System 31 (2003) 393–415

www.elsevier.com/locate/system

Cognitive styles in the service of language

learning1
Madeline Ehrmana,*, Betty Lou Leaverb
a
Foreign Service Institute, US Department of State, Washington DC, USA
b
Coalition of Distinguished Language Centers, 784 Northridge PMB 293, Salinas, CA 93906, USA
Received 25 May 2003; accepted 3 June 2003

Abstract
This article presents a new approach to understanding and using cognitive styles to enhance
individual language learning. Ehrman and Leaver (2002). Development of a profile approach
to learning style diagnosis. Unpublished manuscript; Ehrman, M.E. (2001). Bringing learning
strategies to the learner: the FSI language learning consultation service. In: Alatis, J.E., Tan,
A. (Eds.), Language in Our Time: Bilingual Education and Official English, Ebonics and
Standard English, Immigration and the Unz Initiative. Washington DC, Georgetown Uni-
versity, pp. 41–58] have established a learner profile schema usable for diagnosis and advising
language learners with ten cognitive style dimensions, most of them well-known (e.g., field
independence, leveling-sharpening, random-sequential), The model also includes a super-
ordinate construct, called synopsis–ectasis to avoid confusion with earlier names and con-
structs like ‘global-analytic.’ The article introduces the Ehrman–Leaver Construct and
illustrates it with two student cases.
# 2003 Published by Elsevier Ltd.
Keywords: Learning styles; Cognitive styles; Sharpening; Leveling; Sequential; Reflective; Impulsive; Field
independence; Learner autonomy; Foreign Service Institute

1. Overview of the academic context

1.1. Cognitive styles

Work on cognitive styles has been continuous since the 1940s and 1950s, when
Witkin and his colleagues (Witkin and Goodenough, 1981) began work on perception

* Corresponding author: 1501 S. Randolph Street, Arlington, VA 22204, USA.

E-mail addresses: [email protected] (M. Ehrman), [email protected] (B.L. Leaver).
1
The content of this article does not represent official policy of the U.S. Department of State; the
observations and opinions are those of the author.

0346-251X/$ - see front matter # 2003 Published by Elsevier Ltd.

doi:10.1016/S0346-251X(03)00050-2
394 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

of the vertical, later to become field independence, and the ego psychologists
attempted to understand ‘‘cognitive controls’’ (Jonassen and Grabowski 1993) from
which came many of the still-used cognitive style models like leveling-sharpening.
Most of the subscales used in the Ehrman–Leaver cognitive styles model described
in this paper are adapted from the work of those early researchers. Other work on
individual differences has of course looked at sensory channels (e.g., Reid, 1987),
language aptitude (e.g., Carroll and Sapon, 1959; Wesche, 1981), personality—the
Myers-Briggs Type Indicator (Myers et al., 1998) or the Hartmann Boundary
Questionnaire (Hartmann, 1991), physical preferences while learning (e.g., Dunn
and Dunn, 1978), and other cognitive measures (e.g., Keefe and Monk, 1986). There
is a substantial body of literature now for the affective domain: anxiety (e.g., Hor-
witz and Young, 1991), motivation (e.g., Oxford and Shearin, 1994; Dörnyei, 2001),
and self-efficacy (e.g., Huang and Chang, 1996). Because the literature in individual
differences is so rich, the citations above are only samples; there is much more to
read in this field.

1.2. Application of cognitive styles and learner autonomy

The literature is full of research on individual differences (much of it is cited in

this paper and in this issue). However, there is relatively little about the application
of the findings to real life learning. Here, we describe a new model of cognitive
styles and its application to enhancing the effectiveness of intensive language
training.
Unlike the literature on learning style research, the literature on applications,
especially in the context of language advising, is quite sparse. There is considerably
more being done in Europe and the Pacific Rim (notably Hong Kong, Australia,
and New Zealand) with ‘language advising’ than in the US, but even there, this kind
of work is concentrated in relatively few centers, for example, the University of Hull
and the CIEL Project in the UK, the Language Centre of Helsinki University in
Finland, and Hong Kong Polytechnic University (see Hong Kong Association for
Self Access Language Learning, September 2002). US resources tend to focus much
more on generic learning, either in association with a university counseling center or
independently (Griggs, 1991). University-affiliated examples include the Oklahoma
Institute for Learning Styles (no date) and Pace University (1998). There are other
learning help resources available through the Internet that appear to be independent
of institutions of higher education, such as Support4Learning (no date), a site that
promotes study skills and awareness of learning styles.
Learner autonomy has received a great deal of attention in Europe and Asia, for
example through the venerable CRAPEL in France and Hong Kong Polytechnic
University (see Hong Kong Association for Self Access Language Learning, Sep-
tember 2002, and the AILA Learner Autonomy Project Website), but again, it has
been the object of less effort in the US. The self-access learning center is the best-
known vehicle; it attempts to train learners and provides materials that learners can
use at least partially on their own (Benson and Voller, 1997; Cotterall and Crabbe,
1999; Dickinson, 1995; Fitzgerald et al., 1996; Wenden, 1991, 1995).
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 395

2. The Ehrman–Leaver cognitive styles construct (E&L)2

The E&L model consists of a superordinate construct, synopsis–ectasis, and ten

subscales. It originated from our dissatisfaction with what was commonly referred
to as the ‘global-analytic’ distinction, which was leading us to misdiagnoses and
confusion about the meanings of terms. We therefore selected ten scales that were
informative to us in our work with students, shown in Fig. 1.
Like previous researchers in this field, we saw a systematic, conceptual link among
the scales. We initially thought that global, inductive, random, leveling, etc. were
ways for learners to group the information or treat it all at once and so used the
term ‘synopsis’ to refer to that set of poles. A teacher of Greek provided the oppo-
site of synopsis, ectasis (adj. ectenic),3 meaning to stretch out, which we adopted for
the other pole.
Prior to researching initial student preference responses, we assumed that field
independence would cluster with particular, analytic, etc. on the ectenic pole. How-
ever, it correlated clearly with the synoptic poles instead and led to a revision in the
definition of ‘‘synoptic–ectenic’’ (Ehrman and Leaver, 2002). Because field indepen-
dence is probably a perceptual and therefore not fully conscious process, we now
hypothesize that ectasis controls consciously what synopsis accomplishes through
preconscious or unconscious processes. In short, synoptics ‘‘trust their guts,’’ and
ectenics tend not to. This phenomenon seems especially clear in the impulsive–
reflective subscale, where reflectives need to ‘‘think about it’’ first, whereas impul-
sives are likely to respond quickly, with little thinking.4
We followed in the path of countless researchers and theorists who have attemp-
ted to assemble various cognitive styles into clusters. For example, they would argue
that a field-dependent person was also likely to be global, socially-adept, etc.,
(Messick, 1984; Miller, 1987; Riding and Cheema, 1992; Witkin and Goodenough,
1981). Such clusterings, while noticed by many of those studying cognitive styles,
unfortunately have never been well researched. Although some of these them have
been proposed many times under a variety of names (e.g., active–passive by Davis,
1967; holistic–analytic by Kirby, 1988 and Riding and Rayner, 1998); global–ana-
lytic and levelers-sharpeners5 by Schmeck, 1981, 1988; right-left hemisphere, by, e.g.,
Hermann, 1982, Sinatra and Stahl-Gemake, 1983, Torrance et al., 1977), we decided
to use new terminology because our construct is different. For example, an opposi-
tion like holistic–specific focuses on the level of detail; one like gestalt–analytic
indicates the level of analysis. In contrast, the synoptic–ectenic distinction addresses
the degree of conscious control of learning desired or needed, to the best of our
knowledge a distinction not found elsewhere in the cognitive style literature.

2
The E&L model, instrument, and explanatory material are copyrighted by Ehrman and Leaver
(2002).
3
‘‘Ectasis’’ is stressed on the first syllable; the adjective ‘‘ectenic’’ has stress on the second syllable
(source of ‘‘ectasis’’ was Panagiotis Sapountzis, personal communication, 1977).
4
Other factors, such as pattern matching, intelligence, previous knowledge base, and the like may
determine whether the impulsive’s quick response is accurate or not.
5
Schmeck (1981) referred to these categories as lumpers and splitters.
396 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

We developed an instrument, the Ehrman and Leaver Learning Styles Ques-

tionnaire (2002; Ehrman, 2001), to make the theory operational. Consisting of three
items for each of the ten subscales, the instrument and model have been undergoing
validation at the Foreign Service Institute for several years, where they have been in
active use as part of the battery of questionnaires offered to participants in the
Learning Consultation Service there.6 The E&L construct includes a profile design
providing detailed and individualized information about students but with a con-
struct and a format that make profiles comparable across students and, potentially,
over time. Besides the NASSP instrument (Keefe and Monk, 1986), it may be the
only other multidimensional profile available for learning styles.7

2.1. E&L subscales

Each of the subscales included within synopsis–ectasis has its own body of litera-
ture, often extensive, in both in general psychology and in second language learning.
We provide an overview and sampling of the available literature here.

2.1.1. Field independent and field sensitive

Field independence (FI) addresses the degree to which an individual focuses on
some aspect of experience and separates it from its background. (The word ‘‘field’’
or ‘‘ground’’ is used for this kind of background; the term ‘‘figure’’ is sometimes
used to indicate what receives focus and is thus pulled into the foreground.) Some
extend the concept to refer to the ability to conduct abstract cognitive operations on
the material that receives focus (Witkin et al., 1977). Morgan (1997) describes find-
ings that when the field is not clearly organized, individuals who tend to field inde-
pendence are relatively likely to impose their own structure on the material, whereas
field dependents (FD) often accept it as it is. Ehrman (1996b, 1997) indicates that a
field independent learner is adept at focusing a spotlight on data, distinguishing and
focusing deeply on some specific aspect of the material being learned. Such a learner
can look at the forest and pick out exactly the kind of tree in which she or he is
interested. A field independent learner is likely to be relatively skilled at chunking
information and working further with it.
The first studies in field independence-field dependence were conducted by Witkin
(Witkin, 1969; Witkin and Goodenough, 1981). It has been among the most com-
monly used language learning style dimensions (e.g., Chapelle, 1992; Ehrman, 1997;
Jamieson, 1992). Early studies that applied this concept to foreign language learn-
ing, e.g., Stansfield and Hansen (1983) found that field independent learners were
better at classroom learning, as tested by discrete item instruments. However, the
construct has been little tested with communicative outcomes.

6
The current version of the E&L questionnaire has been administered to well over 500 students at FSI,
most of whom have received face-to-face interpretation. Some preliminary statistics are available, and
further statistical analysis is pending.
7
Others appear to use a collection of styles that may or may not be held together with a complex
instrument.
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 397

The term ‘‘field dependence’’ is used in two ways in the literature: absence of the
kind of discrimination referred to as field independent and awareness of the entire
field. Since field dependence is always measured by tests of field independence; it can
safely be defined only as absence of field independence. However, because learners
need to be able to be aware of background activity as well as bring information into
focus and reorganize it, there is a positive aspect to what is traditionally called ‘‘field
dependence’’, which can enhance functioning in complex social situations. Complex
social situations are in turn often involved in real language use, so this kind of ‘field
dependence’’ is likely to play a constructive role.
Although many, beginning with Witkin and his collaborators (Witkin and Good-
enough, 1981) have used the term ‘‘field dependence’’ for such positive responsive-
ness to the surrounding background, following Ehrman (1996b, 1997), we treat this
kind of processing as a separate style, called ‘‘field sensitivity’’.8 In contrast to a field
independent learner, a field sensitive learner makes skilled use of a floodlight to
maintain awareness of the entire forest, registering the presence of all the flora,
fauna, and moment-to-moment changes in the environment. Both perspectives are
useful, and some learners can apply both, though probably not simultaneously.
Absence of field sensitivity is ‘‘field insensitivity’’, for which, like field dependence,
there exists no direct measure, though language teachers report encountering many
such learners.
The E&L Questionnaire adopts Ehrman’s (1996b, 1997) model of field indepen-
dence versus field dependence and field sensitivity versus field insensitivity. A stu-
dent can be both field independent and field sensitive, one or the other, or neither.9
Field sensitivity is closely allied to the constructs measured by MLAT part II (Car-
roll and Sapon, 1959) and the Hartmann Boundary Questionnaire (Hartmann,
1991).
Ehrman’s (1996b, 1997) introduced a model of field independence and field sensi-
tivity (see Table 1) that produces four possible types.
By ‘unpacking’ field dependence, this model makes it possible for a person of Type
1 to have skills associated with both field independence and field sensitivity.

Table 1
Ehrman model of field independence and field sensitivity

Type 1 Field independent and field sensitive Can learn from material in and out of context
Type 2 Field independent and field insensitive Comfortable with out-of-context material
Type 3 Field dependent and field sensitive Comfortable with in-context material
Type 4 Field dependent and field insensitive Has difficulties with both kinds of material

8
The term ‘‘field sensitivity’’ was originally used by Ramı́rez and Castañeda (1974) as a substitute for
field dependence, which they considered derogatory, and in an attempt to suggest a positive opposite to
field independence.
9
The best language learners are often both field independent and field sensitive; that is, they can work
with material that is not embedded in context or can see what is most important, and they can also pick
up language in a relatively global way by being exposed to it.
398 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

2.1.2. Random (non-linear)-sequential (linear)

‘‘Random’’ (non-linear) and sequential (linear) learning relate to how a learner
structures information, by internal criteria, i.e., random, or external ones, i.e.,
sequential. Random-sequential differences, frequently addressed in the learning styles
literature, have been addressed by Gregorc (1982) as part of a bi-dimensional system,
Pask (1976) as part of the holist and serialist distinction, and Esch (1977), who applied
the serialist-holist distinction to serialist learners in foreign language programs.
Random learners usually work out their own learning sequences; they can be very
systematic learners, but their systems are often idiosyncratic, and their approach
may seem ‘‘random’’ to the outsider. Effective random learners tend to tolerate
ambiguity relatively well and embrace surprises that might disrupt the learning of
others.
Sequential learners want to learn step-by-step, following an order usually pro-
vided by a curriculum or textbook. They may become frustrated when faced with
very open-ended classroom activities. They often prefer—or even need—to master
one thing before going on to the next and want the variables in learning exercises to
be controlled. A positive aspect of sequential learning is that such learners seldom
miss important points, because they make sure that all the material is covered; they
are often systematic and, at their best, good planners.
Ehrman points out that sequential processing provides a kind of control by redu-
cing the likelihood of surprises that may tax the learner’s ability. On the other hand,
random learning is also a form of control, as it enables one to count on one’s own
resources even in an unfamiliar or ambiguous situation (Ehrman, 1996b).

2.1.3. Global-particular
Global processing focuses on the ‘big picture’ and processes ‘‘top down,’’ focusing
on overall meaning first. Leaver (1998) suggests that global students try to put meaning
to everything, thus sometimes missing details. If they miss enough details, the
meaning that they ‘‘invent’’ can stray quite far from reality. In contrast, students
who prefer particular processing attend readily to discrete items and details and
process ‘‘bottom up,’’ with form before general meaning. They may find details
important without regard for larger concepts.
Global learning styles are frequently contrasted with analytic styles (Messick,
1994; Schmeck, 1988). The Reading Style Inventory (Carbo, 1997), and the Style
Analysis Survey (Oxford, 1993) use this distinction. However, Ehrman and Leaver
(2002, Ehrman, 2001) have pointed out that the characteristics of global learning are
not in opposition to those of analytic learning but rather to another set of traits that
they label ‘‘particular.’’ Other than literature produced by the authors of this article,
the particular (contrasted with global) learning style has only been discussed in
learning styles literature to date as ‘‘analytic’’ learning.

2.1.4. Inductive-deductive
This scale is fairly common in the literature (e.g., Messick, 1984, Nickel, 1984).
Induction begins with the data and extracts generalizations, seeking to find a theory
(rule) by looking for patterns in data. Deduction begins with a rule and applies it to
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 399

specific cases, perhaps testing a theory (e.g., a rule or generalization) against the
facts.10
Inductive learners form hypotheses and test them. They often want multiple
examples and may even find teacher explanations intrusive. As a result, they often
have better control of their classroom output the day following its introduction.
Deductive learners prefer to study the rules, and then practice applying them to
examples. They may not know what to do with a multitude of examples and thus
prefer teacher explanations.

2.1.5. Synthetic-analytic
As mentioned earlier, analytic learning (Kagan et al., 1963) is often treated as the
opposite of global style. We believe that a synthetic-analytic scale better reflects a
true cognitive processing dichotomy than does the better-known global-analytic
distinction.
Synthetic students like to use pieces to build new wholes. They easily put together
disparate ideas and may even develop new models with them. Synthesis as a learning
style has several characteristics: (1) hypothesis formation is intuited; (2) process and
product are experienced as simultaneous, and (3) the synthetic learner proceeds from
insight to construction. A synthetic learner might combine input from reading with
other thoughts and information to form new ideas but be unable to retrieve details
because they have been re-formed into new information.
Analytic students like to disassemble wholes into parts; they are at ease working
with all the pieces and tend to see the compositional structure of the whole. They
work well with rules that they can break down into component parts and use to explain
phenomena, and they like word study that permits separating words into etymological
pieces: roots, stems, affixes. Analysis as a learning style has several characteristics: (1)
hypotheses are built up consciously; (2) process and product are experienced as con-
secutive, and (3) the analytic learner proceeds from construct to insight.

2.1.6. Analogue-digital
This is the one subscale that is not common in the literature. It is adapted from
Smith and Berg (1987), who coined the terms digital and analogical from telephone
and computer processing systems. In their construct, digital thinking involves logi-
cal, sequential processing and uses an on/off mechanism. Analogical thinking is non-
linear, ‘‘artistic,’’ and uses a qualitative mechanism of more or less. As far as we
know, the Smith and Berg construct that Ehrman and Leaver (2002, Ehrman 2001)
adapted for their model has not previously been applied to learning styles theory.11

10
Some sources reverse the meanings of these terms.
11
Littlemore (2001, 2003) has investigated the effects of metaphorical competence on language learning,
treating it as a representative of the holistic style which is otherwise disadvantageous in the academic
classrooms where her investigations took place (though not, she suggests, in real communication). A
related concept is tight and loose analogical reasoning, based on Holyoak’s (1984) distinction between
literal (compared material is closely similar) and deep analogies (information types are quite
disparate); loose analogical reasoning would be closer to analogue processing or Littlemore’s metaphoric
competence.
400 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

The traditional clock face uses a physical metaphor (numbers in a circle) to

represent a quantitative concept, time of day; we (Ehrman and Leaver, 2002, Ehr-
man 2001) use the term ‘‘analogue’’ for such metaphoric representation applied to
learning. An example is a student who had difficulty with traditional grammar
explanations and conceptualized the complex grammar system of an Asian language
as an engine that he could assemble and disassemble (Ehrman, 1996b). This was an
extended metaphor. Analogue learners frequently have a strong preference for
learning material in context.
In contrast, a digital clock is something of a literal representation of the quanti-
tative idea of time of day. Working with things as they are on the surface can thus
also be called ‘‘digital’’. Digital processing tends to take a fairly literal approach to
information, without attempting to seek or impose additional meaning. For exam-
ple, a student could watch taped discussions on related topics in order to identify the
discourse style of the interaction (formal, informal, etc.). The more ‘digital’ students
would have more difficulty with this task, because they would tend to focus on the
words and not the accompanying implications and connotations.
Analogue learners like stories, parables, analogies, and metaphors, and tend to use
‘‘deep’’ strategies (e.g., association and elaboration), whereas digital learners want to
hear it ‘‘like it is,’’ without what may seem like extraneous or fanciful embellish-
ment, and may rely mostly on surface strategies like memorizing or word lists.

2.1.7. Concrete-abstract
Concrete learners normally prefer some kind of sensory contact with what is being
learned, a relationship with direct experience, and experiential learning. They often
want activities using the language to do something, not just talking about the lan-
guage. Ehrman (1996b) describes a student who likes to play learning games in class,
read aloud, take field trips, do role-plays, do pronunciation and grammar drills, talk
as much as possible, and write things down. All of these are concrete activities. He
rejects classroom discussion of abstractions like cultural values, and finds explana-
tions and grammar rules to be of little help to him. Because hypothetical questions
are difficult for him, he brings his response back to his direct experience.
Abstract learners show a preference for grammar rules, systems, discussion of
abstract topics, and attention to accuracy. Such a student can become lost in theory
and never achieve language use. Another student in Ehrman (1996b) has a strong
interest in political science and wants to talk about it, only to be tripped up by her
lack of language to match her preference for the abstract. Sometimes she does not
understand what her teachers are saying to her because she sometimes starts think-
ing about a word or grammar rule, rather than just using it. Abstract learners
sometimes indicate more interest in the system underlying the language than in
actual language use for communication.

2.1.8. Leveling-sharpening
The leveling and sharpening scales reflect how students perceive, store, and recall
information (Gardner et al., 1959; Lowery, 1982; Messick, 1984). This was one of
the first cognitive style dimensions, originally in perceptual processing (Holzman
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 401

and Gardner, 1960; Holzman and Klein, 1954). According to Holzman and Klein
(1954), levelers rely on past memories and merge current experiences with them,
whereas sharpeners rely relatively little on memory. Levelers often blur similar
memories and tend to over-generalize. Sharpeners distinguish small differences and
separate their memory of prior experiences relatively easily from current ones; if
they get in trouble, it is for over-discriminating. As learning styles, leveling-
sharpening differences represent what learners pay attention to and how they store
it in memory.
When learning new information, levelers meld together information that may be
distinctly different and that may come from a number of sources. They overlook
distinctions intuitively and frequently see only similarities. They can have trouble
with contrastive analysis and analogies, because these are based on differences.
Leveling may assist in language learning because it promotes data clumping and
thus reduction of cognitive load, and it may encourage fluency if the leveler is not
anxious about making errors. However, levelers can also ignore the important dis-
tinctions that are critical for linguistic and sociolinguistic accuracy.
Sharpeners may readily retrieve details and fine grammatical or lexical distinctions
because they store them separately. Sharpening helps language learning because it
contributes to making important distinctions among speech sounds, grammatical
forms, and fine differences of meaning. It may be less helpful if it leads to hair-
splitting.
Leveling-sharpening differs from global-particular. The former has to do with
attention, perception, and storage in memory, and the latter has to do with direction
of processing (top-down and bottom-up). Leveling lumps information and reduces
distinctions, whereas global learning seeks the underlying meaning and overall pat-
tern of an experience. Sharpening seeks significant distinctions, whereas particular
learning focuses on specifics without necessarily making distinctions of category.
FSI experience suggests that leveling and sharpening play important roles in lan-
guage learning: sharpeners, especially those who are synoptic sharpeners, tend to
reach the required professional levels of language proficiency with greater facility
than other students. In interviews with very highly proficient language learners,
Leaver and Atwell (2002) found that most reported having sharpening preferences
or having developed sharpening skills and considered these, among other factors, to
be critical to their learning success. One individual, for example, when asked what
advice he had for others who would like to reach native-like proficiency levels said
they should learn how to observe well, and if they did not know how to do this
intuitively, they should be taught to do it (Leaver, 2003).12

12
Leaver (1986) conducted a study of successful language learners in Russian at the FSI. She followed
102 students through a full-year of learning, including some that were in an advanced course leading to
Level 4 (advanced professional proficiency). She identified them as ‘‘left-hemisphere-dominant ’’ (a
metaphor that has some strong parallels with ectenic learners) or ‘‘right-hemisphere-dominant’’ (which
has some strong parallels with synoptic learners). ‘‘Right-hemisphere’’ learners learned easily at beginning
levels and quickly reached Level 3 on the compensatorily-scored FSI test [which allows some grammatical
402 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

2.1.9. Impulsive-reflective
This scale refers to speed of processing a response to a stimulus. Impulsive lear-
ners respond very rapidly and tend to complete their work quickly but often with
less accuracy than do reflective learners. This kind of learner is frequently an active
participant in class, contributing remarks on most topics and jumping right in to
new projects or tasks. Such students may think aloud and learn by trial and error.
One who makes relatively few careless errors is referred to as a ‘‘fast-accurate’’
impulsive learner (Ehrman, 1996b; Salkind and Wright, 1977).
In contrast, reflective learners prefer to think and then respond. They often benefit
from relatively complex thinking and tend to work accurately. However, their
slowness can result in incomplete work, especially on timed tests or other ‘real-time’
activities like speaking in a foreign language. In some academic programs, a reflec-
tive student might be labeled ‘‘slow’’. A very reflective learner may have difficulty
finishing a test on time; however, the completed portions are probably correct.
When called on, such a learner often says, ‘‘Let me think.’’
The impulsive-reflective difference was an early discovery in learning styles
(Kagan, 1966; Keogh and Donlan, 1972; Zelniker and Jeffrey, 1976) Impulsivity was
associated with immaturity, especially in the 1950s, and more recently with such
disorders as attention deficit (Morgan, 1997). However, because there are efficient
and inefficient versions of both impulsivity and reflectivity (Salkind and Wright,
1977), neither is necessarily immature or dysfunctional. We (Ehrman and Leaver,
2002; Ehrman, 2001) distinguish between efficient (fast-accurate) and inefficient
(error-prone) impulsivity, and between efficient (the usual meaning of reflectivity as
a learning style) and inefficient reflectivity (compulsive, lacking learning skills). A
summary of the subscales is provided in Table 2.

3. The institutional context

The E&L model and questionnaire are used at the Foreign Service Institute (FSI)
regularly as part of the Learning Consultation Service in FSI’s School of Language
Studies (SLS). FSI provides full-time intensive training in over 60 languages to adult

error even at Level 3 (professional proficiency) if other components of speech are very strong)]. This,
which turned out to be a plateau from which very few rose higher while in training, even though they had
reached Level 3 weeks earlier than their ‘‘left-hemisphere-dominant’’ counterparts. ‘‘Left-hemisphere-
dominant’’ learners had tended to have immense difficulty in the beginning. Half of them simply never
reached Level 3; the other half reached Level 3 significantly later than their right-hemisphere-dominant
counterparts, but in the few remaining days and weeks went on to achieve Levels 3+ and 4. The actual
situation may be more complex: sharpening may be the critical element for reaching Level 4, as opposed
to the whole scale. Consequently, synoptic learner with these sharpening and particular skills preferences
may be equally capable of reaching Level 4. The E&L Construct may explain the difference between the
‘‘left-hemisphere’’ learners who dropped out of the race and those who continued, and plans call to test
the study population with the E&L (Leaver, 2003).
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 403

members of the US foreign affairs community13 usually with a goal of proficiency at

which employees can perform their duties (Interagency Language Roundtable [ILR]
Level 3). There is increasing emphasis on achieving levels ‘beyond three’, including
near-native ILR Level 4.
Although their educational level is generally high, and a substantial number of
students entering language training have previous language learning experience,
many FSI students are not sophisticated about language learning. To meet the needs
of such students, Ehrman established a language advisory service, called the Learn-
ing Consultation Service, basing it on her research during the 1990s (Ehrman, 1993,
1994a,b, 1996a, 1997–1999; Ehrman and Leaver, 2002). Initial experiments with
group learning strategies workshops had not worked well, an experience that corre-
sponds to others documented elsewhere (Chamot and Rubin 1994; Rees-Miller
1993, 1994; Wenden 1995), so Ehrman sought a way to make the teachers the source
of strategy intervention on a ‘‘just-in-time’’ basis. However, with over 300 language
teachers; it would not be feasible to train even most of them to be make sophisti-
cated strategies interventions in a short time. Instead, a small group of carefully
trained specialist teachers and a few supervisors, called Learning Consultants,
received training to help with learning strategies, using what is known about the
individual’s learning style and motivation. The success of the Learning Consultation
Service (LCS) depends on the cooperation of the regular classroom teachers; the
language-training supervisors who are the language program managers and oversee
the work of the teachers; a small corps of ‘‘counselors’’ who administer and interpret
the learning style questionnaires; and the Learning Consultants in almost every
language section. Student participation is confidential and entirely voluntary. Box 1
shows the learning consultation process.

3.1. Learning strategy advice

The program basis its strategy on Schmeck’s model (1988) modified by Biggs
(1992): the Consultation Service calls student attention to the distinction among
surface, achievement, and deep strategies. Surface strategies are those for a specific
task problem and nothing more and entail minimum cognitive or emotional invest-
ment. Achievement strategies aim at a good grade or relationships with teachers or
other students. Deep strategies make elaborations and associate between what is
new and what is known and among concepts and experiences. Deep strategies most
directly result in long-term retention, although achievement strategies can make the
use of deep strategies possible. For example, a discussion with one’s teacher
(achievement) might result in suggestions of effective ‘‘deep’’ learning techniques.

13
Students at FSI are about 40 years of age on average, and about the middle two-thirds are between
30 and 50. They come from a variety of agencies, though most are from the Department of State. Others
come from the Department of Defense, the Department of Commerce, and other agencies. Training gen-
erally consists 5–6 h of daily teacher contact time plus homework; courses are 24–88 weeks in duration,
depending on the difficulty of the language for English speakers.
 404
Table 2
Subscale definitions

Synopsis Ectasis

Synopsis relies on unconscious or preconscious and thus may Ectasis (stress first syllable) seeks conscious control of processing
result in perception of phenomena as wholes. and thus may result in perception of phenomena as composites.

Field sensitivity as learning style: prefers to address material as Field insensitivity: makes little or no use of the whole context and

M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

part of context and often picks up material by ‘‘osmosis.’’ It relates often excludes ‘‘incidental’’ learning.
to everything-foreground and background together and can be
compared to illumination by a floodlight that shows the whole scene.
Field independence as learning style: prefers to separate what is Field dependence treats foreground and background as the same and
considered important from context, setting priorities and hierarchies. does not select out what is important for focus.
It can be compared to a spotlight that focuses sharply on one thing,
in contrast to field sensitivity.
Random (non-linear): follow internally developed order of processing Sequential (linear): follows externally provided order of processing,
that may appear ‘‘random’’ to others. such as that in a textbook.
Global processing: attends to gestalts and ‘‘big picture’’, is aware of Particular processing: attends to discrete items and details, is aware
‘‘forests’’ (vs. trees), oriented toward processing from the ‘‘top down’’. of ‘‘trees’’ (vs. forests), oriented towards processing from ‘‘bottom up’’.
Inductive: goes from specific to the general, generalizes from experience, Deductive: goes from the general to specific, applies generalizations to
and begins with examples rather than rules or theories. experience, starts with rules, theories rather than specific examples.
Synthesis: comprehends through assembly of components into a constructed Analysis: comprehends through understanding componential structure;
whole; creates. disassembles.
Analogue: qualitative or metaphoric approach to interpreting experience; an Digital: quantitative or literal approach to interpreting experience; a
analogue clock represents time metaphorically through a circular image. digital clock shows only numbers directly, not representing them
as a metaphor.
Concrete: interacts with the world directly, learns through application, Abstract: interacts with the world through cognitive constructs, learns
often physical, of knowledge. Experiential. from formal rendition of knowledge. Theoretical.
Leveling: often does not notice disparities and may seek to reduce them; Sharpening notices disparities and differences and seeks to explore and
looks for similarities. account for them.
Impulsivity: reacts quickly in acting or speaking with little or no conscious Reflectivity: ‘‘thinks it through’’ before action; often does not trust ‘‘gut
‘‘thinking it through’’; acts on ‘‘gut’’; thought often follows action. reaction’’; action usually follows thought.

The E&L model, instrument, and explanatory material are copyrighted by Ehrman and Leaver (2002).
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 405

Box 1: The Learning Consultation Process:

Making the Consultation Service available. Students are invited on input day to
participate in the Consultation Service. They are assured that it is voluntary;
and it is for people who are not having trouble as well as to help those who are
having difficulties.
Completing the Diagnostic Questionnaires. If in languages participating in the
accelerated personalized training (APT) initiative, students complete the ques-
tionnaires on their first day of training. Others may ‘‘walk in’’ to the Research,
Evaluation, and Development Division at any time and are given the ques-
tionnaires at that time. Language sections sometimes refer students who are
having difficulties.
Interpretation of Questionnaires. When possible, this step takes place first in
group sessions so the counselors do not have to repeat the same information for
each student, and then in individual sessions to apply the generalizations to the
student’s situation.
Follow-up. This is usually done in the language section by the student’s desig-
nated Learning Consultant, who is expected to take responsibility for ensuring
that information is used to the student’s benefit by other teachers and by the
student (e.g., advice on preparation and classroom strategies). Students may
choose to return for follow-up consultations with a counselor on special
learning strategies or management of anxiety, for instance.

3.2. Consultation service instrumentation

Although the description of cases later in this paper will focus on the E&L model,
it is also interpreted in the light of the other instruments administered with it. A
simple biographical data questionnaire addressing the student’s educational and
language-learning history is used in an interview to reveal a great deal about learn-
ing strategies, what is working well (or not), and the student’s motivation, anxiety,
and self-efficacy. In addition to the Ehrman and Leaver Learning Style Ques-
tionnaire (E&L), other instruments currently used are the Modem Language Apti-
tude Test (MLAT, Carroll and Sapon, 1959; language aptitude, possibly field
independence), Motivation and Strategies Questionnaire (MSQ, Ehrman, 1996b;
strategies, motivation, self-efficacy, anxiety), the, the Myers-Briggs Type Indicator
(MBTI, Myers, McCaulley, et al., 1998; general personality and style), and the Hart-
mann Boundary Questionnaire (HBQ, Hartmann, 1991; tolerance of ambiguity,
receptivity).
The Ehrman and Leaver Learning Style Questionnaire (E&L, Ehrman and Leaver
2002), which is the focus of this article, uses self-reported behavior to determine
cognitive style preferences. During the individual feedback interview, this ques-
tionnaire evokes discussion of both learning styles and learning strategies. For
example, an interviewer may suggest creating paragraphs in writing to a student who
self-reports a strong preference for analyzing but who needs to develop more
406 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

Fig. 1. E&L scoring and feedback grid for Mark.

‘‘synthetic’’ skills. (The E&L model, instrument, and explanatory material are
copyrighted by Ehrman and Leaver, 2002).

4. The E&L in use

The following two short case descriptions represent use of the E&L construct with
actual learners, adults in FSI’s high-stakes training.14 Each begins with a ‘blind’
discussion of the survey information, using the E&L as a framework. This is fol-
lowed by information derived from the learner’s interview with a learning counselor
and conversation with teachers. The cases are not clear-cut examples of synoptic and
ectenic learners. Instead, they were selected as typical of the kind of ambiguous,
apparently contradictory information that a ten-scale profile can present and which
must be interpreted in a way that is useful for the student.

4.1. Mark: a complex synoptic learner

4.1.1. The questionnaire data

Mark has been studying Japanese for about 8 weeks. His E&L results are pre-
sented in Fig. 1. He is field dependent and may also be field insensitive. This is a
combination that suggests difficulties right away: he is likely to have difficulty
determining what is important, discriminating it from the background, and setting

14
Preservation of assignment to a desired post, large bonuses, and even retention in the Foreign Service
can depend on achievement of designated proficiencies at the end of training. Many of these students are
also high achievers and very competitive, so consequences of what they perceive as failure are personally
costly as well.
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 407

priorities (field dependence). His digital preference, associated with use of surface
strategies and preference for a clear field of study, is consistent with the field
dependence.
Mark’s responses on the field sensitivity dimension fail to reveal a tendency, so the
counselor needs to look at other instruments that provide triangulation on the E&L
scores. His MLAT subscale scores support the notion that he may be field depen-
dent and suggest that he may have some field sensitivity, but it is probably not
strong. The HBQ provides similar information: Mark may have somewhat thin ego
boundaries relative to other FSI students and thus be open to input and tolerate
ambiguity, but his response scores are not clear-cut.
Mark’s results on three other E&L subscales are inconclusive: global-particular,
synthetic-analytic, and leveling-sharpening. Information from other questionnaires
suggests that the indeterminacy of the first two subscales is probably real—Mark
does not have a clear preference for either pole. However, we shall see from the kind
of difficulty Mark is having with reading that he is probably leveling out distinctions
that he needs to make among written symbols.
Additional evidence for Mark’s flexibility as a learner is the fact that he has
reported himself a random learner, and less clearly, as an inductive one. The ques-
tion for the counselor to pursue is whether his field dependence may interfere with
the effectiveness of his ability to make use of ‘‘random’’ strategies, non-sequential
input, and the kind of pattern seeking that is part of induction.
Information from the MSQ confirms his inductive and non-sequential self-reports
on the E&L. On the MBTI, his preferences (INTP) support random processing and
induction, but his digital style is unusual for this type. Because he is primarily a
synoptic learner, the counselor’s initial hypothesis is that his difficulties are likely to
come from his field dependence and poorly developed learning strategies (reflected
in digital processing). As an impulsive learner, he may not be taking the time he
needs for metacognitive strategies: planning, monitoring, evaluating, and re-plan-
ning. He has indicated some preference for kinesthetic and auditory learning over
visual; kinesthetic preference and his relatively clear concrete processing preference
indicate that he would respond well to field trips and role plays. Unfortunately,
these are not frequent in FSI programs, especially in the early weeks of instruction.

4.1.2. Information from the interview and the biographic background questionnaire
It turns out that Mark is a relatively inexperienced language learner. He had
Spanish classes in high school, but they taught him very little about language
learning strategies, especially for Japanese. He is also in an occupational category
that is commonly characterized by thick ego boundaries, some ethnocentricity,
avoidance of metaphoric thinking, and frequent difficulty with language learning.
In the interview, his very first words were that he felt as if he was drinking from a
fire hose. While they are studying in FSI’s intensive language programs, language
learning is the students’ full time job; the courses cover a great deal of material very
rapidly; and they require several hours of homework/self-study as preparation for
the next day in class. Mark’s biggest difficulty was with the Chinese characters used
in Japanese; it seems that they were taking so much of his cognitive capacity that he
408 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

was struggling with all the other aspects of learning—even aspects that were nor-
mally not particularly difficult for him. He reported having difficulty focusing, which
often occurs with kinesthetic-concrete learners who have to sit still too much. He
confirmed his concrete learning preference when he indicates that he liked sponta-
neous, real-life language use and using language to communicate.
Because of the uncertainty about Mark’s field sensitivity, there was a question
about whether he should be considered a Type 3 learner (field sensitive and field
dependent) or a Type 4 (field insensitive and field dependent) in Ehrman’s field
independence typology (1996b). If his digital processing is a result of inexperience; it
is possible that with more appropriate learning strategies, he might well develop
skills for coping better with random, unpredictable, natural input (note his

Table 3
Recommendations for Mark

Need or issue Sample suggestions

Coping with field dependence Consciously prioritizing topics and activities; selection of
important items from texts; use of sequential materials as
scaffolding until he is confident enough with the material
to approach it more ‘‘randomly’’
Overuse of digital, surface strategies Reading for the whole gist rather than parts (step by step
procedure as scaffolding)
Difficulty with Chinese character Mind maps; manipulate characters physically
recognition
Kinesthetic/concrete learning Include movement and manipulation in study at home;
participate actively in any role plays or field trips included
in the language program; when more advanced, seek
opportunities to use Japanese outside the classroom; find
something physical to do with his hands when his mind wanders

Fig. 2. E&L scoring and feedback grid for Nell.

 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 409

random, concrete, inductive preferences). In fact, if he were not on cognitive over-

load with Chinese characters, he might well be coping more like a Type 3—com-
fortable with natural context but possibly in need of external structure and
guidance—than like a Type 4, for whom everything is difficult and requires a great
deal of hard, conscious work.

4.1.3. Suggestions for the student

The counselor, taking the questionnaires into account, together with the interview
and other information, has suggested specific strategies for Mark (see Table 3) in
each of the categories that were indicated as in need of work:

4.2. Nell: a synoptic-ectenic mixture

4.2.1. The questionnaire data

Fig. 2 displays Nell’s E&L results. She is field independent and field insensitive, a
Type 2 learner in Ehrman’s (1996b) typology.15 We would expect from this combi-
nation of the two subscales a learner who prefers to learn material abstractly, who is
comfortable with learning out of context, and who has little trouble with prioritiz-
ing, but may take in relatively little unconsciously. Such a learner is likely to favor
accuracy over fluency.
Nell’s preferences for particular and analytic learning are consistent with her being
a Type 2, who is interested in specifics (though they may be specifics of a fairly high
order, such as structural rules) and who learns by analyzing and understanding the
components, rather than the whole. Also consistent with being a Type 2 are her
relatively thick ego boundaries (thin ego boundaries correlate with field sensitivity,
and Nell is field insensitive). Her responses on the MSQ indicate a student who
searches for perfection, which is consistent with the Type 2 classification.
A more extensive look at Nell’s profile indicates that her learning style is in fact is
more complex than the analysis above would indicate, however. First of all, she is a
concrete learner, one who likes to apply what she learns through experience, rather
than the predicted theoretical classroom learner. She also expresses a preference for
analogue learning, which is metaphorical, looking behind the words. Leveling and
field independence are another unusual combination, though certainly not impos-
sible, since a field independent may perceive what is high priority to him or her, but
at the same time remember by lumping things together. She reports her MBTI type
as ENTJ. Her synoptic preferences are consistent with her intuitive type (N), which
correlates with synopsis (Ehrman and Leaver, 2002).

15
This type of learner is surprisingly rare in the Consultation Service, despite the large number of
learners with years of formal education and experience with the kinds of out-of-context learning that
classrooms frequently require. Most students who come to the Service report themselves as Type 3 or
Type 4, with the occasional fortunate Type 1 (who accesses both field independence and field sensitivity).
It is possible that Type 2 learners are well matched to FSI classrooms, and this certainly would have been
the case before communicative approaches came in, but pure Type 2 learners could probably run into
trouble when required to use language in context and in real time.
410 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

Deductive learning can be consistent with leveling, however, inasmuch as both can
serve as ways to reduce cognitive load: a deductive learner avoids the work of fig-
uring out rules from possibly chaotic raw material, and a leveler has fewer things to
keep in mind. Impulsivity, too, can be a way of reducing memory load: get it done
now, and move on to the next thing.
Nell’s self-reported random learning is possibly consistent with her stronger pref-
erence for concrete (experiential) learning, in which there are usually no external
guidelines to follow. Concrete, real-time learning often requires dealing with the
unexpected, and that is part of the random learner’s stock in trade. On the MSQ, she
indicates that mechanical exercises and rote memorization are not helpful for her.
Fig. 1 shows that Nell is probably somewhat more synoptic than ectenic, but she
relies on quite a few ectenic strategies. She needs to take conscious control of her
learning of details and structure, possibly in order to build a solid base for less pre-
dictable language use.

4.2.2. Information from the interview and the biographic background questionnaire
Unlike Mark, Nell had some experience informally learning her second language,
Polish, when she was in Warsaw previously. As a result, she entered her Polish
training with something of a head start. In Warsaw, she spoke some Polish with
friends and worked with a tutor, she says, but she regrets not having taken greater
advantage of the learning opportunity overseas. She is very clear about her need for
accuracy and ‘‘getting the pieces right.’’ She considers language learning to be
mostly a task of vocabulary building, and says the key is memorizing words, a view
that is somewhat consistent with her particular and analytic preferences. She gets
into a downward spiral of anxiety when she forgets words, and she reports con-
sistent anxiety about maintaining her vocabulary and is tempted to give up when she
finds herself forgetting a lot. She relies on writing words to help her remember. Her
teachers say that she is on track in her training, not having any particular difficulty.
They report that she speaks easily on topics for which she is confident of her voca-
bulary control, and listening comprehension is one of her strengths.

4.2.3. Suggestions for the student

Nell is not having any particular difficulty in class. In this case, the counselor is
more concerned with what will happen after Nell arrives at her overseas post, par-
ticularly in view of her strong reliance on atomistic, word-based strategies. The
recommendations for Nell (see Table 4), then, focus on preparing her for what is to
come after her language training and to build on her synoptic preferences.
Both Mark and Nell appreciate a well-designed syllabus but can manage reason-
ably well when the teacher deviates from it. If they were to be studying the same
language and were placed in the same class, the teachers would need to make some
adaptations to each: Mark might need help with pinpointing important content;
Nell might make this too much of a good thing. If they should get along well, they
might well end up helping each other. Of course, both would have to accept that
they cannot have everything exactly as they want it, even though the teachers can
meet their individual needs much of the time. In addition, the teachers could work
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 411

Table 4
Recommendations for Nell

Need or issue Sample suggestions

Coping with field insensitivity Listening for main points; reading for story and letting
herself stay open to new peripheral learning; find polish
speaking places and ‘‘hang out’’; listening for meaning;
seeing if she can listen ‘‘between the lines’’
Excessive perfectionism Take some specific amount of time daily to aim at fluency;
ignore errors except when they interfere with communication
of meaning
Going beyond vocabulary learning Working with material that is above her head, so she has to
to higher level language units guess and infer; writing or speaking using only a limited set
of words (to put the focus on making the most of the words);
writing paragraphs and short passages
Managing anxiety about errors and Reframing them as necessary to the learning process
forgotten words (if a learner isn’t making mistakes and forgetting things,
she or he is not being challenged)

with them to adapt their self-study activities both to tap into their preferences and to
extend their versatility into areas that are less preferred. For example, both learners
can benefit from work on coping with the kinds of unstructured situations that they
will meet daily after leaving the classroom and arriving at their overseas posts.

5. Conclusion

The E&L construct offers a rich variety of information about language learners
and language learning. It is useful alone, but even more so in tandem with other
information, especially that which comes from discussions and interviews with stu-
dents that add value, because they permit the interviewer and the learner to discuss
unusual or apparently contradictory profiles and make sense of them in the learner’s
current context.
This is not a simple model to use because of the many different profiles it can
yield, though some profiles are probably very rare (profile distributions comprise an
area for future research). However, those who are using it are finding it increasingly
rewarding as they come to understand it more deeply. The Canadian Foreign Ser-
vice Institute has even expressed interest in some teacher training so that its staff can
make use of the E&L Construct in their work with students (Claude Altschuler,
personal communication, May 24, 2003).
The E&L construct has the advantages both of generality and of specificity. The
synopsis-ectasis construct level can be used when a learner has a clear set of pre-
ferences tending to the right or the left of the chart and can thus allow more concise
descriptions with less detail from the subscales, which is not possible with mixed
profiles of the sort provided above short cut much unnecessary detail. At the same
time, it provides a more granular approach to individuals than most other models
because of the interplay among the ten subscales. All ten of the subscales, though
412 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

correlated in greater or lesser degree, can also operate independently of each other,
thus creating a multitude of possible combinations and permitting considerable dif-
ferentiation among individuals. And yet, the multiplicity of profiles still falls within
the same relatively standardized system, for comparability.
Through the counselor-student dialogue, the construct has served as the spring-
board for insight into individuals’ learning styles and teaching styles. There is much
research to be done on the E&L construct itself, its applications to such areas of
interest as very high level language learning, and on its relevance in multiple set-
tings—in classrooms and outside of them. In addition to FSI and the Canadian
Foreign Service Institute, the E&L model is also proving useful the National For-
eign Language Center, where it is the central learning style approach to assist users
in finding appropriate learning strategies in the federally funded LangNet project
(Catherine Ingold, personal communication, 2003). It is also in use in at least one US
university’s ESL program (John Green, personal communication, April 28, 2003). We
look forward to the opportunity to continue to explore its potential, its value, and its
validity in a variety of settings, including some outside the language field.

References

AILA Learner Autonomy Project Website. Available from http://lc.ust.hk/ailasc/activities.html#lapi.

Benson, P., Voller, P. (Eds.), 1997. Autonomy and Independence in Language Learning. Longman,
Harlow, England.
Biggs, J.B., 1992. Why and how do Hong Kong students learn? Using the Learning and Study Process
Questionnaires, Education papers #14. Faculty of Education, University of Hong Kong, Hong Kong.
Carbo, M., 1997. Reading styles times twenty. Educational Leadership 54, 38–42.
Carroll, J., Sapon, S.M., 1959. Modern Language Aptitude Test. Psychological Corporation, New York.
Chamot, A.U., Rubin, J., 1994. Comments on Janie Rees-Miller’s ‘‘A Critical Appraisal of Learner
Training: Theoretical Bases and Teaching Implications’’. TESOL Quarterly 28 (4), 771–776.
Chapelle, C., Green, P., 1992. Field independence/dependence in second language acquisition research.
Language Learning 42, 47–83.
Cotterall, S., Crabbe, D. (Eds.), 1999. Learner Autonomy in Language Learning: Defining the Field and
Effecting Change. Peter Lang, Frankfurt.
Davis, J., 1967. Technical Report #32: Concept Identification as a Function of Cognitive Style, Com-
plexity, and Training Procedure. Center for Cognitive Learning, Madison WI.
Dickinson, L., 1995. Autonomy and motivation: a literature review. System 23 (2), 165–174.
Dörnyei, Z., 2001. Teaching and Researching Motivation. Pearson Education Limited, England.
Dunn, R., Dunn, K., 1978. Teaching Students through their Individual Leaning Styles. Reston Publishing
Co, Reston, VA.
Ehrman, M.E., 1993. Ego boundaries revisited: toward a model of personality and learning. In: Alatis,
J.E. (Ed.), Strategic Interaction and Language acquisition: Theory, Practice, and Research. George-
town University, Washington, DC, pp. 331–362.
Ehrman, M.E., 1994a. The Type Differentiation Indicator and language learning success. Journal of
Psychological Type 30, 10–29.
Ehrman, M.E., 1994b. Weakest and strongest learners in intensive language training: a study of extremes.
In: Klee, C. (Ed.), Faces in a Crowd: Individual learners in Multisection Programs. Heinle & Heinle,
Boston, pp. 81–118.
Ehrman, M.E., 1996a. Psychological type and extremes of training outcomes in foreign language reading
proficiency. In: Homing, A., Sudol, R. (Eds.), Understanding Literacy: Personality Preferences in
Rhetorical and Psycholinguistic Contexts. Hampton Press, Cresskill, NJ, pp. 231–273.
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 413

Ehrman, M.E., 1996b. Understanding Second Language Learning Difficulties. Sage Publications, Thousand
Oaks, CA.
Ehrman, M.E., 1997. Field independence, field dependence, and field sensitivity. In: Reid, J. (Ed.),
Understanding Learning Styles in the Second Language Classroom. Regents Prentice Hall, Englewood
Cliffs, NJ, pp. 62–70.
Ehrman, M.E., 1998. The Modern Language Aptitude Test for predicting learning success and advising
students. Applied Language Learning 9 (1&2), 31–70.
Ehrman, M.E., 1999. Ego boundaries and tolerance of ambiguity in second language learning. In: Arnold,
J. (Ed.), Affect in Language Learning. Cambridge University Press, New York, pp. 68–86.
Ehrman, M.E., 2001. Bringing learning strategies to the learner: the FSI language learning consultation
service. In: Alatis, J.E., Tan, A. (Eds.), Language in Our Time: Bilingual Education and Official
English, Ebonics and Standard English, Immigration and the Unz Initiative. Georgetown University,
Washington DC, pp. 41–58.
Ehrman, M.E., Leaver, B.L., 2002. Development of a profile approach to learning style diagnosis.
Unpublished manuscript.
Esch, E.M. (Ed.), 1977. Self-Directed Learning and Autonomy. Department of Linguistics and Nancy,
France, CRAPEL, Cambridge, UK.
Fitzgerald, S., Morrall, A., Morrison, B., 1996. Catering for individual learning styles: experiences of
orienting students in an Asian self-access centre. In: Dickinson, L. (Ed.), Autonomy 2000: The Devel-
opment of Learning Independence in Language Learning. King Mongkut’s Institute of Technology
Thonburi, Bangkok, Thailand, pp. 55–69.
Gardner, R., Holzman, P.S., Klein, G.S., Linton, H., Spence, E., 1959. Cognitive control: a study of
individual consistencies in cognitive behavior. Psychological Issues 1 (4). (whole issue).
Gregorc, A., 1982. Learning style/brain research: harbinger of an emerging psychology. In: Keefe, J. (Ed.),
Student Learning Styles and Brain Behavior. National Association of Secondary School Principals,
Reston, VA, pp. 3–10.
Griggs, Shirley A., 1991. Learning Styles Counseling (ERIC Digest ED341890).
Hartmann, E., 1991. Boundaries in the Mind: A New Psychology of Personality. Basic, New York.
Hermann, N., 1982. The creative brain: part II. Training and Development Journal (November) 32–34.
Holyoak, K., 1984. Analogical thinking and human intelligence. In: Sternberg, R. (Ed.), Advances in the
Psychology of Human Intelligence. Lawrence Erlbaum, Hillsdale, NJ, pp. 199–230.
Horwitz, E.K., Young, D.J., 1991. Language Anxiety: From Theory and Research to Classroom Impli-
cations. Prentice Hall, Englewood Cliffs, NJ.
Holzman, P.S., Gardner, R.W., 1960. Leveling-sharpening and memory organization. Journal of Abnormal
and Social Psychology 61, 176–180.
Holzman, P.S., Klein, G.S., 1954. Cognitive system—principles of leveling and sharpening: individual
differences in assimilation effects in visual time error. Journal of Psychology 37, 105–122.
Hong Kong Association for Self-Access Learning and Development. Self Access Language Learning
(newsletter). Available from http://lc.ust.hk/HASALD/newsletter/newsletterSept02.pdf.
Huang, S.C., Chang, S.F., 1996. Self-Efficacy of English as a Second Language learner: An Example of
Four Learners. Eric Research Report (143). ED396536.
Jamieson, J., 1992. The cognitive styles of reflection/impulsivity and field independence/dependence and
ESL success. The Modern Language Journal 76 (4), 491–501.
Jonassen, D.H., Grabowski, B.L., 1993. Cognitive Controls. Handbook of Individual Differences,
Learning, and Instruction. Lawrence Erlbaum Associates, London, England.
Kagan, J., 1966. Reflection-impulsivity: the generality and dynamics of conceptual tempo. Journal of
Abnormal Psychology 71, 17–24.
Kagan, J., Moss, H.A., Sigel, I.E., 1963. Psychological significance of styles of conceptualization. Mono-
graphs of the Society for Research in Child Development 28 (2). (Serial No. 86).
Keefe, J.W., Monk, J.S., 1986. Learning Style Profile: Examiner’s Manual. National Association of
Secondary School Principals, Reston, VA.
Keogh, B.K., Donlan, McG., 1972. Field dependence, impulsivity, and learning disabilities. Journal of
Learning Disabilities 5, 331–336.
414 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415

Kirby, J.R., 1988. Style, strategy, and skill in reading. In: Schmeck, R. (Ed.), Learning Strategies and
Learning Styles. Plenum, New York, pp. 230–274.
Leaver, B.L., 1986. Hemisphericity of the brain and foreign-language teaching. Folia Slavica 8 (1), 76–90.
Leaver, B.L., 1998. Teaching the Whole Class, fifth ed. Kendall/Hunt, Dubuque, IA.
Leaver, B.L., 2003. Motivation at native-like levels of foreign-language proficiency: a research agenda.
Journal for Distinguished Language Studies 1 (1).
Leaver, B.L., Atwell, S., 2002. Preliminary qualitative findings from a study of the processes leading to the
Advanced Professional Proficiency Level (ILR4). In: Leaver, B.L., Shekhtman, B. (Eds.), Developing
Professional-Level Language Proficiency. Cambridge University Press, Cambridge, pp. 260–279.
Littlemore, J., 2001. Metaphoric competence: a language learning strength of students with a holistic
cognitive style? TESOL Quarterly 35 (3), 459–491.
Littlemore, J., 2003. The communicative effectiveness of different types of communication strategy.
System 31 (3), 331–347.
Lowery, R., 1982. Letteri’s information processing as related to cognitive structure. Paper presented at
Washington Junior High/Middle School Principals’ Association. Seattle, WA.
Messick, S., 1984. The nature of cognitive styles: problems and promise in educational practice. Educa-
tional Psychologist 59, 59–74.
Messick, S., 1994. The matter of style: manifestations of personality in cognition, learning, and teaching.
Educational Psychologist 29 (3), 121–136.
Miller, A., 1987. Cognitive styles: an integrated model. Educational Psychology 7, 251–268.
Morgan, H., 1997. Cognitive Styles and Classroom Learning. Praeger, Westport, CT.
Myers, I.B., McCaulley, M.H., Quenk, Naomi L., Hammer, A.L., 1998. MBTI Manual: A Guide to the
Development and Use of the Myers-Briggs Type Indicator1, third ed. Consulting Psychologists, Palo
Alto, CA.
Nickel, H., 1984. Begriffsbildung im Kindesalter: zum Verhaltnis von Stilen und Faehigkeiten. (The
Relationship of Abstraction in Childhood to Style and Ability.). H. Huber, Bern.
Oxford, R.L., Shearin, J., 1994. Language learning motivation: expanding the theoretical framework.
Modern Language Journal 78, 12–28.
Oxford, R.L., 1993. Style Analysis Survey. University of Alabama, Tuscaloosa, AL.
Pace University Counseling Services Website, 1998. Available from http://www.pace.edu/em/counseling/
learning.htm.
Pask, G., 1976. Styles and strategies of learning. British Journal of Educational Psychology 46, 128–148.
Ramı́rez III, M., Castañeda, A., 1974. Cultural Democracy, Bicognitive Development and Education.
Academic, New York.
Rees-Miller, J., 1993. A critical appraisal of learner training: theoretical bases and teaching implications.
TESOL Quarterly 27 (4), 679–689.
Rees-Miller, J., 1994. The author responds (response to Chamot and Rubin, 1994). TESOL Quarterly 28
(4), 776–781.
Reid, J.M., 1987. The learning style preferences of ESL students. TESOL Quarterly 21, 87–111.
Riding, R., Cheema, I., 1991. Cognitive styles—an overview and integration. Educational Psychology 11
(3 and 4), 193–215.
Riding, R., Rayner, S., 1998. Cognitive Styles and Learning Strategies. David Fulton Publishers, London.
Salkind, N., Wright, J., 1977. The development of reflection-impulsivity and cognitive efficiency. Human
Development 20, 377–387.
Schmeck, R., 1981. R, Improving learning by improving thinking. Educational Leadership 38, 384.
Schmeck, R.R., 1988. Learning Strategies and Learning Styles. Plenum, New York.
Sinatra, R., Stahl-Gemake, J., 1983. Using the Right Brain in Language Arts. Charles C. Thomas,
Springfield, IL.
Smith, K.K., Berg, D.N., 1987. Paradoxes of Group Life: Understanding Conflict, Paralysis, and Move-
ment in Group Dynamics. Jossey-Bass, San Francisco.
Stansfield, C.W., Hansen, J., 1983. Field dependence–independence as a variable in second-language cloze
test performance. TESOL Quarterly 17 (1), 29–38.
Support4Learning Website. Available from http://www.support4learning.org.uk/education/lstyles.htm.
 M. Ehrman, B.L. Leaver / System 31 (2003) 393–415 415

Torrance, E.P., Reynolds, C.R., Reigel, T., Ball, O., 1977. Your style of learning and thinking. The Gifted
Child Quarterly 21, 565–573.
University of Oklahoma Institute for Learning Styles. Available from http://arapaho.nsuok.edu/ oil/.
Wenden, Anita L., 1991. Learner Strategies for Learner Autonomy. Prentice-Hall, Englewood Cliffs, NJ.
Wenden, Anita L., 1995. Learner training in context: a knowledge-based approach. System 23 (2), 183–
194.
Wesche, M.B., 1981. Language aptitude measures in streaming, matching students with methods, and
diagnosis of learning problems. In: Diller, K.C. (Ed.), Individual Differences and Universals in
Language Learning Aptitude. Newbury House, Rowley, MA.
Witkin, H.A., 1969. Embedded Figures Test. Consulting Psychologists, Palo Alto, CA.
Witkin, H.A., Goodenough, D.R., 1981. Cognitive Styles: Essence and Origins: Field Dependence and
Field Independence. International Universities, New York.
Witkin, H., Moore, C.A., Goodenough, D., Cox, P.W., 1977. Field-dependent and field independent
styles and their educational implication. Review of Educational Research 47, 1–64.
Zelniker, T., Jeffrey, W.E., 1976. Reflective and impulsive children: strategies of information processing
underlying differences in problem solving. Monographs of the Society for Research. Child Development
41 (Serial No. 168).