LINGUIST List 9.1026

Tue Jul 14 1998

Review: James: Errors in Language Learning

Editor for this issue: Andrew Carnie <>

What follows is another discussion note contributed to our Book Discussion Forum. We expect these discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for discussion." (This means that the publisher has sent us a review copy.) Then contact Andrew Carnie at


  1. WATTERS Paul Andrew, Errors in Language Learning.

Message 1: Errors in Language Learning.

Date: Tue, 14 Jul 1998 12:39:29 +1000 (EST)
From: WATTERS Paul Andrew <>
Subject: Errors in Language Learning.

James, Carl. (1998). Errors in Language Learning and Use: 
Exploring Error Analysis. London: Longman. ISBN 0 582 
25763-8 (Paperback). RRP: 14.99.

The analysis of errors in cognitive, linguistic and other 
psychological processes has a long history dating back to 
the introduction of signal detection theory in psychophysics 
and behavioural learning paradigms. Although many 
applied linguists now favour interlanguage paradigms for 
second-language acquisition, error analysis (EA) is still 
widely used in language classes. The attraction of EA lies 
in one's ability to isolate variability in responses, such as 
distinguishing true errors from "mistakes", which simple 
"correct/incorrect" paradigms tend to discard. In this new 
book on EA in language learning and usage, Carl James 
builds on his earlier work on contrastive analysis and 
applied linguistics to further explore the role that 
performance errors play in language acquisition (particularly 
second-language acquisition). The book consists of a 
historical overview of EA in applied linguistics, and then 
embarks on an ambitious attempt to both define and 
constrain the scope and methodology of EA in language 
processing and language learning. This methodology 
includes typologies for classifying and understanding how 
errors arise, as well as algorithmic specifications for the 
diagnosis and error correction in clinical and educational 

The first chapter aims to give a historical overview of EA in 
the context of its origins, its inspirations, its competitors, 
and its influences on second language teaching. James 
defines an error as "an unsuccessful bit of language" (p.1) 
which seems to be as succinct and compact a description as 
I've ever read! However, this very readable style of writing, 
whilst appearing informal, is maintained through later 
chapters where discussion of technical issues could easily 
have been obfuscated by a poor writing style. First- and 
second-order paradigms within language learning are 
described in detail in this chapter, with the interlanguage 
and crosslinguistic approaches compared with alternatives 
such as EA and contrastive analysis. Idiosyncratic and 
language-specific difficulties in language learnability are 
also covered in the context of linguistic change and 
metalinguistic influences on successful language 
acquisition. Several methods for collecting data in EA are 
informally introduced in this chapter (e.g., error elicitations 
such as the "broad trawl"), which naturally leads into the 
second chapter on defining the scope of EA in language 

The second chapter begins with an enlightening discussion 
of popular conceptions of what "proper" language is (such 
as the King's English), and catalogues the many failed 
attempts to enforce a "correct" dialect of English both in 
Britain and Asia (the "complaints" tradition). This issue is 
clearly relevant for defining exactly what an error is, given 
the absolutist attributions made by some educators and 
policy makers about the tenability or correctness of certain 
forms of spoken and written English (i.e., "standard" 
English). James outlines some typologies for understanding 
language norms based on geographical and historical 
constraints, but correctly identifies deficiencies in these 
schemes (particularly the failure, for example, to understand 
the role of colonialism in language preferences). This issue 
is taken up with respect to the issue of power and authority 
of native speakers with respect to non-native speakers of 
English, and conversely how the desire to speak a second 
language can unwittingly result in language loss and native-
language change.

Chapters 3, 4 and 5 focus on the definition and description 
of errors within the EA paradigm, having defined the focus 
of the EA methodology in the previous chapters. James 
begins by defining learners' ignorance of a target language 
in terms of four categories of deviance: grammaticality, 
acceptability, correctness and strangeness. It is a clear 
advantage of the authors' approach to EA that both 
grammatical/rational and performance/empirical approaches 
to language acquisition are covered by his typology, thus not 
"taking sides" with one viewpoint or the other. This 
rationale is based on the idea that EA is a methodology 
rather than a theoretical prescription. This focus continues 
with a discussion of error detection methods, in the context 
of locating and describing such errors in different parts of 
speech and indeed with respect to discourses longer than 
single sentences or phrases. The importance of a pluralistic 
approach which is tolerant of differences in dialects is 
emphasised, whilst ensuring that objective and stationary 
criteria are applied to utterances and writing within each 
dialect group. Error taxonomies, such as feature and surface 
structure approaches, are outlined in detail with worked 
examples, which are one of the key design features of the 
authors' pedagogical approach. Computer-assisted analysis 
of errors is also discussed, as are specific algorithmic 
approaches to rating levels of error in lexical and 
grammatical processes.

The next two chapters focus on diagnosing errors and 
evaluating their seriousness and impact for second-language 
learners in particular. Possible negative influences, such as 
interlingual errors arising from conflicts between the target 
language and mother tongue, are treated in detail, as are 
intralingual errors and inconsistencies which the non-native 
speaker encounters for the first time in the target language, 
such as over-generalisations and false analogies. In 
addition, the role of culture in influencing and perhaps 
determining some aspects of linguistic behaviour is 
discussed, for example, how native speakers might "gate" a 
non-native speaker. Error gravity and comprehensibility are 
also covered, as are some amusing examples given for "the 
irritation factor". The sociopragmatic consequences of error 
production in social situations, and the potentially negative 
outcomes for non-native speakers, are also discussed. 

Chapter 8 discusses pragmatic strategies for using EA to 
correct errors in speech and writing for second language 
learners. These are enhanced by a number of case studies 
presented in chapter 9. The issues covered in these last two 
chapters are non-trivial for applied linguistics: is second-
language teaching effective? If so, which approaches are 
best suited to particular kinds of students? Are 
formal/grammatical or informal/conversational approaches 
superior? Although James provides no magic answers for 
any of these questions, he does present a coherent 
methodology for answering these kinds of questions for 
individual situations, which is the great appeal of this book. 
This book would be suitable as an undergraduate or 
graduate text in applied linguistics or TESL programmes, 
but will be an invaluable reference for researchers in related 
fields such as psycholinguistics and machine translation, 
who might be searching for a more formal methodology for 
understanding error production in their respective fields. 
This book will be an indispensable addition to every 
linguist's library.

Reviewed by: Paul A. Watters, Department of Computing, 
School of Mathematics, Physics, Computing and 
Electronics, Macquarie University NSW 2109, 
AUSTRALIA. Tel.: +61-2-9850-9541; Fax: +61-2-9850-
9551; E-mail: Paul A. Watters 
is a research officer at Macquarie University in Australia, 
and is currently working on computational representations 
of semantics in models of language and speech production, 
as well as developing pragmatic approaches to machine 
translation. He is an Associate Editor of the South Pacific 
Journal of Psychology.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue