LINGUIST List 11.1598

Sat Jul 22 2000

Review: Doughty & Williams: Focus on Form in SLA-part 1

Editor for this issue: Andrew Carnie <carnielinguistlist.org>

Directory

1. Ronald Sheen, Review part 1

Catherine Doughty and Jessica Williams (Eds.) (1998) "Focus
on form in classroom Second Language Acquisition" (The
Cambridge Applied Linguistics Series) Cambridge University
Press: Cambridge, U.K. 301 pages. (1)

Ronald Sheen, University of Quebec in Trois Rivieres.

The late 70's and 80's saw the comprehensible input
hypothesis as a driving force in bringing communicative
language teaching to full bloom and the relegation of
explicit grammar teaching to a minor if not non-existant
role. The 90's have seen the questioning of the
sufficiency of the comprehensible input hypothesis and a
return to a concern with a focus on form. However, as
Doughty and Williams point out (p. 2) this does not
constitute "a return to discrete point grammar
instruction.". (2) The term focus on form derives from
Long (1991, for example) in which, whilst maintaining the
validity of his interaction hypothesis, he pointed out the
need for a focus on form to be enmeshed in communicative
activity and motivated by communicative need. As such, it
is contrasted with traditional (termed "synthetic" by Long
following Wilkins, 1976) syllabuses based on the teaching
of a list of grammatical elements.

The book under review, "Focus on form in classroom SLA",
brings together ostensible examples of the new focus on
form approach though, in reality, most of the contributions
implicitly manifest differences with Long's principles.
The editors, Catherine Doughty and Jessica Williams,
provide a very good opening chapter as introduction to the
various contributions and an excellent closing chapter, of
which more later. Between these, are chapters by Long and
Robinson (L&R), DeKeyser, Swain, White, Doughty and Varela,
Williams and Evans, Harley, and Lightbown, in that order.
As such, they present a representative selection of current
thinking in this area though the absence of a chapter on
VanPatten's input processing is a loss.

The term "focus on form" is problematic because, on the one
hand, it has the restrictive meaning intended by Long and,
on the other, it has taken on the more general meaning of
a return to a concern with grammatical accuracy within a
communicative context. Thus, one finds DeKeyser using the
term even though he writes from an entirely different
perspective to L&R and one which the latter proscribe. The
editors address this terminological issue in their
introduction. They opt to use "formS-focused instruction"
to refer to the teaching of grammar in isolation (without,
however, defining what they actually mean by "isolation")
and the acronym, "FONF instruction", to contrast with it,
whilst excluding the term "form-focused instruction" even
though one of the contributors, Lightbown, has embraced the
term. In this review, I will adopt those two terms whilst
using "focus on form" in inverted commas to specify the use
by L&R, and without inverted commas to refer to the
movement in general.

In spite of this classificatory clarification, the terms
still fail to capture the divergence in the implicit
underlying principles upon which the contributors base
their writings. Thus, we have views stretching from
DeKeyser's advocating skills-learning-based synthetic
syllabuses founded on declarative knowledge, controlled
and automatic processing and even accepting a role for
"formS focused" lessons (p. 58) to L&R's advocacy of
"focus on form" with its implicit proscription of what
DeKeyser and other contributors propose. However, close
reading of the nine substantive articles reveals a clear
division between L&R with its doctrinaire arguments,
putatively justifying certain proscriptions, and the other
eight which opt for instructional strategies covered under
the broad purview of FONF instruction. That is, they all
implicity reject L&R's proscriptions for they all accept
the necessary inclusion in a syllabus of language forms to
be covered and nowhere do they express agreement with the
principle of the necessity for teachers to negotiate with
students the aims of focal attention. This said, it needs
to be pointed out that they manifest varying degrees of
difference with L&R which possibly arise from the extent
to which they accept the principle proposed by Doughty &
Williams (p. 261) which rightly contends that the range of
possibilities offered by FONF instruction "...should not
be theoretically proscribed."

For the rest of this review, I will deal with each chapter
separately, in the order they appear, adding some
comparisons and contrasts between them and conclude with
general summarizing comments.

"Focus on form: Theory, research and practice" by Michael
Long and Peter Robinson.

As do most of the contributions by Long to the field, this
chapter provides a comprehensive and critical account of
much of the experimental and quasi-experimental research
of relevance to the issue at hand. As such, it
constitutes a major contribution. There are, however,
aspects of this chapter of questionable legitimacy in
terms of the perception of past failures and the
practicalities of classroom language learning. It is on
these which this review will concentrate.

The authors divide methodological approaches into two
broad types: synthetic and analytic. The former are
characterized as being based on the breaking down of
language into discrete parts, presenting them deductively
or inductively, in linear additive fashion and assuming
that learners will "synthesize the pieces for use in
communication" (p. 15-16). Analytic approaches, on the
other hand, are characterized as being based on the
assumption that "people of all ages learn languages best,
inside or outside the classroom, not by treating the
languages as an object of study, but by experiencing them
as a medium of communication.". (p. 18). Grammar
Translation, ALM, the Silent Way and TPR are assumed to
be synthetic whereas the Natural Approach and the
Procedural Syllabus are assumed to be analytic but
meaning-based whilst Content-based language teaching and
the Process Syllabus are analytic but based on "focus on
form" - that is the approach advocated by the authors.
The chapter then goes on to explain why the synthetic and
the analytic meaning-based approaches have resulted in
failure.

L&R set themselves, then, a worthwhile task for if there
has been failure, the source thereof needs to be
identified.. However, it is a fundamental error to base
the accomplishment of that task on pedagogical assumptions
derived from the putative nature of formal foreign
language learning which has not been validated by any
successful long-term implementation in the classroom.
Their doing so results in their lumping together of
Grammar Translation, ALM, The Silent Way and TPR and
ascribing their assumed failure to the single fact that
they are synthetic methods whilst ignoring fundamental
differences. Setting aside the fact L&R do not
demonstrate failure in all these cases, it is surely
unjustified to take two diametrically opposed methods
such as Grammar Translation and ALM and ascribe their
failure to the fact that they are both based on synthetic
syllabuses whilst ignoring fundamental differences between
them.

The failure of the analytic meaning-based methods
(otherwise termed "non-interventionist") such as The
Natural Approach is ascribed to the lack of a "focus on
form". The authors, however, propose to retain the
essentials of such a method but emphasize the importance
of interaction and the accomplishment of tasks as a means
of promoting it. "Focus on form" is implemented by
means of pedagogic tasks intended to bring formal features
to learners' attention. This may even at times result in
attention to individual linguistic features when problems
of production or comprehension arise. They justify this
by contending that it is similar to what occurs with
native speakers when they need to think consciously about
a particular form to use. (p. 23) This is weak if not
invalid as argument. When L2 learners have such a problem,
it is usually because they do not have the requisite
knowledge and look, therefore, to the teacher or some
other source to provide it. When native speakers encounter
such a problem, it is not because they do not have the
necessary knowledge but because they need to think
consciously about forms they usually use unconsciously.
It is, therefore, unfounded to assume a similarity between
such a classroom learning problem and the difficulty
encountered by a native speaker in the use of the L1.

However, perhaps the two clearest examples of the authors'
failure to show an awareness of the realities of formal
classrooms is first, their proscription of a pre-planned
syllabus and, second, their proposal that how focal
attention should be allocated needs to be negotiated
between students and teacher (p. 24). It is surely
self-evident that most formal classrooms entail
examinations as a means of evaluation of coverage of the
syllabus. This being so, course-content will often need
to include grammatical items (which must necessarily be
pre-planned) and is consequently non-negotiable. However,
apart from these practical considerations, there are
research findings casting doubt on the feasiblity of the
authors' proposals simply because the students, themselves,
may not want such arrangements. Willing (1988) in a
survey of learning preferences, discovered an overwhelming
desire for explicit grammar teaching and the teacher-
controlled classroom. This finds resonance in the
students identified in Carrell, Prince & Astika (1996) as
the "Sensing-Thinking-Judging" types who constituted the
overwhelming majority of students in their study and whom
the authors expect "...to be guided by concrete facts and
sequential learning rather than by abstractions, to prefer
logic, rules and examples over social interaction, and to
prefer order, organization, and formalized, structured
instruction" (p. 96), a position echoed in Gefen (1993:136),
- and manifest in features one would associate with
synthetic syllabuses.


There is also something of an irony in the proscriptions
of these two authors. They write with justifiable
disparagement of "unproductive pendulum swings" (p. 21)
whilst being apparently unaware that the fundamental cause
of such swings is the unquestioning acceptance of
doctrinaire approaches putatively justifying the
proscription of specific teaching practices - in other
words, the type of approach that they themselves advocate.
Moreover, there is further irony in the following
statement in Larsen-Freeman & Long (1991:290): "We must
guard against overzealousness on the part of theorists or
their devotees who feel that they have a monopoly on the
truth". A reading of Long & Crookes (1992) and this
chapter inevitably brings such zealous theorists to mind.
It is such zealotry which leads applied linguists to feel
justified in proscribing teaching practices and, in the
case of Long (1988:136), in characterizing some of them as
"neanderthal", whilst seemingly ignoring the fact that
they have facilitated countless numbers of learners'
achieving high levels of language proficiency (see, for
example, Von Elek & Oskarsson, 1973, Diller 1975, Strevens,
1987, Chastain, 1988, Spolsky, 1989, Cook 1991, Stern,
1992, Palmer, 1992, Ur, 1988, 1993, for a discussion of
aspects of the effectiveness of such practices).

The most serious failure, however, in L&R's advocacy is
the lack of provision of any positive findings based on
long-term trialling in normal classrooms of what they
propose. (See Beretta, 1986, for a call for such research
in preference to the largely experimental research to
which L&R appeal). This is contrary to Long's and L&R's
justified insistence on the need for language learning
prescriptions to be supported by classroom-based research
L&R:41). Their specific advocacy is based largely on
theorizing on the nature of language learning and the
findings of experimental evidence. Past experience has
demonstrated that this approach does not lead to success
in the classroom. Yet, these two authors, whilst
justifiably complaining of the pendulum swings provoked
by such advocacies, repeat those same mistakes of the
past.

This would be understandable if this were the first time
that these proposals had been published. This is not the
case here for essentially the same advocacy was made in
Long and Crookes (1992). One might reasonably expect that
the intervening six years would have provided sufficient
time for their proposals to be trialled in normal
classrooms, particularly as the need for Long & Crookes
to carry out such trialling had already been raised in
the literature (Sheen, 1993, for example). Unfortunately,
this chapter does not address that need and provides no
reports on the findings of such classroom-based research.

In summary, this chapter manifests a general feature which
has been instrumental in preventing the field of applied
linguistics/second language acquisition from fulfilling
its implicit mandate of improving the efficacy of
classroom language teaching/learning. This feature
entails developing a theory of SLA and then assuming
that that theory represents mental reality and as such
justifies prescriptions and proscriptions for language
teaching and learning. It further entails an assumption
that the implementaton of those proscriptions and
prescriptons will result in a substantial improvement
in the efficacy of classroom learning. In fact, this
has characterized most innovations in the field whether
it be in education, in general, or language teaching,
in particular. Unfortunately, most innovations have
proven to be failures (Adams and Chen, 1981; Brumfitt,
1981; Fullan, 1982). In fact, Markee (1993:231), given
the high risk of failure, argues that "...innovations
should be resisted rather than promoted because their
adoption may be more harmful than beneficial." Valette
(1991:325), indeed, argues, with supportive test scores,
that the innovations of the previous twenty five years
had resulted in the worsening of the proficiency
standards of seniors graduating from college. There is,
therefore, little reason to feel optimistic about the
chances of success of the approach proposed here
particularly as it fails take into account the realities
of normal classrooms and students' preferences.

"Beyond focus on form: Cognitive perspectives on learning
and practising second language grammar" by Robert M.
DeKeyser.

The juxtaposing of this chapter to that of L&R highlights
the need for terminological clarity for to use the term
focus on form to characterize what is proposed in both
chapters, as is the case, makes a nonsense of the term
and this, because what Dekeyser proposes is proscribed
by L&R. The following from Dekeyser amply illustrates
this. He proposes that the teaching of grammar should
be based on cognitive-psychological "skill theory" and
says of it "...that declarative knowledge should be
developed first, before it can be proceduralized. This
means that, if grammar is to be taught, it should be
taught explicitly, to achieve a maximum of understanding,
and should be followed by some exercises to anchor it
solidly in the student's consciousness, in declarative
form, so that it is easy to keep in mind during
communicative exercises." Thus, DeKeyser is in actual
fact proposing an initial formS focused instruction,
the type which Long has already characterized as
"neanderthal".

DeKeyser's argument is, in fact, so far removed from
L&R's stance that it fails to respect the one underlying
principle proposed by the editors and one with which all
the other contributors agree. That is that all
instruction should be "embedded in primarily communicative
activities" (p. 2). His primary concern is to demonstrate
how declarative knowledge may be transformed into
prodedural and then automatized knowledge. In fact,
reading Dekeyser's is in some respects like a time-warp
experience for, as he readily admits, methodologically, he
is closest to the cognitive code learning of the 70's.
Where he differs is in the nature of the activities he
proposes for the promotion of proceduralization and
automatization. He emphasizes the need for those
activities to be communicative and not mechanical, a
point of view with which few or any would disagree.

DeKeyser also addresses the question of differences in the
amenability of rules to explicit instruction arguing that
some rules may be such that they can only be acquired
implicitly. Unfortunately, he does not exemplify his
argument. Given his proposals for instruction, it would
have been enlightening to have had an explanation of how
he proposes to integrate grammar acquired implicitly into
the proceduralization process. It should also be pointed
out that just as L&R fail to support their advocacy with
empirically verifiable successful implementation in normal
classrooms, DeKeyser limits his appeal to empirical
findings derived from largely experimental studies.

"Focus on form through conscious reflection" by Merrill
Swain

Swain's chapter is as far removed from L&R's position as
is DeKeyser's and is even more radical than his (given
the contemporary commitment to the communicative
principle). It describes the use of the dictagloss in
order to demonstrate the relationship between knowledge
of the metalanguage and language learning and is,
therefore, more radical for in recent decades the former
has been considered, at best, unnecessary, and, at worst,
harmful. Furthermore, one of her conclusions throws a
positive light on the knock-on effect of metatalk on
solving students' linguistic difficulties in "...making
forms and meaning the focus of their attention." (P. 79).
However, given this conclusion, it would appear to me
that an obvious consequence is that the provision of
explicit grammatical explanation may be considered in
a positive light. Unfortunately, she does not consider
this possibility, limiting herself to the discussion of
options in terms of collaborative tasks to produce metatalk.
One thus finds that the most direct means of providing
metalinguistic knowledge and stimulating metatalk is
precluded without any attempt made to justify that
decision. And this, perhaps, because she does not in fact
preclude it. (See Allen et al., 1990:115) for support
for this conclusion and Carroll & Swain (1993), the
findings of which showed that the group receiving explicit
metalinguistic feedback outperformed the other groups in
the study. Her position is, perhaps, more clear in Swain,
1988, for there she states that teachers should "help
learners undertake the sort of form-function analysis
needed to be effective communicators in the second
language." cited by Stern, 1990:97.

"Getting the learners' attention" by Joanna White

This chapter describes a rigorous 10-hour quasi-
experimental study which investigates the effects of
implicit instruction on the acquisition of possessive
determiners such as "her" and "his". There were three
experimental groups but no control group. One group
received a typographically enhanced input flood with
additional extensive reading and listening; another
received the enhanced input flood without the additional
input; the third group received the unenhanced input flood
again without the additional input. The enhancement was
achieved by means of enlargement, bolding, italics and
underlining. It was hypothesized that thanks to the
instructional means used, group one would advance the most
quickly and group two more quickly than group three.
White provides an excellent detailed analysis of various
aspects of the results which basically demonstrated that
"the findings did not support the hypotheses of this
study" (p. 101) and concludes that the type of instruction
used was inadequate to address learning problems resulting
from L1-L2 contrasts.

This conclusion raises a number of issues related to the
approach adopted by White and other authors of chapters in
this book. That approach entails a certain selectivity in
the research they appeal to and an apparent ignoring of
much other relevant research. Thus, White, inspired by
the proposed necessity of attention and noticing (Schmidt
1993) limits her study to an implicit instructional
approach. Yet, there is already ample evidence that in
terms of tackling learning difficulties caused by L1-L2
contrasts, an explicit instructional strategy cannot be
ignored (see, for example, Von Elek and Oskarsson, 1973,
Kellerman, 1977, 1979, Kellerman and Sharwood Smith, 1980,
James, 1980, Fisiak, 1981, Ringbom, 1987, Nickel, 1989,
Kupferberg & Olshtain, 1996,and Sheen, 1996). Given that
this study is concerned with L1-L2 contrasts, it is
surprising that not one reference in the chapter deals
with such problems or with relevant studies. It is even
more surprising that, given the documented success of an
explicit approach compared to an implicit approach in this
domain, White chose to compare two implicit instructional
strategies and not include a group which received
explicit instruction.