2003-dornyei-skehan-hsla.pdf

(1354 KB) Pobierz
18 Individual Differences
in Second Language
Learning
ZOLTAN DORNYEI AND
PETER SKEHAN
1
Introduction
An appropriate starting point for a discussion of individual differences in
second language learning is what might be termed "the correlational chal-
lenge." This is that individual differences in second language learning, prin-
cipally foreign language aptitude and motivation, have generated the most
consistent predictors of second language learning success. Correlations of apti-
tude or motivation with language achievement range (mostly) between 0.20
and 0.60, with a median value a little above 0.40. Since aptitude and motiva-
tion do not show particularly high correlations with one another, they com-
bine to yield multiple correlations which are frequently above 0.50.
Aside from age of onset, no other potential predictors of second language
learning success consistently achieve such levels. Yet it is fair to say that learner
differences, such as aptitude, style, and strategies, as a sub-area of second
language acquisition, and applied linguistics more generally, have not been
integrated into other areas of investigation, and have not excited much theor-
etical or practical interest in recent years. Other sub-areas, principally those
associated with universal processes, have generated a much higher level of
empirical research, for example, route of second language development, or
features of input or interaction hypothesized to promote second language
development. Other SLA areas have also been more central to theoretical
developments, for example, processability theory. And finally, more universal
areas have enjoyed clearer perceptions of practical, classroom-oriented relev-
ance, for instance, task-based instruction.
This is a curious state of affairs. It is difficult, after all, to ignore such im-
pressive correlations. We are left to conclude that the study of most areas of
individual differences in language learning is simply not fashionable, and has
590
Zoltdn Dornyei and Peter Skehan
been avoided because other areas have appeared to have greater promise. It
will be the aim of this chapter to try to redress this imbalance. We will survey
work in several areas of individual differences, and will conclude that there
are now signs that the study of areas such as aptitude and motivation is ready
for reintegration into mainstream SLA, as well as being closer to generating
active intervention in the language classroom. For reasons of space and duplica-
tion, some individual differences will not be covered in this chapter. The age
factor is addressed by Hyltenstam and Abrahamsson (this volume). The broad
sub-domain of personality is not explored here, partly because it is so exten-
sive, and also because progress in this area has been slow, in terms of both
methodology and systematic patterns of results. A recent review of the area
can be found in Dewaele and Furnham (1999).
The chapter is organized into four main parts. We review research and theor-
izing in the areas of foreign language aptitude (sections 2-5), learning style
(section 6), learner strategies (section 7), and motivation (section 8), in turn. It
will be seen that research into aptitude has languished somewhat, but is now
gathering pace again. A significant influence upon this is that aptitude re-
searchers are now exploring relevant constructs against the sort of progress in
SLA reported in other chapters in this volume. Learning style and cognitive style
are still elusive concepts, and, not for the first time, are assessed as containing
more promise than attested relevance. Learner strategies research, which was
extremely active a few years ago, seems to be losing vitality somewhat. Finally,
the study of language learning motivation, which has generated by far the
most research in this area, will be seen to have changed character in significant
ways in the last 10 years. A concluding section offers generalizations which
indicate links between the various areas of individual difference research.
2 Foreign Language Aptitude
2.1 Preliminaries
The central claim in foreign language aptitude research is very simple. It is
that there is a specific talent for learning foreign languages which exhibits
considerable variation between learners. Expressed in these terms, it would
be unremarkable in nature, and comparable to the myriad other domains
where there is variation between human beings. The complexities with aptitude
derive from a number of related questions:
i Is such a talent innate?
ii Is it relatively fixed?
iii If it is not fixed, is it amenable to training?
iv Is foreign language aptitude a distinct ability, or does it relate to more
general abilities, such as intelligence, effectively functioning as a subset
of a more general view of human variation?
Individual Differences in L2 Learning
591
v Could such a talent be used as the basis for prediction of language learn-
ing success? If so, how effective might it be for such prediction, and how
would predictions based on it compare with predictions made from other
sources?
vi Could such a talent be used as the basis for adaptation of instruction?
vii Does such a talent always apply in a similar manner, without influence of:
a learning context (e.g., FL vs. SL);
b learning methodology;
c LI to L2 combination?
viii Is such a talent undifferentiated, or does it have sub-components?
ix What
is
the theoretical basis for any such talent or sub-talents?
Finally, and in a sense, more importantly, and most mundanely:
x Can such a talent be measured effectively?
This section will address these questions, providing answers to some, and at
least surveying what is known about the others. It will be seen that the con-
cept of aptitude, long regarded as out of date, has much to offer, but needs
new conceptualizations to link it to insights and findings from SLA research. It
also merits an active research program.
2.2
Carroll's initial work
It is appropriate to start the discussion of aptitude with a review of the work
of the American psychologist J. B. Carroll. Rarely has a sub-area been so domin-
ated by one person. Carroll researched foreign language aptitude (as well as
an enormous range of other phenomena) and established the parameters within
which the sub-field still operates. It is instructive to explore his ideas about
aptitude, as well as his methods of inquiry.
Carroll conducted the relevant research during the 1950s. Together with co-
researcher Stanley Sapon, he devised a practical (and commercially available)
aptitude test battery (Carroll and Sapon, 1959). In the fullest account of the way
this was done, Carroll (1965) reported how he and Sapon started by devising
a large number of potential predictor tests of foreign language learning. They
then administered these potential tests (over 40 of them) to learners, and
gathered data on the achievement scores of the learners at the end of the
course of instruction. Armed with those data, Carroll and Sapon then examined:
i which potential aptitude sub-tests correlated with one another highly;
ii which sub-tests actually correlated highly with end-of-course performance
on achievement tests.
As a result of this work, those sub-tests which did not correlate with end-of-
course performance were eliminated, along with those which correlated with
592
Zoltdn Dornyei and Peter Skehan
Carroll's four-component model of aptitude
Table
18.1
Component name
Phonemic coding ability
Nature and function
Capacity to code unfamiliar sound so that it can
be retained over more than a few seconds and
subsequently retrieved or recognized
Capacity to identify the grammatical functions
that words fulfill in sentences
Capacity to extract syntactic and morphological
patterns from a given corpus of language
material and to extrapolate from such patterns
to create new sentences
Capacity to form associative bonds in memory
between LI and L2 vocabulary items
Grammatical sensitivity
Inductive language
learning ability
Associative memory
one another, whatever their correlations with achievement, since they were
clearly duplicating one another. In other words, only the best of such "clus-
ters" were retained. In this way, a small group of sub-tests was selected, each
of which made sufficiently separate contributions to the prediction of end-of-
course performance.
This entirely pragmatic aim of predicting language learning success went
hand in hand with a parallel, and much more interesting and enduring, aim of
understanding the components of foreign language aptitude. On the basis of
an analysis of skills required in the groups of tests which survived the "statis-
tical winnowing" in the research project, Carroll (1965) proposed the com-
ponents shown in table 18.1.
Clearly, these four components have an intriguing relationship to one an-
other, a relationship which has been clarified in a series of papers by Carroll
(1973, 1979, 1981, 1991). Phonemic coding ability represents an interesting
perspective on the auditory component of foreign language learning. Earlier
approaches had focused on simple sound-discrimination tasks, based, essen-
tially on minimal pairs. Carroll, in contrast, realized that perceiving sound
discriminations was not enough. It was more relevant to focus on
stretches
of
sound, and then on the coding (analysis) procedures which operated upon this
sound, that is, the processes which made encoding and retrieval of material
more likely. Thus, there was a memory link-up, even to the auditory compon-
ent of aptitude. The second and third components are both concerned with
the processing of language material. Grammatical sensitivity focuses on the
capacity to analyze language material, and consequently has a rather passive
quality. Inductive language learning is more active, in that it requires learners
to go beyond the information given and to generalize, so that new language
can be produced. Finally, associative memory concerns the linkages that are
Individual Differences in L2 Learning
593
formed in memory. Clearly, this component was strongly influenced by
associationist accounts of memory prevailing in psychology when Carroll's
research program was operating. The emphasis is simply on memory as bonds.
As we shall see, this is a limited conception.
Surprisingly, in reading Carroll's work, one has the impression that the
separation into components was a tactical affair, accomplished because the
identification of related but distinct components had the most effective out-
come in terms of prediction equations derived from regression analyses. The
actual test battery which resulted from the research (Carroll and Sapon, 1959)
consisted of five sub-tests,
but those sub-tests were mainly hybrid mixtures of the
different underlying components.
In other words, understanding and construct
validity were sacrificed in favor of predictive validity. This decision has
occurred at other times with aptitude research (Petersen and Al-Haik, 1976),
and while it has enabled more predictive tests to result, it has had a disastrous
impact upon the place of aptitude within applied linguistics over the years. It
has led to the lack of appreciation of the explanatory contribution that foreign
language aptitude can make to the field of SLA.
2.3
Post-Carroll research
Since Carroll's influential work, the story of aptitude has not changed very
much. In fact, it is only in very recent years that interesting and challenging
reconceptualizations of aptitude have emerged. We will return to these below.
First, however, it is worth briefly surveying how the fields of applied linguistics
and of language teaching have positioned themselves with respect to aptitude,
and what research has actually been completed within the framework estab-
lished by Carroll.
Two major influences have caused the study of aptitude to become a
marginal activity over the last 30-40 years. First, aptitude has been poorly
regarded within language teaching. One reason for this has been that aptitude
is perceived as anti-egalitarian, in that if a fixed, immutable interpretation of
aptitude is taken, it is seen as potentially disadvantaging many learners, with
no hope offered of overcoming the handicap of low aptitude. It may not be a
logical reaction, but many researchers have turned away from the study of
aptitude as a result of drawing essentially this conclusion. Another negative
response within the language teaching profession derives from the place of
learner differences more generally. Even though virtually all teachers would
quickly agree that learners differ from one another (with the acceptable face of
these differences often being referred to as "mixed ability teaching"), the bulk
of language teaching materials have assumed that all learners are the same.
Certainly a major feature of the language teaching profession over the last 20
years or so has been the rise and rise of the main coursebook series. These
series, now produced with immense care and resourcing, necessarily assume
that all learners are essentially the same (thereby maximizing sales potential),
and so downplay how the individual learner may be catered for. As a result,

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin