The Functions of Safety in Psychotherapy: An Integrative Theoretical Perspective Across Therapeutic Schools

Abstract

Objective

There is a certain consensus in the psychotherapeutic literature that safety plays a central role in human development and psychotherapy and that lack of safety undermines mental health. However, the role of safety in psychotherapy has not yet been thoroughly examined. In this article, we identify and integrate the diferent functions of safety in psychotherapy on a theoretical basis.

Method

We made a panoramic overview of the concept of safety across some of the main psychotherapeutic schools that represent major paradigms in contemporary psychotherapy (psychodynamic, cognitive-behavioral, humanistic). We then analyzed, compared, and synthetized the findings to identify the common functions that safety plays both in ontogenesis and in clinical practice across different therapeutic orientations.

Results

Our analysis showed that safety is indeed rightly prioritized across psychotherapy schools because of its developmental value in promoting change and adaptation both in ontogenesis and clinical settings. The findings suggest that the main functions of safety are to secure survival, facilitate restoration, promote exploration, sustain risk-taking, and enable integration, with these functions being complementary and dependent on the context. However, safety seems to be in a dialectical and paradoxical relationship to psychotherapy and human development. Adequate ontogenetic development and treatment progress do not appear to require continuous maintenance of maximum possible safety. Rather, they seem to require enough safety, adequately and timely modulated according to developmental needs and treatment phases.

Conclusions

Although safety provides the necessary basis that enables restoration, fuels exploration, and facilitates treatment progress, safety’s misdosage (e.g., lack, excess), misconstruction (e.g., misattunement, misinterpretation), or misuse (exploitation, idealization) may hinder the healthy development of attachment, identity, autonomy, self/co-regulation as well as the ability to tolerate and cope with dangers, risks, insecurities, or frustrations. Future research is suggested to further explore the role of safety in psychotherapy.

The importance of an individual’s sense of safety in psychotherapy seems to be, at least implicitly, widely recognized in the field of psychotherapy. Safety is considered to permeate the human brain and behavior largely at an automatic and implicit level (Porges, 2020, 2021; Schore, 2003). Moreover, several authors have shown that an inadequate experience of safety during early human development has a significant correlation with psychopathology (Cassidy & Shaver, 2016; Gilbert, 2006; Schore, 2003). Finally, the possibility of experiencing adequate safety during the treatment is considered an important element of the psychotherapeutic process and its outcome (Norcross & Lambert, 2019).

Unfortunately, the psychological and clinical sciences suffer from conceptual and terminological fragmentation (Salvatore, 2011). Probably for this reason, although several authors have recognized the importance of safety in psychotherapy, the scholarly literature still seems to lack a more coherent and consistent account of it. In fact, it seems that what is meant by safety, how it is explained, to what extent and why it can be relevant in psychotherapy, how it functions, and through which mechanisms it impacts the clinical process is still fuzzy. As such, the concept of safety is far from homogeneous and remains subject to a certain variability due to the variety of theoretical perspectives and orientations characterizing the different psychotherapeutic schools (Podolan, 2020). Thus, developing a more integrated and coherent understanding of safety in psychotherapy requires a panoramic overview taking into account different psychotherapeutic approaches.

