Does Intensive Explicit Grammarinstruction Make All The Difference

Language Teaching Research 10,3 (2006); pp.
297–327
Does intensive explicit grammar

instruction make all the difference?
Ernesto Macaro and Liz Masterman
University of Oxford, UK
This paper investigates the effect of explicit grammar instruction on

grammatical knowledge and writing proficiency in first-year students of
French at a UK university. Previous research suggests that explicit grammar
instruction results in gains in explicit knowledge and its application in
specific grammar-related tasks, but there is less evidence that it results
in gains in production tasks. A cohort of 12 students received a course in
French grammar immediately prior to their university studies in order to
determine whether a short but intensive burst of explicit instruction, a
pedagogical approach hitherto unexamined in the literature, was sufficiently
powerful to bring about an improvement in their grammatical knowledge
and performance in production tasks. Participants were tested at three points
over five months, and the results were compared with a group which did not
receive the intervention. Our results support previous findings that explicit
instruction leads to gains in some aspects of grammar tests but not gains in
accuracy in either translation or free composition. Reasons for these findings
are discussed in relation to theories of language development and the
limitations of working memory.
I Introduction
The research reported in this paper is set within an international debate
on the value of an explicit focus on grammar in second language (L2)
classrooms (see Doughty and Williams, 1998 for a comprehensive review
of the theme). It is also set within a (UK) national modern foreign lan-
guage (MFL) context which has seen repeated calls, by both theorists and
practitioners, for a return to more explicit grammar teaching in schools
(Marsden, 1999; Metcalfe et al., 1998; Mitchell, 2000; Wright, 1999), but
Address for correspondence: Ernesto Macaro, Department of Educational Studies, University of

Oxford, 15 Norham Gardens, Oxford OX2 6PY, UK; email: [email protected]
© 2006 Edward Arnold (Publishers) Ltd 10.1191/1362168806lr197oa

298 Does grammar instruction make all the difference?
with a distinction made between low-and high-achieving students

(McLagan, 1994; Wright, 1999), and for those preparing to study languages
at university (Hurman, 1992; Klapper, 1997; 1998; Sheppard, 1993).
What exactly is meant by teaching grammar explicitly is, of course,
highly dependent on the viewpoint of the person advocating it or other-
wise. For the purposes of this paper, we define it in the way that we
believe is meant by many of the authors above:
Establishing as the prime objective of a lesson (or part of a lesson) the explanation of
how a morphosyntactic rule or pattern works, with some reference to metalinguistic ter-
minology, and providing examples of this rule in a linguistic, though not necessarily a
functional, context.
The research, carried out at the University of Oxford, took place

within the context of a mismatch between the university MFL curricu-
lum’s emphasis on grammatical awareness and the grammatical knowl-
edge of some of the applicants (see Macaro and Wingate, 2004, for
further contextualization). This perceived mismatch may persist even
though there was little difference in applicants’ overall linguistic per-
formance in external examinations in that a criterion for admission to the
university is the achievement of grade A in the nationally standardised
Advanced Level MFL examination.
To demonstrate that they can meet this required standard of grammat-
ical knowledge, applicants wishing to study MFL at Oxford must take
an admissions grammar test (henceforth AGT1) in their chosen lan-
guage(s). The test consists of (inter alia) a gap-fill test and translation of
isolated sentences into the L2. The intensive grammar course which
forms the subject of the present study represented an attempt by the uni-
versity to bridge this perceived ‘grammar gap’ for students who had
achieved the lowest scores in the AGT (some nine months earlier) but
who had nevertheless been accepted on other criteria.
The prime motivation for our study, therefore, was to investigate
whether high-achieving students of French who were about to embark
on a university course, but had performed relatively badly in an admis-
sions grammar test, might benefit from intensive and explicit instruction
in a number of grammatical structures.
With this context in mind, we will now examine two aspects of the
research literature relevant to our study. These are: (i) the relationship
between explicit and implicit knowledge and (ii) the effectiveness of
different approaches to L2 grammar instruction.
Ernesto Macaro and Liz Masterman 299
II Background
1 The role of explicit and implicit knowledge
There has been considerable focus of attention on the relationship
between explicit (analysed) grammatical knowledge and implicit
(unanalysed) grammatical knowledge and how this might relate to lan-
guage development. It is generally accepted that explicit knowledge is
acquired through controlled processes in declarative memory, while
implicit knowledge is acquired through much less conscious or even
subconscious processes.
The implications of these two types of knowledge for L2 instruction
are twofold. First, if grammar is taught explicitly can it then become
automatic so that language can be understood and produced without
constant recourse to the rules that generated the explicit knowledge in
the first place? Conversely, can language that is acquired implicitly be
reflected on if and when the language situation or task demands it?
Second, if different language programmes want to measure these
different types of knowledge, can they be measured validly?
The degree to which implicitly acquired knowledge is accessible has
been questioned (Han and Ellis, 1998), for example in production tasks,
where knowledge and behaviour are not easily distinguishable.
Grammaticality judgement tests (GJT), where subjects have to declare
whether a sentence is correct or incorrect, may be more appropriate for
distinguishing between knowledge and performance. However, whether
such tests really measure implicit (as opposed to explicit) knowledge is
now considered to be dependent on a number of conditions, of which the
pressure to respond within a given time, and the use of ‘rule’ rather than
‘feel’, appear to be the most important (Bialystok, 1979; R. Ellis, 2005).
A difficulty also resides in measuring knowledge about language, in
that learners cannot be said to lack explicit knowledge simply because
they do not possess the required metalinguistic competence to articulate
it. Nevertheless, studies have explored the relationship between metalin-
guistic knowledge and general language proficiency. For example,
Alderson et al. (1997) carried out such research among undergraduates
and found ‘no evidence to support the belief that students with higher
levels of metalinguistic knowledge perform better at French, or that they
improve their French proficiency at higher rates than other students’
(1997: 118) and concluded that ‘metalinguistic knowledge and linguistic
proficiency appear to constitute two separate factors of linguistic ability’

(1997: 118). This pattern of findings recurs in the study by Han and Ellis
(1998), who administered a number of measures, including GJTs and pro-
duction tasks, to 48 adult learners of English (L2). They claimed that
their results lend support to Bialystok’s (1982) view that different types
of language use draw on different types of knowledge and that metalin-
guistic knowledge appears unrelated to general language proficiency.
Ellis (2004) has further argued for a distinction between possessing
the knowledge itself and the ability to verbalize it, regardless of whether
the learner possesses the metalanguage. Does a GJT that also asks the
respondent why a sentence is incorrect provide evidence of explicit
knowledge, or simply an ability to express that knowledge? For
example, Green and Hecht (1992) required 300 German learners of
English to correct 12 errors in 12 separate sentences and to offer
acceptable explanations of the rule. They found that learners were able
to correct the errors without being able to offer an explanation of the
rules, suggesting implicit knowledge. However, no attempt was made to
tap into the learners’ thinking processes which may well have provided
evidence of controlled reflection on language patterns, rather than
implicit knowledge.
These considerations also raise the question as to whether there is a
difference between implicit acquisition and the expression of intuitive
knowledge. Grammatical knowledge that is acquired through subcon-
scious processes is not necessarily the same knowledge that is applied
intuitively when, for example, one is monitoring one’s own production.
All kinds of other bits of input and knowledge may be competing for
prominence when monitoring intuitively. Ellis (2004: 231) has argued
that ‘it may be possible to proceduralize explicit knowledge to the point
that it cannot be easily distinguished from implicit knowledge’. If this is
so, then it may be possible for proceduralized knowledge to be brought
back to selective attention for examination because it once was in declar-
ative form. Although these transformations may be true, they do create
the situation where the relationship between how knowledge was
acquired and how it is demonstrated is not a simple one.
Attention has also centred on the extent to which explicit and implicit
knowledge may contribute to variability between a learner’s competence
and their performance (Ellis, 1985; 1990a; 1990b; Tarone, 1988). In
other words it raises the question of how we deal with the discrepancy
between learners’ knowledge about the L2 and their ability to use the
L2 in different situations (Bialystok, 1982), leading to an uncertainty
regarding how to measure interlanguage development. Macrory and
Stone (2000) investigated pre-intermediate learners’ acquisition of the
perfect tense in French and found not only considerable within-subject
variability in the use or non-use of the auxiliary, but also that subjects
used an auxiliary (both correctly and incorrectly) in gap-fill exercises
while frequently omitting it completely in oral and written production
tasks. On a related theme, Hulstijn and de Graaf (1994) theorized that
explicit knowledge facilitates the acquisition of implicit knowledge
under certain conditions. An interesting speculation of theirs, resulting
from this line of reasoning, is that explicit instruction may have a greater
effect on comprehension tasks than on production tasks.
In sum, although the distinctions between the explicit and implicit
acquisition of knowledge, and between intuitive and reflective demon-
stration of knowledge are still being established, it may nevertheless be
the case that the two types of knowledge, although acquired differently,
do in fact interact in long term memory. This is sometimes known as the
interface position (Ellis, 1994; McLaughlin, 1987), a position also held
by N. Ellis (2005). It is this interface that continues to interest
researchers in that, as a consequence, explicit instruction may indeed
have a part to play in developing L2 acquisition.
2 Is explicit L2 grammar instruction effective?

The debate regarding whether grammar should be taught explicitly has
been a constant one ever since the introduction of the Direct Method in
the late-nineteenth century (Richards and Rogers, 1986), which ques-
tioned its effectiveness. Despite this questioning, explicit grammar
instruction persisted in various forms throughout the twentieth century
even though pedagogies of the 1980s and 1990s continued to view
explicit grammar teaching with some caution as can be seen from the
work of theorists (Prabhu, 1987; Rivers, 1983; Widdowson, 1978); from
syllabuses and exam requirements (ILEA, 1979; Edexcel, 2004; AQA,
2005) and from textbooks (Soars and Soars, 1987; Elston and McLagan,
1997). Space does not allow an in-depth analysis of why the debate has
persisted, but we can summarize the link between pedagogy and
research by identifying two strands of research.
The first strand concerned itself with whether grammatical

development was constrained by innate faculties, as in the series of mor-
pheme studies (e.g. Bailey et al., 1974; Larsen-Freeman, 1976) and as in
the subsequent interest in teachability (Pienemann, 1984) and in devel-
opmental readiness (Spada and Lightbown, 1999; Mackey and Philp,
1998). The second strand concerned itself with the power of input and
interaction to deliver acquisition of the rule system without explicit
grammar instruction (Krashen, 1985; Long, 1981; 1983; Swain, 1985).
Counter to these research strands there has been evidence that, despite
immersion in a language, learners continue to make grammatical errors
(Harley, 1989), make insufficient progress with competence in low-input
courses (Sharwood-Smith, 1981; 1994; Mitchell, 2000), and certain
grammatical forms cannot be acquired solely on the basis of comprehen-
sible input (White, 1987). There has also been the notion that academi-
cally gifted students might benefit (or at least derive enjoyment) from a
teaching style where the focus is on the analysis of the L2 grammar (e.g.
Cook, 2001). This argument that academically gifted students may have
a special aptitude for linguistic analysis, is one very much put forward
by some of the authors cited in the introduction to the context of our
study.
Studies of explicit grammar instruction that are relevant to our
research fall into two categories: (a) those which have simply compared
the relative effectiveness of different instructional approaches on learn-
ers’ explicit grammatical knowledge, and (b) those which have investi-
gated the relationships between instructional approaches and both
explicit grammatical knowledge and production. We will start by
reviewing studies in the first category.
Scott (1989) taught two university-level classes French relative
clauses and the subjunctive using, alternately, an explicit method or an
implicit method, and tested them using aural and written gap-filling
exercises. The post-tests suggested that both classes made more progress
when taught explicitly, but only in the written exercises. However, the
author acknowledges that the students had probably been taught previ-
ously via an explicit approach and would have been familiar with it.
Doughty (1991) investigated the acquisition of English relative clauses
under rule-oriented instruction and under meaning-oriented instruction,
and whether acquisition of harder (‘marked’) structures would facilitate
acquisition of easier structures. She found that both groups improved
significantly against a control group with no advantage for the

rule-oriented instruction. However, instruction in marked relativization
did appear to generalize to the acquisition of less marked aspects of the
form, suggesting some advantage for focus on form (FonF). Fotos and
Ellis (1991) compared explicit instruction in a number of grammatical
elements with simply raising university students’ consciousness of those
elements via communicative pair and group tasks. They concluded, on
the basis of GJTs, that consciousness-raising tasks ‘appear to be an
effective type of classroom activity’ (1991: 623) even though they did
not necessarily result in the same levels of longer-term learning. In a
later investigation, Fotos (1993) found that both instructional
approaches were equally effective in promoting subsequent ‘noticing’
(Schmidt, 1990) of targeted features of the language in written texts and
developing longer term knowledge. No measurement of production was
administered.
An investigation into more general grammar learning and over a
longer period was carried out by Klapper and Rees (2003), who fol-
lowed a sample of 57 students of German over their four-year degree
course. Forty were single- or joint-honours students (Majors) and were
taught through an explicit focus on forms (FonFS) approach. The other
17 were studying German ‘in combination with a degree in Commerce,
Social Science or Law’ (2003: 190) and were taught via a FonF approach.
We are not told the weighting of the non-language part of the degree in
relation to the language part but we are told that both groups received
‘almost identical’ amounts of language instruction. At the start of the
course, there were no significant differences between the two groups, as
measured by a C-test and grammar test. By the end of year 2, the FonFS
group scored significantly higher in the two types of test. However, at
the beginning of year 4, after both groups had studied abroad, the FonF
group significantly outperformed the FonFS group in the tests. So,
despite the clear advantage of the FonFS group in having received an
instructional approach that matched the type of test given, in the long
run, the explicit grammar instruction received during the first two years
of the course was not the key independent variable.
Studies measuring the impact of instruction on both explicit grammat-
ical knowledge and production have produced conflicting findings.
Frantzen (1995) investigated whether explicit grammar teaching and
corrective feedback improved grammatical knowledge and accuracy and
fluency of writing, as measured by a discrete-point grammar test and an

essay before and after the intervention. Both treatment and comparison
groups made significant progress in both areas. However, the experimen-
tal group outperformed the comparison group on the grammar test only.
A similar study but without a comparison group (Manley and Calk,
1997) found that although some error reduction followed the treatment,
there was no holistic improvement in written production.
More positive findings are reported by Leow (1996) who tested under-
graduate beginner students of Spanish after 6 hours and after 35 hours
of formal exposure to the L2. Significant correlations at around the
R ⫽ 0.6 level were registered between the GJTs and production tasks,
suggesting an association between knowledge of the language and per-
formance in it. However, the production tasks were heavily constrained;
i.e. students were posed a set of essentially closed questions, rather than
being required to generate and monitor their own language.
Using an input-processing approach (VanPatten, 1996; VanPatten and
Cadierno, 1993), Benati (2001) investigated the acquisition of the future
tense in Italian by three groups of university students. The first group
was taught via focus on positive evidence of the inflected form in the
input, the second via paradigms to explain the rules followed by output-
based practice, and the third, a control, received non-systematic expo-
sure to the target feature. Both the treatment groups outperformed the
control group in tests of implicit knowledge, explicit knowledge and oral
production. However, in none of the tests did the ‘explicit group’ outper-
form the ‘input processing group’, suggesting no advantage for the
explicit explanation of rules. It should be noted, additionally, that the
future tense in Italian is a comparatively easy rule.
The line of enquiry regarding easy and hard rules was pursued by
Robinson (1996). In a controlled experiment, subjects were taught rules in
four different conditions: they viewed sentences and were told it was a
memory test; they viewed sentences and were told to look for meaning;
they viewed sentences and were told to try to identify rules; they read
through the rules that were the focus of the study, then saw some sentences
and were asked metalinguistic questions about them. Results showed that
simple rules were indeed learnt more easily under all conditions.
Norris and Ortega (2000) concluded, from their meta-analysis of the
relative effectiveness of different instructional approaches, that FonF
and FonFS are both equally effective and durable. However, they
acknowledged the limitations of conclusions based on a meta-analysis

that incorporated vastly different contexts and different outcome meas-
ures. They also pointed to variations in the way specific instructional
approaches were operationalized, the tendency of explicit treatments to
incorporate more instructional components, thereby giving recipients
more exposure to the target form than participants in implicit treatment
groups, the need for longer post-intervention observation periods to
detect significant changes possibly resulting from implicit treatments,
and decontextualized, memory-based outcome measures which de facto
favour the recipients of explicit instruction.
3 Summary
Whilst there is research evidence that some focus on the grammatical
features of the L2 is beneficial to developing the interlanguage of a
learner, the evidence with regard to the explicit teaching of grammatical
features is not sufficiently conclusive to be able to influence pedagogy
directly. Particularly inconclusive is the issue of whether being taught
rules explicitly leads to successful internalization of those rules. This
appears to be linked to the considerable uncertainty over the nature of
the relationship between implicit and explicit knowledge, and between
implicitly acquired knowledge and the application of knowledge which
is, at least in part, intuitive. Nevertheless, the evidence from research
tends to support an interaction between these constructs rather than a
complete dissociation. For that reason alone, it would seem to be worth
continuing to ask questions about explicit grammar instruction.
However, there are additional reasons for continuing to investigate the
effectiveness of explicit grammar teaching.
First, it remains to be determined whether explicit knowledge can
become sufficiently automatic to enable both fluency and accuracy in
production tasks. Conversely, we need to explore whether explicit
knowledge can be easily brought back under selective attention in order
to monitor for possible mistakes in production tasks.
Second, the relationship between implicit and explicit knowledge
appears to account for variability between tasks which measure different
things. Although GJTs are not without controversy with regard to their
validity (Birdsong, 1994; Ellis, 1990a; Gass, 1994), they continue to be
used, alongside other types of tasks, in order to test for grammatical
knowledge of some kind and in a variety of both research and pedagog-

ical contexts. In this study, a GJT was used in combination with other
discrete-point grammar tests to measure knowledge of grammar irre-
spective as to whether it was intuitive or otherwise and compared this
knowledge to that expressed through written production.
Third, evidence from intervention studies suggests that if learners are
taught rules explicitly they will perform better in grammar tests. There
is tentative evidence that learners will perform better in highly structured
production tasks. There is little evidence so far that they will perform
better in less structured production tasks. We compared a discrete-point
grammar test with two different production tasks: one highly structured
and the other less structured.
Fourth, although we have identified many studies involving treatments
carried out over a relatively brief period, we were unable to identify any
that investigated the effect of a short, but intensive burst of explicit
grammar instruction before starting a language course with a strong
grammar focus.
Lastly, to our knowledge, no studies tested this kind of instruction on
comparatively high-achieving language learners: that is, learners who
had been admitted to an ‘elite university’ through a highly selective
admissions process. A related study (Macaro and Wingate, 2004)
showed this particular sub-population to be highly motivated to perform-
ing well on a grammar-oriented first year of a four-year language pro-
gramme at that university. In other words, if this particular type of
intensive, explicit instruction would be effective with anybody, it would
surely be effective with these participants and in these circumstances.
Given the national context, other similar institutions may be considering
an intensive course of this type. It was therefore appropriate for this type
of intervention to be investigated.
III Research questions and rationale

The context outlined above, together with the review of previous
research, led us to ask the following inter-related research questions:
1) Is an intensive course in explicit French grammar given to
high-achieving first-year undergraduates a sufficiently powerful
intervention to bring about (a) an improvement in their grammatical
knowledge, and (b) a reduction in their written production errors?
2) Is any immediate improvement sustained over a longer period?

3) Is any reduction in production errors brought about without a
detrimental effect on other aspects of writing proficiency?
4) Is any detected improvement significantly different from the long-
term progress made by a comparison group not receiving the
intervention?
We laid stress on the concept of sufficiency in formulating these
research questions as we recognized that other factors beyond our con-
trol would almost certainly also influence students’ progress, for exam-
ple, the fact that the students were subsequently taught by different
teachers. This variation in learning experience at university is not exclu-
sive to Oxford. Many universities offer additional supported self-study.
Moreover, independent self-study, over a period of four to five months,
may vary enormously according to a university student’s personal moti-
vation. Thus what we would look for would be reliable evidence that the
intensive grammar course, 30⫹ hours of tuition, made a significant dif-
ference over and above these factors. If the treatment was ‘sufficiently
effective’, it should result in significant improvement, not only in stu-
dents’ ability to demonstrate knowledge in grammar tasks, but also in
controlled and less controlled production tasks. In the less controlled
production tasks, we would look to see whether improved accuracy was
not obtained at the expense of other aspects of writing proficiency. We
formulated research question 2 because of the longitudinal nature of the
study with unequal time points of data collection (see below).
IV Method
1 Participants
Participants were drawn from the population of first-year students study-
ing French at the university. The context of the study required us to
adopt a purposive strategy by recruiting exclusively those students from
a percentile who had achieved the lowest scores in the AGT. The inter-
vention group was recruited by invitation in summer 2003 and numbered
12 students. For the comparison group we identified the tranche of stu-
dents who had achieved the next lowest scores in the AGT and sent them
invitations on their arrival in October 2003. We recruited 10 students
who received payment for taking part. All students in the overall sample
Table 1 Summary of AGT scores
Group N Lowest Highest Mean SD
Intervention 12 2.4 6.0 5.11 1.01

Comparison 10 4.7 6.6 5.79 0.63
had obtained a grade A (the highest possible) in French in the national

examinations at age 18, as required for admission to the university.
Table 1 shows the range of AGT scores within the two groups.
With regard to the AGT scores, a Kolmogorov-Smirnov test showed
that the distribution of the whole sample was significantly non-normal
( p ⫽ 0.013), and so we could only use non-parametric tests for
between-group comparison. However, the distribution within each group
was normal, making it possible to conduct parametric tests for repeated
measures.
Although the students were not randomly allocated to the two groups,
a Mann-Whitney test conducted on their AGT scores showed that the
difference between the two groups was not significant (U ⫽ 35.000,
p ⫽ 0.098). As we shall see, our own pre-tests also did not show any
significant difference between the two groups. This, together with their
results in national examinations, would suggest that the purposive
sampling strategy did not undermine the comparability of the two
conditions.
Students in both groups were free to withdraw from the study at any
time, but none did so.
2 The intensive grammar course

The course tutor, highly thought of and a French native speaker, adopted
a FonFS approach, with a preliminary session in which the tutor rein-
forced students’ knowledge of grammatical terminology. The daily pro-
gramme consisted of three hours of oral and written activities in the
classroom (a total of approximately 30 hours), which were designed to
develop students’ grammatical knowledge by focusing on a list of
grammatical topics.
For example, one 60-minute session observed by the second author on
day 3 of the course comprised the following sequence of activities, in
which the tutor developed the topic of relative pronouns, which he had
begun the previous day, and also revised other grammatical elements
which had been covered in earlier sessions:
1) Working individually or in pairs, students composed sentences
incorporating que, dont, lequel, etc. (continuation of previous day’s
topic) (9 minutes).
2) Tutor used a sentence written by a student in a previous exercise to
explain the agreement of the direct object in relative clauses with
the passé composé (3 minutes).
3) Tutor explained the use of the relative pronoun dont and set an
exercise for students to compose sentences. He then coached indi-
vidual students during the exercise and gave feedback to the whole
class (9 minutes).
4) Tutor set an exercise for students to work in pairs to reconstruct
fragmented sentences involving relative pronouns and the future
and conditional verb tenses (15 minutes). Tutor then asked students
to read out their work and included explanations of grammatical
rules in his feedback. He also asked them to write the sentences in
their notebooks (7 minutes).
5) Tutor set a computer-based exercise for students to read articles on
the Website of Le Monde, identify constructions containing relative
pronouns and copy them into their notebooks (8 minutes).
6) Tutor gave a dictation and asked students to underline occurrences
of grammatical elements studied during this session and on the
previous day, including demonstrative adjectives and agreement of
the past participle (9 minutes).
The daily classroom activities were reinforced by independent study
tasks, including translating, essay writing and memorizing verb forms.
Students also had twice-weekly sessions, in pairs, with a postgraduate
student to help them consolidate what they had learnt.
3 Grammar tests
Three grammar tests were administered to participants during the study
as follows:
• Pre-test: Intervention group – immediately before the start of the

intensive grammar course (September 2003). Comparison group – as
soon as practicable after starting their studies (October 2003).
• Interim test: Intervention group only – one week after the end of the
intensive grammar course and immediately prior to starting their
studies (October 2003).
• Post-test: Both groups: after 11⁄2 terms’ tuition (12 weeks) (February
2004).
The timings enabled us to measure the progress made by the interven-

tion group students, and to gauge the impact of the grammar course plus
term-time tuition versus term-time tuition alone.
Although the intensive course covered a number of areas of French
grammar, the tests concentrated on four categories only: verbs/tenses/
aspect, relative clauses, agreement and prepositions. These had given
rise to the greatest proportion of errors, in the first-year exams in June
2003, according to university teaching staff.
The tests were written by the research team as no standardized tests of
this type at this level were available. They were checked for accuracy,
relevance and consistency, and the ‘model answers’ for Parts 1–3 were
approved by the grammar course tutor. The pre-test was piloted with stu-
dents studying French in the final year of state secondary education and
whose grammatical knowledge could be considered close to that of the
sample. They were able to complete the test in the time allowed and
gained scores that were considered acceptable.
The tests each lasted 11⁄4 hours and had an identical four-part structure.
The structure and scoring system were as follows:
a Grammaticality judgement (Part 1)
• Purpose: Use explicit and/or implicit grammatical knowledge to

identify and correct grammatical errors.
• Structure: 20 sentences (5 per grammatical category), 12 of which
contained a grammatical error.
• Examples: *Elle souhaitait qu’il écrirait plus souvent. (Error in
verb tense: écrivît/écrive). La porte par où vous êtes entrés date du
XIIe siècle. (Testing recognition of use of relative pronoun).
• Scoring: Sentences with errors: 1 mark for identifying error ⫹ 1
mark for correcting it (max. score: 24). Error-free sentences: 1 mark
for leaving sentence unchanged (max. score: 8).
b Error correction and rule explanation (Part 2)
• Purpose: Show evidence of grammatical knowledge by correcting

grammatical errors and show evidence that grammatical rules taught
have been retained.
• Structure: 5 sentences, each containing a grammatical error.
• Example: *Mon oncle a été mordu d’un énorme tigre blanc. (Error:
preposition should be par, not de. Rule: ‘The agent of a passive
verb describing an action or event is always preceded by the prepo-
sition par.’)
• Scoring: Error correction component: up to 2 marks for each error
correction. Rule explanation: 1 mark for correctly stating (or para-
phrasing) the grammatical rule (regardless of whether grammatical
terminology is used) ⫹ 1 mark for the use of some appropriate
grammatical terminology in that explanation. (Max. score: 20.)
c Translation (Part 3)
• Purpose: Test accuracy in students’ productive use of French in a sit-

uation of relatively controlled output.
• Structure: Five sentences, which together contained a total of three
instances of each of the four grammatical categories being tested.
• Examples: The people I met in France during the summer will be
coming to England for Christmas. Les gens que j’ai rencontrés
[connus] en France pendant l’été vont venir [viendront] en
Angleterre pour [à] Noël. (Verb tense: vont venir [viendront].
Agreement: rencontrés [connus]. Preposition: en (two occur-
rences), pour [à]. Relative pronoun: que.)
• Scoring: 1 mark for each correctly translated structure (remainder of
each sentence was disregarded). Incorrect vocabulary was admissi-
ble if the gender was correct. (Max. score: 12.)
d Narrative composition (Part 4)
• Purpose: Test students’ accuracy and general writing proficiency in

a production task in which they were relatively free to form their
own output.
• Structure: Re-tell a story sequence of six pictures as a third-person
narrative in the past tense, in at least 250 words.
• Scoring: (a) Holistic scores: accuracy, lexical diversity, range of

expression, using a nationally standardized marking scheme
(Oxford Delegacy of Examinations) with a maximum score of 10
each; (b) Computed scores: accuracy (percentage of correct
noun and verb phrases); lexical diversity using Measure D (Malvern
et al., 2004).
Part 1 and the error-correction component of Part 2 were scored by the
second author in accordance with the model answers approved by the
course tutor. Students’ rule explanations in Part 2 were open to subjec-
tive interpretation and were therefore scored by both authors, with an
average inter-rater reliability of 75%. Agreement was then reached over
the remaining items.
In the narrative task (Part 4), when scoring the compositions
holistically, the two authors marked the compositions separately,
then reached agreement on the scores by discussion. The computational
methods were intended to triangulate these scores and to counteract
subjective elements in the holistic marking. We were unable, however,
to identify a satisfactory computational measure for range of expression.
All scores were entered on SPSS version 11.
V Results
1 Discrete-point grammar tests within groups
In order to begin answering research questions 1 and 2, we first analysed
the within-group change over time. The discrete-point grammar tests
(Parts 1 and 2) were computed for frequencies and their mean scores
compared at each time point using a paired samples t-test. These
are summarized in Table 2. Part 1 in the table is split in two, to reflect
the different marking schemes used with erroneous and error-free
sentences.
The mean ability of the intervention group to spot and correct errors,
and to provide explanations for their corrections, improved significantly
over the period of the tests. However, it should be noted that the ability
to spot errors improved significantly between interim- and post-test
rather than between pre- and interim-test. Moreover, the improvement in
their ability to recognize error-free sentences was less marked, and
even declined between the interim and post-tests. It should also be noted
Table 2 Within-groups mean scores (and standard deviations) for parts 1 and 2 of
grammar tests.
Cond’n Pre- Interim Post- Pre- to Interim- Pre- to

Interim to Post- Post-
1a. Erroneous I 5.00 7.58 12.54 t ⫽ 1.705 t ⫽ 3.224 t ⫽ 5.387

sentences (3.84) (5.20) (3.90) p ⫽ 0.116 p ⫽ 0.008 p ⫽ 0.000
(max. 24) C 7.90 9.15 t ⫽ 1.168
(0.74) (4.26) p ⫽ 0.273
1b. Error-free I 2.33 4.08 3.17 t ⫽ 2.836 t ⫽ 3.224 t ⫽ 1.820
sentences (1.37) (1.78) (1.70) p ⫽ 0.016 p ⫽ 0.008 p ⫽ 0.096
(max. 8) C 2.90 3.00 t ⫽ 0.218
(2.59) (1.49) p ⫽ 0.832
1c. Combined I 7.33 11.67 15.71 t ⫽ 2.422 t ⫽ 2.226 t ⫽ 6.204
score (3.89) (5.31) (4.15) p ⫽ 0.034 p ⫽ 0.048 p ⫽ 0.000
(max. 32) C 10.80 12.15 t ⫽ 1.125
(4.18) (4.20) p ⫽ 0.290
2. Error I 9.13 9.29 12.63 t ⫽ 0.133 t ⫽ 1.984 t ⫽ 2.440
correction & (3.21) (3.96) (4.21) p ⫽ 0.897 p ⫽ 0.073 p ⫽ 0.033
rule expl’n C 10.75 8.20 t ⫽ 1.980
(max. 20) (2.06) (3.29) p ⫽ 0.079
Notes : Boldface indicates a significant difference.

I ⫽ Intervention condition; C ⫽ Comparison condition.
that the intervention group’s ability to correct errors and provide rule
explanations improved significantly only between pre- and post-test, and
not between pre- and interim test.
In contrast, the comparison group’s scores did not improve signifi-
cantly and, in the error correction and rule explanation task, declined.
Standard deviations indicate some wide variations in individual scores
and this issue will be address later.
2 Production tasks within groups

Table 3 summarizes the results of the translation.
Neither group improved significantly between pre- and post-test, and
there was even a slight decline in the intervention group’s mean per-
formance between interim and post-test. Analysis of individual scripts
confirmed that students rarely used avoidance strategies (i.e. simplifying
the text to be translated). This was detected in only 15 out of 672 cases
Table 3 Within-groups mean scores (and standard deviations) for translation (part 3)
Cond’n Pre- Interim Post- Pre- to Interim- Pre- to

3. Translation
(max.12) I 6.58 7.12 6.92 t ⫽ 1.096 t ⫽ 0.405 t ⫽ 0.471
(1.72) (1.90) (2.19) p ⫽ 0.297 p ⫽ 0.693 p ⫽ 0.647
C 7.35 7.80 t ⫽ 0.927
(1.73) (1.75) p ⫽ 0.378
(i.e. 2.23%). In other words, the translation was indeed measuring

controlled output.
Table 4 summarizes both the holistic and the computed scores for all
the assessments performed on students’ narrative compositions, showing
that, overall, both groups’ performance improved significantly over the
period of the study, and with generally smaller variations than in the pre-
ceding three tasks. It should be noted that the intervention group did not
improve on grammatical accuracy (either measured holistically or com-
puted) between pre- and interim tests.
In terms of within-groups differences, the computational scores pro-
duced similar scores to the holistic marking, with the exception of the
intervention group’s accuracy, in which the computation detected a
slight dip from pre- to interim test, and the comparison group’s lexical
diversity, which the Measure D calculation does not show as having
expanded significantly.
3 Between-group comparisons
In order to further answer research questions 1, 2 and 3, and specifically
answer question 4, we carried out between group comparisons.
Mann-Whitney calculations were performed for the reasons given
above. Table 5 shows the results for each part of the tests. The second
and fourth columns show which group obtained the higher mean
score for the relevant part (for the actual figures, please refer to the
preceding tables).
In the pre-test, none of the differences were significant between the
two conditions, reflecting the absence of a significant difference in the
AGT scores of the two groups. In the post-test, there was a significant
Table 4 Within-groups mean scores (and standard deviations) for narrative composi-
tion (part 4)
Condn Pre- Interim Post- Pre- to Interim- Pre- to

4a. Holistic: I 4.93 5.50 6.50 t ⫽ 1.865 t ⫽ 5.745 t ⫽ 4.710

accuracy (1.56) (0.80) (0.67) p ⫽ 0.089 p ⫽ 0.000 p ⫽ 0.001
(max. 10) C 4.70 6.80 t ⫽ 6.034
(1.06) (0.79) p ⫽ 0.000
4b. Holistic: I 5.58 6.50 6.92 t ⫽ 3.527 t ⫽ 1.449 t ⫽ 4.690
lexical (0.90) (0.67) (0.67) p ⫽ 0.005 p ⫽ 0.175 p ⫽ 0.001
diversity C 6.00 7.40 t ⫽ 3.500
(max. 10) (1.25) (0.84) p ⫽ 0.007
4c. Holistic: I 5.17 6.08 6.75 t ⫽ 2.727 t ⫽ 2.602 t ⫽ 5.062
range of (1.27) (0.79) (0.62) p ⫽ 0.020 p ⫽ 0.025 p ⫽ 0.000
exprn C 5.14 6.70 t ⫽ 3.748
(max. 10) (1.10) (0.68) p ⫽ 0.005
4d. Computed: I 66.58 65.72 73.12 t ⫽ 0.398 t ⫽ 3.674 t ⫽ 3.401
accuracy (9.38) (7.51) (8.33) p ⫽ 0.698 p ⫽ 0.004 p ⫽ 0.006
(max. 100%) C 64.64 73.96 t ⫽ 3.821
(5.85) (6.11) p ⫽ 0.004
4e. Computed: I 70.45 85.25 90.40 t ⫽ 3.157 t ⫽ 0.791 t ⫽ 4.620
lexical (12.45) (12.18) (16.89) p ⫽ 0.009 p ⫽ 0.446 p ⫽ 0.001
diversity: C 73.84 83.70 t ⫽ 1.794
(no max.) (11.20) (18.70) p ⫽ 0.106
Notes : Boldface indicates a significant difference.

I ⫽ Intervention condition; C ⫽ Comparison condition.
between-groups difference in one score only, the combined error

correction and rule explanation task, and this may be accounted for in
part by an actual decline in the comparison group’s mean performance
(see Table 1).
To check whether these results were distorted by the small sample
sizes, we calculated the effect size – Cohen’s d – for each task in the
post-test by dividing the differences between the means of the two
groups by the pooled standard deviation of both groups, as advised by
Norris and Ortega (2000). A d value approaching 1 is deemed large
(i.e. the findings carry weight). In our study, this was the case only
with the ‘erroneous sentence’ component of Part 1 (d ⫽ 0.835), where
the Mann-Whitney test approached significance, and Part 2
(d ⫽ 1.167), although the latter figure needs to be treated with caution
Table 5 Statistical analysis of the differences between the two groups in their pre-
and post-test scores
Pre-test: Pre-test: Post-test: Post-test:

higher between-groups higher between-groups
mean score difference mean score difference
1a. Erroneous C U ⫽ 38.500 I U ⫽ 31.500

sentences p ⫽ 0.159 p ⫽ 0.059
1b. Error-free C U ⫽ 45.500 I U ⫽ 56.500
sentences p ⫽ 0.346 p ⫽ 0.821
1c. Combined C U ⫽ 34.500 I U ⫽ 33.000
score (1a ⫹ 1b) p ⫽ 0.093 p ⫽ 0.080
2. Error correction C U ⫽ 40.500 I U ⫽ 25.500
& rule explanation p ⫽ 0.203 p ⫽ 0.021
3. Translation C U ⫽ 44.00 C U ⫽ 52.50
p ⫽ 0.314 p ⫽ 0.628
4a. Holistic: I U ⫽ 57.50 C U ⫽ 47.00
accuracy p ⫽ 0.872 p ⫽ 0.418
4b. Holistic: C U ⫽ 51.00 C U ⫽ 41.00
lexical diversity p ⫽ 0.582 p ⫽ 0.228
4c. Holistic: I U ⫽ 57.00 I U ⫽ 57.00
range of expression p ⫽ 0.872 p ⫽ 0.872
4d. Holistic: total C U ⫽ 59.00 C U ⫽ 48.00
p ⫽ 0.974 p ⫽ 0.628
4e. Computed: C U ⫽ 57.00 I U ⫽ 52.00
accuracy p ⫽ 0.872 p ⫽ 0.628
4f. Computed: C U ⫽ 46.00 I U ⫽ 45.00
lexical diversity p ⫽ 0.381 p ⫽ 0.346
Note : Boldface indicates a significant difference.
as it lies just outside upper bound (1.13) of the 95% confidence level
computed by Norris and Ortega in their meta-analysis of metalin-
guistic tasks. In other words, sample size did not appear to influence
results.
4 The intensive grammar course: a sufficiently powerful

intervention?
On the basis of the results reported so far, any impact of the intensive
grammar course appears to have been limited to an improvement in stu-
dents’ ability to identify and correct ungrammatical sentences and to
identify and explain errors when sensitized to their presence (as shown
by scores from Parts 1 and 2 of the tests). In our research questions,

however, we drew attention to the concept of sufficiency and to the
notion of sustained improvement.
To shed light on these, we asked the whole sample, after the post-test
(i.e. four months into their studies), whether they had attended weekly
grammar classes organized centrally by the University in addition to the
language tuition which they received. Thirteen said that they had. We
then mapped students’ total scores for Parts 1 and 2 only in each test to
their response to this question, as shown in Figure 1.
A feature of Figure 1 is the contrast between the general pre- to post-
test improvement shown by the intervention group, and the mixed per-
formance by the comparison group students, a substantial number of
Figure 1 Students’ total scores for parts 1 and 2 in the tests (grammaticality judge-
ment and error correction/rule explanation). Asterisks denote students who attended
additional grammar classes.
whom actually performed worse on post-test after four months of uni-

versity instruction.
With the exception of student I05, the students in the intervention
group appear to fall into two roughly equal categories in relation to their
improvement in the discrete-point grammar tasks: those who showed the
most improvement immediately after the grammar course, i.e. at interim
test (I01, I08, I03, I04, I12) and those who showed little immediate
improvement (or even an initial decline) but subsequently leaped forward
(I02, I06, I09, I11, I10, I07). For students in the second category, either
the course had a delayed effect or its benefit was combined with subse-
quent experience of teaching and study. Effect size calculations using the
intervention group’s pre-, interim and post-test scores from Figure 1 were
even more explicit, with Cohen’s d calculated as 0.696 (i.e. moderate to
strong) for the contrast between pre- and interim tests and 1.109 (i.e. very
strong) for the contrast between interim and post-tests.
It would appear that many of the ‘most-improved’ students were also
attending the extra grammar classes. Indeed a significant correlation was
discovered between the claimed attendance and the percentage improve-
ment in students’ test scores between October 2003 and February 2004
(i.e. interim to post-test for the intervention group, pre- to post-test for
the comparison group): Spearman’s rho ⫽ 0.503, p ⫽ 0.017, N ⫽ 22.
VI Discussion
The study described in this paper set out to answer four inter-related
questions regarding whether an intensive course in French grammar
given to high-achieving first-year undergraduates, prior to starting their
degree programme proper, was a sufficiently powerful intervention to
bring about an improvement in their grammatical knowledge, both in the
short and long term, a reduction in their production errors without any
detrimental effect on other aspects of written production, and these
improvements as compared to a group who did not receive the intensive
course.
Our findings suggest that the intensive grammar course was not a
sufficient factor to bring about a significant improvement in their gram-
matical knowledge as there was no greater ability to make judgements
overall of grammaticality when compared to the comparison group.
There was, however, evidence of a greater ability to correct errors in
sentences once these had been judged ungrammatical. Whether students

were applying explicit knowledge gained from the course, or whether
they were correcting ‘by feel’ cannot be ascertained, as there was no
apparent parallel improvement in their ability to articulate rules.
The intensive grammar course did not bring about a reduction in pro-
duction errors in either controlled or uncontrolled production tasks.
There is no evidence that explicit grammar taught on the course led to
effective production monitoring. Our earlier review would suggest that
this production monitoring could have resulted from one or more of
three possible factors: (i) the proceduralization of explicit knowledge so
that certain elements would now ‘feel’ incorrect; (ii) the ‘bringing back’
of proceduralized knowledge under selective attention and the conse-
quent application of a particular rule; (iii) the direct application of
explicit knowledge which had never become proceduralized.
Interestingly, the intervention group did not outperform the comparison
group in the translation task, a controlled output task that tested the same
range of grammatical structures as the GJT.
With specific regard to research question 2, we investigated whether
any immediate improvement resulting from the intervention was sus-
tained over the longer period. In the majority of measures there was no
significant improvement in the intervention group’s performance on the
interim-test. We have noted above that the intervention group was
divided roughly into those who made some progress immediately after
the course and those whose progress was delayed. For those students
who demonstrated an immediate improvement after the intervention,
this was not sustained over the longer period. Since the interim test was
taken only one week after the end of the grammar course, it is clear that
the short-term impact of explicit instruction was minimal and, in some
cases, negative. On the other hand some students did appear to make
progress between interim and post-test but we attribute this to a possible
awareness raising brought about by the intervention, which was then
consolidated by later grammar study. In the absence of other data, we
can only speculate that individual differences may be the cause of the
different effects of the intervention. Developmental readiness for the
acquisition of certain elements of grammar may have been spurred by
the intervention and consolidated only if those elements were reinforced
through attendance at subsequent grammar classes or through self-study
with a grammar focus. If true, this speculation would lend support to
previous research on developmental readiness, although a qualitative

analysis through case studies of each student’s interlanguage develop-
ment would be necessary to provide the desired insights.
Our findings, then, provide tentative support to studies previously
cited (such as Green and Hecht, 1992) that there may be some link
between explicit learning and its application in discrete-point grammar
tasks. However, this evidence appears to hold for only some students
rather than being a relative but general trend in all students. For some
students it is possible that a burst of intensive exposure to a large number
of grammatical forms and rules may actually have served to confuse
them in the short term.
Our findings also lend support to the conclusions of previous studies
(e.g. Bialystok, 1982) that linguistic knowledge manifests itself differ-
ently according to task requirements. We should note that in our case this
applies to both the highly structured translation task and to the freer
composition. The question is why?
One possible explanation may reside in the limitations associated with
selective attention and working memory (Baddeley, 1986; 1997) and in
the effect of multi-tasking (Skehan, 1998). When attempting to identify
incorrect sentences, the subject’s cognitive processes are almost entirely
focused on identifying errors, while in production tasks the attention
is distributed over a number of problems to be solved almost
simultaneously.
A second explanation may be related to variability in interlanguage
(Ellis, 1985) and to retrieval processes. Allowing for the co-existence of
more than one form for a particular element in long-term memory, dif-
ferent psychological mechanisms may be involved (a) in processing an
element produced by others and perceived through the visual medium,
and (b) in processing an element retrieved and then produced by the self,
monitored (at least in the first instance) through the phonic medium for
correctness. In other words, a receptive skill such as reading may allow
for more direct trace connections to the stronger of the systematic vari-
ables than do production skills. The stronger of the systematic variables
may well have been influenced by visually displayed positive input. We
should also note that, when reading, the learner decodes first via visual
connections and then, if necessary, via phonic connections (especially in
a GJT). In contrast, when rehearsing for production a subject may be
encoding via phonic connections before visual connections. Given the
lack of orthographic transparency of the French language, this feature

may be particularly accentuated by language specificity. A related expla-
nation might be that, when reading, a learner is being exposed directly
to sentences and asked to judge their correctness, while in production
they may first be generating phrases or utterances which compete with
subsequent sentences being monitored (either in rehearsal or on the
page). These observations support the earlier conclusions by Hulstijn
and de Graaf (1994) cited above. Further research may illuminate these
hypotheses.
Our prime concern in measuring range of expression and lexical
diversity was to ensure that other aspects of the writing process were not
adversely affected by explicit grammar tuition in case the post-test
results proved more positive for grammatical accuracy in free produc-
tion (research question 3). Since this was not so, these data are largely
redundant for the purposes of this paper.
VII Limitations
Our study, of course, was not without its limitations. These were
imposed largely by the non-random nature of the selection and the small
number of participants. Both of these factors lay outside the researchers’
control due to the ‘remedial purpose’ of the grammar course and the
reliance on volunteers for participation in the comparison group. It was
thus a matter of fortune rather than of deliberate design that the differ-
ences in the AGT scores between the two groups proved statistically
non-significant, as did the pre-tests. Even though effect size calculations
reinforced the significance of our principal findings, our lack of control
over other variables – such as the quantity and quality of regular
language teaching received by students and individual learner differ-
ences in approaches to self-study – means that we cannot generalize
the results with confidence beyond our sampling frame of Oxford MFL
students.
VIII Conclusions
The intensive course of explicit grammar teaching was not a sufficiently
powerful independent variable in bringing about the intended structural
change in the intervention group’s interlanguage. In other words, it did
not ‘make all the difference’.
There are a number of implications that we feel we can draw despite

the limitations above.
First, this type of course may be inappropriate – and even
ineffective – for university courses where the prime objective is students’
development of the four language skills. However, it might prove
useful for university courses which emphasize what we might call
an awareness of the structures of language, if only to boost the
confidence of students who have not received this kind of explicit tuition
previously.
Second, our findings do not lend support to the argument that it is the
linguistically most able students who are most likely to benefit from
intensive and explicit grammar instruction. Since exposure to language
analysis in an intensive course such as this did not result in a great
improvement in grammatical awareness, and none in production, the
implication is that this sub-population of students should not be treated
differently unless for course-specific purposes.
Third, since our findings support previous evidence which has
suggested that the development of grammatical accuracy (i) cannot
easily be hurried, (ii) is individually developed, and (iii) requires
continuous exposure to both positive and negative evidence in both
receptive and productive tasks, then the ‘short, sharp shock’ of an
intensive two-week course does not match these developmental patterns.
However, this does not necessarily mean that in combination with
other pedagogical interventions it might not usefully contribute to
development. For example, were it to be associated with subsequent
sustained FonF instruction, it might indeed prove beneficial. On the
basis of our findings, we hope to be able to investigate this possibility in
the future.
Acknowledgements
We acknowledge funding for both the grammar course and the associ-
ated research study from the Higher Education Funding Council for
England (HEFCE). We also thank the following for their contribution to
the study: Michaël Abecassis, Kate Tunstall, Ursula Wingate, Suzanne
Graham, Brian Richards, Robert Vanderplank and Lynn Erler.
We would also like to acknowledge the two anonymous reviewers for
their constructive comments in helping to shape this article.
Note
1 It should be made clear that neither the authors nor the university consider the AGT
a standardized test of grammatical knowledge. It is merely a component in
the admissions process designed to provide an indicator of current grammatical
knowledge.
IX References
Alderson, J.C., Clapham, C. and Steel, D. 1997: Metalinguistic knowledge,
language aptitude, and language proficiency. Language Teaching
Research 1(2): 93–121.
AQA 2005: Qualifications and subjects. Retrieved 24 February 2005 from
http://www.aqa.org.uk/qual/.
Baddeley, A. 1986: Working memory. Clarendon Press.
—— 1997: Human memory: theory and practice. Psychology Press.
Bailey, N., Madden, C. and Krashen, S. 1974: Is there a ‘natural
sequence’ in adult second language learning? Language Learning 21:
235–43.
Bautier-Castaing, E. 1977: Acquisition comparée de la syntaxe du Français
par des enfants francophones et non francophones. Études de linguis-
tique appliquée 27: 19–41.
Benati, A. 2001: A comparative study of the effects of processing instruction
and output-based instruction on the acquisition of the Italian future
tense. Language Teaching Research 5(2): 95–127.
Bialystok, E. 1979: Explicit and implicit judgements of L2 grammaticality.
Language Learning 29(1): 81–103.
—— 1982: On the relationship between knowing and using forms. Applied
Linguistics 3: 181–206.
Birdsong, D. 1994: Decision making in second language acquisition. Studies
in Second Language Acquisition 16(2): 169–82.
Brown, R. 1973: A first language: the early stages. Harvard University Press.
Cook, V. 2001: Second language learning and language teaching, third
edition. Arnold.
Doughty, C. 1991: Second language instruction does make a difference.
Studies in Second Language Acquisition 13(4): 431–69.
Doughty, C. and Williams, J., editors, 1998: Focus on form in classroom
second language acquisition. Cambridge University Press.
Dulay, H. and Burt, M. 1974: Natural sequences in child second language
acquisition. Language Learning 24: 253–78.
Edexcel 2004: Edexcel qualifications. Retrieved 24 February 2005 from
http://www.edexcel.org.uk/qualifications/.
Ellis, N. 2005: At the interface: dynamic interactions of explicit and implicit

language knowledge. Studies in Second Language Acquisition 27(2):
305–52.
Ellis, R. 1985: Sources of variability in interlanguage. Applied Linguistics
6(2): 118–31.
—— 1990a: Grammaticality judgements and learner variability. In Burmeister,
H. and Rounds, P.L., editors, Variability in second language acquisition:
Proceedings of the Tenth Meeting of the Second Language Research
Forum. University of Oregon, Department of Linguistics, 25–60.
—— 1990b: A response to Gregg. Applied Linguistics 11: 384–91.
—— 1991: Grammaticality judgements and second language acquisition.
Studies in Second Language Acquisition 13: 161–86.
—— 1994: The study of second language acquisition. Oxford University
Press.
—— 2004: The definition and measurement of L2 explicit knowledge.
Language Learning 54(2): 227–75.
—— 2005: Measuring implicit and explicit knowledge of a second language:
a psychometric study. Studies in Second Language Acquisition 27(2):
141–72.
Elston, T. and McLagan, P. 1997: Génial. Oxford University Press.
Fotos, S. 1993: Consciousness-raising and noticing through focus on form:
grammar task performance vs. formal instruction. Applied Linguistics
14(4): 385–407.
Fotos, S. and Ellis, R. 1991: Communicating about grammar: a task-based
approach. TESOL Quarterly 25: 605–28.
Frantzen, D. 1995: The effects of grammar supplementation on written
accuracy in an intermediate Spanish content course. Modern Language
Journal 79(3): 329–55.
Gass, S. 1994: The reliability of second language grammaticality judge-
ments. In Tarone, E., Gass, S. and Cohen, A.D., editors, Research
methodology in second language acquisition. LEA, 303–22.
Green, P. and Hecht, K.-H. 1992: Implicit and explicit grammar: an
empirical study. Applied Linguistics 13(4): 168–84.
Han, Y. and Ellis, R. 1998: Implicit knowledge, explicit knowledge and
general language proficiency. Language Teaching Research 2: 1–23.
Harley, B. 1989: Functional grammar in French immersion: a classroom
experiment. Applied Linguistics 10: 331–59.
Hulstijn, J. and de Graaff, R. 1994: Under what conditions does explicit
knowledge of a second language facilitate the acquisition of implicit
knowledge? AILA Review 11: 97–112.
Hurman, J. 1992: Performance in the A level speaking test by candidates

with GCSE training: oral examiners’ views. Language Learning
Journal 5: 8–10.
ILEA 1979: Graded objectives for modern foreign languages. HMSO.
Klapper, J. 1997: Language learning at school and university: the great
grammar debate continues (I). Language Learning Journal 16: 22–27.
—— 1998: Language learning at school and university: the great grammar
debate continues (II). Language Learning Journal 18: 22–29.
Klapper, J. and Rees, J. 2003: Reviewing the case for explicit grammar
instruction in the university foreign language learning context.
Language Teaching Research 7(3): 285–314.
Krashen, S. 1982: Principles and practice in second language learning and
acquisition. Pergamon.
Krashen, S. D. 1985: The input hypothesis: issues and implications.
Longman.
Larsen-Freeman, D. 1976: An exploration of the morpheme order of second
language learners. Language Learning 26: 125–34.
Leow, R. 1996: Grammaticality judgements tasks and second-language
development. In Alatis, G., Straehle, C., Rankin, M. and Gallenberger,
B., editors, Georgetown University round table on language and lin-
guistics 1996. Georgetown University Press, 126–39.
Long, M. 1981: Input, interaction and second language acquisition. In
Winitz, H., editor, Native language and foreign language acquisition.
Annals of the New York Academy of Sciences, Vol. 379. New York
Academy of Sciences, 259–78.
—— 1983: Native speaker/non-native speaker conversation and the
negotiation of comprehensible input. Applied Linguistics 4: 126–41.
Macaro, E. and Wingate, U. 2004: From sixth form to university: motivation
and transition among high achieving state-school language students.
Oxford Review of Education 30(4): 467–89.
Mackey, A. and Philp, J. 1998: Conversational interaction and second
language development: recasts, responses and red-herrings? Modern
Language Journal 82: 283–324.
McLagan, P. 1994: Grammar and the ‘less able’ learner. In King, L. and
Boaks, P., editors, Grammar! A conference report. CILT, 69–73.
McLaughlin, B. 1987: Theories of second language learning. Edward
Arnold.
Macrory, G. and Stone, V. 2000: Pupil progress in the acquisition of the per-
fect tense in French: the relationship between knowledge and use.
Language Teaching Research 4(1): 55–82.
Malvern, D., Richards, B., Chipere, N. and Durán, P. 2004: Lexical diver-
sity and language development: quantification and assessment. Palgrave.
Manley, J. and Calk, L. 1997: Grammar instruction for writing skills: do stu-
dents perceive grammar as useful? Foreign Language Annals 39: 73–83.
Marsden, R. 1999: Go for the ‘a-ha!’ factor in grammar learning. Deutsch:
Lehren und Lernen 19: 15–16.
Metcalfe, P., Laurillard, D. and Mason, R. 1998: ‘It’s just a word’: pupils’
perceptions of verb form and function. Language Learning Journal 17:
14–20.
Mitchell, R. 2000: Applied linguistics and evidence-based classroom
practice: the case of foreign language grammar pedagogy. Applied
Linguistics 21(3): 281–303.
Norris, J.M., and Ortega, L. 2000: Effectiveness of L2 instruction: a
research synthesis and qualitative meta-analysis. Language Learning
50: 417–528.
Pienemann, M. 1984: Psychological constraints on the teachability of lan-
guages. Studies in Second Language Acquisition 6(2): 186–214.
Prabhu, N.S. 1987: Second language pedagogy. Oxford University Press.
Richards, J.C. and Rogers, T.S., editors, 1986: Approaches and methods in
language teaching. Cambridge University Press.
Rivers, W.M. 1983: Communicating naturally in a second language: theory
and practice in language teaching. Cambridge University Press.
Robinson, P. 1996: Learning simple and complex second language rules
under implicit, incidental, rule-search and instructed conditions. Studies
in Second Language Acquisition 18(1): 27–68.
Schmidt, R. 1990: The role of consciousness in second language learning.
Applied Linguistics 11: 129–58.
Scott, V. 1989: An empirical study of explicit and implicit teaching strategies
in French. Modern Language Journal 72: 14–22.
Sharwood-Smith, M. 1981: Consciousness raising and the second language
learner. Applied Linguistics 2: 159–69.
—— 1994: Second language learning: theoretical foundations. Longman.
Sheppard, R. 1993: Getting down to brass syntax: German teaching and the
great standards debate. German Teaching 8: 3–9.
Skehan, P. 1998: A cognitive approach to language learning. Oxford
University Press.
Soars, J. and Soars, L. 1987: Headway. Oxford University Press.
Spada, N. and Lightbown, P.M. 1999: Instruction, first language influence,
and developmental readiness in second language acquisition. Modern
Language Journal 83(1): 1–22.
Swain, M. 1985: Communicative competence: some roles of comprehensible

input and comprehensible output in its development. In Gass, S. and
Madden, C., editors, Input and second language acquisition. Random
House, 235–53.
Tarone, E. 1988: Variation in interlanguage. Edward Arnold.
VanPatten, B. 1996: Input processing and grammar instruction. Ablex.
VanPatten, B. and Cadierno, T. 1993: Explicit instruction and input process-
ing. Studies in Second Language Acquisition 15: 225–43.
White, L. 1987: Against comprehensible input: the input hypothesis and the
development of second language competence. Applied Linguistics 8:
95–110.
Widdowson, H.G. 1978: Teaching language as communication. Oxford
University Press.
Wright, M. 1999: Grammar in the languages classroom: findings from
research. Language Learning Journal 19: 33–39.

Does Intensive Explicit Grammarinstruction Make All The Difference

Uploaded by

Copyright:

Available Formats

Does Intensive Explicit Grammarinstruction Make All The Difference

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Does Intensive Explicit Grammarinstruction Make All The Difference

Uploaded by

Copyright:

Available Formats

Language Teaching Research 10,3 (2006); pp.

Does intensive explicit grammar

This paper investigates the effect of explicit grammar instruction on

Address for correspondence: Ernesto Macaro, Department of Educational Studies, University of

© 2006 Edward Arnold (Publishers) Ltd 10.1191/1362168806lr197oa

with a distinction made between low-and high-achieving students

The research, carried out at the University of Oxford, took place

proficiency appear to constitute two separate factors of linguistic ability’

2 Is explicit L2 grammar instruction effective?

The first strand concerned itself with whether grammatical

significantly against a control group with no advantage for the

fluency of writing, as measured by a discrete-point grammar test and an

acknowledged the limitations of conclusions based on a meta-analysis

knowledge of some kind and in a variety of both research and pedagog-

III Research questions and rationale

2) Is any immediate improvement sustained over a longer period?

Group N Lowest Highest Mean SD

Intervention 12 2.4 6.0 5.11 1.01

had obtained a grade A (the highest possible) in French in the national

2 The intensive grammar course

• Pre-test: Intervention group – immediately before the start of the

The timings enabled us to measure the progress made by the interven-

a Grammaticality judgement (Part 1)

• Purpose: Use explicit and/or implicit grammatical knowledge to

b Error correction and rule explanation (Part 2)

• Purpose: Show evidence of grammatical knowledge by correcting

• Purpose: Test accuracy in students’ productive use of French in a sit-

d Narrative composition (Part 4)

• Purpose: Test students’ accuracy and general writing proficiency in

• Scoring: (a) Holistic scores: accuracy, lexical diversity, range of

Cond’n Pre- Interim Post- Pre- to Interim- Pre- to

1a. Erroneous I 5.00 7.58 12.54 t ⫽ 1.705 t ⫽ 3.224 t ⫽ 5.387

Notes : Boldface indicates a significant difference.

2 Production tasks within groups

Cond’n Pre- Interim Post- Pre- to Interim- Pre- to

(i.e. 2.23%). In other words, the translation was indeed measuring

Condn Pre- Interim Post- Pre- to Interim- Pre- to

4a. Holistic: I 4.93 5.50 6.50 t ⫽ 1.865 t ⫽ 5.745 t ⫽ 4.710

Notes : Boldface indicates a significant difference.

between-groups difference in one score only, the combined error

Pre-test: Pre-test: Post-test: Post-test:

1a. Erroneous C U ⫽ 38.500 I U ⫽ 31.500

Note : Boldface indicates a significant difference.

4 The intensive grammar course: a sufficiently powerful

by scores from Parts 1 and 2 of the tests). In our research questions,

whom actually performed worse on post-test after four months of uni-

sentences once these had been judged ungrammatical. Whether students

previous research on developmental readiness, although a qualitative

lack of orthographic transparency of the French language, this feature

There are a number of implications that we feel we can draw despite

Ellis, N. 2005: At the interface: dynamic interactions of explicit and implicit

Hurman, J. 1992: Performance in the A level speaking test by candidates

Swain, M. 1985: Communicative competence: some roles of comprehensible

You might also like