7

Second-Language Teaching Methods

7.1. Characterizing the Essentials of Methods

Secon d- Ianguage teachi ng is a field which providcs an cxcellent meeti ng
ground for many of the theoretical and practical aspects of psycholi nguistics
to come together. It is here tha t we have a chance to see how ideas of human
language and human learnin g interconnect .
In ou r view, language-t eaching rn ethods may be conveniently character-
ized according to five principal dimens ions (Steinberg, 1993):

I. Language Focus: Speech Communication vs, Literacy

2. Meaning Learning: Direc t Experience vs. Translation
3. Gramma r Learni ng: Ind uction vs. Explicatio n
4. Psychological O rientation: Mentalist vs, Behaviourist
S. Linguistic Orientation: Me nta list vs. Strucruralist

T hese dime nsio ns involve theories which have been realized in principal
second-la nguage teaching methods. Abrief description of each of these dimen-
sions, some aspects of which have been described elsewhere in this book,
folIows.

7.1.1. Language Focus: Speech Commu nication vs. Literacy

Methods can be divided into rwo categories of focus, those which teach
langnage throu gh the speech of the targe t language (the 'target language'
being th e language to be learn ed) and those which approac h the target
language th ro ugh reading and writing. Except for G rammar-Translation,
which focuses on reading, writing, and the trans lation of written words,
rnost other meth ods focus on speech and th e use of speech in communication .
 SECOND-LA NGUAGE TEACHI NG METHODS 191

The principal aim of Gra mmar-Translation is typically ro get srudents to

be able to read, and, ultirnately, to read literary works. Other proponents of
the meth od see Iireracy as a foun dation and a means for approaching speech
communication. T he prob lem with starring out with literacy when the goal
is speech is that students may never get to the speech stage unle ss th ey go to
university where the y may com e int o contacr with fluent instructo rs. Even
at university, though, the focus may rernain on literacy, Such is typically the
case in Japan.
Ge nerally. the proponents of speech- based methods regard G rammar-
Translation as their ultimate enemy, since th ey consider speech cornrnunica-
tio n to be primary in the learn ing of Ianguage. Speech-bas ed rneth ods attempr
to provide a speech environ ment in which students may lear n th e target
Ianguage. Reading and writi ng may be used, but only to reinforce what is
initially learned in speech.

7.1.2. Meaning Learning: Direet Experience VS. Translation

In providing the meaning of tar get language items, tr anslation may be used,
as is commonly the case with the G rammar-T ranslation method. For ex-
arnple, English-speaking students srudying Italian may be told thar 'libro'
means 'book', and that 'Come sta?' means 'H ow are you?' T hus the native
language (in th is case, English) is used to provide the meaning for the rarget
language (Italian). T he meanings of single vocabulary items and entire phrases
and sentences may be learn ed in th is way.
This is very different, though, frorn acquirin g meaning by being exposed
to actual objects, events , or situations in which the tar get language is used.
For exarn ple, th e lcarner can be shown a book and hear the teacher say
'libro', or sec rwo persons meet , with one saying to the ot her 'C ome sta?'
Meani ng here is to be learned through direct experience and not by the use
of th e native language to provide translation.

7.1.3. Gramm ar Learning: Induetion VS. Explication

Explication involves explanatio n, in the native language, of the grarn matical

rules and stru ctu res of the second language. For example, a teacher can
explain to japanese studen ts in the Japanese language that English has a
Subject + Verb + Ob ject ordering of basic sente nce constituents. (Japanese
has a Subject + Object + Verb ordering.)
In learning the same by induction , however, students would have to dis-
cover th ar order of constituents on their own. It would be necessary for
them to hear sent ences of the sort, 'Mary caught the ball', while experi -
encing a siruation in which such an action (or a picrure of the action)
19 2 PSYCHOLINGUISTICS

occurs. In this way they would discover for themselves, throu gh self-analysis,
i.e. induction , that English has a Subject + Ver b + O bject or dering.

7.1.4. Psychological Orientation: Mentalism VS. Behaviourism

Thc psychological presum ptions of a method can have a great effect on how
that meth od is formulared and used. A Behaviourist would prefer, for example,
to mechanically dr ill srudcnts on sente nces while a M entalist wou ld prefer to
have students th ink about sentences and their structure and learn abou t th em
in this way. For the Behaviourist, there is nothin g for a Iearner to th ink about;
thinking is irrelevant for language learning, only habit formati on is important.
As stated by Brooks (1964), 'The single paramounr fact about language learn-
ing is that it concerns, not problem solving, but the format ion and perform-
ance of habits' (p, 49). Fortunately, not many nowadays hold such a view.
O n th e other hand, in a Mentalist approach to language, students may
be given more time to puzzle over speec h and less time for dri ll. For the
M entalist, a sentence is mo re than a sequence of overt words, for un derlying
those words is an abstract mental structure th at involves a lot of abstracr
opera tions in its formation.

7.1.5. Linguistic Orientation: Mentalist vs. Structuralist

What one bclieves to be the concept of a senrence, and what grammat-
ical rules and structures may underlie th e sent ence, will affect grea tly what
one teaches , T he linguistic Stru cruralist is the counterpart of th e psycho-
logical Behaviourisr, According to the Strucmralists (Bloomfield, Fr ies, Pike),
a sente nce like 'T he dog jumped' would be analysed as a simple order of
word classes (Article + Noun + Verb or at best a sequence of phrases (Noun
Phrase (th e dog) + Ver b P hrase (j umpedj). Yet, as C homsky (1957, 1965)
po inted out in his or iginal attac ks on Strucrural Linguistics in the 1950s
and 1960s, ot her sent ences having the same observable structu re, such as
'J ohn is easy to please' and 'Jo hn is eager to please', canno t be explained
by a simple listing of word classes or even phrase structu res since both of
these sen tences are identical in this regard, i.e. N oun + Verb + Adjective
+ Preposition + Verb,
A Me ntalist grarnmarian would explain th ese sente nces by discussing the
syntactic or sernantic relations th at underlie th ose sente nces. T hus, a Mental-
ist could say th at in 'John is easy to please', '[ohn' is the underlying object
of 'pleasc' , while in 'J ohn is eager to please', 'J ohn' is the underlying subject
of 'please' . In practical terms, a teacher would have quite different con-
ceptions to offer students with respect to such sentences .
D epend ing on the linguistic orientation of a theor ist or teacher, th e very
nature of sente nces will be conceivcd of quite diffcrently. Such an orientation
will inevitably affect how senten ces are to be taught or presented to studen ts,
 SECOND -LA NGUAGE TEACH I NG MET HO DS 193

7.2. Traditional Methods: Grammar-Translation, Natural,

Direct, Audiolingual
With the above fi ve dimensions in mind, let us now examine some of the
major second-language teaching methods. For a more in-depth treatment
of the thr ee oldest methods: Gram mar-Translation, the Na tural Method,
and the Direct M ethod, the reader should refer to the works of Kelly
(1969), T itone (1968), and Darian (1972). For consideration of the histor -
ical movemcnt between the old and new methods see Howatt (1984), and
for an overview of most current method s see Richards and Rodgers (1986).
Since the 1980s little that is new has happened in teaching methods.
Under the heading of Traditional Meth ods, we shall consider the fol-
lowing: ( I) Gra mmar-Translation Method, (2) Na tural Met hod, (3) Direct
Method, and (4) Audiolingual Method.

7.2.1. The Grammar-Translation Method

Features cf Grammar-Translation
T hc Gra mmar-Translation (GT) method cssentially involves two com-
ponents: (I) the explicit explanation o[ grammatical rnles using the native
language, and (2) the use of translation, in the native language, to explain
the meaning of vocabulary and structures. Translation is thc oldest of the
components and is probably the oldest of all formal teaching methods,
having been used in ancient Greece and Rome and elsewhere in the ancient
world. The grammar aspect of GT was rather limited in those times since
gramrnat ical knowledge itself was limited. It was later in Euro pe, particularly
in the seventeenth century, that intensive and detailed studies of various
languages were conducted. Wi th this spirit of the Renaissance came an
interest too in the understanding and teaching of ordinary (non-Classical)
languages.

Effeetiveness in mass education

The modern form of the GT method is an attem pt to devise a method of
foreign language teaching for the purpose of mass education. In medieval
times, the learning of languages by individual scholars had consisted of the
study of grammar and its subsequent application to reading with the aid of
a dictionary. T his style of learning by individual mature adults had to
be adapted to thc mass education of youth as schools for adolescents began
to appear in the eighteenth century. The grammatical explication and trans-
lation aspects were retained because rhe reachers had already been trained
in these techniqu es through their individual studies in thc Classics.
H owatt (1984) explains that grammar and translation were merely
borrowed from earlier scholarly methods, and that the significant change
19 4 PSYCHO LI N GU ISTICS

which occurred was from un derstanding th e meaning of enti re texts to that

of the ind ividual sente nce as th e basis of langu age education. O ut-of-
context sentences from th e text serv ed as exarnples of the gram mar po int
being taug ht for the day's lesson . Senten ces were also th e focus because
they were conside red to be easier for th e young stu de nt to un ders tand tha n
entire texts. T hus , whil e the scho lar could spend hour s of stu dy on a single
text, an institute of educa tion, wit h many d asses sched uled an d a Iimit ed
amo unt of time for each dass, would need a sma ller un it of analysis for
srudents .
T he teaching o f gram mar went han d-in-hand with translation for the
teaching of a seco nd langua ge, with both relying on th e use of the native
language to imp art knowledge. With th e growth of grammatic al kno wledge,
however, th e grarnrnatical co mpone nt played a greater role in teaching,
eventu ally do minating the tr anslation aspect. By the end of the eight-
eenth centu ry in Europe it had become a full partn er in the met hod . T he
gr owth of the grammatica l component continues to th e present day. Rules
are explained by the teacher, the n memorized, recited, and applied by
th e srudent,

lntroduction of uernacular languages

T he a irn of GT has change d over the years. Originally it had two pr incipal
aims: (1) th e study of the literarure o f the seco nd langu age, which was
typ ically C lassical Greek or Latin , and (2) th e developm ent of analyt ical
skills th rou gh th e srudy o f gramma r, T he latter airn, whic h was mo tivated
by Ca rtes ian gram ma rians who believed th at a universal logic und erlies all
langu ages, has largely disapp eared , along with th e notion that by teaching
a lan gu age one could train th e m ind in logical processing.
Wi th th e Enlightenment, vernacular languages other th an G ree k and
Lat in were introdu ced and some atten tion was given to pron unciation and
listeni ng. However, for the most part, GT's orientation to literarure rernains
with its em phasis on reading and writing in the seco nd langu age.

The modern GT approacb

Typica lly, modern textbooks using the G T Metho d have lessons which
in d ude: a read ing passage in the target lan guage, a list of vocabulary iterns
and th eir translations, and an explanation in th e native lan gn age of import-
an t poin ts of gram mar exem plified in the text , T he lesson ofte n ends with
aseries of exer cises, ra nging from straight translation to qu estion s on poin ts
of gramma r. T ranslation is typ ically don e from th e targe r lan gu age int o
th e native lan gu age, with reverse translation (from the native language int o
the ta rget langu age) seldo m being done. T he teache r will spend most of the
dass time explaining th e gra m ma r point s, while occasionally questioning
stu dents abo ut a particular translation or having stud ents read aloud and
explain the meani ng of what they have rea d.
 SECO N D- LANGUAG E TEAC HI NG M ETHO D S 195

Advantages of GT
Despite the method 's ind ifferen ce to speech and ora l communication, and
despite its being disparaged by leading language educators for such an indif-
ference, the GT method has enjoyed and conti nues to enjoy acceptance in
ma ny countries around th e world. This may seem a mystery, until one looks
at the advantages of GT.
(a) Non-fluent teacbers can teacb Im'ge classes. T he meth od can be applied by
teachers (1) who lack verba l fluency in the target language, both in terms
of understanding and producing sp eech , and by teachers (2) who have an
incomplete know ledge of th e langu age. T his situa tion is common in many
countries, typ ically underdeveloped ones, where kn owledgeable teac hers are
scarce.' It is not uncommon in such countries for teachers to be place d in a
dass with 40 , 50, and more students.' In effect, Ianguage learn ing is treated
as a mass lecture course where, rypically, students only meet once a week.
(b) Self-study, T he me tho d also lends itsel f well to self-study. By using
boo ks, students can study on thei r own outside of the dassroom. There
is much that they can learn from srudying and reading on th eir own. Of
importance , too, is the fact that the me thod is appropriate for all levels
of learn ers. From the introductory to th e very advanced, th ere is an abund- .
ance of materials available for dassroom use.
(c) A daptability to cbanging linguistic and psycbological tbeories. One of GT's
strongest points is its capacity to adapt to ever-cha nging linguisti c and psycho-
logical theories. The distinguish ing feature of the me thod, the explication
of grammar, can easily be adapted to new ideas and theories. G ram mati cal
explanations can be couched in th e linguistic theory of the day. Wheth er a
grammatical point is to be explained according to C ho rnsky' s or Bloomfield's
theory of granunar is of no concern to the me thod - GT is neutra l with
respect to any specific grammar. Whatever grammar it is fed, that is the
grammar it will explain. Similarly GT is neutr al about whether a ßehaviou rist
or aMentalist psycho logical th eory is applied,
In th is way, GT need never becom e obsolete from a linguistic or psycho-
log ical point of view. The fact that it th rived under Structural linguistics and
ßehaviouristic psychology did not prevent it from thriving under M enta lism.

Success and [ailure

Almost everyone who has stu died second-language teaching me thods has
criticized GT. In fact, we are not aware of any known theorist who is a
proponent of GT! Yet, despite sustai ning over a cen rury 's attack by a host
of opposing methods (which we will discuss), GT has survived . Although
we, too, are members of the oppositio n we do recognize th ar GT is not
a failure. Ma ny srudents do learn a goo d part of a second language th ro ugh
GT.
Where GT fails, however, relates to its secondary treatm ent of COIll-
mun icative ora l skills. Students who pass through ma ny year s of strict GT
196 PSYCHO LING UISTICS

tr aining o ften come out unable to comprehen d or utter sen tences at a level
that allows th em to engage in even simple co nversat ions.
A lirnitation of GT, which sho uld be not ed, is that it can not be used with
young children, for young children cannot read or write and are una ble
to understand gra mma tical explan ations. Perhaps this is a blessing in disguise
for countries which are pre disposed to GT, such as J apan. Since J apan ese
children in th e early grades are goi ng to be taugilt Eng lish and other lan-
guages, and since this cannot be don e through GT, mor e narur al spee ch-
co mmunication -based techn iqu es are go ing to have to be used.

7.2.2. The Natura l Method

Tbe Natural Metbod as a product of tbe Enligbtenm ent
T he Na rura l M ethod (NM) devcloped as areaction to Gra mmar-Translation
and was th e outgrowth of scientific thought on th e narure of langu age
and language learning. Such knowledge flowered in Europe with inspiratio n
fro m the work of Comenius (1568), Rousseau (1780), and oth er theorists
such as Pes talozzi (1801). The philosophy of the En lighte nme nt duri ng th e
eigh teenth cenrury was part icularly concerne d with th e natural srate of
hu man beings. Q uest ions abo ut the natu ral developrnen r of hu mans and
th eir langua ge became of grea t inte rest,
NM began to be form ed early in the ninet eenth cenrury and by the latter
part of th at cenrury the meth od had become firm ly esta blished th rough the
writings of such as Sauveu r (1878) and Go uin (1880). Go uin observed chil-
dre n learn ing language and noticed that this occur red wirhin the context
of meaning- rel ated siruations (see C hap ter I). T his observation of children's
language learn ing was then applied to secon d-language teachin g methods
for chil dren and adults.

N atural order of language learning

T hat approach to language learn ing, whe re ' natu ral is best', so to speak,
led to a method of teaching th at stressed the value of intro ducing a seco nd
lan guage to a learn er exactly as the native language had been experienced.
T he mo del for the N arura l Method of second -language learni ng was the
child learn ing its native language. T his rnean t adherence to the na tu ral
sequence of the child's acquiri ng its first language, i.e. (I) speec h COIn-
pr ehension , (2) speech pro duction, and, mu ch larer, (3) reading an d (4)
writing. Grammar was not taugh t directly. Rath er, grammatical rules and
strucrures wer e to be learn ed th rough ind uction (self-analysis) by experi-
encing speech in a siruational contcxt , Me aning was to be gaine d th rou gh
experience and exposure to objeets, siruatio ns, and events; translation was
to be avoided .
Typically, teachers would not use prepare d siruatio ns or material. Learn-
ing was through 'spo nta neo us' co nversat ion and derno nstrutio n, all of whic h
 SECO ND -LA NG UAGE TEACH IN G MET HODS 197

was done in the target language and supported with gestures and actions.
The teacher used language appropriate to the students' level of understanding,
much in the way parents would with a child. The method was totally oriented
towards the acquisition of oral skills. Student participation in sirnational
activities was the essence of this kind of second-language learning.

Advantages and disaduantages

The great advantage of NM was that by exposure to natural language in
a natural context, learners could acquire a speech capability both in under-
standing and produ ction . H owever, one prob lem for this method is that
it requires th e teacher to create interes ting situations so that students may
be naturally exposed to langnage. T his, and the reliance on sponta neous
speech, places an extre mely heavy burd en on even the best of teachers.
Besides possessing an undue arnoun t of ingenuity teachers must, of course,
be fluent in the target language. Such a demand cannot always be met,
particularly if mass education is involved. Class size, too, could be a prob-
lem, since the numb er of students must be quite smalI, usually less than 15.
Acrually, the probl ems mentioned here are not unique to NM . Indeed,
all speech-based method s have similar problems, given their cmphasis on
exposure to natural speech and student participation in a variety of com-
municative situations.

7.2.3. The Direct Method

The Direct Method develops fr om the Natural Method

T he Di rect Method (DM), appearing in the late nineteenth and early rwen-
tieth centuries, developed from the Na tura l M ethod. Like the N atural
Method , it emphasized the learning of speech, acquiring meaning in envir-
onrnental context, and learn ing grammar th rough induction.
T he advocates of DM, while approving of the Natural Method, sought
to improve upon it by providing systematic procedures based on scientific
knowledge oflinguistics and psychology. For example, in psychology, Franke
in the 18805 argued for the exclusive use of the second language in the
classroom and discussed the importance of the direct connectio n between
meaning and form in the second language. (T hese were the days of W illiam
j ames's psychology, with its connections bcrwecn ideas and behaviour.) The
native language was not to be used as an interm ediary in any way. The
narnc, Direct M ethod , incidentally, refers to this direct connection between
the second language and meaning.
DM theorists believed that by applying scienti fic knowledge from psy-
cho logy arid linguistics, language learnin g could be made more efficient,
with the result that students would learn faster than they would under the
sponta neous and unplanned lessons of the N atural Method. H aro ld Palmer
(1922) was perh aps its most articulate and eminent advoca te,
198 PSYCH OLI NGUI STICS

DM aduocates natural learning but tuitb graded materials

Like the Natural Me thod, DM is menta listically oriented since it presumes
that the learn er is a th inking being who can learn abstract language ideas.
Also, like the Natural Method, DM relies on learnin g the language by induc-
tion . However, unlike the Natural Method, language materials for teach ing
in DM are explicitly preselecte d and graded on the basis of linguistic com-
plexity. Simple sente nces, for example, prece de those with relative clauses or
in the passive construction. All of this is done for the purpose of making the
acquisition rask easier for the learner. While there is still much spontaneous
use of speech by the teacher, it is considerab ly less than is the case for the
Natural Method.

Dia/oglle and action materials

Lessons in DM are mainly devoted to oral communication and follow (as
with N M) the acquisition order of the first language. T hus speech under-
standing precedes speech prod uction, which is then followed by reading and
then by writing.
Elementary social dialogues are introduced almost immediately: ' H ow are
you?', 'Fi ne, rhanks', as are questions: 'Where is . .. ?', 'When is . . . ?', 'Who
is . .. ?', and comma nds for action : 'Sta nd up', 'Sir down', and 'G ive the book
to Mary'. (The similarity here to the fundamenta ls of the Total Ph ysical
Response Me thod which was proposed some 50 years later is important to
note and will be discussed later.)
Sometimes oral pattern drills and memorization of dialogues were also
included in DM lessons. Such techniques were devised and applied for the
purpose of giving practice in speech pro duction . Interestingly, these same
techniques later came to be used (perhaps more accurately over-used) by
proponents of the Audiolingual Me thod .
Sometimes, too, trans lations might be given verbally, as might gramrn at-
ical explanations. However, these were used sparingly. For the most part,
DM is typified by its reliance on natural speech in context and on the
smde nts' ment al powers of inductio n.

Teacber ftuency and dass size

The structure d nature of the Direct Method is such that, in the hands of a
good teacher, it can be used in relatively large classes of even 40 srudents,
with teachers getting students to speak in chorus. Still, like the Natural
Method, DM requires a teacher with high fluency in the second language.
Some schoo l systems may find it diffi cult to find a sufficient numb er of such
teachers .

One psycbo/ingllist 's experience untb DM

Desiring to experience what the child experiences in learning a language, the
psycholinguist Roger Brown (1973, p. 6) enr olled in a two-week intensive
 SECO ND - LAN GUAGE TEACHING M ETH OD S 199

langu age programme which used DM . H e progresse d very weil while he

had th e sarne teac her, but as soo n as the teac hing sched ule changed and
he had new tea chers who were not fami liar with his specific knowled ge, he
co uld no longe r perform at the level he had reached with his first teacher.
T he teache rs failed to coordinate the ir materials, He came out of th e course
saying that he had little speech at his com man d. He may, though, have
understood much more than he was aware of. In any case, given that the
langu age was j apanese, a language whose gra mmar is rad ically different from
any Euro pean Ianguage thar Brown had experienced, the rwo non-Japanese
authors of this hook th ink that his expectations were muc h roo high - for
a two -week course'

Demise of DM
With th e adve nt of th e Audiolingual Method, DM was crus he d. Ir has
alrnost disappea red . O n the other hand, it should be recognized that a number
of cur rent meth ods such as Total Physical Response and the Natural Approach
reflect the essential ideas of DM arid hence may be viewed as development s
of th e Direct Method.

7.2.4. The Audiolingual Method

Despite the Direct Merhod's lon g and widespread acceptance (in other
than Grammar-Translation circles), it was overshadowed and th en virtually
wiped out with the advent of the Audiolingual Metho d (ALM). AU,,! was
advocated by suc h th eorists as Fries (1945, 1949) and Lado (1957).

Popularity of A m erican linguistics and psycbology and tbe rise of A LM

T he phenomena l rise of ALJV! was d ue to the popularity of the American
lingu istic and psychological theories which it incorporated into its founda-
tions. The great popularity and influen ce of America itself in the world,
follow ing the end of the Second World War, is a factor here. T he langnage
ana lyses provided by Arnerica n Structural linguists and the stim ulus and
response learning psychology provided by American Behaviourists endowed
ALM with great cre dibiliry, The Direct Method, which imp lied aMentalist
psycho logy, went out of fashion, except in Contineutal Europe.

ALM incorporates stmaural linguistics

Strucrural linguists such as Fries regarded sent ences as sequences of gram-
matical word classes or phrases, New sentences wou ld be created by sub-
st ituting words within a word dass. For example, a sequen ce such as Artide
+ Adjective + Noun + Verb + Artide + Noun coul d yield a large nu mber of
sentences suc h as 'T he rich hoy bough t a car' and 'The friendly girl kissed
the boy', by su bstituting members of the sarne gra mrnati cal dass. Because
Behaviou rist psychologists, too, regarded sentences as the simp le association
200 PSYC HO LI NG UISTICS

of key words (Skinner) or word c1asses (Staats), it was not much of a step for
ALM to adopt sente nce patterns as the learning fundamenta ls for language.
Unfortu nately for the theory, as we point out in Chapter 12, substitution
cannot prevent the creatio n of sequences like 'T he happy dust memorized
the table', or 'A poor mountain elapsed the wine' , which also fit the pattern
for the sentc nce 'The rich boy bought a car'. T here were ot her more serious
problems with the theory which Chomsky (1957, 1959) pointed out. Struc-
tur alist th eory could not account for a spcaker's ability to genera te gram-
matical sente nces of no fixed length or number (which were not defined by
such sente nce pattern s), nor could it account for sente nce synonymy ('Jo hn
sang then he danced' and 'Jo hn sang then danced') and structur al ambiguity
('T hc shoo ting of the hunters was terrible'. Different answers can be given
to the questions of 'Who was shot?' and 'Who did the shoo ting?'). (These
probl ems too are discussed in Chapter 12.)

A LM incorporates Bebauionrist psycbology

ALM incorporated Behaviourist psychology (W atson, 1924; Thorn dike, 1932;
Skinner 1957), which was rhe dom inant school of psychology in America for
most of the first half of thc twentieth centu ry. Behaviouri st psychology
regarded mind and thinking ro be irrelevant for rhe und erstanding and
productio n of speech. Language learn ing was regar ded as no different from
other types of learnin g in which a stimulus and response paradigm was
operating. Repetition and mechanical drills involving word s as stimuli and
responses were considered to be the essence of learn ing.
T he defec ts of such a view concern ing language and psychology were
demonstrated by C hornsky during the 1950s and served as the basis for the
subsequent collapse of Stru crural linguistics in the 1960s and the downfall
of Behaviou rism as the principal explanation for linguistic behaviour in the
1970s.

Features 0/A LM
T he Audiolingua l Meth od incorporated into its methodology many of the
same features which the Dir ect Method had developed, namely, planned
situa tions, graded materials, and such techniqu es as pattern drills and dia-
logue memorization (Brooks, 1964). In contrast with DM , th e Audiolingual
Me thod almost entirely dropped the use of natur al situatio ns and sponta ne-
ous speech . There was even a tendency for some ALM advocates, such as
Mo ulto n, to reduce the meaningfulness of the speech that was taught - a
practice which was frowned on by Fries, one of the founders of ALM.

Success 0/A LM
In its tim e ALM generated an enormous amount of ent husiasm. T eachers
everywhere lined up to teach second languages accord ing to principles which
reflected the latest scien tific word on how humans learn langu age. In the
 SECON D- LANG UAG E TEACH I NG METH ODS 20 1

1950s the U niversity of Mi chigan at Arm Arbor, where both Fr ies and Lado
taught, was the cent re of the ALM univ erse.
H owever great the populariry of ALM, the fact is th at ALM failed to
produce the fluent communicating spea kers it had promised. This might
have been overlooked by th e second-language teaching commu nity, though ,
since no othe r rnethod had pro ved that it could do better. H owever , a
revolution was in th e offing and this revolution , which was starred by a single
person , Noam C ho msky, was ro rip away th e th eoretical un derpinni ngs of
ALM and destroy it as a cohe rent movemen t,

7.3. The Chomskyan Revolution Brings Down the

Foundations of the Audiolingual Method

T he attacks on Structural linguistics and Behaviouristic psychology which

the linguist Noam C ho msky rnade in th e 1950s and 1960s were so devastat-
ing thar virtual ly sing le-handed ly he was able to bring down both of these
paradigms. (H is essent ial argum ents regardi ng linguistics are deta iled in
C hapter 12.)
C homsky's attraction was the formulatio n of a powerful grammar cast as
a set of recursive (repeata ble) rules. T hese rules were of a conce prual abstract
nature arid differed qualitatively fro m th e observable words of a sentence.
H e argued for a grammar that requ ired mind and mental operation: th e very
sort of ent ities which Behaviou rist psycholog ists and Structu rallinguists had
opposed and had founde d th eir argu me nts against. C homsky convincingly
demonstrated th e adequacy of his noti ons. (A detailed discussion of the
adequacy of Beh aviourism as contrasted with that of M entali sm is presented
in C hap ter 10.) The outcome of C homsky's theo rizing was that if langnage
use and language learn ing were to be explained, both lingui stics and psy-
cho logy wou ld require a M entalistic base.
W ith both the lingu istic and psycho logical foundation s of its method
undennined, the imp acr on the Audiolingua l M eth od was devastatin g, All
thar rerna ins of ALM today is th e occasional use of pat te rn practice drills as
. an auxiliary exercise in the second-language classroom .
It migh t be noted th at while C homsky has prop osed ideas concerning
first- language acquisition, he has avoided speculation rega rd ing the teachin g
and learning of a second language. H e has left it to orhe rs to atte rnpt to
apply his ideas in the second-language field. The effects of his ideas, never-
th eless, have been pro found . T hese effects range fro m how grammar prob-
lem s are to be explained in rhe Grammar-Translation meth od to how ru les
sho uld be presented in th e Natural Approach. (The Natural Approach, which
will be discussed later in this chapter, sho uld not be confused with th e
Na tura l M ethod which was discussed earlier in the chap ter.) C homsky's basic
202 PSYCHO LI NG U ISTICS

ideas in linguistics and psychology have been absorbed by second-language

theorists in a variety of ways. Few are the theorists today, for example, who
do not take a Menralistic approac h to second-Ianguage problems.

7.4. Offbeat Methods Appear then Disappear:

Cognitive Code, Community Language Learning,
Silent Way, Suggestopedia
Since the downfall of the Audiolingual Method in the 1960s, a number of
new mcth ods have arisen. H owever, only a small number have managed
to survive, and fewer still have managed to thrive. Four that have not sur-
vived are Cognitive Co de, Co mmunity Langnage Learn ing, Silent Way, and
Suggesto pedia. Of course, there are a few adhere nts sprinkled abour, just as
there are some Audiolinguists around, bur they are disappearing. Later we
will discuss more viable methods: Total Physical Response, Communicative
Language Teaching, and the Na tura l Approac h.
Let us now begin with a brief description and assessment of (I) Cogn it-
ive Code, (2) Com munity Language Learn ing, (3) Silent Wa y, and (4)
Suggesto pedia.

7.4.1. Cognitive Code

Cog nitive Co de (CC) arose in the 1960s as one of thc first reactions to the
Audiolingual Me thod and one of the first to apply Chomsky's ideas to the
teaching of a second language. W ith changes in psychology and linguistics, a
new approac h to second-Ianguage learn ing was needed .
T heo rists who are associated with this or ientation, e.g. Ausubel (1964),
Chastain (1969, 1971), and Donaldson (1971), are typically Me nta list in their
philosophy, advocates of generative grammar in their Iinguistics, and eclectic
in their methodology. T he teaching of grammatical rules was permi tted
thro ugh both inductive and explicative means. And there was no strict
sequencing, where specch had to precede literacy. T here was no special order
to reading, writing, and speaking. T he teacher could mix them by saying
sentences and writing those sente nces on rhe board . However, proponents
of Cognitive Code developed little in the way of a distinctive method. T he
ideas espoused by Cog nitive Code theorists are now used to support other
methods which advocate the use of meaningful language-use for learning.

7.4.2. Community Language Learning

Co mmunity Language Learning (CLL), or Counseling Learning as it
is sometimes called, was originared in the 1960s by Charles A. Cur ran, a
 SECOND· LANG UAGE TEACHING METHODS 203

coun sellor-therapist and priest who regarded the second-language learn ing
situation fro m the point of view of small-gro up dynsm ies and counselling
(Curran, 1972, ( 976).
T he teacher takes the role of a counsellor while the learner takes the ro le
of a di ent. The d ients are to inte ract independently with one anot her, with
the cou nsellor's ro le being only to foster that interaction. In effect, th is is
reduced to the counsellor translating into the target language whatever it
is that the d ient s wish to say to one ano the r. The d ients sit in a cirde arid
converse using only the target langua ge. T he counsellor stands behind th e
dient who is to speak. T he di ent teils the counsellor in the native language
what it is that he or she wants to say, and the counsellor provides the
translation. T he dient the n utters tha t translated piece of target language to
one or more of the other dients, who, in turn, are obliged to respond. T he
counsellor is obliged to run around the room giving d ients translations for
whatever it is th at the y want to say. Discussion of grammatical points and of
the language is kept to a minimum.
Since Cur ran's original CLL pro posal (and his fantastic unsubstanti ated
clairns for the meth od's success), numerous versions of the meth od have arisen,
some of which differ so greatly that perhaps all they have in common is the
establishing of some sort of small-group interaction (La Forge, 1983). For
example, th e sentences utt ered by the group may be recorded, transcribed,
and then given to the srudents for study and rnemorization. The students
reflect upon the interaction and frankly express their feelings about the session.
T hey th en raise question s about th e grammar and ot her aspects of language
and the teacher provides detailed explanations (Stevick, 1980). H ere, CL L
does not differ in fundamen tals from the traditiona l G ramma r-Tra nslation
meth od. The small-gro up interaction serves to maintain a certain distinct-
iveness for C LL, but it was not sufficient to keep this met hod afloat.

7.4.3. The Silent Way

Rationale: student speaks, teacber Silent

T he Silent W ay (SW), developed by G att egno (1972, 1976), is based on the
radical notion tha t the teacher is to be as silent as possible while the students
employ their own abilities to discover and creat e the language themselves.
Whi le other methods, such as the Na tural Meth od, view the pro cesses of
second-language learnin g as similar to that of first-langu age learning,
Ga ttegno argu ed that the processes are different because a second-language
learn er already knows a first language and has adult cognitive abilities. Con-
sequently, teachin g must 'replace a "natural" approach hy one that is very
"artificial" and, for some purposes, strictly contro lled' (Gattegno, 1972, P: 12).
T he underlying approach to this method is said to be based on the 'cre-
ativc' aspect of language learn ing, where learn ing is viewed as a process of
2 04 PSYCHOLIN GUISTICS

discovery or creation on the part of the stu dent, The students are to guess
on their own the grammatical rules and str ucrures which are inherent in the
situations presented to the m. T his is particularly difficult for students be-
cause the teacher is typically silent and so the studenrs have little speech data
to analyse. H aving the teacher silent is especially peculiar since learning
from a model , be it spoken or written, is essential to every other teaching
method ever devised.

Production precedes comprebension

In contrast to other speech-based methods, the Silent Way virrually reverses
the natural sequence in first-Ianguage learn ing by having speech produ ction
precede speech comprehension. The teacher says little but rather encour-
ages the students to talk. As might be expecred, this is especially difficult and
stress-provo king since the students do not know how to say anything in the
beginnin g. T he instructor does not generally model the pronunciation but
points to lett ers on a special chart, and waits for good pronunci ation from
someo ne in the d ass and lets that serve as a mod el, T he teacher requires the
students to produ ce as much speech as possible, and as early as possible.

M aterials
T he teacher then uses a certain set of physical objects, such as the coloured
rods that Ga ttegno specified, in order to convey the meanin g of the words
and the gram rna tical rules by which sentences are to be construc ted. Some
participants of Silent Way d asses are enthusiastic about the method. I-Iowever,
a good many srudents react quite negatively to the stress of having to discover
grammatical mi es with no speech model present . Although the learn ers are
expected to 'work cooperatively rather than competitively' (Richards &
Rodgers, 1986, p. 106), because they cannot rely on the teacher but must rely
on their classmates for learnin g, competition is often the norm.

All aduenture untb SW

T he third aut hor of this book some time ago participated in a Silent Way
course in Chinese. H e found that of the five students participating one
excelled, anot her did weil, the next hung on, and the last (Wo gave up. The
meth od does not take into account the individual differences in the students'
learnin g styles. Addition ally, although students are supposed to take the
initiative in the learning process under the Silent Way, there is little that
a student can do when the teacher is not pr esent, It is the teacher who points
at the wall chart, moves objects, etc, Although the method may be success-
ful to some degree with some srudents within tightly controlled settings,
any success may not be easily tran sferable outside of those setti ngs (Lantolf,
1986).
SW must be one of the most ridiculous language-teaching meth ods every
devised, next to Suggestopedi a!
 SECOND-LANGU AG E TEACH ING M ETHOD S 205

7.4.4. Suggestopedia

Super memory induced by relaxation

If one believes th e claim s made abo ut Suggestop edia by its Bulgari an foun der
Lozanov (I 978), then it is the closest thin g to the 'magic method' which
everyone has been lookin g for in seco nd-language teachin g. Briefly, Sugges t-
opedia purports to produce in stude nts an altered state of consciousness which
is conducive to learning. T his state, terme d 'hypermnesia' (super mernory),
is bro ug ht abou t by certa in relaxation techniques which serve to build the
con fidence of the learn er and thu s to break down th e 'antisuggestive barriers'.
Relaxation is achieved thro ugh listening to cert ain specified passages
of classical mu sic, The music rnust be played at a specific tem po to enable
it to induce th e desired state of men tal readiness. Sorne claims about the
best music to be used have been made by 'East German researchers
of Sugges to pedia at [the former] Kar! Ma rx University in Leipzig [who]
observed th at slow movernen ts from Baroque instrum ental music featu ring
stri ng instruments gave the very best results' (Ostrander, Schroe der, &
Ostrander, 1974, p. 115).

Role of the teacber and[antastic claims

Furtherm ore, th e learn ers must be provided with armchairs and pleasan tly
decorated rooms. O ther possible relaxers such as alcohol, however, ar e dis-
courage d. T he confidence of the learn er is built up by what th e teacher says
and does. T he tea cher is to suggest th ings to th e learn er and to act in a
highl y authori tarian and confide nt way. (These were the days of Stalinist
Bulgar ia!) As a result, according to Lozanov, seco nd-Iangu age learn ers can
learn 1800 words , speak within the framewerk of a whole essen tial gra mmar,
and read any text - all in 24 days!

Grammar- Translation is fundamental

Teaching involves the presentation of dialogues and vocab ulary which the
stu dent is to stu dy and memori ze. The materials are presented first in writ-
ten th en in spoke n fonn . A translation is offered alon g with the written
form. The uni que aspect of teaching lies in the way materia ls are then
presente d in co njunction with certai n learn er behaviou r and environrnental
events. While the students are relaxing in their armchairs, the teacher reads
each dialogue aloud th ree tim es, in a special way.
T he special way of reading is important and in clud es 'varyi ng int on ations
and a coordination of sound and printed wor d or illustrati on' (Bancroft, 1972,
p. 17). H owever, this special way of reading has no t been explained in anyth ing
but vague rerrns. In discussing th is reading style, Stevick (1976) notes the lack
of specificity of tec hnique: 'The precise way of using voice quality, in ton a-
tio n, an d tim ing are appa rently both imp ortan t and intri cate. [However] I
have found no one who could give a first-band acco unt of th ern' (p. 157).
206 PSYCH OLIN GUI STIC S

On th e first presenration the students follow by reading. On the second

and third readin gs th e students listen o nly. It is on the third reading thar the
mu sic is played, sup posedly inducing hypermnesia and learning o n th e part
of the srudent, In effect, Suggestopedia is little more tban Grammar- Translation
witb music.
What can we say abo ut the extraordinary claims of success which have
been made by Lo zanov arid his small group of supporte rs (Bancro ft, 1972,
1978; Stevick, (9 76)? There is certainly nothin g wron g with th e idea of
memory enhance rn ent, If a second -Ianguage teachin g method comes along
and claims, as does Suggestopedia, to greatly en hance mem ory by relaxarion
and music th ereby allowing for an enormous arnou nr of language to be
acquired in just a matter of weeks, it should not be dismissed out of hand,
The fact of th e matter is that, almosr 30 years afte r its introduction, th c
method, which has been given a fair try in many countri es, has still not
provided convincing evidence in support of its extravaga nt claims.
T hus, there seems to be little reaso n to amend Scovel's (1979) evaluatio n
of th e meth od: 'suggesropedy, taken as a self-conta ined method for language
instruction, offers at best nothing much that can be of ben efit to presem day,
eclectic EF L programs, and at wo rst nothing more than an overso ld package
of pseud oscientific go bbledygook!' (I'. 258).
Suggestop edia's on ly legacy today seems to be that some teachers play
mu sic before th ey begin class in order to calm stu dents down.

7.5. Contemporary Methods: Total Physical Response,

Communicative Language Teaching, Natural
Approach
7.5. 1. Total Physical Response

Rationale of tbe metbod

T otal Ph ysical Respon se, frequently referred to as T P R, is very much a
'natural'- type meth od: speech unders tanding precedes speech production ,
which, in turn, precedes readin g and writing. Only th e target language is
used in the classroorn and meaning is derived from actual objects and siru-
at ions, Srudcnts are enco uraged 10 induce ru les on th eir own and speak when
they are read y. Again, as with o th er natural -type rnethods, thi ngs go besr
with a sma ll number of students .
James Asher, th e founder of TPR in th e 1970s, con siders its unique
characteristic to be the learn ers' perfo rmance of physical actions in respon se
10 th e teach er's commands in th e targe t language (Asher, 1966, 1969, 1977;
Asher, Kusud o, & O e La Torre, 1974). H is idea is th at mem ory will be
enhanced by rnot or activiry with th e result th ar language will be mo re easily
 SECOND- LANG UAGE TEACHING METH OD S 207

re membered and accessed. Interestingly, th is idea and the ot her rnajo r ideas
comp rising T P R are to be found in the Di rect Method, particularly with the
Palm ers (Palme r & Palmer , 1925) in thei r book, Langllage Tbrougb A ctions.
Asher, though, has em phas ized phys ical activity much more tha n did Palm er.
In any case, there is no do ubt that TPR is a very useful meth od and one
which deserves at tention.

Classroom matertals and actiuities

In itially, in a dassroom of beginn ers in English for examp le, commands are
given such as 'Sta nd up', 'Sit down', 'Open the door', 'Walk to th e ta ble', ' Po int
to the table', ' Poi nt to the door ', 'Where is the table?', 'Where is th e book?',
etc, Soo n after, someti mes even with in the same dass hou r, statements o r
questions are paired with commands: 'This is a book. Give the book to Susie',
'T he book is o n th e table. Put the boo k on the cha ir', 'Who has the boo k?
You? All right. Give th e boo k to Arme', 'Where is the ball? O n the table? All
righ t. Tony, bring me th e ball'. After the proper groundwork has been laid,
students are presented with more co mplex sentences, like 'Give th e book to
Bob and give the pen to j ean', 'Walk to the tablc and then turn around', 'T ake
the yellow card and place it under th e boo k', 'If yo u have a blue card th en
raise your hand ', ' If you have the big card th en place it un der th e sma ll car d' ,
From the beginn ing th e stu dent is introduced to whole sentences in con-
rexr. T he teacher demon strates the mea ning of th e words and sentences by
pointi ng to the ob jects and by acting on th e com man ds for all to see. It is
claimed that with this meth od a student can easily learn around 25 new lexical
item s in an hour, along with a variety of struc rures , We believe th is to be
tru e. In fact, with rega rd to vocab ulary, the number co uld be mu ch higher.

A demonstration project:] apanese students learn German

Japanese srude nts in a psychol ingu istics d ass in J apan , which was taught
by the firsr author of this book, were given a T P R dcm onstrati o n lesson in
German by a colleague . T he studcnts, who had no t learn ed Germa n befo re,
learn ed to understand mor e than 50 different words as weil as a variety of
imp era tive sen tences ('Stand up', 'Turn aro und', 'Open the door', 'Close the
doo r', 'Give the ball to Karen and give the book to Emi l') in just a little over
an hour. Wh ile hesitan t in th eir action at first, th e students soon gained in
confide nce, performing th eir tasks swiftly and with assura nce . Such behav-
iour is a direct measure of their progress in speech com pre hens ion . Interest-
ingly, whe n a videotape of this TPR German lesson was shown to J apanese
stu dents in o ther d asses, th ey learn ed about the same num ber of items.
T hey did not per form any act ions but simply observed what was happ enin g
on th e tape. Observation was sufficient for learn ing . Wheth er th e stu de nts
who performed the actio ns retained more over time th an th e students who
simply observed th e actio ns was not measur ed, unfornmately. T P R wou ld
predict th at doing the action would solidify memory.
208 PSYCHO LINGUIS TICS

Advancing witb T PR
Afte r the teacher has determined tha t the srude nts are firm in un derstandi ng
what they have learn ed, they are th en enco uraged to speak. T hey are asked
to give command s to th eir c1assmates with th eir c1assmates performing th e
actio ns. Ga mes can be devised to encourage speaking.
T P R has essentially the same advantages and limitation s as the D irec t
Method. Srudents do learn to com mun icate in speec h in a natural way and
also relatively qu ickly. In order for this to happen , however, they must have
fluent and crea tive teac hers. Nowadays, th ough, perhaps th e teacher need
not be especially crea tive since a great dea l of curricu lum material has been
develo ped and pu blished for T P R instructi on.
T PR is best used for the int roductory phases of second- language learn -
ing . W ith mo re advanced langua ge know ledg e, actions become less useful
and relevant to communication. T hen, too, there is the probl ern of horne-
wor k. O nce out of th e c1assroom, th ere is nothing a student can do to review
or gain know ledge. In this rega rd, adopting the G rammar-T ranslation meth od
along with T PR wo uld be one goo d solution.

Cbildren vs. adults

One problem with T PR relates to its special reliance on act ion ('Physical
Response'). Fo r social reasons, many adu lts, more so than childre n, feel
em barrassed marching aro un d a room do ing things. While th e required
action couId be modified to lessen this probl em, there is not mu ch else a
teac her can do to remedy th is siruation, Adults may become rnore accepting
in time, th ou gh, especially after th ey see th eir teacher doing th e same thin gs
th at th ey are obliged to do .
T P R is best viewed as a teaching tech nique which can be applied in
beginning to interme diate c1asses. It works especi ally weIl with children , and
with adu lts it may be best util ized in combi na tio n with other meth ods. TPR
sho uld no t be viewed as a self-contai ned method app licable to all language-
teaching co ntexts. With such flexibility, it may weil be considered the best
of th e speec h-based teach ing meth ods.

7.5.2. Comm unicative Language Teaching

In the early 1970s, W ilkins (1972) proposed a system of dividing comm unicat-
ive speech into two aspects : functions and notion s. Functions are things like
requ ests, denials, complaints, excuses, etc, (They are called Speec h Acts in
lingu istics.) T hey are expressed th rou gh who le sentences. Essentia lly th e
learn er is provided with a mea ns for performing a given function. Fo r ex-
am ple, learners may be told that there are vario us ways to make a req uest:
they may be told 'S hut th e window', 'Please shut the windo w', 'Would you
shut th e window?', 'Would you rnind shutting the window?', 'W ill you be so
kind as to shut th e window? ', etc. (Wilkins, 1976, p. 5I).
 SECO ND- LANGUAGE TEACHI N G M ETHO D S 209

Notions are expressions of frequency, quantity, Iocation , etc, These are

typically words or ph rases within a sente nce . For example, students may
learn 'I often go to the movies', 'I have a lot of friend s', and 'He's standing
by the tuindoui',
Communicative Language Teaching (C LT ) presumes that students wanr
to comrnunicate and it enables the m to do just that. Lessons often start with
the simultaneo us reading and hearin g of a dialogue based on a real-life
everyday situatio n, such as greeting a friend or buying something in a shop.
Initi ally, there is no translation and no explanation of the structures involved,
although the method does not exclude native language aids if rhar is what
th e stu dents fee! th ey need for a par ticular point, There is to tal re!iance on
situa tio ns and the students' desire to communicate with in those situa tions.
Since this kind of teachin g stresses comm unicatio n, it has developed a
f1exibility which allows anything into the classro om so lon g as it will furth er
th e communicative ability of the student. This can include translation and
gra mmati cal explanatio ns in the native language, if the teacher believes that
this will be beneficial. And, if a teacher feeIs th at an Audio lingua l technique
such as drilling a phrase a number of times might help a srudent, th en this is
done , so long as th at phrase is later used in a mea ningful siruation ,
O ften, th ere are phrases or sentences which the stu dent has started to
create , but is having tr oubl e with . For example, if a stu dent would like to say
in Eng lish someth ing like 'I wish I could have gone' but can get out only
'I wish . . .', the teacher might mode! th e whole sentence a few times, let
the srudent repeat it a few times, and then retu rn to th e situatio n in which
the stu dent was trying to use it and let him or her use it. (There is some
similarity here to the counselling role suggested by Cur ran in Commun ity
Language Learnin g.)
Later, th ere migh t even be an explanatio n of the gramm ar involved, or
even a struc tu re drill, such as letti ng th e student substitute other past par-
ticiples in a sente nce : 'I wish I could have eaten it', 'I wish I could have done
it', 'I wish I could have seen it' , H owever , such techniques are only employed
in the interest of assisting th e srudents to communicate thei r ideas.
In comparing Co mrnunicati ve Language T eaching with strictly speech-
or iente d methods such as the Direct M eth od , Total Ph ysical Response, and
th e Natural Appro ach (to follow), we can see that there are marked differ-
ence s. C LT permi ts reading and writing almost imrn ediarely, as long as it
serves the cause of cornmunication, It also permi ts grammatical explana-
tio ns, not relying totallyon th e student learning by induction . Fur the rmo re,
it permits translation.
G iven th e above, it would appear thar C LT is not so much a particular
meth od as an edectic meth od which borrows, as it does, aspects of other
meth ods, such as G ramma r-Translation, Audiolingua l, and T PR. T he con-
cern of CLT 's advocates is to get peopl e to communicate by any means
possible. It is probably because of its eclecticism th at C LT has becom e one
210 PSYCHO LI N GUI STICS

of the most widespread of teaching methods in use today. T his is especially

so in the Unired Kingdom where so many of its originators and developers
have been active (W ilkins, 1976; A1exander, 1978; Wid dowson, 1978; Brumfit
& j ohnson, 1979; Yalden, 1983). However, to our knowledge , research srud-
ies which dernonstrate the effectiveness of the method are not yet available.

7.5.3. The Natural Approach

Speed) understanding to precede production

The Na tura l Approach ( A) is the name given by Terrell (1977,1982) and
Krashen (Krashen & Terrell, 1983) to their ' new philoso phy of language
teaching' develop ed in the early 19805. It is to be distinguished from the
nineteenth -cen tury Natural Method, although NA has a numbe r of similar-
ities with that and with oth er natura l speech- based meth ods such as the
Direct Me thod and TP R. (Really, not so 'n ew' after all.) Yet, perh aps the
Natural Approach is more of an attc mpt to provide a theoretical description
of the processes involved in second-language acquisition than it is a body of
innovative techniques for teaching.
In accord with the Natural Me thod , Direct Method, and T PR, the
importa nce of listening cornprehension and delayed speech production is
stressed in the Natural Approach. Pro duction is delayed unti l the student is
believed to be ready. T he idea that you can only effectively pro duce speech
that you already understand is in keeping with the unders tandi ng-p recedes-
production aspect of native-langnage acquisition.

Graded materials and syutax by induction

As for grammatical structures and rules, these are seldom explained and are
expected to be acquired by receiving appropri ate language input. In this
respect, sentences are presented in a simple-to-co rnplex grading and at a
level that may be slightly higher than students can understand . T his is very
similar to the Direct Me thod and TP R.
NA define s itself as a method for developing basic perso nal communica-
tive skills, oral and written. Go als of rhe method would include the ability to
engage in simple conversational exchanges, to unders tand announcernents
in public places, to read newspapers, write personalletters, etc. Like most
other speech-based merhods, teachers of the Natural Approac h make ample
use of picrure s, objects, charts, and situat ions in the classroom as the source
of language inp ut ,

Tbe affictivefilter
Such personal learning factors as motivation, self-confidence, and anxiety
are given special considerati on in NA . T hese constitute what Krashen calls
the learner's 'Affective Filter' and play a significant role in inß uencing the
 SECO ND-LANGUAG E TEACH ING METHODS 21 1

acquisition /lea rning of a language. A 'Iow' condition of the Affective Filte r is

said to be rnost desirable, for in such a case students would be highly motiv-
atcd, very confident, and under Iittle stress. Such dcsirable cond itions can be
fostered if, for exarnple, students are allowed to corn mun icate in sirua tio ns
witho ut having to wor ry abo ut any gra mmat ical misrakes they may m ake.
O n th e othe r hand , a 'high ' condition of the Affective Filter wou!d have the
oppos ite effect, effectivcly blocking any learning thro ugh too much anxiery
and low moti vation,
H owever, the re are othe r views which hold that Affective Filter facrors
such as anxiety can acrually be useful for language learning, Scovel (1978)
conte nds that a student may have 'facilitating' anxiety, which pushes th ern to
greater efforts, rather than 'de bilitating' anxiery which is rhe type of anxiety
tha t the 'Affective Filter' is most conce rne d with. T hen , too , according to
Skehan (1989), high motivation may be a result of language learning under
con dit io ns of little stress as weil as a cause of th at learning as in the premise
of th e Affective Filter. Fo r examp le, a learner with high motivation who
succeeds in learning would, as a result, pro bably have the ir mo tivation
increase even more. T he opposite would be true for a learner with initial low
moti vation: th e low mo tivation wou ld lead to failure in learning, decreasing
mot ivatio n for furth er learning, It is thus not clear to what extenr motivation
is a cause (as Kr ashen and T errell posit) or an effect of langu age-learning
success. While it is probably th e case that stu dents Iearn better when th ey
are motivated, not over -anxious, and when th ey feel relaxed and receive
enco uragement for their efforts , to label this an 'Affective Filter' is rather
pret enti ous, but harm less, academic jargo n.
It is od d tho ugh thar Terrell and Krashen , who are such advocates of
simulating a child's natu ral language acquisitio n, sho uld posit an Affect ive
Filter ar all, since even children who are raised in anxiety-ridden hornes
learn th eir native language, O nly in extreme cases, muc h more extre me tha n
th e anxiety of the ordinary classroom siruation experienced by stu dents,
would langu age learning fail to oecur.

Tbe Monitor Hypothesis: tbe acquisition- learning distinction

T he Na tura l Approac h differentiates between acquiring and learning a sec-
ond langnage (Kr ashen, 1982). Acquisition is said to involve a kind of induc t-
ive process similar to that which OCCutS in th e acquisition o f the native
langn age. Such a process is claim ed to be auto matic and unconseious. Learn-
ing, on th e other hand , is said to involve a formal process by whieh on e
conseiously learns rules such as those raught by a teacher, According to
Krashen, language knowledge which is 'learned' never beeo mes un con seious
or auto matic as does know ledge which is 'ac quired', ( We would like the
read cr to note that we make no such distinction, and th e terms 'Iea rning'
and 'acquisition ' are used interchangeably through out this book.)
212 PSYCH OLIN GUI STICS

T he distinction is based on Krashen's so-called Monitor I-Iypoth esis. Ac-

cording to the hypothesis, 'learned' mies are always monitored, i.e, consciously
applied in the prod uctio n of sentences . No such 'mo nitoring' of speech
production, however, is said to occur with a gramma r that has been 'acquired'.
It is bccause of the monitoring process that Krashen claims that once
stude nts 'learn' gram mar (instead of 'acquiring' it) they will be unable to use
it unconsciously, and thu s effortlessly, in pro ductio n.
T he ' Iearned' system can only be applied under certain conditions adequate
for 'mo nitoring'. Only when learners have adequate time, are focused on
gra mmatica l form, and know th e rule of th e gram mar can they prod uce
speech using what they have 'learned ', For exarnple, the 'Iearned' system
might be used during a language test or during writing. T he consequence of
the limited application of th e 'learned' system is that the teaching of gram -
mar mies by explication is frowned on by advocates of thc Na tura l Approach.

Criticism 0/tbe Monitor Hypothesis

T he Mo nito r Hypothesis has been subjected to severe CrJ tlCISm by many
theo rists, such as Gregg (1984), McLaugh lin (1978), Bialystok (1979, 1981)
and Steinbe rg (1993). Krashcn has not really answered his critics, nor has
he provided convincing evidence in support of his claim that knowledge
gaincd from the presentation of m ies and their explanations cannot become
unconscious and autornatic (Krashen & Scarcella, 1978; Krashen & Pon,
1975). T he validity of th c Mo nitor H ypoth esis, th erefore, is very much in
doubt.

Tbe distinetion is counterintuitiue

An alternative to th e Mo nitor I-Iypoth esis is provided by a model of learn ing
proposed by Bialystok (1979, 1981) in which consciously learn ed language
can become automatic and unconscious through practice. Certainly, Krashen 's
claim is counte rintuitive to what many people experience when they produ ce
sente nces in a second language. Fo r example, according to Krashen, Eng-
lish spcakers who are told in the ir first ] apanese lesson that ] apanese has a
Subject + Object + Ver b erdering would continue to consciously monitor this
orderi ng even after six weeks or more. T his was certa inly not th e experience
of the two non-]apanese authors of this book, Initial awareness disappears
rather quickly. Of course, there are times when secon d-Ianguage learners do
become aware of applying certai n grammatical rules in the construction of
sente nces. H owever, this typically occurs only in the early stages when the
learner has not yet integrated that knowledge weil eno ugh.

Monitaring atuareness also occurs uiitb 'acquisition'

Acrually, monitoring auiareness may occur euen tuhen m ies are acquired by induc-
tion. Language problems are forever occurri ng to the second -Ianguage learner
 SECO N D- LANGUA GE TEACHIN G METH O DS 2 13

and such prob lems will ofren be considered consciously. For example, an
English learn er ofJ apan ese could have figured out on his or her own what a
certain verb ending in J apanese rneans and the n properly use that ending.
This might prompt the learn er to consciously think mo re abo ut other verh
endings and how they relate to one ano the r. Such 'monito ring' will aid, not
hind er, acqu isition .

Aritbmetic and tbe acquisition- learn ing distinction

Let us approach the adequacy of Kr ashen' s acquisition- learn ing distinc-
tion from another point of view, one which includ es an area of knowledge
different from but still quite relevant to language, that is, ar ithmet ic, Sup-
pose we ask you now to divide 954 hy 6, and to do it as quickly as possible.
(You can do it on paper or in your head.) Do you have the answer? We
will wait.
Now, were you conscious of every step of th e process whereby you came
up with this answer? Let us ask you, if, according to one common method
of division that we shall use, YOll wer e aware that your first step was to begin
by consideri ng the single leftmos t (not rightmost) digit of 954, the nu mber
to be divided? Thar is, 9.
Then, were you aware of decidin g thar, since 6 is equal to or less than 9,
you must sub tract 6 and so have 3 remaining? W ere you aware th at because
the rema ind er was less than 6 you would write a 1 for the beginning of YOllt
answer? W hat did you the n do with that 3? W ere you co nscious of having
to place it in fron t of the next leftmost sing le digit of 954, which is 5, and
then tre at the two digits of 3 and 5 as 35?
Next you divided 35 by 6 and got an answer of 5. Bur how did you do
tha t? W ere you co nscious of dipping into the multiplication table (I x 1 = I ,
1 x 2 = 2, 1 x 3 = 3 . . . 6 x 4 = 24, 6 x 5 = 30, 6 x 6 = 36, etc.), which YOll had
me morized years ago in elementary school?
You needed th at knowledge to determ ine that the product of 30, which is
produced by 6 x 5, will bring you closesr to 35 witho ut exceeding it; 24
wou ld not be as close as 30, and 36 would exceed that number, and th erefore
you selected 5 as an answer and you then placed tha t 5 to the rightmost of
YOllt answer of I. You wou ld then have 15 and be on your way to cornplet-
ing the answer.
Were you aware of all the ste ps tha r would bring you to the answer
of 159? Not likely! Yet, all of these steps were taught to you explicitly
in the c1assroom, and, th rou gh your teachers' explanation of the pro cess,
you 'learned' the pro cess. Now, certainly, while the initial learn ing was
formal and presented through explicatio n, nevertheless, th rough time
and pra ctice, the ru les of the pro cess became largely unconscious and
automatic.
Kr ashen's claim, therefore, that learn ing never becom es uncon scious and
auto matie is one thar cannot be up held with rega rd to arithmetic knowledge.
2 14 PSYCHOLINGUISTICS

This bein g so, th ere is no reason to believe th at a special case should be

mad e for one particular kind of knowledge, langu age kno wledge. C learly,
mu ch of th e learning tha t is gained in a formal situatio n can becom e un con -
scious and auto mati c. Since Krashen 's acquisitio n-Iearn ing distinctio n is
not a valid o ne, th ere is no good reason to suppose th at teaching gra mmar
by explication in a second-la nguage situa tio n can no t, at tim es, be bene fi-
cial. Such grammar knowledge may late r becom e uncon scious and automat-
ically used.

