LINGUIST List 11.1599

Sat Jul 22 2000

Review: Doughty & William: Focus on Form in SLA- part 2

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. Ronald Sheen, Review part 2

Message 1: Review part 2

Date: Fri, 21 Jul 2000 12:46:51 -0400
From: Ronald Sheen <Ronald_SheenUQTR.UQuebec.CA>
Subject: Review part 2

[Review of Doughty & Williams continued]

"Communicative focus on form" by Catherine Doughty and
 Elizabeth Varela.

This chapter is different from the others in that it
addresses the use of FONF instruction in a content-based
science class. The instruction was achieved by means
of the teacher's (Varela) using recasts as a means of
corrective feedback on the use of past-time reference.
Testing of students' oral and written production
revealed statistically significant progress compared to
a control group which received no form of correctve

This as an excellent study and an exemplary model of how
an applied linguist might collaborate with a teacher in
an action research project. My one reservation concerns
a position which has become almost axiomatic in the focus
on form research. That is the principle that any focus
on form must be motivated by communicative situations and
carried out as part of communicative activities.
However, this is stated more as an act of faith than as
an empirically-supported argument. There is no research
finding that demonstrates this. This study offered an
ideal opportunity to put it to the test. It would have
been possible, for example, to have chosen another
grammatical item and to have begun each class with a
separate treatment of how it should be used. Then, when
errors in its use occurred, timely prompts could bring
about self-correction. See Lightbown's chapter for an
example of this type of intervention I would predict that
this would be as effective if not more so than the
incidental correction used in this study. The sooner such
comparative studies are carried out providing reliable
findings, the sooner we can abandon this oft-repeated
proscription of focus on formS or retain it.

"What kind of focus and on which forms?" by Jessica
Williams and Jacqueline Evans.

This chapter asks and partially answers the questions
that should have been addressed from the time that Long
proposed his "focus on form". That is, given that FONF
instruction will be of two broad types, implicit and
explicit, we need to know which of the two is generally
most effective and if the degree of effectiveness will
vary between the times when it is used and with the types
of grammatical items involved. In actual fact, there is
already ample empirical evidence to demonstrate that
explicit instruction is the most effective option if the
aim is to improve accuracy in production (See, for
example, Von Elek & Oskarsson 1973, Rivers, 1986, Palmer,
 1992, Kupferberg and Olshtain, 1996, and Sheen, 1996.)
However, as each new trend in applied linguistics tends
to reinvent the wheel and, in doing so, leave aside
findings of the past, this chapter, without referring to
such previous comparative studies, compares the
effectiveness of an input flood without corrective
feedback but increased frequency of occurrence of the
items under investigation (participial adjectives and the
passive) with another group which received the same input
flood combined with the presentation of rules and an
emphasis on the form, meaning and use of the items. A
control group had the same input flood without the
increased frequency. Predictably, the group receiving
explicit instruction had generally superior results though
the superiority was less pronounced in the case of the
more complex passive than in the case of the participial

The authors' discussion of the results is illustrative of
the problems faced by researchers in this field. The
results do not lend themselves to straightforward
conclusions as they would if all students in one group
did uniformly better than those in another group. In
the group receiving explicit instruction, some made little
to no progress in the use of the passive; others made
good progress. The authors, therefore, resort to
speculation as a means of explaining the apparent
inconsistencies. They speculate that the results make a
good case for "individual readiness for instruction
containing focus on form." (p. 151) However, such a
conclusion is only one possibilty and not the most
plausible. Two other possibilities are that, one, the
students who had done so poorly had little aptitude or
motivation to progress in their language learning, an
assumption supported by their low pre-test scores and, two,
more importantly, that their L1's caused greater
interference (see below). In fact, I would argue that
no experiment which did not minimally demonstrate that
instruction of some grammatical item was unsuccessful at
some time but was successful at a later time, could be
used as support for the readiness principle. Yet, the
authors offer their results as support for the validity

The issue of the influence of the L1 is of some relevance
in heterogeneous ESL classes as a means of accounting for
differential progress. It is, unfortunately, seldom
addressed and this possibly because of the minimalizing of
its importance in second language learning. (See Dulay
et al. 1982:5 for an example of this) This study is no
exception. Given the ubiquitous influence of the L1,
discussion of the results such as those of this chapter
cannot ignore (but does) the different degrees of
interference exerted by different languages. The
students in this study had as L1's the following
languages: Arabic, Cantonese, Spanish, Tagalog, Gujarati,
Hindu, Khmer, Korean, Mandarin, Polish, Russian, Thai,
Urdu and Vietnamese. Just to take two examples, Spanish
and Arabic, their respective passive structures are
radically different. That of Spanish is so close to that
of English that often a word for word translation is
correct as in, for example. "Esta ventana fue rota ayer"
(This window was broken yesterday) whereas in Arabic word
order presents crucial differences between it and English.
Unlike in English, in Arabic, the respective order of
performer and performed upon may vary according to
inflectional endings. Most of the non-Indo-European
languages involved in this study, present such
difficulties and will, therefore, constitute much greater
obstacles to progress than does Spanish. This factor
is not considered by the authors. It would have been of
interest to know the languages of the students who
 progressed most with those who made little to no
progress in the use of the passive.

