8 Perception, Attitude and Motivation: Jean-Marc Dewaele
8 Perception, Attitude and Motivation: Jean-Marc Dewaele
8
Perception, Attitude and Motivation
Jean-Marc Dewaele
8.1 Introduction
In this contribution I will focus on one of the most fascinating and messy areas of
applied linguistic inquiry, namely the perceptions that people have of their own
language(s) and foreign languages, the attitudes that they develop towards differ-
ent languages and their motivation to acquire foreign languages. The ‘messiness’
arises from the fact that this area of research is highly interdisciplinary: it has
attracted the interest of ethnographers, psychologists from different areas, second
language acquisition (SLA) researchers and second language (SL) teachers,
sociolinguists, psycholinguists, conversation analysts, etc. These researchers have
applied their own distinctive epistemologies and methodologies or have created
their own unique blends by importing concepts and approaches of neighbouring
disciplines. It is therefore appropriate to start this introduction with a quick
presentation of the two major epistemological approaches available to SLA
researchers interested in perception, attitude and motivation, and the methodo-
logical implications of the choice of a particular epistemological approach.
The first decision SLA researchers in this area of research have to make is where
to situate their analysis on the etic–emic dimension. This epistemological opposi-
tion was introduced by Pike (1964) in ethnographical research. He extended
the phonetic/phonemic distinction in linguistic meaning to cultural meaning.
He pointed out that researchers’ interpretive (etic) frameworks differ from cultur-
ally specific (emic) frameworks used by members of a society/culture for inter-
preting and assigning meaning to experiences.
Etic analyses and interpretations are based on the use of carefully defined and
relatively stable concepts from the analytic language of the social sciences (Pike,
1964). This makes them useful for comparative research across languages, situa-
tions and cultures. Emic analyses, on the other hand, incorporate the participants’
perspectives and interpretations of behaviour, events and situations using the
descriptive language of participants (Pike, 1964). A debate has been raging in the
SLA literature on these ontological and epistemological questions (Gregg, 2006;
Firth and Wagner, 2007). Current research in SLA varies widely in terms of prefer-
ence of emic and etic perspectives, or combinations of both.
164 Jean-Marc Dewaele
Defending the emic perspective, Firth and Wagner (2007) complain that
methodologies and theories within SLA reflect ‘an imbalance between cognitive
and mentalistic orientations, and social and contextual orientations to language,
the former orientation being unquestionably in the ascendancy’. They plead for
an enhanced awareness of the contextual and interactional dimensions of lan-
guage use and an increased emic sensitivity towards fundamental concepts. It is
not entirely clear whether Firth and Wagner (2007) want the balance to shift
completely. Previous defences of emic perspectives in SLA, like Watson-Gegeo
(1988), stated that emic and etic analyses were complimentary: ‘a carefully done
emic analysis precedes and forms the basis for etic extensions that allow for cross-
cultural or cross-setting comparisons’ (Watson-Gegeo, 1988: 581–582).
A stronger focus on the emic perspective is characteristic of the Conversation
Analysis (CA) paradigm and the postmodernist study of L2 socialization. CA
researchers are interested in an ‘emic reality’ (Ten Have, 2007: 37). In order to
uncover ‘the procedural infrastructure of situated action’ (ibid.: 37) they carry out
an inductive search for patterns of interaction in episodes of naturally occurring
interactions. No attempt is made to generalize findings, as researchers are commit-
ted to ‘elucidate the local logic’ (ibid.: 199). This approach has been successfully
applied to SLA (Mondada and Pekarek, 2004). Postmodernists also insist on the
uniqueness of participants, hence their resistance to generalize findings (see below
for more on this topic).
Most attitudinal research in SLA privileges a more etic perspective: opinions
and attitudes of participants are heard but these are collected through research
instruments designed and formulated by the researchers. The first aim in this type
of research is to gain an insight in relationships between variables in particular
settings, with the ultimate aim of comparing these patterns with studies carried out
in different settings. While these researchers acknowledge the uniqueness of
L2 learners, their aim is to explain variation in the data through statistical analysis
using carefully defined categories: ‘there are probably as many factors that might
account for individual differences in achievement in a L2 as there are individuals.
However, they may be grouped into one of the two classifications of cognitive or
affective variables’ (Gardner and MacIntyre, 1992: 212).
Psychologists such as Manstead and Fisher have talked about the mutual exclusivity
of approaches and the ‘incommensurability of findings generated by standardized
and essentially quantitative methods, on the one hand, and idiosyncratic and essen-
tially qualitative research methods, on the other’ (Manstead and Fisher, 2002: 3).
It is therefore not surprising that SLA research on perception, attitude and
motivation is also the prime battleground between groups who could stereotypi-
cally be labelled ‘relativists’ and the ‘positivists’ (see Block (1996) and Gregg
(2006) for both sides of the debate). Spolsky (2000) has pointed out that in SLA
motivation is most often seen in positivist and utilitarian terms. Indeed, the pri-
mary goal of motivation research in SLA has been to maximize the effectiveness of
L2 education through manipulation of personal attributes or contextual variables.
Kinginger (2004) concurs with this view and regrets that among American foreign
language teachers motivation is solely understood in positivist, utilitarian terms.
For them, motivation exists in some stable and unitary form and they need to
‘find, build, nurture, or match it’ in order to increase the effectiveness of the
teaching (ibid.: 161). The one move that these positivists do normally not envisage
is to question the very existence of the construct, which she perceives to be an
‘overstuffed can of worms’ (ibid:161).
Leppänen and Kalaja (2002) also criticize what they perceive as a positivist bias
in SLA motivation research and defend a social approach instead:
varieties and practices become imbued with value or devalued in the linguistic
marketplace’ (ibid.: 284).
Pavlenko therefore rejects the ‘reductionist, static and homogeneous view
of culture’ (ibid.: 280). She also criticizes the lack of explanatory validity of quanti-
tative sociopsychological approaches concerning the social causes of particular
attitudes, motivations and beliefs (ibid.: 280). She argues that constructs such as
‘identity’, ‘in-group membership’, ‘self-identification’ or ‘accommodation’ cannot
be considered to be explanatory because they are themselves in need of explana-
tion (ibid.: 280).
Pavlenko’s third objection relates to the ‘causal, unidirectional and stable
nature attributed to such constructs as motivation, attitudes, or social distance’
(ibid.: 280). She points to the continuous shaping and reshaping of motivation
and social contexts, where initial success may strengthen the individual’s determi-
nation in learning the target language while a series of disappointing results may
sap that individual’s learning motivation (ibid.: 280).
Pavlenko’s fourth point concerns the artificiality of the separation assumed in
sociopsychological approaches between social factors and the individual, or
psychological factors. She argues that many individual factors, such as age, gender
or ethnicity, are also socially constituted. As a consequence, the understanding
and implications of age, gender or ethnicity vary across communities and cultures.
Moreover, ‘attitudes, motivation or language learning beliefs have clear social
origins and are shaped and reshaped by the contexts in which the learners find
themselves’ (ibid.: 280–281). She points out that the ‘understanding and implica-
tions of being Jewish or Arab, young or old, female or male are not the same across
communities and cultures’ (ibid.: 281).
The methodology used in quantitative sociopsychological studies constitutes
Pavlenko’s (2002) fifth target for criticism. The validity of questionnaires is in
doubt, she argues, because it is not clear ‘what exactly was measured by the multi-
ple questionnaires that attempted to quantify language attitudes, motivation,
acculturation or language proficiency, in particular, when the latter was reduced to
self-evaluation’ (ibid.: 281).
Pavlenko’s sixth point concerns the fact that most research within the socio-
psychological paradigm has been carried out in English-speaking environments in
Europe and North America, raising the issue of generalizability of the findings.
She points out that a very different picture may emerge in different environment
(ibid.: 281).
Pavlenko’s seventh criticism concerns ‘the idealised and decontextualised
nature attributed to language learning, which is presented as an individual
endeavour, prompted by motivation and positive attitudes, and hindered by nega-
tive attitudes and perceptions’ (ibid.: 281). As an alternative, she proposes the
poststructuralist approaches which provide ‘a more context-sensitive way of theo-
rising social impact on L2 learning and use’ (ibid.: 295). This allows researchers to
examine complex situations in which L2 users are legitimate speakers who move
between different contexts or create rich, hybrid identities. Poststructuralists recast
language attitudes and language learning beliefs as ideologies, hence ‘illuminat-
ing the socially constructed nature of beliefs previously seen as individual’
168 Jean-Marc Dewaele
8.3.1 Definitions
The data Gardner obtained from participants through the Attitude Motivation
Test Battery (AMTB) (for more on its development, see Gardner and Smythe,
1981) are quantitative, namely a value on a 5-point Likert scale accompanying
a list of statements relating to possible reasons why the participants want to learn a
L2 (‘Studying the L2 can be important for me because it will allow me to travel to
L2 areas; . . . it will allow me to learn about myself; . . . it will help me find a better
job; . . . it will allow me to appreciate L2 minority problems’). These items are
linked to 11 scales (integrative orientation, attitudes towards the target group,
interest in foreign languages, teacher evaluation, course evaluation, motivational
intensity, desire to learn the language, attitudes towards learning the language,
language class anxiety, language use anxiety and instrumental orientation. These
scales form the basis of five constructs, namely integrativeness, attitudes towards
the learning situation, motivation, language anxiety and instrumentality (Gardner,
2006: 246). These dimensions emerge from standard use of factor analyses of
individual items in the AMTB. The researcher at that point has to interpret and
name the dimensions. Gardner named these dimensions as he did, and though
they have a logical basis, they could only emerge because of the choice of the par-
ticular items included in the questionnaire. In other words, these dimensions did
not exist a priori. This approach is common in psychology. For example, factor
analyses of items in personality questionnaires typically propose solutions based on
five super-dimensions (Furnham and Heaven, 1999). It is not entirely clear to me
whether the inclusion of different types of items related to affective aspects of
language learning and language perception might not cluster on a new dimen-
sion. Statistical analyses allow the calculation of the score of each participant on
the different dimensions, and in comparing that single score to group means the
researcher can provide a unique numerical profile of that person’s attitudes and
motivation, or of that group of individuals.
170 Jean-Marc Dewaele
Communication
Layer I L2
Behaviour
Use
2
Layer II Behavioural Intention
Willingness to
Communicate
3 4
Layer III Desire to State Situated Antecedents
communicate with Communicative
a specific person Self-Confidence
5 6 7
8 9 10
Layer V Affective-Cognitive Context
Intergroup Social Communicative
Attitudes Situation Competence
11 12
Figure 8.1 The Pyramid Model of WTC (MacIntyre et al, 1998), ‘Conceptualising willing-
ness to communicate in a L2: A situational model of L2 confidence and affiliation’, The
Modern Language Journal, 82, 545–562
172 Jean-Marc Dewaele
Motivation is clearly linked to individual and societal factors. The basis of the
pyramid (Layer VI) refers to intergroup climate and personality. MacIntyre (2007)
explains that this layer exists
even before the individual is born, because they capture enduring intergroup
and genetic influences handed down from one generation to the next. The
individual has little influence over these factors and generally they play a some-
what indirect role in language behaviour. (ibid.: 567)
The next layer (V) of the pyramid includes intergroup attitudes, the social situation
and communicative competence. It thus refers to the individual’s typical affective
and cognitive context. MacIntyre (2007: 567) explains that this layer sets ‘the tone
for motivation to learn the L2’, namely ‘the tension between a desire to approach
the target language group and a sense of hesitation or fear of the implications of
doing so’.
The next level (Layer IV) ‘Motivational Propensities’ is the last to refer to endur-
ing influences: ‘Intergroup motives stem directly from membership in a particular
social group and interpersonal motives stem from the social roles one plays within
the group. Issues of affiliation and control (broadly defined) are the most basic of
motives, exerting their effects throughout the system. Roles and motives combine
with L2 self-confidence; perceptions of communicative competence coupled with
a lack of anxiety’ (ibid.: 568).
Layer III of the pyramid refers to situational influences: ‘The sense of time is
coming to focus on the here-and-now. At this level of the pyramid model is the
desire to communicate with a specific person as well as a state of self-confidence’
(ibid.: 568).
The next layer (II) refers to the behavioural intention to speak if one has the
opportunity, or to remain quiet, in other words, an individual’s willingness ‘to initi-
ate second language discourse on a specific occasion with a specific person, remem-
bering that the person represents a different social group’ (ibid.: 568). Layer II,
willingness to communicate ‘represents the final psychological step in preparation
for L2 communication’ (ibid.: 568), while the top of the pyramid (I) represents
the actual communication in the L2.
What is clear from the pyramid model is that the social dimension plays a crucial
part in determining an individual’s motivation to learn an L2 and to use that L2 at
some point. The authors do not consider motivation and attitudes as idealized and
decontextualized, on the contrary, they are grounded in the context of enduring
influences from specific affective, historical, social, political and geographical
factors.
While these attitudes can be relatively stable within speech communities, they
can suddenly shift as a consequence of political circumstances (war), or migration
patterns, when a new language suddenly emerges in the local linguistic landscape
and becomes associated with a particular style of music or activity.
Perception, Attitude and Motivation 173
Attitudes towards languages may also vary widely within a particular country.
Dörnyei and Clément (2001) looked specifically at the attitudes of thousands of
13- and 14-year-old Hungarian schoolchildren towards five different languages –
English, German, French, Italian and Russian. Except for marked gender differ-
ences (with girls scoring higher on most attitudinal/motivational measures), the
researchers found strong regional variation: schoolchildren in the capital scored
highest on most variables, those living near the Austrian border scored higher on
German, those in the east scored higher on French and Russian and Italian was
most popular in Budapest. The authors conclude that ‘macrocontextual, geopolit-
ical factors significantly affect people’s language attitudes’ (ibid.: 423). In a further
study Dörnyei, Cziser and Németh (2006) analysed the effect of sociopolitical
changes at the end of the 1980s in Eastern Europe on attitudes towards foreign
languages. The researchers investigated whether the strong increase of intercul-
tural contact in Hungary had affected the attitudes of Hungarian schoolchildren
towards five different languages – English, German, French, Italian and Russian.
They found that the rank order between the different languages had remained
unchanged after the fall of the Iron Curtain: English was the most popular lan-
guage, followed by German, followed by French and Italian at similar levels, with
Russian at the bottom of the table (ibid.: 143). Gender differences, which had
been strong in earlier studies (English, German and Russian being preferred by
males, French and Italian being preferred by females) were weaker and had disap-
peared for English in 2004. French and Italian were found to have overtaken
German in the capital (ibid.: 144).
Rich quantitative sociopsychological research into language attitudes has also
been emerging from trilingual regions. Bernaus, Masgoret, Gardner and Reyes
(2004) used Gardner’s Attitudes and Motivations Test Battery (AMTB) to investi-
gate the effect of the cultural background of immigrant learners on their attitudes
and motivation to learn Catalan, Spanish and English. Participants were grouped
in four categories: Spanish, South American, African and Asian. Cultural back-
ground turned out to be linked to very few differences. Attitudes and motivation
were less positive for Catalan than for both Spanish and English, with little differ-
ence between Spanish and English. A factor analysis showed that integrative moti-
vation was generally language-specific (ibid.: 75). One could argue that the
categorization is problematic as it compares a group of Spaniards, in other words,
a subgroup of Western Europeans with a huge and extremely diverse group of
Africans and Asians. It is also unclear to what extent South Americans differ from
the Spaniards as they share the same language.
Lasagabaster (2005) and Ibarraran, Lasagabaster and Sierra (2007) have con-
sidered the attitudes of students in a Basque university towards their three lan-
guages (i.e. Basque, Spanish and English) and found that the most positive attitudes
were towards English and their own L1 (Basque or Spanish). The first language
174 Jean-Marc Dewaele
determined the attitude towards English: Spanish L1 participants had more favour-
able attitudes than those with Basque as L1. In an attempt to reach more generaliz-
able conclusions on language use and attitudes, Lasagabaster and Huguet (2006)
asked researchers from nine multilingual European states/areas to apply the same
research design in their unique context and have them analyse the data using the
same techniques in order to answer common research questions. The resulting
studies took into consideration the attitudes towards their own majority and minor-
ity languages as well as towards the ‘foreign’ languages.
Dewaele (2005) found that Flemish high school students’ attitudes towards
English (the L3) were much more positive than those towards French (the L2),
despite the longer and more intense formal instruction in French. Students
who described themselves as more Flemish (i.e. their regional identity) than
Belgian (i.e. their national identity) displayed more negative attitudes towards
French. The more negative attitude towards French was attributed to the tense
socio-political relations between the Dutch and French speaking communities in
Belgium while English is generally perceived to be a ‘cool’ lingua franca. Personal-
ity dimensions were not found to be significantly linked to attitudes to French or
English.
Attitudes towards foreign languages are often influenced by local stereotypes
and beliefs. Diab (2006) studied foreign language students’ attitudes and beliefs
about learning English and French in Lebanon. Diab found that students per-
ceived English as an easy language and French as a difficult one, which reflects a
popular belief in Lebanon. The instrumental motivation to learn English was
found to be stronger than that for learning French.
It would be wrong to consider sociopsychological research as monolithic in
its methodological approach. Ushioda, one of the eminent researchers in the
field, carried out a longitudinal qualitative study into 20 Irish learners of French
(Ushioda, 2001). She relied on interviews in which she asked participants to reflect
on their motivation for learning French (ibid.: 99). She distinguished eight dimen-
sions in the interview material: academic interest, language-related enjoyment,
desired levels of L2 competence, personal goals, positive learning history, personal
satisfaction, feelings about French-speaking countries or people, and external
pressures (ibid.: 102). Her longitudinal approach allowed her to see motivation
not as a cause or the product of specific learning experiences but rather an
on-going process (ibid.: 122). She found that the most successful participants
engaged in intrinsic motivational processes more often, reminding themselves
about their past and future successes and their aspirations. This allowed them to
‘take control of their affective learning experience and to sustain their involve-
ment in language learning’ (ibid.: 122). Less successful learners focused more on
external incentives and blamed factors beyond their control for their lack of
progress.
Some sociopyschological studies have tried to control for the effect of the local
sociopolitical context by gathering data from participants in different countries
and on different continents. Dewaele (in press) carried out a quantitative study,
partly based on the attitudes questionnaire by Lasagabaster (2005), among NS and
Perception, Attitude and Motivation 175
NNS students of French. The study focused on the perceived characteristics of the
French language and of Francophones, as well as the perception of areas of diffi-
culty in the French language. NS and NNS were found to differ significantly in
their judgements of the characteristics of the French language. NNS judged French
to be less difficult overall than French NS. Both groups also disagreed on what area
of French is the most difficult. The NNS judged French to be significantly more
poetic, more useful and more romantic than NS of French. NNS judged Franco-
phones to be significantly more cultivated and polite, but not more open or nicer.
It was argued that the NNS’ more positive attitudes could be linked to their
on-going investment in the foreign language learning process (cf. Peirce, 1995)
and the fact that they may have idealized certain aspects of the target language and
target language community.
An even more de-contextualized study of language attitudes was carried out in
Dewaele (2007a, b). These studies were based on a database of 1,579 multilinguals
who filled out the Bilingualism and Emotions Questionnaire (BEQ) (Dewaele and
Pavlenko, 2001–2003). Participants were multilinguals living all over the world with
more than 70 different first languages. The sample contained a strong proportion
of female, highly educated multilinguals. The first part of the questionnaire con-
tained 13 questions relating to participants’ sociobiographical and linguistic back-
ground. The second part of the questionnaire consisted of 13 close-ended
Likert-type questions on language choice for the expression of various emotions
with various interlocutors, on code-switching behaviour in inner and articulated
speech, on the use and perception of swear-words, on attitudes towards the differ-
ent languages and finally on communicative and foreign language anxiety in the
different languages. The last part of the BEQ presented 5 open-ended questions
which asked about: (1) the weight of the phrase ‘I love you’ in the participants’
respective languages; (2) their linguistic preferences for emotion terms and terms
of endearment; (3) emotional significance of their languages; (4) language for
emotional interactions; (5) ease or difficulty of discussing emotional topics in lan-
guages other than the first. The data elicited through the open questions yielded
a corpus of about 150,000 words. The complete BEQ has been incorporated as an
appendix in Pavlenko (2005: 247–256). Rather than focusing on specific languages,
languages were categorized according to the order of acquisition (L1 to L5). The
aim of the research was to investigate whether there are any systematic differences
between perceptions and frequency of use for languages acquired first and later in
life. Dewaele (2007a) compared perception of usefulness, colourfulness, richness,
poetic character, emotionality and warmth in the L1 and the L2 of participants.
Between one half and three quarter of participants choose the statement ‘abso-
lutely’ to describe the positive characteristics of their L1 compared to between a
third to two thirds for the L2. Statistical analyses revealed that the values for the L1
were significantly higher for all language characteristics compared to the L2.
One of the research questions in the study was whether the perception of
emotionality and usefulness of a language predicted self-rated proficiency in that
language. It emerged that the L2 was felt by three quarters of participants to be
maximally useful, this fell to less than half for the L3 and hovered around a third
176 Jean-Marc Dewaele
for the L4 and the L5. The perception of emotionality was quite different: slightly
more than a third of participants judged the L2, L3, L4 and L5 to be maximally
emotional. Participants’ perception of usefulness and emotionality of a language
were found to be significant predictors of self-rated proficiency in the L2 and the
L4, with usefulness also being a significant predictor in the L3.
SLA researchers from different research traditions generally agree that attitudes
towards foreign languages are always shaped by the people who speak these
languages, and that some inherent characteristic of the language may affect the
attitude of a group of people or individuals towards that language.
This dissociation between the language and the cultural values that the language
represents has been amply documented in the literature. Some teachers want
learners to acquire competence in the foreign language and culture without actu-
ally feeling any sympathy for the speakers of that language. In other words, an
instrumental motivation is encouraged but an integrative one is strongly discour-
aged. An illustration of this peculiar situation is provided by Pavlenko (2003) who
describes her first English language class in the former Soviet Union in the mid-
seventies. Her teacher framed the objective of the class as the necessity to master
English, the language of the imperialists, in order to better defeat this enemy:
A similar attitude existed towards Russian during the Cold War in the West, or
towards German during the two World Wars (Pavlenko, 2003).
Perception, Attitude and Motivation 177
Dörnyei (2005) reflects on his many years in Russian classes in Hungary which
led to only minimal competence. He attributes his lack of progress in that
language to a general resistance among teachers and students towards the lan-
guage of the oppressor. Nikolov (2001) came to a similar conclusion in her study
on unsuccessful Hungarian foreign language learners: of the 87 participants who
had studied Russian, none ‘reported any proficiency in Russian or a liking of the
language and culture’ (ibid.: 162). Attitudes towards certain languages and target
language communities can also be ambiguous rather than completely positive or
negative. LoCastro (2001) reports that her Japanese students had positive attitudes
towards English but were unwilling to adopt certain English pragmatic norms that
differed from Japanese ones because of their desire to retain a Japanese identity.
More negative attitudes exist in Iran and much of the Middle East towards English,
yet, it remains typically one of the frequently taught foreign languages (Ahmadi,
2008). That language is taught without the cultural values and attitudes of that
group. This so-called language teaching ‘with a clothespin on the nose’ is dictated
by political or economical necessity. It is also dependent on the political climate
and can evolve quite rapidly.
Dutch in Belgium suffers from negative attitudes by many young Francophone
Belgians who feel they need to learn Dutch because knowledge of both French
and Dutch is often required for jobs in Belgium (Mettewie et al., 2006). These
Francophone Belgians thus rate Dutch as useful in the Belgian context but they
often find it a difficult and ugly language, with very little international stature. Not
surprisingly, the motivation to learn a language perceived as ugly and not very
prestigious will be low. Emotional attitudes towards languages may thus be quite
independent from purely instrumental evaluations.
While attitudes towards foreign languages are often shaped by the global geo-
political and sociopolitical context, they can also be quite unique to the individual.
Indeed, ‘FL learners are not passive “consumers” of institutionalised ideologies
and may engage in resistance and opposition’ (Pavlenko, 2003). Her own act of
resistance as a pupil was walking away from the English language class and opting
for French instead. This had obviously nothing to do with her attitude towards
English (which must have been positive considering her mother was a teacher of
English), but rather with the attitude towards the teacher. Mettewie (2004)
similarly reported that the attitude towards the foreign language teacher deter-
mines to a large extent the attitude towards the foreign language, especially for
languages which are rare in the usual linguistic landscape of the learners. Noels
(2003) pointed out that the teachers’ communicative styles also have a strong
effect on students’ continued motivation to study the L2. The decision to learn a
particular foreign language and to soak up the cultural and political values attached
to that language can be an act of resistance in itself. Pavlenko (2003) refers to
Natasha Lvovich, a translator and professional teacher of French in the former
Soviet Union who bitterly regretted being unable to travel to France and for whom
French became an intellectual escape from the oppressive communist regime:
‘Associating French with intellectualism, sophistication, and nobility, she created
an imaginary French identity for herself, learning to speak with a Parisian accent,
memorising popular French songs, reading French classics and detective stories in
178 Jean-Marc Dewaele
argot, mastering numerous written genres, cooking French food (from locally
available ingredients)’ (Pavlenko, 2003: 326).
The attraction of a foreign language and culture is not necessarily linked to an
oppressive home environment, it can simply be a matter of falling in love with it.
Kaplan, an American academic, describes her infatuation with French and France
as a form of addiction: ‘sometimes I don’t want to need French so much. I want to
be free of it’ (Kaplan, 1993: 207).
The famous Latin saying De gustibus et coloribus non est disputandum (One mustn’t
quarrel about tastes and colours) is also applicable to the reaction of people
towards foreign languages. These tastes can of course be affected by stereotypes
about the languages. Some languages sound repulsive, other sound sexy and attrac-
tive in the ears of potential learners. Piller (2002) coined the term ‘language
desire’ to describe the attraction of language learners. In her study of linguistic
practices of bilingual couples, Piller (2002: 269) notes that some participants
were in love with English or German as an L2 long before they actually met their
partners. However, languages may have completely different effects on different
people. The American philosopher Richard Watson, a specialist in Descartes,
could read French but not speak it. After an invitation to give a paper at a confer-
ence in Paris marking the anniversary of Descartes’ Discours de la Méthode, he
decided it was time to learn to speak French. Six months of conversation classes
with a tutor had very little effect and the conference presentation turned out disas-
trously. Among the affective obstacles to speaking French, he mentions the fact
that French sounded ‘syrupy’ and ‘effeminate’ and a language that ‘Real Men’
would not speak (Pavlenko, 2005: 67).
A similar report of ambiguous attitudes of Americans towards French is reported
in Kramsch (2003). She analysed metaphors and subjective construction of beliefs
of foreign language learners. One student of French at UC Berkeley wrote that
learning French is ‘like eating regurgitated paté’. Kramsch (2003) notes that this
metaphor might evoke and be linked to a cluster of other ambiguous attitudes and
beliefs such as ‘France is the country of good food’ as well as ‘learning French is
like having phlegm stuck in the back of my throat’.
The more ambiguous attitude towards French has also been present among
British schoolchildren for decades, with clear gender differences in the preferred
languages (Williams et al., 2002). Participants explained the boys’ preference for
German in terms of French being a more feminine language and German more
masculine: ‘As a high proficiency year 9 boy put it, “French is the language of love
and stuff” while German is “the war, Hitler and all that”’ (ibid.: 520). A girl from
the same year group confirmed: ‘I reckon girls are really into French: they like the
way that French sounds more than boys do’ (ibid.: 521).
In sum, whatever researchers’ epistemological or methodological stance, there
is a broad agreement that perceptions and attitudes towards foreign languages
and motivation to learn them are linked to both the macro context (geopolitical,
social and historical factors resulting in specific opinions within groups of people)
and the micro context, an individual’s like or dislike of the language teacher,
admiration or distaste for specific speakers of the target language. This complex
Perception, Attitude and Motivation 179
interaction of macro and micro factors can produce ambiguous attitudes, and
specific types of motivation (i.e. purely instrumental ones) or even a total lack of
motivation.
Studies on the year abroad have considered the effect of a period of total immer-
sion in the target language culture on learners’ interlanguages and identities, as
well as on their social and cultural understanding of the target language culture
(Freed, 1995; Coleman, 1998; Collentine and Freed, 2004; Dufon and Churchill,
2006). SLA researchers working in an etic perspective have been particularly
interested in development of fluency, accuracy, lexical richness, grammatical intui-
tions of study abroad students (Howard, 2005; Towell and Dewaele, 2005). A fair
amount of research has also been carried out on the acquisition of sociolinguistic
and sociopragmatic norms in the target language through interactions with NS
(Barron and Warga, 2007; Kinginger and Blattner 2008; Regan et al., 2009). Recent
research has also focused on the development of speed and accuracy of pragmatic
comprehension in the target language during the stay abroad (Taguchi, 2008).
Learners who become frequent users of the L2 in authentic interactions with NSs
typically make the best progress, though not always in terms of accuracy. The will-
ingness to communicate in the L2 is highly variable among learners/users in the
target language community (MacIntyre et al., 1998) which might explain the wide
inter-individual variation on the return from the year abroad. Some students come
back from the year abroad with extremely positive attitudes towards the target
language community, others come back loathing every speaker of the target lang-
uage (see also Mike Byram’s chapter in this volume).
Kaplan (1993) presents the case of Edna, an American student who spent a year
in France and who has fallen in love with every aspect of the target language and
culture:
Edna is just back from a year of study abroad where she took the majority of
her classes at ‘Sciences Po’ – the French institute for the study of political
science. She is imbued with the style and the tone of a French Sciences Po
student. She went even farther than I did to make herself over; her French is
indistinguishable from that of a student at Sciences Po. She is every French
professor’s dream! She acquired all the micro-traits of intellectual francophilia:
in her notebook, she underlines her main points, marked with Roman numer-
als ‘I.’, ‘II.’, ‘III.’, in contrasting colored-pencil colors, and for every big point,
she’s got a little point ‘a.’, a little point ‘b.’, sub-points ‘i.’, ‘ii.’, ‘iii.’ This,
I imagine, she got from sitting in French university amphitheaters where the
professors actually say ‘aspect number three, second part, small a’ as they’re
talking, to assist the students taking notes. (Kaplan, 1993: 175)
180 Jean-Marc Dewaele
Edna seems to have been driven, in Gardner terms, by extreme integrative motiva-
tion. Not only does she manage to sound exactly like her French peers at Science
Po, she seems also to have adopted some specific French academic traits. It is
perfectly possible, of course, that she possessed these traits before her stay in
France, but that Kaplan attributes them to her contact with French academic
culture.
Kinginger and Farrell (2005) investigated the stories of three American
students who spent a year in France with a focus on their confrontation with
gendered practices and ideologies. The authors found that the students examined
and interpreted their observations of gendered practices through the lens of
American ideologies of French gender. For one female student ‘this confrontation
was embedded in and added to a general sense of alienation from the social
context and from language learning’ (ibid.: 15). The student rejected gender
practices which she identified as sexual harassment and ‘used these as contribu-
tions to her rationale for rejection of emotional investment in French language
competence’ (ibid.: 15). A second female student accommodated to a certain
degree to these gender practices but learned how to defend herself from unwanted
sexual advances. Rather than being deterred, she learned how to manipulate them
and used them as opportunities for social interaction in French. She also started to
question her own image and behaviour in order to enhance her success as
a L2 user of French (ibid.: 15). The third, male, student stuck to his American
gender identity in the new social environment and was able to interact freely with
men and women, creating a story of his own gallantry (ibid.: 15).
The three students considered in Kinginger and Farrell (2005) form part of a
group of 24 American students that Kinginger (2008) followed during their year
in France. Data on the individual experiences and evolving emotional states were
collected through pre- and post-study abroad interviews, and by participants’
diaries. Kinginger (2008) was inspired by Kramsch’s observation that ‘language
learners are not just communicators and problem solvers, but whole persons with
hearts, bodies, and minds, with memories, fantasies, loyalties, identities’ (Kramsch,
2006: 251). The study shows that attitudes towards the French language and the
motivation to learn it vary wildly among certain participants. These attitudes are
typically associated with specific interlocutors, like the members of the host family.
One of the students called Beatrice reports being mocked by the host sisters about
her accent in French. Her final journal entry reads as follows:
Let me just say this, I am ready to leave my host family. Although it used to be
cute and funny when they would imitate my French accent, it now pisses me off.
I have worked my ass off here while I have been in france [sic] to learn their
language and I know deep inside that I have made a lot of progress and I really
do not find it funny anymore when they mock me 5 times a day. . . . I sure as hell
do not want to learn your language when you speak to me like that. I think
I came to France with a pretty good attitude. Talk about disenchanted!
(Kinginger, 2008: 179)
Perception, Attitude and Motivation 181
Kinginger (2008) concludes that despite the fact that each student’s experience
was unique, a certain thematic coherence emerged from the assembled stories of
individuals.
Garrett and Young (2009) present a longitudinal case-study based on interviews
throughout an 8-week Portuguese course for beginners. The first author described
her language learning experiences to the second author. Sessions were transcribed,
then coded and analysed. The authors developed a theoretical model grounded in
the learner’s experiences in order to understand the learner’s affective responses
to the language learning process, the events from which her affect sprang, and her
affective trajectory over the 8 weeks. The authors point out that the learner and
her experiences were unique, and that other learners may respond differently to
similar experiences. They do feel that their observations are relevant, however, as
affective responses of foreign language learners and responses to events in the
classroom clearly determine individual learner’s affective states and the trajectory
of change in those states throughout the language learning process (Garrett and
Young, 2009).
Initially positive attitudes towards a foreign language can fade quickly for
participants in study abroad programmes. Isabelli-García (2006) reports the
cases of four American students in Argentina, including Jennifer, who entered
the program with a positive attitude and a strong instrumental motivation to use
Spanish. Encountering housing difficulties at the start of her stay and unfamiliar
gender-related practices, she soon developed a negative attitude towards the host
culture ‘and her low motivation to learn the language hindered her from includ-
ing more Argentines in her social network; she lacked any investment to learn the
target language’ (ibid.: 254).
As Isabelli-García (2006) shows, attitudes towards the language and the target
language country can vary over a period of months but they can also fluctuate
on a daily basis according to specific events in the students’ life. Isabelli-García
refers to the U-shaped development of attitudes, starting with (often unrealisti-
cally) positive attitudes prior to travelling to the given country, which often weaken
at the beginning of the stay. This could be the consequence of the unforeseen
negative effect of the cultural and language barriers experienced while engaged
in intercultural communication (Isabelli-García, 2006). Students who stay long
enough, may overcome the negative attitudes and end up with more realistic
positive attitudes towards the target language speakers and country (Stangor et al.,
1996).
In sum, research on the effect of study abroad on SLA in the etic, quantitative
perspective typically shows moderate development on a range of linguistic and
pragmatic variables in the L2. The high degree of inter-individual variation on the
return to the home country has puzzled quantitative-oriented researchers. It is
here that a more emic, qualitative-oriented research into perception, attitudes and
motivation proves extremely fruitful. Detailed case studies show that attitudes, per-
ception and the investment in the learning of a foreign language sometimes
depend on a number of unrelated factors: on sheer chance (staying with a nice
182 Jean-Marc Dewaele
8.6 Discussion
The review of the literature on language attitudes and motivation shows a field
that is evolving in different directions, with a fair degree of cross-fertilization.
I would argue that the monolingual and monocultural bias in quantitative socio-
psychological approaches is largely a thing of the past. Research in the European
context, for example, is often based on participants from mixed ethnic and
linguistic backgrounds who typically learn multiple foreign languages. A Euro-
barometer survey in 1997 showed that more than three quarters of the 9,400 young
Europeans interviewed declared to speak a foreign language and about two
thirds desired to learn another (European Report on the Quality of School Education).
(Incidentally only 45 per cent of British participants reported knowledge of a
foreign language and a 62 per cent desire to acquire another language.)
Although concepts such as ‘integrative motivation’ do appear in the attitudinal
studies on SLA in the European context, it obviously does not imply the loss of
the L1 culture. The Hungarians, Basques, Catalans, Frisians learning English may
want to integrate a nebulous community of English L2 users rather than a specific
target group of NSs of English. The same holds for Flemings learning foreign
languages. They may want to broaden their sociocultural horizon by learning
foreign languages, thus joining the legions of young European multilinguals,
but this learning process does not imply the sacrifice of the Flemish culture.
It comes therefore as no surprise that Dörnyei, a European researcher, proposes to
abandon the concept of ‘integrativeness’ and focus instead on identification
aspects and on the learner’s self-concept (Dörnyei and Cziser, 2005; Dörnyei,
2005).
Dörnyei’s recent focus on theoretical, epistemological and methodological
issues in attitude and motivation research (Dörnyei, 2005, 2007) could be inter-
preted as an answer to Pavlenko’s criticism concerning the lack of explanatory
validity of basic constructs in sociopsychological research and by her doubts about
the validity of questionnaires. The fact that these are based on self-evaluation do
not render the outcomes invalid, but everything evidently depends on what claims
are made from the data obtained through questionnaires. I completely agree with
Dörnyei (2003, 2007) about the advantages of quantitative approach in attitude
and motivation research. It is ‘systematic, rigorous, focused, and tightly controlled,
involving precise measurement and producing reliable and replicable data’
(Dörnyei, 2007: 34). However, Dörnyei is aware about the limitations of quantita-
tive research namely its limited general exploratory capacity and he therefore
Perception, Attitude and Motivation 183
Pavlenko’s claim that motivation, attitudes are causal, unidirectional and stable
in sociopsychological research seems shaky. Even in his early work, Gardner
insisted on the dynamic character of attitudes and motivation in a complex social
context. The pyramid model of WTC shows how attitudes and motivation are
linked to both enduring social factors, psychological dimensions, situational
factors which may ultimately affect the decision to communicate in an L2. Motiva-
tion and attitudes are therefore both sources of input and output in the pyramid.
Dörnyei’s (2001a) process-oriented approach had already highlighted the fact
that attitudes and motivation are highly dynamic, fluctuating over different time-
frames and related to various actional/engagement contingencies. The pedagogi-
cal implications have been highlighted by Dörnyei (2002) who pointed out that,
even if certain learners do not enjoy a particular language learning task, they might
still try their best if they have positive attitudes towards the foreign language
course (i.e. the teacher, the learning materials, fellow students . . .). On the
other hand, some language tasks may be perceived by learners to be particularly
motivating, and could help them maintain high levels of motivation in the learn-
ing of the foreign language. Dörnyei (2001b) has also described ways to strengthen
the learner’s determination in learning the foreign language and maintain the
motivation of disappointed learners.
Pavlenko’s criticism that the separation assumed in sociopsychological
approaches between social factors and psychological factors is artificial is not one
that seems to have overly worried quantitative researchers. Personality psycho-
logists recognize that an individual’s personality is formed by a combination of
biological and social factors (Furnham and Heaven, 1999). Sociopsychologists
seem less worried that age, gender or ethnicity may also be socially constituted.
However, they do accept that motivation or language learning beliefs are constantly
shaped by the social and political contexts in which the learners find themselves
(MacIntyre et al, 1998).
Pavlenko’s criticism about the Anglo-centrism characterizing most research in
Europe and North America, thus raising the issue of generalizability of the find-
ings, seems to have been heard by Gardner: Bernaus, Masgoret, Gardner and Reyes
(2004) and Gardner (2006) used the AMTB to gather data from immigrants and
from foreign language learners in nine European countries to demonstrate the
stability of the factor structure emerging from the data. Moreover, Yashima,
Dewaele, Dörnyei and colleagues, Lasagabaster and his colleagues have analysed
Perception, Attitude and Motivation 185
8.7 Conclusion
Note
1 This paper, nominated by the editorial board of the Modern Language Journal, was
awarded the K. W. Mildenberger Prize for distinguished publication in the field of
language education by the Modern Language Association.