A comprehensive and systematic review of safety – including the coexistence of its various meanings, underpinnings, functioning mechanisms, impacts, and types (e.g., emotional, psychological, social, cognitive, and behavioral) – across the array of different academic sciences is beyond the scope of this article. Rather, the present paper represents an initial attempt to provide an integrative theoretical perspective on safety in psychotherapy concerning its functions in both ontogenesis and clinical practice. To this end, we limited ourselves to three different psychotherapeutic approaches by referring to the literature of the main authors of each who have provided references and/ or insights into the role of safety in ontogenesis and/ or clinical practice: psychodynamic approach (e.g., classical psychoanalysis, ego psychology, object-relations theory, relational and intersubjective psychoanalysis), cognitive-behavioral approach (e.g., acceptance and commitment therapy, dialectical-behavior therapy, mindfulness-based cognitive therapy, compassion-focused therapy), and humanistic approach (e.g., Gestalt therapy, existential therapy, person-centered therapy, emotion-focused therapy). Aware that these approaches do not exhaust the vast range of psychotherapy models, we selected them because we agree they represent some of the major paradigms in contemporary psychotherapy and, therefore, may be appropriate for the initial proposal of an integrative theoretical account (for other relevant nonsystematic reviews focusing on these three main therapeutic approaches, see Castonguay & Hill, 2012; de Jong & DeRubeis, 2018). Moreover, focusing on approaches that are paradigmatically different (see Ford & Urban, 1998) – notwithstanding the possible commonalities between them (e.g., Wampold, 2015) – may provide a better ground to provide an integrative theoretical account of the functions of safety in psychotherapy.

In the first part of the paper, we provide a panoramic overview of how these major psychotherapeutic approaches have addressed the concept of safety concerning its psychological role and clinical value. Then, we draw on this overview to identify some basic and common functions that safety may play across these different therapeutic orientations in both ontogenesis and clinical practice. We conclude by suggesting future lines of research.

Psychodynamic Approaches

From a psychodynamic perspective, safety may be defined as a sense of physical and emotional well-being, free from the pressure of need and anxiety (Greenberg, 1991). Psychodynamic approaches were the first to emphasize the primary role of safety in the ontogenesis and therapy of the psyche (Podolan, 2020). In terms of the ontology of the psyche, Freud was the first one to assert that “the ego is governed by considerations of safety” (Freud, 1966; p. 199). Since then, Alfred Adler’s safeguarding tendencies, Anna Freud’s defense mechanisms, Karen Horney’s safety devices, Joseph Sandler’s background of safety, and Harry Sullivan’s security operations have referred to basic mechanisms of safety in different ways, as to overriding aspects of the psyche that secure its survival, protect it against dangerous and unbearable experiences, organize its defensive and coping mechanisms, and manage all of its perceptions, pleasures, and fantasies. Safety has also been associated with affective, motivational, and behavioral systems that seek, maintain, or use various forms of safety to maintain homeostasis, secure survival, provide restoration, enable exploration, promote resilience, and enable growth (Cassidy & Shaver, 2016; Lichtenberg et al., 1996; Panksepp & Biven, 2012).

Psychodynamic ontogenesis and clinical practice prioritize safety as well. Imre Herman’s research of primate’s clinging instinct or William Blatz’s developmental security theory led Mary Ainsworth (1985) and John Bowlby (1988) to introduce the concept of attachment, which represented a paradigm shift in our understanding of our early needs for safety and their role in ontogenetic development. Bowlby and Ainsworth have convincingly demonstrated that attachment proximity to the caregiver is not only a primal need but also a critical source of an infant’s safety (Ainsworth, 1985; Bowlby, 1998). This is because the infant’s attachment relationship with the caregiver secures protection (safety) against hunger, disease, predators, and other dangers.

They showed that the infant is most vulnerable and exposed to danger during the earliest stages of development and that attachment aims to safely protect the infant by providing the infant with safety during these early times. It became clear that untimely and inadequate provision of safety (e.g., too little, too late, too short) during these most critical developmental stages seriously undermines healthy ontogenesis. Sroufe and Waters (2017) later clarified that the goal of any type of attachment is for the infant to survive by feeling secure. Moreover, it was found that the infant is more likely to feel safe with a mother classified as continuous-secure or earned-secure than with a mother classified as insecure (Saunders et al., 2011).

Simultaneously, safety has also been recognized as a precondition for effective psychodynamic psychotherapy. For a good therapeutic process it is essential that clients feel safe and trust the therapist. This can be better achieved when the therapist has a secure attachment style (e.g., Levy et al., 2018). Additionally, maintaining confidentiality and therapeutic boundaries is crucial for creating a safe environment (Gabbard, 2016). Furthermore, the therapist’s ability to pass clients’ tests (Rappoport, 1997; Siegel & Hilsenroth, 2013) and deactivate their defenses (Frederickson, 2020) is key to enabling clients safely disclose and explore potentially dangerous material. Thus, psychodynamic literature emphasizes the role of safety in terms of survival, human development, and effective psychotherapy.

It is important to note that psychodynamic literature highlights that safety is linked not only with survival, homeostatis, and defense, but also with soothing and restoring a distressed organism through a secure relationship. The interpersonal aspect of safety that soothes, calms, and heals has been, for example, addressed by Donald Winnicott (1965). His concept of the holding environment refers to the mother-infant relationship as a symbolic union where safety is provided not only through the protection but also through her love, support, stability, continuity of being as well as the integration of the infant’s inner emotional and cognitive processes.

Later, Ainswowrth (1985) used the term secure attachment to characterize loving, trustful, warm, soothing, sensitive, and reliable care, correct interpretation of the infant’s signals, and prompt, timely, and appropriate responding to his or her needs (Ainsworth, 1985). Psychodynamic theories developed a plethora of relational concepts that encompass various forms of relational safety (e.g., containment, basic trust, holding environment, extra-uterine matrix). These concepts go beyond just protecting the infant and also include recognizing, reciprocating and holding the infant’s needs in the mind of the caregiver, establishing trust and well as organization, integration, consolidation, naming, and symbolization of his or her experiences. Additionally, psychodynamic theories emphasize the importance of soothing, calming, co-regulating the infant’s inner processes as well as providing inner and outer boundaries to encourage exploration and growth. Therefore, from a psychodynamic perspective, safety not only protects but also establishes trust and soothes, calms, restores, and heals a distressed organism.

An additional important psychodynamic aspect that is inseparably interconnected with safety is the concept of exploration. Attachment theory posits that infants only become able and motivated to seek and explore the outer and inner world freely after experiencing sufficient safety. This allows them to do so without having to rely on defense mechanisms. Exploratory activity not only enables survival (e.g., through seeking food and shelter) but also facilitates the expansion of existing functions and growth (e.g., by gaining new information or learning new skills). Here, the caregiver functions primarily as a secure base (Bowlby, 1988), which provides emotional fuel for the infant’s exploratory endeavors, as well as a beacon of orientation (Mahler et al., 1975), which provides reassurances for new directions and more specific work amid uncertainties and painful emotions.

Psychodynamic theories warn against idealization and devaluation of safety. They posit that psychotherapy may be not effective when a client feels too unsafe (hyper- or hypoarousal states with the fight-flight-freeze system being activated) or too safe (comfort state with no emotionally distressing experiences and dysregulated arousal states) (Ogden, 2009). For Bromberg (2006), the enhancement of resiliency and expansion of affect tolerance requires that therapy is “safe but not too safe” (Bromberg, 2006, p. 4). In this respect, Allan Schore (2003) coined the term regulatory boundaries (also called the window of affect tolerance) to describe an optimal arousal zone within which emotions can be effectively experienced, processed, and integrated and within which progress and growth occur. Psychotherapy should provide experiences of safety to encourage both client and therapist to experience optimal frustrations and tolerable disappointments (Kohut & Orstein, 2011) while remaining in a position of safe uncertainty (Mason, 2015). The therapeutic dyad should also engage in certain levels of tension and danger to achieve therapeutic progress and change because lingering in safety may be problematic (Eldridge, 2018). In psychodynamic approaches, the client and therapist in therapy are encouraged to take acceptable risks and to avoid devaluing/idealizing safety. Safety does not appear to be therapeutic when it is misconstrued (e.g., perceived as overconfidence, overcontrol, overindulgence), underdosed (e.g., neglect, ignorance), overdosed (e.g., pampering, over-nurturing), mistimed (e.g., missed or applied at an inappropriate time), exploited (e.g., used to satisfy one’s own needs at the expense of others), misused (e.g., employed to avoid socialization/commitment/ risk-taking) or perplexed with other non-therapeutic experiences (e.g., tiptoeing around the risks, surfing on top of vulnerabilities, coasting through transferences or countertransferences). Thus, after gaining sufficient safety, psychodynamic clients are encouraged to avoid lingering in or clinging to safety and emotional comfort. This allows them to develop better pain, risk, ambiguity, frustration, or affect tolerance (Dufourmantelle, 2019; Ogden, 2009; Podolan, 2020; Segalla, 2018).

In summary, the psychodynamic perspective underlines the relevance of such a level of safety that fuels both parties with trust. This sense of safety encourages them to face challenges, cope with uncertainties, tolerate frustrations, take risks, and explore vulnerable, threatening, or unknown domains of experience.

Another no less important aspect of safety addressed by psychodynamic literature relates to the integration of the self. On an unconscious level, integration of the self is thought to occur, partly through dreams which serve as a safe environment, safe place, safety valve, or secure base for our overburdened brain. The safety of dreaming is crucial as it enables our psyche to (i) creatively and playfully contain and regulate our emotions and integrate our daily experiences into stabile self-image (Hartmann, 1995), (ii) weave in new material into similar experiences that feel the same way to preserve our emotional memory (Payne, 2010), or (iii) adapt to trauma, stress, and the problems of life (Sørensen, 2018). In terms of the early development of the self, Kohut similarly showed that the cohesion and integration of the infant’s emerging self requires a great deal of safety provided through interpersonal mirroring, idealizing, twinship, and sustenance (Kohut & Orstein, 2011). These experiences contribute to the development of internal ideals and feelings of cohesiveness, wholeness, consistency, resilience, and coherent image of oneself (inclusive of any fragmented parts) that are gradually integrated and internalized, thus forming the foundation of the infant's sense of security (Kohut & Orstein, 2011). Another safety-related concept that facilitates the integration of the self is mentalization (Fonagy et al., 1991). When being mentalized by the caregiver (e.g., thought about, felt, understood, and recognized), the child feels safer because mentalization fosters the child’s ability to think, understand, differentiate, organize, and consolidate experiences about oneself, others, and the world (Fonagy & Allison, 2014). The safety of the interpersonal integration of the self could also be characterized by predictable interactions marked by rhythmic attunement (Kestenberg), affective resonance (Stern), moments of meeting (Sander), synchronous interactions (Levy, Tronick), or directional fittedness (Boston Change Process Study Group). Siegel (2010) further argued that the mind’s safety and neural integration arise from not only interpersonal but also intrapersonal attunement and integration (ability to perceive the mind of oneself and another), which he called mindsight. Thus, in essence, dreaming as well as the therapist’s mirroring, mentalizing, and mindsight seem to support the development of the client’s capacity to safely observe, organize, and integrate one’s own self with regard to the self of another person.

According to psychodynamic theories, safety is viewed as means of regulating, balancing, and integrating psychic experiences within relationships. Such approaches share the perspective that safety plays a vital role in ensuring psychic survival (e.g., through defenses, coping mechanisms, or security operations), facilitating healing and restoration (e.g., through secure attachments or attuned/synchronized relationships), promoting exploration (e.g., by creating a secure base, secure boundaries, and deactivation of client’s defenses), sustaining risk-taking (e.g., by sensitive interventions and fluid oscillations between danger and safety), and/or enabling integration (e.g., through mirroring, mentalization, or mindsight). To create safety in therapy, the therapist must first establish empathic holding, trust, mirroring, attunement, containment, mentalizing, prompt and appropriate responding, correct interpretation of the client’s signals, and observation of therapeutic boundaries. Only if enough safety has been timely created within the therapeutic relationship may the parties explore the client’s vulnerable areas and engage in more risky and painful areas of the client’s life to promote transformative and integrative processes.

Cognitive-Behavioral Approaches

Approaches from cognitive-behavioral therapy (CBT) rely heavily on evolutionary and neurobiological research indicating that the human brain and mind developed on a continuum between defense and safety (Gilbert, 2004, 2006). One of the leading contemporary theories on safety – the polyvagal theory – postulates that one’s sense of safety is perceived and regulated through bidirectional (top-down and bottom-up) neural pathways among specific brain regions and peripheral areas, whereby the vagal complex plays a crucial role in regulating and maintaining safety by inhibiting neurobiological fight-flight-freeze mechanisms and fostering social exploration and learning (Porges, 2021). Thus, the detection and differentiation between safety or danger cues occur without awareness mainly through our sensory channels, including auditory, kinesthetic, organic, visual, gustatory, olfactory, cutaneous, and vestibular channels (Gilbert, 2004, 2006; Porges, 2021). Recent research on safety in psychotherapy provides evidence that the therapist’s non-sensory signals (e.g., language and words) seem to elicit less safety in the client than the therapist’s voice intonation and other non-verbal signals, such as posture, facial expressions, or eye contact (Mair, 2021).

Experiences of safety within relationships seem to have been necessary for brain development (Allison & Rossouw, 2013; Gilbert, 2004, 2006) and for the brain’s capacity to detect safety through any relationships that reduce threats and provide a certain form of well-being (Porges, 2021). Such relationships may provide safety through attachment (protection and care through bonds), domination (power over subordinates), submission (protection by dominant ones), competition (knowing each other’s strengths and weaknesses), herding (protection through being part of a larger group), and cooperation (sharing of resources and co-regulation of aims) (Ivaldi, 2016). From an evolutionary and relational perspective, the detection and sense of safety in relationships seem to have been the primary goal in the course of human evolution.

It should also be noted that CBT has been increasingly integrating and using various relational aspects of safety (e.g., empathy, unconditional positive regard, cooperation, validation, mentalization, attunement, presence, and congruence) from various other psychotherapy schools. It considers these different facets of safety (and their differentiation from danger) to be central to effective therapy and the related work with clients’ distorted cognitions and emotions (Bennett-Levy et al., 2015). Concerning danger and safety, CBT differentiates between the perception of danger signals (cues indicating that a dangerous event may occur) and/or safety signals (cues indicating that there is no threat or that an aversive event will not occur) (Lohr et al., 2007). CBT also developed the concept of safety behaviors that – analogously to defense mechanisms – refer to those behaviors intended to detect, avoid, escape, neutralize, or reduce fear or anxiety (Hayes & Hofmann, 2018). In this context, CBT therapies strive to create safety in therapy by deactivating defense mechanisms (Bryant, 2006), satisfying the client’s attachment or control needs (Epstein, 1998), acknowledging and accepting unpleasant feelings (Hayes et al., 1999), and promoting self-soothing (or the elicitation of soothing from others) and regulation (or co-regulation) of inner states (Porges, 2021). Clients’ senses of safety and control can be facilitated through various techniques including psychoeducation, creating a safe place, sustaining hope, utilizing structured cognitive techniques (e.g., exposure techniques; Hezel & Simpson, 2019), sourcing of personal qualities, mindfulness of bodily sensations (Sipe & Eisendrath, 2012), giving advice, providing assistance of some sort or another, offering encouragement, or various forms of enhancement of client’s personal, social, or economic resources (Norcross & Lambert, 2019; Rappoport, 1997).

Through relational safety, CBT clients are encouraged to engage in vulnerable behaviors, be more agentic and authentic, voice disagreements, accept and normalize the expression of negative and positive emotions, or directly discuss and address dysfunctional interactions or escalation of conflicts.

CBT approaches also recognize that safety not only reduces defenses but also promotes the activation of non-defensive exploratory behaviors. Paul Gilbert (2004) showed that the sense of safety may be misunderstood and misinterpreted with states associated with safety behaviors and corresponding fight-flight-freeze and other defensive strategies. When these defensive strategies are successful, anxiety is typically reduced. The resulting physical and psychological relief may be misinterpreted as therapeutic or safe. However, Gilbert (2004) clarified that a true feeling of safety does not trigger or depend on safety behaviors. For this reason, he introduced the term safeness behaviors, which refers to curious exploration with relaxed, open, and non-defensive attention without the need to rely on safety behaviors. The work of Bennett-Levy et al. (2015) added that the feeling of safety promotes, in addition, self-exploration and self-reflection, which are likewise considered essential for psychotherapy.