"The role of focus on form tasks in promoting child L2
acquisition" by Birgit Harley

Harley's chapter sets it apart from the others for it
is the only one which deals with child L2 acquisition.
It is radical (used to mean divergence from current
positions) for it takes an age group, 7- to 8-year olds,
which according to received wisdom is not susceptible
to formal instruction, and demonstrates that by use of
various game-like strategies bringing attention to form,
such students can be brought to use far more correct
gender determiners than did a control group not
benefiting from such instruction. Furthermore, that
success was sustained over a six-month period, leading
the author to conclude that there is merit in pre-
planned instruction of parts of the curriculum which are
known to cause learning problems even with such young
learners. Harley's initiative is welcome for it
implicitly refuses to accept the validity of received
wisdom without putting it to the test. It is to be hoped
that her future plans include adding to her game-like
strategies judicious explicit grammatical instruction.
My experience as a teacher of this age group demonstrates
that the latter option should not be dismissed out of hand.

"The importance of timing in focus on form" by Patsy

Lightbown addresses two issues in her chapter: one
concern concerns the question of the desirablity of
concerns matching teaching interventions with assumed
developmental stages; the other addresses the question of
the need to integrate FONF instruction into communicative
activities while still retaining the possiblity of
momentary (no more than a minute p. 193) separation
therefrom. As to the first, she provides an excellent
summary of the relevant research and implies that such
stages have psychological reality but, at the same time,
wisely advises caution in implementing such findings in
the classroom because of the ubiquitous heterogeneity of
the majority of classes of students in terms of their
language development..

Her treatment of the second issue is worthy of particular
attention for it reveals both weaknesses in the research
studies of her own she cites and in the manner in which
she interprets them for in so doing she fails to heed her
own advice of "...emphasizing the need to avoid over-
interpretation of results." (Lightbown, 1984:241). She
uses those results to define her own conception of focus
on form, a term she uses without explicit
differentiation from L&R's use. However, she (personal
 communication, July, 1999) differentiates the form-
focused instruction she advocates from "focus on form".
(See Spada, 1997 for clarification of this point) The
bedrock of that conception is represented in the
following (p 192): "Research in intensive ESL classes
with young francophone learners has shown that teachers
who focus learners' attention on specific language
features during interactive, communicative activities of
the class are more effective than those who never focus
on form or who do so only in isolated "grammar lessons".
(Lightbown, 1991; Lightbown & Spada, 1990; Spada &
Lightbown, 1993). These effective teachers tend to
provide focus on form on the fly, without causing the
interaction to be interrupted or learners to be

It is instructive to submit this citation to some analysis
in part because the Lightbown & Spada research is so
frequently quoted in support of some position or other
related to FONF instruction. it is problematic in a
number of ways. The problems arise first from the
misleading nature of the first statement. The research
referred to is not only on ESL intensive classes but also
on regular classes of two hours a week of audiolingual
sessions. (See Lightbown 1991:203). Thus, Lightbown is
comparing two entirely different forms of instruction from
two different studies and drawing a conclusion based on
an invalid implicit comparison thereof. Furthermore, she
states that because the drills used in the audoliingual
lessons devoted to teach students to use "there's" did not
prevent their over-using "have" once they learned it,
they were not effective. She then uses this finding to
conclude that isolated grammar lessons are not as
effective as other instructional options. The reasoning
here is flawed. If one wishes to evaluate the value of
isolated grammar lessons, one does not take audiolingual
drilling as examples of such lessons. The method
comparison research of the 60's and 70's overall showed
audiolingual instruction to be less effective than
cognitive code learning type instruction (See Von Elek
and Oskarsson, 1973, for an extensive review of that
research). Lightbown needs, therefore , to take a more
effective option for the teaching of grammar in
isolation before dismissing it out of hand as being not
sufficiently effective.(3) Then, in order to carry out
 a rigorous comparison, apart from the obvious one of
comparing like-groups in like-situations which Lightbown
fails to do, one would take account of the difficulties
caused by the L1 interference of other forms. (Any
teacher of experience of francophones learning English is
well aware of the problems caused by the confusion
between "be" and "have" and, therefore, would not teach
them isolated from each other.) Then one would compare
the learning outcomes of such a treatment with that of
another group benefiting from another instructional
treatment. In the case in point, Lightbown compared the
learning outcomes of a regular audiolingual group of
some years previously with those of an intensive ESL
class without indicating how much time each group spent
on the grammatical point - an immediate disqualifier,
particularly given the contrast between the appeciably
briefer class time of the audiolingual class compared
with the ESL intensive class over a short time period.
Furthermore, the "instruction" of the ESL class in
question was ostensibly limited to instruction providing
corrective feedback. However, in providing this account,
Lightbown is obliged to rely on the anecdotal evidence of
a single teacher whose providing of that corrective
feedback she did not observe and, therefore, only assumes
that the "drumming into their little heads" (1991:207)
was carried out "...on the fly, without causing the
interaction to be interrupted or learners to be
discouraged.". However, not having observed the classes
in question, Lightbown is not justified in drawing the
conclusion she does. (4) In fact, the editors in their
final chapter, characterize this supposed intervention
"on the fly" somewhat differently. They state (p. 207)
"Her technique was to correct learner errors and remind
students of a metalingustic lesson that she had
apparently provided at an earlier class session". Thus,
Lightbown whilst being somewhat disparaging of isolated
grammar lessons, is, at the same time, citing as an
example to emulate the instruction of a teacher who based
her corrective feedback on such lessons previously given.

Yet, this particular "finding" is continually cited
without demur. (See, for example, chapters in the volume
under review). Thus we have a "finding" of doubtful
validity being continually blackboxed (ie a citation
which may not support what it is purported to do) (See
Latour, 1987; Block, 1996; Griffiths & Sheen, 1992;
Sheen, 1999, for comments on the dangers of such
blackboxing) to support the value of various means of
corrective feedback in an ESL intensive course and of its
being favourably compared with the teaching of "there is"
in a two-hour a week audiolingual class. This, Lightbown
then uses (1991:209) to critique the teaching of forms in
isolation and praise the drawing of attention to form
during communicative activities. If the same evaluative
criteria which have been applied to the method
comparison research of the 60's and 70's (See, for
example, Long, 1980, Lightbown, 1992 and Spada, 1997)
were applied to the Lightbown findings and interpretations
cited here, those findings would not be afforded credence.
Yet, Lightbown uses that anecdotal evidence and reasoning
as part of the justification for the specific type of FONF
instruction she proposes.

The position adopted by Lightbown (5) is of vital
relevance both to the book under review and to language
teaching in general for it has all the appearance of
becoming yet one more myth of language teaching. Nowhere
in the literature is there to be found research findings
which demonstrate that the approach advocated by Lightbown
is more effective than that proposed by DeKeyser, for
example. Until numerous studies have compared the
differential effectiveness of these two approaches, for
example, and demonstrated the necessity for all classroom
activity to be focused solely on communicative activity,
Lightbown' restrictions on FONF instruction will remain
more a position of faith than proven effectiveness.

"Pedagogical choices in focus on form" by Catherine Doughty and Jessica Williams

In an extended final chapter of some 65 pages, the editors
of this volume provide a thorough and excellent account
of the previous chapters, the focus on form literature in
general and the issues it must address in the future.
They arrive at the thoroughly justified conclusion that
some type of FONF instruction is justified whether it be
implicit or explicit, reactive or proactive. There are,
however, disquieting features which from this reviewer's
perspective reveal a tendency both in this chapter and in
some of the others to consider only what came after Long's
introduction of the idea of focus on form to be of any
relevance and, this, in spite of stating (p. 1) that the
issue has a long history. It is this long history which
moved Phillipson et al. (1991:8) to note: "There is a
definite need among professional applied linguists for
a heightened awareness of our historical and scientific
origins.". It would appear, however, that in the eyes of
these two editors that that history has little to offer
of value before 1988. Thus, in a chapter containing some
250 references, all but some 15% are 1988 or post 1988
and all but 2% are post 1980. Furthermore, major figures
in the history of applied linguistics and language
teaching are not mentioned. The following among numerous
others come to mind: Richards, Rivers, Stevick, Strevens,
Spolsky, McLaughlin, James, Carroll, Nation, Diller,
Faerch, Stern, Cohen, Ur, Chastain, Von Elek, Oskarsson,
Lado. Is it really plausible that Long and one of the
editors, Doughty, would merit over a dozen references each
while the contributions of the above-mentioned figures
are ignored. This attitude on the part of these two
authors is not uncommon in this field. Ostensible new
fields of inquiry lead the participants therein to
consider the past to have little relevance and large
sections of the applied linguistic world to have nothing
to offer. This might be justified in this case if this
new initiative had brought about marked improvement in
learning and implicitly revealed past research findings
to have no relevance. This is not the case here. All
the above-above mentioned unnoted figures have made
contributions of relevance to the issues discussed as
have the largely unmentioned many European-based applied
linguists. (See, for example, Kettemann & Wieden, 1993
and Fisiak, 1981). The approach of these two authors
results in a form of exclusionary parochialism which
leads them to adopt positions for which there is
substantial counter-evidence which is readily available
if one adopts a broader view of the field. Take the
following two examples from this chapter:

1. As far as the choice between a reactive approach
(trouble-shooting from the hip, so to speak) and a
proactive approach (resulting in a planned grammatical
syllabus, for example), the authors declare (p. 198):
"the jury is still out". It is only still out if the
members are blind to the history of the field. There
is ample evidence to demonstrate that proactive
approaches have produced countless successful language
learners. However, as stated in my comments on the L&R
chapter, in spite of Long's advocacy of a reactive
approach for over a decade, he has not as yet produced
reliable empirical findings which demonstrate that if
implemented in real classrooms over, say, a period of
five years, a solely reactive approach can produce
successful language learners. This lack of readily
available empirical evidence is recognized by the authors
(p. 206); yet, in spite of there being substantial
evidence demonstrating the positive potential of
proactive approaches, the authors puzzingly feel
justified in stating (p. 211) "At the present time,
there is no definitive research upon which to base a
choice of one over the other." There probably is not
if one ignores the history of the field.

2. The authors accept that "vocabulary is best acquired
in purely meaning-focused instruction" but add that "it
is likely that focus on form can enhance lexical
acquisition." (p. 212) and cite Paribakht and Wesche
(1997). However, this fails to reveal the fact that the
article cited finds a focus on formS plays an important
role in the most effective means to use. (p. 194-195).
Furthermore, they fail to refer to the work of one of
the leaders in the sub-field of vocabulary learning,
Nation who in Nation (1982:23) (a revew of empirical
research on vocabulary acquisition) states with respect
to the principle that vocabulary should be taught in
context, "this idea remains a statement of belief rather
than a conclusion based on experimental evidence." and
adds further "In fact, we know that words in isolation
are retained very well indeed, both in large quantities
and over long periods.".

Thus, though the chapter is excellent in terms of the
confines of its own remit, it is remiss in failing to
bring out the relevance to the issue of focus on form
of the extensive literature on methodological choices
covering at least this century. Part of that literature
provides evidence of the value of explicit grammar
instruction without its being motivated by some immediate
communicative need. Yet, both this final chapter and
all the others, with the exception of DeKeyser's and that
of Williams and Evans, tend to pay little attention to
this option.


This a timely book for it brings together representative
examples of the focus on form movement providing accounts
of a range of instructional exponents of FONF instruction
and relevant research findings. However, apart from
Dekeyser's chapter, the instructional option of providing
declarative grammatical knowledge in the most direct way
and how it may be used to improve proficiency is not
addressed. It is associated with so-called traditional
methods and, therefore, unjustifiably assumed to be less
effective than the options proposed. This is a pity.
There is ample evidence to demonstrate that options akin
to what DeKeyser proposes can be effective in certain
situations and with certain types of learners. It is,
therefore, imperative that those in the focus on form
movement carry out comparative studies involving what
they propose and other more direct and explicit options
as do Wlliams and Evans in their chapter. It is to
such a strategy that I believe that Allen et al. (1990)
were referring in the context of different forms of
communicatve language teaching in saying: "As the next
step, we would like to undertake research in classes
where more extreme contrasts can be established through
the use of the COLT to see if we can find further
differences in student proficiency under these more
prototypic conditions." (p. 115). It is only. then, and
preferably based on the findings of mutiple replication
studies, that advocates of a focus on form approach will
be able to base their dismissal of options on the basis of
empirical evidence and not on theorizing on the nature of
SLA and flawed contemporary wisdom.


1. Given the length restrictions, this review is a
substantially shortened version of the original.

2. This type of hedging on the nature of a return to
FONF instruction is nicely pointed out by Stern (1990:
97). He writes: "Analytic ("synthetic in L&R terms)
teaching is very familiar, but it is not clearly
formulated. In the treatment studies too, the views about
it are implicit or stated in asides rather than being
specifically defined. The analytic approach does not have
the excitement that is offered by the experential or the
communcative approach. Hence we tend to be negative or
at least ambivalent about it.".

3. The danger of this sort of comparison is emphasized by
Stern (1990: 98) when he writes: "It is also important
that we do not compare experiential teaching at its best
with analytic ("synthetic" in L&R's terms) teaching at its
worst.". This, indeed, what Lightbown does.

4. Lightbown is, of course, aware of this problem of not
having observed what one is reporting upon (1990:82).
However, in this case, she does not make clear that it is
crucially relevant.

5. In actual fact, I suspect that the position expressed
in this chapter is more extreme than the one that
Lightbown truly advocates which is far more eclectic.
Her reservations about isolated grammar lessons are
motivated by a fear of too much of a swing back to FONF
instruction (See Lightbown, 1990: 90-92).
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue