|
"I
never thought of losing, but now that it's happened, the only thing is
to do it right. That's my obligation to all the people who believe in
me. We all have to take defeats in life." (Muhammad Ali, 1973; emphasis added).
Combat
sport is unique insofar as it is characterised by two combatants in a
fixed environment (Lane, 2006). Although there is risk of injury (a physical beating)
associated with participating in combat sports (Bledsoe et al., 2006) this review is directed toward the psychological "beating"
that participants may direct towards themselves in the aftermath of a
bout. We use the term "psychological beating" colloquially to
signify the maladaptive thoughts that may occupy athletes' reflections
in the aftermath of competition, typically although not exclusively associated
with losing. As the above quote by Ali hints at, individual athletes may
manage the adversity associated with a defeat in a more or less constructive
manner. Indeed, how athletes respond to both adversity and success is
one component of a framework of mental toughness proposed by Connaughton
and Hanton, 2009. Although mental toughness is often considered an important
psychological characteristic contributing to performance excellence, there
is little understanding of how mental toughness is acquired and developed
(Connaughton et al., 2008a; 2008b; Jones et al., 2002). Similarly, there is surprisingly little research directed
toward the way in which athletes generally, and combat sport participants
particularly, "manage" defeat. Although sport behaviour occurs
in the context of time there is something of an asymmetry associated with
the examination of participants' psychological state - research being
directed primarily toward athletes' pre-event (prospective) cognitions
and feelings as opposed to their post-event (retrospective) thoughts and
affect (e.g., Cerin et al., 2000). Phrased more precisely it is not so much about when
the examination of athletes' thoughts occur, more about the temporal perspective.
For example, in the week leading up to a fight, a judoka may have thoughts
about the forthcoming competition alongside thoughts directed toward a
past defeat. Retrospective cognitions include both prolonged, recurrent
and repetitive reflections about one's self, one's concerns, and one's
experiences (Harvey et al., 2004; Watkins, 2008), as well as reflections of a more immediate and intuitive
nature (e.g., Vallerand, 1983). Such retrospective cognitions form an important aspect
of athletes' "mental time travel" (Tulving, 2002); they are common to the experience of many sport participants
(e.g., Thelwell et al., 2007) including those engaged in combat sports (Devonport,
2006).
Intuitively there is some ambivalence about how to respond to losses in
sport. On the back of a resounding defeat, reflecting on what's happened
may appear somewhat fruitless insofar as it changes nothing about the
outcome. Such a stance is reflected to some degree in the concept of mindfulness
(Bishop et al., 2004), an approach that, at least in part (Gardner and Moore,
2007), advocates a somewhat detached, objective and non-evaluative
reference in relation to competitive outcomes. This stands in contrast
to the assumption that there can sometimes be a good deal to be gained
about reflecting on past defeats, as intimated by Devonport, 2006. However, close inspection of the theoretical and empirical
foundation for this assertion reveals a paucity of literature. Accordingly,
the primary aim of this paper is intended to provide a stimulus for future
theorising and research to be directed towards the retrospective cognitions
that athletes may exhibit about competitive events. To accomplish this
objective we firstly define each of four types of retrospective cognition
namely: (re)appraisal (Vallerand, 1987), attribution (Allen et al., 2009), rumination (Maxwell, 2004) and counterfactual thinking (Dray and Uphill, 2009). In the second section we provide a "caricature"
of this literature, highlighting the principle features of the content
and consequences of each type of cognition. In the final section we summarise
the observations and conclude with some directions to move the field forward.
Characteristics
of appraisal, attributions, rumination and counterfactual thinking
This section provides a succinct overview of the central characteristics
of each of the four types of cognition. Appraisals are in essence evaluations
about the significance of events for individuals' well-being (e.g., Lazarus,
1991). As elaborated upon below, appraisals comprise assessments
broadly about what is at stake for an individual in an encounter and his/her
ability to cope with the demands of a situation (cf. Lazarus, 1991; 1999). While appraisals are concerned with evaluating the significance
of what's happened, attributions relate to individuals' attempts to explain
the occurrence of events (Anderson et al., 1996; Rees et al., 2005). Two boxers in the semi-final of the Olympics who both
experience a technical knockout might appraise the loss very differently
(e.g., "I've done well to get this far" versus "I've really
let my family and teammates down"), yet attribute their loss in a
similar way (e.g., "I wasn't good enough on the day"). Both
appraisals and attributions are prototypically directed to what has happened
in the past. In contrast, counterfactual thinking describes the process
of imagining how things could, would and perhaps should have turned out
differently in the past (Roese and Olson, 1995). Arguably counterfactual thinking shares some overlap
with appraisal insofar as repeated iterations of appraisal (i.e., reappraisal)
may alter the significance of an event for an individual (e.g., Jones,
2003), just as imagining how things may have turned out better
or worse (e.g., Medvec et al., 1995, and see below) influences athletes' reactions to success
and defeat (Dray and Uphill, 2009). Indeed counterfactual thinking could conceivably be
viewed as a type of reappraisal. However, reappraisal need not involve
counterfactual thinking. Altering the emotional significance of an event
may include a variety of cognitions (e.g., perspective taking) that might
not involve a consideration of how things could have turned out differently.
A wrestler may alter the significance of a defeat by considering it was
his first fight at this weight division (and by implication there is hope
for improvement, rather than disappointment at defeat), as opposed to
how things may have turned out differently.
Rumination involves repetitive, intrusive and negative cognitions (Papageorgiou
and Wells, 2004). Although rumination has typically been examined in relation
to depression (e.g., Nolen-Hoeksma, 1991), models have posited theoretical relations between ruminative
thinking and sadness (Conway et al., 2000), stress-reactivity (Alloy et al., 2000), and post-event processing (post-mortem thinking) in
relation to social interactions (Clark and Wells, 1995) and anger (e.g., Maxwell, 2004). This broadening of the literature, suggests that in
comparison to the content of thoughts defining rumination, it is the process
of thinking preserveratively (see also Brosschot et al., 2006), about ones feelings and problems that characterise rumination.
Spellman and Mandel, 1999 highlighted that rumination may be associated with counterfactual
thinking, although counterfactual thinking need not involve repetition,
a fundamental aspect of rumination (Sukhodolosky et al., 2001).
Even this cursory glance at the literature illustrates commonality between
the operational definitions of appraisal, attribution, rumination and
counterfactual thinking. These constructs have emerged as part of distinct
research traditions with different precursors and foci. We believe it
is desirable to consider these constructs collectively for at least two
reasons. Firstly, the appropriateness of distinguishing between constructs
versus aggregating constructs with seemingly similar features can be traced
to the earliest stages of scientific psychology (cf., Judge et al., 2002). In presenting a position that counteracts the prevailing
trend of considering these constructs in relative isolation, we hope to
prompt theoretical debate and research as a strategy to strengthen enquiry
in athletes' post-competition cognitions generally.
Secondly, there is some overlap in the antecedents of the respective constructs.
Although attributions, counterfactual thinking, appraisals and rumination
can all occur in response to positive outcomes (e.g., Dray and Uphill,
2009; Ellsworth and Smith, 1988; Smith and Alloy, 2009; Weiner, 1985), it is typically the case that the propensity for these
types of thought are prompted in relation to negative affect (cf. Taylor,
1991). In practice then, if not yet in theory, there remains
a strong likelihood that these thoughts are intimately related and involved
in reciprocal relations, as alluded to above. For practitioners to assist
athletes in constructively reflecting on performance it is desirable to
understand more about the types of post-event retrospective cognitions
exhibited, and their consequences. It is to this end that attention is
now turned.
A
caricature of the literature
Given the constraints of this paper and the availability of comprehensive
reviews of the literature elsewhere (e.g., Biddle et al., 2001;
Mandel et al., 2005; Rees et al., 2005; Smith and Alloy, 2009; Vallerand and Blanchard, 2000) we paint something of a caricature of the extant literature.
Drawing upon the description of caricatures by Sparkes, 1992, it is inevitable that subtleties and areas of debate
will be obscured in presenting what is admittedly an oversimplification
of the nuances that exist. Moreover the intention is not to cartoon the
research or ideas of others. Rather, in recognising and highlighting those
areas that characterise the prevailing trends of research in each of these
four areas, we utilise this caricature to highlight the central and recognisable
features to provide a foundation for the thesis that follows.
Appraisal
The content and outcomes of appraisals are described by a number of appraisal
theories which although differing slightly in their detail, typically
describe the role of appraisal (or evaluations) in eliciting emotions.
The basic premise of appraisal theories generally is that emotions appear
to be related to how people evaluate events in their lives. More specifically
appraisal theories posit that (a) the meaning of a situation or event
influences that individual's emotional reaction and (b) the meaning that
an individual ascribes to a situation or event can be regarded as a composite
of individual appraisal components (Bennett et al., 2003). Where the theories differ is in the specification of
the appraisal components or dimensions.
Roseman and colleagues (e.g., 1984, 1990) suggested that (a) motivational state (extent to which
individuals were motivated to avoid punishment versus motivated toward
reward), (b) situational state (motive-consistent, versus motive- inconsistent),
(c) probability (occurrence of an outcome is certain or uncertain), (d)
power (extent to which the individual is appraised as strong or weak),
(e) agency (whether the event is caused by something impersonal, another
person, or the self) and (f) legitimacy (whether the outcome is deserved)
were implicated in the elicitation of particular discrete emotions (e.g.,
anger, joy, relief, hope, regret). Sharing some overlap, Smith and Ellsworth,
1985 proposed eight appraisal dimensions considered important
in differentiating between discrete emotions: pleasantness, anticipated
effort, attentional activity, certainty, responsibility, control, legitimacy
and perceived obstacle. In a later study Ellsworth and Smith, 1988 observed appraisals of human agency or situational control
were associated with levels of anger, sadness and guilt; however certainty
and attention did not predict differences between emotions.
In sport, Lazarus' cognitive motivational relational (CMR) theory (2000)
has been the model of choice for researchers investigating relations between
appraisal and emotion (e.g., Skinner and Brewer, 2004; Uphill and Jones, 2007). Briefly stated, Lazarus posited two classes of appraisal:
primary and secondary. Primary appraisals relate to whether an event is
relevant to the athlete and consists of three judgements: goal relevance,
goal congruence and type of ego involvement. Goal relevance concerns an
assessment of whether the individual has something important at stake
in an encounter. A judoka who experiences a defeat against an opponent
in training may not become upset or angry if the defeat is not perceived
to matter (e.g., "it's just training"). Goal congruence describes
the extent to which an event or situation impedes or facilitates the attainment
of a goal. For example, a martial artist who receives a "bye"
in the first round because of participant illness could evaluate that
occurrence in a negative manner ("I won't be suitably warmed up")
versus a more positive manner ("I'll be fresh for my next bout").
Finally, type of ego-involvement addresses the type of goal that is invested
in a particular encounter. A boxer may want to win the fight, accomplish
the win in style via a knockout in an early round, and avoid getting hurt.
Depending on the extent to which these goals are perceived to have been
attained, will influence the boxer's post-match response.
Secondary appraisals are broadly concerned with an athlete's perceived
coping options and comprise evaluations about blame/credit, ability to
cope, and future expectations. Briefly stated, a boxer who receives a
cut around the eye and considers himself to blame, is having difficulty
against a "rangy" opponent, and suspects things are likely to
get worse, rather than better, would be hypothesised to feel differently
in comparison to the boxer who appraises the cut eye as a creditable punch
by his opponent, has the defensive technique to avoid further injury,
and believes that he is ahead on points. While still in its infancy, research
has provided some support for the applicability of components of Lazarus'
CMR theory to sport broadly (Skinner and Brewer, 2004; Uphill and Jones, 2007).
Appraisal approaches have been subject to a number of criticisms. Critics
have questioned the likelihood that (a) elaborate evaluations about the
significance of an event are formed in the few milliseconds it can take
for an emotion to arise (Scherer, 1999), and (b) whether cognition is a necessary precursor to
the elicitation of emotion (Zajonc, 1984). However, if one subscribes to the view that emotions
are not a product of either the person or situation, but as a combination
of the two, then cognition plays a significant role in the aetiology of
emotion (cf., Bennett and Lowe, 2008). Moreover, the debate about the involvement of cognition
in the generation of emotion has become more semantic than substantive,
with the answer to the question of whether appraisals are involved in
emotion, dependent upon how appraisal and emotion are defined (e.g., Lazarus,
1999;
Scherer, 1999).
In summary, not only do appraisal theories represent parsimonious accounts
of inter- and intra-individual differences in emotion experience, specific
appraisal components associated with Lazarus' CMR theory have been implicated
in athletes' emotions (Skinner and Brewer, 2004; Uphill and Jones, 2007). Although appraisals are typically utilised to explain
individuals' emotions, understanding how assessments of goal incongruence,
or appraisals of whether things are likely to get better/worse for example,
relate to individuals' self-efficacy and motivation would represent a
useful extension.
Attributions
One assumption of the majority of attribution research is that the explanations
individuals give for events are underpinned by a dimensional structure
(e.g., Abramson et al., 1978; Weiner, 1985). Moreover, by categorising those explanations into dimensions,
one can better understand those explanations (Rees et al., 2005). Five principle dimensions have been proposed (see Biddle
et al., 2001;
Rees et al., 2005). Dimensions include controllability (those causes that
are affected by the individual or not affected by the individual), locus
of causality (causes perceived as residing within or without the individual),
stability (causes perceived as being stable or transient over time), intentionality
(causes deemed to be either deliberate or accidental), universality (extent
to which the cause is perceived to be common among others, or specific
to the individual) and globality (causes deemed to be perceived as localised
or occurring across many situations). As Biddle et al., 2001 observe, these dimensions have been accepted largely uncritically
by researchers, and have not been studied extensively. For example, although
the dimensions of locus of causality and controllability can be separated,
(genetics may be considered internal but not controllable), there is often
some overlap between where a cause lies and by whom it is controlled (Ingledew
et al., 1996; Rees et al., 2005).
When assessing individuals' attributions then, one can examine the attributional
elements (the specific reasons individuals give) or attributional dimensions
that the responses may represent (see Biddle et al., 2001 for a review; Biddle and Hanrahan, 1998). There are difficulties associated with either of these
stances. With regard to the former, it is sometimes difficult to be sure
what an athlete means by certain words or phrases. A Taekwondo participant
may explain a defeat because "an opponent played better". It
is unclear whether this example represents an unstable or stable attribution.
In the latter case problems can arise when researchers attempt to categorise
attributions along dimensions or otherwise assume the dimensional categories
of attributions (Russell, 1982). For example, Weiner, 1986 indicated that effort could be conceived as stable or
unstable, and luck as either internal or external to the person. Accordingly,
it is generally recommended that participants rate their attributions
along causal dimensions to avoid these interpretational ambiguities.
Importantly how one explains the (perceived) causes of events along these
various dimensions has implications for emotional reactions (Weiner, 1986), expectancies (e.g., self-efficacy, Bandura, 1977), and motivation (Biddle et al., 2001). With regard to emotional reactions, Weiner et al., 1978 differentiated between outcome-dependent emotion (e.g.,
generally being pleased or happy with the outcome) and attribution-dependent
emotion, which related to the perceived cause or reasons for the outcome.
A similar dichotomy was proposed by Vallerand, 1987. In a summary of the literature on attribution-dependent
emotions, Biddle et al., 2001 suggest (a) self-esteem emotions (e.g., pride) are associated
with an internal causality dimension, (b) emotions related to expectancy
(e.g., hope) are associated with the stability of attributions, (c) social
emotions (e.g., guilt) are related to the controllability of the outcome.
With regard to expectancies, Weiner, 1986 suggested that the stability of the attribution is central
in determining changes in expectancy, a statement, or in his terms "law"
that has three corollaries: (a) if the outcome of an event is ascribed
to a stable cause then that outcome will be anticipated with increased
expectancy in the future, (b) if the outcome is ascribed to an unstable
cause then the expectancy of that outcome may be unchanged or be different
from the past, and (c) outcomes ascribed to stable causes will be anticipated
to be repeated with a greater degree of certainty than outcomes ascribed
to unstable causes.
However, Grove and Pargman, 1986 observed that effort (a relatively unstable attribution)
led to the highest expectancy in both success and failure conditions.
A "weak" interpretation of this finding is that stability is
not the only attribution dimension implicated in changes in expectancy.
A "stronger" interpretation is that controllability rather than
stability is important in influencing expectations.
Rees et al., 2005 highlight that the unpredictability of competitive sport,
coupled with changing interactions with opponents and the environment,
affords a focus on controllability rather than stability. Specifically
Rees et al. advocate examining the interactive effects of attribution
dimensions. For instance, a recent study by Coffee and Rees (2008) found that for individuals who perceived performance
as less successful, had higher self-efficacy when they viewed the causes
of performance as controllable and when the causes were perceived to be
global. Higher levels of controllability were associated with higher levels
of self-efficacy.
Although it may sometimes be considered a subtle distinction, attributions
to controllable factors, in contrast to factors that have an internal
locus of control have been recommended (Rees et al., 2005).
In summary, literature on attributions suggests that how one explains
the causes of events can influence our emotion reactions, expectations
of future success and motivation. Whereas literature on appraisal has
critically considered the extent to which appraisals are unconscious/automatic
versus deliberate/conscious, literature on attribution has been relatively
more concerned with assessment of the construct, generating a number of
instruments (cf. Allen et al., 2009), and recognised the importance of athletes' perspective
on assessing these constructs.
Counterfactual
thinking
Counterfactuals in the broadest sense are thoughts or statements that
include at least some premises believed to be contrary to fact (Mandel
et al., 2005). Indeed, logicians (e.g., Goodman, 1947;
Lewis, 1973) have been particularly interested in how knowledge could
be derived from false conditional premises (Mandel et al., 2005). Psychologists by contrast, have typically focused on
how things could, would and perhaps should have turned out differently
in the past (Roese and Olson, 1995).
In attempts to understand characteristics of individuals' counterfactual
thinking researchers have typically examined the "direction"
(e.g., Roese, 1994), "structure", (Markman et al., 2007) and "content" (Mandel, 2003) of counterfactual thoughts. Direction of counterfactual
thinking refers to the tendency to imagine how things could have turned
out better (upward counterfactual thinking) compared to imagining how
things could have turned out worse (downward counterfactual thinking).
A judoka could conceivably simulate how a loss may have been averted "if
only she had not been injured in the week leading up to the fight"
(i.e., an upward counterfactual). Although historically upward and downward
counterfactual thinking was associated with an emotional contrast effect
(e.g., Roese, 1994), that is imagining how things could have been better
is associated with feeling worse (and vice versa), more recent literature
has suggested that upward and downward counterfactuals can elicit a contrast
and assimilation effect (McMullen and Markman, 2002). An assimilation effect occurs when judgements are pulled
toward the counterfactual comparison (McMullen and Markman, 2002). For instance, a boxer who is awarded victory by a split
decision may suggest a downward counterfactual ("that was too close
for comfort") and exhibit negative affect. Extending this example,
the defeated opponent may suggest an upward counterfactual ("you
almost did it") and experience positive affect. One explanation for
why assimilation and contrast effects occur is the mode of thinking individuals
engage in with reflection considered to elicit an assimilation effect
and evaluation (or comparison) evoking a contrast effect (Markman and
McMullen, 2003).
The structure of counterfactual thought refers to the addition of something
that did not happen in the past versus removing something that did. For
example, an additive counterfactual could involve a wrestler imagining
how the addition of a change in stance might have contributed to a critical
throw; in contrast a boxer might rue (and remove) a punch which left him
vulnerable to a counter-attack. Counterfactual content refers to the "target"
- whether in imagining how things could have turned out differently, the
individual changes something about themselves, others or the environment
(e.g., Dray and Uphill, 2009).
Similar to literature on appraisals and attributions, it is differences
in direction, structure and content of counterfactuals that influences
affect (e.g., Markman et al., 2007) emotions (Niedenthal et al., 1994), self-efficacy (Tal-Or et al., 2004), and persistence (e.g., Markman et al., 2008). One limitation of the majority of the literature on
counterfactual thinking has been the use of vignette studies in which
participants rate how things might have been different in response to
hypothetical scenarios (cf. Rye et al., 2008). Indeed as has been contended in relation to attributions
(Faulkner and Finlay, 2005), there is some doubt about whether counterfactuals generated
in the laboratory approximate those that are exhibited in a more ecologically
valid context.
Rumination
On the one hand, Martin and Tesser, 1996 have proposed that ruminative thoughts are instigated
by a discrepancy between one's current position and desired goal, and
by reducing perceived discrepancies within individuals' lives, rumination
is adaptive. On the other hand Nolen-Hoeksma (1996) has conceptualised rumination as maladaptive insofar
as it involves thoughts directed towards ones negative affective states,
rather than toward resolving problems or reducing goal discrepancy (see
also Wade et al., 2008).
Rumination has been considered as either a response style (Nolen-Hoeksma,
1996) or as a state (Wade et al., 2008). Thus, while some individuals may exhibit a propensity
to ruminate, rumination may also differ across situations in response
to varying antecedents. A number of questionnaires operationalising the
measurement of rumination are evident in the literature (Smith and Alloy,
2009). In a review of ruminative thinking, Smith and Alloy,
2009 indicate that factor analyses of rumination measures suggest
evidence for a dichotomisation of repetitive thinking about the self:
broadly speaking harmful or helpful or brooding- and reflective-type rumination
(e.g., Treynor et al., 2003).
Predictions about what individuals think about when they ruminate differs
between theories (cf. Smith and Alloy, 2009). Some authors suggest that rumination is directed toward
the negative feeling states and/or the circumstances associated with that
emotion (Nolen-Hoeksma, 1996; Trapnell and Campbell, 1999). Other authors contend that rumination focuses on discrepancies
between one's current and desired status (e.g., Martin and Tesser, 1996). Finally, other models contend that it is negative themes
of uncontrollability and harm in metacognitions that are important (Smith
and Alloy, 2009). Smith and Alloy recognise however that few studies have
been directed toward the analysis of the content of ruminative thought.
Indeed, although a high number of causal words in written accounts of
rumination has been reported (indicative of a search for the antecedents
of current distress, Watkins, 2004), one cannot be sure that written accounts, mirror cognitive
content (Smith and Alloy, 2009).
In sport there is a paucity of research examining rumination and because
little is known about whether it is a type of thinking that athletes engage
in, the function of such a cognitive strategy is somewhat speculative.
There is literature to suggest that rumination is related to emotional
avoidance strategies (e.g., alcohol misuse). Further, individuals who
engage in post-event processing tend to avoid social situations that are
similar to the one that initiated rumination (Mellings and Alden, 2000;
Rachman et al., 2000). Extrapolating from these observations, Smith and Alloy,
2009 contend that rumination may impede more adaptive experiencing
of negative affect.
In summary, literature on rumination highlights the frequency and repetitiveness
of thoughts as being implicated in both adaptive (e.g., problem-solving)
and maladaptive (e.g., depression) outcomes. From this perspective it
is not just the content of individuals' thoughts that are important in
facilitating an understanding of the constructive reflection on defeat
but the manner in which these thoughts occur (frequency, persistence etc).
|
|
and Katie Dray