According to Gilbert (2004), an inner sense of safety promotes an individual’s ability to engage in vulnerable activities, take risks, face dangers, and develop more complex repertoires for thinking and behavior. Lohr et al. (2007) similarly suggest that the gradual coping with danger signals – where one feels safe enough to cope with danger (as well as the reduction of excessive reliance on safety signals) – contributes to the reduction of pathological fears, dependency, isolation, or energy consumption. Both these authors warned against a continuous or excessive reliance on safety signals (e.g., familiar places or persons, cellular phones) because it may contribute to the maintenance and exacerbation of anxiety, as well as to the development of various maladaptations that employ defensive mechanisms even in non-threatening situations. CBT clients seem to achieve changes and growth if their contact with danger occurs within their safety zone, which is represented by an individual’s ability to regulate and control his or her responses to threats (Freeman & Dolan, 2001). In essence, CBT approaches teach clients to cope with threats, insecurities, and uncertainties, to take acceptable risks, to reduce over-reliance on safety behaviors, and to assume that situations are safe unless there is clear evidence to the contrary.

Cognitive-behavioral literature also posits that our sense of safety arises from a cognitive process called security priming (Baldwin, 2007). Security primes may include not only memories but also pictures, portrayals, text messages or images, or subliminal representations of available and supportive attachment figures (Cassidy & Shaver, 2016). The security primes are activated through techniques such as guided imagery and visualization of a safe place, mindfulness, or other methods that all allow clients to draw upon or think about mental states (of oneself and others) as well as fantasies about union with another or memories of unconditional social connection (Baldwin, 2007).

Security priming enhances the integration of the self, including how we see ourselves, others, and our relationship with them (Rowe & Carnelley, 2003). Priming with mental representations of security-enhancing attachment figures also fosters a person’s sense of security, activates a sense of attachment security, helps to understand internal working models, improves mood, and has a calming and soothing effect (Gillath & Karantzas, 2019). In particular, psychotraumatologists postulate that an infant/client needs first to be provided with sufficient safety to become resilient and able to successfully process painful (traumatic) experiences and integrate them into a new and more adaptive self (Shapiro, 2017). An internalized sense of safety is also needed for the integration of opposites (e.g., dynamic synthesis of emotional and cognitive processes) or to resolve internal conflicts (Linehan, 2015). In psychotherapy, therapists teach clients how to prime their sense of security in order to feel safer, to down-regulate their hyperarousal, up-regulate their hypoarousal, enhance their positive affects, facilitate their self-efficacy through the self-regulation of thoughts, emotions, and behaviors. Additionally, therapist may help clients to co-construct narratives that support the discovery of new meaning and purpose, as well as consolidation of the self (see Castonguay & Hill, 2012). In summary, cognitive-behavioral approaches – especially those with a specific trauma-focused orientation – see safety as a need enrooted in the nervous system that ensures survival, enables restoration, and facilitates exploration and development. Safety is experienced through co-regulating relationships (developing through the priming of various security-related relational aspects). Being safe also requires psychoeducation (through understanding and oversight, “I feel to have power”), self-efficacy (“I trust in my qualities and abilities to cope with danger”), and the development of the ability to differentiate between safety and danger or between safeness and safety behavior. Cognitive-behavioral approaches seem to integrate various safety-related constructs (e.g., attunement, unconditional positive regard, presence) into their theories. They prioritize the creation and maintenance of safety sources during the process of psychotherapy, in particular when working with traumatized clients (Farina & Liotti, 2013). These modalities also warn against a client’s over-reliance on safety signals and encourage clients to foster inner safety by exposition to and coping with danger signals (see also Schimmenti et al., 2020).

