MEMORY, 2004, 12 (4), 507516

A pilot exploration of the use of compassionate images in

a group of self-critical people
Paul Gilbert and Chris Irons
Kingsway Hospital, Derby, UK
Self-criticism has long been associated with a variety of psychological problems and is often a key focus for
intervention in psychotherapy. Recent work has suggested that self-critics have underelaborated and
underdeveloped capacities for compassionate self-soothing and warmth. This pilot study developed a
diary for monitoring self-attacking and self-soothing thoughts and images. It also explored the personal
experiences of a group of volunteer self-critics from the local depression support group who were given
training in self-soothing and self-compassion. Although using small numbers, this study suggests the
potential value of developing more complex methodologies for studying the capacity for self-compassion,
interventions to increase self-compassion (including imagery techniques), and their effects on mental
health.

Self-criticism and shame have been proposed to

play a key role in anger (Tangney & Dearing
2002), social anxiety (Cox, Rector, Bagby, Swinson, Levitt, & Joffe, 2000), mood disorder (Blatt &
Zuroff, 1992; Gilbert & Miles, 2000; Gilbert,
Clarke, Hempel, Miles, & Irons, 2004b), suicide
(Blatt, 1995), alcoholism (Potter-Efron, 2002),
post-traumatic stress disorder (Brewin, 2003),
psychotic voice hearing (Gilbert et al., 2001),
affect regulation and personality disorders (Linehan, 1993), and interpersonal difficulties (Zuroff,
Moskowitz, & Cote 1999). Vulnerability to
shame-based self-criticism is commonly rooted in
feeling memories of the self being rejected, criticised, and shamed (Gilbert 1989, 1998, 2002;
Kaufman, 1989; Tomkins, 1987), and/or abused
(Andrews, 1998). Shame memories can be intrusive (Kaufman, 1989). Reynolds and Brewin
(1999) found that depressed people often have
intrusive memories of being shamed, rejected,
and/or abused. Internalising these experiences can
result in seeing and evaluating the self in the same

way others have; that is as flawed, inferior,

rejectable, and globally self-condemning (Gilbert,
1998, 2002; Tangney & Dearing, 2002). Irons,
Gilbert, Baldwin, Baccus, and Palmer (2004a)
found a significant association between recall of
parents as rejecting and low in warmth, and level
of self-criticism in students.
When self-criticism emerges from a sense of a
shamed self, people can feel beaten down and
depressed by their own self-criticisms (Greenberg,
Elliott, & Foerster, 1990). Indeed, intense selfcriticism has been viewed as a form of internal
harassment that is stressful and undermining of
the self (Gilbert, 2004). Gilbert et al. (2001)
explored self-critical thoughts in depressed people
and malevolent voices in voice hearers, in regard
to their ``critical'' qualities such as anger, intrusiveness, and the ``felt power'' of a criticism/
attack. The study reported here expands on that
methodology by piloting the use of a diary for selfcritical people to monitor and report on the triggers and forms of their daily self-criticisms. A less

Correspondence should be sent to Professor Paul Gilbert FBPsS, Mental Health Research Unit, Kingsway Hospital, Derby DE22
3LZ, UK. Email p.gilbert@derby.ac.uk
We would like to acknowledge the enormous help of the Derby Depression Alliance Self-Help Group for their advice and
participation in this study. We would also like to thank Rakhee Bhundia for her help with collating and analysing the diaries. This
project was supported by NHS Executive funding.

# 2004 Psychology Press Ltd

http://www.tandf.co.uk/journals/pp/09658211.html

DOI:10.1080/09658210444000115

508

GILBERT AND IRONS

explored aspect of shame-based self-condemning

is the degree to which there is a relative inability to
generate caring and self-soothing/reassuring,
thoughts, feelings, and images (Gilbert, 2000a,
2000b). One source of self-soothing and selfreassuring is access to emotionally textured feeling memories of others who have been soothing
and reassuring, e.g., loving attachment figures
(Bowlby, 1969, 1973; Kohut, 1977). Mikulincer,
Gillath, and Shaver (2002) found that threat can
prime access to attachment inner working models
and memories that are used for coping (for a
review see Gillath, Shaver, & Mikulincer, in
press). Irons et al. (2004) found that self-soothing
and reassuring abilities in students were significantly associated with recall of parents as
warm/affectionate and low on rejection.
A key problem for some self-critical people
may be that they do not have access to feeling
memories of being affectionately cared for
(soothed), and their self-care abilities have been
understimulated, underdeveloped, and underelaborated (Gilbert & Irons, in press). Some evidence for this was found in a study by Gilbert et al.
(2004). They used an imagery task to explore how
easy or difficult it was to imagine a self-critical/
attacking part of the self, and a soothing, compassionate, and accepting part of the self. Those
high in self-criticism found it relatively easy to
imagine a self-critical part of self that was
experienced as hostile, powerful, and controlling,
while low self-critics found this imagery task more
difficult. Self-critics found compassionate selfimagery more difficult, while low self-critics found
this relatively easy to do. These data may indicate
that self-critical people have more ready access to
hostile self-to-self thoughts and feelings, and less
automatic and easy access to self-soothing
systems.
There is evidence that high self-critical
depressed people may not improve as much as low
self-critical depressed people in standard cognitive
therapy (Rector, Bagby, Segal, Joffe, & Levitt,
2000). Helping people generate self-compassionate images and focus on feelings of warmth for the
self may therefore be a useful therapeutic endeavour (Gilbert, 2000a; Gilbert & Irons, in press).
Indeed, McKay and Fanning (1992) made selfcompassion central to their cognitive behavioural
approach for building self-esteem. Dialectic
behaviour therapists also recommend developing
compassion for the self (Linehan, 1993). Developing compassion for the self has a long tradition in
Buddhist healing practice (Salzberg, 1995). Self-

compassion differs from self-esteem in that it is

focused on affects of warmth and sympathy
directed at self (Gilbert & Irons, in press; Neff,
2003). In Buddhist practices, developing compassion for self and others can use highly structured
images that are practised repeatedly (Dagsay
Tulku Rinpoche, 2002; Ringu Tulku & Mullen, in
press). The use of images to stimulate brain
pathways for compassion may be powerful. For
example, images have powerful emotional effects
(Hackmann, 1998, in press) and are increasingly
used in fMRI research to explore neurophysiological systems involved in certain kinds of memory,
thoughts, and feelings (e.g., George, Ketter, Parekh, Horwitz, Hercovitch, & Post, 1995; Schwartz
& Begley, 2002). To date, however, no study has
explored how people might generate their own
images of compassion to self, how they may try to
imbue them with certain qualities (e.g., warmth
and acceptance), and whether they find working
this way helpful or difficult.
This pilot study did not aim to focus on the
effects of giving intensive training in compassionate mind work, but to explore steps before that
to see how people experience their self-criticism
on a day-to-day basis, to see what type of compassionate imagery they would be able to generate
for themselves, and to explore whether they
thought ``practising self-compassion'' could help
counteract self-criticism. Clearly, one would have
to think carefully about developing a psychological treatment that patients thought was inappropriate or unlikely to work.
There is increasing recognition that in investigating how patients may experience a disorder,
processes associated with a disorder, or interventions, researchers should seek patient collaborative involvement in guiding and informing the
research. Once patients understand what knowledge is sought they can offer insights from ``the
inside'' (Goodare & Lockwood, 1999). Hence,
given the nature of this research we recruited the
help of a local self-help group for depression. The
aims of this pilot study were:
1. To invite people attending a depression
support group who have problems with self-criticism to take part in a collaborative research project investigating their inner self-critical and selfsoothing processes.
2. To use a diary method to explore the triggers
and forms (e.g., degree of intrusiveness and the
power) of naturally occurring self-criticism, in this
group of people.

COMPASSIONATE IMAGES AND SELF-CRITICISM

3. To explore the ability to generate and use

compassionate imagery, and obtain views of how
helpful this may be for this group of people.
4. To explore the types of images generated
and the experiences of working with compassionate imagery.

METHOD
Participants
A self-help depression group, with whom the
authors have worked closely over a number of
years, was advised of our study at one of their
larger meetings. Those who regarded themselves
as self-critical were invited to take part in this
study exploring self-criticism and the use of compassionate imagery to help reduce it. Of the 18
people at that meeting, most expressed interest
but 9 (2 men and 7 women) were able to take part
and attend four 112 hour evening meetings. However, our data are based on eight participants due
to incomplete data from one person.
All nine participants verbally reported that
they had had at least one diagnosed depressive
episode (diagnosed by a psychiatrist). All were
currently on anti-depressants. All participants
completed the Hospital Anxiety and Depression
Scale (HADS; Zigmond & Snaith, 1983). The
group mean for the depression subscale was 9.00
(SD = 5.1) and for the anxiety subscale was 11.83
(SD = 3.6). For the depression subscale, scores of
< 8 indicate ``non-cases'', 811 doubtful/possible
cases, and scores of 11 or above definite cases.
One person scored 10, and four people scored 11
or greater.
All participants reported that they had had
problems for longer than 10 years or ``most of
their lives''. A number of participants had comorbid difficulties such as social anxiety, agoraphobia, and obsessive-compulsive disorder. Our
group was not pre-selected, other than that they
attended a depression self-help group, they saw
themselves as self-critical, and agreed to participate. Our focus was on self-criticism and imagery
development rather than a specific disorder.

Self-attacking and self-reassuring

diary measure
Diaries were constructed based on previous studies exploring hostile and compassionate selfimagery (Gilbert et al., 2004a, 2004b). They are

509

available on request from the corresponding

author.
We chose an interval contingent format for our
diaries (Wheeler & Reis, 1991), which requires
respondents to record their critical thinking over a
set period of time. In this study, the set period was
initially daily (for 2 weeks) and then weekly.
Participants were asked to write down each day
what situations or events triggered their self-critical thinking. Participants were also asked how
these situations made them think about themselves, and how these thoughts made them feel.
While Wheeler and Reis (1991) suggest that this
method is open to retrospective bias, self-critical
thoughts can often be variable in their frequency
and duration, and can be difficult to measure using
alternative diary methods. In the second section of
the diary, participants were asked to give a
quantitative rating of their critical thoughts and
their ability to self-soothe in these situations, on a
110 interval scale. Ratings for self-criticism were
given on: how often it occurred; how powerful,
intrusive, long-lasting, distressing, and angry it
was; and how difficult it was to distance from. This
gave a possible range of 070. Ratings for selfsoothing were given for how easy was it to: selfreassure, self-comfort, self-support, self-care, and
self-soothe, giving a possible range of 050. As
Ferguson (in press) points out, interval contingent
diaries are useful when the subject being recorded
is frequent, may not have a fixed start/end point,
and may be continuous or sporadic.

Procedure
We arranged to meet with the participants for four
evening sessions. During the first session, we
outlined our interest in exploring with them the
day-to-day nature of self-criticism, and how
learning to be compassionate with the self, and
focusing on compassionate imagery, might help to
counteract self-criticism. The focus was to engage
them as joint partners in this project. All participants agreed to the requirements in the spirit of a
collaborative exercise, and signed consent forms
that they were happy to take part. We agreed to
have three consecutive weekly meetings, with a
follow-up 4 weeks later. All participants were free
to contact us if they had any distress associated
with the procedure. None did, and in ongoing
group discussion thought the process was useful.
Session 1: At our first meeting we discussed the
nature of self-criticism. In open discussion, many

510

GILBERT AND IRONS

participants thought that ``a lot of this comes from

childhood''. Participants were asked to fill in
various self-report questionnaires (not reported
here) and to keep diaries of their self-critical
thinking for the coming week. Instruction was
given in how to use the diaries, with examples.
Session 2: Participants handed in their diaries
and discussed how the previous week had gone.
Following this, the first compassionate mind
imagery exercises were conducted. This was
introduced as part of training our minds to focus
and attend to compassionate processes in the self
(Gilbert & Irons, in press) in the following way:
When we attack ourselves we stimulate certain
pathways in our brain, but when we learn to be
compassionate and supportive of our efforts we
stimulate different pathways. Sometimes we are
so well practised in stimulating inner attacks that
our ability to stimulate inner support and warmth
is rather underdeveloped. What we would like to
do today is see if we can generate some compassionate images and ways of thinking that you can
practise using over the next week and see how this
may help you.

The group then discussed the nature of compassion, the value of compassion for the self, and
key elements of compassion such as empathy,
sympathy, warmth, and self-acceptance. The
group had a discussion about whether developing
these qualities for the self would be helpful, and
the importance of training/practice in trying to
generate these aspects for the self.
Following this, we engaged in imagery work.
First, the group was taken through a short (34
minutes) relaxation process that focused on
breathing and tension release. We then asked
participants to imagine an inner place of safeness,
which would allow them to do this work. They
were then invited to ``focus on an image of compassion that contains the attributes we discussed'';
to ``allow images to come to your mind that capture these qualities''. Following this exercise,
which lasted for about 10 minutes, the researchers
asked each participant to share and discuss their
images with the group. Participants were encouraged to practise their compassionate imagery as
often as possible, and in particular to try to elicit it
when they had self-critical thoughts. At the end of
the session, participants were each given another
diary to record their critical thoughts in the
coming week.
Session 3: Participants handed in their diaries,
and took possession of diaries that could be

completed weekly rather than daily. As the next

meeting was to be a month later, many felt it
unreasonable to keep diaries each day for a
month. We obtained feedback from participants
on their experience of using compassionate mind
techniques. We then took participants through the
same process, again with the relaxation exercise
followed by compassionate imagery practice,
focusing on generating specific qualities of compassion.
Session 4: We met the group for the final time 6
weeks later. We had intended to meet earlier but
holidays and other commitments prevented this.
During this time participants were asked to keep
weekly diaries of their self-critical thinking and
abilities to be compassionate to themselves. Two
participants were unable to come to the final
session due to illness and child-care commitments.
Again, we talked to participants about their
experiences of using compassionate mind imagery, including specific aspects that they found
helpful or difficult. A final form was given, asking
participants five questions: What was the image
that you have used over this research period? How
did the image appear to you? What was the most
difficult aspect? How much time were you able to
practise? How helpful was it using the image? At
the end of this session, participants were thanked
for their time and help, and the researchers
answered questions that were posed.

RESULTS
Triggers and forms of self-critical
thoughts
Table 1 provides exploratory qualitative data
based on the first three diary questions from Week
1 recordings. Two questions focused on types of
thoughts and a third focused on what people felt
as a result of what they thought.
Self-criticism was linked to a multiple array of
activities and social interactions. In particular,
many of the situations that activated critical
thoughts were to do with relationships (including
partner, family member, friends, and colleagues)
and negative comparisons with others. Many selfcritical thoughts were triggered by day-to-day
occurrences, such as ``housework'', ``visiting a
friend'', ``given a gift from a client'', ``at the gym'',
``being awake at 3 am'', and ``having a headache''.
Also of interest is the wide range of critical
thoughts and feelings reported about the self,

COMPASSIONATE IMAGES AND SELF-CRITICISM

511

TABLE 1
Self-critical themes
Question 1:
What situations/events brought
them about?

Question 2:
What sort of things did you
think/feel about yourself?

Question 3:
How did your thoughts about
yourself make you feel?

Family
Visiting/socialising with friend
Given a gift
Waking at 3am
Having headache
Relationships
Work
Housework
Gym/body image

Inadequate
Incompetent
Angry
Frustrated
Negative body images
Unattractive
Lack of control
Irritated
Lack of organisation

Inability to meet required standards

Unhappy
Anxious
Inferior
Lonely
Disliked
Dejected
Failure
Hurting
Weak

Examples of self-critical themes elicited from diaries over week 1 (pre compassionate mind training) ``Looking back over today,
please could you carefully think about any critical thoughts you may have had''.

including anger, frustration, inferiority, and

depression. Some people felt harassed by their
self-criticism: ``it was always there whatever I
did''. More comprehensive data for each participant are available from the authors.

Qualities of self-criticism
Alongside the qualitative diary information, we
also asked participants to give quantitative ratings
of their self-criticism (e.g., its power, intrusiveness, and hostility) and their ability and ease of
self-soothing. Table 2 gives each participant's
scores for baseline depression, self-criticism and
self-soothing, and post compassionate mind
training scores (after 1 week of practice) for selfcriticism and self-soothing. We had hoped to
obtain diary data from the fourth session to see
how compassionate mind training had progressed

over 6 weeks. Unfortunately, all participants had

experienced problems in keeping diaries over this
time (e.g., losing diaries, forgetting to fill them in)
and so this set of data is unreliable. We would
advise researchers to use shorter time periods or
more frequent sampling points.
A paired t-test revealed the small reduction in
scores for self-criticism was non-significant: mean
score baseline = 42.35, (SD = 13.7) mean score
post compassionate mind training = 37.46 (SD =
11.2); t(7) = 1.32, p = .22. One patient who had
been more self-critical in the week (participant 6)
felt this was related to unforeseen life events.

Self-soothing
In regard to self-soothing/compassion, there was a
significant increase in the ease of generating these
images and soothing oneself in a self-critical

TABLE 2
Participant HADS depressions scores pre-training, and mean criticism and compassion scores pre and post compassionate mind
(CM) training
Participant
1
2
3
4
5
6
7
8

HADS depression
subscale score

Criticism
pre CM training

Criticism
post CM training

Compassion
pre CM training

Compassion
post CM training

5
0
16
13
11
10
12
7

46.57
38.18
56.50
64.50
44.86
22.43
34.88
30.91

22.85
39.53
58.25
46.77
39.50
28.73
32.68
31.40

19.25
24.33
5.00
7.00
12.71
29.14
20.00
7.17

30.87
32.50
12.25
20.32
18.84
29.00
20.00
6.40

512

GILBERT AND IRONS

situation: mean score baseline = 15.57 (SD = 9.0),

mean score post compassionate mind training =
21.27 (SD = 9.2); t (7) = 2.94, p = .02.

Experiences of compassionate mind

training
On the final meeting, we obtained data from six
participants in response to questions regarding
their thoughts and feelings about compassionate
mind training. These are given in Table 3.
Participants' images varied greatly. Some individuals focused on personified images, whereas
others did not. Most images were visual. Those
who found compassionate imagery helpful
described their images as having calming, soothing, and caring effects. Those who said they found
it less helpful had found it difficult to bring an
image to mind, hold it in imagination, and practise
it. One participant found her ``compassionate
image'' turning into the stomach of a well-rounded male who reminded her of her ex-husband,
which made the experience unpleasant. Intrusive
negative images, at times linked with memories,
when one is trying to create a positive image can
be distressing. In Buddhist meditation, should this
happen the person is invited to let the image go
and become gently mindful of the compassionate
image again (Dagsay Tulku Rinpoche, 2002;
Ringu Tulku & Mullen, in press). However, this
lady felt that she needed to work through her
anger towards her ex-husband.
We explored participants' images, focusing on
different components of compassion. Sometimes
an image would change with practice. Participants
discussed how different blends of compassion
components (e.g., warmth, acceptance) were more
or less difficult. One participant said that he could
imagine warmth, but not acceptance. This may
have been related to unresolved hostility, and he
seemed to hold a ``Groucho Marx'' belief that ``I
wouldn't want to be a member of a club that
accepted me as a member''. Some participants
noted that it was difficult to hold a compassionate
image, and that feelings of warmth or acceptance
were often ``only fleeting''.

DISCUSSION
This pilot study explored the use of a diary to
monitor typical elicitors of self-criticism and their
qualities, such as their felt power, intrusiveness,
and distressfulness, and builds on earlier work

(Gilbert et al., 2001). Participants felt their selfcriticisms were automatic, powerful, intrusive,
distressing, and difficult to distract from (Table 1).
Participants felt able to keep the diaries, and
found them revealing of just how much they did
self-criticise. They suggested that diaries like this
could be useful in helping people monitor their
self-critical thoughts, although participants may
not have been able to discriminate the various
qualities of self-soothing and this requires further
study.
A second key question concerned how easy or
difficult it is for people to learn to generate and
use compassionate feelings and images for the self.
One participant found that her compassionate
image changed into something unpleasant and she
could not hold a ``nice'' image in mind. Another
felt that images were difficult to generate or
engage with. However, the other six participants
felt they had benefited from their efforts, and two
participants felt it had been a ``great'' help,
although all thought they needed more help and
support to practice, and more work as a group.
We found that there was a significant
improvement in the reported ability to self-sooth.
One cannot attribute this necessarily to the compassionate mind imagery work because participants also felt that working as a group and sharing
their self-critical thoughts and efforts to be kinder
to themselves had been helpful. Imagery work
might be helpful in that it enables people to ``carry
their images'' with them and use them outside a
group setting. We would also suggest that therapists need to explore the functions of self-criticism
and fear of giving up self-criticism (Gilbert &
Irons, in press).
In Buddhist meditation, developing compassion for the self involves giving people specific
images to focus on (Dagsay Tulku Rinpoche,
2002; Ringu Tulku & Mullen, in press). However,
this pilot study was based on guided discovery and
we were interested in how people generate their
own images and work with them. Table 3 offers
insights into the kinds of images created and how
they were used. In discussion, some participants
felt it might have been easier if they had been
given specific images to focus on, while others
thought they would prefer to work on their own
images. For example, one person started with a
religious image of a Buddha giving her compassion but could not make this ``work'' for her. She
then generated her own image of a bush in bloom
and found this very helpful. More research is
needed in this area. We have no data on whether

TABLE 3
Full reported experiences of using compassionate imagery from the six participants at follow-up
Question 2:
How did the image
appear to you?

Question 3:
What was the most
difficult aspect?

Question 4:
How much time were
you able to practise?

Question 5:
How helpful was it
using the image?

Participant
Number

Question 1:
What was the image
that you used over
the research period?

Floating in warm sea

Comforting
sensation
Sights and feelings

Visual
Feelings
Sound

Conjuring up image
when needed

When needed

Recognising selfcritical thoughts

when they occur
Diverting/stopping
self-critical thoughts

White bush with

comforting arms

Visual
Sense of warmth

Concentration
Stopping whirling
thoughts

5 mins to all day

Focus on things and

self not being bad
Helps ease pain of
being high/low
Need to be alone to
succeed

Rainbow
Candles underneath
instead of dark sky

Visual
Nice feelings

Hard to get feelings

Easier to get picture
Used chanting to try
and get image

Couple of times per

week

Not yet achieved

compassion with
self, but may with
more practice

Spiritual/Jesus
Love/caring/ support
In the air
Sunset/stars/flowers/
mountains

All senses
Visual
Beauty of
flowers, sky, sunsets
waterfalls
The love of friends
Peace

Would find
imagination
impossible because
it is unreality
I would be living in
an unreal world,
unreal ideals
Lying to myself,
pretending
dangerous for me
Enjoy daydreaming

Several times per

day
Whenever needed
Situations where am
giving myself putdowns, feel
inadequate etc
Pull up using
compassionate mind
my spiritual friend
reality

Made aware how

little I think about
myself
Changing thoughts
round
Would not have
survived life without
my Lord God

Sun
Feeling of warmth

Visual
Brightness
Open space
Warmth

Breaking lifetime's
habit of feeling bad
about self
Lack of
compassionate
people around me
Hard to conjure up
image felt remote
and cold

Only occasionally

Relaxation aspect
helped calm anxiety
Aware of benefits of
being compassionate
but unable to do.
Frustrated that
unable to do it
Hard to do with no
support

Arm round my
shoulders

Visual
Sense/feeling of
warmth

Image turned into

stomach of wellrounded male.
Brought back
difficult memories of
ex-husband

510 mins

Not at all
All good destroyed
by that (second)
image

513

514

GILBERT AND IRONS

an image of ``a person'' with compassionate qualities would work better than these non-person
images. On this more research is needed.
Some self-critical people may have few caring
and soothing memories to call on (Gillath et al.,
in press; Mikulincer et al., 2002). Thus, the selfcare and self-compassionate system may be
underelaborated (Gilbert & Irons, in press). If
people cannot utilise memories of caring others
to be self-soothing, then an important research
question is whether training people to generate
self-soothing imagery is possible, can be helpful,
and can be laid down as memories for subsequent recall. Lee (in press) has suggested that
compassionate imagery can be directed to that
of a ``perfect nurturer'' that has distinctive
features including sensory ones. These features
may aid the ease of accessibility from memory
on subsequent occasions, in the context of selfcriticism. Moreover, Lee has outlined how compassionate imagery can be helpful with people
suffering from post-traumatic stress disorder,
marked feelings of shame.
In regard to developing compassion for the self,
participants agreed with one member who said,
``this will take time as it is breaking the habits of a
life time.'' A number of participants reflected that
even as children they could not recall parents
being particularly kind or compassionate to them,
but more often cold or critical. Participants noted
that ``being kind'' to themselves was not ``something they were used to'' and ``at times it seemed
strange'' to them. However, all agreed that if they
could develop compassion for themselves this
would help them. Our research is clearly very
preliminary given the small numbers, but suggests
that some self-critical people can see the benefits
of attempting to become more self-compassionate, can generate a range of varied images with
different features, and find it a helpful process.
Questions arise about personified and non-personified images, and distinctions between feelings
of warmth, acceptance, and strength that are part
of compassion but can also vary from person to
person.
This study suffered from small numbers, and
also the fact that participants did not keep
their diaries adequately for the full 6 weeks.
Nonetheless, as a pilot study it points to the
value of diaries, especially for monitoring
forms of self-criticism and self-soothing, the
acceptability of this intervention for patients,
and the indications that, with development, it
may be a helpful intervention for some

patients. Future research may focus on the following:

1. What are the most useful, distinctive features of compassionate imagery?
2. Exploring how developing and practising
compassionate imagery may aid people who have
few memories of others being compassionate
towards them.
3. Exploring how development in the articulation and accessibility of compassionate images
may reduce the influence of self-criticism and help
alleviate various emotional difficulties associated
with it.
4. Investigating how a compassionate image(s)
may change with practice and the impact of such
change on self-criticism and affect self-regulation;
5. How to build this process into an established
psychotherapeutic approach, such as cognitive
therapy.

REFERENCES
Andrews, B. (1998). Shame and childhood abuse. In P.
Gilbert & B. Andrews (Eds.), Shame: Interpersonal
behavior, psychopathology and culture (pp. 176
190). New York: Oxford University Press.
Blatt, S. J. (1995). The destructiveness of perfectionism:
Implications for the treatment of depression.
American Psychologist, 50, 10031120.
Blatt, S., & Zuroff, D. (1992). Interpersonal relatedness
and self-definition. Two prototypes for depression.
Clinical Psychology Review, 12, 527562
Bowlby, J. (1969). Attachment: Attachment and loss
(Vol. 1). London: Hogarth Press.
Bowlby, J. (1973). Separation, anxiety and anger:
Attachment and loss (Vol. 2). London: Hogarth
Press.
Brewin, C. R. (2003). Post-traumatic stress disorder:
Malady or myth? New Haven, CT: Yale University
Press
Cox, B. J., Rector, N. A., Bagby, R. M., Swinson, R. P.,
Levitt, A. J., & Joffe, R. T. (2000). Is self criticism
unique for depression: A comparison with social
phobia. Journal of Affective Disorders, 57, 223228.
Dagsay Tulku Rinpoche (2002). The practice of Tibetan
meditation: Exercises visualizations and mantras for
health and well-being. Vermont: Inner Traditions
International.
Ferguson, E. (in press). The use of diary methodologies
in health and clinical psychology. In J. N. V. Miles &
P. Gilbert (Eds.), A handbook of research methods in
clinical and health psychology. Oxford: Oxford
University Press.
George, M. S., Ketter, T. A., Parekh, P. I., Horwitz B.,
Hercovitch P., & Post R. M. (1995). Brain activity
during transient sadness and happiness in healthy
women. American Journal of Psychiatry, 152,
341351.

COMPASSIONATE IMAGES AND SELF-CRITICISM

Gilbert, P. (1989). Human nature and suffering. Hove,

UK: Lawrence Erlbaum Associates Ltd.
Gilbert, P. (1998). What is shame? Some core issues and
controversies. In P. Gilbert & B. Andrews (Eds.),
Shame: Interpersonal behavior, psychopathology and
culture (pp. 336). New York: Oxford University
Press.
Gilbert, P. (2000a). Social mentalities: Internal `social'
conflicts and the role of inner warmth and compassion in cognitive therapy. In P. Gilbert & K. G.
Bailey (Eds.), Genes on the couch: Explorations in
evolutionary psychotherapy (pp. 118150). Hove,
UK: Psychology Press.
Gilbert, P. (2000b). Overcoming depression. New York:
Oxford University Press.
Gilbert, P. (2002). Body shame: A biopsychosocial
conceptualisation and overview, with treatment
implications. In P. Gilbert & J. Miles (Eds.), Body
shame: Conceptualisation, research & treatment (pp.
354). London: Brunner-Routledge.
Gilbert, P. (2004). Depression: A biopsychosocial,
integrative and evolutionary approach. In M. Power
(Ed.), Mood disorders: A handbook of science &
practice (pp. 99142). Chichester, UK: Wiley.
Gilbert, P., Baldwin, M., Irons, C., Baccus, J., & Clark,
M. (2004a). Evolution, relational schemas and the
internalisation of hostile and reassuring selves: Their
relation to depression. Manuscript submitted for
publication.
Gilbert, P., Birchwood, M., Gilbert, J., Trower, P., Hay,
J., Murray, B. et al. (2001). An exploration of
evolved mental mechanisms for dominant and subordinate behaviour in relation to auditory hallucinations in schizophrenia and critical thoughts in
depression. Psychological Medicine, 31, 11171127.
Gilbert, P., Clarke, M., Hempel, S., Miles, J. N. V. &
Irons, C. (2004b). Criticising and reassuring oneself:
An exploration of forms, styles and reasons in female
students. British Journal of Clinical Psychology, 43,
3150.
Gilbert, P., & Irons, C. (in press). Compassionate
mind training, for shame and self-attacking, using
cognitive, behavioural, emotional and imagery
interventions. In P. Gilbert (Ed.), Compassion:
Conceptualisations, research and use in psychotherapy. London: Brunner-Routledge.
Gilbert, P., & Miles, J. N. V. (2000). Sensitivity to put
down: Its relationship to perceptions of shame, social
anxiety, depression, anger and selfother blame.
Personality and Individual Differences, 29, 757774.
Gillath, O., Shaver, P. R., & Shaver, P. R. (in press). An
attachment-theoretical approach to compassion and
altruism. In P. Gilbert (Ed.), Compassion: Conceptualisations, research and use in psychotherapy.
London: Brunner-Routledge.
Goodare, H., & Lockwood, S. (1999). Involving patients
in clinical research: Editorial. British Medical Journal, 319, 724725.
Greenberg, L. S., Elliott, R. K., & Foerster, F. S (1990).
Experiential processes in the psychotherapeutic
treatment of depression. In C. D. McCann & N. S.
Endler (Eds.), Depression: New directions in theory,
research and practice (pp. 157185). Toronto: Wall &
Emerson Inc.

515

Hackmann, A. (1998). Working with images in clinical

psychology. In A. Bellack & M. Hersen (Eds.),
Comprehensive clinical psychology (pp. 301317).
London: Pergamon.
Hackmann, A. (in press). Compassionate imagery in the
treatment of early memories in axis I anxiety disorders. In P. Gilbert (Ed.), Compassion: Conceptualisations, research and use in psychotherapy.
London: Brunner-Routledge.
Irons, C., Gilbert, P., Baldwin, M., Baccus, J., & Palmer,
M. (2004). Parental recall, attachment relating and
self-attacking/self-reassurance: Their relationship
with depression. Manuscript submitted for publication.
Kaufman, G. (1989). The psychology of shame. New
York: Springer.
Kohut, H. (1977). The restoration of the self. New York:
International Universities Press.
Lee, D. (in press). The perfect nurturer: A model to
develop a compassionate mind within the context of
cognitive therapy. In P. Gilbert (Ed.), Compassion:
Conceptualisations, research and use in psychotherapy. London: Brunner-Routledge.
Linehan, M. (1993). Cognitive behavioral treatment of
borderline personality disorder. New York: Guilford
Press.
McKay, M., & Fanning, P. (1992). Self-esteem: A
proven program of cognitive techniques for assessing, improving, and maintaining your self-esteem.
Second Edition. Oakland, CA: New Harbinger
Publishers.
Mikulincer, M., Gillath, O., & Shaver, P. R. (2002).
Activation of the attachment system in adulthood:
Threat-related primes increase the accessibility of
mental representations of attachment figures. Journal of Personality and Social Psychology, 83,
881895.
Neff, K. D. (2003). Self compassion: An alternative
conceptualisation of a healthy attitude to oneself.
Self and Society, 2, 85102.
Potter-Efron, R. (2002). Shame, guilt and alcoholism:
Treatment issues in clinical practice. New York: The
Haworth Press.
Rector, N. A., Bagby, R. M., Segal, Z. V., Joffe, R. T., &
Levitt, A. (2000). Self-criticism and dependency in
depressed patients treated with cognitive therapy or
pharmacotherapy. Cognitive Therapy and Research,
24, 571584.
Reynolds, M., & Brewin, C. R. (1999). Intrusive
memories in depression and posttraumatic stress
disorder. Behaviour Research and Therapy, 37,
201215.
Ringu Tilku & Mullen, K. (in press). The Buddhist use
of compassionate imagery in Buddhist meditation. In
P. Gilbert (Ed.), Compassion: Conceptualisations,
research and use in psychotherapy. London: Brunner-Routledge.
Salzberg, S. (1995). Loving-kindness: The revolutionary
art of happiness. Boston, MA: Shambhala.
Schwartz, J. M., & Begley, S. (2002). The mind and the
brain: Neuroplasticity and the power of mental force.
New York: Regan Books.
Tangney, J. P., & Dearing, R. L. (2002). Shame and
guilt. New York: Guilford Press.

516

GILBERT AND IRONS

Tomkins, S. S. (1987). Shame. In D. L. Nathanson (Ed.),

The many faces of shame (pp. 133161). New York:
Guilford Press.
Wheeler, L., & Reis, H. T. (1991). Self-recoding of
everyday life events: Origins, types and uses. Journal
of Personality, 59, 339354.

Zigmond, A. S., & Snaith, R. P. (1983). The Hospital

Anxiety and Depression Scale. Acta Psychiatrica
Scandinavica, 67, 361370.
Zuroff, D. C., Moskowitz, D. S., & Cote, S. (1999).
Dependency, self-criticism, interpersonal behaviour
and affect: Evolutionary perspectives. British Journal of Clinical Psychology, 38, 231250.