Ru/es can be tflught

T he re is th crefore no sound reason to ban th e formal teac hing of all rules in
th e classroom . Simple ru les can be taught direc tly and the learner can inter-
nalize the rn so th at th ey can be used later in an automatic and un con scious
mann er. For example, suppose j apanese learn ers of Eng lish are told thar
Eng lish requires: a Sub jecr-Verb-O bject orde ring, a plura l mark er on noun s
when more th an one count able objec t is involved, thar verb forms of 'be'
cha nge with person , or that preposition s are placed before nouns (and not
after nouns, as in j apanese). Surel y thi s will not hampe r acqui sition of th ese
grammar points, Of course, if a reneher mainly spends tim e on explaining
rules, the results will be as T errell and Krash en predict,
T hen, to o, the re are some very goo d ways to teach more compl ex rules
tha n by tr aditional stateme nt and example. T he stu dent can be presented
with data and given a chance to discover th e ru le on his or her own . These
are what Ellis (1994) refers to as 'co nscio usness raising' tasks. After the
stu dcnts have had an opportu nity to figu re o ut the ru le for th emselves, the
teach er can th en make certain that everyone understands the rule und er
co nsideration. The sensible teacher will strike a balanc e with dir ect speech
experience, consciousness- raising tasks, and rule explication.

7.6. Same Research Studies Camparing Effectiveness

of Methads

T here are many probl ems in doing research compari ng methods. In part icu-
lar, th e teachers using Metho d 1 must be of equal proficiency, be equally
desirou s of success, be equally attractive, etc, as those teac he rs of M ethod 2.
This is not easy to achieve, but it is possible. T hen , materi als presented to
th e srudenrs rnust be in some way compa rab le. This is alrnos t imp ossible to
achieve. Then , too, the students given Metho d 1 should be equal in all
respects to th e srudents given M cth od 2. T his is easier than having to deal
with the othe r rwo variables. T estin g ret ention over tim e is another relevant
variable to be considered.
 SECO N D- LANGUAGE TEA CHI NG M ETHO D S 2 15

7.6.1. Grammar-Translation and Audiolingual Compared

In a comparison of the G rammar-T ranslation method with the Audiolingual
metho d, Scherer and W ertheimer (1964) foun d that GT produ ced higher
scores in reading and writing while test scores in speaking and listenin g were
highest for AL. That is, unsurprisingly, higher scores were found on the
factors which the method emphasized.
Other srudies have shown sirn ilar results. In a notable srudy known as
the Penn sylvania Project, Smith (1970) compared three groups of srudents
on three method s: essentially the G rammar-T ranslation meth od , the Audio-
lingual rn ethod, and a combination of the two. On testing the stude nts on all
four skills of reading, writing, listening, and speaking, students receiving GT
had a better reading ability altho ugh they also had lower speaking scores.
N o other differences were found. Again we see that the focus of a method
will produce higher scores for the meth od's specific emphasis.
T he above facts are import ant too when one considers that most language
testing consists of written exarninations rather than oral exams. This is often
the case because of the ease with which written exams can he administered to
large groups of srudents and the great expense in time of testing the spoken
language. I-Iowever , such testing favours students who have learned by GT.

7.6.2. Total Physical Response and Audiolingual Compared

O ther meth od comparisons have been made by Asher (Asher et al., 1974),
comp aring Total Physical Response and the Audiolingua l Method. T he
findings in this sho rt-term srudy dernonstrated a superiority for T PR for
beginning srudents. Whether T PR could maintain this edge over AL and
other met hods with intermediate and advanced srudents, and over a Ionger
period of time , has yet to be demonstra ted.

7.6_3 . Natural Approach and Grammar-Tra nslation Compared

In a stu dy comparing the Natura l Approach with G rammar-T ranslation
using Spanish learners, Hammond (1988) found that srudents studying
under the N atura l Approac h scored only slightly high er than those under
Gra mmar-T ranslation . In terms of grammar learn ing, NA did as weil as
GT: thu s, even witho ut explicit grammar teaching, NA srudents learned by
indu ction as weil as did the GT srude nts. T hus, srudents can learn gram mar
thro ugh induction just as weil as they can thro ugh explication . H owever, as
to the use of grammatical structures in actual communication, there would
likely be an addi tional benefit of learning for NA learners since teaching
through induction usually pro vides the learner with a communicative ability
that is often superior to that of srudents taught through GT. Further srudy
might weil confirm this supposition.
2 16 PSYCH OLI NGUI STICS

7.7. Goals Must Be Considered In the Selection

of a Method

It is safe to say th at srudents will learn some thing from any method. No
method is a tota l failure because, in all methods, students are expose d to
th e data of a seco nd language and are given th e opportu nity to learn the
langnage. H owever , to the disappointment of all, there is no magic meth od.
No method has yet been devised that will permi t people over the age of
12 or so to learn a second language as effortlessly as th ey did their native
language. Still, tea chers can do mu ch to make the exper ience for a learn er
rewa rding and enjoyable, whatever method is employed.
In judging th e relative merits of teaching meth ods, one must consider
goals. Just what is the purpose of having people learn a second language?
If th e ability to speak and un derstand a second language is th e primary goal,
th en a spee ch-based method wou ld be best for them. If, on th e other hand,
the ability to read and write is th e primary goa l, th en Grammar-Translatio n
should be th e meth od of choice .
T he goa ls of a nation are imp ortant in determining second- Ianguage
teaching programm es in the schoo l system. O ne country may wish to pro mote
the stu dy of reading and tra nslation of scientifi c mater ial from a second
language, and would, th erefore, wish to stress th e knowledge tha t is gained
through readi ng. In such a case, the Grammar-Translatio n Method may
weil be appropriate. Other countries, however, may regard communication
through speech as the highest priori ty. As such, speech-based meth ods may
be preferred, providin g, of course, th at adequate finances are available for
th e specia lized training of teache rs in such methods and th ar th e school
system can afford teac hing classes with small numbers of stu dents. When
large numb ers of stu dents are to be taught and few teachers are available,
G rammar-T ranslation might weil be chos en by default, since, practically
speaking, no othe r cho ice is viable.
A teach er who can afford th e luxury of selecting a method might weil
consider putti ng tog ethe r a person al meth od of second-language teaching.
Fo r example, with both speech and literacy as objectives, one cou ld adopt
Co mmunicative Language Teaching and th en suppleme nt it with physical
activities (from Total Ph ysical Respon se), pattern practice drills (fro m th e
Audiolingu al Metho d), and explication and translation (from th e Grammar -
T ranslation meth od), Most methods will have some fearure which can be of
benefit to the language learner.
 SECOND-L ANGUAGE TEACH ING METHO DS 217

Notes

I. Wh ile Japan is not an underdcveloped country , its rnethods of teaching

instruction in rhe public schools are still largely focused on G rammar-
Translation. At the starr of the twentieth cenrury, the traditional method
of teaching a second language had been based on yaklldokll (translation
reading). This method was developed in Japan as rneans of understanding
Chinese classical texts and ent ails the translation of word s, their reo rdering,
and final formulation into correct j apanese (Law, 1995; H ino, 1988).
(T he Japanese regarded Chinese classics the way that Euro peans regarded
Latin and Greek classics.) As we enter the rwenry-first century, Gra mmar-
Translation is still weil in contro!.
2. The first author recalls (with horror !) the first English 'conversation' class
which he taught in Japan. He walked into a class with over 60 students in it!