On one of his frequent trips to Japan, Rod Ellis, author of SLA Research
and Language Teaching and co-author of Impact Grammar, agreed
to an interview with Kent Hill. Rod is presently director of the Institute
of Language Teaching and Learning at The University of Auckland.
KH: The gap between SLA research and teaching pedagogy is recurrently
mentioned in SLA Research and Language Teaching. What can materials
development do to narrow the gap?
RE: A possible way is through materials development. In fact there is,
particularly in British applied linguistics, a long tradition of researchers
or applied linguists involving themselves in materials development. Perhaps
the most obvious example is Widdowson's involvement in the Focus
series, which was an English for Specific Purposes (ESP) set of books. The
approach to teaching ESP that those books embodied derives directly from
Widdowson's own theory of language teaching as reflected in Teaching
Language as Communication and other publications that he produced at
the end of the 19 70s and beginning of the 1980s. So Widdowson is a classic
example of someone prepared to suggest ways in which his theoretical ideas
can be fed into language teaching, and the approach he followed is materials
development.
I've tried something similar. Much of my recent work has been on form-focused
instruction, in particular how one can do form-focused instruction in a
way that might be compatible with how learners learn. This led recently
to the development of grammar teaching materials in conjunction with Stephen
Gaies.
KH: Should materials be required to have some form of validity or credibility
the same as all research has?
RE: People can do materials evaluation for two rather different purposes.
They can carry out an evaluation to decide whether the materials achieve
what they are supposed to achieve; that's a kind of aims evaluation. Or,
alternatively, they can carry out an evaluation because they want to obtain
some understanding of what students do when they use the materials and to
improve the materials. With the first aim, the issues of reliability and
validity do apply, because if your evaluation is not valid or reliable,
then you can't say whether the materials are or are not achieving their
aim. But with the second aim, the issues of reliability and validity are
not really so crucial. What is more crucial is informativeness, the extent
to which the evaluation that you've undertaken has given you helpful insights
as to what's going on when you use these materials with students.
KH: Would a teacher's conducting action research-- in the sense of a
materials microevaluation of a task--be useful to SLA research and teaching
pedagogy?
RE: Materials evaluation can also be of two kinds: it can take the form
of a macroevaluation, where we might try to evaluate a whole book or a whole
set of materials designed for a particular course, or it can involve what
I call microevaluation, where we take a particular class or perhaps the
materials needed for a single lesson and proceed to subm it those to a fairly
detailed evaluation What I've suggested is that doing microevaluations of
materials is potentially an effective way of doing action research. Teachers
very often think of their teaching in terms of what materials am I going
to use tomorrow? and therefore asking teachers to do action research
following the traditional idea -- identifying problems and possible solutions
to problems -- is not something that perhaps comes easily to them.
So, instead of taking the problem-based approach, teachers could take
a materials-based approach to action research. That is, to find out a bit
more about what happens when students actually do a task. They can base
their action research on some particular task, to try to find out what's
going on, or to find out whether the aims of the task are actually being
met when the task is used in the classroom. In general, my postgraduate
students found microevaluations very revealing. They enabled them to get
inside their teaching and to discover things that they had not been aware
of before. They also led them to propose ways in which they could modify
the task, in order to try it again -- so in that respect, the microevaluations
worked very well as action research.
KH: You're a researcher, a teacher educator, and a materials developer.
How do you see these three roles to be interrelated?
RE: In my work in general, these things have been interrelated. First,
as a researcher, I've tended to research issues that are of theoretical
interest in SLA research but are also of potential pedagogical importance.
The same is true for my work as a materials developer, although I have
to be pretty honest here: When one develops materials, there is always a
tension between the sorts of materials that one might want to develop on
the basis of one's theoretical understanding of SLA and the kinds of materials
that publishers think that they can sell. There are necessary compromises
that have to be made.
I can give one very concrete example of that kind of compromise. Recently,
together with Stephen Gaies, I developed a set of materials for teaching
grammar through awareness raising rather than through production practice,
and one of the types of activities that we have in these materials we call
noticing activities -- interpretation tasks where students are given a text
which is gapped. They have to listen to a recording of the text and fill
in the gaps through processing the aural input.
Now, one of the things that we discussed when we were developing these
materials was whether we should give some guidance as to what students needed
to do when they filled the gaps -- for example, by providing the verb in
the unmarked form -- or whether one should leave the gap without any clue
as to how it should be filled in. My preference was for not giving any clues,
but forcing learners to listen very closely to the text in order to work
out what the exact word was. But the editor of these materials felt that
the learners for whom these materials were primarily intended, Asian learners
of English, needed some assistance in processing the forms that they were
supposed to listen for. Therefore, he insisted strongly that it would be
better to give a clue in the form of the unmarked verb, and eventually Stephen
and I agreed to do this.
Retrospectively, though, many teachers who are using these materials
have asked, "Why do you give them the verb? If you give them the verb
they don't really need to listen as they can just use the context to put
the verb into the correct form. And we're back into the usual type of blank
filling exercise, whereas if you didn't give them the verb then it's necessary
for them to process the text, to listen carefully to the text."
So in this case, my own inclination as a theorist and teachers' own understanding
of how the materials might work best actually concurred. But the editor
was concerned more with the marketability of the materials. Publishers sometimes
insist on compromises that really aren't necessary, and if they were prepared
to be a little bit more adventuresome, in fact there would be no negative
impact on sales.
KH: Are publishers slow to take risks and make innovations? Does this
hinder putting the results of SLA research to use?
RE: Are publishers slow to accept innovations in materials? The answer
is definitely yes. Nevertheless, publishers are aware of what constitute
the best sellers in the particular area. If we take grammar practice books,
they are aware that the grammar practice books that have made the most money
are grammar practice books like, say, Murphy, and therefore there is a tendency
for them to produce some kind of clone of Murphy, so that they can get their
share of the market in this particular area. This tends to preclude producing
grammar practice books that are radically different.
Recently, I undertook an analysis of the methodological options that
were used in six grammar practice books, one of which was Murphy, another
of which was Eastwood -- both best-selling grammar practice books. One of
the things that struck me about the results of my analysis was that the
principal methodological options being used in these state-of-the-art 1990s
grammar practice books were exactly the same as the principle methodological
options being used in grammar practice books produced in the 1950s or the
1960s, and what constitutes a grammar practice book -- certainly from the
methodological perspective -- hasn't really moved on very much. The two
main methodological options were to give the learners an explicit description
of a grammar point and then give them some practice exercises, typically
of a very controlled nature, such as filling in the blanks.
Why haven't grammar practice books moved on? There are many possible
reasons, but one reason is that grammar practice books that utilize those
two methodological options have proved very successful commercially. Therefore,
publishers are perhaps reluctant to change them, so the changes were more
cosmetic changes than methodological changes. Asking publishers to publish
a grammar practice book that is radically different to the Murphy or the
Eastwood format is a challenge to them, and they are somewhat wary of doing
so.
Then the question that you wanted me to address was how easy had I found
it to incorporate the theoretical ideas that come from my work in SLA research
into my materials writing and materials publishing . . .
KH: How receptive publishers are to it . . .
RE: Perhaps understandably, publishers are very often reluctant to make
radical changes to formulas that have led to publishing successes in the
past. I've been lucky, because with Stephen Gaies I have managed to publish
Impact Grammar, which I think is very different from traditional
grammar practice books. It's interesting to consider why this was possible.
There are two major reasons why Impact Grammar got to be published.
One was that it rode on the coattails of the other books in the Impact
series. So the publisher, Addison Wesley Longman, felt more confident in
publishing it, because they felt that teachers would buy it simply because
it had the Impact name.
The other reason why it got to be published was that it was developed
under rather unique circumstances. It wasn't developed directly by Longman,
but by a private company run by Michael Rost that was contracted by Longman
to develop the Impact series. Mike Rost, as well as being the owner
of a publishing company, is also an applied linguist. He is a researcher
and has written books on applied linguistics. Mike Rost, perhaps more than
anyone, has one foot in each camp. He has a foot in the researcher/theoretician
camp and also a foot in the practical world of teachers. This meant that
we were developing a book with someone who had a very solid understanding
of the theoretical principles that underlay it.
This did help to produce a book that was innovative. It's probably rare
that textbook writers have the opportunity to work with editors who are
themselves applied linguists, and the situation that we enjoyed in developing
Impact Grammar in that respect is unique. One only wishes that this
sort of situation existed more generally.
KH: It's almost action publishing. Why does Impact Grammar use
listening as the form of input?
RE: One of the ways in which Impact Grammar is innovative is that
the data learners work with is aural input, rather than written. If you
look at grammar practice books, you'll find that very few of them actually
provide learners with the opportunity to listen to texts, as opposed
to that of reading texts. We chose listening texts because we felt
that one of the things that learners need to do when they are learning grammar
is to notice particular grammatical features as they occur in input.
