USER
what cite or direct qoute supports this? Psychological research further supports this idea. According to self-determination theory, individuals who focus on intrinsic motivation—activities they find passion in experience greater fulfillment compared to those who chase external rewards. This the information you can chose from: construct of passion. However, like flow, these concepts do not
make the distinction between two types of interest or talent that
reflect different types of engagement while the passion conceptu-
alisation does make such a distinction (i.e., harmonious and ob-
sessive passion). Finally, intrinsic motivation also shares some
conceptual similarity with passion, as both involve interest and
liking toward the activity. However, intrinsically motivated activ-
ities are typically not seen as being internalised in the person’s
identity (Deci & Ryan, 2000) and are best seen as emerging from
the person-task interaction at the short-term level (Koestner &
Losier, 2002). Furthermore, extrinsic motivation does not entail
performing the activity out of enjoyment, but for something out-
side of the activity. A fundamental difference between extrinsic
motivation and passion is thus the lack of liking for the activity
(see also Vallerand et al., 2003, Study 2).
In sum, while the passion framework does share some concep-
tual similarities with other motivational constructs, it also differs
from them in significant ways. Noteworthy, is that a complete
understanding of people’s engagement in an activity (such as the
cause of Canadian psychology discussed previously) necessitates
the distinction of two different types of heavy activity involve-
ment. We now turn to an investigation of the empirical evidence
for the Passion Model.
Research on Passion
Over the past few years, we have conducted a number of studies
on passion. These studies pertain to a variety of activities, settings,
participants, and outcomes. Below, we briefly review the results of
some of these studies that pertain to the concept of passion,
psychological adjustment, physical health, interpersonal relation-
ships, performance, and the development of passion.
On the Concept of Passion
In the initial study (Vallerand et al., 2003, Study 1), we sought
to develop the Passion Scale and to test the validity of our defi-
nition of passion. To that end, college students (n 539) com-
pleted the Passion Scale with respect to an activity that they liked,
that they valued, and in which they invested time and energy (i.e.,
the passion definition), as well as other scales allowing to test
predictions derived from the Passion Model. Interestingly, 84% of
our participants indicated that they had at least a moderate level of
passion toward an activity in their lives. Passionate activities
ranged from physical activity and sports and music to watching
movies and reading. Participants reported engaging in their pas-
sionate activity for an average of 8.5 hours per week and had been
engaging in that activity for almost 6 years. Thus, clearly passion-
ate activities are meaningful to people and do not simply reflect a
fleeting interest. These results also suggest that for most partici-
pants, initiation with the passionate (self-defining) activity started
around adolescence. This would be in line with the position that
adolescence is a crucial period of identity construction (e.g.,
Eccles, Barber, Stone, & Hunt, 2003; Erikson, 1968).
Research from the Vallerand et al. (2003, Study 1) study has
provided empirical support for several aspects of the passion
conceptualisation. First, results from exploratory and confirmatory
factor analyses provided strong support for the existence of two
constructs corresponding to harmonious and obsessive passion.
The Passion Scale consists of two subscales of seven items each,1
the Obsessive (e.g., “I almost have an obsessive feeling toward this
activity”) and Harmonious subscales (e.g., “This activity is in
harmony with other activities in my life”). Subsequent research has
supported the bifactor structure validity of the scale in a number of
life contexts including sports (Vallerand et al., 2006, Study 1),
gambling (Castelda, Mattson, MacKillop, Anderson, & Donovick,
2007; Rousseau, Vallerand, Ratelle, Mageau, & Provencher, 2002),
and work (Houlfort & Vallerand, 2006; Vallerand & Houlfort, 2003).
Furthermore, internal consistency analyses supported the reliability of
the scale.
Second, results from the Vallerand et al. (2003) Study 1 also
revealed that both harmonious and obsessive passions were posi-
tively associated (partial correlations, controlling for the common
variance between the two types of passion) with measures of
activity valuation and measures of the activity being perceived as
a passion, thereby providing support for the definition of passion.
In addition, while both types of passion were seen as being part of
one’s identity, only obsessive passion was positively related to a
measure of conflict with other life activities. Third, empirical
evidence (Vallerand et al., 2003, Study 1) has also shown that
harmonious and obsessive passions were associated with different
affective experiences. For instance, results from partial correla-
tions between the two types of passion and affective variables have
shown that when controlling for obsessive passion, harmonious
passion is positively associated with positive experiences such as
flow and positive emotions during activity engagement, and pos-
itive emotions and the absence of negative affect following task
engagement. On the other hand, when controlling for harmonious
passion, obsessive passion is positively associated with negative
emotions (especially shame), both during and following activity
engagement. Particularly interesting is the finding that obsessive
passion has been found to be associated with negative affect
(notably shame and anxiety) and rumination when the person is
prevented from engaging in the passionate activity while harmo-
nious passion is unrelated to these negative experiences (Vallerand
et al., 2003, Study 1). These latter findings suggest that obsessive
passion entails some negative feelings akin to dependence toward
the activity, while harmonious passion does not. Finally, additional
research in the initial publication has also shown that obsessive
(but not harmonious) passion leads to rigid persistence in ill-
advised activities such as cycling over ice and snow in winter
(Vallerand et al., 2003, Study 3) and pursuing one’s engagement
in gambling even though it has become pathological in nature
(Vallerand et al., 2003, Study 4). We will return to this issue in a
later section.
In sum, initial research provided support for the concept of
harmonious and obsessive passion. We now turn to different lines
1 The original scale (Vallerand et al., 2003) consisted of two 7-item
subscales. A slightly revised scale consisting of two 6-item scales is now
used. These subscales correlate very highly with their respective original
subscale (r .80 and above) and yield the same findings with determinants
and outcomes. However, the correlation between the harmonious and
obsessive passion subscales is lower than that between the original ones. In
addition, we have used a 3-item version (Vallerand et al., 2007b, Study 2)
and even a 1-item version (Philippe & Vallerand, 2007) of each subscale
with much success.
3PSYCHOLOGY OF PASSION
of research that have explored some of the processes associated
with the passion construct.
Passion and Psychological Adjustment
An early study (Vallerand et al., 2003, Study 2) revealed that
engaging in the passionate activity (football) out of harmonious
passion predicted an increase in general positive affect in one’s life
(as measured by the PANAS; Watson, Clark, & Tellegen, 1988)
over the course of the entire football season. Conversely, partici-
pating in the activity out of obsessive passion predicted an increase
of negative affect in one’s life during the same time interval. If the
two types of passion are conducive to increases in general affect
over time, it is thus possible that passion also affects one’s psy-
chological adjustment. Research provides support for this hypoth-
esis. For instance, in a study with elderly individuals, Rousseau
and Vallerand (2003) showed that harmonious passion positively
predicted positive indices of psychological adjustment (life satis-
faction, meaning in life, and vitality) but negatively predicted
negative indices (anxiety and depression). Conversely, obsessive
passion positively predicted anxiety and depression, was nega-
tively related to life satisfaction, and was unrelated to vitality and
meaning in life. Research with young adults and teenagers using
different measures has yielded similar findings. Specifically, har-
monious passion was positively related to life satisfaction and
vitality (Vallerand, Salvy, et al., 2007, Studies 1 and 2; Vallerand
et al., in press, Study 2), while obsessive passion was either
negatively related (Vallerand, Salvy, et al., 2007, Study 2) or
unrelated (Vallerand, Salvy, et al., 2007b, Study 1; Vallerand et al.,
in press, Study 2) to these indices.
If passion affects psychological adjustment, then what are the
processes mediating such effects. We feel that at least two pro-
cesses might be at play in this relationship. A first deals with the
repeated experience of situational (or state) positive affect during
the course of engagement (as well as after engagement) of the
passionate activity which should lead people to have access to a
broader set of cognitive and social skills, thereby facilitating
psychological adjustment (Fredrickson, 2003). Research on pas-
sion (e.g., Mageau et al., 2005; Vallerand et al., 2003, Study 1;
Vallerand et al., 2006, Studies 2 and 3) has shown that harmonious
passion does facilitate positive emotions both during and after
activity engagement. Furthermore, Fredrickson (1998, 2001, 2003)
has shown that the experience of positive emotions have the virtue
of broadening people’s thought-action repertoires (Fredrickson,
2001), thereby leading to the use of more adaptive behaviours, and
thus better psychological adjustment (on this issue, see Frederickson
& Joiner, 2002). Since passionate individuals engage on average for
about 8 hours per week in their passionate activity, this means that
harmonious passion can lead people to experience each week 8 hours
of cumulative positive affect and adaptive-related repertoire on top of
what may be experienced in other life domains. Such additional
cumulative positive experiences represent roughly 10% of our waking
time and may indeed spice up our week and have facilitative effects
on psychological adjustment.
Our research under this heading has used path analysis in order
to look at the role of positive affect experienced during task
engagement as a mediator of the harmonious passion-psychological
adjustment relationship. A recent research by Rousseau and Vallerand
(in press) provided support for this hypothesis with passion toward
physical activity. At Time 1, participants completed the Passion Scale
with respect to physical activity, as well as measures of psychological
adjustment. At Time 2, immediately following an exercise bout, they
completed situational measures of positive and negative affect. Fi-
nally, at Time 3, they completed measures of psychological adjust-
ment again. Results from a path analysis revealed that harmonious
passion positively predicted positive affect which led to increases in
psychological adjustment from Time 1 to Time 3. On the other hand,
obsessive passion was unrelated to positive affect but positively
predicted negative affect. The latter did not predict psychological
adjustment. These basic findings were replicated in another study
(Vallerand & Rousseau, 2007). Overall, these findings provide strong
support for the role of situational positive affect experienced during
task engagement as a mediator of the effect of harmonious passion on
psychological adjustment.
The above research suggests that harmonious passion positively
contributes to psychological adjustment through the repeated ex-
perience of situational positive affect. However, such research also
showed that although obsessive passion is negatively related to
psychological adjustment, neither situational positive nor situa-
tional negative affect mediate the obsessive passion-psychological
adjustment relationship. So, what is the mediator of such a rela-
tionship? We posit that rigid persistence in activities with negative
returns for the person represents the second process underlying the
negative influence of obsessive passion on psychological adjust-
ment. Because the passionate activity is very dear to the heart of
those who engage in it (after all it is part of their identity), people
are likely to persist in them for a rather long period of time.
However, as described previously, there would seem to be some
differences in the type of persistence associated with the two types
of passion. With harmonious passion, the person is in control of
the activity. As such, the person can decide when to and when not
to engage in the activity and should even be able to drop out of the
activity if the latter has become permanently negative for the
person. Thus, behavioural engagement and persistence can be seen
as flexible. Such is not the case with obsessive passion. Typically,
because the activity has taken control of the person, obsessive
passion would also be expected to lead to persistence. However,
such persistence can be seen as being rigid because it can take
place not only in the absence of positive emotional experience, but
even when important costs are accrued to the person. Such rigid
persistence can lead the person to persist in the passionate activity
even though some permanent negative consequences are experi-
enced, eventually leading to low levels of psychological adjust-
ment.
Recent research of ours in the realm of gambling provides
supports for this analysis. Such research has shown that obsessive
passion predicts pathological gambling while harmonious passion
is unrelated to it (Ratelle et al., 2004). Furthemore, in a recent study
on the prevalence of gambling problems (Philippe & Vallerand,
2007), it has also been shown that obsessive passion predicts the
occurrence of pathological gambling, presumably because obsessive
passion entails a rigid persistence in the activity that is out of the
person’s control. But is it the case? Is rigid persistence toward the
passionate activity really at play in psychological problems. In one
particular study (Vallerand et al., 2003, Study 4), we tested this
hypothesis more directly. Specifically, in that particular study, we
compared the two types of passion of regular casino gamblers (who
played at least once a week) with those of people with gambling
4 VALLERAND
problems so severe that they end up asking the Montreal Casino to bar
them from entry. Clearly, these individuals do have important prob-
lems (93% of the self-exclusion sample displayed pathological gam-
bling vs. 37% for the regular casino players). As such, the self-
exclusion gamblers should have stopped gambling a long time ago.
And yet, they have not. The question is why?
The results revealed that the self-exclusion gamblers reported
significantly higher levels of obsessive passion than regular casino
gamblers. No difference existed on harmonious passion. Further-
more, for the self-exclusion group, their obsessive passion was
significantly higher than their harmonious passion. There were no
differences between the two scales for the control group. It would
thus appear that obsessive passion has led self-exclusion gamblers
to persist while they should not have. When one takes into con-
sideration the fact that pathological gambling has been found to
lead to a sleuth of problems that include divorce, losing one’s job,
and even suicide (Walker, 1992), clearly obsessive passion for an
activity such as gambling can send one on a path of severe
psychological problems. Longitudinal research involving other
types of negative activities (e.g., using drugs, Internet sex) is
needed in order to replicate the Vallerand et al. (2003, Study 4)
study findings.
In sum, preliminary evidence reveals that harmonious passion
may positively contribute to psychological adjustment through its
impact on situational positive affect. Obsessive passion, on the
other hand, does not contribute to psychological adjustment and
may even detract from it through rigid persistence in ill-advised
activities such as gambling.
Passion and Physical Health
Passion may also affect one’s physical health in a number of
ways. One of these entails leading one to engage in risky behav-
iour. Take cycling for instance. Cycling in the spring, summer, and
fall can be a lot of fun and can promote one’s health. However, the
reality in the winter is drastically different (at least in the Province
of Quebec). The roads are icy and full of snow, and they make
cycling a very hazardous affair that may lead to falls and injuries.
Clearly, it would be advisable not to cycle under such conditions.
If our hypothesis on the rigid persistence of obsessive passion is
correct, then obsessive passion should lead one to engage in risky
behaviours such as winter cycling. On the other hand, if we are
correct with respect to the flexible persistence of harmonious
passion, then the latter should not lead to engage in such a
behaviour. In a recent study (Vallerand et al., 2003, Study 3), we
tested these hypotheses. Cyclists (n 59) completed the Passion
Scale in August with respect to cycling. Six months later, they
were contacted again through e-mail to determine who was still
cycling in the dead of winter (in February). Results showed that
only 30% of participants were still cycling in winter. It was found
that those persistent cyclists had reported higher levels of obses-
sive passion 6 months earlier than those who did not cycle in the
winter. No differences were found with respect to harmonious
passion. Thus, obsessive passion may potentially affect people’s
health by leading them to engage in certain risky activities while
they should not (such as cycling in subzero temperature on icy
roads).
The above study (Vallerand et al., 2003, Study 3) while inter-
esting, did not show that obsessive passion led to injuries as such.
It only showed that obsessive passion puts people at risk of
experiencing injuries when engaging in the passionate activity
when they should not. A recent study with dancers (Rip, Fortin, &
Vallerand, 2006) has looked at the passion-injury relationship
more closely. The important question with dancers is not whether
or not they get injured, but rather what happens once they do.
Obsessive passion, as we have seen earlier, is associated with rigid
persistence. Therefore, when injured, obsessive passion should
lead people to continue dancing, thereby leading to chronic inju-
ries. On the other hand, with harmonious passion, the person is in
control of the activity. Thus, persistence can be seen as flexible. As
such, when dancing out of harmonious passion, the dancer can
decide to stop dancing when injured if there is a risk of developing
a chronic injury. In the Rip et al. (2006), university dancing
students (n 80) completed the Passion Scale as well as questions
pertaining to their coping behaviour when injured. Results from
partial correlations revealed that both types of passion were neg-
atively related to acute injuries. This was to be expected as the
more passionate the dancers, the more likely they are to practice,
keep fit, and thus prevent the occurrence of acute injuries. The real
test of our hypothesis deals with chronic injury. These findings
appear in Table 1. Results revealed that obsessive passion was
positively related to the number of weeks missed because of
chronic injuries, while harmonious passion was unrelated to
chronic injuries. Furthermore, analyses with the behavioural items
revealed that harmonious passion was positively related to seeking
information on the injury and its treatment and being able to
completely stop dancing, but was negatively related to ignoring the
pain and hiding an injury. Conversely, obsessive passion was
unrelated to hiding an injury but positively predicted ignoring the
pain and having one’s pride interfere with treatment. These latter
two types of behaviour are particularly troublesome as they lead
dancers to neglect their physical condition and may lead to injury
aggravation.
In sum, harmonious and obsessive passion can lead to positive
and negative effects on one’s health, respectively, through the
different types of persistence and engagement in risky behaviour
that they engender. Future research is needed in order to reproduce
the present findings with more objective measures of injuries such
as medical records.
Table 1
Partial Correlations Involving Harmonious and Obsessive
Passion, Injuries, Coping, and Prevention
HP OP
Duration of suffering from:
Acute injuries .44** .25 ns
Chronic injuries .01 .42**
Coping with injury: when injured
I seek info on the injury and its treatment .34 *** .02
I completely stop dance activities .26* .06
I ignore the pain .33*** .20*
I hide the injury .34*** .05
I partially stop dance activities .12 .30**
Pride prevents adequate treatment .19 * .40****
Note. Adapted from Rip et al. (2006).
*p .10. **p .05. ***p .01. ****p .001.
23 dancers reported having at least 1 acute injury; 26 dancers reported
having at least 1 chronic injury.
5PSYCHOLOGY OF PASSION
Passion and Interpersonal Relationships
Our conceptualisation on passion posits that having an obsessive
passion toward an activity should lead to conflict and problems in
other life activities, while this should not be the case for harmo-
nious passion. Results from the Vallerand et al. (2003, Study 1)
provided preliminary evidence for this hypothesis by showing that
obsessive (but not harmonious) passion for an activity was posi-
tively associated with experiencing conflict between activity en-
gagement and other aspects of one’s life. We believe that this basic
hypothesis has important implications for the quality of interper-
sonal relationships that people may experience outside the realm of
the passionate activity. Let’s take people who have a passion for
surfing on the Internet. They may surf the net at night, and if
obsessive passion is at play, they may forget what time it is, even
ignoring their spouse’s call to come to bed for the night. If done on
a repeated basis, such behaviour is a recipe for disaster in a
couple’s relationship. Such should not be the case with harmonious
passion because the person can let go of the passionate activity
when needed and thus should not experience conflict between
the passionate activity and the couple’s relationship. A recent
study of ours (Se´ guin-Le´ vesque, Laliberte´ , Pelletier,
Blanchard, & Vallerand, 2003) has specifically addressed this
issue. The results showed that controlling for the number of hours
that people engaged in the Internet, obsessive passion for the
Internet was positively related to conflict with one’s spouse, while
harmonious passion was unrelated to it. A subsequent study with
English soccer fans (Vallerand, Ntoumanis, et al., 2007, Study 3),
revealed that having an obsessive passion for supporting one’s
soccer team predicted conflict between soccer and the quality of
the couple relationship. Conflict, in turn, negatively predicted
satisfaction with the relationship. Harmonious passion was unre-
lated to conflict. Furthermore, individuals who were single were
asked to indicate if their passion for soccer was responsible for
their being single. Results revealed a strong positive correlation
between obsessive passion and this measure, but a negative cor-
relation for harmonious passion.
The above findings reveal that depending on one’s passion for a
given activity, one can experience conflict (or not) between that
activity and one’s relationships outside of the passionate activity,
thereby affecting the quality of such relationships. One question
not addressed by such research is whether one’s passion for a
given activity can affect the quality of relationships that one
develops in the purview of this activity. Research by Fredrickson
(2001; Waugh & Fredrickson, 2006) has shown that the experience
of situational positive affect is conducive to high quality of rela-
tionships. This is so because positive affect facilitates smiles,
positive sharing of the activity, connection, and openness toward
others that are conducive to positive relationships. Because har-
monious passion leads one to experience positive affect during
engagement in the passionate activity (e.g., Vallerand et al., 2003,
Study 1; Vallerand et al., 2006, Studies 2, 3), one would then
predict that it should therefore indirectly lead to high quality
relationships within the passionate activity. Conversely, because it
is typically unrelated to positive affect and correlated to negative
affect, obsessive passion would be expected to negatively affect
the quality of relationships that develop within the purview of the
passionate activity. Results from a series of studies conducted in a
variety of settings, including work and sports, has provided sup-
port for these hypotheses (Philippe, Vallerand, & Houlfort, 2007).
A third and last area where passion can affect relationships
pertains to the impact one’s passion for the loved one can have on
the quality of the intimate relationship. Past research (e.g., Hatfield
& Walster, 1978) has typically looked at passionate love as
roughly the equivalent of obsessive passion. However, as seen so
far, one can still be harmoniously passionate for an activity, and
this should apply to love relationships as well. Furthermore, as
found in past research on passion for activities, harmonious pas-
sion for the love relationship should predict a high quality rela-
tionship, while obsessive passion should be either negatively re-
lated or unrelated to the quality of the relationship. A first study
(Ratelle, Carbonneau, Mageau, & Vallerand, 2007, Study 1)
showed that harmonious passion positively predicted all dimen-
sions of the Spanier Dyadic Adjustment Scale (1976) while ob-
sessive passion was typically unrelated to the various subscales. In
a subsequent study (Ratelle et al., 2007, Study 3), both partners
completed the Passion Scale toward the love relationship, as well
as the Relationship Satisfaction Scale (Fletcher, Simpson, &
Thomas, 2000). Results revealed that both partners’ passion made
a significant contribution to the level of satisfaction toward the
relationship. For instance, both women and men’s harmonious
passion positively predicted women’s general satisfaction with the
relationship. Even better, men’s harmonious passion proved to be
a better positive predictor of women’s satisfaction with their sex
life than women’s own harmonious passion. Men’s obsessive
passion for the loved one was a significant negative predictor of
women’s satisfaction with their sex life. Similar findings were
obtained with men’s relationship satisfaction, although the predic-
tion from women’s passion was not as strong. These findings
suggest that having a balanced harmonious passion toward the
loved one can affect not only the quality of the relationship as we
experience it, but also as experienced by our partner. Future
research on this issue would appear particularly exciting, espe-
cially as pertains to the identification of the behaviours and pro-
cesses through which such interpersonal effects take place.
Passion and Performance
Research on expert performance reveals that high-level per-
formers spend several years of considerable engagement in an
activity with clear goals of improving on certain task components,
in order to reach excellence in their chosen field of expertise (what
is called deliberate practice, see Ericsson & Charness, 1994;
Starkes & Ericsson, 2003). What is the underlying motivational
force that leads individuals to spend so much time in perfecting
their skills in a given activity in order to achieve high proficiency?
We believe that passion represents one answer to that question.
Indeed, if one is to engage in the activity for long hours over
several years and sometimes a lifetime, one must love the activity
dearly and have the desire to pursue engagement especially when
times are rough. Thus, the two types of passion (harmonious and
obsessive) should lead to engagement in deliberate practice that, in
turn, should lead to improved sport performance. This model was
tested in a study with basketball players (Vallerand et al., in press,
Study 1). Male and female basketball players completed scales
assessing their passion as well as deliberate practice (based on
Ericsson & Charness, 1994). Coaches independently rated the
6 VALLERAND
athletes’ performance during the first game of an important tour-
nament. A path analysis provided support for the basic model.
Results revealed that both types of passion led to engagement in
deliberate practice that, in turn, led to objective performance.
These findings were replicated in a prospective design with dra-
matic arts performers (Vallerand, Salvy, et al., 2007, Study 1).
Also of interest is the finding that in the preceding study, harmo-
nious passion toward dramatic arts was positively and significantly
related to life satisfaction, while obsessive passion was unrelated
to it. This is in line with our work on passion and psychological
adjustment. It thus appears that both types of passion positively
contribute to deliberate practice and thus, indirectly, to perfor-
mance at least in the short term. However, it might very well be
that obsessive passion leads to some sense of “suffering” (or lower
levels of life satisfaction than harmonious passion) in the process
of pursuing high performance levels.
The results of the two performance studies presented above
established a direct relationship between passion and deliberate
practice, and an indirect relationship between passion and perfor-
mance (through deliberate practice). We conducted an additional
study (Vallerand et al., in press, Study 2) in order to examine the
psychological processes through which passion directly contrib-
utes to deliberate practice, and indirectly contributes to perfor-
mance. In line with Elliot (1997), we proposed that achievement
goals should represent important mediators between passion and
deliberate practice. Elliot and colleagues (Elliot & Church, 1997;
Elliot & Harackiewicz, 1996) have distinguished between three
types of achievement goals: mastery goals (which focus on the
development of personal competence and task mastery), perfor-
mance-approach goals (which focus on the attainment of personal
competence relative to others), and performance-avoidance goals
(which focus on avoiding incompetence relative to others). Passion
has been found to relate to affective and cognitive investment in an
activity, thereby implying that the individual is committed to
engaging in that activity in a competent manner. Harmonious
passion, being a rather pure autonomous form of regulation is
predicted to be positively related to mastery goals but not to
performance goals of either type. On the other hand, obsessive
passion, being a more pressured, internally controlling form of
regulation is likely to lead the individual to feel compelled to seek
any and all forms of success at the activity, and may even evoke
concerns about doing poorly. As such, obsessive passion should be
positively related to mastery and performance-approach goals, as
well as to performance-avoidance goals.
A study with water-polo and synchronized swimmers (including
some who were part of the junior national teams) was conducted
over an entire season to test the above model (Vallerand et al., in
press, Study 2). At Time 1 (the beginning of the season), individ-
uals completed the Passion Scale, the Achievement Goals Scale,
and scales assessing psychological adjustment. At Time 2, they
completed the Deliberate Practice Scale. Finally, at Time 3 (end of
the season), coaches assessed individuals’ performance over the
entire season. Results of a path analysis yielded support for the
proposed model. The results appear in Figure 1. As can be seen,
harmonious passion was found to lead to mastery goals that, in
turn, led to deliberate practice that positively predicted objective
performance. On the other hand, obsessive passion was positively
related to all three goals. While performance-approach goals did
not predict any variables in the model, performance-avoidance
goals negatively predicted performance. Finally, as in the Vallerand,
Salvy, et al. (2007, Study 1) study with the dramatic arts performers,
harmonious passion was positively associated with psychological
adjustment while obsessive passion was unrelated to it. This basic
model was replicated in another study involving students who had a
passion toward studying psychology as their future profession and
with objective exam scores in a psychology course as a measure of
performance (Vallerand, Salvy, et al., 2007, Study 2).
The above findings had been obtained with teenagers and young
adults who were at best at the national level (some of the water
polo players in the Vallerand et al. [in press, Study 2] study were
on the Junior Canadian team). It is thus of particular interest that
these findings have been replicated with adult high elite perform-
ers. Specifically, in a study with national and international classi-
cal musicians, Bonneville-Roussy and Vallerand (2007) were able
to replicate the model of Vallerand et al. (in press, Study 2—see
Figure 1). Thus, these findings would appear rather robust, even
with very high-level performers. Furthermore, once again, harmo-
nious passion positively predicted life satisfaction. Clearly, har-
monious passion does appear to be an important predictor of both
performance and psychological adjustment.
In sum, it appears that harmonious passion contributes to ob-
jective indices of performance. Such effect seems to take place
through mastery goals that lead to deliberate practice, that, in turn,
leads to performance. The role of obsessive passion in perfor-
mance is rather complex as it positively predicts mastery goals
(which leads to performance through deliberate practice) but also
performance-avoidance goals which may negatively influence per-
formance. These findings suggest that having the goal of beating
others at all costs may not only have some negative effects on
intrinsic motivation (e.g., Vallerand, Gauvin, & Halliwell, 1986)
but may also undermine performance. Because of the obvious
implications of these findings, additional research is warranted to
test its generalisation in other fields of endeavor over longer
periods of time.
On the Development of Passion
Thus far, we have documented the impact that passion can have
on a variety of processes and outcomes with harmonious passion
typically leading to more adaptive outcomes than obsessive pas-
Figure 1. The passion and performance model. From Vallerand et al. (in
press, Study 2). Printed with permission, Psychology of Sport and Exercise.
7PSYCHOLOGY OF PASSION
sion. If passion makes a difference in people’s lives, then it would
appear worthwhile to determine how it develops. Three processes
are hypothesised to influence the development of passion toward
an activity: activity selection, activity valuation, and the type of
internalisation of the activity representation in one’s identity. Ac-
tivity selection refers to the person’s preference for the activity
over other activities. To the extent that the person feels that such
selection reflects true choice and interests and is consonant with
one’s identity, it should promote the development of passion
toward that activity. Activity valuation (or the subjective impor-
tance given to the activity by the person) is expected to play an
important role in the internalisation of the activity in identity and
self. Research has indeed shown that when the object of interest is
highly valued and meaningful, one is inclined to internalise the
valued object, to make it part of him or herself (Aron, Aron, &
Smollan, 1992; Deci et al., 1994). The more important (or valued)
the activity, the more the activity will be internalised in the
person’s identity, and, consequently, the more passionate the per-
son will be toward this activity. Thus, activity valuation can be
seen as the intensity (or quantity) dimension (the fuel) underlying
activity internalisation and the development of passion.
It is further proposed that once an interesting activity becomes
highly valued, the type of passion that will ensue is determined by
the type of internalisation that takes place. This last process can be
seen as affecting the “quality” dimension or the type of passion
that will take place. A controlled internalisation of the activity
representation is expected to lead to the development of an obses-
sive passion and an autonomous internalisation to a harmonious
passion. It is further proposed that one important determinant of
the internalisation process is the extent to which the social envi-
ronment promotes the person’s autonomy (Deci & Ryan, 1987)
toward activity selection and activity valuation. Much research has
shown that autonomy support (or promoting choice and self-
initiation of another person’s behaviour) from parents and teachers
facilitates children’s autonomous internalisation of values and
regulations of noninteresting activities such as school (see
Grolnick & Ryan, 1989). Similarly, it is proposed that autonomy
support will facilitate the autonomous internalisation of the inter-
esting activity in one’s identity and self, thereby leading to har-
monious passion. Conversely, once a given activity has been
selected by the individual a controlling behaviour from important
social agents should facilitate a controlled internalisation of the
activity in one’s identity, thereby leading to an obsessive passion.
A recent study by Mageau et al. (2007, Study 3) has tested the
above hypotheses. In this study, first-year high school students
who had never played a musical instrument before and who were
taking their first music class completed a series of questionnaires
early in the term assessing activity selection and valuation (per-
ceived parental activity valuation and both perceived parental and
children activity specialisation), autonomy support from parents
and music teachers, as well as identity processes. By following
participants who were registered in their first music class over the
course of a full semester, we sought to predict who would develop
a passion for music at the end of the term, and, among those who
did, predict those who would develop a harmonious or an obses-
sive passion. Results of a discriminant analysis revealed that
students who ended up being passionate for music (36% of the
sample) at the end of the term had reported higher levels of activity
valuation and specialisation, identity processes, and parental and
teacher autonomy support earlier in the term than those students
who didn’t turn out to be passionate. Furthermore, it was found
that high autonomy support from close adults (parents and music
teachers) and activity valuation from the students’ port were condu-
cive to the development of harmonious passion. However, high levels
of parental perceived valuation for music and lack of autonomy
support were found to predict the development of obsessive passion.
In sum, the results of the Mageau et al. (2007, Study 3) provided
support for the role of activity selection and valuation, identity pro-
cesses, and autonomy support from significant adults in the develop-
ment of a passion for music in general, and harmonious and obsessive
passion in particular.
A second determinant of the internalisation process deals with
one’s personality (see Vallerand & Houlfort, 2003; Vallerand &
Miquelon, 2007; Vallerand et al., 2006). Past research (see Vallerand,
1997, 2001) has shown that an autonomous personality orientation
(having a tendency to do things out of pleasure and/or choice) leads
to the internalisation of noninteresting activities in the self. On the
other hand, having a controlled personality orientation (to do things
out of external and/or inner pressure) leads to the pressured internal-
isation of noninteresting activities in the person (Guay, Mageau, &
Vallerand, 2003; see also Vallerand, 1997, 2007). It thus appears that
an autonomous personality leads to an autonomous internalisation
style while a controlled personality facilitates a controlled internalisa-
tion style. In light of the above, to the extent that one highly values an
enjoyable activity, people with an autonomous personality should be
more likely to internalise the activity in their identity in an autono-
mous fashion (i.e., willingly, without any external or internal pres-
sure) thereby leading to harmonious passion. Similarly, a controlled
personality should be more conducive to the internalisation of an
enjoyable and valued activity in a controlled manner, thus leading to
obsessive passion.
Recent research by Vallerand et al. (2006, Studies 1 and 3)
conducted with athletes supported the above hypotheses. For in-
stance, Vallerand et al. (2006, Study 1) showed that results from a
path analysis revealed that valuation of the sport activity coupled
with an autonomous internalisation style (as assessed by the
Global Motivation Scale; Guay, Mageau, & Vallerand, 2003)
predicted harmonious passion. On the other hand, a controlled
internalisation style coupled with valuation of the activity led to an
obsessive passion. These findings appear in Figure 2. Furthermore,
these findings were replicated in a second study (Vallerand et al.,
2006, Study 3) using a short longitudinal design. Thus, personality
factors do play an important role in the prediction of the type of
passion that will develop.
Overall, results presented in this section provide support for our
position on the development of passion. Additional research is
needed in order to replicate these findings over a longer period of
time and to determine more clearly the psychological processes
leading to the development of passion toward a new activity.
Future Research
So far, in this article, we have reviewed several studies that
provide strong support for the proposed approach on passion with
respect to a number of important variables and with a variety of
activities and populations. While the research conducted to date is
indeed encouraging, additional research is necessary in order to
probe further the role of passion in people’s lives. Below, we
8 VALLERAND
present certain directions for future research that would appear
particularly exciting.
On the Stages of Passion
A first area of research that would appear important deals with
the potential existence of stages of passion toward a given activity.
Is there a universal sequence such that initially the passion toward
a given activity is, for instance, obsessive, and then later on
becomes more harmonious? In other words, do the two types of
passion follow some stages? We do not have any conclusive
empirical evidence on this issue so far. Preliminary evidence
seems to indicate that such is not the case. In the Vallerand et al.
(2003, Study 1), we did not find any relationship between length of
involvement and the two types of passion. However, these findings
are limited in scope as the design used was not longitudinal in
nature. What is needed is to follow individuals from their begin-
nings in a given activity up to their decision to become heavily
involved in that activity. In other words, a follow up to the Mageau
et al. (2007) study discussed previously would appear in order.
Such a study would allow us to determine if a sequence exists and
if it changes over time.
On Passion, Competence, and Future Selves
A second fruitful area for future research deals with the potential
role that perceptions of competence and progression toward the
activity may have on the development of the ensuing passion.
Among other things, we need to assess whether perceptions of
competence and those of progression lead to the same impact on
passion. Indeed, people may see themselves as currently compe-
tent but not as progressing enough to reach the highest levels of
performance. Such perceptions might curtail the development of
one’s passion because people might feel that feelings of compe-
tence and excitement will be no longer forthcoming at some point.
However, individuals who perceive themselves as competent and
who expect a bright future in the passionate activity might be more
likely to increase or at least maintain their passion for the activity.
Furthermore, perceiving oneself as progressing and eventually
reaching a certain level of proficiency cannot only allow one to
satisfy the basic need for competence (Deci & Ryan, 2000) but also
lead one to see the activity as representing a meaningful “future self”
(Markus & Nurius, 1986), and eventually as contributing to one’s
identity. Since identity construction plays an important role in the
development of a passion toward a given activity (Mageau et al.,
2007), self-perceptions of progression may indirectly facilitate the
evolution of one’s passion through its impact on the identity construc-
tion process. Future research is needed on this issue.
On Long-Term Involvement and Psychological Adjustment
Another important research issue pertains to long-term involve-
ment in a given activity and the psychological adjustment that may
ensue. Research reveals that it may take as much as 10 years of
sustained and intense training in a given field to reach the highest
levels of one’s discipline such as music or sports (Ericsson &
Charness, 1994). What type of passion may best allow one to
sustain such long-term involvement and reach such high levels of
performance? What is the long-term impact on one’s psychologi-
cal adjustment to persist for such a long time in environments that
may be particularly competitive? Our research so far suggests that
there are two roads to excellence, the harmonious and the obses-
sive ones. While both may lead to high-level of performance, the
obsessive road may lead individuals to suffer psychologically
along the way, relative to harmoniously passionate performers.
Research in this area tends to support this hypothesis (Bonneville-
Roussy & Vallerand, 2007; Vallerand et al., in press, Study 2;
Vallerand, Salvy, et al., 2007). Thus, these findings would appear
rather robust, even with high-level performers. However, such
research was not longitudinal. Clearly future research is needed in
order to replicate the above findings over a sustained period of
time.
One issue raised by the above findings pertains to the possibility
that obsessive passion leads participants to experience lower levels
of psychological adjustment and thus to eventually drop out before
reaching the highest levels of performance of the activity. How-
ever, recent research by Amiot, Vallerand, and Blanchard (2006)
with teen elite hockey players suggests that this may not neces-
sarily be the case. Results from their study revealed that obses-
sively passionate individuals were found to psychologically suffer
only in nonobsessive environments (less competitive leagues)
wherein task engagement was more limited. Obsessively passion-
ate individuals who were in obsessive environments (the more
competitive leagues) where they could devote themselves fully to
the passionate activity (hockey) fared much better psychologically
than harmoniously passionate hockey players in a similar environ-
ment. Thus, it is unclear at the moment whether obsessively
passionate people would indeed be at risk psychologically when
going up the ladder of excellence and if they would be more likely
to drop out of the activity. Based on past research on rigid
persistence (e.g., Vallerand et al., 2003, Studies 3 and 4), it is also
possible that they may stick with the activity no matter what
allowing them to be among the fortunate few who make it to the
top. Clearly, future longitudinal research is needed to address the
issue of long-term persistence and high-level performance and
psychological adjustment.
Figure 2. The role of valuation and personality processes in the predic-
tion of passion. From Vallerand et al. (2006, Study 1). Printed with
permission, Journal of Sport & Exercise Psychology.
9PSYCHOLOGY OF PASSION
On Passion for a Cause
For a number of people, the passionate activity they engage in
is related to a cause or an aspiration. For instance, the Red Cross
and Doctors Without Borders are organisations that have human-
itarian interests at heart. People who espouse such goals and values
often have a passion for the cause promoted by such organisations.
What are the consequences both for individuals and organisations
of having passionate people involved in such activities? Who is
more likely to take chances with one’s lives as well as that of
others under extreme conditions in order to reach objectives re-
lated to the cause? Who is more likely to come back with post-
traumatic stress experiences following such missions? Based on
the findings presented in this article, one would be inclined to
predict that obsessive passion would be more likely to lead to such
negative outcomes. Furthermore, because the cause is so important
for the individual, passion for the cause may lead at times to
extreme behaviours in order to reach the cause. This would be
expected to be the case particularly for obsessive passion at least
in part because the passionate activity then provides individuals a
high sense of self-esteem when they reach their goal (Mageau &
Vallerand, 2007). For instance, in a recent study with highly
politically involved individuals, Rip, Vallerand, Grenier, Lafre-
nie`re, and Charbonneau (2007) found that harmonious passion was
positively related to acceptable behaviours to reach the cause (e.g.,
participating in discussion groups to persuade people of the im-
portance of the political cause) but not to extreme ones (e.g., being
part of subversive groups to reach the political cause). The exact
opposite findings were obtained with obsessive passion where
strong positive relationships were found between obsessive pas-
sion and extreme forms of political behaviour. Because obses-
sively passionate individuals’ sense of self-esteem appears to be
contingent on reaching the cause (Mageau & Vallerand, 2007),
failure to reach such cause may not be an option for them. In the
end, the end justifies the means, and obsessively passionate indi-
viduals may engage in extreme behaviours in order to reach their
goal. Future research on these issues may have important impli-
cations for a host of areas, including politics and activism.
Conclusion
The purpose of the present article was to present the Dualistic
Model of Passion and review some of the related research. Passion
is defined as a strong inclination toward self-defining activity that
one loves, finds important, and devotes significant amount of time
and energy. Furthermore, two types of passion are proposed de-
pending on how the activity representation has been internalised in
one’s identity. While harmonious passion entails control of the
activity and an harmonious coexistence of the passionate activity
with other activities in identity, obsessive passion entails the
relative lack of control over the passionate activity, rigid persis-
tence, and conflict with other activities in one’s life. Strong sup-
port for the model was obtained. Specifically, a passion scale was
developed, assessing harmonious and obsessive passion. In addi-
tion, harmonious passion was found to promote more adaptive
outcomes than obsessive passion on a number of cognitive, affec-
tive, behavioural, interpersonal, and performance outcomes, on a
variety of activities and with various populations ranging from
children to the elderly. Finally, the processes through which pas-
sion develops were also discussed and directions for future re-
search were also proposed.
In concluding, two caveats are in order. First, much of the
research presented herein is correlational in nature. Thus, one
should not at this point make conclusive statements about causal-
ity. However, the evidence is highly consistent, always pointing in
the same direction (harmonious passion being positively correlated
with adaptive outcomes and obsessive passion with less adaptive
ones). Furthermore, research by Carbonneau, Vallerand, Fernet,
and Guay (in press) using a cross-lagged panel design showed that
while passion leads to changes in outcomes (e.g., work satisfac-
tion) over time, outcomes do not lead to changes in passion. This
last finding would tend to support the view that passion can
produce outcomes. A second caveat is that we would not want to
portray obsessive passion as being completely negative. While it
may not lead to outcomes as adaptive as those derived from
harmonious passion, obsessive passion is still more adaptive than
being nonmotivated or amotivated (Deci & Ryan, 2000). Indeed,
among other things, it promotes long-term commitment and persis-
tence in the passionate activity. Furthermore, future research may
show that it facilitates other kinds of positive outcomes. Thus, future
research is needed to place the passion construct in a broader moti-
vational framework, allowing researchers to more firmly determine
the positivity or negativity of the outcomes engendered by the two
types of passion relative to other motivational constructs.
In sum, positive psychology (Seligman & Csikscentmihalyi,
2000) has recently asked the question “How people’s lives can be
most worth living?” It is believed that one answer to that question
is by having in one’s life a harmonious passion toward a mean-
ingful activity or a cause. A prime example would be the harmo-
nious passion toward the cause of Canadian Psychology displayed
by colleagues on the CPA board of directors. I believe that such a
passion helps explain their intense involvement toward the cause
and why they have done so well advancing it. No wonder that with
such harmoniously passionate people my involvement in CPA was
so positive and memorable.