Dörnyei, Z. (2010). Researching motivation: From integrativeness to the ideal L2 self. In S.

Hunston &
D. Oakey (Eds.), Introducing applied linguistics: Concepts and skills (pp. 74-83). London: Routledge.

CHAPTER/

Researching motivation: from integrativeness

to the ideal L2 self
Zoltdn Dornyei

CONTENTS Language teachers frequently use the term 'motivation' when they describe successful
The starting point: or unsuccessful learners. This reflects our intuitive belief that during the lengthy and
'Integrativeness'as a
motivational
often tedious process of mastering a foreign/second language (L2), the learner's
factor 74 enthusiasm, commitment and persistence are key determinants of success or failure.
Towards the'L2 Indeed, in the vast majority of cases, learners with sufficient motivation can achieve a
Motivational Self
System' 75
working knowledge of an L2, regardless o/their language aptitude, whereas without suf-
Validating the L2
ficient motivation even the brightest learners are unlikely to persist long enough to
Motivational Self attain any really useful language ('you can lead a horse to water, but you can't make it
System 80
drink').
Conclusion 82
Because of the central importance attached to it by practitioners and researchers
References 82
alike, L2 motivation has been the target of a great deal of research in Applied Linguistics
during the past decades. In this chapter I describe a major theoretical shift that has
recently been transforming the landscape of motivation research: the move from the
traditional conceptualization of motivation in terms of an integrative/instrumental
dichotomy to the recent conceptualization of motivation as being part of the learner's
self system, with the motivation to learn an L2 being closely associated with the learner's
'ideal L2 self. For space limitations I cannot provide a detailed review of the relevant
literature (for recent summaries, see Dornyei 2005; Dornyei and Ushioda 2009); instead,
my focus will be on illustrating how such a major paradigm shift has emerged through
a combination of theoretical considerations and empirical research findings.

The starting There has been a long-lived (and inaccurate) understanding in the L2 profession that
point: language learning motivation can be divided into two main dimensions: integrative
Mntegrativeness' motivation and instrumental motivation. The former refers to the desire to learn an L2
as a of a valued community so that one can communicate with members of the community
motivational and sometimes even to become like them. Instrumental motivation, on the other hand,
factor is related to the concrete benefits that language proficiency might bring about (e.g.
career opportunities, increased salary). Thus, broadly speaking, it was thought that we
learn a language either because we like it and its speakers or because we think it will be
useful for us. . : .
The integrative/instrumental distinction has been attributed (again somewhat inac-
curately) to the influential work of Canadian social psychologist Robert Gardner (1985,
2001), who did indeed introduce these terms but whose theoretical motivation
construct was much more elaborate than this simplistic duality. Furthermore, Gardner
hardly ever discussed the nature and impact of instrumental motivation, because he was
 Researching motivation 75

almost exclusively interested in the interpersonal/emotional aspect of motivation that

. . ' he termed 'integrativeness'. He characterized this motivational dimension as follows:

Integrativeness reflects a genuine interest in learning the second language in order to come
closer to the other language community. At one level, this implies an openness to, and respect
: '- ' for other cultural groups and ways of life. In the extreme, this might involve complete iden-
-; tification with the community (and possibly even withdrawal from one's original group),
but more commonly it might well involve integration within both communities.
(Gardner 2001:5)

The concept of integrativeness/integrative motivation has become a popular and much

•' researched concept in L2 research, but starting in the 1990s an increasing number of
scholars began to raise issues about how generalizable the term was. In a multicultural
setting such as Montreal, where Gardner first developed his theory, it made sense to talk
about potential 'integration', but in learning situations where a foreign language is taught
only as a school subject without any direct contact with its speakers (e.g. teaching English
or French in Hungary, China, Japan or other typical 'foreign language learning' contexts),
•' : . the 'integrative' metaphor simply did not make sense. In such environments what exactly
would be - to quote Gardner (2001) - 'the other language community1 that the learner
would want to 'get closer to'? In many language learning situations, and especially with the
learning of world languages such as English or French, it is not at all clear who 'owns' the
L2, and this lack of a specific L2 community undermines Gardner's theoretical concept
of integrativeness. This view has been shared by several scholars worldwide (e.g. Coetzee-
Van Rooy 2006; Lamb 2004; Yashima 2000; for a review, see Dornyei 2005), and, as a result,
. .'.-. - over the past decade 1 have been trying to find a broader interpretation of the notion that
goes beyond the literal meaning of the verb 'integrate' but which also builds on the relevant
knowledge and considerable body of research that we have accumulated in the past.

Towards the In 2005,1 proposed a new motivation construct (Dornyei 2005) - the 'L2 Motivational
*L2 Self System' - that builds upon the foundations laid by Gardner (1985) but which at the
Motivational same time broadens the scope of the theory to make it applicable in diverse language
Self System* learning environments in our globalized world. The proposed model, which attempts
to synthesize a number of influential new approaches in the field (e.g. Ushioda 2001;
, Noels 2003), has grown out of a combination of empirical research findings and the-
oretical considerations (for a detailed description, see Dornyei, 2009). Let us look at
these more closely, starting with the former.

Empirical findings pointing to the need to reinterpret Integrativeness

Over the past 15 years I have been heading a research team in Hungary with the
objective of carrying out a longitudinal survey amongst teenage language learners by
administering an attitude/motivation questionnaire at regular intervals so that we can
gauge the changes in the population's international orientation. So far three successive
waves of data collections have been completed (in 1993, 1999 and 2004) involving
over 13,000 learners (for a detailed summary, see Dornyei et al. 2006). The survey
76 Zoltan Dornyei

questionnaire targeted attitudes towards five target languages: English, German,

French, Italian and Russian. It was originally developed in collaboration with one of
Robert Gardner's closest associates, Richard Clement, and therefore integrativeness
and instrumentality had a prominent place in it, but we also measured several other
attitudinal/motivational dimensions, such as Direct contact with L2 speakers (i.e.
attitudes towards actually meeting L2 speakers and travelling to their country),
Cultural interest (i.e. the appreciation of cultural products associated with the par-
ticular L2 and conveyed by the media; e.g. films, TV programmes, magazines and pop
music), Vitality ofL2 community (i.e. the perceived importance and wealth of the L2
communities in question), Milieu (i.e. the general perception of the importance of
foreign languages in the learners' school context and in friends' and parents' views)
and finally Linguistic self-confidence (i.e. a confident, anxiety-free belief that the
mastery of an L2 is well within the learner's means).
In an analysis of the first two waves of the survey (Dornyei and Csizer 2002), we
computed correlations of the various motivation components with a criterion measure,
Language choice, which referred to the degree of the learners' desire to learn a particular
L2 in the next school year. Correlation is a conceptually straightforward statistical
procedure: it allows us to look at two variables and evaluate the strength and direction
of their relationship or association with each other. To do so, we compute a 'correlation
coefficient' between the two variables, which can range from -1 to +1, with a high
correlation indicating a positive relationship, zero correlation no relationship, and a

English/UK English/US German French Italian Russian

1993 1999 1993 1999 1993 1999 1993 1999 1993 1999 1993 1999

Integrativeness .43* .33* .43* .33* .47* .43* .42* .44* .43* .43* .25* .32*

Instrumentality .28* .25* .28* .25* .30* .30* .27* .30* .29* .31* .20* .21*

Attitudes towards L2 ,23* .16* .17* .16* .33* .30* .31* .33* .32* .31* .12* .21*
speakers/community
Vitality of the .12* .09* .12* .09* .11* .12* .13* .16* .16* .18* .07* .10*
community

Cultural interest .14* .09* .12* .10* .20* .17* .20* .21* .26* .23* .12* .17*

Milieu .12* .12* .12* .12* .01 -.00 .03 .04 .01 -.00 -.05* -.10*

Linguistic .07* .06* .07* .06* -.00 .01 .03 - .02 -.01 -.02 -.02 -.04
self-confidence

Multiple
correlations .44* .34* .44* .34* .49* .45* .44* .46* .45* .45* .27* .34*

p < .001

Table 7. i Correlations between the attitudinal/motivational scales and Language choice in the Dornyei and
: Csizer (2002) study
 Researching motivation 77

negative correlation an inverse relationship (Dornyei 2007). Thus, for example, learners'
IQ is expected to have a high positive correlation with their mathematics grades and
zero correlation with, say, the love of chocolate. Table 7.1 presents the results.
As can be seen in Table 7.1, three variables stand out consistently across the
languages and the data points: Integrativeness, Instrumentality and Attitudes towards L2
speakers/community. This was, actually, to be expected given our previous under-
standing of L2 attitudes and motivation, but what surprised us was that when we
computed multiple correlations (i.e. correlations between language choice and all the
motivational variables together), the joint correlation was hardly higher than the cor-
relation associated only with Integrativeness. For example, the correlation of the choice
of English (UK) in 1993 was .43 with Integrativeness and .44 with all the altitudinal
variables together. This suggested that Integrativeness played a principal role in
determining the extent of a learner's overall motivational disposition.
To test the prominent position of Integrativeness, Dornyei et al. (2006) submitted
the data from all the three waves of the survey to a more complex statistical procedure,
structural equation modelling (SEM). SEM is very useful to interpret the relationship
among several variables within a single framework. Its strength is that we can specify
directional paths (i.e. cause-effect relationships) amongst the variables and SEM then
produces various goodness-of-fit indices to evaluate the feasibility of the whole model.
In conducting the analysis, we took each language and each year separately (so we
computed separate models for, say, German in 1993 and French in 2004), but the
various models converged and with minor variations produced the same overall result.
Figure 7.1 presents the schematic representation of the final construct.

Figure 7.1
Schematic
representation
of the structural
equation model
in Dornyei et al.'s
(2006) study

Attitudes toward
L2 speakers
/8 Zoltan Dornyei

The final model that emerged from our study, presented in Figure 7.1, confirms our
earlier observation based on correlation analysis that Integrativeness plays a key role in
L2 motivation, mediating the effects of all the other attitudinal/motivational variables
on the criterion measure Language choice (and we obtained exactly the same results with
another criterion measure, Intended effort to study the L2). Curiously, the immediate
antecedents of this latent variable were Attitudes towards L2 speakers/community and
Instrumentality, thus, the three variables that the correlations in Table 7.1 highlighted
emerged as the central motivational components in the SEM model as well, and the
model also gave us an indication about how these variables related to each other and
to the criterion measure. What is more, this was a very consistent finding because it
applied to all the different target languages and all the three waves of our survey. The
only problem was that what we found did not make much theoretical sense:
'Integrativeness' turned out to be the principal motivation factor in an environment
where 'integrating' was not very meaningful (since there was nothing really to integrate
into) and, furthermore, integrativeness was closely associated with two very different
variables: faceless pragmatic incentives and personal attitudes towards members of the
L2 community. It was clear that we needed a new theory to accommodate these
findings.

Theoretical considerations
Parallel to conducting the empirical research outlined above, I became familiar with an
intriguing new theoretical approach in psychology that looked particularly promising
with regard to applying it to L2 motivation: the conceptualization of possible selves. First
introduced by Markus and Nurius (1986), the concept of the possible self represents
an individual's ideas of what they might become, what they would like to become and
what they are afraid ofbecoming. That is, possible selves are specific representations of
one's self in future states, involving thoughts, images and senses, and are in many ways
the manifestations of one's goals and aspiration. From a motivational point of view, two
types of possible selves - the ideal self and the ought self - seemed particularly relevant
(Higgins 1987). The former refers to the representation of the attributes that someone
would ideally like to possess (i.e. representation of hopes, aspirations or wishes),
whereas the latter refers to the attributes that one believes one ought to possess (i.e. a
representation of someone's sense of duty, obligations or responsibilities) and which
therefore may bear little resemblance to desires or wishes. The motivational aspect of
these self-guides was explained by Higgins's (1987, 1998) self-discrepancy theory, pos-
tulating that motivation involves the desire for people to reduce the discrepancy
between their actual and ideal/ought selves.
This self framework not only made intuitive sense to me but it also seemed to
offer a good explanation of our Hungarian findings. Looking at 'integrativeness' from
the self perspective, the concept can be conceived of as the L2-specific facet of one's
ideal self. If our ideal self is associated with the mastery of an L2, that is, if the person
that we would like to become is proficient in the L2, we can be described in Gardner's
(1985) terminology as having an integrative disposition. Thus, the central theme of
the emerging new theory was the equation of the motivational dimension that has
 Researching motivation 79

traditionally been interpreted as 'integrativeness/integrative motivation' with the Ideal

L2Self. . - - . - .
Looking back at Figure 7.1, please recall that our Hungarian data showed that the
immediate antecedents of the Ideal L2 Self in the model were Attitudes towards the L2
speakers/community and Instrumentality. Does this make sense? Yes it does: with regard
to Attitudes towards the L2 speakers/community, we must realize that the actual L2
speakers are the closest parallels to a person's idealized L2-speaking self, which suggests
that the more positive our disposition towards these L2 speakers, the more attractive
our idealized L2 self. Or, to turn this equation around, it is difficult to imagine that we
can have a vivid Ideal L2 Self if the L2 is spoken by a community that we despise. With
regard to Instrumentality, because the ideal language self is a cognitive representation
of all the incentives associated with L2 mastery, it is naturally also linked to professional
competence that often requires the knowledge of the L2. Thus, to put it broadly, in our
idealized image of ourselves we may want to be not only personally agreeable but also
professionally successful.
We should note here, however, that from a self perspective the term 'instrumentality'
can be divided into two distinct types. In conceptualizing the ideal/ought self dis-
tinction, Higgins (1987, 1998) highlighted a crucial difference between the two
dimensions, a contrasting approach/avoid tendency: ideal self-guides have a promotion
focus, concerned with hopes, aspirations, advancements, growth and accomplishments
(i.e. approaching a desired end-state); whereas ought-to self-guides have a prevention
focus, regulating the absence or presence of negative outcomes, concerned with safety,
responsibilities and obligations (i.e. avoidance of a feared end-state). With this dis-
tinction in mind, we can see that traditionally conceived 'instrumentality/instrumental
motivation' mixes up these two aspects. When our idealized image is associated with
being professionally successful, 'instrumental1 motives with a promotion focus - for
example, to learn English for the sake of professional/career advancement - are related
to the Ideal L2 Self. In contrast, instrumental motives with a prevention focus - for
example, to study in order not to fail an exam or not to disappoint one's parents - are
part of the Ought-to L2 Self. . . . . : ,

The *L2 Motivational Self System*

As we have seen above, both the empirical findings and the theoretical considerations
seemed to support a reconceptualization of L2 motivation as part of the learner's self
system. I have come to believe that the two elements discussed before, the ideal and the
ought selves, are central components of this system, but I also felt that we needed to add
a third major component, which concerns the direct impact of the learning envi-
ronment (and which will not be discussed in this chapter in detail; for reviews, see
Dornyei 2001; Manolopoulou-Sergi 2004). Accordingly, I proposed (Dornyei 2005) that
the 'L2 Motivational Self System' was made up of the following three components (for
more details on the self approach, see Dornyei 2009; Dornyei and Ushioda 2009):

Ideal L2 Self, which is the L2-specific facet of one's 'ideal self. If the person we would
like to become speaks an L2, the 'ideal L2 self is a powerful motivator to learn the L2.
8o Zoltan Dornyei

Ought-to L2 Self, which concerns the attributes that one believes one ought to possess
to meet expectations and to avoid possible negative outcomes.
L2 Learning Experience, which concerns situated motives related to the immediate
learning environment and experience (e.g. the impact of the teacher, the curriculum,
the peer group, the experience of success).

Validating the Over the last two years my research students - Stephen Ryan, Tatsuya Taguchi and
L2 Michael Magid - and I have been conducting large-scale survey research in Japan and
Motivational China to validate the L2 Motivational Self System (for details of the surveys, see Ryan
Self System 2009; Taguchi et al. 2009). We believed that if we can support the main tenets of the
theory by data coming from foreign language contexts that are very different from the
Hungarian learning environment that the L2 self approach originated from, this would
be a powerful validity argument of the construct. The studies did indeed confirm that
our assumptions were correct; in the following I present some of the key findings: (a)
correlations between traditional Integrativeness and the Ideal L2 Self to check whether
the two constructs can indeed be equated; (b) correlations of Integrativeness and the
Ideal L2 Se/fwith criterion measures to see which variable does a better job at explaining
motivated behaviour; (c) correlations between aspects of Instrumentality and the
Ought-to 12 Self to check whether traditional instrumentality can indeed be divided
into two distinct types.

The relationship between the Ideal L2 Self and Integrativeness

Table 7.2 presents correlations between Integrativeness and the Ideal 12 Self, in three
different East Asian surveys involving over 5,300 participants, each consisting of
different subsamples. As can be seen, there are substantial positive correlations in every
subsample, suggesting that the two variables do indeed tap into the same underlying
construct domain. The average correlation across the subsamples is .54, which is very
high in L2 motivation research. These consistent results leave no doubt that the two
concepts are very closely related.

„', 1 . t " '• ' , , : ' ' " '

University University
': **.*" *"," ", ~i "". " -- '•"' • • -"- : ", "*< - . • students students Secondary
'•"•'•.• 11* *'•'*'•'"•*"'-•. Total (non-English (English school Adult
"'•"" T/,*. sample majors) majors) pupils learners

Correlations 2,397 Japanese learners .59 .54 .53 .61

between (Ryan forthcoming)

1,586 Japanese learners .63 .63 .48

(Taguchi et al. forthcoming)

1,328 Chinese learners .51 .46 .46 .66 .53

(Taguchi et al. forthcoming)

Note: All figures are significant at the p < .001 level.

 Researching motivation 81

The correlation of the Ideal L2 Self and Integrativeness

with Intended effort
Table 7.3 presents the correlations of Integrativeness and the Ideal L2 Self with Intended
effort in the same samples as in Table 7.2. These figures allow us to compare which
variable does a better job at explaining the criterion measure, effort. Although
Integrativeness does a consistently good job at accounting for the variance in the
criterion measure, the Ideal L2 Self exceeds it in all but one subsample. The average
variance (which is the average of the squared correlation coefficients) explained by
Integrativeness across the different samples is 32 per cent, which can be considered high,
but the same figure for the Ideal L2 Self is 42 per cent, which is almost 30 per cent
higher!

University University
students students Secondary
Total (non-English (English school Adult
sample majors) majors) pupils learners

2,397 Japanese learners Ideal L2 Self .77 .74 ~7^

- r- ' 1
"7C
. /o
<
, M
• , • . -•&
, I-
(Ryan forthcoming) Integrativeness .65 .61 .54 .71

1 ,586 Japanese learners Ideal L2 Self .71 .71 .61 ' ' • • ;• ' . ', -j
(Taguchi et al. forthcoming) Integrativeness .63 .64 .49 ' - • • •;• - •:;

1 ,328 Chinese learners Ideal L2 Self .55 .52 .51 .69 .51
(Taguchi et al. forthcoming) Integrativeness .52 .47 .53 .63 .44

Note: All figures are significant at the p < .001 level.

Table 7.3 Correlations of Integrativeness and the ideal L2 Se//with Intended effort

The case of Instrumentality

Table 7.4 presents correlations that allow us to examine whether Instrumentality can
indeed be divided into two types as outlined above. The data are drawn from Taguchi
et al.'s (2009) Japanese and Chinese samples where the promotion and the prevention
aspects of Instrumentality were measured separately. If Higgins's (1987, 1998)
promotion/prevention distinction applies to our data, then we would expect to find
higher correlations of the Ideal L2 Self with Instrumentality-promotion than with
Instrumentality-prevention-, and the Ought-to L2 Self should display the opposite
pattern. This is exactly the case in Table 7.4. Furthermore, if the promotion and the pre-
vention aspects are separate from each other then we would not expect a high
correlation between them, and indeed both correlations are modest (with even the
higher one explaining less than 12 per cent of the variance). Thus, these figures provide
unambiguous confirmation that the traditionally conceived 'instrumental motivation'
can indeed be divided into two distinct types: one relating to the Ideal L2 Self, the other
82 Zoltan Dornyei

to the Ought-to L2 Self. These two types are only moderately related to each other anc
show a distinct correlation pattern with the two self dimensions.

Table 7.4 Ideal L2 Self Ought-to L2 Self Instrumentality -

Correlations : " promotion
between the
Japan China Japan China Japan China
ideal L2 Self,
Instrumentality Instrumentality- .63* .46* .28* .45* - -
and the Ought- promotion
To L.2Self'm
Instrumentality- -.01 -.13* .53* .63* .34* .26*
Taguchietal. prevention
(forthcoming)
Japanese and * p < .001
Chinese samples

Conclusion This chapter discussed a major theoretical shift that has been taking place within the
field of L2 motivation research. I described how a new paradigm has emerged from
both theoretical considerations and research results, and then presented the main com-
ponents of the newly proposed 'L2 Motivational Self System'. In the second part of the
chapter I provided empirical data from three different surveys involving over 5,300 par-
ticipants to validate the new construct. The correlational results clearly indicated that:
(a) Integmtiveness and the Ideal L2 Self lap into the same construct, but the Ideal L2 Self
does a better job at explaining variance in the criterion measures; (b) the traditionally
conceived concept of Instrumentality mixes up two types of pragmatic motives (with
a promotion vs. a prevention focus) that show a rather different relationship pattern
with the Ideal and the Ought-to L2 Selves. These results are all in accordance with the
proposed theory and thus provide a strong validity argument for it. We should reiterate
here that in the current study the third main component of the 'L2 Motivational Self
System1, the L2 Learning Experience, was not measured.

References Coetzee-Van Rooy, S. (2006) 'Integrativeness: Untenable for world Englishes learners?' World
Englishes, 25: 437-50.
Dornyei, Z. (2001) Teaching and researching motivation, Harlow: Longman.
Dornyei, Z. (2005) The psychology of the language learner: Individual differences in second
language acquisition, Mahwah, NJ: Lawrence Erlbaum.
Dornyei, Z. (2007) Research methods in applied linguistics: Quantitative, qualitative and
mixed methodologies, Oxford: Oxford University Press.
Dornyei, Z. (2009) 'The L2 Motivational Self System', in Z. Dornyei and E. Ushioda (eds),
Motivation, language identity and the 12 self, Clevedon: Multilingual Matters.
Dornyei, Z. and Csizer, K. (2002) 'Some dynamics of language attitudes and motivation:
Results of a longitudinal nationwide survey', Applied Linguistics, 23: 421-62.
Dornyei, Z. and Ushioda, E. (eds) (2009) Motivation, language identity and the L2 self,
Clevedon: Multilingual Matters.
 Researching motivation 83

Dornyei, Z., Csizer, K. and Nemeth, N. (2006) Motivation, language attitudes and global-
isation: A Hungarian perspective, Clevedon: Multilingual Matters.
Gardner, R.C. (1985) Social psychology and second language learning: The role of attitudes and
motivation, London: Edward Arnold.
Gardner, R.C. (2001) 'Integrative motivation and second language acquisition', in Z. Dornyei
and R. Schmidt (eds), Motivation and second language acquisition, Honolulu, HI:
University of Hawai'i Press.
Higgins, E.T. (1987) 'Self-discrepancy: A theory relating self and affect', Psychological Review,
94:319-40.
Higgins, E.T. (1998) 'Promotion and prevention: Regulatory focus as a motivational
principle1, Advances in Experimental Social Psychology, 30: 1-46.
Lamb, M. (2004) 'Integrative motivation in a globalizing world', System, 32: 3-19.
Manolopoulou-Sergi, E. (2004) 'Motivation within the information processing model of
foreign language learning', System, 32:427-41. . • .,•'",=.
Markus,H. andNurius, P. (1986) 'Possible selves' American Psychologist, 41: 954-69.
Noels, K.A. (2003) 'Learning Spanish as a second language: Learners' orientations and per-
ceptions of their teachers' communication style' in Z. Dornyei (ed.), Attitudes,
orientations and motivations in language learning, Oxford: Blackwell.
Ryan, S. (2009) 'Self and identity in L2 motivation in Japan: The ideal L2 self and Japanese
learners of English', in Z. Dornyei and E. Ushioda (eds), Motivation, language identity
and the L2 self, Clevedon: Multilingual Matters.
Taguchi, T, Magid, M. and Papi, M. (2009) 'The L2 motivational self system amongst
Chinese, Japanese and Iranian learners of English: A comparative study', in Z. Dornyei
and E. Ushioda (eds), Motivation, language identity and the L2 self, Clevedon:
Multilingual Matters.
Ushioda, E. (2001) 'Language learning at university: Exploring the role of motivational
thinking', in Z. Dornyei and R. Schmidt (eds), Motivation and second language
acquisition, Honolulu, HI: University of Hawaii Press.
Yashima, T. (2000) 'Orientations and motivations in foreign language learning: A study of
Japanese college students', JACETBulletin, 31: 121-33.