H.(
1977)Acquisit
ionofEngl
ishst
ruct
ures:
acasest
udyofanadul
tnat
ivespeakeri
nan
Engl
i
sh-
speaki
ngenv
ir .LanguageLear
onment ni
ng27:75-92.

HARRI
S, 1969)Test
D.( ingEngl
i
shasaSecondLanguage.NewYor
k.McGr
awHi
l
l.

HARTNETT, D.(1974)TheRel
ati
onofCognit
iveSt
yleandHemispher
icPr
efer
encet
oDeduct
iveandI
nduct
iveSecond
LanguageLearning.MAThesi
s,Depar
tmentofEngl
ish,(
TESL)
,UCLA.

HATCH, E.(
1972)Somest
udi
esi
nsecondl
anguagel
ear
ning.UCLAWor
kpaper
sinTeachi
ngEngl
i
shasaSecond
Language6:29-
36.

HATCH,E.(
1976)Languagei
nout
erspace.Paperpr
esent
edatt
heUCLA-
USCSecondLanguageAcqui
sit
ionFor
um,
Fal
l
,
1976.

HATCH,E.(
1978a)Di
scour
seanalysisandsecondl
anguageacqui
sit
ion.I
nE.Hat
ch( )SecondLanguageAcqui
Ed. sit
ion.
Rowl
ey,
Ma.:NewburyHouse.pp.401-435.

HATCH,
E.(
1978b)I
ntr
oduct
ion.I
nE.Hat
ch( )SecondLanguageAcqui
Ed. sit
ion.Rowl
ey,
Ma.
:Newbur
yHouse.pp.1-
18.

194
194
194
HATCH,
E.(
1979)Appl
ywi
thcaut
ion.St
udi
esi
nSecondLanguageAcqui
sit
ion2:123-
143.

HATCH,E.
,SHAPI
RA,
R.andGOUGH,
J.(
1978)"
For
eigner
-t
alk"di se.I
scour TL:Rev
iewofAppl
i
edLi
ngui
sti
cs39-
40:
39-
60.

HAWKI
NS, 1978)Def
J.( ini
tenessandI
ndef
ini
teness:ASt
udyi
nRef
erenceandGr
ammat
ical
i
tyPr
edi
cti
on.London:Cr
oom
Hel
m.

HENDRICKSON,J.(
1978)Er
rorcor
rect
ioni
nforei
gnlanguageteaching:
recenttheor
y,r
esear
ch,
andpr
act
ice.I
nK.Cr
oft
(
Ed.)Readi
ngsonEngl
ishasaSecondLanguage.Cambridge,
Ma. :
Wi nt
hrop.pp.153-
175.

HOUCK,N.
,ROBERTSON, J.andKRASHEN, S.(
1978)Ont hedomai
noftheconsci
ousgr
ammar
:mor
phemeor
der
sfor
cor
rect
edanduncor
rect
edESLst udentt
ranscr
ipt
ions.TESOLQuar
ter
ly12:335-
339.

HOUCK,N.,
ROBERTSON,
J.andKRASHEN,
S.(
1978b)Whathappensi
ner
rorcor
rect
ion.Abst
ractsubmi
tt
edt
o1978
TESOLConfer
ence.

HYLTENSTAM,
K.(
1977)I
mpl
i
cat
ional
pat
ter
nsi
nint
erl
anguagesy
ntaxv
ari
aton.LanguageLear
i ning27:383-
411.I
NHELDER,

B.
,andPIAGET, 1958)TheGr
J.( owt
hofLogi
calThi
nki
ngf
rom Chi
l
dhoodt
oAdol
escence.NewYor
k:Basi
c
Books.

JOHNSON, T.andKRUG,K.(1980)I
ntegr
ati
veandinst
rument
almoti
vati
ons:i
nsear
chofameasur
e.I
nJ.Ol
l
erandK.Per
kins
( )Resear
Eds. chinLanguageTesti
ng.Rowley,
Ma.:
NewburyHouse.pp.241-
249

JONES, R.(
1979)Per
for
mancet
est
ingofsecondl
anguagepr
ofi
ciency
.InE.Br
ier
eandF.Hi
nof
oti
s( )Concept
Eds. sin
LanguageTesti
ng.Washi
ngt
on:
TESOL.pp.50-
57.

JORDENS,P.andKELLERMAN, E.(
1978)Invest
igati
onint
othest
rat
egyoft
ransf
eri
nsecondl
anguagel
ear
ning.Paper
present
edatAI
LAconference,
Montreal
,August,1978.

KAYFETZ,J.(
Full
er 1978)Nat
)( uralandMoni
tor
edSequencesbyAdul
tLear
ner
sofEngl
i
shasaSecondLanguage.Ph.
D.
di
ssert
ati
on.Fl
ori
daStateUni
versity
.

KELLERMAN,E.1978)Givi
nglearner
sabreak:
nat
ivel
anguagei
ntui
ti
onsasasour
ceofpr
edi
cti
onsaboutt
ransf
erabi
l
ity
.
Worki
ngPapersonBil
ingual
ism 15:59-
92.

KESSLER,
C.andIDAR,I
.(1977)Theacquisi
ti
onofEngl
i
shsy nt
act
icst
ruct
uresbyaVi
etnamesechi
l
d.Paperpr
esent
edatt
he
LosAngel
esSecondLanguageAcquisi
ti
onForum,UCLA,1977.

KLEI
NMAN,
H.(
1977)Av
oidancebehav
iori
nadul
tsecondl
anguageacqui
sit
ion.LanguageLear
ning27:93-
107.

KLI
MA,E.andBELLUGI,
U.(1966)Sy
ntacti
cregul
ari
ti
esinthespeechofchil
dren.I
nJ.Ly
onsandR.Wal
es(
Eds.
)
Psy
chol
ingui
sti
cPaper
s.Edinbur
gh:
EdinburghUni
versi
tyPress.pp.183-
208.

KOUNIN,T.andKRASHEN,S.( 1978)Appr
oachingnat
ivespeakercompetencefrom t
wodiff
erentdi
rect
ions.InC.
Bl
atchf
ordandJ.Schacht
er( )OnTESOL'
Eds. 78:EFLPol
i
cies,
Programs,
Practi
ces.Washi
ngt
on:
TESOL.pp.205-212.

KRASHEN,S.(
1976)For
mal
andi
nfor
mal
li
ngui
sti
cenv
ironment
sinl
anguagel
ear
ningandl
anguageacqui
sit
ion.TESOL
Quar
ter
ly157-
168.

KRASHEN,S.(
1977)Someissuesrel
ati
ngt
ot heMonit
orModel.I
nH.D.Brown,C.Yor
ioandR.Crymes( )OnTESOL
Eds.
'
77:Teachi
ngandLear
ningEngl
ishasaSecondLanguage:Tr
endsinResear
chandPract
ice.Washi
ngt
on:TESOL.pp.
144-
158.

KRASHEN, S.(
1978)Indi
vi
dualvari
ati
onintheuseoft
heMoni
tor
.InW.Ri
tchi
e( )Pr
Ed. inci
plesofSecondLanguage
Lear
ning.NewYork:AcademicPress.pp.175-
183.

195
195
195
KRASHEN,S.(
1980)Thetheoret
ical
andpr
acti
calrel
evanceofsi
mplecodesi
nsecondl
anguageacqui
sit
ion.I
nR.Scar
cel
l
a
andS.Kr
ashen( )Resear
Eds. chi
nSecondLanguageAcquisi
ti
on.Rowl
ey,Ma.
:Newbur
yHouse.pp.7-
18.

KRASHEN, 1981)SecondLanguageAcqui
S.( sit
ionandSecondLanguageLear
ning.Oxf
ord:
Per
gamonPr
ess.

KRASHEN,S.(
1982)Newmar
k's"
Ignor
anceHy
pot
hesi
s"andcur
rentsecondl
anguageacqui
sit
iont
heor
y.Unpubl
i
shed
manuscr
ipt
.

KRASHEN,S.
,SELI
GER,
H.andHARTNETT,
D.(
1974)Twost
udi
esi
nsecondl
anguagel
ear
ning.Kr
it
ikonLi
tt
erar
um 3:
220-
228.

KRASHEN,
S.andPON,
P.(
1975)Aner
roranal
ysi
sofanadv
anceESLl
ear .Wor
ner kingPaper
sonBi
l
ingual
i
sm 7:125-
129.

KRASHEN, S.
,MADDEN, C.andBAILEY,N.(1975)Theor
eti
calaspect
sofgrammat
ical
sequenci
ng.InM.Bur
tandH.Dul
ay
( )Secondl
Eds. anguageLear
ning,
Teaching,andBil
i
ngualEducati
on.Washi
ngt
on:
TESOL.pp.44-54.

KRASHEN,
S.andSELI
GER,
H.(
1975)Theessent
ial
char
act
eri
sti
csoff
ormal
inst
ruct
ion.TESOLQuar
ter
ly9:173-
183.

KRASHEN,S.andSELIGER,H.(1976)Therol
eoffor
mal andinf
ormal
li
ngui
sti
cenv
ironment
sisadul
tsecondl
anguage
l
ear ng.I
ni nter
nat
ionalJour
nalofPsychol
ingui
sti
cs3:15-21.

KRASHEN, S.
,SFERRLAZZA,V.FELDMAN, L.andFATHMAN, A.(1976)Adul
tperf
ormanceontheSLOPEt
est
:mor
eev
idence
foranat
uralsequencei
nadultsecondl
anguageacquisi
ton.LanguageLear
i ning26:145-
151.

KRASHEN,S.
,HOUCK,N.,
GIUNCHI,P,
BODE,S.
,BI
RNBAUM,R.andSTREI,J.(
1977)Diff
icul
tyor
derf
orgr
ammat
ical
morphemesf
oradul
tsecondl
anguageper
for
mersusi
ngf
reespeech.TESOLQuarter
ly11:338-
341.

KRASHEN,S.,
BUTLER,J.
,BI
RNBAUM,R.,andROBERTSON,
J.(
1978)Twost
udi
esi
nlanguageacqui
sit
ionandl
anguage
l
ear ng.I
ni TL:Rev
iewofAppl
iedLi
ngui
sti
cs39-40:
73-
92.

KRASHEN,S.,andSCARCELLA,
R.(
1978)Onr
out
inesandpat
ter
nsi
nlanguageacqui
sit
ionandper
f mance.Language
or
Lear
ning28:283-
300.

KRASHEN,S.,
ZELINSKI,S.,JONES,
C.andUSPRI
CH,
C.(
1978)Howi
mpor
tanti
sinst
ruct
ion?Engl
i
shLanguage
Teachi
ngJour
nal32:257-261.

KRASHEN,S.
,LONG,M.andSCARCELLA,
R.(
1979)Age,
rat
eandev
ent
ual
att
ainmenti
nsecondl
anguageacqui
sit
ion.
TESOLQuart
erl
y13:573-
582.

KUNIHARA,S.andASHER,J.(
1965)Thestr
ategyofthet
otal
phy
sical
response:
anappl
i
cat
iont
olear
ningJapanese.
I
nter
nati
onalRevi
ewofAppli
edLingui
sti
cs4:277-
289.

LADO, 1964)LanguageTeachi
R.( ng:ASci
ent
if
icAppr
oach.NewYor
k:McGr
awHi
l
l.

LADO,
R.andFRI
ES, 1958)AnI
C.( ntensi
veCour
sei
nEngl
i
sh.AnnAr
bor
:Uni
ver
sit
yofMi
chi
ganPr
ess.

LAMBERT,
W.andTUCKER, 1972)TheBi
G.R.( l
ingualEducat
ionofChi
l
dren.Rowl
ey,
Ma.
:Newbur
yHouse.

LAMENDELLA,
J.(
1979)Lect
urespr
esent
edatt
he1979TESOLSummerI
nst
it
ute,
UCLA,
Engl
i
sh272K.LARSEN,

D.(
1975)ARe-
eval
uat
ionofgr
ammat
ical
str
uct
ur ng.OnTESOL'
esequenci 74.Washi
ngt
on:
TESOL.

LARSEN-FREEMAN,D.(1975)TheAcquisi
ti
onofGrammat
ical
Mor
phemesbyAdul
tLear
ner
sofEngl
i
shasaSecond
Language.Ph.
D.di
sser
tati
on,Uni
ver
sit
yofMi chi
gan.

196
196
196
LARSEN-FREEMAN,D.(
1979)Theimpor
tanceofinputi
nsecondl
anguageacqui
sit
ion.Paperpr
esent
edatt
heLi
ngui
sti
c
Soci
etyofAmeri
ca,
LosAngeles,
December,1979.

LAWLER, JandSELI
NKERL.(
1971)Onpar
adoxes,
rul
es,
andr
esear
chi
nsecondl
anguageacqui
sit
ion.Language
Lear
ning.21:27-
43.

LEE,R.
,McCUNE,L,andPATTON,
L.(
1970)Phy
siol
ogi
cal
responsest
odi
ff
erentmodesoff
eedbacki
npr
onunci
ati
ont
est
ing.
TESOLQuart
erl
y4:117-122.

LENNEBERG,E.(1962)Under
standi
ngl
anguagewi
thoutabi
l
ityt
ospeak:
acaser
epor
t.Jour
nalofAbnor
malandSoci
al
Psychol
ogy65:419-425.

LEVI
N, 1972)Compar
L.( ati
veSt
udi
esi
nFor
eign-
LanguageTeachi
ng.St
ockhol
m:Al
mqv
ist&Wi
ksel
l
.

LI
GHTBOWN,P.,
SPADA,N.andWALLACE, R.(
1980)Someef f
ect
sofinst
ructi
ononchil
dandadolescentESLl
ear
ners.I
n
R.Scar
cel
l
aandS.Kr
ashen( )Resear
Eds. chinSecondLanguageAcqui
sit
ion.Rowl
ey,
Ma: Newbur
yHouse.pp.162-172.

LIGHTBOWN,P.Expl
oringthe r
elat
ionshipsbet
ween development
aland inst
ruct
ionalsequencesi n secondlanguage
acqui
sit
ion.I
nH.Sel
igerandM.Long( Eds.)Cl
assr
oom LanguageAcqui
sit
ionandUse:New Per spect
ives.Rowley
,Ma.:
NewburyHouse.I
npr
ess.

LoCOCO,
V.(
1975)Ananal
ysi
sofSpani
shandGer
manl
ear
ner
'ser
r s.Wor
or kingpaper
sonBi
l
ingual
i
sm 7:96-
124.

LONG, 1980)I
M.( nput
,Int
eract
ion,
andSecondLanguageAcqui
sit
ion.
Ph.
D.di
sser
tat
ion,
UCLA.

LUKMANI
,Y.(
1972)Mot
ivat
iont
olear
nandl
anguagepr
ofi
ci .LanguageLear
ency ning22:261-
273.

MACHA, D.(
1979)Readi
ngcompr
ehensi
onofnon-
nat
ivest
udent
sinEngl
i
shcomposi
ti
onatt
hef
reshmanl
ev .TESOL
el
Quar
ter
ly13:425-
427.

MaCNAMARA,
J.(
1972)Cogni
ti
vebasi
sofl
anguagel
ear
ningi
ninf s.Psy
ant chol
ogi
calRev
iew79:1-
14.

MAKI
NO, 1980)Acqui
T.( sit
ionOr
derofEngl
i
shMor
phemesbyJapaneseAdol
escent
s.Toky
o:Shi
noz
aki
Shor
inPr
ess.

MASON, C.(
1971)Therel
evanceofi
ntensi
vet
rai
ningi
nEngl
i
shasaf
orei
gnl
anguagef
oruni
ver
sit
yst s.Language
udent
Lear
ning21:197-
204.

MI
LON,
J.(
1974)Thedev
elopmentofnegat
ioni
nEngl
i
shbyasecondl
anguagel
ear .TESOLQuar
ner ter
ly8:137-
143.

MINOURA,Y.(1979)Anexami
nati
onoft
herol
eofaccul
tur
ati
oninsecondl
anguageacqui
sit
iont
hroughmul
ti
var
iat
e
anal
ysi
s.Paperpr
esent
edatTESOLsummermeeti
ng,UCLA,Jul
y,1979.

MUELLER, T.(1971)Theef
fect
ivenessoftwolear
ningmodel
s:theaudi
o-l
i
ngualhabi
ttheor
yandthecognit
ivecode-
learning
theory
.InP.PimsleurandT.Quinn( )ThePsy
Eds. chol
ogyofSecondLanguageLear
ning.Cambr
idge:CambridgeUniversit
y
Press,pp.113-
122.

MURAKAMI ,
M.(1980)Behavi
oralandat
ti
tudinal
corr
elatesofpr
ogr
essinESLbynat i
vespeaker
sofJapanese.I
nJ.Ol
l
erand
K.Per
kins( )Resear
Eds. chinLanguageTest
ing.Rowl
ey ,
Ma. :
NewburyHouse.pp.227-
232.

NELSON, 1980)LanguageSy
J.( stemsi
nAdul
tInf
ormalSecondLanguageLear
ner
s.Ph.
D.di
sser
tat
ion,
McGi
l
lUni
ver
sit
y.

NEWMARK, L.(1966)Hownottoint
erf
erewithlanguagelear
ning.LanguageLear
ning:TheI
ndi
vi
dualandt
hePr
ocess.
I
nter
nat
ionalJournalofAmer
icanLi
nguist
ics40:77-83.

197
197
197
NEWMARK, L.(1971)Amini
mallanguaget
eachi
ngprogr
am.InP.PimsleurandT.Qui
nn( )ThePsy
Eds. chol
ogyofSecond
LanguageLear
ning.Cambri
dge:
Cambr i
dgeUniv
ersi
tyPr
ess.pp.11-
18.

NEWMARK,L.
,andREI
BEL,D.(1973)Necessi
tyandsuffi
ci
encyi
nlanguagelearni
ng.I
nM.Lest
er( )Readi
Ed. ngsi
n
Appl
i
edTr
ansfor
mati
onalGrammar .NewYork:Hol
tRinehar
tWinst
on.pp.220-244.

NEWMARK, E.,GLEI
TMAN,H.andGLEI
TMAN, L.(
1977)Mother
,I'
drat
herdoitmy sel
f:someeffect
sandnon-ef
fectsof
mater
nalspeechstyl
e.I
nC.SnowandC.Fer
guson( )Tal
Eds. ki
ngtoChil
dren.Cambridge:Cambri
dgeUni
versi
tyPress.pp.
109-
149.

OLLER, 1979)LanguageTest
J.( satSchool
.London:Longmans.

OLLER,J.,BACA,L.andVIGI
L,A.(1977)Att
it
udesandatt
ainedpr
ofi
ciencyi
nESL:
asoci
oli
ngui
sti
cst
udyofMexi
can-
Americansinthesout .TESOLQuar
hwest t
erl
y11:173-
183.

OLLER, J.andPERKI
NS,K.(1978)I
ntel
li
genceandlanguageprof
ici
encyassourcesofvar
iancei
nselfr
eport
edaffect
ive
var
iables.InJ.Ol
l
erandK.Perki
ns( )Languagei
Eds. nEducati
on:Testi
ngtheTests.Rowl
eyMa.:NewburyHouse.pp.
103-122.

OLLER,J.,
PERKINS,K.andMURAKAMI ,
M.( 1980)Sev
enty
pesofl
ear
nervari
ablesi
nrel
ati
ont
oESLl
ear
ning.I
nJ.Ol
l
erandK.
Per
kins( )Resear
Eds. chi
nLanguageTesti
ng.RowleyMa.
:Newbur
yHouse.pp.233-
240.

OLSSON,
M.( 1969)I
mpli
citandexpl
ici
t:anexper
imentinappl
i
edpsycholi
ngui
sti
cs,assessi
ngdi
ffer
entmethodsofteachi
ng
gr
ammatical
struct
uresi
nEnglishasaforei
gnlanguage.GUMEReport#3,Got
henburgSchoolofEducat
ion.Sweden.

OSTRANDER,
S.andSCHROEDER,
L.(
Eds.
)1976)TheESPPaper
( s.NewYor
k:Bant
am Books.

OYAMA, S.(
1976)Asensi
ti
veper
iodf
ort
heacqui
sit
ionofanon-
nat
ivephonol
ogi
cal
sy em.Jour
st nalofPsy
chol
i
ngui
sti
c
Resear
ch5:261-285.

OYAMA,
S.(
1978)Thesensi
ti
veper
iodandcompr onofspeech.Wor
ehensi kingPaper
sonBi
l
ingual
i
sm 16:1-
17.PATKOWSKI
,

M.(1980)Thesensi
ti
veper
iodf
ort
heacqui
sit
ionofsy
ntaxi anguage.LanguageLear
nasecondl ning30:
449-
472.

PAULSTON,
C.(
1972)St
ruct
ural
pat
ter
ndr
il
ls:
acl
assi
fi
cat
ion.For
eignLanguageAnnal
s4:187-
193.

POSTOVSKY, V.(1974)Ef
fect
sofdel
ayi
nor
alpr
act
iceatt
hebegi
nni
ngofsecondl
anguagel
ear
ning.Moder
nLanguage
Jour
nal58:229-239.

PublicServi
ceCommi ssi
onofCanada,St
affDev
elopmentBranch,St
udi
esDi
vi
sion,
Suggest
opedi
aPr
ogr
am (
1975)A
teachingexper
iencewit
htheSuggest
opedicMethod.Ot
tawa,Canada.

PURCELL,E.
.andSUTER,
R.(
1980)Pr
edi
ctor
sofpr
onunci
ati
onaccur
acy
:ar
e-exami
nat
ion.LanguageLear
ning30:
271-
287.

RACLE,
G.(
1979)Psy
chopédagogi
epr
ofondeenensei
gnementdel
langes.Medi
um 4:
91-
110.

RACLE,
R.(
1980)Psy
chopédagogi
epr
ofondeenensei
gnementdel
langes.(
sui
t .Medi
e) um 5:73-
83.

RAMSEY,
C.andWRI
GHT,
E.(
1974)Ageandsecondl
anguagel
ear
ning.Jour
nalofSoci
alPsy
chol
ogy94:115-
121.RAVEM,
R.

(1974)Thedevel
opmentofwh-quest
ionsinf
ir
standsecondl anguagel
earner
s.I
nJ.Ri
char
ds( )Er
Ed. ror
Analysi
s:Per
spect
ivesonSecondLanguageLear
ning.London:Longman.pp.134-
155.

198
198
198
REBER,
A.(
1976)I
mpli
citlearni
ngofsynt
het
icl
anguages:
ther
oleofi
nst
ruct
ional .Jour
set nalofExper
iment
alPsy
chol
ogy
:
HumanMemoryandLearning.2:88-
95.

RI
VERS, 1968)Teachi
W.( ngFor
eignLanguageSki
l
ls.Chi
cago:
Uni
ver
sit
yofChi
cagoPr
ess.

RI
VERS,
W.(
1979)For
eignl
anguageacqui
sit
ion:
wher
ether
eal
probl e.Appl
emsl
i i
edLi
ngui
sti
cs1:48-
57.

ROBI
NSON, 1980)ESP:
P.( Engl
i
shf
orSpeci
fi
cPur
poses.Oxf
ord:
Per
gamonPr
ess.

SCARCELLA,
R.Dev
elopi
ngconv
ersat
ional
compet
encei
nasecondl
anguage.For
thcomi
ng.

SCARCELLA,R.andHI
GA,C.I
nputandagedi
ff
erencesinsecondlanguageacqui
sit
ion.I
nS.Krashen,R.Scar
cell
aandM.
Long(
Eds.)Chi
ld-
Adul
tDi
ff
erencesi
nSecondLanguageAcquisi
ti
on.Rowley,Ma.
:Newbur yHouse.Fort
hcoming.

SCHACHTER,
J.(
1974)Aner
rori
ner
roranal
y s.Languagel
si ear
ning24:204-
214.

SCHACHTER,
J.,
TYSON,
A.andDI
FFLEY,
F.(
1976)Lear
neri
ntui
ti
onsofgr
ammat
ical
i
ty.LanguageLear
ning26:67-
76.

SCHERE,
G.andWERTHEI
MER, 1964)APsy
M.( chol
i
ngui
sti
cExper
imenti
nFor
eignLanguageTeachi
ng.NewYor
k:
McGrawHil
l
.

SCHLUE,K.(
1977)Aninsi
dev
iewofint
erl
anguage.I
nC.Henni
ng( )Pr
Ed. oceedi
ngsoft
heLosAngel
esSecondLanguage
Resear
chForum.UCLATESLDepar
tment.pp.342-
348.

SCHUMANN, 1978a)ThePi
J.( dgi
nizat
ionPr
ocess.Rowl
ey,
Ma:
Newbur
yHouse.

SCHUMANN, J.(
1978b)Theaccult
urat
ionmodelforsecond-l
anguageacqui
sit
ion,
InR.Gingras( )Second-
Ed. Language
Acqui
sit
ionandFor
eignLanguageTeachi
ng.Ar
li
ngton,Vir
gini
a:Centerf
orAppli
edLingui
sti
cs.pp.27-50.

SCHUMANN, J.(
1979)Theacqui
sit
ionofEngl
ishnegat
ionbyspeaker
sofSpanish:arev
iewoft
hel
it
erat
ure.I
nR.Ander
sen
( )TheAcqui
Ed. sit
ionandUseofSpanishandEngl
ishasFir
standSecondLanguages.Washi
ngt
on:
TESOL.

SCHUMANN, J.(
1980)Theacqui
sit
ionofEngl
ishr
elat
iveclausesbysecondlanguagelearner
s.I
nR.Scar
cel
l
aandS.Kr
ashen
( )Resear
Eds. chi
nSecondLanguageAcquisi
ti
on.Rowley,
Ma. :NewburyHouse.pp.118-131.

SCHUMANN, J.andSCHUMANN, F.(

1977)Diar
yofalanguagel
earner
:ani
ntr
ospect
ivest
udyofsecondl
anguagel
ear
ning.I
n
H.D.Brown,C.Yori
oandR.Cr y
mes( Eds.)OnTESOL'77:Teachi
ngandLearni
ngEngli
shasaSecondLanguage,Tr
endsin
Resear
chandPracti
ce.Washi
ngton:TESOL.pp.209-
249.

SCHWABE,T.(
1978)Sur
viv
alEngl
i
shf
orESLst
udent
sinAmer
icaneducat
ional
inst
it
utons.CATESOLOccasi
i onalPaper
s
4:79-
87.

SCOVEL,
T.(
1979)Geor
giLoz
anov
:Suggest
ologyandout
li
nesofSuggest
ol .TESOLQuar
ogy ter
ly13:255-
266.

SELI
GER,H.(1975)Inducti
vemet
hodanddeduct
ivemet
hodi
nlanguaget
eachi
ng:
areexami
nat
ion.I
nter
nat
ionalRev
iewof
Appl
iedLi
nguist
ics13:1-18.

SELI
GER,
H.(
1979)Ont
henat
ureandf
unct
ionofl
anguager
ulesi
nlanguaget ng.TESOLQuar
eachi ter
ly13:359-
369.SELI
GER,

H.
,KRASHEN,S.andLADEFOGED, P.(1975)Mat ur
ati
onal
constr
aint
sint
heacqui
sit
ionofanat
ive-
li
keaccent
i
nsecondl
anguagelear
ning.LanguageSciences36:209-
231.

SELI
NKER,
L.(
1972)I
nter
languageI
nter
nat
ionalRev
iewofAppl
i
edLi
ngui
sti
cs10:209-
231.

SHIPLEY,E.
,SMITH,C.andGLEI
TMAN,
L.(
1969)Ast
udyi
ntheacqui
sit
ionofl
anguage:
freer
esponsest
ocommands.
Language45:322-
342.

199
199
199
STAFFORD,
C.andCOVI
TT,
G.(
1978)Moni
torusei
nadul
tsecondl
anguagepr
oduct
ion.I
TL:Rev
iewofAppl
i
edLi
ngui
sti
cs39-
40:
103-
125.

STAUBLE,
A.(
1978)Thepr
ocessofdecr
eol
i
zat
ion:
amodel
forsecondl
anguagedev
el .LanguageLear
opment ning28:29-
54.

STEVI
CK, 1976)Memor
E.( y,Meani
ng,
andMet
hod.Rowl
ey,
Ma.
:Newbur
yHouse.

STEVI
CK, 1980)Teachi
E.( ngLanguages:AWayandWay
s.Rowl
ey,
Ma.
:Newbur
yHouse.

SWAFFER,
J.andWOODRUFF,
M.(
1978)Languagef
orcompr
ehensi
on:
focusonr ng.Moder
eadi nLanguageJour
nal62:27-
32.

SWAI
N,M.(
1974)Fr
enchi
mmer
sionpr
ogr
amsacr
ossCanada:
resear
chf
i ngs.Canadi
ndi anModer
nLanguageRev
iew31:117-
129.

TERRELL,
T.(
1977)Anat
ural
appr
oacht
osecondl
anguageacqui
sit
ionandl
ear
ning.Moder
nLanguageJour
nal6:325-
337.

TUCKER,
G.R.andSAROFI
M,M.(
1979)I
nvest
igat
ingl
i
ngui
sti
caccept
abi
l
itywi
thEgy
pti
anEFLst s.TESOLQuar
udent ter
ly13:29-
39.

ULUN, J.andKEMPEN,G.(
1976)Therol
eofthefi
rstl
anguageinsecondlanguagereadi
ngcomprehensi
on--
someexper
iment
alev
idence.
Proceedingsoft
heFour
thI
nter
nati
onalCongr
essofAppli
edLingui
sti
cs.Stutt
gart
:Hochschul
Ver
lag.pp.495-
507.

UPSHUR,
J.(
1968)Fourexper
iment
sont
her
elat
ionbet
weenf
orei
gnl
anguaget
eachi
ngandl
ear
ning.LanguageLear
ning18:111-
124.

VALETTE, 1977)Moder
R.( nLanguageTest
ing.NewYor
k:Har
cour
tBr
aceJov
anov
ich.

v
anNAERSSEN,
M.(
1981)Ph.
D.di
sser
tat
ion,
Depar
tmentofLi
ngui
sti
cs,
Uni
ver
sit
yofSout
her
nCal
i
for
nia.

VARVEL,
T.(
1979)TheSi
l
entWay
:panaceaorpi eam?TESOLQuar
pedr ter
ly13:493-
494.

vonELEK,T.andOSKARSSON,M.(1975)Comparati
veMethodExper
iment
sinFor
eignLanguageTeachi
ng.Depar
tmentofEducat
ional
Resear
ch.Mőlndal
(Got
henbur
g)SchoolofEducat
ion.Sweden.

WAGNER-GOUGH,
J.andHATCH,
E.(
1975)Thei
mpor
tanceofi
nputdat
ainsecondl
anguageacqui
sit
ionst es.LanguageLear
udi ning25:
297-
308.

WALBURG,H.,
HASE, K.andPINZURRASHER,
S.(
1978)Engl
i
shacqui
sit
ionasadi
mini
shi
ngf
unct
ionofexper
iencer
athert
hanage.
TESOLQuar
ter
ly12:427-437.

WHITE,
L.(
1977)Er
roranal
ysi
sander
rorcor
rect
ioni
nadul
tlear
ner
sofEngl
i anguage.Wor
shasasecondl kingPaper
sonBi
l
ingual
i
sm 13:
42-
58.

WI
DDOWSON,
H.(
1977)Thesi
gni
fi
canceofsi
mpl
i
ficat
ion.St
udi
esi
nSecondLanguageAcqui
sit
ion1.

WI
GGI
N,B.(
1979)Comment
sont
heTOEFLt .TESOLQuar
est ter
ly13:292-
294.

WI
NN-
BELLOLSEN, 1977)Communi
J.( cat
ionSt
art
ersandOt
herAct
ivi
ti
esf
ort
heESLCl
assr
oom.SanFr
anci
sco:
AlemenyPr
ess.

