8/6/2019 Error Analysis Ellis1985

1/2

Error Analysis

Sridbar (1981) points out that Error Analysis has a long tradition. Prior to the early 1970s, however,

Error Analysis consisted of little more than impressionistic collections of common errors and their

linguistic classification (e.g. French 1949). The goals of traditional Error Analysis were pedagogicerrors

provided information which could be used to sequence items for teaching or to devise remedial lessons.The absence of any theoretical framework for explaining the role played by errors in the process of SLA

led to no serious attempt to define error or to account for it in psychological terms. Also as the

enthusiasm for Contrastive Analysis grew, so the interest in Error Analysis declined. In accordance with

Behaviourist learning theory, the prevention of errors (the goal of Centrastive Analysis) was more

important than the identification of errors. It was not until the late 1960s that there was a resurgence of

interest in Error Analysis. A series of articles by Corder (e.g. 1967; 1971; 1974) both traced this resurgence

and helped to give it direction.

The procedure for Error Analysis is spelled out in Corder (1974). It is as follows. (1) A corpus of

language is selected. This involves deciding on the size of the sample, the medium to be sampled, and

the homogeneity of the sample (with regard to the learners ages, L1 background, stage ofdevelopment, etc). (2) The errors in the corpus are identified. Corder (1971) points out the need to

distinguish lapses (i.e. deviant sentences that are the result of processing limitations rather than lack

of competence) from errors (i.e. deviant sentences that are the result of lack of competence). He also

points out that sentences can be overtly idiosyncratic (i.e. they are ill formed in terms of target

language rules) and covertly idiosyncratic (i.e. sentences that are superficially well formed but when

their context of use is examined are clearly ungrammatical). (3) The errors are classified. This

involves assigning a grammatical description to each error. (4) The errors are explained. In this stage

of the procedure an attempt is made to identify the psycholinguistic cause of the errors. For example,

an attempt could be made to establish which of the five processes described by Selinker (1972) is

responsible for each error. (5) The errors are evaluated. This stage involves assessing the seriousness

of each error in order to take principled teaching decisions. Error evaluation is necessary only if the

purpose of the Error Analysis is pedagogic. It is redundant if the Error Analysis is carried out in order

to research SLA.

The context for the new interest in errors was the recognition that they provided information

about theprocess of acquisition. As I pointed out earlier, this term has two meanings. Two questions

can be asked, therefore. What light can the study of learner errors throw on the sequence of

developmentthe interlanguage continuumthrough which learners pass? What light can errors

shed on the strategies that the learner uses to assimilate the rules of L2? Both of these questions are of

central importance to the main theme of this chapterthe natural order of development.

Error Analysis provides two kinds of information about interlanguage. The firstwhich is

relevant to the first of the two questions posed aboveconcerns the linguistic type of errors produced

by L2 learners. Richards (1974), for instance, provides a list of the different types of errors involving

verbs (e.g. be + verb stem instead of verb stem aloneThey are speak French). However, this

type of information is not very helpful when it comes to understanding the learners developmental

sequence. Error Analysis must necessarily present a very incomplete picture of SLA, because it

focuses on only part of the language L2 learners producethat part containing idiosyncratic forms.

Describing interlanguage requires identifying what the learner can do in toto by examining both

idiosyncratic and non-idiosyncratic forms. Also because SLA is a continuous process of development,

8/6/2019 Error Analysis Ellis1985

2/2

it is doubtful whether much insight can be gained about the route learners take from a procedure that

examines language-learner language at a single point in time. Error Analysis provides asynchronic

description of learner errors, but this can be misleading. A sentence may appear to be non-

idiosyncratic (even in context), but may have been derived by means of an interim rule in the

interlanguage. An example might be a sentence like Whats he doing? which is well formed but may

have been learnt as a ready-made chunk. Later, the learner might start producing sentences of the kindWhat he is doing?, which is overtly idiosyncratic but may represent a step along the interlanguage

continuum. For these reasons an analysis of the linguistic types oferrorsproduced by learners does

not tell us much about the sequence of development.

The second type of informationwhich is relevant to the question about the strategies used in

interlanguageconcerns the psycholinguistic type of errors produced by L2 learners. Here Error

Analysis is on stronger ground. Although there are considerable problems about coding errors in

terms of categories such as developmental or interference, a study of errors reveals conclusively

that there is no single or prime cause of errors (as claimed by the Contrastive Analysis hypothesis)

and provides clues about the kinds of strategies learners employ to simplify the task of learning a L2.

Richards (1974) identifies various strategies associated with developmental or, as he calls them,

intralingual errors. Overgeneralization is a device used when the items do not carry any obvious

contrast for the learner. For example, the past tense marker, -ed, often carries no meaning in context,

since pastness can be indicated lexically (e.g. yesterday).Ignorance of rule restrictions occurs when

rules are extended to contexts where in target language usage they do not apply. This can result from

analogical extension or the rote learning of rules.Incomplete application of rules involves a failure to

learn the more complex types of structure because the learner finds he can achieve effective

communication by using relatively simple rules.False concepts hypothesizedrefers to errors derived

from faulty understanding of target language distinctions (e.g. is may be treated as a general marker

of the present tense as in He is speaks French). Perhaps the most ambitious attempt to explain SLA

by analysing the psycholinguistic origins of errors, however, is to be found in George (1972). George

argues that errors derive from the learners need to exploit the redundancy of language by omitting

elements that are non-essential for the communication of meaning. Implicit in the types of analysis

provided by both Richards and George is the assumption that at least some of the causes of errors are

universal. Error Analysis can be used to investigate the various processes that contribute to

interlanguage development.

The most significant contribution of Error Analysis, apart from the role it played in the

reassessment of the Contrastive Analysis Hypothesis, lies in its success in elevating the status of

errors from undesirability to that of a guide to the inner workings of the language learning process. As

a result of interlanguage theory and the evidence accumulated from Error Analysis, errors were no

longer seen as unwanted forms (George 1972), but as evidence of the learners active contribution to

SLA. This contribution appeared to be broadly the same irrespective of differences in learners

backgrounds, suggesting that the human faculty for language may structure and define the learning

task in such a way that SLA, like L1 acquisition, was universal in nature. However, the conclusive

evidenceproof that there was a natural route of developmentwas not forthcoming from Error

Analysis, although with a number of methodological developments the cross-sectional analysis of

language-learner language could provide such evidence.

(Ellis, Rod. 1985. Understanding second language acquisition pp.51-54. Oxford University Press.)