In particular, we felt that they need to be able to notice these features
when they occur in aural input. In general, learners -- perhaps Asian
learners in particular -- have very considerable difficulty in processing
aural input. They have difficulty in processing it for meaning, and because
of this, they have very little processing space left to actually notice
the grammatical features that are present in the input. Current theories
of SLA argue that noticing is essential in order for acquisition to take
place. Therefore, one of the things that the Impact Grammar materials
try to do is train the skill of noticing in aural input.
We also want learners to be able to process grammatical features in real
time -- not in a very controlled fashion, but as they hear them. Obviously,
aural input is essential, because with written input learners have the opportunity
to read the text five times, to translate it, to engage in what I call control
processing, but when learners are listening to aural input, they have
to process a feature as they hear it. Of course, they can replay the tape,
but even then, they're still processing it again in real time, so you are
forcing automatic processing. One of the conditions for successful acquisition
is that learners engage in automatic processing. Hence our use of listening
rather than written texts.
KH: You label your tasks consciousness raising (CR). There has
been a substantial amount written about conscious experiences being too
subjective and therefore making external observation of CR impossible. Could
you explain why you stick with the term CR?
RE: The term CR has probably been the preferred term in the literature.
It goes back to the 1980s, when Sharwood-Smith used it. Also, Rutherford
used it, I've used it, and it has been fairly widely accepted now in teaching
circles. Sharwood-Smith has argued that researchers or teachers can not
necessarily raise consciousness in learners minds, because that's something
that learners can only do for themselves. All that the researchers or teachers
can do is to fiddle around with the input that the learners are exposed
to, hoping that the input will raise consciousness in some way. For this
reason, Sharwood-Smith has proposed that we use the term input enhancement
rather than CR. I stuck with CR, because the term has become generally
accepted by teachers. I don't find too many teachers using input enhancement.
I do find them using the term CR.
I've also stuck with it because I wanted a term that contrasted with
the notion of practice. When I first introduced the term CR, I wanted it
to refer to a different approach to grammar teaching. Practice materials
are directed at getting learners to produce the target language structure;
consciousness-raising materials are directed at developing awareness of
how the grammatical structure works.
One of the general problems of applied linguistics is that it's full
of terms and at various points people come along and say, "Well, perhaps
this isn't the best term for x, and maybe this will be a better term for
x." People have to decide whether they want to go with the new term
or stick with the older term.
KH: Earlier, you touched on the relationship between meaning- and form-focused
tasks. Do you feel that they have different goals, or can materials focus
on both simultaneously?
RE: This is perhaps one of the essential questions about language teaching
facing us at the start of this century. That is, the relationship between
meaning-focused and form-focused instruction. There are problems with trying
to integrate the two. When you try it, you run the risk of compromising
the communicative part of your program. Students perceive the entire program
as requiring accuracy and a display of knowledge, rather than efforts to
communicate meaning.
Recently, Michael Long has been arguing that the best way to integrate
a focus on meaning and a focus on form is methodologically, rather than
through design. From a materials point of view, this suggests that all we
need are meaning-focused materials and then methodological guidance to the
teacher as to how a focus-on-form (FonF) can be incorporated in the context
of doing the meaning-focused activities. What Long has in mind is that the
FonF should occur through the feedback that the teacher gives to learners
as they attempt to do communicative activities. For example, if learners
make errors, the teacher can step in and model the correct form by means
of recasts. If one does it this way, then in essence one would rely entirely
on a communicative task-based syllabus.
KH: Perhaps some methodological outline could be provided in the teacher's
book.
RE: Yes the teacher's book could include possible grammatical structures
that the teacher might look out for, to see if students perform them correctly,
but by and large the teacher would have to act responsively, in the sense
of responding to grammatical problems that arose when students were trying
to perform a particular communicative activity.
KH: Is that only half of an approach then? Isn't there a proactive situation
as well as a reactive situation?
RE: Right. The two ways of dealing with form focused instruction are
both necessary. We could have separate components for meaning- and form-focused
instruction. The form-focused component of our curriculum would follow a
structural syllabus, based on the kinds of problems that we know learners
are likely to make. There would be grammar lessons in such a program. Then,
in the meaning-focused part of the program, there would still be the opportunity
for FonF to be introduced methodologically, through teachers' responses
to grammatical problems.