Teachers of English to Speakers of Other Languages, Inc.

(TESOL)

There Is No Best Method-Why?

Author(s): N. S. Prabhu
Source: TESOL Quarterly, Vol. 24, No. 2 (Summer, 1990), pp. 161-176
Published by: Teachers of English to Speakers of Other Languages, Inc. (TESOL)
Stable URL: http://www.jstor.org/stable/3586897 .
Accessed: 21/08/2013 09:44

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Teachers of English to Speakers of Other Languages, Inc. (TESOL) is collaborating with JSTOR to digitize,
preserve and extend access to TESOL Quarterly.

http://www.jstor.org

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
 TESOL QUARTERLY,Vol.24, No.2, Summer1990

ThereIs No BestMethod-Why
?
N. S. PRABHU
NationalUniversity
ofSingapore

Thispaperexaminesthepossiblesubstanceof a statement, often

heardbutrarelyexplained,inthediscussion oflanguageteaching,
namely,thatthereis no bestmethod.Threepossibleexplanations
ofthestatement arediscussed,based on a broadinterpretation
of
thetermmethod:(a) thatdifferent methodsare bestfordifferent
teachingcontexts;(b) thatall methodsare partially
trueorvalid;
and (c) that the notionof good and bad methodsis itself
misguided.The first twoexplanations areshownto be unhelpful;
the thirdis discussedin some detail,withan explorationof the
concept,teachers'senseofplausibility.

It is uncommonthese days to have a sustaineddiscussionon lan-

guage teachingwithoutsomeone at some point declaringthatthere
is no best method,or words to thateffect.Such a declarationusually
occurs at a late stage in an indecisive debate about different
methods,and has thegeneraleffectof alteringthe orientationof the
debate itself,somewhat abruptly. It also carries the ring of an
incontrovertible statement-or a statement so tolerant and
reconciliatoryin spiritthat to dispute it would be professionally
churlish.As a result,one rarelysees a detailed examinationof what
it mightmean to say thatthereis no best method.
I thinkit helpfulto see thestatementas havingan illocutionaryas
well as a propositionalmeaning.As an illocutionaryact, it seeks to
terminatea debate withoutreaching any substantiveconclusion,
and it does so not by admittingdefeat in the effortto reach a
conclusion,but by appearing to raise the debate itselfto a higher
level, thushelpingto save professionalface all round.It suggestsnot
justthatproponentsof different methodsagree to disagree,but that
theygive up theirpursuitof agreementand disagreementas being
unproductive.It manages to reconcileconflictingviews by defining
a positionto which all can honourablysubscribe,and by strikinga
philosophical note on which enlighteneddiscussion can properly
end. It is thusa convenientdevice forlegitimizingnonresolutionof
methodologicalissues.

161

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
 The propositional contentof thestatement, on theotherhand,
callsfora greatdeal of clarification in termsof
and substantiation,
thesensein whicha methodcan be consideredgood, thereason
whyno methodcan be thought or shownto be betterthanothers,
the consequenceof thatpositionforthe theoryand practiceof
pedagogy,therevaluation it demandsof theassumptions currently
underlying conflictand debate betweenmethods,and theformsin
which debate or discussionmay stillbe of use fromthe new
perspective.What is involvedis not just the termination of a
discussion, butthe beginning ofa new one-or ratherthebeginning
ofa newphaseinthediscussion, as a positiveoutcomeoftheearlier
phase. It is thispropositional contentof the statement thatis the
concernof thispaper;I hope to clarify thesenseinwhichtherecan
be said to be no bestmethodoflanguageteaching.
First,however,a word about thetermmethod.I use the term
inclusively,toreferbothto a setofactivities to be carriedoutinthe
classroomand to the theory,belief,or plausible concept that
informsthoseactivities.The readerwill thusfindone or both of
these aspects the focus of particularparts of the discussion.I
considerthis"global" interpretation of the termappropriateto
analysing a statement thatis equallyglobalin spirit.
There are, I think,threegenerallinesof argumentthatcan be
advanced in supportof the statement to be discussed.These are
examinedin turn.

IT ALL DEPENDSON THE TEACHINGCONTEXT

Ifthosewhodeclarethatthereisnobestmethod areaskedwhy,
themostimmediate andfrequent answeris likelytobe "Becauseit
all depends,"meaningthatwhatis bestdependson whomthe
method is for,inwhatcircumstances,forwhatpurpose, andso on.
Thatthereisnobestmethod thereforemeansthatnosinglemethod
isbestforeveryone,as there
areimportantvariationsintheteaching
contextthatinfluence whatis best.The variationsare of several
kinds, relatingto social situation(language policy, language
environment, economicand
linguisticand culturalattitudes,
etc.), educational
ideologicalfactors, (instructional
organisation
objectives, constraintsof time and resources,administrative
classroom
class-size,
efficiency, factors
ethos,etc.),teacher-related
(status, belief,autonomy,
training, skill,etc.),and learner-related
factors learning
previous
(age,aspirations, experience, to
attitudes
tocategorizesuch
etc.).Therehavebeenseveralattempts
learning,
(e.g.,see Brumfit,
and comprehensively
variablessystematically
1984),buteventhebriefand randomlistingabove showsthatthey

162 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
are at different levelsof generality, as well as of discreteness and
tangibility.Moreover, recent and current work in the field seems to
be addingnew factorsand categoriesto theinventory, in theform
of varied learningstyles,communication strategies,personality
factors,and psychological processes.Togetherwithslightly earlier
workon variationsin learners'needs and purposes,thisseemsto
suggestthatvariability on suchdimensions is infinite, thusstrongly
challenging the notion that any given method can be good for
everyone. "It all depends," and what itdepends on is a vastnumber
ofthings.
Notice,however,thatto say thatno singlemethodis best for
everyoneis also to saythatdifferent methodsare bestfordifferent
people-or for different teaching contexts. This impliesthat,for
anysingleteachingcontext, there is in fact a methodthatisbestand,
further, we are able to determine what it is. If we are unclearor in
disagreement about what methodis best fora specificcontext,
there is need for discussion,debate, and interactionbetween
differing perceptions; we shouldbe seekingto further pedagogic
debateratherthanto terminate it witha face-saving formula.It is
perhapsnotwithoutsignificance thatstatements like"Thereis no
bestmethod"are made mostoftenas defensivepostures-asways
of sayingnot onlythatone does not agree withwhat is being
argued,butthatone refusesto engageanyfurther in theargument.
Suchstatements succeedinpreserving theconversational peace,but
cause a loss of theproductivepotentialof professional debate.For
suchstatements to act as a contribution to debate,we willneed to
interpret themas proposingonlya narrowing of thescope to some
singleteachingcontext,so thatdependenciesbetweencontextual
variablesand methodologicaloptionsmightbe explored,and a
searchforthebestmethodmightbe continuedwithgreaterfocus.
The statementthatthereis no best methodwould thenbe an
assertion, notof thefutility of lookingforthebestmethod,but of
the desirability of askingwhat methodis best forsome specific
context;contextualvariabilitywould serve not as a means of
avoidingmethodological issues,butas a possiblenew approachto
resolving them.
I thinkit is importantto realisehow complex it can be to
determinedependenciesbetweencontextualfactorsand instruc-
tional methods.To begin with,many of the factorsthat are
discussedareneither easytoidentify norsimpletoassess:Theytend
suspiciouslyto fitdifferent slotsin different taxonomies.Formal
environment, for instance,may referto classroomlearning,as
againstlearningthroughsocial exposure,or to the formalschool
system, as against private language instruction,or to relative

THERE IS NO BEST METHOD-WHY? 163

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
formalityin teacher-learnerrelations,as against informalrelations,
or even to teacher-fronted activities,as againstgroup work among
learners. Motivation may mean anything from future career
ambitions,to a desire forgroup approval, to a passing interestin a
particularclassroom activity.What are identifiedas attitudesare
seldom clear-cut or unmixed or even stable. When we come to
factorslike preferredlearning styles,sociocultural influences,or
personalityfactors,we are faced with unclear and overlapping
distinctions,and are thereforeforced to simplifyand stereotype,
oftenin preconceived ways. Further,even when some contextual
factorsare clearlyidentifiable,theirconsequence for instructional
procedurescan be farfromclear. Do older learnersneed a different
method of teachingfromyoungerones, and if so, how fundamen-
tallydifferent?Is a change in methodmore likelyto succeed in the
hands of experienced teachers, or less? If there is a mismatch
between officiallanguage policy and learners'personalgoals,which
should have what weightin the choice of an instructionalmethod?
If we identifycertainlearningstrategiesthatlearnersnaturallytend
to employ, do we conclude, for that reason, that they are good
learning strategies?If not, are learners likely to learn more by
followingtheirown strategies,which may be less thangood in our
view, or by adopting strategieswe consider more conducive to
learning,though they may go against such natural tendencies? If
earlier experience has conditioned learnersor teachers to certain
perceptions of learning and teaching, does that constitutean
argumentagainstchange, or indicatea greaterneed forchange?
The pointI am makingis not just thatour knowledge is uncertain
at thistime; the more importantpointis thatit is onlywhen we can
show a relationshipbetween a contextualfactorand a methodolog-
ical decision that the contextual factor becomes significantfor
pedagogy. What we need is notjust an identificationand projection
of variationbut, equally, some way of determiningwhich formof
variationmattersto instructionand how, and which does not. If we
look for variation merely on the assumption that the teaching
context matters for teaching methodology,we are sure to find
indefinitevariationon many dimensions,thusmakingit impossible
to justifyany instructionalmethod forany singlegroup of learners.
If all physiologicalvariationamong individuals (includingfinger-
prints)were assumed to call formatchingdifferentiation in medical
treatment,no medical practice would be justifiable.
It is, of course, possible to obviate the problem of relating
contextual factors to instructionalmethods by giving contextual
factors a central role in pedagogy and treating instructional
methodsas a kind of logical derivationfromthem.This is the move

164 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
oftenadvocated: froma preoccupationwithteachingmethodsto an
effortat "curriculumdevelopment"or "course design."The assump-
tion is that,with such a move, decisions concerningmethods will
eitherbe renderedunnecessaryor play onlya smallpartin theenter-
prise. As one applied linguistputs it,"The importantissuesare not
which methodto adopt but how to develop proceduresand instruc-
tional activitiesthatwill enable programobjectives to be attained"
(Richards,1985,p. 42). The proceduresenvisagedare thoseof situa-
tional analysis,needs analysis,analysis of "authentic"samples of
targetlanguage use, surveysof opinionsand attitudes,estimatesof
resources,etc.-that is to say, compilationsof differentkindsof in-
formationabout learners,teachers,theschool,and thesociety,witha
view to determininginstructionalobjectives. Instructionalproce-
duresare seen to follow,or be easilydeterminable,fromtheprofile
of contextualfactorsand the statementof objectives. This is, in
effect,a kind of discoveryprocedureformethods:That methodis
best,it seems to say, whichresultsfroma carefulimplementation of
theprocedure,thesoundnessof themethodbeing guaranteedby the
soundnessof theprocedureleadingto it.
There is, however, a price to pay for this simplificationof
pedagogy. The instructionalprocedures most directlyderivable
from a specificationof needs, wants,and objectives are those of
supplying to learnersthe relevant tokens of language, or getting
them to rehearse target language behaviour in simulated target
situations.Any concept of developing in learners a more basic
capacity for generatingtokens of language when needed, or for
adapting to unforeseentarget language behaviour as necessary,
leads one toward ideas about the natureof language abilityand the
process of language acquisition-complex methodological issues
that the discovery procedure seeks to avoid. Besides, a more
elaborate analysisof contextualfactorsresultsin a correspondingly
largerset of criteriato be metby instructional
contentor procedure,
and the larger the set of criteriato be met, the fewer the choices
available in meeting them. Language instructionthat attemptsto
cater directly to social objectives, learning needs, target needs,
learners' wants, teachers' preferences,learning styles, teaching
constraints,and attitudesall roundcan end up as a mereassemblage
of hard-foundpieces of content and procedure-a formulathat
manages, with difficulty, to satisfymultiplecriteriaand therefore
cannot affordto let itselfbe tamperedwith.What sets out to be a
wideningof attentionto varied educationalconsiderationscan thus
end up as an abridgementof choice and flexibility in thepracticeof
pedagogy. In avoidingadherence to a singlemethod,one will have
arrivedinsteadat a single,fixed,teachingpackage.

THERE IS NO BEST METHOD-WHY? 165

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
 Avoiding adherence to a single methodhas a certainideological
aura to it. It suggestsliberationfroma monolithicmould, a refusal
to be doctrinaire,an espousal of plurality.It is, however, also a
denial of the role of understandingin language pedagogy, which is
necessarilya matterof ideation:We understandsomethingwhen we
have a set of ideas or principlesthatcohereto make up a conceptual
model, or theory.Althoughtheoryis only an abstraction,it is about
the only instrumentwe have (at any rate, the most accessible
instrument) for making sense of complex phenomena and
conveyingthatsense to one another.Theory,as we know,arisesnot
froma cataloguing of diversity,but froma perceptionof unityin
diverse phenomena-a single principle, or a single system of
principles,in termsof which diversitycan be maximallyaccounted
for.If thetheoriesof language teaching(thatis to say,methods)that
we have at presentfail to account sufficiently for the diversityin
teaching contexts,we ought to tryto develop a more general or
comprehensive(and probably more abstract)theoryto account for
more of the diversity,not reject the notion of a single systemof
ideas and seek to be guided insteadby diversityitself.Pointingto a
bewilderingvarietyof contextualfactorsas a means of denyingthe
possibilityof a singletheorycan onlybe a contributionto bewilder-
ment,not to understanding.
There is also the fact that a concentrationon dissimilarities
between teachingcontextsis likelyto obscure similaritiesbetween
them;in the case of language learning,we are dealing witha human
abilitythatconstitutesa definingcharacteristicof the species. It is
true that variations in social situation,institutionalorganisation,
individuals'histories,attitudes,or intentionscan all have the effect
of limiting or extending opportunities for desired forms of
pedagogic action. However, to implythattheycall fora matching
differentiation in pedagogic theoriesis to make theverylarge claim
thatthe process of language acquisition-a basic humanattribute-
itselfvaries accordingto contextualfactors.

THERE IS SOME TRUTH TO EVERY METHOD

I have discussed in some detail one formof substantiationof the
statementthat there is no best method, namely, that different
methods are best fordifferentteachingcontexts.A differentform
namely,thatthereis
of substantiationis also heard fairlyfrequently,
some truth(or value or validity)to every method-or, at any rate,
to several differentmethods-even though the methods may be
conceptually incompatible. This, as we know, is an argumentfor
eclecticismin language pedagogy-not an argumentthatdifferent

166 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
contextsshould use differentmethods,but an argumentthat the
same contextshould use a numberof differentmethods,or perhaps
parts of differentmethods. There is an immediate appeal to
common sense in thisstance: If everymethodis a partialtruth,then
it seems clear thatnone representsthe whole truth;to adopt any
singlemethodis to settleformuchless thanone can getby adopting
all or several of them.
As a commenton our state of knowledge at thistime (or indeed
at any time), the suggestionthatno method containsmore than a
partial truthis clearlyunexceptionable.We continueto engage in
the professional activity of research, concept development,
discussion,and debate because all of ourunderstandingof language
learningand teachingis at best partial,and forany of us operating
with a theory,which can representonly a partial truth,it remains
entirelypossible thatothertheoriesrepresentpartialtruthsas well.
However, this philosophical perception of imperfectknowledge
does not help us to see which theoryrepresentswhich part of the
truth-or whichpartof a giventheoryrepresentsthetruth.Were we
able to see that,we would no longerbe operatingwitha theoryof
which onlya partis known to representthe truth,and missingthose
partsof othertheoriesthatare knownto representotherpartsof the
truth.Our knowledge would make a leap toward the whole truth-
whatever that might mean. The fact, however, is that the
understandingone has of a phenomenonat a giventime-the theory
one is operating with-represents for one the whole truth.Other
theories are true to the extent they share the understanding
representedby one's own theory.In this sense, each theorycan
claim to representthe partialtruthsof othertheories:If thereis an
overlap of understandingbetween it and certain parts of other
theories,thenit can be said to containthose parts of othertheories
thatconstitutepartialtruths.
But the statementthatthereis some truthto everymethodneeds
to be seen not just as an epistemologicalobservation,but as a plea
for an eclectic blending of all or several methods. Now, any such
blending of differentmethods is eitherdone with a perceptionof
what is true about each method, or it is done withoutany such
discrimination.If there is a perception of which part of what
method is a partialtruth,thenthatperceptionconstitutesa theory,
which happens to have an overlap of understandingwith various
othertheories.It thereforerepresentsa method,which is like any
othermethod,withan overlap of understandingwithothers.There
is no reason to think,on the strengthof itsbeing a blend, thatit has
any more of the truththanany othermethod.It is simplyone of the
methodsthatsharesome of theirconceptsor procedureswithother

THERE IS NO BEST METHOD-WHY? 167

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
methods;how muchofthetruth itrepresentsis a matternotofhow
muchblendingitdoes,butofwhatparticular perception makesthe
blendingpossible. What we have,therefore,is notan eclecticblend,
buta different method-orelse,all methodswhichhappento have
partialoverlapswithothersare equallyeclectic.
If,on theotherhand,an eclecticblendingofdifferent methodsis
done not withany particularperceptionof what partsof those
methodsrepresent thetruth, butratherinthehope thatwhateveris
trueaboutthemwillbe capturedin theblending,thentheeclectic
blenddoes notconstitute a method,butinsteadan act of gambling
or a hedgingof bets:It can onlyhavetreatedall partsof different
methodsas beingequallylikelytobe trueoruntrue, and,as a result,
thereis as stronga possibilityof itsbeinga blendof theuntruthful
partsof differentmethodsas thereis ofitsrepresenting thetruthful
parts (whatevernotionof truthand untruthwe may care to
employ).Further,such indiscriminate blendingof methodsadds
nothingto our pedagogic understanding,since it offersno
perceptionof whatmay be trueabout whichmethod.It simply
plays it safe-as safe fromtruthas fromuntruth.An eclectic
blendingthatconstitutes a formofpedagogicunderstanding atleast
offersus an additionalmethod,thoughitmakesan unjustified claim
to beingmorethanan additionalmethod;an eclecticblendingthat
does notconstitute an additionalmethodinthatsenseleadsus away
fromanyfurtherance of understanding, whileoffering us a chance
at whatmaybe called"truth by accident."

WE NEED TO RETHINKWHAT"BEST"MIGHTMEAN
Let me now turnto a thirdpossibleway of substantiating the
statement thatthereis no best method.This is thatwe have no
adequate notionof what"best"mightmean-or thatthenotionof
good and bad needsto be reexamined and clarified.
A prevalentnotionofthebestmethodis thatitis themethodthat
yieldsthebestresultsin termsof learningoutcomes.Sincetheaim
of all teachingis to bringabout as muchlearningas possibleas
quicklyas possible,it seems self-evident thatteachingmethods
shouldbe judgedby theamountsoflearningtheycan lead to,in a
given period of time.This appears to call for a comparisonof
methodsand a quantification of learningoutcomes,throughwell-
designed, controlled experiments, in keepingwiththe spiritof
objective,scientific
enquiry.It is truethatsuchobjectiveevaluation
is so difficult
to implement thatall attemptsat it in thepast have
resultedin a wider agreementon the difficulties of doing an
evaluation than on the resultingjudgementon methods. It is also

168 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
truethatargumentshave been put forwardfora possible alternative
to the experimentaldesign. Nevertheless,we generallycontinueto
assume, more or less consciously,that there is a method that is
objectively the best, thatit is in principlepossible to demonstrate
that fact, and thatonce demonstrated,the superiorityof the best
method will lead to its widespread acceptance in the profession.
That is to say,we generallysee ourselvesas workingto thatideal, on
the tacit premise that what is unrealised is not necessarily
unrealisable,and thatall our professionalendeavour is a formof
progresstoward it. Alternativessuch as tryingto constructcompre-
hensive descriptions of methods-as "illuminative" evaluation
(Parlett& Hamilton,1977)-involve an abandonmentof thatideal,
thusthreateningto disorientour professionalthought.We preferto
retainthe ideal as the basis of all our professionaleffort.
Seen in thiscontext,the statementthatthereis no best methodis
a questioning of the current concept of the best method-an
argumentthatthe ideal of objective and conclusive demonstration
is notonlyan unrealisedone, but an inherently unrealisableone, ard
thatworkingwithsuch an ideal is unproductiveforthe pedagogic
profession.Brumfit(1984), forinstance,has stronglyquestionedthe
notionthatteachingmethods,whichare essentiallyconcernedwith
human interaction,can usefullybe subjected to the processes of
objective testingand prediction,which are part of the scientific
method. He argues, in summary,(a) that a teaching method in
operation is necessarily an embodiment of certain general
pedagogic principlesinto a varietyof specific contextualfeatures
(including participants'psychological states); (b) that predictive
testing of a method demands manipulation and control of the
manifold contextual features; (c) that many of the contextual
features are either difficultor impossible to control; and, most
important,(d) that any success actually achieved in controlling
contextualfeatureswill have only the effectof disembodyingthe
method, as it were, of its actual, operational form,thusrendering
the outcome of the testinginapplicable to the operation of the
method in any specificcontext.Brumfitcomments:
A claimthatwe can predictcloselywhatwillhappenin a situation as
complexas [theclassroom]can onlybe based on eithertheview that
humanbeingsare moremechanicalin theirlearning responsesthanany
recentdiscussionwouldallow,or thenotionthatwe can measureand
predictthequantitiesand qualitiesof all thesefactors.Neitherof these
seemsto be a sensiblepointofviewto take.(pp. 18-19)
While Brumfit'sargumentis based largelyon the complexityof
the pedagogic operation, it is also possible to point to the

THERE IS NO BEST METHOD-WHY? 169

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
complexityof assessinglanguageattainment as such. Examining
aspects of thislatter complexity further underscores thefutility of
to
attempts objectify method evaluation. To begin with, an
importantconsiderationfor language teachingmethodsis the
qualityof learningto be promoted,as distinctfromthequantity.
The questionof qualityhas been a recurrentconcernfor the
professionthroughthe ages, beingconceptualisedand verbalised
variouslyas grammar incontrast to practice,knowledgein contrast
to skill,explicitknowledgein contrastto implicitknowledge,
accuracyin contrastto fluency, learningin contrastto acquisition,
to
ability display in contrast to abilityto deploy,etc.Theremaybe
disagreements about how the different kinds of knowledgeor
abilityarerelated to each other, but itis remarkable how,whenever
a distinctionis made betweendifferent formsof knowledgeof a
language, it is the less conscious, less observable, and less
quantifiableformthatis seen to be of greatervalue or validity.
Objectiveevaluationof methods,however,necessarily relieson a
quantificationoflearning outcomes,and therefore tendsto measure
themorequantifiable formofknowledge.Thismeansthatthemore
objectivetheevaluationis,thelesslikelyitis toassesslearning ofthe
desiredquality,and vice versa.
Second,a perceptionof languageabilityas an implicitformof
knowledgeis linked to a perceptionof its developmentas an
internal,unobservableprocessthatis organicratherthanadditive,
and continuous ratherthanitemisable.Thismeansthatat anystage
of thegrowthprocess,thereis notonlythegrowthachievedso far,
but a potentialfor furthergrowthachieved as a part of it-a
potential that can be thoughtof in terms of inchoationor
incubation.Our mostambitiouseffort at languagetestingcan only
hope to giveus evidenceon theactualgrowthachievedat thestage
of testing,noton thepotentialgeneratedforfurther growth, since
knowledgeinan inchoative stateis evenlessaccessibletoelicitation
and quantification than implicitknowledgeas such. Again,not
everyonein theprofession mayregardknowledgeof a languageas
beingequallyimplicitin natureor organicin itsdevelopment, but
thepointis thatan objectiveevaluationofmethodsis unlikely to be
able to cope withconceptsof implicitness and inchoation and,as a
result,unlikelyto providewidelyacceptabledecisions.
Third,an objectiveevaluationof methodsis notjustan assess-
mentoflearners'languageattainments; italso involvesan objective
attributionof thelearningthathas takenplace to theteachingthat
has been done. However, the relationshipbetween language
teachingand languagelearningbecomesless and lessdirectas one
perceives language as being an implicit ability and an internal

170 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
development:What is less conscious and less observable is also less
directlyteachable. The more indirectthe relationshipis between
teaching and learning, the more difficultit is to attributeany
specific piece of learningto any specific piece of teaching. It is,
ultimately,difficultto tellwhat learninghas takenplace as intended
by the teaching,and what has taken place independentof it (or,
indeed, what in spite of it). We have problemsenoughmaintaining
a subjective perceptionof general causationbetween teachingand
learningin the developmentand discussionof particularmethods.
An evaluativecomparisonof different methodscalls fora degree of
objectivity and in
specificity cause-effect
relationsthatmay well be
unreasonableto expect in the fieldof language pedagogy.
More generally,the notion behind an objective evaluation of
methods is that thereis somethingin a method that is by itself-
independent of anyone's subjective perception of it-superior or
inferiorto what thereis in anothermethod. If some method were
shown by such evaluation to be superiorto all others,then that
method would be expected to benefitall (or a large number of)
classrooms,regardlessof how it is subjectivelyperceived by the
differentteachers involved. A method, in this view, is a set of
procedures thatcarriesa predictionof results;the fulfilment of the
predictiondepends only (or mainly) on an accurate replicationof
the procedures, not on any perceptions of those who do the
replication--ratherin the way the replicationof a procedure in
chemistryyields the predicted result,regardless of the chemist's
thoughtsor feelingsabout it. No doubt the idea looks fairlyabsurd
when put in thisform:It reduces teachingto a faithfulfollowingof
highlyspecified routine--somethingof a pedagogic ritual.I am,
however, unable to see how a seriouspursuitof objective method
evaluation can be sustained without some such idea. The only
alternativeto it is to maintainthatthe method thatis shown to be
objectively superior will somehow carry with it the subjective
perception thatlay behind its development and, equally, that the
perceptionconcernedwill thenreplace the differingperceptionsof
all the teachers who may be led to adopt that method on the
strengthof the objective evaluation. This implies, among other
things,that teachers'pedagogic perceptionsare as easily replace-
able as classroom procedures, an idea that could hardly be less
absurd.
It is useful to ask why it looks absurd to suggest that a good
teaching method can be carried out, without loss, as merely a
routine.We finditnecessaryto thinkof good teachingas an activity
in which thereis a sense of involvementby the teacher.When we
encounteran instanceof reallybad teaching,it is most oftennot a

THERE IS NO BEST METHOD-WHY? 171

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
case of the teacherfollowinga methodwithwhichwe disagree,but
ratherof the teachermerelygoing throughthe motionsof teaching,
with no sense of involvement.When a method considered to be
good has been implementedon a large scale and laterthoughtnot
to have "worked," an importantpart of the reason identifiedhas
been that teachers followed the method "mechanically,"with no
sense of understanding or identification. Indeed, the more
"efficiently"a method is implemented (that is to say, with all
possible measures to ensure that teachers will carry out the
procedures envisaged), the more likely it is that mechanical
teachingwill turnout to be the main impedimentto success.
Perhaps, then, there is a factor more basic than the choice
between methods, namely, teachers' subjective understandingof
the teachingtheydo. Teachers need to operate withsome personal
conceptualisationof how theirteachingleads to desired learning-
witha notionof causationthathas a measureof credibilityforthem.
The conceptualisation may arise from a number of different
sources, including a teacher's experience in the past as a learner
(with interpretations of how the teachingreceived at thattime did
or did not supportone's learning),a teacher'searlierexperience of
teaching (with similar interpretationsfrom the teaching end),
exposure to one or more methodswhile trainingas a teacher (with
some subjective evaluationof the methodsconcerned and perhaps
a degree of identificationwith one or another of them), what a
teacher knows or thinksof otherteachers'actions or opinions,and
perhaps a teacher's experience as a parent or caretaker.Different
sources may influencedifferentteachersto differentextents,and
what looks like the same experience or exposure may influence
differentteachersdifferently.
The resultingconcept (or theory,or, in a more dormantstate,
pedagogic intuition)of how learningtakes place and how teaching
causes or supportsitis what may be called a teacher'ssense ofplau-
sibilityabout teaching.This personal sense of plausibilitymay not
only vary in its contentfromone teacher to another,but may be
more or less firmlyor fully formed, more or less consciously
considered or articulated,between differentteachers.It is when a
teacher's sense of plausibilityis engaged in the teachingoperation
thatthe teachercan be said to be involved,and the teachingnot to
be mechancial. Further,when the sense of plausibilityis engaged,
the activityof teachingis productive:There is thena basis for the
teacher to be satisfiedor dissatisfiedabout the activity,and each
instance of such satisfactionor dissatisfactionis itselfa further
influenceon the sense of plausibility,confirmingor disconfirming
or revisingit in some small measure,and generallycontributingto

172 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
its growthor change.I also thinkthatthe greaterthe teacher's
involvement in teachingin thissense,themorelikelyitis thatthe
sense of involvement will conveyitselfto learners,gettingthem
involvedas well and helpingto create that elusivebut highly
regardedcondition intheclassroom:teacher-learner rapport.It is of
coursepossiblethatotherfactorshavea role,too,in thecreationof
rapport,such as learners'own perceptionsof learningand their
interpretations oftheteachingactivity (Allwright,1984).Mypointis
thatan engagement of theteacher'ssenseof plausibility is a major
conditionfor classroomrapport,whetheror not it is the only
condition.It is commonto hear thatlearningis enhancedwhen
learnersenjoyclassroomactivity, butenjoyment is a broadnotion
and is often equated with some formof light entertainment
interspersed withmoreseriousactivity.I thinkthereis a formof
enjoymentarising from teacher-learnerrapport that is less
conspicuousbutmoreintegral to classroomactivity, and moretruly
productiveoflearning.
The pictureofclassroomactivity thatengagestheteacher'ssense
of plausibilityis no doubt closer to an ideal than to a factual
description of muchof theteachingthatactuallygoes on. Butthat
does notdetractfromthesuggestion I am making,namely,thatthat
ideal is more worthour while to pursuethanthe notionof an
objectively bestmethod.The questiontoaskabouta teacher'ssense
of plausibilityis notwhetheritimpliesa good or bad methodbut,
morebasically,whetheritis active,alive,or operational enoughto
createa senseof involvement forboththeteacherand thestudent.
Mechanicalteachingresultsfroman overroutinisation of teaching
activity,and teachingis subjectto greatpressures ofroutinisation. It
is,afterall,a recurrent patternofprocedureson regularly recurrent
occasions.It is also a formof recurrent social encounter between
teachersand learners,withself-images to protect,personalities to
cope with,etc. And,like all recurrent social encounters, teaching
requiresa certaindegreeof routineto makeit sustainableor even
endurable.Thereare,in addition,variedfeelingsof adequacy or
confidence among teachers, varied degrees of concern for
maintaining status,stressof overwork, threatof peercomparisons
or of expectationsfromsuperiors,etc., all of whichcan use the
protection offeredby role-defining routines.Morefundamentally,
thereis thefactthatlearning by learners is essentiallyunpredictable
and, consequently, it is an unacceptablyhighriskforteachingto
have to justify itselfin directrelationto learning: Thereis need for
a way to claim thatthe teachingexpectedhas been performed,
thoughthe learningexpectedmay stillnot have occurred;and
teachingdefinedas a routinepreciselymeetsthatneed.
THERE IS NO BEST METHOD-WHY? 173

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
 An activesenseofplausibility is verydifficult to maintain among
such pressureson routinisation, and can easily become frozen,
ossified,or inaccessiblysubmerged,leaving only a scheduleof
routines.Whenteachersprofessto believe in some methodthey
have been following-perhaps to thepointof swearingby it-they
may well be merelydemonstrating how frozentheirsenseof plau-
sibilityis and, as a result,how insecure theyfeelagainsta threatto
theirteachingroutines.When a teacher'ssense of plausibility is
activeand engagedin theteaching, itis necessarily to
open change,
howeverslowlyor imperceptibly, in the processof the ongoing
activityof teaching.Such teachingcan perhapsbe regardedas
being"real,"in contrast to teachingthatis mechanical.We can then
say that a distinction between "real"and mechanicalteachingis
moresignificant forpedagogythananydistinction betweengood
and bad methods.The enemyofgoodteachingisnota bad method,
butoverroutinisation.
If it is important fora teacher'ssenseof plausibility to remain
alive and therefore to
open change-not frozen but fluidin some
an
degree-then important goal for the pedagogicprofession is to
seek waysin whichthesenseof plausibility in as manyteachersas
possible can be helped to remainas alive as possible,though
necessarilyin variedforms.It is truethattheongoingactivityof
teachingis itselfa sourceofcontinual influence on a teacher'ssense
ofplausibility, thushelpingtokeepitalive,butwe havenotedhow
the ongoingactivityof teachingis, at the same time,subjectto
variedpressures ofroutinisation, whichcan havea deadeningeffect
on the sense of plausibility. A second sourceof influenceon the
senseof plausibility-perhaps themostimportant one outsidethe
classroom-is interaction betweendifferent sensesof plausibility.
Thisinteraction can arisefroman articulation and discussion among
teachersof one another's pedagogicperceptions, fromprofessional
readingor writing, and in other,moreorlessformal, ways.
A specialist-level debatebetweendifferent methodsis,in fact,an
interaction betweendifferent sensesofplausibility, seekingto exert
an influence on all thosewho participate in it (moreorlessovertly)
through a processofsharing, sharpening, strengthening, weakening,
changing,or helpingto develop further the different formsof
understanding involved.A specialistwho advocatesa methodis on
thesamefooting as a teacherwhooperateswitha senseofplausibil-
ity,exceptthatthespecialistcan be saidto haveachievedfullerand
morecommunicable articulation ofa particular senseof plausibili-
ty,perhapsin the courseof wider(or longeror moreintensive)
interactionwith other similarlywell-articulated perceptions.If
ossification is lesslikelyto occurat thespecialist's level,itis onlyto

174 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
the extentthat the specialisthas more of a commitmentto
professionalinteractionand is more continuallyengaged in
exploringand articulatingsome senseof plausibility.
The resulting
well-developed and senses
well-articulated of (thatis to
plausibility
say,methods) have value not as desirable for
replacements many
teachers'sensesof plausibilitybut forwhat may be called their
powerto influence-toinvoke,activate,interact with,alterinsome
way, and generally
keep alive-differentteachers'differingsenses
thushelpingto promoteand enlargetheoccurrence
of plausibility,
of"real"teaching.A method,fromthispointofview,is notgood or
bad inanyobjectivesense,buthasmoreorlesspedagogicpowerto
influenceteachers'subjective understanding of teaching;and
debatesbetweendifferent methodsareimportant fortheprofession
because they help to give expressionand opportunity to the
pedagogicpower of differentmethods.

CONCLUSION
To summarise, ifwe regardourprofessional as a searchfor
effort
thebestmethodwhich,whenfound,willreplaceall othermethods,
we maynotonlybe workingtowardan unrealisable goalbut,inthe
process, be misconstruing the nature of teachingas a set of
procedures thatcan by themselves a
carry guaranteeof learning
outcomes.To say thatthebest method,in thissense,variesfrom
one teachingcontextto anotherdoes nothelpbecauseitstillleaves
us witha searchforthe best methodfor any specificteaching
context.To say thatthereis some truthto everymethoddoes not
help either,because it stilldoes not tell us whichpartof which
methodis true.Objectivemethodevaluationhas eitherto assume
thatmethodshavevalue forlearningindependent of teachers'and
students'subjectiveunderstanding of them,thusperpetuating an
unrealisablegoal and reinforcingthemisconstruction ofpedagogy,
or to tryto takeintoaccountteachers'subjectiveunderstanding of
teaching,thusceasingto be objectively evaluative.If,on theother
hand, we view teachingas an activitywhose value depends
centrally on whetherit is informedor uninformed by theteacher's
subjectivesenseof plausibility-onthedegreeto whichit is "real"
or mechanical-itbecomesa worthwhile goal forourprofessional
effortto helpactivateand developteachers'variedsensesof plau-
A methodis seen simplyas a highlydevelopedand highly
sibility.
articulatedsenseof plausibility,witha certainpowerto influence
otherspecialists'or teachers'perceptions.Perhapsthebestmethod
variesfromone teacherto another,but onlyin the sensethatit
is best foreach teacherto operatewithhis or her own sense of

THERE IS NO BEST METHOD-WHY? 175

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions
 at any giventime.There may be some truthto each
plausibility
method, onlyin so faras each methodmayoperateas one or
but
anotherteacher'ssenseof plausibility,
promoting themostlearning
thatcan be promotedby thatteacher.The searchforan inherently
bestmethodshouldperhapsgivewaytoa searchforwaysinwhich
teachers'and specialists'pedagogicperceptionscan mostwidely
interact
withone another, so thatteachingcan becomemostwidely
and maximally real.

ACKNOWLEDGMENT
A versionof thispaper was presentedas a plenaryaddressto the23rdannual
meetingoftheInternationalAssociation
ofTeachersofEnglishas a ForeignLan-
guage(IATEFL) heldinCoventry, England,April1989.

THE AUTHOR
N. S. Prabhuteachesappliedlinguistics
at theNational ofSingapore.
University
His earlierworkin India includedthedevelopment of task-based
language
teaching and theconceptof a procedural in thecontext
syllabus, of whatis
generallyknown as theBangalore
Project.

REFERENCES
R. L. (1984).The importance
Allwright, in classroom
ofinteraction lan-
guagelearning. 5(2),
AppliedLinguistics, 156-171.
C. J.(1984).Communicative
Brumfit, methodologyinlanguage teaching.
Cambridge: Cambridge Press.
University
M., & Hamilton,
Parlett, D. (1977).Evaluation
as illumination:A.new
approach to the of
study innovatory programmes. In D. Hamilton,
D. Jenkins,
C. King,B. Macdonald, & M. Parlett(Eds.),Beyondthe
numbersgame: A readerin educationalevaluation(pp. 6-22).
Macmillan.
Basingstoke:
Richards, C.
J. (1985).The context of languageteaching. Cambridge:
Cambridge UniversityPress.

176 TESOL QUARTERLY

This content downloaded from 141.161.91.14 on Wed, 21 Aug 2013 09:44:03 AM

All use subject to JSTOR Terms and Conditions