DOCUMENT RESUME

AL 002 764

ED, 048 579

AUTHOR
T1ZTLE

PUB DATE
NOTE

EDRS PRICE
DESCRIPTORS

Richards, Jack
Error Analysis and Second Language Strategies.
24 Feb 71
27p.

EDRS Price MF-$0.65 HC-$3.29

Child Language, *English (Second
Patterns, Interference (Language
Theories, Linguistic Competence,
Performance, *Psycholinguistics,
Learning

Language), *Error
Learning), Learning
Linguistic
*Second Language

ABSTRACT

Discussed in this paper are reasons why people who

speak second languages may not speak or write them with
native-speaker-like fluency. These second-language deficiencies may
be the results of (1) interference, the use of aspects of another
language at a variety of levels; (2) strategies of learning such as
over overgeneralization and analogy by means of which the learner
tests out his hypotheses about the structure of the language; (3)
strategies of assimilation, in which the learner makes his learning
task easier; and (4) strategies of communication, whereby the learner
adapts what he knows into an efficient communication model, producing
an optimal utility grammar from what he knows of the language. In
addition to understanding these deficiencies, it is necessary to
distinguish between performance and competence errors. The former are
occasional and haphazard and are related to such factors as fatigue
and memory limitations. The latter are systematic and may represent
either a transitional stage in the development of a grammatical rule
or the final stage of the speaker's knowledge. While knowledge about
second language learning is still largely speculative, excluding the
possibility of prescribing recipes for teachers, this account of
errors and learning strategies may suggest some of the reasons for
errors. (Author/AMM)

4 sl

ERROR ANALYSIS AND SECOND LANGUAGE STRATEGIES

U.S. DEPARTMENT OF HEALTH, EDUCATION

& WELFARE
OFFICE OF EDUCATION
THIS DOCUMENT HAS BEEN REPRODUCED
EXACTLY AS RECEIVED FROM THE PERSON OR

ORGANIZATION ORIGINATING IT. POINTS OF

VIEW OR OPINIONS STATED DO NOT NECESSARILY REPRESENT OFFICIAL OFFICE OF EDUCATION POSITION OR POLICY.

JACK RICHARDS
Department of Linguistics

University Laval, Quebec, Canada

The text of an invited lecture given at

Indiana University, Bloomington
February 24, 1971

i.

TABLE OF CONTENTS

INTRODUCTION

1.

"Errors" in First Language Learning

2.

The Significance of Errors in Second Language Learning

3.

Interference

4.

Overgeneralization

5.

Performance Errors

6.

Markers of Transtional Competence

7.

Other Forms of Interference

8.

Strategies of Communication and Assimilation

9.

Errors and the Teacher

CONCLUSIONS
REFERENCES

ti

ERROR ANALYSIS AND SECOND LANGUAGE STRATEGIES

INTRODUCTION

The field of error analysis may be defined as dealing with the

differences between the way people learning a language speak, and the
way adult native speakers of the language use the language. Such
differences may create interest for a variety of reasons. We may begin
with the interests of those who study language "for its own sake."
Since language is not simply a more complex instance of something found
elsewhere in the animal world, Chomsky suggests that the study of human
language is the most fruitful way of discovering what constitutes human
intelligence.
Some of the most insightful notions about what language
is have come from observing how language is acquired by children. By
looking at children's speech, comparing it with adult speech, and trying
to account for the differences, psycholinguists have been able to
speculate about the nature of the mental processes that seem to be
involved in language.
While the mother-child relationship and other language crucial
activities of the child's experience do not include any consciously
incorporated instructional strategies, in the field of second language
teaching elaborate instructional procedures are often defended as being
essential components in successful second language learning. Since the
goal of the language course is to lead the learner towards adult uses of
the new language, differences between the way the learner and the native
speaker speak the language have been studied in the hope that methods of
overcoming these difficulties might be devised. Errors in second
language learning, it is sometimes said, could be avoided if we were to
make a comparison of the learner's mother tongue and the target language. The sum of the differences would constitute his learning difficulties, and it is here that teaching strategies would be optimal.
Alternatively, interpretation of errors in second language learning
along the lines of errors in first language learning suggests that
second language errors are not, by nature, different from those made by
children learning English as a mother tongue, hence they should not be
of undue concern to language teachers.
The purpose of the present paper
is to look at recent and less recent ideas on errors in second language
learning, in the hope that such an examination might illuminate the
experience of second language teachers.

1.

"ERRORS" IN FIRST LANGUAGE LEARNING

What sort of ideas about language and language learning have been
deduced from differences between children's and adult speech? Firstly,
considerable support for current notions of language have been found in
studies of child language development.
It used to be thought that
speaking was. simply the exercise of our individual verbal habits, and
that these were acquired through repetition, reinforcement, and conditioning, in much the same way as animals can be trained to perform
certain tasks through the use of appropriate conditioning techniques.
It is suggested that this is an inadequate account of language and of
language learning. In children, it appears that the process of formulating language is an active and creative process, yet a process which
follows similar patterns in children across quite differing learning
circumstances. All children learning English as a mother tongue, seem
to follow a similar sequence in their acquisition of grammar for instance.
If we listen to the speech of English speaking children at
about three or four years old for example, we hear them using question
forms like this:'
What he can ride in?
Where I should put it?
Why he is doing it?
These questions share a common structural feature which makes them
different from adult questions.
In these questions, that part of the
sentence which normally comes aft -r the subject in the statement form
- the can in He can do it, - has been left in this position in the question form, instead of being put before the subject as in the adult
sentence What can he do? Now at about the same time as children are
producing sentences like this, they are able to make questions which
don't require a wh word such as where or why.
They have no difficulty
in saying Can he ride in a truck? but when they use a wh word they fail
to change the word order and produce What he can ride in?
It is clear from instances like this in children's language that
the children are not simply imitating the speech of their parents, for
sentences like this do not appear in adult speech. If we were to compare the sentences which the child is capable of producing at this
stage, with those of an adult, we would have an illustration of the
child's competence at age 3 or 4 compared with adult competence.
Child's competence

Adult competence

He possesses the rules

permitting questions
with the form:

He possesses the rules

permitting questions
with the form:

- 2 -

Can he go?
Where he can go?

Can he go?
Where can he go?

He cannot produce:

Where can he go?

By looking at the differences between the child's sentences and those

of the adult we see evidence of the way the child appears to be formulating hypothesis about the English language, some of which he will
eventually abandon in favor of the rules of adult language.
Since the
differences between children and adult speech are systematic, found
wherever children learn English, psycholinguists have been able to postulate universals that seem to be crucial for language development.2
Now the crucial elements in first language acquisition would seem
center on the psychology of learning, that is, those strategies
employed by the child as he teaches himself his mother tongue, the development of his other faculties such as intelligence, cognition, perception and so on, and the structure and rules of the particular language he is acquiring, in this case, English. These would appear to
shape and formulate the sentences he produces in a systematic way. It
has been suggested that the differences between the way a second language is often spoken and the way the language is spoken by native
speakers are systematic, just as children's language follows a definite
norm and developmental sequence of its own, as we saw with the example
of the use of questions.
As evidence that second language learning is
not entirely haphazard, Table 1 includes a sample of typical errors in
the English verbal group made by people learning English as a second
language.3 These represent common errors made by people with quite
different mother tongues.
They cannot be dismissed as occasional, accidental, or as mere performance slips. Corder believes that such errors
in second language speech reveal a systematic attempt to deal with the
data, and that they should play the same role in our study of second
language learning as differences between child and adult speech play in
the study of first language acquisition.
"It is in such an investigation that the study of learner's errors (in second language learning)
would assume the role it already plays in the study of child language
acquisition, since the key concept... is that the learner is using a
definite system of language at every point in his development, although
it is not the adult system in the one case, nor that of the second
language in the other. The learner's errors are evidence of this system
and are themselves systematic".4
to

- 3 -

TABLE

TYPICAL ERRORS IN ENGLISH AS A SECOND LANGUAGE: THE VERB GROUP 3

(a) be + verb stem for verb stem

We are live in this hut.
The sentence is occurs...
We are hope...
He is speaks French.
We are walk to school every day.
(b) be + verb stem 4- ed for verb stem -I- ed

Farmers are went to their houses.

One day it was happened.
The teacher was told us.

-7 Wrong form after do

He
He
He
He

did not found...

did not agreed...
did not asks me.
does not has...

(d) Wrong form after modal verb

Can be regard as...
We can took him out.
They can used it.
She cannot to go.
We must made...
We can to see..,
e

be omitted before verb 4-stem 4- ed

He born in England.
He disgusted.

be omitted after be 4- sarticigle verb stem

artici le

The sky is cover with clouds.

He was punish.
1[8

be omitted be ore verb 4-in.

They running very fast.
The'cows also crying.
At 10.30 he going to kill the sheep.

NO verb stem for stem 4-'s

......

He always take a lot.

He come from India.

6
- 4 -

THE SIGNIFICANCE OF ERRORS IN SECOND LANGUAGE LEARNING

2.

If learner's errors in second language acquisition are systematic,

in what ways are they organized, and what do they suggest about the
nature of second language acquisition? Selinker calls the speech output
in a second language an Interlanguage since it invariably differs from
the target language, and he uses the term fossilization to refer to
permanent characteristics of the speech of bilinguals irrespective of the
age at which the second language is acquired or the amount of instruction
or practice in it. He characterizes fossilization in the following way:
"it is my contention that the most interesting phenomena in interlanguage
performance are those items, rules and subsystems which are fossilizable
If it can be experimentally demonstrated that fossilizable items,
rules and subsystems which occur in interlanguage performance are a
result of the native language then we are dealing with the process of
language transfer; if these fossilizable items, rules and subsystems are
a result of identifiable items in training procedures, then we are dealing with transfer-of training; if they are a result of an identifiable
approach by the learner to the material to be learned, then we are dealing
with strategies of learning; if they are the result of an identifiable
approach by the learner to communication with native speakers of the
target language, then we are dealing with strategies of communication;
and finally if they are the result of a clear overgeneralization of target
language rules then we are dealing with the reorganization of linguistic
materials.
I would like to hypothesize that these five processes are
central processes in second language learning and that each process forces
fossilizable material upon surface Interlanguage utterances, controlling
to a very large extent the shape of these utterances".5
.

I should like to focus on Selinker's description of Interlanguage

characteristics as a basis for an account of typical errors in second
language communication in English. As data I will begin with an analysis
of samples of English speech elicited from two speakers, one whose mother
tongue is European French, the other whose mother tongue is Czech. To
obtain the six speech samples presented here the speakers were given a
number of short texts in English; they were instructed to read the texts,
and then asked to relate the content of each text in their own words
without referring to the texts.
In what ways does their performance of
this task illustrate systematic approaches to second language communication?
SAMPLES OF SECOND LANGUAGE SPEECH
SAMPLE 1

(French)

The fact that land and minerals are very

cheap in inaccessible region' and the
development of new techniques has allowed

- 5 -

1 Interference (plural not

pronounced in French)

to capitalist mane to invest the money3

in this region and exploit the minerals4.
It's
this is occurs5 in Australia
where man has exploited huge mounting6 of
minerals in this region. And they have
the possibility to do great profits7 in
this part of the country.
.

2 Interference (a permis au
capitaliste de
)
.

3 Interference (i'argent)
4 Interference (les mineraux)
5 Overgeneralization
6 Overgeneralization

7 Interference (ont la possibilite de faire de grands

profits)
SAMPLE 2 (Czech)

In the first part of the article the

author give' us reasons for investment
in inaccessible regions. The reason for
to invest in this regions2 is the
possibility to buy3 land, minerals and
deposits at very low cost. Another factor
which permits to invest4 in this region
is the development of new technology
which permits to connected5 this inaccessible regions6 in short time7 with the
regional
the regions which are
civilized.
.

