Chapter 3

Individual Differences, Learning Styles and CELL

3.1 Introduction
The argument being proposed in this work is based on the premise that learners differ in

their approach to language learning in numerous ways, including their preferred

channels of perception, the learning processes they activate, their background

experience and education, their aptitude and motivation for language learning or

learning in general, and their age and level of maturity. These differences in turn affect

choices made by both learners and their teachers as to which materials, resources,

teaching and learning approaches, and forms of interaction are most effective for

language learning to proceed. Broadly speaking, these differences can be classified as

being either internal or external to the learner, though there are areas of obvious cross-

over such as learning styles where, as will be discussed below, external features of

tasks, for example, can influence features of perception which are essentially internal to

learners. The extent to which learner choices and differences can be accommodated

within a learning environment has a critical influence on the level of success in

language learning that individual learners achieve.

3.2 Relationship between CELL and learner differences

In a CELL environment, particularly where the computer resources are being used for

self-access purposes, it is important to be able to identify possible sources of individual

difference, in order better to tailor self-access resources and materials to cater for this

range. In such an environment, where there is much less opportunity for teachers to

mediate the learning processes directly, it becomes critical that individual differences

which have been identified in the research literature are incorporated into the

instructional design of computer-based resources. By implementing this approach,

teachers can participate in the mediation of learning, to create an environment that does
enhance language learning. The incorporation of multimedia (sound, video, graphics

PhD 130 D.Hoven

and text) into a CELL environment also provides learners with access to a wider range

of media in the self-access situation. In order for learners to take advantage of multiple

channels of perception, they need to be aware of their preferred styles of perception, the

extent to which the available media correspond to these, and how best to exploit this.

In addition, as mentioned in the previous chapters, the capacity for computers to be

used by learners to work in their own time and at their own pace is one of the principal

advantages in their use. However, in order for learners to use computer resources in this

way, they need to be aware of the variations that have been identified as regards

successful language learning, and to possess and be able to use skills in self-directed or

autonomous learning. Some of these skills and understandings can be acquired by

learners through awareness-raising and practice, while others cannot. The more

information learners have of these differences and variations, the more chance they will

have to make strategically effective choices in their language learning programs.

From a teacher’s point of view, decisions on the investment of money for purchase of

equipment, time for the training of staff and learners, or the development of materials,

must be grounded in considerations of the effectiveness of computer technology over

other media, or to complement them, and the implications of these factors for the

language learning programme. In Chapter 1, issues surrounding the roles of computers,

teachers, and learners, and the appropriateness of the use of technology in language

learning, have already been discussed. In the following sections, the human side of the

decision to incorporate computer technology in language learning programs will be

examined, particularly as regards individual differences and learning styles. This will

then be incorporated into the discussion in the following chapters of certain aspects of

learners’ awareness of these factors, and approaches to raising this awareness.

PhD 131 D.Hoven

3.3 Individual differences and learning styles – what are they?
Individual differences can be defined as those features or factors within individual

learners which influence differential success in language learning among learners, or

which can be identified as accounting for such differences. As will be discussed in more

detail in later sections, some of these features can be changed, modified, or developed

through instruction and awareness-raising, whereas others are less amenable to change.

This amenability to change becomes a critical component of a CELL, or indeed any,

learning environment, as it raises the dual question of which changes are desirable, and

whether such changes can be achieved through instruction.

The terms ‘learning style’ and ‘cognitive style’, though often used interchangeably, can

be differentiated, with ‘cognitive style’ being seen as a subset of the more

comprehensive ‘learning style’. Thus cognitive style has been defined as:

general characteristics of intellectual functioning that influence how one approaches a

learning task (Brown, 1987); the manner in which people perceive, conceptualize,
organize and recall information (Ellis, 1986).
(Wenden, 1991: 36)

Learning style, however, includes:

cognitive, affective and physiological behaviors that indicate learners’ characteristic and
consistent way of perceiving, interacting with and responding to the learning environment; more
concrete than cognitive style (Willing, 1988).

(Wenden, 1991: 36)

In their introduction to the discussion of explanations for differential success among

second language (L2) learners, Larsen-Freeman and Long (1991: 153) start with the age

factor as being ‘one of the most obvious potential explanations for the comparative lack

of success of second language learners’. L2 learners simply start later than first
language learners. Larsen-Freeman and Long go on to say, however, that there is ‘a

PhD 132 D.Hoven

host of other factors which have been proffered to explain differential success among

SL [second language] learners’. Among these other factors, they list: ‘language

aptitude, social-psychological factors [motivation and attitude], personality, cognitive

style, hemisphere specialization, learning strategies, and a few others.’

In addition to these, Altman (1980) offers previous experience with language learning,

sex of the learner, proficiency in the native language, and sense modality preference, as

well as ‘sociological preference’, by which he means preferred level of social

interaction (whether in groups, alone, or in pairs). These latter two factors correspond

to features mentioned by Wenden (1991: 37) as comprising ‘learning style’. Skehan

(1989) also deals with most of these factors, though from the perspective of their

implication in theories of second language acquisition (SLA). All of this has led Ellis

(1994: 471) to comment that, while ‘the study of IDs [individual differences] has

attracted a lot of attention in SLA research, and has made considerable advances’, the

constructs used to interpret research findings ‘are often vague and overlap in

indeterminate ways’, making it ‘difficult to synthesize the results of different studies,

and even more difficult to arrive at a coherent overall picture.’

While there is obviously some confusion and overlap among the findings of these

approaches to research and theory in this area, as can be seen above, there is sufficient

consistency on the main features for these to be considered and catered for in language

teaching and learning. In the development of an effective CELL package, therefore, it is

necessary to examine the factors within the individual which relate to success in

language learning in general. Only then can we identify those factors of general

language learning which impinge specifically on the use of computers to enhance

language learning, particularly in self-access contexts.

PhD 133 D.Hoven

3.3.1 Individual differences and developmental factors
The field of psycholinguistics has been a fruitful area of research into individual

differences in SLA. In this field, the focus has been on individual variation in terms of

developmental factors, both of cognitive maturity and ‘stages’ of acquisition of

morpho-syntactic features of language. While cognitive maturity will be discussed in

detail below in the section on age, developmental stages in morpho-syntactic

acquisition require some examination here, for their relation to the issue of teachability

and learnability mentioned above in section 3.3. In brief, developmental factors in SLA

include such features as stages, rate or order of acquisition, natural sequences (usually
of syntactic morphemes – Dulay and Burt, 1974; Meisel, Clahsen & Pienemann, 1981),

and interlanguage (Selinker, 1969) development.

It was Selinker who first coined the term ‘interlanguage’ to describe the form of the

target language (TL) produced by individual learners of it as comprising:

(1) knowledge underlying production in his native language (NL); (2) knowledge –
presumably at more than one level of consciousness – about the TL; and (3)
knowledge underlying attempted production of a TL norm.

(Selinker, 1971: 37)

Other terms which are used to refer to interpretations of interlanguage, as described

above, include ‘transitional competence’ (Corder, 1967) and ‘approximative systems’

(Nemser, 1971). It can be seen, therefore, that researchers from a range of paradigms

have recognised the importance of examining learners’ perceptions of the target

language, and their attempts to approximate the TL in production.

In 1974, two researchers into the emergence of natural sequences, Dulay and Burt, were

able to confirm their hypotheses from earlier studies, that the ‘same sequence of
acquisition of 11 functors, obtained by three different methods, provides strong

PhD 134 D.Hoven

evidence that children exposed to natural L2 speech acquire certain structures in a

universal order’ (1974: 50). Dulay and Burt call the process by which children are able

to do this ‘creative construction’ in which, ‘guided by universal innate mechanisms’,

they ‘formulate certain types of hypotheses about the language system being acquired,

until the mismatch between what they are exposed to and what they produce is

resolved’.

Meisel, Clahsen and Pienemann (1981) similarly, for adults learning German as a

second language in natural settings, were able to identify developmental stages in their

learners’ acquisition which were subject to ‘readiness’ constraints. To account for the

variations found among learners at the same developmental level, Meisel and his

colleagues postulated the interference of socio-psychological factors such as group

identification and acceptance of group values, as proposed by Schumann in his

acculturation theory discussed below. In other words, the acquisition of some features is

dependent on, for example, the learner being positively disposed towards, or valuing,

linguistic accuracy.

On the basis of this and previous studies, Meisel and colleagues also suggested the

implementation of staged instruction of morpho-syntactic features to match learners’

developmental stages. This suggestion has subsequently been further developed by

Pienemann as his ‘teachability hypothesis’ (1984: 35), whereby the order of

presentation of morpho-syntactic features in classroom instruction is designed to

correspond to the natural order of acquisition of those features. One of the major issues

investigated in relation to classroom instruction and the teaching of syntax has been this

complex question of the best way to teach learners syntactic rules, and which rules are

learnt best in what way: in other words, whether it is better in language instruction to

take an inductive or a deductive approach, or whether certain rules are better learnt by

one approach rather than the other.

PhD 135 D.Hoven

In an attempt to specify which morpho-syntactic rules might be easier to learn as

explicit knowledge, Green and Hecht (1992) examined a group of German learners of

English as the L2. They found that, in terms of explicit knowledge instruction, rules

relating to easily recognised categories, that lend themselves to mechanical application,

and are independent of the larger context, are easy to learn. Alternatively, when rules

were not subject to ‘simple exhaustive descriptions’, or were governed by features of

the broader context, such as aspect rules, they were found to be much more difficult to

learn.

Figure 3.1 The Role of Explicit Knowledge in L2 Acquisition

(Ellis, 1992: 97)

Code-oriented
Instruction

Explicit
Knowledge

noticing comparing Filter

INPUT INTAKE IL System

(Implicit Knowledge) OUTPUT

Studies such as this lead Ellis (1992: 97-8) to hypothesise a weak interface model to

explain the role of explicit knowledge in SLA. According to this model, illustrated in

Figure 3.1 above, explicit knowledge helps learners ‘to notice features in the input that

might otherwise be ignored’ (Ellis, 1992: 97), and as an ‘acquisition facilitator’

(Seliger, 1979) also assists learners to notice the gap between the input they are exposed
to and their own interlanguage output.

PhD 136 D.Hoven

Another researcher, Schmidt (1990; 1993), goes even further to suggest that ‘noticing is

the necessary and sufficient condition for converting input to intake’ and that ‘paying

attention is probably facilitative, and may be necessary if adult learners are to acquire

redundant grammatical features’ (1990: 129). This issue of conscious noticing will be

discussed in greater detail later in section 3.4.4.4 on the development of the concept of

strategies. As Slimani (1992) found in her studies of ‘uptake recall’, 89 percent of

‘uptaken’ items recorded on her student classroom charts were focussed upon at some

point during instruction. However, viewed differently, 11 percent of uptaken items only

occurred as part of classroom discourse, among peers, and even within the 89 percent,

there was considerable individual variation regarding which aspects of items were

attended to by her students. Taken altogether, Slimani’s findings seem to provide

further evidence for the sociocultural construction of meaning, as discussed in Chapter

1, in that it is within the interaction among users of the language, even novice users,

that the construction of meaning emerges. The combination of teacher and peer

discourse in the classroom provides both the context, and language and discourse

models, for learners.

Also in the area of classroom acquisition of syntactic structures, Felix (1981) found that

the acquisition of his German high school learners of English as a second language

paralleled L1 and naturalistic L2 acquisition. However, because of the constraints of

classroom instruction, his learners ‘were continuously forced to produce structures for

which, developmentally, they were not yet ready’ (1981: 87). To compensate for this,

they adopted ‘two basic strategies: (a) they followed principles of naturalistic

acquisition; or (b) they randomly selected any one structure from a finite repertoire’.

From this range of studies on developmental stages or natural sequences, it can be seen

that instruction or teaching has minimal effect on acquisition of morpho-syntactic

features, unless the learner is developmentally ready to acquire those features. As these

PhD 137 D.Hoven

stages are seemingly so unamenable to change, they do not warrant further discussion

in this work, except to the extent that learners and teachers alike need to be informed of

the existence and nature of this phenomenon. In addition, as the focus here is not

morpho-syntax, but rather the development of strategic listening, through the use of

authentic materials, the introduction of morpho-syntactic features is not controlled.

In this sense, the materials simulate a more ‘naturalistic’ acquisition environment. If

learners are at the appropriate stage to make meaning from a piece of authentic

material, and use the structures that appear in it, then they will be able to learn or

acquire those structural elements. While some of the listening tasks focus on specific

functional features of the language presented, such as distinguishing language used to

express cause and effect, for example, there is no explicit focus on syntactic elements.

From the evidence found in the research studies discussed above, however, this focus

on functional features and the negotiation of meaning required of learners in this CELL

environment should make it possible for them to form tenable hypotheses, whether

conscious or not, about the rules relating to the syntactic elements of such features.

As Schmidt (1990: 149) found, ‘incidental learning is certainly possible when task

demands focus attention on relevant features of the input’. In addition, Ellis (1994: 89)

stresses that ‘many formal language lessons are not directed at teaching new properties

but at enabling the learners to use features they have already partly acquired with

greater accuracy’. The listening and viewing material in the program described in this

work is thus not graded to cater for the progressive emergence of morpho-syntactic

features, but is rather directed at focussing attention on models of language as it is used,

on the assumption based on Schmidt’s findings above, that morpho-syntactic features

can be learnt incidentally under these circumstances.

PhD 138 D.Hoven

3.4 Some individual difference characteristics less amenable to change
Though considerable discussion continues about the identification and distinctiveness of

the various individual differences introduced earlier in section 3.1.1, the main interest

we have in these is the extent to which they influence learners’ capacity to learn from

the computer-enhanced materials being designed. It is therefore necessary to attempt to

identify, among these areas of difference, those features which are more amenable to

change. In order for optimal learning to be promoted for different learners using the

CELL materials, these features will need to be taken into account in the design of the

CELL program, and also it is these features that will need to become the focus of

awareness-raising within the materials themselves.

This process will also allow us to move away from those features which are less

amenable to change. It is necessary for us to be aware of their existence and the

influence they have in order to choose more appropriate materials, or direct learners to

make use of the materials in certain ways. In addition, as will be discussed in the next

chapter, through education and awareness-raising in a range of learning strategies, it is

possible to help learners compensate for less desirable or less effective characteristics

which they may have. However, within the materials themselves, and within a program

of instruction or learning little change can be effected in learners with regard to these

features. A learner may, for example, be beyond an age when language learning can

easily be achieved through the use of certain techniques, but he/she can be made aware

of, and encouraged to employ, other techniques or strategies which have been shown to

be more effective for older learners.

The factors or characteristics being discussed here are specific to individual learners,

and include age and maturity, personality, aptitude for language, hemispheric

processing orientation, motivating influences, learning and cognitive styles, and

strategies. As mentioned above, some of these individual and developmental factors are
obviously more amenable to being changed, developed, or influenced through

PhD 139 D.Hoven

education than others. Furthermore, the techniques learners use which have been

identified in the literature as ‘learning strategies’ are the most changeable or ‘trainable’

features. Broadly speaking, these include metacognitive strategies used to organise and

plan learning, cognitive strategies used to process language, and socio-affective

strategies employed to promote language learning through interpersonal interaction.

Language learning strategies will become the focus of the next chapter.

Specifically, developmental factors discussed earlier, and age, personality, language

aptitude, and hemispheric orientation are less changeable, while attitude, motivation,

possibly learning and cognitive style, and learning strategies are more open to

influence. As features less amenable to change have an effect on language learning,

however, it is important to outline briefly why this is so, and what these effects are,

before discussing in greater depth those aspects of individual learners which can

become the focus of education and awareness-raising. The area of learning strategy

identification and training will then form the major part of the discussion in the

following chapter, as such training forms an integral part of the development of learner

autonomy which is necessary to the efficient use of a learner-centred CELL package.

3.4.1 Age as an unchangeable factor

As outlined by numerous researchers and theorists in SLA, age is a contentious factor

when it comes to language learning (McLaughlin, 1987: 29; Genesee, 1988; Skehan,

1989: 137; Larsen-Freeman & Long, 1991: 154). For younger learners, age may be a

positive factor, not from the simplistic point of view that younger learners learn

languages better (Dulay & Burt, 1974), but rather from the view that they have a longer

period of time to learn when starting younger (Genesee, 1988: 105). In addition, when

children start learning a language at a younger age, their cognitive maturation

(Donaldson, 1978) is developing simultaneously with their language learning.

PhD 140 D.Hoven

Some researchers see the lack of cognitive maturity as an advantage, claiming that the

absence of Piagetian ‘formal operations’ allows children to learn their second language

using processes much like those involved in learning their first language (Krashen,

1982a). Others, however, find that the cognitive maturity of older learners favours the

way languages are generally taught in schools, using more abstract processes, and

decontextualised language (Genesee, 1988: 104). Genesee also concludes that

naturalistic approaches to language teaching produce superior learning regardless of the

age of learners, while matching instructional style with the cognitive development and

interests of learners (Stevens, 1983) also improves ultimate achievement.

A considerable number of research studies has been undertaken in an attempt to

determine if there is an optimum age to start learning a language (Genesee, 1976;

Snow, 1983; Ellis, 1985), and if so, what this age might be (Snow and Hoefnagel-

Hohle, 1978; Dulay and Burt, 1974), whether there might be a point beyond which

language learning abilities stultify (Snow, 1987; Penfield and Roberts, 1959;

Lenneberg, 1967; Krashen, 1974; Neufeld, 1979), the differing approaches to language

learning employed at varying ages (Skehan, 1986, Wong-Fillmore, 1979; Chesterfield

and Chesterfield, 1985; Krashen, 1982; O’Malley et al., 1985), and the relative

effectiveness of beginning to learn a language at an age that is not the ‘optimum’

(Skehan, 1989: 137; Hatch, 1983; McLaughlin, 1984; Krashen, Long and Scarcella,

1979).

To date, the conclusions seem to consistent with the oft-cited adage ‘older is faster, but

younger is better’ (Krashen et al., 1979; Larsen-Freeman & Long, 1991: 155), where

‘faster’ refers to short-term gains found with older learners, while ‘better’ refers to the

ultimate attainment of native-like proficiency that is possible for younger learners to

eventually approximate. However, there have been several studies, particularly in early

and later bilingual programs, that indicate that early adolescence may be the optimum

time to begin (McLaughlin, 1984; Swain and Lapkin, 1982). In relation to types of
instruction, as mentioned above (Genesee, 1988), on the basis of data from immersion

PhD 141 D.Hoven

programs in Montreal, naturalistic language learning environments have been shown to

be more successful, regardless of the age of the learners.

Studies of the differences between child L2 learning and that of older learners have

frequently hypothesised the existence of a critical or sensitive period, after which

native-like competence in a second language learning is difficult, if not impossible.

Many of these studies have used data collected from immigrants learning their L2 while

living in the target culture, where the age of arrival provides a clear reference point.

Several of these have indicated that age of arrival in the target language culture is

inversely proportionate to ultimate level of proficiency (Oyama, 1976; Fathman, 1976).

In terms of phonology, other studies indicate that it may even be impossible to achieve

native-like speech if language exposure begins after the age of about six years (Payne,

1980; Tahta et al., 1981; Major, 1987), though some exceptions have been found

(Neufeld, 1979).

In his examination of the role of consciousness and attention in language learning,

Schmidt (1990: 145) was intrigued to note that the age range during which the shift

occurs from child to adult consciousness ‘approximates to the sensitive period for

language acquisition’. Whereas child consciousness is characterised by ‘a passive mode

that includes an open awareness of the environment’, adults exhibit a ‘more controlled

mode that includes the strategic allocation of attention’. Schmidt used studies by Ceci

and Howe (1982) and Miller (1985) on strategic attention and incidental learning in

children, together with his own 1983 study of adults’ failure to learn grammar through

communicative activities, and compared these with the observation of McLaughlin et

al. (1983) that ‘children learn the rules of grammar as a by-product of trying to
communicate’. From these, he proposes the hypothesis that it is children’s lack of

strategic attention which causes them to notice grammar features that are

communicatively less important. Adults, therefore, having more control over their focus

PhD 142 D.Hoven

of attention, tend not to pay attention to ‘redundant and communicatively less important

grammatical features’.

Several other explanations have been offered for the age-related changes in capacity for

language learning referred to above. Larsen-Freeman and Long (1991: 163-7) classify

these explanations as follows:

1. socio-psychological (involving ethnic and group identification, acculturation,

and accommodation theories, which will be discussed in more detail in section

3.4.3 on personality)

2. input-dependent (based on features of the input and discourse models discussed

below)

3. cognitive (involving Piaget’s maturational stages mentioned earlier)

4. neurological (implicating a sensitive or critical period mentioned above and

discussed in detail in section 3.4.2).

Input-dependent explanations (point 2. above) derive from the observed differences in

the classroom environments and teacher/learner interactions among various levels of

education, namely pre-school and primary, contrasted with secondary and adult

contexts (Curtain and Pesola, 1988; Wankowski, 1973). In pre-school and primary

learning environments, language play is encouraged, and learners are exposed to more

input that is concrete and meaningful, providing simple syntactic models for them to

follow.

Another argument that is advanced to explain the perceived advantages of the kinds of

input to which younger learners are exposed, particularly in naturalistic learning

environments, is that the second language input experienced by younger learners more

closely approximates that of their first language (Krashen, 1981; Larsen-Freeman &
Long, 1991). The latter argument is premised on the principles that, as children acquire

PhD 143 D.Hoven

their L1 in naturalistic settings, if teachers can set up an L2 learning environment that

approximates the characteristics of a L1 environment, the learners could then acquire

their L2 using processes similar to those activated in the acquisition of the L1 (Krashen

and Terrell, 1983; Larsen-Freeman and Long, 1991: 140).

Explanations based on principles of cognitive maturation (point 3 above) hinge on

developmental features as represented in Piaget’s concrete and formal operations

stages. Children and adults would therefore learn second languages differently: the

former, still having access to the Language Acquisition Device (LAD - Chomsky,

1965) used for L1 acquisition, could utilise this in much the same way for their second

language; the latter, having more mature cognitive skills, could approach the second

language as a series of problems to solve (Larsen-Freeman and Long, 1991: 163).

This problem-solving element is manifest in the cognitive inductive, deductive,

analytic, and reasoning strategies adolescents and adults use in higher frequency and

number than children (O’Malley & Chamot, 1990). Particularly in instructed learning

environments, problem-solving has been identified as one of the two basic processes,

along with memorisation, involved in the acquisition of explicit knowledge (Ellis,

1994b: 90). This strategic approach has been considered in the choice and design of

tasks in developing the taxonomy of listening comprehension skills discussed in the last

chapter and Chapter 4, as some researchers claim that such ability can develop quite

early (Karmiloff-Smith, 1986).

Currently, however, insufficient investigation has been carried out into the differences

between children and adults as regards the language learning processes operating. As

Larsen-Freeman and Long comment, ‘longitudinal child-adult interlanguage

comparisons will clearly be the key, yet are virtually non-existent to date’ (1991: 166-

7). Larsen-Freeman and Long acknowledge that biological constraints on SLA, such as
that espousd in the critical (Penfield and Roberts, 1959; Lenneberg, 1967) or sensitive

PhD 144 D.Hoven

(Krashen, 1973; Lamandella, 1977; Seliger, 1978; Scovel, 1988; Breathnach, 1993)

period hypothesis, might exist as a plausible explanation for the ‘maturational

constraints on child language development’ found in numerous studies. However, they

maintain that, in the absence of interlanguage comparisons, and the ‘lack of any known

neurological changes clearly coinciding with changes in SLA ability’, they cannot

assume that child and adult SLA processes are different.

3.4.2 Hemispheric orientation and lateralisation

Information on the preferred modes or channels of perception and processing of

different learners helps us to ensure that the language learning materials we design cater

for these preferences, thereby providing as many learners as possible with optimal

learning opportunities.

Studies of neurological factors in second language acquisition are hampered by the

difficulty of obtaining reliable data on the mechanisms and processes involved (Jacobs,

1988). An unfortunate factor in many of these studies is that the investigation of the

cause of subjects’ neurological dysfunction is frequently the purpose behind the

investigation of their processing. While this has to be borne in mind as regards the

applicability or transferability of the findings to normally functioning people,

conclusions and speculations emerging from these studies illuminate an area of interest

to us: hemispheric lateralisation.

Several researchers (Joiner, 1984; Fountain & Fillmer, 1987; Danesi, 1988; Curro,

1995) into the effects of hemispheric lateralisation on language learning advocate

activities and a learning program that promote the integration of processing from both

hemispheres of the brain, in order for learners to realise their maximum potential. At

the same time, these researchers and others (Walsh & Diller, 1981) maintain that older

learners can take greater advantage of the ‘higher-order’ language functions that
develop with cognitive maturity. As Walsh & Diller comment (1981: 18): ‘As people

PhD 145 D.Hoven

grow older and cognitively more mature, their increasing higher-order cortical

functions [such as semantic relations] allow them to do more than they could before

with their lower-order aptitudes and functions’. This provides further evidence for the

importance of the awareness-raising for metacognitive and cognitive learning strategies

incorporated into the software package presented in this work.

However, as mentioned earlier, the exact nature of any relationship between

neurological changes and differing processing techniques in language learning remains

speculative. This is because, despite advances in the area, equipment for measuring

neural activity is still not accurate enough, and because tests are so indeterminate. Thus,

for example, even though the most modern techniques for mapping the activity of

language areas of the brains of healthy, normally-functioning human beings can

confirm the earlier findings of much of the research based on brain-dysfunction

individuals (Begley, et al. 1992; Damasio, 1992), this information cannot yet tell us

how or why the events and activity identified relate to the actual learning process. In

addition, in the collection of their data, researchers are necessarily restricted to

investigating neural activity in living, breathing people, and are therefore constrained in

the range and nature of tests they can conduct. There remains, therefore, no reliable

means of determining what exactly they are seeing, and how this relates specifically to

language learning.

Nevertheless, it is important to examine other factors involved in hemispheric

lateralisation, both for the purpose of demonstrating the range of such factors that

contribute to individual differences among learners, and to eliminate some of these

from our scope in the design of a learner-centred software package. In addition to those

mentioned above, studies of hemispheric lateralisation have focussed on the influence

of lateralisation on modes of teaching and learning, script nature and direction, and type

of first or second language. It is the first of these areas of focus that interests us here for
the design of a software package that caters for individual differences.

PhD 146 D.Hoven

On the issue of mode of teaching/learning, Genesee (1982) proposes, for example, that

‘(w)hat distinguishes bilinguals from monolinguals when neurolinguistic differences

occur, is their use of strategies that deploy the specialised processes of the two

hemispheres differently’. Specifically, late L2 acquirers deploy left- and right-

hemisphere-based processes differently, and these strategy differences may characterise

both L1 and L2 processing.

The research suggests that children exhibit more right hemispheric involvement in

language processing than adults, and that this is also true for beginning second language

learners. In addition, it is hypothesised from the studies that while left hemisphere

processing always predominates for language functions, there left hemisphere

processing seemed greater in formal learning contexts than in informal ones. Therefore

Genesee (1988) both cautions that such studies examine language processing, not

learning, and cites evidence for greater left hemisphere involvement in more formal
learning contexts, with greater right hemispheric processing in more informal contexts.

Possibly the most important aspect of the research into hemispheric lateralisation and

differential processing is the implications of these findings for teaching and learning.

Curro (1995), frames her discussion of this as the necessary consideration of bimodality

in syllabus planning and task selection and design. As she explains, on the basis of

research findings (Goldberg & Costa, 1981; Danesi, 1988; Palotta, 1993):

the right hemisphere has a more compatible anatomical structure for new material.
Given its intricate network of neuronal connectivity, it can accommodate information
for which no preexistent codes or schemes are available. [...] According to Danesi, the
construct of bimodality does not constitute a syllabus format or method. ‘It simply alerts
the teacher to certain fundamental requirements of human learning, such as the need to
start a novel learning task in a sensorial and concrete way and to end it in a more
formalized fashion (the learning flow principle)’ (Danesi, 1991: 28).

(Curro, 1995: 88)

PhD 147 D.Hoven

Furthermore, a bimodal approach is also indicated because of the differential perceptual

processing in the right and left hemispheres, and particularly in teaching and learning

involving both listening and viewing. As Danesi (1987, citing Sperry) reminds us,

while neuroscience has fairly clearly established that verbal thinking is represented in

the left hemisphere and non-verbal thinking in the right, our current teaching and

learning practices tend to neglect the non-verbal. This again reinforces the findings of

Hurley (1992) and Kellerman (1992) that language learning materials incorporating the

non-verbal or co-verbal aspects of language need to be provided for learners. This

applies equally to computer-enhanced language learning materials, particularly those

incorporating multimedia.

3.4.3 Personality and aptitude

These two areas of individual difference will be dealt with together in this section, as

several researchers have observed certain correspondences between personality

characteristics such as extroversion/introversion with aptitude and classroom language

learning success (Hamayan et al., 1977; Wong-Fillmore, 1982). Furthermore, teachers

and researchers have hypothesised about the desirability of matching teaching style and

the nature of classroom interaction with the personality characteristics of individual

learners. While it is proposed that such matches will promote learning, little evidence

has been provided for any consistency in the findings (Hamayan et al.; 1977; Wong-

Fillmore, 1982).

In addition, personality and aptitude are both features which are relatively difficult to

influence or change (Skehan, 1989: 38 - 39; 71 - 72) within a language programme. A

further complication resides in the experience of researchers that tests of the existence

and nature of personality traits are unreliable (Larsen-Freeman & Long, 1991;

McLaughlin, 1985). Despite these difficulties, studies of personality and aptitude

provide us with valuable information on the range of influencing features in the

PhD 148 D.Hoven

learning process, and which characteristics might be more desirable in the promotion of

learning than others, particularly as they relate to individual learning. This information

then enables us to make provision in language learning programs and materials for the

range of characteristics that learners bring to them.

3.4.3.1 Personality
The personality traits which occur most frequently in the literature surveyed are

extroversion/introversion. In studies on the desirability of these traits in relation to

success in general academic performance, it was predicted that introversion would show

higher correlations with academic achievement (Entwistle and Wilson, 1977). This

prediction was based on the assumption that introverted learners would be more self-

reliant and more prepared to study alone. This had, in fact, been shown to be the case

(Entwistle and Entwistle, 1970), though the cross-cultural applicability of the inventory

used (the ‘approaches to studying’ inventory developed by Ramsden and Entwistle,

1981) has been questioned (Watkins, 1982). In Hattie and Watkins’ (1981) study of

Filipino tertiary students, for example, the researchers failed to find evidence of the

study dimensions upon which the inventory is based.

In a study of the relative advantages of extroversion and introversion at primary and

secondary school, Wankowski (1973) found that before puberty, extroversion was

associated with higher academic achievement, with introversion being more dominant

an indicator after puberty. The explanation he offered for this was that after puberty, or

at high school, the nature of the teaching and testing approaches changed to favour

introverted learners. In the primary school, tasks were more often oriented towards

group work, while at high school, with greater subject specialisation, the emphasis was

more on individual work and homework. In a self-access environment for learners of

secondary and post-secondary ages, these findings indicate the necessity to provide

activities that range across contexts of working alone, in pairs, and in groups, and to
provide activities which help extroverted learners develop strategies for working alone.

PhD 149 D.Hoven

In the field of language learning, particularly in naturalistic or communicative language

learning contexts, extroversion seems to be the desirable trait, because of the emphasis,

discussed in Chapter 1, on interaction and communication. Where successful language

learning is gauged by the proficiency with which the learner can negotiate meaning and

interact in the target language, characteristics such as sociability and seeking out

occasions for interaction with speakers of the target language assist in the acquisition of

the target language (Strong, 1983; Krashen, 1985; Long, 1985).

A study by Swain (1985) also indicates that not only is it necessary for a language

learner to have useful or comprehensible input (Krashen, 1982) from speakers of the

target language, but that it is equally important for the learner to experiment with

communicating meaning in interactive encounters (‘comprehensible output’ – Swain,

1985). Such behaviours are typical of extroverted personalities. As the ‘Good Language

Learner’ studies of Naiman and colleagues (1978), and others (Wong-Fillmore, 1979;

Rubin, 1975; 1987) have indicated, being active and pursuing occasions for interaction

are common characteristics of successful language learners. These and other

characteristics will be discussed in more detail in the next chapter.

Alternatively, where language proficiency is gauged by more decontextualised means

such as discrete point grammar tests, pronunciation and standardised reading tests

(Suter, 1976; Busch, 1982), the correlations favour characteristics of introversion over

extroversion. As communicative language teaching, which favours learner interaction,

production, and engagement with the second language has been the major teaching

approach since the early 80s (Swan, 1985a, b), we would expect higher correlations

with language proficiency for learners with stronger extroversion than introversion

characteristics. It may be the case that the relationship between extroversion and

introversion and language learning has not yet become clear, because studies are
hampered by the predominantly decontextualised nature of the tests of language

PhD 150 D.Hoven

proficiency usually employed in such studies (Skehan, 1989). Although more

communicative tests have been developed, such as those of Ingram (1984) and

McNamara (1996), these have not been the tests used in studies of

extroversion/introversion. Until such communicative tests are included in comparative

studies of these traits in relation to language proficiency, the conclusions remain

questionable. Indeed, both Skehan (1989) and Ellis (1994) argue for studies using a

more naturalistic methodology in that, in addition to showing ‘the dynamic nature of

the interaction between the more malleable aspects of individual difference’, these can

‘provide measures of learning based on natural language use rather than on tests or

ratings’ (Ellis: 1994: 524).

Aspects of gender have also been studied in relation to personality factors in language

learning environments involving computers. Abraham and Liou (1991), Underwood et

al. (1991), Dye (1995), and Meunier (1996) have all found that females tend to be
dominated by males in mixed-gender language learning activities. However, as Meunier

comments (1996: 50): ‘the poor performance of female students seems to be related

more to interaction styles and personality differences than to their level of computer

literacy’. In her study of mixed- and single-gender dyads working at a computer,

Meunier finds personality to be a stronger predictor than gender of language learning

and keyboard control, and, together with ‘position at the computer’, personality was

also a stronger predictor of verbal interaction.

As a result of these findings, she recommends that (1996: 67): ‘foreign language

software designers should attempt to integrate more than one path of data exploration

along with a large diversity of information to accommodate personality and gender

differences’. Additional recommendations include: pairing linear and non-linear

learners in hypertext environments ‘for the purpose of being exposed to other cognitive

styles’ (Meunier, 1996: 67); and incorporating a ‘road map’ to provide linear learners
with some structured support in their exploratory learning. While issues of cognitive

PhD 151 D.Hoven

style will be discussed in more detail in section 3.4.4, in relation to Meunier’s findings,

the software package described in Chapters 5 and 6 of this work has been designed to

make provision for learners’ different cognitive styles and to take into account the

difficulties linear learners can experience in exploratory multimedia and hypertext

environments. This is achieved partially through transparent menu and navigation

systems, and partially through the division of the package into three progressively less

structured layers between which learners can alternate as their needs determine.

3.4.3.2 Related personality characteristics

Apart from the introversion/extroversion trait, other affective or personality

characteristics, many of which are inter-related, have been investigated for their

influence on second language learning. The ones which will be dealt with here include:

self-esteem, anxiety, risk-taking, sensitivity to rejection, empathy, inhibition, and

tolerance of ambiguity (Larsen-Freeman and Long, 1991). Anxiety in particular is

mentioned by Ellis (1994: 524), along with motivation, as being one of the ‘more

malleable aspects of individual difference’. These will be discussed below.

Self-esteem and anxiety have been shown to have only low to moderate correlations in

predicting language learning success, a certain level of each characteristic being

necessary, and too much being deleterious. Self-esteem can be defined as ‘self-

judgement of worth or value, based on feelings of efficacy’ (Scarcella & Oxford, 1992:

57), and, in common with anxiety described below, can be either a trait or global

characteristic, or a task- or situation-dependent state. For example, self-esteem has been

found to be relatively task- or instructor-dependent (Hyde, 1979; Chastain, 1975;

Scovel, 1978), with different ratings emerging for the same learners with different

teachers and subject areas. For learners with low global self-esteem, it can be built up

through the use of such affective learning strategies as ‘taking your emotional

temperature’ and ‘encouraging yourself’ (Oxford, 1990:21), as well as the various

PhD 152 D.Hoven

metacognitive strategies involving planning and organising learning (Oxford, 1990:

20).

In a self-access CELL environment, such as the one described in Chapters 5 and 6, the

issue of low self-esteem is addressed by providing a context in which learners can

succeed at their own level. Referred to in the literature as ‘individualised learning’, this

is one aspect of the oft-cited advantage of computers: that they allow learners to work

at their own pace and level (Underwood, 1984: 38; Batley & Freudenstein, 1991: 10,

12-13; Weible, 1987: 81; Wyatt, 1987: 91). However, this can only apply if the

appropriate materials and activities are provided and if the learner has an understanding

of how to choose appropriately, as illustrated in Chapters 5 and 6.

Anxiety in language learning follows the patterns already illustrated in mainstream

learning, where a critical point is reached at which learning and anxiety are at an

optimum balance in relation to one another (Biggs and Telfer, 1981: 113 - 5). After this

point, performance drops off rapidly as anxiety increases. Researchers in the area of

language learning have defined anxiety as being either facilitating (Scovel, 1978), or

debilitating. The former characteristic, which can also be described as ‘positive tension’
(Scarcella & Oxford, 1992: 54), promotes learning, while the latter can produce worry

and self-doubt, causing the learner to participate less and perform less well. Horwitz

(1990) sees the role of facilitating anxiety as being limited to simple tasks which may,

therefore, exclude language learning (Williams, 1991). Other researchers (Scarcella &

Oxford, 1992: 54) argue that ‘a certain amount of positive tension is helpful for

language learners’. In this light, Scarcella and Oxford make a claim for the use of the

term anxiety to be restricted to negative contexts only, as this is how it is normally

construed.

Performance anxiety, engendered by having to produce language or ‘perform’ in front

of others, is most commonly exhibited in classroom contexts This is considerably eased

PhD 153 D.Hoven

in a self-access situation as there are no witnesses to a learner’s performance. Where the

only ‘witness’ to one’s language interaction and performance is a computer, it has been

suggested that learners build a special relationship with the computer, experiencing a

sense of privacy, and increasing their motivation to learn and their sense of

achievement (Batley & Freudenstein, 1991: 11). It is in contexts such as this that

working with computers is regarded as being less threatening than public (i.e.

classroom) performance.

In common with anxiety, risk-taking can be a valuable characteristic in language

learning (Rubin, 1975; Ely, 1986), although it may result in reduced performance when

over-used (Beebe, 1980, 1983). Risks here are exemplified variously by guessing

meanings, producing spoken utterances in an attempt to communicate, even though

mistakes are also produced, and initiating conversations with native speakers of the

target language. Such risks are a normal part of learning a language in naturalistic

settings, and have been shown to characterise successful language learners (Oxford,

1990; O’Malley & Chamot, 1991). They are also an essential component of the

hypothesis-testing approach to language learning (Bialystok, 1983) which is consistent

with SLA research findings in studies of learners’ development of their interlanguage

(Selinker, 1969; 1971; 1972). Naiman (1978), on the other hand, has found little

correlation, either positive or negative, between sensitivity to rejection, a trait intrinsic

to risk-taking, and language learning success. One possible explanation for this could

be that the test of language proficiency is not an appropriate instrument for measuring

the usefulness or otherwise of this characteristic.

As well as being difficult to test accurately, empathy has produced mixed correlations

when matched with language learning. In 1972 (b), Guiora and others hypothesised the

‘permeability of language ego boundaries’ in an attempt to explain the strong facility

found in children for accurate pronunciation, mentioned earlier. Children, who have
low ego boundaries, have less difficulty with, and attain greater accuracy in,

PhD 154 D.Hoven

pronunciation of a second language than older learners. In fact, as ego boundaries

increase in strength with age, this facility for accurate pronunciation diminishes. Guiora

and colleagues also investigated the influence of inhibition (Guiora et al. 1972a) but

found little firm evidence for this in either direction. Variations of both these

characteristics are implicated in theories of acculturation and assimilation, which will

be discussed further in the section on socio-psychological factors.

Lastly, tolerance for ambiguity, together with risk-taking, seem to exhibit the highest

correlation with success in second language learning (Naiman, 1978; Chapelle and

Roberts, 1986). Interestingly, it is this tolerance which underpins success in humanistic,

top-down, or holistic approaches to language learning where the focus is on the

meaning of a whole text or discourse of a text (Canale & Swain, 1980; Canale, 1983),

rather than that of individual words. In order to be able to function on such a holistic

basis, a high tolerance for ambiguity is necessary to avoid the anxiety which might

ensue when the meaning of individual words within the text may be unknown.

Tolerance for ambiguity has also been linked to self-esteem (Scarcella & Oxford,

1992), in that learners with high self-esteem are more likely to feel confident enough to

accept initial lack of total comprehension or possible mismatches between their existing

schema and the input they are experiencing.

If we accept this, the development and encouragement of compensation strategies such

as paraphrasing and guessing (Oxford, 1990: 19) may help minimise the perception of

risk on the part of students with lower self-esteem. On the other hand, too high a

tolerance for ambiguity might actually inhibit language learning by not allowing

learners to perceive and analyse some rules of language. As will be discussed in more

detail in the next chapter, such learners need exposure to cognitive strategies such as

practising, analysing and reasoning, and creating structure for input and output (Oxford,

1990).

PhD 155 D.Hoven

It can be seen, therefore, that while elements of all of the characteristics mentioned in

this section are typical of successful language learners, moderation in each seems to

produce the best effects. Thus, while some anxiety, in the form of ‘positive tension’,

seems to assist performance, it cannot be allowed to reach debilitating levels. Whereas

good language learners have sufficient self-esteem to take risks in interpreting and

producing their own utterances, they also need to cultivate a critical internal monitor of

the language to which they are exposed, in order to formulate the appropriate

hypotheses about its structures and functions.

3.4.3.3 Aptitude
When we come to the question of the role of aptitude in language learning, Skehan

(1989: 136) comments: ‘(s)ome people are endowed with better cognitive/linguistic

ability than others’. But what constitutes aptitude for language learning, and can it be

improved through instruction, experience, acculturation, or some other means?

Skehan’s understanding of language aptitude can be defined in terms of learners’ ability

to deal with disembedded or decontextualised language – the kind of language, in fact,

that was typical of traditional grammar-translation language classes and tests. Based on

Carroll’s (1965) comprehensive testing in the area, Skehan goes on to claim that:

Aptitude is multi-componential [...] there are (fairly) independent components of

aptitude [... such as] a language analytic capacity; memory ability; and phonemic
coding ability.
(Skehan, 1989: 137)

Though Skehan lists the above three components as being the most critical, Carroll also

mentions a fourth component, grammatical sensitivity, which he interprets as the

capacity to benefit from grammatical instruction when it is offered. There does,

however, seem to be a close relationship between the processes inherent in grammatical

sensitivity and those of a language analytic capacity, and all four components seem to

PhD 156 D.Hoven

be related to cognitive perception or learning styles, which will be discussed in the next

section.

The fact that these components of aptitude are identifiable provides strong grounds for

the design of instructional materials that would cater for the range of individual learner

needs represented. In commenting about the strong relationship found by Cummins and

Nakajima (1987) between the L1 and L2 reading skills of Japanese immigrant students,

McLaughlin (1990) makes the claim that some of the strength of this relationship is

based on general aptitude. However, he goes on to argue that:

[...] aptitude should not be viewed as a static personality trait; novices can become
experts with experience. I believe there is an interdependence between first and second
language in the cognitive/academic domain because experience with one language gives
the learner strategies and metacognitive skills that generalize to subsequent languages.
[...] Teachers, in my view need to do more than provide “comprehensible input”. They
need to make these strategies and metacognitive skills available to learners.

(McLaughlin, 1990: 173)

McLaughlin’s comments here are echoed in the research reported in Chapter 1 of this

work in the area of socio-cultural ‘apprenticeships’ in the learning of a second language

through adopting the social norms, practices, and patterns of the target culture, and

thereby also the target language (Lantolf, 1994; Donato & McCormick, 1994). For

McLaughlin, then, aptitude can improve with experience that derives from transfer of

skills from the first language, and also through the development of strategies and

metacognitive skills. As we will see in the next chapter, most researchers in the field of

strategies have found metacognitive skills or strategies to be particularly susceptible to

training (O’Malley & Chamot, 1991). This being so, there is a place for education or

awareness-raising among learners in the utilisation of more productive metacognitive

strategies.

PhD 157 D.Hoven

In their discussion of aptitude, Larsen-Freeman and Long (1991) cite earlier work of

Politzer and Weiss (1969) and Carroll (1979) as showing evidence of aptitude being

unamenable to training, while Neufeld (1978) found aptitude to be dependent on

experience. However, while accepting the difficulty of training learners in certain skills

in order to improve their aptitude, Carroll (1965) recognises that some improvement

can be achieved with high quality instruction. As Larsen-Freeman and Long (1991)

argue, when instruction is of lower quality, learners are forced to fall back on their own

resources, and it is in these cases that higher aptitude makes a difference.

In a study of adults studying French in Canada, Wesche (1981) found positive effects

for matching learners with appropriate teaching methods and materials based on their

learning style preferences exhibited in personality profiles. On the basis of this and

Carroll’s findings above, there seem to be clear advantages for a range of learners in

using multimedia CELL materials which are learner-driven. These materials should

vary in the channels of perception they activate (visual, textual, auditory), and learners

should be able to choose the types of tasks and materials they feel most comfortable

using. In this model, more challenging tasks and materials would be those which

exploit less preferred channels of perception or strategies. As will be discussed in

Chapters 5 and 6, the software package developed as a prototype of this model, for

example, makes available to learners three levels of working through the material: a

highly structured level, a level which focusses on cognitive demand, and an exploratory

level for more autonomous learners.

Given that aptitude is an important factor in language learning, and that learners’

aptitude varies, it is important that language learning materials reflect an

accommodation for this variability. For a learner-centred CELL package to make such

accommodation, and in keeping with Meunier’s (1996) ‘road map’ suggestion

mentioned earlier, it is necessary for learners to be able to see an overview of the

structure of the lessons so that they can make informed choices about activities

PhD 158 D.Hoven

appropriate to their needs. Similarly, it is important for the instructions to indicate

clearly the choices and options available. In addition, the feedback mechanisms within

the package need to be task- and choice-specific and to give learners access to

information about the right and wrong answers and the context of and background to

these.

3.4.4 Learning Styles

Much work has been done in the area of learning or cognitive styles and strategies in

both general learning and second language learning. Work by Hartnett (1975 – cited in

Larsen-Freeman & Long, 1991: 210) has also shown that cognitive style may, in fact,

be related to hemispheric specialisation as discussed above. Most of this research has

been driven by the same motivating force that lies behind research in the other areas of

individual difference discussed so far: to determine what level of interaction can be

found between modes of instruction and learning or cognitive style. Similarly, the

impetus behind the work on learning strategies has been the extent to which educating

learners in the use of more effective strategies enables them to improve their learning.

As a prelude to this latter area of research, it has first been necessary to identify and,

usually, classify, such strategies. Subsequently, having identified and classified the

desirable strategies, it is proposed that learners of lower aptitude, or those who do not

normally employ such strategies, can then be educated in their efficacy and use.

The distinction between styles and strategies has only been articulated since studies

were begun on the differing performances of students on inventories of ‘approaches to

studying’ (Ramsden and Entwistle, 1981) in study, school or learning contexts (Marton

and Säljö, 1976; Svensson, 1977; Laurillard, 1979), rather than in the previous ‘natural’

contexts (Biggs, 1976; 1978; Entwistle et al., 1979). Questions on the existence of

cognitive or learning styles, their nature, and their capacity for being influenced

through teaching or training, have been the subject of inquiry since the early work of
Witkin (1954; 1967), Witkin et al. (1967; 1977) and Goodenough (1976; 1981) on

PhD 159 D.Hoven

field-dependence/independence, one of the most frequently researched bipolar cognitive

styles, discussed in detail below.

It is important that learning or cognitive styles, which refer to ‘the student’s preferred

way of tackling learning tasks in general’ be distinguished from learning strategies

which, as will be discussed in detail later, ‘concern the way a student elects to tackle a

particular learning task’ (Watkins, 1982: 78). It is only since the mid eighties that this

distinction has begun to clarify, with the work of Oxford on language learning strategy

identification and inventories (1986; 1990), Willing (1985) on learning style

identification, Dickinson (1987), Holec (1987), and Wenden (1991) on strategies useful

for self-instruction and autonomy, and O’Malley and Chamot (1990) on the relationship

between learning style, strategies, and classroom instruction.

Early studies of learning and cognitive style began to show that there were similar,

identifiable, dichotomous dimensions to cognitive style, in both the study and natural

contexts, but that another dimension ‘characterized by a highly organized approach to

study and high achievement motivation’ (Watkins, 1982: 77) also seemed to be

operating. Thus, for example, in a qualitative study of student descriptions of how they

approach learning tasks, Marton and Säljö (1976) were able to classify the responses

generally according to whether the students were employing surface or deep levels of

processing. These levels were characterised by a mechanical approach, focussing on

parts rather than the whole in surface processing; and a holistic, structure-oriented

approach in deep processing. However, another dimension relating to achievement

orientation and attitude also emerged as an influencing factor in this and other studies

examining students’ approaches to learning. Aspects of achievement orientation and

attitude will be discussed in detail in section 3.5.

As illustrated by the variety and nature of terms used by researchers into learning
styles, a lot of confusion resulted from early analyses of learner approaches to study.

PhD 160 D.Hoven

This confusion derived from problems of identification, measurement, and defining the

nature of these styles: whether they were innate and fixed, or whether they were task-

or context-dependent. Many of these problems arose because researchers had not yet

realised the differences and interactions between ‘perceptual’ and application or

‘strategic’ processing, discussed in the following sections, which had not yet been

adequately described.

The dichotomous ‘surface <–> deep’ dimensions mentioned above have now been

clarified further, and redefined as cognitive or learning styles, while an additional

dimension fits the description of what we now call learning strategies, which will be

discussed in detail in the next chapter. As will become clear by the next chapter, even

though the recognition and accommodation of learning styles and strategies is now

fairly widely accepted as playing a crucial role in the teaching and learning process, we

are still not certain of their exact nature and the interaction between them. First,

however, let us examine the critical features of cognitive and learning styles as

identified in various research studies.

3.4.4.1 Perception vs application

As will be discussed in detail below, early research in the area of cognitive or learning

styles was focussed on modes of perception and differences in style among individuals

for making meaning from input gathered by the senses, particularly sight. Later studies

attempted to extend the implications of these perceptual studies into an application or

implementation phase, by examining students’ approaches to learning tasks (Brumby,

1982; Lam, 1982). In these studies, the basic premises were that the differences in

cognitive style would affect individuals’ performances and/or modes of processing or

working through learning tasks. It was at this stage that clear discrepancies began to

emerge as to whether learning styles are stable and static, or variable and dynamic

(Brumby, 1982), as well as whether they are conscious or unconscious (Bialystok,

1985).

PhD 161 D.Hoven

The implication of the latter concern is that, if conscious, they can be trainable or

changeable. The emergence of these discrepancies has led to the distinction being made

between learning styles, which have come to be regarded as mostly unconscious and

innate in individuals for particular contexts (see also Larsen-Freeman & Long, 1991),

and learning strategies, which are, or can become, conscious, and which therefore

could be trainable. As Bialystok (1985: 256) states: ‘Consciousness itself is not a

criterion in the identification of strategies of learning. [....] [but] those strategies which

are under conscious control are of course easier to identify, to discuss, to manipulate

and to teach.’

Schmidt (1990: 133), on the other hand, argues that consciousness and intentionality

need to be separated, as ‘we often become aware of things we do not intend to notice’.

While denying the possibility of subliminal perception from his data, Schmidt makes a

strong claim for the possibility of incidental learning taking place while the focus of

learning is on other, relevant and salient points. His necessary criterion in this process is

understanding of the material encountered. Therefore, while accepting Bialystok’s point

about consciousness being critical to the teaching of strategies, Schmidt’s refinements

to the phenomenon of consciousness allow the possibility, for example, of new learning

strategies being acquired by learners when the focus of material presented is listening

comprehension tasks incorporating such strategies. This is, in fact, the paradigm

underlying the simultaneous presentation of learning strategies and listening and

viewing comprehension tasks in the CELL package described in Chapters 5 and 6.

3.4.4.2 Perceptual studies

Initially, Witkin’s conception of cognitive style was as a bi-polar dichotomy for

perception. These poles, which he termed field-dependence (FD) and field

independence (FI), were identified as tendencies only, though quite stable over time.
Females were generally represented slightly more highly at the dependent end, possibly

PhD 162 D.Hoven

because of socialisation factors (Witkin, et al. 1977). Characteristics of a global, FD

cognitive style included perception being ‘strongly dominated by the surrounding

field’, a low capacity for segregating the self from the non-self, greater use of social

frames of reference which influence attitudes and feelings, an orientation to be ‘with

people’, and closer personal space. The articulated, FI cognitive style, on the other

hand, was characterised by a sense of separate identity, more impersonal orientation,

not being ‘sensitive to social undercurrents’, and individualism.

More recently, in the SLA field, FI/D has been defined as ‘a cognitive style, a bipolar,

stable trait affecting how one thinks, feels, and behaves’. Field Independent people are

characterised as being ‘analytic, confident, and self-reliant’, whereas Field Dependent

people are holistic, uncertain, and dependent upon others’ (Chapelle & Green, 1992).

Chapelle and Green go on to note that it is important to distinguish cognitive style from

ability because, among other reasons, ‘a cognitive style is a bipolar trait, indicating

equal value for each pole’ (1992: 79).

As can be seen in Figure 3.2 below, other researchers in the area of cognitive style have

used various terms to describe FD, including global (Witkin, 1977), holistic (Pask and

Scott, 1972), intuitive (Bruner, 1960), intuitive/random (Scarcella & Oxford, 1992),

and divergent production (Guilford, 1967). These are contrasted with articulated style

(Witkin, 1977), serialist (Pask & Scott, 1972), analytic (Bruner, 1960),

sensory/sequential (Scarcella & Oxford, 1992), and convergent production (Guilford,

1967) respectively. One aspect on which all researchers seem to agree with regard to

cognitive style is that extremes of the dichotomy are equivalent in learning potential. In

other words, individuals can achieve equally on intellectual tasks, regardless of which

cognitive style they exhibit, though their means of achieving the same end product may

differ.

PhD 163 D.Hoven

 Figure 3.2 Cognitive Style Dimensions

Witkin (1967) Field Dependent Field Independent

Bruner (1960) intuitive analytic

Guildford (1967) divergent production convergent production

Pask & Scott (1972) holistic serialist

Witkin et al. (1977) global articulated

Scarcella & Oxford (1992) intuitive/random sensory/sequential

However, some researchers in the SLA field (Felder & Henriques, 1995) believe that

this equality can be compromised when mismatches occur between the learning styles

of members of a class and the teaching style of the teachers, where the teacher’s

cognitive style predisposes him or her towards presenting, teaching, illustrating, and

testing in certain ways. Other researchers, while recognising the need for awareness on

the part of teachers of their learners’ preferred styles, emphasise the importance also of

challenging their learners to experiment with and acquire new styles or extensions of

their preferred styles (Smith & Renzulli, 1984; Reid, 1987; Larsen-Freeman & Long,

1991).

Although there has been some agreement on the existence of bipolar cognitive styles,

other researchers are less convinced of the distinctiveness of field

dependence/independence as a construct (Vernon, 1972; Griffiths & Sheen, 1992).

Vernon, for example, began by questioning the effectiveness of available instruments to

measure the construct. On the basis of the variety of tests used to measure FD/I, Vernon

PhD 164 D.Hoven

(1972: 386), though ultimately positive, cautioned that ‘it is doubtful whether different

versions of the commonly used [FD/I] tests measure the same thing, and whether they

do not also involve general intelligence’, with only the Rod-and-Frame test1 possibly

showing the involvement of a distinctive visuo-kinesthetic factor. This comment was

made in response to the conflicting findings of researchers attempting to link Locus of

Control (Rotter, 1966), male-female role identification (Sherman, 1967; Vaught, 1965),

cultural or ecological differences (Berry, 1966; Vernon, 1969), and personality

differences, to measures of FD/I.

Along similar lines, Griffiths and Sheen (1992: 145) make the claim, on the basis of ‘so

few substantive findings’ and the lack of consistency in measurement instruments, that

FD/I ‘is a wasteland, bereft of meaningless hypotheses for L2 researchers’. In addition,

Reid (1987) has questioned the cross-cultural validity of the construct on the grounds

that, in her study of ESL learners, differing educational backgrounds produced a

different range of learning style preferences. However, she uses these findings to

advocate greater variety of teaching and learning styles in classrooms. Finally, in

response to Griffiths and Sheen, and after a comprehensive overview of the nature of

the construct and the instruments for measuring it, Chapelle concludes that:

life is both too short and too complex to ignore concepts such as cognitive style which
express some of our intuitions about students and which facilitate appreciation for the
divergent approaches to thinking and learning. [...] But, personally, I believe that the
most important and relevant human constructs are those which are neither interesting to
‘authorities’ nor measurable at present.

(Chapelle, 1992: 381)

1The apparatus used here was Oltman's (1968) portable daylight test as described in Vernon
(1972: 369).

PhD 165 D.Hoven

The general conclusion to be drawn from this and other findings discussed above

therefore seems to be that the original construct of Witkin et al. (1977) should be

approached circumspectly, and that careful consideration should be given to the

applicability of various methods postulated to test FD/I. Furthermore, most researchers

have acknowledged that there do seem to be certain characteristics among learners

which conform to the FD/I distinction, or some similar form of these characteristics.

3.4.4.3 Application studies

In spite of the equivocation over the exact nature of the FD/I construct, numerous

researchers have applied the construct in various studies. Some of the most prominent

include examinations of male-female differences in cognitive style related to second

language achievement (Hansen, 1984), cognitive style nurtured by socio-culturally-

determined education systems (Hansen, 1984; Berry and Dasen, 1974), advantages for

one cognitive style rather than the other on certain kinds of second language tests

(Readance et al., 1980; Stansfield and Hansen, 1983), and possible impacts of

mismatch between cognitive styles of teachers and learners (Frank and Davis, 1982;

Hansen and Stansfield, 1982). This last issue of cognitive style mismatch is of most

interest for us in the context of providing self-access CELL materials. Particularly in a

self-access context, in order for learners to feel comfortable using the materials and to

be able to progress at their own pace, it is important that the organisation of the

materials and the navigation through them correspond to their expectations.

Hansen and Stansfield (1982: 263), for example, discuss the finding that ‘field

independence, as opposed to field dependence, has been shown to play a helpful though

minor role in the development of second language proficiency in a formal

environment’. However, from the perspective of cognitive style affecting teaching

effectiveness, ‘little evidence exists that either style, in and of itself, produces better all-

round teaching or learning’ (Hansen and Stansfield, 1982: 264). While urging further
investigation of the causal factors in their study, Hansen and Stansfield found that the

PhD 166 D.Hoven

most significant advantage emerged for FI females working with an FI instructor, with

only minor advantages for FD males working with an FD instructor. Alternatively, in

their study, Frank and Davis (1982) found that teachers with FI preferences produced

much better performances among both FI and FD learners than did teachers with FD

preferences. These researchers hypothesised that FD learners were more aware of the

stylistic differences between them and their (FI) teachers, and therefore attempted to

model those characteristics to compensate for the absence of these characteristics in

themselves.

Three major conclusions to emerge from the studies mentioned above are that analytic,

inferencing tasks of the kind represented by cloze tests favour FI-style learners, that

these kinds of tests are therefore culturally biased against members of some societies

which are more FD-oriented, and that teachers should be more aware of the variety of

their own learning style preferences and of those of their learners. As Witkin and

Goodenough (1981) found, people in agrarian and authoritarian societies are generally

more inclined towards FD than FI. Hansen (1984) also found that, of the five Pacific

Island cultures he studied, learners from Hawaii, the least agrarian society, showed the

highest incidence of FI. One additional point made by Readance et al. (1980) was that

although FD-style learners experience more difficulty than FI-style learners in

analysing information to solve a problem, they can be trained to use an analytical,

hypothesis-testing approach in appropriate situations. These findings fit very

comfortably with learning strategy training, and principles of learner-centred and

communicative language teaching which favour greater focus on learners as

individuals. Thus, materials reflecting a range of different cognitive styles will be

beneficial for both types of learners, by providing them with opportunities both to work

within their own stylistic comfort zones, and to extend themselves in new ways.

In an early investigation of the interaction between cognitive style and learning

strategies, in a learning environment involving computers, Jamieson and Chapelle

PhD 167 D.Hoven

(1987) used computerised spelling and dictation lessons. The learners in this study were

from diverse L1 backgrounds learning ESL. Jamieson and Chapelle, in common with

Poulisse (1989-90, found that strategy use was task-specific in that more strategies were

evident for the more difficult task of dictation. Only one strategy, monitoring input,

was used more by lower proficiency learners, when they listened repeatedly and

carefully to the dictation passage. As regards cognitive style, the findings showed that

FI learners used less advance preparation and more output monitoring (self-reflection)

than did FD learners, while another individual difference factor,

impulsivity/reflectivity, showed reflective learners employing advance preparation and

output monitoring more than impulsive learners.

Jamieson and Chapelle collected their computer data records on learners’ working

styles and learning strategies, using a record-keeping program running in the

background to their computerised spelling and dictation lessons. Advance preparation

was inferred from the length of time a learner took to respond to a question, output

monitoring was inferred from the number of times answers were edited, while the

number of times learners chose to have the audio segment repeated was used to infer

input monitoring. However, these findings should not be extended to apply to all tasks

involving computers, and should be considered with caution from two viewpoints.

Firstly, the tasks used for this study were very restrictive and computer-managed. A

different model of instructional design such as the one advocated in Chapters five and

six would produce different data. Secondly, as discussed by Pusack and Otto (1995),

computer records such as key strokes and time between strokes provide much

extraneous data that is very uninformative. In addition, such records do not have the

capacity to record more useful information such as the reasons for students’ repeated

listening to a particular audio segment, or the source of their difficulty. Information on

these reasons could only be collected using the kinds of prompt questions and learner

PhD 168 D.Hoven

self-report protocols employed, for example, by Liou (1995), in her examination of

learning strategies discussed in detail in the next chapter.

3.4.4.4 Development of the concept of strategies

When other researchers (Pask, 1976; Brumby, 1982) attempted to expand the

application of cognitive style dimensions from merely perception to the actual

implementation of tasks and more generally to learning and memory (Goodenough,

1976), they found that other factors were operating, though it was not until later that

these other factors were identified as strategies which could possibly be trained. Though

he distinguished between styles and strategies, and identified that underlying his

‘holistic’ and ‘serialist’ strategies lay distinct learning styles, comprehension and

operation learning respectively2, Pask (1976) failed to specify elements of his strategies
in such a way that they could be classified. For Pask, for example, the difference

between learning styles and strategies was that styles were more generally exhibited,

while strategies are exhibited in strictly constrained circumstances in the form of his

‘conversational’ systems (Pask, 1976: 133.).

Schmidt (1990) examines the relationship between language processing and learning

from the perspective of consciousness. While assuming that both conscious and

unconscious processing takes place in language learning, Schmidt separates

2 Marton and Säljö (1976), following Craik and Lockhart (1972), make the distinction between
'surface' and 'deep' processing in memory and learning. In this approach to depth of processing,
perception is built up through a progression or series of processing stages, where increased
semantic or cognitive analysis is referred to as having greater depth. Greater depth of
processing, in turn, produces a stronger memory trace. For Marton and Säljö, the level of
processing brought into operation depends on students’ perception of the demands of
assessment: ‘the type of test expected affects the kind of processing that subjects engage in’
(1976: 115). Pask, on the other hand, claims that the requirement for understanding is ‘as
strong as, or stronger than, the requirement for deep-level processing’ (1976: 133). He goes on
to hypothesise that in classroom or self-study contexts, the requirement for understanding is
relaxed, allowing learners to be more flexible in their approach, choosing to act like ‘holists’
(comprehension learners) or ‘serialists’ (operation learners), or varying their approach,
depending on the subject matter.

PhD 169 D.Hoven

consciousness into three different types: awareness, intention, and knowledge. This

separation helps to clarify some of the dilemmas faced by second language researchers.

Within consciousness as awareness, in turn, three crucial levels can be distinguished:

perception, noticing (focal awareness), and understanding. From an earlier study

(Schmidt & Frota, 1986) Schmidt concludes that, while there is a close association

between noticing and emergence in production, the evidence is inconclusive that

noticing is a sufficient condition for learning (Schmidt, 1990: 141).

It is Schmidt’s contention that it is at the understanding level of consciousness that

problem solving, and all types of metacognition, belong. This becomes a crucial

distinction for the discussion in the next chapter on whether, and to what extent,

learning strategies can be taught. As much of the research in this area shows, it is the

metacognitive and problem solving strategies that lend themselves to successful

instruction, and these strategies are also implicated in successful language learning.

In his discussion of consciousness as intention, Schmidt provides a useful corollary to

Bialystock’s (1984) argument against consciousness as a criterion for strategic action,

by clarifying the ambiguity between passive awareness and active intent. For Schmidt,

as for us, ‘intentions may be either conscious or unconscious [...] and we may become

aware of things we do not intend to notice’ (1990: 133). Thus, learners may act

strategically, with unconscious intent, or only become aware of their strategy use when

it has noticeable positive or negative effects. In other words, ‘unconscious learning [...]

may be seen as an unintended by-product of communicative interaction’ (Schmidt,

1990: 135).

Later, Brumby (1982), as a result of her research, saw the dichotomies of cognitive

style as falling into two groups: those relating to the perception of a problem; and those

involved in the process of integration with existing knowledge. While dichotomies in

the former group, as mentioned earlier, were assumed to be equivalent in effectiveness,

PhD 170 D.Hoven

those in the latter group were perceived as being hierarchic, with one pole being

distinctively more effective than the other. In this latter group, Brumby (1982: 245)

lists Wertheimer’s (1945) ‘rote-memorising’ versus ‘productive thinking’, Ausubel’s

(1968) ‘rote-learning’ versus ‘meaningful-learning’, Marton’s (1975) ‘surface- and

deep-processing’ and Pask’s (1976) ‘operation learning’ versus ‘comprehension

learning’. Thus, as illustrated in Figure 3.3 below, characteristics in the right-hand

column are hierarchically ‘higher’ in that they are regarded as producing ‘better’

learning outcomes than those in the left-hand column.

Figure 3.3 Integration-based cognitive styles (Brumby’s Group 2)

rote-memorising (Wertheimer, 1945) productive thinking (Wertheimer, 1945)

rote-learning (Ausubel, 1968) meaningful-learning (Ausubel, 1968)

surface-processing (Marton, 1975) deep-processing (Marton, 1975)

operation learning (Pask, 1976) comprehension learning (Pask, 1976)

Brumby, however, could not overcome her dilemma of whether cognitive styles were

innate or learnt. In the light of the findings by Readance et al. (1980) that certain

‘styles’ can be trained, and considering that by ‘styles’ they were referring to what we

would now call strategies, it can be determined that while styles may be innate (though

subject to variations as described earlier), at least some strategies can be trained. Based

on the research of Willing (1989: 139 - 146), it may in fact be the case that there are

learning strategies that can be categorised as, and that are used for, becoming

consciously aware of one’s preferred learning styles in various language learning

PhD 171 D.Hoven

contexts, which in itself is an application of metacognitive strategies (Oxford, 1990:21),

as outlined in the next chapter.

Willing (1985, 1988) has identified four major learning styles as being useful for ESL

teachers of migrants to help them develop sensitivity towards their students’ individual

needs and learning styles. He classified these styles as: communicative, authority-

oriented, concrete, and analytical, where a person may change styles, but generally

tends more toward one of these styles than another (see Figure 3.4 below). Inherent in

an identification of teachers’ tendencies towards any of these styles is the expectation

that they will then be able to modify consciously their behaviour towards students and

choose tasks for their classes that cater for a range of learning styles in addition to their

own preferred styles.

Figure 3.4 Willing’s learning style classifications and their interaction

Communicative Authority-oriented

Concrete Analytical

Other researchers have focussed on sensory preferences such as visual, auditory, kinesic

and tactile, particularly as regards cultural norms, values, and expectations (Wesche,

1981; Soudek & Soudek, 1985; Reid, 1987). Four major points which emerge from

Reid’s study of cross-cultural learning style preferences among ESL learners are that:

1) learners learn best when their styles match those of their teachers, 2) learning style

preferences show strong cultural determinants, 3) because learners differ significantly

in their preferences of sensory mode, so that these need to be catered for, and 4) the
optimal approach is to provide learners with a range of activities incorporating different

learning styles. Meanwhile, Soudek and Soudek (1985: 113) conclude that ‘much of

PhD 172 D.Hoven

non-verbal communication is culture-specific’ and that ‘by accumulating and sharing

knowledge of the non-verbal dimensions of communicative competence, we endeavor

to help our students not only switch from language to language, but also to switch

kinesic and proxemic patterns’.

Felder and Henriques (1995) also support the need for multiplicity of exposure to

different learning styles, having first identified what those styles might be in any given

classroom. As they show:

matching teaching styles to learning styles can significantly enhance academic

achievement, student attitudes, and student behavior at the primary and secondary
school level (Griggs & Dunn 1984; Smith & Renzulli 1984), at the college level
(Brown 1978; Charkins et al. 1985), and specifically in foreign language instruction
(Oxford et al. 1991; Wallace & Oxford 1992). This is not to say that the best thing one
can do for one’s students is to use their preferred modes of instruction exclusively.
Students will inevitably be called upon to deal with problems and challenges that
require the use of their less preferred modes, and so should regularly be given practice
in the use of those modes.
(Felder & Henriques, 1995: 28)

For Gremmo and Riley (1995), it is important that we avoid categorising learners, but

instead concentrate on using the research findings on learning styles to help us set up a

range of appropriate learning tasks, or a learner-friendly learning environment. In this

way, we can ‘help learners come to terms with their strengths and weaknesses, to learn a

language efficiently in ways which are compatible with their personalities’ (1995: 158).

Similarly, Ehrman and Oxford (1995: 70) advocate learners ‘using strategies that are

not necessarily directly related to their preferred styles’ and that ‘indeed, to learn

effectively, every student must be able to do this some of the time’. This also reflects

PhD 173 D.Hoven

Kumaravadivelu’s (1993) insistence on the need to design learning environments that

are learning-driven in order to cater for learner diversity.

This, then, is the approach taken in the software package discussed in Chapters 5 and 6:

a range of tasks and media are provided, and these tasks are designed to incorporate a

diversity of learning styles as well as give learners practice in numerous learning

strategies, some of which may not come naturally to them. Returning to the question of

the distinction between ‘styles’ and ‘strategies’, it may be the case that styles, which

emerge at the perceptual level, are innate and unchangeable preferences, but that

learners can be trained to become aware of, though not prefer, different styles. Other

‘styles’ emerge at the application level and can be varied by any individual learner,

according to the tasks and context. As identified above, these styles are better identified

as ‘strategies’. From a teacher’s or material designer’s point of view, instructional

materials need to be designed to cater for differences in perceptual style. For application

styles or ‘strategies’, however, learners can be trained, or made aware of, the range of

more effective strategies that can be used when dealing with tasks of varying cognitive

or contextual demands (Oxford, 1993). Lawrence (1984) and O’Malley and Chamot

(1993: 109) also make the claim for ‘a tendency to use certain learning tools (learning

strategies) and to avoid others’, as one interpretation of the term learning style.

As Scarcella and Oxford comment:

when left to their own devices, and if not overly pressured by their environment to use a
certain set of strategies, students typically use learning strategies that reflect their basic
learning style (Ehrman & Oxford 1989). They can, however, learn to develop additional
learning strategies and test the value of ones they already use.

(Scarcella & Oxford, 1992: 63–4)

PhD 174 D.Hoven

Which strategies are trainable and to what extent will be discussed in the next chapter

on learning strategies. Scarcella and Oxford’s comment above raises critical questions

of how additional learning strategies can be developed, how learners can be brought to

want to develop or change strategies they already use, which strategies can and should

be developed, and whether, or to what extent, this can be achieved.

More specifically, as some of the researchers mentioned above comment, motivation,

orientation, and attitudes towards and beliefs about language learning, or the learning of

the language of a specific culture, are also critical in the use of learning styles and

strategies for successful language learning. In fact, according to Gremmo and Riley ‘it

is important to keep in mind the fact that any form of self-directed learning scheme is

an intervention in the social knowledge system, that is, in the set of structures and

functions through which a society manages the creation, organization, distribution,

legitimization and storage of knowledge’ (1995: 161). The next section will deal with

some aspects of these factors in relation to their susceptibility to change or training.

3.5 Characteristics more amenable to change

Having reviewed some of the characteristics of learners that are less amenable to

change, we now turn our discussion towards more changeable characteristics.

Prominent among these are learners’ attitudes towards each other, and towards the

target language and target language culture, as well as their motivations to learn. For

our purposes in designing a learner-centred CELL instructional design model, these

changeable characteristics are equally important, if not more so, than those that are less

amenable to change. Where there is the possibility for learners to make changes within

themselves, and for the software program to enhance this process, the functions of the

instructional design model change from that of providing flexibility and variety, to

more awareness-raising and pedagogic functions.

PhD 175 D.Hoven

3.5.1 Attitude and motivation
Motivation is generally regarded by teachers, researchers in SLA, and learners as a

necessary, if not sufficient, condition of language learning. As Scarcella and Oxford

point out: ‘regardless of whether we believe motivation to be the single primary factor,

one of several equally important factors, or an intermediary factor, it is impossible to

dismiss the importance of motivation in language learning’ (1992: 52). Crookes and

Schmidt (1989) define motivation as comprising two equally important factors: external

or behavioural, and internal or attitudinal factors. The behavioural features include:

decision-making (to participate in an activity or not), persistence (both over time and

after interruptions), and activity level. Four factors characterise the internal perspective

on motivation, namely: interest (based on learners’ attitudes, experience, and

background knowledge); relevance (to learners’ personal needs such as achievement);

expectancy (of success or otherwise); and outcomes (rewards perceived by learners,

which may be intrinsic or extrinsic).

Both motivation and attitude are important to the learning process, whether before,

during, or after formal instruction (Scarcella & Oxford, 1992; Skehan, 1989). In turn,

there are a variety of factors which have been shown to influence motivation, though

questions as to the direction of this influence are frequently raised. Thus, while studies

have included parental and teacher attitudes (Feenstra, 1969; Schumann and Schumann,

1977; Buttjes, 1991; Keller, 1991), politics (Byram, 1991), and materials orientation

(Risager, 1991), it is still unclear whether these factors increase or decrease motivation,

or, alternatively, whether it is the level of motivation which affects the factors studied.

Skehan (1989: 49) hypothesises a distinction between Intrinsic motivation, which stems

from an inherent interest in a given task, and Resultative motivation, which is

influenced by the success or otherwise achieved on a task, or in a subject area. Results

from the latter area of research have been relatively inconclusive, with surprisingly low
correlations found between motivational levels and success. One explanation for these

PhD 176 D.Hoven

varying results may actually be the range of contexts and types of instrumentally

motivating factors investigated. Many of these studies have also been criticised for their

lack of rigour in method and design (Skehan, 1989: 64–7). Such studies have, for

example, examined factors as diverse as dominance of interaction (Strong, 1984),

working with Spanish kindergarten children, to Hermann’s (1980) comparison of the

attitudes of first and fifth year students towards the target culture (German).

Gardner (1980) has postulated various models to account for influencing factors in

motivation, including instrumental versus integrative orientation, and motivation being

a sum of three factors: effort, desire to achieve a goal, and attitudes (Gardner, 1985: 9–

10). For Gardner, integrative motivation, the desire to achieve integration into the target

culture through learning the language, was preferred over instrumental motivation, or

learning the language to achieve a goal more effectively. However, he has since

(1985:169) modified this preference in favour of a more general advocacy of the

importance of motivation being aroused, regardless of its source. More recently,

Gillette (1987) has shown that Rubin and Thompson’s (1982) argument for pragmatism

(i.e. instrumental motivation) may be well-founded, particularly, as is the case of

learning Indonesian in Australia, where the second language is being learnt in a foreign

language context, that is, the language being learnt is not the language of the majority

population.

In a comprehensive overview of the relationship between autonomy and motivation,

Dickinson (1995 166) finds evidence that ‘learning autonomy increases motivation and

consequently increases learning effectiveness’, both from theories of extrinsic/intrinsic

motivation (Deci & Ryan, 1985) and attribution theory. In addition, he cites the claims

of the Carnegie Project (deCharms, 1984) that ‘motivation can be enhanced through

encouraging learners to exert personal control over their learning and to take

responsibility for it’ (Dickinson, 1995: 166). As has been shown earlier, learners can
achieve this through an understanding of the values of such an approach, an awareness

PhD 177 D.Hoven

of techniques available to them to take advantage of self-direction and autonomy, and a

supportive environment.

However, the issue of orientation (positive or negative) and attitude towards the target

language and the target language culture remains a strong influencing factor in extrinsic

or instrumental motivation, and to a lesser degree in intrinsic or integrative motivation.

The latter, however, is more difficult to define, with researchers differing in their

attributions of the same factors. Attitude and orientation are the most extensively

researched aspects of this problematic area, though Peirce (1995) has strong arguments

for a redefinition of the terms under which these aspects are investigated, to include the

multiple social identities of the individual. Peirce (1995: 11–12) argues that ‘although

muted, there is an uneasy recognition by some [SLA] theorists that current theory about

the relationship between the language learner and the social world is problematic’.

She then extends this statement to cite the ambiguities found in various studies of

anxiety (Scovel, 1978), and personality variables and language achievement (Gardner

and McIntyre, 1993), as deriving from the fundamental misapprehension that ‘language

learners can choose under what conditions they will interact with members of the target

language community’ (Peirce, 1995: 12), and that this choice is learner-motivated.

However, for Peirce, many of these ambiguities can be addressed when the learner’s

multiple social identities, across time and different sites, are taken into account. Before

taking full consideration of Peirce’s theory, and its conjunction with the sociocultural

model proposed earlier in this work, we need to look at the other major theories

purporting to account for, or in part explain, the findings in studies of attitude,

motivation, and orientation.

3.5.2 Acculturation and Accommodation Theories

Considerable research has been carried out in the areas of learners’ identification with,
attitudes towards, and perceptions of, the values and beliefs of the target culture and the

PhD 178 D.Hoven

extent to which these factors influence learners’ achievement and level of proficiency in

the language of this culture. Thompson (1991), for example, in a study on

pronunciation and the Critical Period Hypothesis, suggested Weinreich’s (1953)

‘interlingual’ identification as a possible explanation for the fact that some learners just

do not want to sound like native speakers of the target language. She also proposed

learners’ desire to maintain their L1 as another possible cause. Two major hypotheses

have emerged to articulate an understanding of these findings of Thompson and others:

Speech Accommodation Theory (Giles & Byrne, 1982; Beebe, 1988) and Schumann’s

Acculturation Model (1978, 1986).

Speech Accommodation Theory has developed to help account for the degree to which

different individuals are able to integrate with the target culture, by ‘explaining the

interaction between speakers in terms of their feelings, values, and motives’ (Beebe,

1988: 62). According to this theory, broadly speaking, the degree of learners’

identification with the people of the target culture is in direct proportion to the degree

of accommodation to the target language made in learners’ speech. As learners become

more linguistically proficient, that is, closer to the speech of the target culture, the more

they are accepted by or integrated with that culture. This in turn produces the desired

social and communicative advantages, as, being accepted, they come to be regarded as

part of the in-group (Tajfel, 1974), thereby benefiting from more positive attitudes on

the part of native speakers of the target language.

Similarly, Schumann’s Acculturation Model (1978) focusses on identification with, and

the relative desirability of, the target culture and language (Schumann uses the terms

social and psychological distance) as determining factors in the degree of language

proficiency acquired. Social distance is used to refer to an individual’s membership of

one social group, and the degree of contact between this group and another, the

members of which speak another language. Psychological distance, on the other hand,
involves affective factors more central to the individual such as culture shock, different

PhD 179 D.Hoven

kinds of motivation as discussed above, and ego permeability as studied by Guiora and

his colleagues (1972a).

As McLaughlin puts it: ‘[...] social and psychological distance influence second-

language acquisition by determining the amount of contact learners have with the target

language and the degree to which they are open to the input that is available’

(1987:111). Thus, when the distance is perceived to be great, the learner may miss out

on the necessary native-speaker input, leading to fossilisation in language development,

and therefore the learner’s interlanguage (Selinker, 1971) develops no further in the

direction of the target language. However, the validity of this argument is questioned as

to the direction of causality between attitude and motivation. It remains unclear, and,

indeed, may vary with individuals, as to whether a positive attitude motivates a learner

to become more proficient, or whether proficiency produces a positive attitude – or

indeed, whether the two factors are mutually supportive. This remains a contentious

issue, while also being difficult to test.

For the purposes of the present work, and the determination of the influences of

motivation on success in language learning in a CELL environment, this problem with

the direction of causality diminishes the value and applicability of Schumann’s model

for us. Namely, if it cannot be determined whether motivation or proficiency is the

causal factor, no effective action can reasonably be taken to improve one or the other.

In other words, there is insufficient evidence to support taking action to increase

proficiency by trying to build more positive motivation in learners. In the meantime, as

advocated by theorists in the field such as Oxford (1994), Gardner and Tremblay

(1994), MacIntyre (1995; 1995a), and Sparks and Ganschow (1995), we must await

further empirical research in the area to test the various aspects of theories currently

being propagated.

PhD 180 D.Hoven

Nevertheless, social/psychological studies have identified several trends in the

relationship between various social and psychological factors and language learning.

Macnamara (1973) and Genesee and Hamayan (1980), for example, found evidence

that younger learners are less prone to fixed attitudes which affect their language

learning than older learners. From his studies of Anglophone students learning French

in bilingual contexts in Canada, Gardner (1980) suggests that attitude seems more

critical in second, as against foreign, language learning contexts, where much more

contact is possible between learners and speakers of the target language. In a contrastive

study of learners in a foreign language context, Dörnyei (1990) finds that for such

learners, beliefs about, and attitudes towards, the target culture are less specific, and

integrative motivation (see section 3.5.1) is much lower than for learners in a L2

context. He also finds that integrative motivation may be necessary for learners to

progress in their language development beyond an intermediate level.

Both these trends are supported by Oxford (1994a), Dörnyei (1994), and Oxford and

Shearin (1994). Oxford and Shearin also explore the implications of motivational

factors from other discipline areas, including Need Theories (Maslow, 1970),

Instrumentality Theories (Atkinson, 1964), Equity Theories (Pritchard, 1969),

Reinforcement Theories Landy (1985), and Social Cognition Theories (Locke &

Latham, 1990). As a result of their analysis, they suggest a number of ways in which

teachers can increase their learners’ motivation, including identifying their motivational

forces, encouraging a positive classroom environment, and informing learners of a

range of (instrumental) motivational factors that may correspond to the needs and

aspirations of different learners. As suggested by these and other researchers in this

field, these are fertile areas for future empirical and ethnographic research. For the

purposes of our CELL instructional design model, these issues of motivation are

addressed through the provision for learner control over the materials. As will be

discussed in detail in the following chapter, learner motivation to learn has been shown
to increase when learners have control over the paths their learning takes (Nakhoul,

PhD 181 D.Hoven

1993), and, when learners are aware of how to use this control (Stevens, 1984;

Robinson, 1991).

To sum up, in the words of Scarcella and Oxford ‘[...] the arguments about instrumental

vs. integrative motivation that have dominated the research literature might be less

important than the absolute degree of motivation possessed by the individual learner’

(1992: 53). As illustrated by Oxford and Nyikos (1989) mentioned above, and Genesee,

Rogers and Holobow (1983), motivation can often be the catalyst needed to progress in

one’s language learning, by stimulating the search for, and acquisition of, more

effective learning strategies. Although some individual difference factors may be

necessary conditions, others may merely constitute ‘typical’ or ‘graded’ conditions

(Ellis, 1994).

3.6 Conclusion
To reframe this discussion of individual differences within a sociocultural framework,

it is useful to return to the contentions of Peirce (1995), that individuals differ, not only

from one another, but also within themselves from one point in time to another, and

from one context to another. Having reviewed the literature on individual differences

from perspectives of age, maturation, hemispheric differentiation, personality, aptitude,

learning styles, and motivation, we see that social factors, as much as psychological or

neurological factors influence a learner’s capacity for, and success in language learning.

Three critical elements emerge from this wide-ranging discussion. Firstly, that learners

can be assisted in compensating for characteristics less appropriate to successful

language learning by helping them develop stronger characteristics in other areas.

Secondly, that learners need to move out of their individual ‘comfort zones’ in order to

participate productively and effectively in the learning process, and to carry the

learning beyond the immediate task to novel situations. Thirdly, that learners need
information, support, and the infrastructure to negotiate this development – that they

PhD 182 D.Hoven

need not only to be given control, but also to be provided with the means by which to

take control on their own terms.

Because of the indeterminacy pervading most of the findings in these areas, we cannot,

therefore, afford to be overly specific or restrictive in the way we conceptualise learner

differences, or in the approaches from which we design instructional or learning

materials purporting to recognise these differences. It is therefore necessary, in a

Computer-Enhanced Language Learning package, to cater for this range of diverse

influences in order that language learning can be enhanced for as many learners as

possible. In a multimedia CELL environment, this can be achieved through the

provision of materials based on a variety of media and content, and utilising a range of

task responses which favour different learning styles and sensory preferences.

As mentioned earlier, the mobilisation of learning strategies, particularly metacognitive

strategies, and more particularly strategies appropriate to individual learning style

preferences, plays a significant role in helping learners towards language learning

success. In the following chapter the conceptualisation, existence, identification, and

teachability of language learning strategies will be examined in the light of research and

theory. When we combine this with our understanding of individual differences,

listening and viewing comprehension theories, and the concept of learner- and learning-

centred instruction, we will then be able to construct the framework for the multimedia

CELL package described in subsequent chapters.

PhD 183 D.Hoven