Trends in

OPEN ACCESS Cognitive Sciences

Opinion

The pattern theory of compassion

1,2,
Shaun Gallagher *, Antonino Raffone 3, and Salvatore M. Aglioti 3,4

Concepts of empathy, sympathy and compassion are often confused in a variety of Highlights
literatures. This article proposes a pattern-theoretic approach to distinguishing This article proposes a pattern theory
compassion from empathy and sympathy. Drawing on psychology, Western phi- of compassion on a model similar to a
pattern theory of emotion.
losophy, affective neuroscience, and contemplative science, we clarify the nature
of compassion as a specific pattern of dynamically related factors that include This pattern theory of compassion is
physiological, cognitive, and affective processes, relational/intersubjective pro- consistent with cognitive and affective
cesses, and motivational/action tendencies. We also show that the dynamic nature neuroscience studies of compassion.

of the compassion pattern is reflected in neuroscientific findings, as well as in com-

The pattern theory of compassion fa-
passion practice. The pattern theory of compassion allows us to make several clear cilitates sharp conceptual distinctions
distinctions between compassion, empathy, and sympathy. among compassion, empathy, and
sympathy.

A clear conception of compassion is

Empathy, sympathy, and compassion helpfully relevant to compassion training
In philosophical and scientific literatures, one continues to find long-standing equivocations and the development of compassion-
based therapies in mental health and
between three closely related but distinct phenomena: empathy, sympathy, and compassion
medical contexts.
(see Glossary). One problem is that terms such as 'sympathy' and 'empathy' have taken on
different meanings since their early uses. Adam Smith, for example, considered sympathy to be
a human innate capacity that serves the processes of social and moral relationships, cooperation,
and communication. More recently it has come to signify something akin to pity or feeling sad for
someone [1]. 'Empathy', a more recent term (originally Titchener's translation of Lipps' [2] term
Einfühlung), at first meant something resembling an automatic, responsive feeling or resonance
towards another person, a meaning still maintained in simulation theories of empathy (e.g., [3]). In
contemporary common usage, however, it is not clear that empathy is viewed as automatic. Instead,
empathy is viewed as being more selective – a special type of affective state that may depend on
understanding the situation of a person.

Our focus in this article is on compassion, and on how to distinguish it from empathy and sympathy.
Clarifying the meaning of compassion is relevant because it is increasingly discussed in clinical
contexts and contexts concerning psychological well-being. Compassion-focused therapy for
mental health problems, for example, attempts to identify interventions that stimulate processes of
compassion (e.g., [4,5]). One finds a similar and growing interest in compassion in education, medi-
1
cine, and other applied contexts (e.g., [6–10]). In these various contexts, however, the concepts Department of Philosophy, University of
Memphis, Memphis, TN, USA
of compassion and empathy are not always distinguished [11], although in compassion-focused 2
School of Liberal Arts (SOLA), University
therapy a distinction is sometimes made, where compassion is considered to be a form of motivation of Wollongong, Wollongong, NSW,
and empathy a competency that may be used for compassionate means [4]. Australia
3
Department of Psychology, Sapienza
University of Rome, Rome, Italy
Drawing on psychology, western philosophy, cognitive and affective neurosciences, and 4
Center for Life, Nano-, and Neuroscience
contemplative science, we clarify the nature of compassion and its differences from empathy (CLN2S), Istituto Italiano di Tecnologia,
Genova, Italy
and sympathy by developing the pattern theory of compassion. In this approach, compassion
consists of a specific set of dynamically related factors that include physiological, cognitive, and
affective processes, relational/intersubjective processes, and behavior/action tendencies. We
*Correspondence:
will also show that the dynamic nature of the compassion pattern is reflected in neuroscientific [email protected]
findings, as well as in compassion practice. (S. Gallagher).

504 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 https://doi.org/10.1016/j.tics.2024.04.005
© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Trends in Cognitive Sciences
OPEN ACCESS

Confusions in scientific, therapeutic, and philosophical contexts Glossary

As numerous authors suggest, consistent definitions of empathy, sympathy, and compassion do Cognitive appraisal: cognitive
not exist in the literature; they are often treated as overlapping concepts [12–16]. Acknowledging evaluation and interpretation of a
phenomenon or event such as an
many different definitions of sympathy and empathy, one paper [17] provides 42 references as emotional situation.
evidence of this definitional diversity. Some theorists consider compassion to be a form of empathy Compassion: an action-motivating
(e.g., [18]); others consider empathy to be a form of compassion (e.g., [19]). form of feeling or caring attitude directed
towards alleviating the suffering of
another person, supported by emotion
In the context of brain-imaging studies, researchers both distinguish between and associate regulation and cognitive processes.
empathy and compassion, equating compassion with sympathy that is understood as a possible Compassion meditation: the
outcome of empathy (e.g., [20]). An empathic response to suffering is said to involve both cultivation of an emotional attitude aimed
at fostering compassion toward oneself
empathic distress and compassion. In addition, it has been shown that empathic distress and
and others.
compassion activate different brain networks that are tied to negative and positive affect, respec- Emotion regulation: attempts to
tively, thereby supporting a distinction between empathy and compassion [20]. Likewise, empa- monitor and modulate one's emotional
thy training is distinguished from compassion training [21]. Although some researchers consider experience through strategies.
Empathy: an affective state isomorphic
compassion to be a type of emotion [19], and in some cases reference Buddhist conceptions of
with the positive or negative affective
compassion [20], some Buddhist thinkers [22–24] refuse to equate compassion with a particular state of another person that is elicited by
emotion. Instead, for them it involves a type of practiced wisdom that acknowledges our interde- observing or imagining the other person,
pendence with others and the obligation to act to reduce their suffering. and when one knows that the other
person is the source of one's own
affective state.
Compassion is increasingly discussed in clinical contexts. In compassion-focused therapy, that is Equanimity: an even-minded mental
included in the treatment of mental health problems [25], there is a clear distinction between com- attitude or dispositional tendency
passion that is understood as a form of motivation, and empathy (or sympathy) that is understood toward all experiences, objects, or
persons (living beings) regardless of their
as a skill or competency which can be used for compassionate means [6]. In other discussions of origin, attributes, or the implicated
therapist competency, however, compassion is not clearly differentiated from empathy [11]. affective valence.
Intentional object: the target of any
cognitive act, such as perception,
Cognitive scientists sometimes identify criteria to distinguish empathy from sympathy, as well as
memory, or imagination.
from theory of mind and contagion [26]. The central idea here is that empathy is based upon Mindfulness: non-judgmental or
simulation, a process that involves the empathizer entering into emotional states that match acceptance-based attentional awareness
those of the person with whom they are empathizing. On this view, when I empathize with some- of experience at the present moment.
Pattern theory: the view that a
one who is angry, I also experience simulated, vicarious anger. By contrast, when I sympathize
particular phenomenon (e.g., an
with someone my emotional state does not need to be isomorphic with the emotional state of emotion) is a pattern of dynamically
the other. I may feel pity, or worry, or sadness if the other person is upset, for example. On the related diverse factors or processes.
one hand, this difference may differentiate empathy and compassion if one equates compassion Phenomenal feeling: the qualitative
feel of experiencing, typically construed as
and sympathy [20]. On the other hand, the Buddhist requirement that compassion involves
'what it is like' to experience something.
motivation for prosocial behavior may distinguish it from both empathy and sympathy [22–24]. Suffering: a severely unpleasant
experience of a psychophysical or
This short review may be sufficient to indicate the serious conceptual confusions that exist in the existential type that may include but is
not reducible to physical pain.
literature. We offer some conceptual clarifications by taking a pattern-theoretic approach to dis- Sympathy: a caring affective state
tinguishing between empathy, sympathy, and compassion. In the next section we explain the intentionally directed towards the
concept of a pattern theory by focusing on emotion as a starting point. We then present the pat- negative experience of another person.
tern theory of compassion and the related neuroscience. This will allow us to distinguish the com- Wisdom: general knowledge in life
management, life planning, and life
passion pattern from the patterns of empathy and sympathy. review, including an ethically grounded
capacity for social-cognitive reflection
The pattern theory of emotion and perspective taking, with an
orientation toward the common good
The pattern theory of emotion claims that emotions are complex patterns of bodily/physiological
and pursuing a deeper truth.
processes, experiences, expressions, and actions, which together make up an emotion pattern.
As such, emotions are individuated by patterns of characteristic features [27–29]. This idea was
developed under the title 'differential emotions theory' [27] which contends that a complex emo-
tion consists of a pattern of more basic emotions. On a different and more dynamic conception,
emotions are considered as patterns of variables that are not themselves emotions [29]. We can

Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 505

 Trends in Cognitive Sciences
OPEN ACCESS

think of such variables, such as neurohormonal, motoric, and experiential processes set up as
evolutionary adaptations, as being interactively related such that emotions are constituted as
dynamic interactions of such constituents, including organism–environment transactions. We
can think of discrete emotions as being dynamically self-organizing in that 'recursive interactions
among component processes generate emergent properties' ([28], p. 15). Thus, different emo-
tions are constituted by different patterns of processes which yield behaviors that vary from
one individual to another, and within individuals over time – where such behaviors/expressions
are components of the emotion. Individual emotions may still combine or coassemble with
other emotions to form new emotion patterns that may stabilize over repeated occurrences.

On this approach, one can provide a catalog of different features which may contribute to specific
patterns that constitute emotions [29]. These include the following.

(i) Physiological processes: for any emotion, autonomic activity is a possible constituent, and the
subject may be aware of autonomic effects or not. Not every emotion has a distinct autono-
mous nervous system (ANS) pattern, and different emotions do not necessarily have different
ANS patterns [30].
(ii) Actions or action tendencies: these are bodily changes in preparation for actions that may be
experienced as urges to perform a particular type of action [31]. Some emotions, for example
happiness, may or may not include this component; others may be typically associated with
specific actions (e.g., freezing, fighting, or fleeing in fear).
(iii) Overt expressions: these include expressive postures and movements, facial expressions,
gestures, and vocal expressions and intonations.
(iv) Phenomenal feeling: the conscious phenomenal feeling component is often part of an emo-
tion, although it is not necessary for every emotional occurrence. In some rare cases, typical
physiological, expressive, and cognitive aspects may be present without the phenomenal
aspect (e.g., in subjects who are disposed to repress fear [32]).
(v) Cognitive processes: these include typical emotion-related attitudes, shifts of attention, and
perceptual changes. Cognitive attitudes may include, for example, belittling thoughts about
one's rival in the case of jealousy, or a judgment that one has been treated unfairly in some
cases of anger [29].
(vi) Intentional objects: an intentional object is the perceived, remembered, or imagined object
that the emotion is about [33]. One can distinguish between self and other orientations:
some emotions are focused on self-related processes, whereas others involve intersubjective
relations.

These features are variables that can take different values and weights in the dynamic constitution
of an emotion. Some values are more or less likely to occur together. In this respect we can dis-
tinguish a typical pattern of affective aspects and values and define an emotion as involving some
set of variations of that pattern. To say that a particular feature is constitutive of an emotion does
not mean that it is an essential component [29]. Although there may be some minimal number of
characteristic features and their values which is sufficient to constitute a particular pattern that
counts as instantiating a particular emotion type (e.g., anger), the same type of emotion can
lack a particular characteristic feature. Not all instances of anger, for example, involve aggressive
action tendencies or narrowed attention. One or more factors may be missing, but there may still
be a sufficient number and organization of factors to allow the pattern to be recognized as anger.

The idea that emotion can be understood as a pattern of this type is consistent with enactivist
conceptions of emotion 'forms' that involve various neural, muscular, and autonomic processes
that together form an integrated pattern ([34], p. 69). A pattern exists not only as a collection of

506 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6

Trends in Cognitive Sciences
OPEN ACCESS

factors but also, importantly, as a set of relations among these elements. The specific value and
weight that each element has in the pattern will depend on its relations with other elements. We
can think of the pattern as a dynamic gestalt in which, if one factor (or value or weight relative to
the whole) is changed above a particular threshold, some or all of the other factors (and perhaps
the whole) adjust. In a dynamic gestalt, what happens with respect to one factor is determined by
what happens with respect to all the others [35].

The pattern theory of compassion

Several psychologists define compassion as a unique emotion [19,36]. On this view, one should be
able to map out an emotion pattern that defines compassion. To be clear, however, we take com-
passion to be (like emotion) a pattern of different factors that include not only affective components
but also other features that are not necessarily the same components as those of an emotion pat-
tern. Accordingly, we understand the pattern theory of emotion to be a useful heuristic for
developing a pattern theory of compassion, one that allows us to characterize compassion without
necessarily equating compassion with an emotion.

We can start our considerations with a consensus list of components or characteristics derived from
several reviews [36,37]. Compassion involves (i) recognition or awareness of suffering, (ii) feeling
'moved' for those who are suffering, (iii) appraisals of the situation of the other person and one's
own ability to respond and tolerate distress, (iv) being motivated to act, or acting to alleviate the suf-
fering, and (v) activation of neural systems that drive such motivations and actions.

We propose that these components can be incorporated into a more comprehensive pattern of
processes and factors that constitute compassion, consistent with the view expressed by Goetz
and Simon-Thomas: 'Although we see compassion as involving a patterned and specific
response, we do not see the components listed here as serial, or occurring in temporal sequence.
We also do not consider the processes underlying these components to be wholly independent;
they probably overlap and occur in parallel, and exert bidirectional influence upon one another in
different configurations throughout life' ([37], pp. 27–28). Like emotion, we can define compas-
sion as a particular recursive gestalt arrangement of processes. Mapping out this compassion
pattern will allow us to clearly differentiate it from empathy and sympathy.

The compassion pattern includes the following factors/processes drawn from scientific,
Western philosophical, and Buddhist psychology sources. As in the case of an emotion
pattern, not all factors are necessarily present in all forms of compassion. Moreover, although
various theories may claim that one or more of these factors is central or essential to compas-
sion, the pattern theory can leave this an open question and still offer some clarification about
the nature of compassion. What is central to the compassion theory is that some set of factors/
processes and their dynamic relations form a particular recursive arrangement of processes
that constitute the compassion pattern.

Physiological processes
Compassion modulates the hormones oxytocin and vasopressin, processes of the parasympathetic
nervous system, and other neurophysiological circuits [16,38,39]. Autonomic/parasympathetic
processes leading to heart rate deceleration can be associated with compassion [19].

Action tendency/motivation
Compassion involves the altruistic motivation to act to alleviate suffering [3,19,40]. This motivation
may be reactive to the experience of the suffering of another person or may be proactively attuned
to improve the lives of others. This prosocial aspect is consistent with neuroscientific evidence [16,41].

Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 507

 Trends in Cognitive Sciences
OPEN ACCESS

Overt expressions
Typical (although not unique) bodily expressions – facial expressions, posture, vocalizations, and
tactile behavior – may characterize instances of compassion [19].

Affective/phenomenal feeling
Feeling 'moved' for those who are suffering, variously described as an experienced emotional
resonance or involuntary arousal (which may be due to activation of the ANS [19]). Compassion
may reflect feelings that are both positive (related to social accomplishment, caregiving) and
negative (about the other's suffering) [37].

Cognitive processes
Compassion involves processes of cognitive appraisal that are attuned to antecedent
signs of undeserved suffering, as well as to one's physical and emotional capacity to ad-
dress it [19]. This depends on an ability to attune to and cope with the distress associated
with witnessing the suffering of others [16]. At a dispositional level, this is not merely an in-
tellectual understanding of our interdependence [23,42] but is also a wisdom that informs
practice [42,43]. Contemplative traditions identify a variety of cognitive processes involved
in compassion practice, including attentional, constructive, and deconstructive strategies
that involve meta-awareness, perspective taking, cognitive reappraisal, and self-inquiry [42]
(Box 1).

Intentional object
Compassion specifically targets the suffering of another person [4]. Compassion is not the
generic sharing of the emotions of another; instead it implicates a feeling of caring about and
being moved by the suffering of another [36]. Sometimes, however, compassion may not have
an intentional object [22,23].

Self/intersubjective orientation
Compassion involves a relational attitude that extends beyond the self to others without being
affected by qualities such as emotional closeness, trust, or kinship. Compassion allows for rela-
tionships beyond kin relationships: it can apply to out-group members [19]. In Buddhist practice,
compassion applies equanimously to everyone [22,44].

Situational aspects
Specific situational aspects or antecedents of compassion may include the presence of others
(including infants or children, as well as animals) in need, distress vocalizations, pain, sadness,
illness, physical or mental disability, homelessness, poverty, and natural disasters [19].

Emotion regulation processes

Compassion requires a reflective and emotion regulation competency that maintains a mental
balance [45] which is capable of steering a course between avoidance of pain, suffering [46,47],
and emotional reactivity involving empathic distress [20]. This type of regulation can be achieved
by mindfulness training aimed at cultivating non-judgmental or acceptance-based attention to
experience at the present moment [46–48].

The processes involved in the compassion pattern are dynamically intertwined, enacted in a
holistic fashion, and should not be thought of as being simply an additive list or following a step-
wise serial or sequential order [37]. Nevertheless, for heuristic purposes one could map out the
set of elements/processes that are involved in the practice of compassion in a multifaceted
cycle (Box 1).

508 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6

Trends in Cognitive Sciences
OPEN ACCESS

Neuroscientific findings related to the compassion pattern

Compassion has received increased attention in neuroscientific studies in recent years
[19,49,50]. One can find data suggesting physiological and neurophysiological correlations that
reflect many of the specific factors and their dynamic interactions in the compassion pattern
and the cycle proposed in Box 1 (Table 1).

A recent meta-analysis of 16 fMRI studies [49] confirms compassion-related activation across

seven broad regions, where the largest peaks are localized to the periaqueductal gray, anterior
insula, anterior cingulate cortex (ACC), and inferior frontal gyrus. An integrative review [50]
revealed that the most frequent associations with compassion across all the studies analyzed
are found in the orbital part of the left inferior frontal gyrus, right cerebellum, bilateral middle tem-
poral gyrus, bilateral insula, and right caudate nucleus. A lower degree of compassion tends to be
associated with either lower neural activity or gray matter volume in neural areas linked to reward,
and is thus related to the motivation or action tendency processes of the compassion pattern.

Box 1. The compassion cycle

We can picture a compassion cycle that includes the recurring aspects listed below (Figure I).

1. An underlying dispositional attitude (which may include wisdom as the result of prior compassion experiences or
compassion training) and related cognition/metacognition based on insights about the nature of the self, the
other, and interdependence.
2. An underlying dispositional motivation to (equanimously) care about and alleviate the suffering of others.
3. Recognition of the pain and suffering of others (intentional object).
4. Empathic negatively valenced phenomenal feeling of pain/suffering.
5. Rapidly established emotion regulation (e.g., based on mindfulness and equanimity) processes that prevent the pain
avoidance drive and reactivity, leading to emotional contagion or empathic distress.
6. Cognitive appraisal of the suffering, its context, and the available resources to help, possibly also involving perspective
taking and further emotion regulation based on cognitive reappraisal.
7. A caring drive with a positively valenced phenomenal feeling of warmth and feeling 'moved' to alleviate suffering.
8. Context-sensitive intention formation toward an appropriate compassionate response.
9. Conscious selection and enaction of an appropriate response.
10. Appraisal of the enacted response consequences.
11. A cognitive monitoring and learning process that links the recognition of the pain/suffering stimulus and the awareness
of the situation/context with the enacted response and its consequences, which inform cognitive and emotional dispo-
sitions towards further encounters with pain/suffering stimuli.

This cycle includes a functional recurring arrangement of dispositional aspects, perceptions, feelings, cognitive processes,
actions, monitoring, and learning processes. We can ask whether these are discrete states or traits, or are instead more
relational or fluid processes [4]. Likewise, we can ask whether, in each case, every step is necessary – something that is
not required by the pattern theory (see Outstanding questions).

This pattern-theoretic compassion cycle is relevant to training for compassion in contemplative, clinical, and educational
contexts. The putatively synergistic involvement of such processes in compassion training may underlie its holistic effective-
ness in improving psychological well-being, including stress-related immune response (e.g., [67,68]), mental and physical
health (e.g., [67,69]), positive affect (e.g., [69]), empathic responses towards others (e.g., [70,71]), positive emotions
towards those who are suffering (e.g., [72,73]), prosocial behavior (e.g., [74,75]), cognitive reappraisal as an emotion
regulation strategy [76], emotional memory retrieval [77], and the reduction of interpersonal conflict [78].

Concerning compassion training in contemplative practices, the pattern theory of compassion highlights the importance of
integrating the training of (i) attention, mindfulness, equanimity, and related emotion regulation skills, (ii) the 'constructive'
strengthening of psychological patterns that foster a caring motivation and a sense of interconnectedness with others,
as well as positive emotions and well-being, and (iii) the 'deconstructive' undoing of maladaptive cognitive patterns and
habits by exploring the dynamics of perception, emotion, and cognition to generate insights into one's understanding of
self, others, and the world [42]. Accordingly, the contemplative traditions identify cognitive processes that involve meta-
awareness, namely a heightened awareness of one's own thinking, feeling, and perceiving, as well as perspective taking
– understanding another's situation and how they feel, cognitive reappraisal, and self-inquiry [42].

Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 509

 Trends in Cognitive Sciences
OPEN ACCESS

Trends in Cognitive Sciences

Figure I. The compassion cycle. A possible ordering of elements and dynamic processes in the compassion pattern in
a multifaceted cycle in compassion practice.

Table 1. Compassion and neural activations: associations between compassion processes, brain region
activations, and factors of the compassion pattern derived from [19]
Compassion process Brain region activations Factor of the
compassion pattern
Detecting the expressions of suffering Temporoparietal junction Intentional object
of another person
Mirroring the emotional experience of Inferior frontal cortex, insula, temporal pole Intersubjective
another person orientation
Assessing the relevance/deservedness Mid-ventral medial prefrontal cortex Cognitive processes
of the sufferer
Coping with empathic distress Dorsal medial prefrontal cortex/inferior frontal Emotion regulation
cortex
Feeling warmth/tenderness towards Periacqueductal gray, substantia nigra, and Phenomenal feeling
others ventral tegmental area
Overarching motivation to approach Heightened left hemisphere activation Action tendency/
motivational aspect

510 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6

Trends in Cognitive Sciences
OPEN ACCESS

Several neuroscientific studies highlight the involvement of the ACC and anterior insula in
compassion. According to an influential model [51], the anterior insula and ACC have a close
functional relationship: they may be considered together as the input (anterior insula) and output
(ACC) regions of a functional system that is engaged across cognitive, affective, and behavioral
contexts, and that interplays with autonomic processes. One possible hypothesis is that the an-
terior insula plays a key role in the sensory and phenomenal feeling aspects of the compassion
pattern, and the ACC in the motivational action-related aspect. Given their interplay with the
ANS (and thus with hormonal regulation), the limbic network, and other executive control regions
in prefrontal cortex, as well as their role in modulating interactions between core brain networks im-
plicated in cognition, emotion, and interpersonal processing [52], the anterior insula and ACC may
serve as crucial brain hubs for compassion. Moreover, motivational action-related processes in the
compassion pattern may involve midbrain periaqueductal gray [53], a phylogenetically old area that
supports mother–infant attachment behaviors and is plausibly implicated in a caregiving orientation
towards those who are vulnerable or suffering, thus motivating compassionate responses.

The interplay with hormones (i.e., oxytocin, estrogens, prolactin, and vasopressin) reflects
a dynamic linking with other core physiological processes of the compassion pattern. The overall
process is highly complex if one considers that each hormone may exert different roles in different
circumstances. For example, it has been demonstrated that the increase of plasma oxytocin
levels contingent upon acute psychosocial stress is buffered by affective and perspective mental
training. Crucially, however, stress-contingent oxytocin levels increase after a short-term
compassion-related loving/kindness/meditation session [54]. In keeping with the allostatic theory
of oxytocin [55], this pattern of results indicates that one important function of oxytocin may be
that of maintaining stability in ever-changing environments. It is also worth mentioning that neuro-
pharmacological and neuroimaging research [49] highlights the combined influence of neural and
hormonal influence on compassionate behavior.

Given the proposal that the multiple processes or aspects of the compassion pattern are interde-
pendently related in a dynamic gestalt, it is important to explore the integration processes that are
reflected in neural mechanisms. A putative mechanism may involve large-scale oscillatory syn-
chronization in the brain, as observed in the gamma band in long-term compassion meditation
practitioners [56]. Dynamic complexity measures, such as causal density and information integra-
tion, may also be relevant to reveal the integrative neural processes that are implicated in the com-
passion pattern, including relevant phenomenal feelings [57] (Box 2).

Revisiting the distinctions between compassion, empathy, and sympathy

The pattern theory of compassion puts us in a better position to discuss how compassion relates
to empathy and sympathy. One strategy would be to provide a pattern theory of empathy and
sympathy, and then perform a comparative analysis. Given that there are extensive philosophical
issues involved in this type of analysis, and given the space limitations of a journal article, in place
of the longer and more complex analysis we propose a shortcut. Simulationist accounts of empa-
thy propose a clear distinction between empathy and sympathy, and identify five conditions that
define empathy [26,58,59]. We take this as a good starting point for an analysis that will distinguish
the pattern of compassion from those of both empathy and sympathy.

The simulationist account offers the following definition in terms of a set of collectively sufficient
conditions for empathy [26]. There is empathy if (i) one is in an affective state, (ii) this state is
isomorphic to the state of another person, (iii) this state is elicited by the observation or imagina-
tion of the affective state of another person, and (iv) one knows that the other person is the source
of one's own affective state ([26], p. 435). The first 'affectivity' condition distinguishes empathy

Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 511

 Trends in Cognitive Sciences
OPEN ACCESS

Box 2. How to empirically test the pattern theory of compassion

A variety of specific research targets and innovative methods in the study of the processes of compassion and its neural cor-
relates, that combine dispositional self-report measures, neuroscientific investigations, behavioral tasks, phenomenological re-
ports, and experience sampling methods, can enhance our understanding of compassion as well as the ecological validity,
precision, and discriminant validity of compassion measures. Accordingly, the pattern theory of compassion can be empirically
tested using the following research methods.

(i) Studying the relationships between dispositional (trait) self-report measures of compassion pattern factors (such as
through scales measuring empathic feeling, sympathetic concern, and perspective taking; e.g., [79,80]), prosocial
motivation and altruistic behavior (e.g., [81,82]), equanimity (e.g., [83]), cognitive reappraisal [84], wisdom [85], and
mindfulness facets (e.g., [86]) by both meditation-naive and meditation practitioners with different types of expertise.
Positive correlations between trait compassion pattern factors and self-reassurance [87], as well as negative corre-
lations with self-criticism [88] and pain catastrophizing [89], can be further assessed.
(ii) The use of focused induction procedures [90–92] to assess causal influences [46] between compassion pattern
states, including compassion-related physiological states.
(iii) The use of modular mental training programs, with modules focused on different aspects of the compassion pattern,
to longitudinally assess influences between measures of multiple factors in the compassion pattern.
(iv) The use of ecological experience sampling [93] to assess relationships between states associated with different
aspects of the compassion pattern, also in association with dispositional self-report measures.
(v) The use of functional neuroimaging to assess the implications of brain regions and networks in association with the
induction of different aspects of the compassion pattern, with particular reference to anterior insula and anterior
cingulate cortex (ACC, main text). Brain functional connectivity, including patterns of causal influences, can be usefully
assessed.
(vi) The use of event-related potentials [94], in combination with subjective reports (e.g., [46]) related to different aspects
of the compassion pattern, to assess the temporal ordering and causal influences of aspects of the compassion
cycle, in response to stimuli about the suffering of others, also in association with dispositional measures and in
longitudinal designs with mental training.

from standard mindreading, whether based on theoretical inference or simulation routines, be-
cause inferring that another person is in pain, for instance, can be distinguished from
experiencing or feeling empathetic pain [59]. The second 'interpersonal similarity condition'
([59], p. 521) distinguishes empathy from sympathy: empathy involves being in the same affec-
tive state as the other person (I feel your pain), whereas sympathy involves being in a different,
although valence-congruent, affective state (e.g., I feel sorry that you are in pain).

To capture the idea that empathy is other-directed, one can add a fifth condition: the caring
condition [58]. In the case of empathy, one cares about the affective life of the target. However,
even if empathy is other-directed, is not clear that the caring condition is essential because it is
also thought that someone who is a good torturer may use an empathic sense to know the
state of his victim, without caring for that person. On a different view [60], the caring condition
operates as a mark of sympathy rather than empathy, and this marks the difference between
the two types of state: sympathy involves concern or caring for the other person, whereas
no such concern is essential to empathy. The conditions defining empathy suggest that empa-
thy involves a specific pattern of processes that overlap in some regards with compassion and
sympathy, but differ in some regards. Both empathy and sympathy involve, or are, emotional/
affective states, although empathy (understood as cognitive empathy) can sometimes be more
of an intellectual attitude. Empathy, in contrast to both sympathy and compassion, requires not
only a relevant affective state in the empathizer, but this state, which can range across a broad
set of affective conditions, from joy to fear, to anger, to pleasure, must match that of the target.
The intentional object of compassion, however, is specified as the suffering of the other person,
and compassion is not itself a form of suffering and in fact is a counterpoint attitude to suffering.
Compassion, like sympathy, involves a caring attitude which is not essential for empathy. An
important difference between compassion and sympathy, however, is that – even if sympathy
does involve a caring attitude – such caring does not necessarily lead to action designed to al-
leviate suffering, whereas compassion involves motivation to engage in exactly such an action.

512 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6

Trends in Cognitive Sciences
OPEN ACCESS

There are three further issues to consider. First, researchers suggest that empathy may lead to
empathic distress resulting in a negative feeling or stress, burnout, or poor health, and thence
to withdrawal and non-social behavior [20]. This type of effect would involve a different kind of
normative attitude (as well as differences in affectivity and bodily processes) and could undermine
a positive caring attitude. Second, concerning the type of intersubjectivity involved, one can cite
what is sometimes regarded as the dark side of empathy, namely that it tends to be discriminatory,
such that it fails for out-groups as much as it succeeds for in-group members. Equanimity is often
not a characteristic of empathy, which may not exist for out-groups or competitor groups
[4,61–65]. These are complications that clearly distinguish empathy from compassion because
the latter is characterized by equanimity. Finally, we note that, like other facets of intersubjectivity,
compassion is influenced by cultural factors such as the degree of social interdependence. It has
been hypothesized that negative social affect has a major influence on the conception, experience,

Table 2. Distinctions between compassion, empathy, and sympathy: contrasts between compassion, empathy, and sympathy with reference to the
pattern theory of compassion
Compassion Empathy Sympathy
Self/intersubjective Equanimity: relational, intersubjective aspects that Discriminatory: it can fail for out-groups as much Like empathy,
orientation extend to out-groups, or 'universally', without as it succeeds for in-group members discriminatory
discrimination
Intentional object Specifically targets the suffering of another person Sharing/matching of the emotional state Variable targeting of the
(e.g., fear, anger, happiness) of another person; emotional states of another
interpersonal similarity condition person; no interpersonal
similarity condition
Situational aspects Specific situational antecedents that define or Various situations involving positive or negative Various situations involving
indicate suffering affect negative affect that elicit a
caring attitude
Cognitive processes Cognitive appraisal processes attuned to Processes involving imagination to form 'as if' or Appraisal processes
undeserved suffering, and to one's capacity to pretense state that matches the state of the attuned to the negative
address it, including a reflective ability to attune to other; this may involve simulation or inference of situation of another person
and cope with the distress associated with the mental state of another person
witnessing the suffering of others; this may involve
a wisdom-informed practice with insights leading
to realization that we are interdependent with
others
Emotion regulation A reflective and emotion regulation competency May involve distress and may be strategically Typically not distressful,
processes that maintains an emotional and mental balance used to cause distress in others and in this respect there is
capable of steering a course between avoidance no need for emotion
of pain and suffering and emotional reactivity regulation
leading to empathic distress
Action Altruistic motivation to act to alleviate suffering as No motivation to act towards the other or their Motivation to act towards
tendency/motivation an essential part; a prosocial behavioral attitude situation; possible withdrawal and non-social the other or their situation is
behavior in the case of empathic distress not necessarily involved
Physiological Modulates the hormones oxytocin and Increased heart rate; possible activation of mirror Contrasting with empathic
processes vasopressin, and processes of the system, including superior/middle temporal distress, lower heart rate
parasympathetic nervous system (e.g., the vagus sulcus/gyrus, and the anterior insula, as well as
nerve), as well as different neurophysiological more cognitive areas: the medial prefrontal
circuits cortex, the temporoparietal junction, the superior
temporal sulcus, and the anterior temporal pole
Affective/phenomenal Emotional resonance; both positive (associated Emotional/affective resonance Emotional/affective
feeling with care-giving and prosocial action) and resonance plus concern
negative feelings (associated with the suffering of
another person)
Overt expressions Typical (although not unique) bodily expressions: Similar Similar; contrasting with
facial expressions, posture, vocalizations, tactile empathic distress,
behavior increased facial responses

Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 513

 Trends in Cognitive Sciences
OPEN ACCESS

and expression of different facets of intersubjectivity [66]. However, a comprehensive model of the Outstanding questions
cultural shaping of compassion is still lacking, and further research on this topic is needed. Does compassion involve discrete
states, or traits, or are rather more
relational or fluid processes implicated?
The similarities and distinctions between compassion, empathy, and sympathy in light of the
Are all discrete states or relational/fluid
pattern theory of compassion are summarized in Table 2. processes in the compassion pattern
implicated in compassion practice?
Concluding remarks
How do the patterns of empathy and
Our focus in this paper has been on compassion. Using a pattern-theoretic approach we distin-
sympathy overlap with the factors and
guish compassion from empathy and sympathy. Drawing on psychology, Western philosophy, processes of the compassion pattern?
affective neuroscience, and contemplative science, we have offered some clarification about the
nature of compassion, and show how neuroscientific studies support the notion of a compassion How can the compassion pattern be
measured? What trait and state self-
pattern, and how the dynamic nature of the compassion pattern is reflected in practice. report scales can be developed for
such purpose? How can the interde-
Acknowledgments pendence of the aspects of the com-
A.R. was supported by grant 272/70 (Advancements on the aware mind-brain: new insights about the neural correlates of passion pattern be assessed?
meditation states and traits) from the BIAL Foundation (Portugal). S.M.A. gratefully acknowledges the support of the Paris
How do self-compassion and compas-
Institute for Advanced Study. S.G. benefited from support as a Visiting Professor at the Sapienza University of Rome. We
sion for others interact in compassion
thank the anonymous reviewers for their valuable comments and suggestions leading to an improved manuscript, and Dr
practice and compassion training?
Mirjam Hartkamp for useful comments.
What are the effects of an equanimous
Declaration of interests stance toward intentional objects of
The authors declare no conflicts of interests. compassion on other aspects and
processes of the compassion pattern?

References How do compassion meditation and

1. Trevarthen, C. (2014) Une sympathie innée: comment les bébés 14. Hall, J.A. and Schwartz, R. (2019) Empathy present and future. mindfulness meditation practices
établissent leurs relations émotionnelles avec les autres. In L'Empathie: J. Soc. Psychol. 159, 225–243 interact to support the training of
Au Carrefour des Sciences et de la Clinique: Proceedings of 15. Lamm, C. et al. (2019) Imaging empathy and prosocial emotions.
compassion? How do meditative
the 2011 Conference at Cerisy-la-Salle (Botbol, M. et al., eds), pp. Neurosci. Lett. 693, 49–53
47–75, Editions John Libbey 16. Stevens, F. and Taber, K. (2021) The neuroscience of empathy and insights about the self and social
2. Lipps, T. (1909) Leitfaden der Psychologie, Engelmann compassion in pro-social behavior. Neuropsychologia 159, 107925 interdependence support compassion
3. Gallese, V. (2001) The 'shared manifold' hypothesis: from mirror neu- 17. Decety, J. and Chaminade, T. (2003) Neural correlates of feeling training?
rons to empathy. In Between Ourselves: Second-Person Issues in sympathy. Neuropsychologia 41, 127–138
the Study of Consciousness (Thompson, E., ed.), pp. 33–50, Imprint 18. Batson, C.D. (1991) The Altruism Question: Towards a Social-
How are the dynamical brain–body–
Academic Psychological Answer, Erlbaum
4. Gilbert, P. (2020) Compassion: from its evolution to a psycho- 19. Goetz, J.L. et al. (2010) Compassion: an evolutionary analysis
environment processes of the com-
therapy. Front. Psychol. 11, 586161 and empirical review. Psychol. Bull. 136, 351 passion pattern integrated and how
5. Matos, M. et al. (2022) Compassion protects mental health and so- 20. Singer, T. and Klimecki, O.M. (2014) Empathy and compassion. do they correlate with neural activity?
cial safeness during the COVID-19 pandemic across 21 countries. Curr. Biol. 24, R875–R878 Do large-scale oscillatory synchroniza-
Mindfulness 13, 863–880 21. Engen, H.G. et al. (2018) Structural changes in socio-affective tion in the brain play a role for the inte-
6. Egan, S.J. et al. (2022) A review of self-compassion as an active networks: Multi-modal MRI findings in long-term meditation
gration of such processes? Can causal
ingredient in the prevention and treatment of anxiety and practitioners. Neuropsychologia 116, 26–33
depression in young people. Admin. Pol. Ment. Health 49, 385–403 22. Aronson, H. (1980) Love and Sympathy in Theravada Buddhism, density and information integration
7. Estrada, M. et al. (2021) Does emotional intelligence influence Motilal Banarsidass measures be successfully applied to
academic performance? The role of compassion and engage- 23. Dalai Lama (1995) The Power of Compassion, HarperCollins the compassion pattern? Do anterior
ment in education for sustainable development. Sustainability 24. Dalai Lama and Carriere, J.C. (1996) Violence and Compassion, insula and anterior cingulate cortex to-
13, 1721 Doubleday
gether play the role of a core hub or an
8. Patel, S. et al. (2019) Curricula for empathy and compassion 25. Gilbert, P. (2014) The origins and nature of compassion focused
training in medical education: a systematic review. PLoS One therapy. Br. J. Clin. Psychol. 53, 6–41
orchestrating role in the brain for the
14, e0221412 26. de Vignemont, F. and Singer, T. (2006) The empathic brain: how, compassion pattern? How do they in-
9. Sinclair, S. et al. (2021) What is the state of compassion educa- when and why? Trends Cogn. Sci. 10, 435–441 teract with each other? How do they
tion? A systematic review of compassion training in health care. 27. Izard, C.E. (1972) Patterns of Emotions: A New Analysis of relate to reward circuits, the autonomic
Acad. Med. 96, 1057–1070 Anxiety and Depression, Academic Press
nervous system, and hormonal modu-
10. Younas, A. and Maddigan, J. (2019) Proposing a policy frame- 28. Izard, C.E. and Ackerman, B.P. (2000) Motivational, organizational,
work for nursing education for fostering compassion in nursing and regulatory functions of discrete emotions. In Handbook of
lation? Is the anterior insula primarily
students: a critical review. J. Adv. Nurs. 75, 1621–1636 Emotions (2nd edn) (Lewis, M. and Haviland-Jones, J.M., eds), implicated in the empathy pattern,
11. Luyten, P. et al. (2020) The mentalizing approach to psychopa- pp. 253–322, Guilford Press and anterior cingulate cortex in the
thology: state of the art and future directions. Annu. Rev. Clin. 29. Newen, A. et al. (2015) Emotion recognition as pattern recognition: sympathy pattern?
Psychol. 16, 297–325 the relevance of perception. Mind Lang. 30, 187–208
12. Breyer, T. (2020) Empathy, sympathy and compassion. In The 30. Prinz, J. (2004) Gut Reactions: A Perceptual Theory of Emotion,
Routledge Handbook of Phenomenology of Emotion (Szanto, Oxford University Press
T. and Landweer, H., eds), pp. 429–440, Routledge 31. Frijda, N.H. (1986) The Emotions, Cambridge University Press
13. Cuff, B.M.P. et al. (2016) Empathy: a review of the concept. 32. Sparks, G.G. et al. (1999) The repressive coping style and fright
Emot. Rev. 8, 144–153 reactions to mass media. Commun. Res. 26, 176–192

514 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6

Trends in Cognitive Sciences
OPEN ACCESS

33. Goldie, P. (2000) The Emotions: A Philosophical Exploration, 61. Hein, G. et al. (2010) Neural responses to ingroup and outgroup
Clarendon Press members' suffering predict individual differences in costly helping.
34. Colombetti, G. (2014) The Feeling Body: Affective Science Meets Neuron 68, 149–160
the Enactive Mind, MIT Press 62. Bloom, P. (2017) Empathy and its discontents. Trends Cogn.
35. Merleau-Ponty, M. (2012) Phenomenology of Perception, Routledge Sci. 21, 24–31
36. Lazarus, R.S. (1991) Emotion and Adaptation, Oxford University 63. Bloom, P. (2017) Empathy, schmempathy: response to Zaki.
Press Trends Cogn. Sci. 21, 60–61
37. Goetz, J.L. and Simon-Thomas, E. (2017) The landscape of com- 64. Richins, M.T. et al. (2019) Empathic responses are reduced to
passion: Definitions and scientific approaches. In The Oxford competitive but not non-competitive outgroups. Soc. Neurosci.
Handbook of Compassion Science (Seppälä, E.M. et al., eds), 14, 345–358
pp. 3–15, Oxford University Press 65. Vaughn, D.A. (2018) Empathic neural responses predict group
38. Saturn, S.R. (2017) Two factors that fuel compassion: the oxytocin allegiance. Front. Hum. Neurosci. 12, 302
system and the social experience of moral elevation. In The Oxford 66. Koopmann-Holm, B. and Tsai, J.L. (2017) The cultural shaping of
Handbook of Compassion Science (Seppälä, E.M. et al., eds), compassion. In The Oxford Handbook of Compassion Science
pp. 121–132, Oxford University Press (Seppälä, E.M. et al., eds), pp. 273–285, Oxford University Press
39. Ebert, A. et al. (2018) Endogenous oxytocin is associated with 67. Davis, M.H. (1983) Measuring individual differences in empathy:
the experience of compassion and recalled upbringing in border- evidence for a multidimensional approach. J. Pers. Soc. Psychol.
line personality disorder. Depress. Anxiety 35, 50–57 44, 113–126
40. Weng, H.Y. et al. (2015) The role of compassion in altruistic help- 68. Lee, E.E. et al. (2021) Compassion toward others and self-
ing and punishment behavior. PLoS ONE 10, e0143794 compassion predict mental and physical well-being: a 5-year
41. Batson, C.D. and Shaw, L.L. (1991) Evidence for altruism: to- longitudinal study of 1090 community-dwelling adults across
ward a pluralism of prosocial motives. Psychol. Inq. 2, the lifespan. Transl. Psychiatry 11, 397
107–122 69. Pace, T.W. et al. (2009) Effect of compassion meditation on neu-
42. Dahl, C.J. et al. (2015) Reconstructing and deconstructing the self: roendocrine, innate immune and behavioral responses to psy-
cognitive mechanisms in meditation practice. Trends Cogn. Sci. chosocial stress. Psychoneuroendocrinology 34, 87–98
19, 515–523 70. Fredrickson, B.L. et al. (2008) Open hearts build lives: positive emo-
43. Gallagher, S. et al. (2024) The self-pattern and Buddhist psychology. tions, induced through loving-kindness meditation, build conse-
Mindfulness 15, 795–803 quential personal resources. J. Pers. Soc. Psychol. 95, 1045–1062
44. Desbordes, G. et al. (2014) Moving beyond mindfulness: Defining 71. Lutz, A. et al. (2008) Regulation of the neural circuitry of emotion
equanimity as an outcome measure in meditation and contem- by compassion meditation: effects of meditative expertise. PLoS
plative research. Mindfulness 6, 356–372 One 3, e1897
45. Wallace, B.A. and Shapiro, S.L. (2006) Mental balance and well- 72. Mascaro, J.S. et al. (2013) Compassion meditation enhances
being: building bridges between buddhism and western psychology. empathic accuracy and related neural activity. Soc. Cogn. Affect.
Am. Psychol. 61, 690–701 Neurosci. 8, 48–55
46. Nicolardi, V. et al. (2024) The two arrows of pain: mechanisms of 73. Förster, K. and Kanske, P. (2022) Upregulating positive affect
pain related to meditation and mental states of aversion and through compassion: psychological and physiological evidence.
identification. Mindfulness 15, 753–774 Int. J. Psychophysiol. 100–107
47. Kabat-Zinn, J. (1990) Full Catastrophe Living: Using the Wisdom 74. Klimecki, O.M. et al. (2013) Functional neural plasticity and asso-
of your Body and Mind to Face Stress, Pain and Illness, ciated changes in positive affect after compassion training.
Delacorte Cereb. Cortex 23, 1552–1561
48. Guendelman, S. et al. (2017) Mindfulness and emotion regula- 75. Karnaze, M.M. et al. (2022) Association of compassion and
tion: Insights from neurobiological, psychological, and clinical empathy with prosocial health behaviors and attitudes in a
studies. Front. Psychol. 8, 220 pandemic. PLoS One 17, e0271829
49. Kim, J.J. et al. (2020) The neurophysiological basis of com- 76. Leiberg, S. et al. (2011) Short-term compassion training
passion: An fMRI meta-analysis of compassion and its re- increases prosocial behavior in a newly developed prosocial
lated neural processes. Neurosci. Biobehav. Rev. 108, game. PLoS One 6, e17798
112–123 77. Diedrich, A. et al. (2014) Self-compassion as an emotion regulation
50. Novak, L. et al. (2022) Neural correlates of compassion – an inte- strategy in major depressive disorder. Behav. Res. Ther. 58, 43–51
grative systematic review. Int. J. Psychophysiol. 172, 46–59 78. Leary, M.R. et al. (2007) Self-compassion and reactions to un-
51. Medford, N. and Critchley, H.D. (2010) Conjoint activity of ante- pleasant self-relevant events: the implications of treating oneself
rior insular and anterior cingulate cortex: awareness and re- kindly. J. Pers. Soc. Psychol. 92, 887–904
sponse. Brain Struct. Funct. 214, 535–549 79. Scarlet, J. et al. (2017) The effects of compassion cultivation training
52. Menon, V. and Uddin, L.Q. (2010) Saliency, switching, attention (CCT) on health-care workers. Clin. Psychol. 21, 116–124
and control: a network model of insula function. Brain Struct. 80. Spreng, R.N. et al. (2009) The Toronto empathy questionnaire:
Funct. 214, 655–667 scale development and initial validation of a factor-analytic so-
53. Noriuchi, M. et al. (2008) The functional neuroanatomy of mater- lution to multiple empathy measures. J. Pers. Assess. 91,
nal love: mother's response to infant's attachment behaviors. 62–71
Biol. Psychiatry 63, 415–423 81. Caprara, G.V. et al. (2005) A new scale for measuring adults'
54. Hoehne, K. et al. (2022) Plasma oxytocin is modulated by mental prosocialness. Eur. J. Psychol. Assess. 21, 77–89
training, but does not mediate its stress-buffering effect. 82. Windmann, S. et al. (2021) Constructing the facets of altruistic
Psychoneuroendocrinology 141, 105734 behaviors (FAB) scale. Soc. Psychol. 52, 299–313
55. Quintana, D.S. and Guastella, A.J. (2020) An allostatic theory of 83. Juneau, C. et al. (2020) Reliability and validity of an equanimity
oxytocin. Trends Cogn. Sci. 24, 515–528 questionnaire: the two-factor equanimity scale (EQUA-S). PeerJ
56. Lutz, A. et al. (2004) Long-term meditators self-induce high- 8, e9405
amplitude gamma synchrony during mental practice. PNAS 84. Gross, J.J. and John, O.P. (2003) Individual differences in
101, 16369–16373 two emotion regulation processes: implications for affect,
57. Seth, A.K. et al. (2011) Causal density and integrated information as relationships, and well-being. J. Pers. Soc. Psychol. 85,
measures of conscious level. Philos. Trans. A Math. Phys. Eng. Sci. 348–362
369, 3748–3767 85. Ardelt, M. (2003) Empirical assessment of a three-dimensional
58. de Vignemont, F. and Jacob, P. (2012) What is it like to feel an- wisdom scale. Res. Aging 25, 275–324
other's pain? Philos. Sci. 79, 295–316 86. Baer, R.A. et al. (2006) Using self-report assessment methods to
59. Jacob, P. (2011) The direct-perception model of empathy: a cri- explore facets of mindfulness. Assessment 13, 27–45
tique. Rev. Philos. Psychol. 2, 519–540 87. Petrocchi, N. et al. (2019) Self-reassurance, not self-esteem,
60. Darwall, S. (1998) Empathy, sympathy, care. Philos. Stud. 89, serves as a buffer between self-criticism and depressive symp-
261–282 toms. Psychol. Psychother. 92, 394–406

Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6 515

 Trends in Cognitive Sciences
OPEN ACCESS

88. Thompson, R. and Zuroff, D.C. (2004) The levels of self-criticism anxiety disorder and healthy controls. Anxiety Stress Coping
scale: comparative self-criticism and internalized self-criticism. 31, 594–609
Personal. Individ. Differ. 36, 419–430 92. Gill, L.N. et al. (2020) Mindfulness induction and cognition: a
89. Sullivan, M.J.L. et al. (1995) The pain catastrophizing systematic review and meta-analysis. Conscious. Cogn. 84,
scale: development and validation. Psychol. Assess. 7, 102991
524–532 93. Runyan, J.D. et al. (2019) Using experience sampling to examine
90. Condon, P. and Feldman Barrett, L. (2013) Conceptualizing and links between compassion, eudaimonia, and pro-social behavior.
experiencing compassion. Emotion 13, 817–821 J. Pers. 87, 690–701
91. Arch, J.J. et al. (2018) Self-compassion induction enhances 94. Rodriguez, K. et al. (2023) Event-related correlates of compas-
recovery from social stressors: comparing adults with social sion for social pain. Soc. Neurosci. 18, 91–102

516 Trends in Cognitive Sciences, June 2024, Vol. 28, No. 6