1 Overgeneralization
2 Overgeneralization
3 Overgeneralization
4 Overgeneralization
5 Overgeneralization
6 Overgeneralization
7 Interference

SAMPLE 3 (French)

The human eye may be compared to a

camera. The camera functions with a
lens
would' enregistratel by a
lens and a screen behind it which
enregistrate3J4 the image. Once the
image is developed it stays here, In
contrary5 in movies the image disappear7
from the screen. The eye-human eye
functions like
the, like camera.
It is composed with8 a lens and behind
the lens is little9 screen coated with
cells and enregistratesl the light.
.

1 Performance error
2 Interference (French
borrowing)
3 Interference (French
borrowing)
4 Overgeneralization (Omission of s)

5 Overgeneralization
6 Interference (au contraire)
7 Overgeneralization
8 Overgeneralization
9 Performance error
to Interference (French
borrowing)

8
- 6 -

SAMPLE 4 (Czech)

The author of the article compares

the function of a camera and the
function of the human eye. In an
ordinary camera we have lens' which
concentrate2 beams of light on the
film which is in the back of camera3.
This light can impress the film and
in this way to fix4 the image of the
film.
The function of the human eye
is very similar. We have the same
lens in our eye and the film which
is found in cameras is replaced in
our eye by retina6, an organic matter
which is composed by7 light sensitive

Interference

2 Overgeneralization
3 Interference

4 Performance error
Interference

6 Interference
7 Overgeneralization

cells.

SAMPLE 5 (French)

Steam enginelis composed with2 a

cylinder in which a piston can move
easily and which fix3 well the
cylinder4. This piston is actionated5
by the steam and it is connected to
a wheel which it makes turned6. The
steam
this engine has been
discovered7 in the 18th century, but
James Watt is the person who
ameliorate8 it and who give9 it
his actual form".
.

1 Performance error
2 Overgeneralization
3 Overgeneralization

4 Interference (fixe bien le

cylindre)

5 Interference (actionne)

6 Interference (le fait

tourner)
7 Interference (a ete ddcouvert)

8 Interference (French
borrowing)
9 Performance error
18 Interference (sa forme actuelle)
SAMPLE 6 (Czech)

The article is about the invention

of steam machine'. Steam machine2
was invented in 17th century3 by
an Englishman, James Watt. The
principle of the steam machine is
the basis of the steam machine
.

1 Interference

2 Interference
3 Interference

9
- 7

is piston4 which is pushed by the

pressure of steam 5. This piston is
connected with6 a wheel by a rod
and in this way the motion of the
wheel is possible.
3.

4 Interference
5 Interference

6 Overgeneralization

INTERFERENCE

Perhaps some of the most apparent examples of fossilizable items

in second language communication are those described as instances of
language transfer or interference. This may be defined as the use of
elements from one language while speaking another, and may be found at
the level of pronunciation, morphology, syntax, vocabulary and meaning.
Examples 1, 2, 3, 4 and 7 in Sample 1 for example, reflect the use of
elements of French morphology and syntax.
In Example 1, the plural s
is omitted, perhaps because plurals are not pronounced in French.
A French structure is used in Example 2
has allowed to capitalist man
following the French structure
a permis au capitalists de
In
Example 3, French article usage is reflected in the use of
the money
.

following the French pour investir l'argent . In Example 7 the

influence of French likewise seems evident; have the possibility to do
great profits follows the French
ont la possibility de faire de grands
profits.
These examples of interference might appear to confirm some of the
claims that have sometimes been made for the possibility of predicting
instances of interference by contrasting the grammatical or other systems
of two languages.
"We can predict and describe the patterns that will
cause difficulty in learning and those that will not cause difficulty,
by comparing systematically the language and culture to be learned with
the language and culture of the student".6 Many such contrasts between
languages have been attempted, though they have been criticized because
they make demands on linguistic theory that our present knowledge about
language is simply not ready to meet. We do not know enough about the
higher level organization of particular languages to make the neat sort
of contrasts that such statements imply, either feasible or meaningful.
Most of the contrasts that have been made have been based on practical
knowledge of two languages rather than on any systematic application of
a theory of contrastive analysis.? Yet the instances of interference we
have looked at seem so evident that it might appear that second language
data can be entirely described in such terms.
Indeed it often has been.
What happens in fact is that in analysing second language data it is
tempting to see all errors as effects of the interference of the mother
tongue, ignoring all other relevant phenomena.
Both Samples 1 and 2
provide examples of errors that require alternative interpretations.

10
- 8 -

4.

OVERGENERALIZATION

In Samples 1 and 2 we find examples of a similar type of error,

that illustrated by this is occurs (Example 5 in Sample 1) and the
author give us (Example 1 in Sample 2). Here we have a similar type of
error from both the French and Czech subjects.
These illustrate what
Selinker calls overgeneralization of target language rules, or the
reorganization of linguistic materials. Jakobovits defines generalization
as
the use of previously available strategies in new situations
In
second language learning some of these strategies will prove helpful in
organizing the facts about the second language, but others, perhaps due
to superficial similarities will be misleading and inapplicable".8 In
Sample 1 the French speaker seems to have generalized the form is
occurs from his experience of forms like it is made of and it occurs.
In the Czech example in Sample 2 the omission of the third person s in
the author give us may result from the pressure of other forms in English
without s.
Oilsckova remarks "Since (in English) all grammatical persons
take the same zero verbal ending except for the third person singular in
the present tense
omissions of the s in the third person singular
in the present tense may be accounted for by the heavy pressure of all
the other endingless forms. The endingless form is generalized for all
.

persons ".9

These examples of overgeneralization are the effects of particular

learning strategies on items within the target language, and since such
learning strategies appear to be universally employed when a learner is
exposed to second language data, it is not surprising that many of the
errors found in second language communication are identical despite the
background language of the speaker. What differs are those effects of
language transfer or interference while those that we find in common are
the results of other learping strategies.
One aspect of generalization
has often been referred to in studies of first language acquisition. It
has often been remarked that children learning English as their mother
tongue will produce forms like corned, and goed, by analogy with past
tense formation in regular verbs. Among children acquiring French as
their mother tongue we likewise find things such as on poudra instead of
on pourra (we will be able
) by probable analogy with on voudra."
Likewise both French speaking children and people learning French as a
second language produce sentences like je serai tras malcontent instead
of je serai tres mecontent presumably on the analogy:
.

heureuse /maiheureuse

content /malcontent.

Similar processes seem to account for common preposition mistakes in

English. The learner, encountering a particular preposition with one
type of verb, attempts by analogy to use the same preposition with
similar verbs. He showed me the book leads to He explained me the book.
He said to me gives He asked to me.
We talked about it gives We dis-

9 -

cussed about it. Ask him to do it produces Make him to do it. Go with
him leads to Follow with him and so on. The pressure of one English
construction on another as the learner tests out his hypotheses about
the structure of English may account for forms like permits to invest and
permits to connected in Sample 2 and the other examples of overgeneralation by analogy noted in the other samples.
(Sample 1, Example 6;
Sample 2, Examples 3, 4 and 5; Sample 3, Example 8; Sample 4, Example 7;
Sample 5, Example 2; Sample 6, Example 6).

In Sample 3 we have an interesting demonstration that the French

speaker is not simply transferring the grammar of his mother tongue into
English.
In Example 8 he uses composed with instead of composed of.
The
Czech speaker in Sample 4, Example 7, likewise uses composed by and in
Sample 5, Example 2, the French speaker again produces composed with
instead of composed of.
Had the French speaker followed the grammar of
his mother tongue here he would have produced the correct English form.
In French the equivalent would be compose de which is the English equivalent of composed of. The French equivalent of the form the French
speaker used (compose avec) would in fact be inappropriate in French in
this context.
Both the French and the Czech speaker are evidently trying
to work out the particular rules of English structure, being guided here
not by the grammar of the mother tongue but by what they already know of
English, and by their own intuitions. As Wolfe comments: "Once the
student grasps the idea that the new language differs from his native
language in many matters of structure, he will not know when it is safe
to operate in terms of his native language (it seldom is) and he may try
to create his own structures on the basis of previous contact with the
new language
Some students, not knowing a correct form, will make
up a form which does not parallel either the native or the target
language. Or a student will persistently fail to make a grammatical
distinction in the target language which he actually does make in his
mother tongue".11 We see examples of this from the French speaker, who
fails to use the past tense in Sample 5, Example 9, although this would
be required were he recounting the text in French.
The psychological
parameters in second language learning thus cannot be identified
exclusively with the linguistic ones.
.

The effect of rules within the target language has been described
in more formal terms by Falk in this way: "Few if any of the rules of the
syntactic, component are completely independent of the other rules.
The
formulation of one rule will invariably affect other rules in the grammar .
Because this is so, the construction of a subgrammar, i.e.
of some subset of the rules for a particular language, is a complicated
task. Some of the rules in such a subgrammar will inevitably be ad hoc
since the limited nature of the undertaking excludes detailed consideration of all the linguistic facts which may affect the rules".12
.

12
- 10 -

So far I have talked about interference and overgeneralization as

if they were independent factors. The facts are not quite so consistent.
In Samples 4 and 6 the Czech speaker consistently omits articles, and
this I have attributed to interference, since articles are not present in
his mother tongue.
Dusckova notes however "Although the difficulty in
mastering the use of articles in English is ultimately due to the absence
of this grammatical category in Czech, once the learner starts internalizing their system, interference from all the other terms of the (English)
article system begins to operate as an additional factor".13 This no
doubt leads to diffidence and hesitancy in the use of articles by the
Czech speaker. The interference of other aspects of the English article
system on someone acquiring English as a second language is illustrated
by F.G. French.14 He notes that in English we say The sparrow is a small
Sparrows are small birds. Since the statements are exactly
bird
parallel a logical substitute for the second sentence would be The
sparrows are small birds. In Burmese the equivalents would be:
sa gale thi
the sparrow
and in the

nge than
small

pyit thi

nget
bird

is

plural

sa gale mya thi

the sparrows

nge thaw

nget mya

small

birds

pyit kya thi

are

Instead of following the form of the mother tongue however, the learner
having first produced The sparrows are from The sparrow is sees a
parallel between sparrows and birds and produces the common error, The
sparrows, are the small birds. A similar example is noted by Aguas, from
Tagalog speaking students.15 In her study she found a number of examples
where sentences such as Robot is like a human being were produced instead
of A robot is like a human being. Aguas rules out the possibility of
transfer from Tagalog since this would have required a determiner before
the noun.

Ang robot(ay) parang tao

(N-m)robot
like human being
Here she suggests the analogy is from a somewhat comparable English
structure, such as Robots function like human beings.

5.

PERFORMANCE ERRORS

If language learning both in a first and second language setting

involves trying out hypotheses about the language from one's experience
of it (and in the case of second language learning
from one's
experience of other languages), and abstracting the rules which permit us

13

to produce sentences in the language, then we will need to go further

than looking at simple examples of interference, overgeneralization and
analogy. These processes are in themselves insufficient to account for
the complexity of language learning. The samples we have looked at do
not enable us to say whether these mistakes are occasional, or represent
permanent states in the speaker's competence.
In Sample 3 for example,
the speaker says (Example 9) behind the 'lens is little screen, omitting
the article.
But what does this error really represent? There are
several possibilities.
Firstly we may exclude interference from the
mother tongue, since this would have required the use of the article in
English.
A second possibility is that the speaker realized the mistake
as he said it but forgot to correct it
that he would have corrected it
if he had had more time to think about it.
In fact when looking at the
transcribed text the speaker did correct this mistake himself.
Now this is something that the child who says What he can ride in?
is unable to do. The child's competence is represented by that particular
sentence, whereas this particular article error by the French speaker is
simply the sort of error anyone is likely to make speaking under normal
circumstances.
It is therefore an error at the level of performance
rather than competence.
Performance errors are quite normal aspects of
language use. When we are tired or hurried, we all make errors of this
type.
There is a similar error in Sample 4, Example 4. The Czech
speaker was able to recognize the mistake in This light can impress the
film and in this way to fix the image of the film.
This error was probably a function of the length of the sentence the speaker was trying to
produce, hence it is related to memory limitations rather than to
competence.
Clearly if we want to have a more certain idea of someone's
competence in a second language, we need to take a much closer and more
detailed look at their speech output to find out if their sentences consistently display particular interlanguage characteristics, that is, to
find out if the speaker has internalized a system of some sort that does
not correspond to that of the native speaker.

6.

MARKERS OF TRANSITIONAL COMPETENCE

Just as we need to be able to distinguish between performance and

competence errors in the analysis of second language data, so it may be
ncesssary,to distinguish between those errors which indicate the learning
sequence by which particular grammatical rules are built up, and those
which represent the final state of the speaker's competence.
In our
example of first language learning for example, sentences like What he
can ride in? seem to be produced by all English speaking children before
they are able to use the form of the adult grammar
What can he ride in?
Are some of the errors observed in second language learning also representative of developmental sequences by means of which the learner masters
the rules of the English grammatical system? This question cannot be

14
- I2 -

answered from the data I have presented here, and I know of no studies
which could confirm or reject such a hypothesis.
What is needed are
detailed longitudinal studies of an adult learner's progress with a
second language, documenting the appearance and development of particular
structures.
The types of short term errors attributed here to overgeneralization and analogy could then be placed within the overall sequence of language development.
It may be that the innate ability to
generate and hypothesize rules, so evident in first language acquisition,
becomes subordinate in adult second language learning to secondary
learning strategies, such as generalization, borrowing, and memorization,
for purely biological reasons.
It should also be noted that I am using
terms like interference, analogy, and generalization, without relating
them to a psycholinguistic model of language. They are used here simply
as convenient ways of classifying observable phenomena at the level of
speech, though this does not explain how they are represented at the
level of language. Psycholinguistic models have been proposed which do
try to account for such factors as interference and generalization, such
as that of Jakobovits.18
If we were to try to locate our Czech speaker's article errors at
the level of competence, we would require a close developmental study of
his article usage as he acquires English, to find out if after some
exposure to English, his use of articles was entirely haphazard, or
whether he had worked out some consistent way of dealing with them.
Is
his learning task the same as that of the child faced with the appearance
of articles in his mother tongue?
Jones, in his study of child language development, places the
development of the article system as coming logically and necessarily
after the development of the substantive or noun.'? It is a further
conceptualization of the substantive, permitting a point of view which
conceptualizes it either as a universal or a particular.
Leopold found
that articles were not used at all by his child subjects in the first two
years but later took on systematic usage.18 With an adult acquiring the
system as a second language however, cognitive development has already
occurred, hence he may have to resort to other strategies to develop
rules to deal with the article system in English.

7.

OTHER

FORMS OF INTERFERENCE

Before looking at other aspects of second language learning strategies, I should like to refer to an aspect of interference which is not
manifest in the particular samples we have looked at so far, but which
is nevertheless quite widespread. This has to do with contrasts between
styles across languages.
We may regard style as the choice we have
within a language of a particular mode of expression, such as formal or
informal, and colloquial or offi-ious.
In some speech communities

15
- 13 -

differences of this sort are so marked and have become so institutionalized

that they may be regarded as quite distinct varieties of the same language.
In Bhagdad, according to Ferguson, the Christian Arabs speak a "Christian
Arabic" dialect when talking among themselves but speak the general
Bhagdad dialect "MuslimArabic" when talking in a mixed group.19 This
phenomenon, where two distinct varieties of a language exist side by side
in a community, with each having a definite and distinct role to play, is
what Ferguson called Diglossia. What typically happens is that particular
functions are assigned to each variety of the language, the most fundamental distinction being between what may be called High and Low uses of
the language.
The High form is generally regarded as superior to the Low
form and is compulsory for certain types of situations.
Ferguson gives
the following sample situations, with indications of the appropriate form
of the language to be used:
sermon in church or mosque
instructions to servants, waiters,
clerks, workmen
personal letter
speech in Parliament
university lecture
conversation with family, friends,
colleagues
news broadcast
radio soap opera
newspaper editorial, news story,
caption on picture
caption on political cartoon
poetry
folk literature

H
L
H
H
H
L
H

L
H
L

Now the differ'ences between the High and Low forms in language which
exhibit this phenomenon, and it is very widespread --Arabic, Modern Greek,
Swiss German, and many Asian languages for example
are much greater than
those between what we may call a formal or informal style in English. In
diglossic communities the High style may have striking differences in
grammar and in word order, and in the area of the vocabulary the High
style may have a much more learned and classical lexicon than the Low.

People who belong to language communities which manifest this

phenomenon may come to expect such switching between distinct language
varieties to be a universal feature of languages. In other words, when
they come to write something in English which-would demand a high style
in their mother tongue, they may feel a pressure to give their English
a correspondingly high style. Thus one finds people who write English
clearly and faultlessly, exhibiting what I call diglossamania in certain
situations.
In English for example, we do not feel that there is a low
status about words of high frequency, in the vocabulary we use in normal

16
- 14 -

everyday situations. A Tamil speaking student however may prefer the

feel of I instructed him to obtain it for me over I told him to get it
for me. There may be a distinct preference for archaic words or for
words of low frequency, whereas the native speaker of English prefers a
simple style.
You may have noticed this in written English from Greek
or Arabic students for example, depending on the topic they choose to
write about.

The effects of the High/Low distinction in the student's mother

tongue may also be evident at the grammatical level in his written
English. Kaplan, in an article dealing with composition difficulties of
Arab students, remarks that Arab students appear to prefer certain
devices of conjoining sentences over others 2u, thus the sentences:
The boy was here.
He drank the milk.

may be more frequently conjoined with and by Arab students, producing

The boy was here and he drank the milk in preference to The boy who was
here drank the milk. This may be attributed to the influence of tne
grammar of a high style in Arabic. This use of balancing devices to
create a more aesthetic style produces among certain Indian students
sentences like:
While I am still a student, and yet I still have plenty of time
Whenever I see him, then I feel happy
Even though I am poor, yet I am happy
Here are two examples from a letter-writing manual for students of English
written by an Asian with a diglossic mother tongue. As a model for a
personal letter he uses a distinctly high style:
With the warm and fragrant breath of Spring, here approaches
the bliss of Eastertide. May it shower joy and happiness upon
you aZZ.
The tide brings to my mind the old memories of
several Easters feted with you and the decorated eggs placed
on my dresser.
For an order however, the writer presents a normal low style as a model:

I want to order from you a copy of the latest edition of

book.
In this connection please let me know what
will be the total cost.
The prevalence of artificially high-style uses of English among
Indians has been called "Babu English" and it is well parodied in a novel
by F. Anstey.21

17

After forming my resolution of writing a large novel, I

confided it to my crony, Mr. Ram Ashootosh LaZZ, who
warmly recommended to persevere in such a magnus opus.
So I became divinely inflated periodically every evening
from 8 to 12 p.m., disregarding aZZ entreaties from
feminine relatives to stop and indulge in a blow-out on
ordinary eatables . .
and at length my colossal
effusion was completed, and I had written myself out:
after which I had the indescribable joy and felicity to
read my composition to my mothers in law and wives and
their respective progenies and offsprings, whereupon,
although they were not acquainted with a word of English,
they were overcome by such severe admiration for my
fecundity and native eloquence that they swooned with
rapture .
.
"
.

In examining instances of interference or language transfer we

thus need to consider more than just the linguistic variables and their
distribution across languages; we need also consider social reactions
to different aspects of language use, since these too may be carried
from one language to another influencing the sort of sentences that may
be formulated in the second language.

8.

STRATEGIES OF COMMUNICATION AND ASSIMILATION

The shape of the utterances produced in the second language may

be influenced by additional factors, not related to interference, or to
aspects of generalization and analogy. Under communication strategies
we may include errors which derive from the fact that heavy communication demands may be made on the second language, forcing the learner to
mold whatever he has assimilated of the second language into a means of
saying what he wants to say, or of getting done what he wants to get
done.
The learner may simplify the syntax of the language in an effort
to make the language into an instrument of his own intentions. Errors
deriving from such efforts may be attributed to strategies of communication.
Errors attributable to the learner's attempts to reduce the
learning burden of what he has to assimilate. may be closely related,
and these may be referred to as strategies of assimilation.
Perhaps the clearest examples of alterations in language structure as a result of strategies of assimilation and communication are to
be found in pidgin languages
-.languages used by people whose mother
tongues are different, in order to facilitate communication between
them.
In the process of becoming a.pidgin a language often loses some
of its vocabulary, or is simplified in its phonology or grammar. Simpli-

18
- 16 -

fication is one way in which speakers of different languages can make a

new language easier to learn and use.22
We frequently find instances of a similar process in a normal
second language setting. The communicative demands made on the second
language may far outpace the speaker's actual competence in the second
language, thus the speaker may have to create the means of expressing
relations for which the language course has not prepared him. The school
English course, for example, may begin with the present tense, the present
continuous, and following concepts of linguistic grading or sequencing,
delay the introduction of other tenses until the present and present
continuous have been mastered. Supposing however, that the English program is that of an English-medium school in a multi-lingual context,
where English provides the only lingua franca between Chinese, Malay and
Tamil-speaking children. The children in their use of English as a lingua
franca outside of the classroom cannot wait until the past and "future"
tenses have been taught before they will begin talking about past and
planned events. They will have to establish their own way of dealing
with past and planned events, and in so doing there will be considerable
simplification of syntax, and extension of the uses of known forms into
other areas. This may produce what has been called "dormitory English"
and "playground English" which can function as an entirely efficient form
of communication: You are not knowing what Boonlat is do.
He is open
stand-pipe
water is not coming, so he go ask .
23
.

This may seem an extreme example, but in any situation where the
second language actually has to be used outside the classroom in real
situations, inevitably the learner finds himself having to cope with
circumstances which the school syllabus has not covered or for which he
may not have the linguistic resources available.
Looking at such language
samples, it is often not possible to say whether a particular error is
attributable to a strategy of communication, or to a strategy of assimilation, that is, an identifiable approach by the learner to the material
being learned.
My own acquisition of French in Quebec has provided me with many
examples of the effects of communication and assimilation strategies,
since on arrival in Quebec extensive demands were immediately made on
whatever I had been able to pick up of the language. As an example of
assimilation strategy, I found the form je vais (I'm going to) easier to
learn than the future tense in French, and I quickly developed the means
of expressing futurity or intention with the use of this construction.
This however led me to use the going to form in situations where the
future tense was appropriate, and now I frequently have to correct a
tendency to use the going to form in sentences like Je vais vous tete=
phoner ce soir (I'm going to telephone you tonight) when what is really
intended is Je vous taaphonerai ce soir (I'll telephone you tonight).

As an example of a communication strategy my acquisition o? the

conditionnel pass tense in French is illuminating. A language need
which soon presented itself was the need to express intentions in the
past which I was unable to fulfill. Since I did not acquire the grammatical means to do this until fairly late, I had to find alternative ways
of expressing the same content. This I did by lexical means, through the
use of a longer construction. Thus instead of saying "I would have liked
to have seen the film last night" J'aurais voulu voir le film hier soir
I would say "I had the intention of seeing the film last night, but ..."

J'avais l'intention de voir Ze film hier soir, mais

Similar strategies may account for the frequent misuse of the present
tense as a narrative form, since the present tense is usually introduced
first in language courses and the additional learning burden involved in
acquiring the past tense can be avoided if the past is simply expressed
lexically. A word like yesterday will thus suffice to locate the time
setting and the speaker will continue in the present. Yesterday we go
for a drive and we stop near the beach and we
.
Thus the speaker is
able to expand the functional capacities of his knowledge of the second
language, while keeping to known or sure ground.
.

A similar explanation may be attributed to the development of the

past continuous form as a narrative instead of the simple past. George
notes that many course books introduce the present and the present continuous in contexts such as these:24
The book is on the desk (present state)
I am putting the book on the desk (present action)
Then the past form of the first sentence may be extensively practised.
The book was on the desk (past state)
By the time the course comes to teach past narrative
We walked outside,
we looked towards the sky .
the students have already developed
their own means of expressing this function, by using the was form as a
past marker and by using the continuous form for verbs of action. In the
afternoon we were going back.
On Saturday we were going down town and we
were seeing a film and after we were meeting my brother . . .
.

The selection and gradation of items in a language course should

therefore take account not merely of linguistic factors, such as the
frequency or complexity of items to be introduced, but also the demands
that will be made on the language and how the learner will adapt what he
is given to his particular needs.

20
- I8 -

9.

ERRORS AND THE TEACHER

Teachers, as Pit Corder remarks, are more concerned with how to

deal with errors than with what causes them, and in this there have been
several schools of thought.
"One is that if we were to achieve a perfect
teaching method the errors would never be committed in the first place,
and that therefore the occurrence of errors is merely a sign of the
present inadequacy of our teaching techniques."25 This attitude is
illustrated in the introduction to a recent elementary English course.
"One of the teacher's aims should be to prevent mistakes from occurring.
In the early stages while the pupils are wholly dependent on the teacher
for what they learn, it should be possible to achieve this aim. "26
It is difficult to reconcile this approach with what we know or can
observe about language learning. Children do not themselves acquire
language by correctly imitating sentences they hear.
"A child learns his
language by interacting with it, by actively coping with and manipulating
his environment. He does this on the basis of unsystematic, usually unplanned language input on the part of his parents."27 Does this lead to
an essential casualness on the part of the teacher towards the nature of
the sentences his students produce? After all, current views of language
learning emphasize that language cannot be taught, but must be learned by
the child. Attempts to teach language by direct imposition of an adult
grammatical model seem psycholinguistically inconsistent.
In assessing the teacher's role in second language teaching we are
however faced with the fact that although the child learning his mother
tongue begins by producing sentences which do not duplicate adult sentences,
gradually building towards an adult grammar, in second language learning
what is usually the end point in the learning process is an Interlanguage,
that is, a form differing from the target language characterized by interference both internal and external to the language. What we would have to
postulate then is the encouragement of those types of sentences that indicate language development, and the minimization of opportunities for
fossilization, for the establishment of permanent deficiencies in the
learner's competence.
In the present state of our knowledge about second
language acquisition this is not a realizable goal, since we have so
little information about which types of errors in second language learning
are positive and which are negative.
A form like What he is doing? may
represent,transitional competence in first language acquisition, but
fossilization in the second language learner's speech. In the present
utate of our knowledge, we need to be careful not to be overoptimistic
about the relevance to second language teaching of studies of first
language acquisition.
At the same time all language teaching techniques can be seen to
produce transfer-of-training effects, where the general effects of memory
on the material taught are apparent in reconstruction, and retroactive
inhibition as well as generalization and analogy.
It is possible to try

to arrange teaching materials to minimize the unwanted effects of such

factors, but we have no way of knowing whether this is ultimately of any
advantage to the learner.
Questions, for example, invariably influence
in unintended ways the form of the sentence the student uses in his
answer. What was the woman saying? used to elicit statements about a
film will tend to elicit The woman saying she will
Ask her how
long it takes will produce How Zong it takes? and so on.
Or if we introduce a text for practising comparative constructions including is as big
as, is bigger than, some of the students will produce is as bigger as,
is as big than.
Such effects might be minimized by separating the forms
which interfere with each other and teaching them at different times, but
we cannot say very much about what the results of particular instructional
approaches will be on the final competence of the learner. It may make
no difference at all, since there is no evidence that the linguistic and
instructional variables
the points we select to teach and the order and
manner in which we teach them
are the same as the psychological variables,
the actual nature of the process by which the language items become part
of the speaker's competence.
.

Despite the inadequacies of our present knowledge about the relevance of particular approaches to language instruction, there are excellent social motivations for teachers' drawing their learners' attention
to examples of fossilization, to those errors which seem to have become
a permanent rather than a transitional feature of their speech. In looking at the social justification for the correction of certain errors we
can keep in mind however that linguistically, we may simply be trying to
modify our learner's performance rules. Even if his competence is represented in sentences like Yesterday I go down town, conscious attention
to the way he speaks may enable him to modify his performance so that he
produces Yesterday I went down town.
If grammatically deviant speech
still serves to communicate the speaker's intentions, why should we pay
further attention to it?
Simply because speech is linked to attitudes and social structure.
Deviancy from grammatical or phonological norms of a speech community
elicits evaluational reactions which may classify a person unfavorably.
In sociolinguistic terms "
our speech, by offering a rich variety
of social and ethnic correlates, each of which has attitudinal correlates
in our own and our listener's behaviour is one means by which we remind
ourselves and others of social and ethnic boundaries, and is thus a part
of the process of social maintenance (or change) ".2C Psychologists have
thus investigated the way listeners will provide a range of reliable
cultural, social class, and personality associations upon hearing speech
samples. These are usually measured by playing recordings of speakers,
and having them rated on a series of characteristics, such as intelligence,
character, good looks and so on.
I propose that the adult second language
learner's deviation from grammatical norms elicits evaluational reactions
.

22

-20-

which can be measured in similar ways.

My evidence is largely anecdotal.
Here is an extract from a recent UNESCO report.
"Individuals who learn
new languages later in life, especially after the age of 15 or so, characteristically have more difficulty with new structures than with new vocabulary, and the difficulty seems to increase with age.
It is not uncommon
to characterize such people as "having a foreign accent" or "speaking
brokenly", even though their vocabulary and general fluency may be quite
satisfactory in the acquired language.
Although they have little difficulty in being understood for practical purposes, they are apt to be
considered as perpetual foreigners or outsiders".29

My observations of native speakers' reactions to grammatical deviancy suggest that not all instances of deviancy, not all errors, are
evaluated in the same way. We don't react to I'm going in Paris next
week in the same way as we react to I is going to Paris next week or to
He come from India.
Omission of a third person s or a plural seems to
grate rather violently, whereas a misplaced preposition may not affect us
so much. Deviancy in article usage may elicit a "baby talk" evaluation.
You may have noticed this in your own relations with non-native speakers
of English. One almost automatically corrects certain types of mistakes
while we let others pass without too much thought.
Native speakers' reactions to systematic variation in grammatical
or other features could be measured using the psychologists' techniques
for measuring reactions to different dialects.
This would demand systematically varying the nature of deviancy in oral or written texts and
having native speakers assess the personality traits of the speakers.
Here is an example of a passage in which article usage is deviant:
I remember war period very clearly. I remember big bomb which
exploded near house I was living in in 1940.
First wall began
to crack and window broke and I hurried out of house just
before chimney fell down. I saw that big tree in front of
house was broken.

Here is the same text with article usage restored but past tense
omitted:
I remember the war period very clearly. I remember a big
bomb which explode near the house I live in in 1940. First
the wall begin to crack and the window break and I hurry out
of the house just before the chimney fall down. I -ee that
a big tree in front of the house is broken.
Information on the reaction of native speakers to particular aspects of
grammatical deviancy would thus enable us to say which examples of fossilization the second language teacher should pay most attention to.

23

- 21 -

CONCLUSIONS

To draw together some of the points I have touched upon here, I

have first of all tried to suggest why people who speak second languages
may not speak or write them with native-speaker like fluency. I have
suggested that deficiencies in their knowledge may be the results of
interference, the use of aspects of another language at a variety of
levels; of strategies of learning such as overgeneralization and analogy
by means of which the learner tests out his hypotheses about the structure
of the language; of strategies of assimilation, in which the learner makes
his learning task easier; and of strategies of communication, whereby the
learner adapts what he knows into an efficient communication model, producing an optimal utility grammar from what he knows of the language.
At the same time we need to distinguish between performance and
competence errors. The former are occasional and haphazard and are
related to such factors as fatigue, memory limitations and so on. The
latter are systematic and may represent either a transitional stage in
the development of a grammatical rule or the final stage of the speaker's
knowledge. While our knowledge about second language learning is still
largely speculative, excluding the possibility of prescribing recipes for
teachers, I hope that this account of errors and learning strategies has
at least suggested some of the reasons for what we hear from our students
of second languages.

24
- 22 -

REFERENCES

1.

Linguistic Mechanisms Underlying Child Speech.

Ursula Bellugi.
Proceedings of the Conference on Language and Language Behaviour,
E.N. Zale (Ed.). New York, Appleton-Century Crofts, 1968. p. 3650.

2.

Dan. I. Slobin.
Suggested Universals in the Ontogenesis of Grammar.
Working Paper 32, Language Behaviour Research Laboratory. University of California at Berkeley, 1970.

3.

Examples taken from Jack Richards. A Non-Contrastive Approach to

Error Analysis. English Language Teaching, 1971.
(in press)

4.

S.P. Corder.
169, 1967.

5.

L. Selinker. The Psychologically-rOevant Data of Second Language

Learning. Paper read at the Secoitd International Congress of
Applied Linguistics.
Cambridge (I ngland), September 1969. p. 8.
See also L. Selinker. Language Tiansfer. General Linguistics 9
(1969)2:67-92.

6.

R. Lado. Linguistics Across Cultures.

Michigan Press vii, 1957.

7.

R. Wardhaugh. The Contrast6e Analysis Hypothesis.

ly 4(1970)2:123-130.

8.

L. Jakobovits.
Foreign Language Learning.
1970.
p. 111-112.

9.

L. DUskova.
On Sources of Errors in Foreign Language Learning.
IRAL 7:11-36, 1969.

The Significance of Learner's Errors.

IRAL 5:161-

Ann Arbor, University of

TESOL Quarter-

Rowley, Newbury House,

Kinzel. Lexical and Grammatical Interference in the Speech of

a Bilingual Child.
Seattle, University of Washington 2-3, 1964.
11.

D. Wolfe. Some Theoretical Aspects of Language Learning and

Language Teaching. Language Learning 3&4:181, 1967.

12.

J. Falk. Nominalizations in Spanish. Studies in Linguistics and

Language Learning V. Seattle, University of Washington Press 4,
1968.
- ".3

13.

Dbskova 1969, 18.

14.

F.G. French. Common Errors in English.

Press 9, 1949.

15.

E.F. Aguas. English Composition Errors of Tagalog Speakers. D.Ed.

dissertation, University of California, Los Angeles 49, 1964.

16.

cp Jakobovits 1970, ch. 4.

17.

R. Jones. System in Child Language.

Wales Press 38, 1970.

18.

W.F. Leopold. Speech Development of a Bilingual OhiZd.

Northwestern University Press 4:64-65.

19.

C. Ferguson.

20.

R. Kaplan.
Contrastive Rhetoric and the Teaching of Composition.
TESOL Quarterly 1(1967)4:10-16.
See also M. Macmillan. Aspects of Bilingualism in University
Education in Sudan, in Language in Education in Eastern Africa
(ed. T. Gorman). Nairobi, Oxford University Press, 1967. p. 146.

21.

F. Anstey.

22.

W.J. Samarin. Lingua Francas, with Special Reference to Africa, in

Study of the Role of Second Languages in Asia, Africa, and Latin
America (ed. F. Rice).
Washington, Center for Applied linguistics, 1962.
p. 54-64.

23.

Example taken from a working paper by H.V. George.

24.

H.V. George. Teaching Simple Past and Past Perfect. Bulletin of

the Central Institutes of English (Hyderabad, India) 2:18-31,

Diglossia.

London, Oxford University

Aberystwyth, University of

Evanston,

Word 15(1959)2:325-340.

A Bayard from Bengal.

London, Methuen, 1902.

1962.
25.

S.P. Corder 1967.

26.

W.R., Lee.

27.

D.F. Moores. Psycholinguistics and Deafness.

the Deaf-39, January 1970.

28.

F. Williams.

The Dolphin English Course; Teachers Companion.

Oxford University Press 3, 1970.

London,

American Annals of

Language Attitude and Social Change, in Language and

Poverty (ed. Williams).

26
- 24 -

29.

R. Noss. Higher Education and Development in South East Asia.

Paris, UNESCO 22.

27
- 25 -