WODE,H.(
1976)Developmental
sequencesi
nnat
ural
i
sti
cL2acqui
sit
ion.I
nE.Hat
ch( )SecondLanguageAcqui
Ed. sit
ion.Rowl
ey,
Ma.
:
Newbur
yHouse.pp.101-117.

YORIO,C.(
1978)Conf
essi
onsofasecondl
anguagespeaker
/l
ear
ner
.Paperpr
esent
edat12t
hannual
TESOLconv
ent
ion,
Mexi
coCi
ty,
Apr
il
,
1978.

ZOBL,H.(1980a)Devel
opment
alandt
ransf
erer
ror
s;t
hei
rcommonbasesand(
possi
bly
)di
ff
erent
ial
eff
ect
sonsubsequentl
ear
ning.TESOL
Quar
terl
y14:469-479.

ZOBL,H.(
1980b)Contact
-i
nducedlanguagechange,l
ear
ner
-l
anguage,
andt
hepot
ent
ial
sofamodi
fi
edCA.Paperpr
esent
edatt
heLos
Angel
esSecondLanguageAcquisi
ti
onResearchForum,UCLA.

ZOBL,
H.(
1980c)Thef
ormal
anddev
elopment
alsel
ect
ivi
tyofL1i
nfl
uenceonL2acqui
sit
ion.LanguageLear
ning30:43-
57.

200
200
200
I
ndex
Accul
tur
ati
on 45-
50 For
get
ti
ngPr
inci
ple66,
133,
168

Achi
evementt
est
ing176-
182 Fi
rstl
anguagei
nfl
uence27-
28

Acqui
siti
on(defi
ned)10 Grammat icalsequencing
doesnotresul
tfrom l
earni
ng83-89 (foracquisit
ion)14,25-26,
68-
70,
128,
131,
134,
136,
notthesameasi nduct
ivel
ear
ning113 138,141,145, 163,165,168
char
acteri
sti
cs18 (f
orlearning)115-116

Aff
ecti
veFil
ter30-
32,73-
76,128,131,
134,
136,
139, Grammart
eachi
ng81,
88,
89,
112,
113-
115,
119-
121,
141-
142,144-145,
164,165,
168-169 137

Age43-
45,
95 Gr
ammar
-t
ransl
ati
on127-
129,
149-
154

Audi
o-l
i
ngual
i
sm 129-
132,
149-
154 GUMEpr
oject150-
154

Av
oidance51 I
mmer
sion170-
171

Backwashef
fect177-
178,
182 I
nduct
iveLear
ning11,
87,
113-
115,
132,
135,
142

"
Car
eless"er
ror
s86,
99 I
nputhy
pot
hesi
s20-
30,
60

Car
etakerspeech22-
24,
64-
65 I
nter
languaget
alk24

Cl
assr
oom,
rol
eof30,
33-
37,
58-
60 I
nter
nal
i
zat
ion81,
84

Cogni
ti
ve-
code84,
113,
132-
134,
149-
154 Lear
ning(def
ined)10,
16,
83-
124,
129,
132,
134,
137,
139,142,
187
Conv
ersat
ion61,
163-
164
Languagel
abor
ator
y184-
185
Conver
sat
ionalmanagement44,
59, 61,76-
79,
129,
131,
134,135-136,
139,142,
164,165,169 Mat
eri
als182-
186

Counsel
i
ng-
Lear
ning32,
126 Met
hodcompar
isons147-
160

Deduct
ivel
ear
ning87,
113-
115 Moni
torhy
pot
hesi
s15-
20

Di
rectmet
hod135-
137 Moni
toruse89-
104

Err
orcorr
ect
ion11,
61,
74-
76,
91-
92,
116-
119,
138, Nat
ural
Appr
oach76,
117,
137-
140
139,141
Nat
ural
order12,
13,
14,
17,
101,
116,
124,
185
"
Eur
eka"ex
per
ience88
Out
put60-
62,
74
Ex
posur
e37-
43
Pl
easur
ereadi
ng164-
167,
183-
184
For
eignert
alk24,
60,
64-
65,
163

201
201
201
Quant
it
yofi nput71,128,131,134, 136,
138-
139,141,163,
165,168,
184 Teacher
,funct
ionof64-
66

Sel
f-
cor
rect
ion104-
112 Teachert
alk24,
64-
65,
137

Si
l
entper
iod26-
27,
71-
72,
81,
157,
171 Tot
alPhy
sical
Response140-
142,
155-
157

Si
l
entWay126 Tr
ansi
ti
onal
for
ms14,
51-
52

Subj
ectmat
tert
eachi
ng167-
175 Vocabul
ary80-
81

Suggest
opedi
a32,
142-
146,
158-
160

202
202
202