Humanistic Approaches

Within humanistic approaches, safety is defined as a need or desire for a secure, familiar, and predictable environment where one is free from illness and danger (Maslow, 1943). Humanistic therapies broadened our understanding of safety. For example, Giddens (1991) developed the term ontological security to refer to a sense of self-identity, a sense of order, a sense of continuity in everyday events, a belief in the continuity of the world, the confidence in the social order, a capacity to find meaning in our lives and the belief that self-realization can be achieved. Eric Erikson (1993) used the term ontological security to denote existential trust in the continuity of relationships with significant others. For Laing (1960), ontological security means the ability to cope with life without the loss of a sense of existence and reality.

Maslow (1943) prioritized safety needs (together with physiological needs) before social and other needs. He proclaimed that “everything looks less important than safety, even sometimes the physiological needs” (p. 376). Importantly, like other authors from different orientations (e.g., Freud, Sandler, Gilbert, or Porges), Maslow (1943) was also convinced that our needs and perceptions of safety “serve as the almost exclusive organizers of behavior, recruiting all the capacities of the organism in their service” (p. 376). Humanistic therapies also recognize that the scope and quality of an infant’s sense of safety develop through contact at the boundaries of the self in the present meetings or dialogue of me-you (becoming “me” through “you”) (Buber, 1958). Ludwig Binswanger (1963) underlined that healthy developmental interactions constitute a dual mode of love, where the love relationship develops our sense of safety and serves as the basis for our further growth and change. In essence, for humanistic psychotherapy, ontogenetic development proceeds primarily through the interaction between the infant and the world. The infant’s feelings of safety are formed in the here and now and in the meetings of minds where the infant is not only affirms his/her own presence and identity but also remains receptive to the presence and identity of another person (Schneider, 2016). The humanistic approach holds that infants require intense experiences of security to develop a healthy adult relationship with the outer and inner environment.

Just like psychodynamic and cognitive-behavioral approaches, humanistic therapies emphasize the role of safety not only in ontogenesis but also in clinical practice. According to existential therapist Irvin Yalom (2002), “nothing takes precedence, I emphasize, over the importance of the patient’s feeling safe in the therapy office and the therapy hour” (p. 189). Angus and colleagues have shown, for example, that one of the most important factors in humanistic therapy consists of the therapist’s ability to build a client’s sense of safety, underlying how slippery and fragile such a sense of safety is (Angus et al., 2015). Therefore, just like psychodynamic and CBT approaches, humanistic approaches also recognize that the sense of safety is primarily provided through relationships. In this respect, it should be noted how humanistic approaches impacted our general conceptualization of psychotherapy by defining certain relational qualities which create or enhance the client’s experiences of safety. Rogers (1995) postulated that psychological safety in psychotherapy arises from an empathic, genuine, authentic, congruent, and non-judgmental therapeutic stance in which the therapist accepts the client recognizing his or her unconditional worth as a person. Humanistic therapists enhance clients’ safety also through an attuned, supportive, kind, and respectful relationship in which the therapists act for the client as a human sanctuary and as a safe container (Wheeler & Axelsson, 2015). In terms of the therapeutic relationship and how therapists can foster a sense of safety for their clients, humanistic approaches have introduced the concept of presence (Geller & Porges, 2014). This concept involves the therapist utilizing their entire being to remain fully engaged and attuned to the client in the present moment. The concept of presence may be compared to the concepts of a secure base, holding environment, or safe place. In other words, the therapists are present when they can listen “without memory or desire” and modulate their “automatic pilots” (i.e., viewing the clients per old automated and processed memory) to adequately attune and focus on the present relational moments with their clients.

Just like in psychodynamic or CBT therapies, humanistic approaches also recognize that a sufficient sense of safety that has been gained through relationships triggers human activity toward the exploration of the inner and outer world (Schneider et al., 2014). Gestalt therapists also proclaim that “only with the assurance of absolute safety within treatment can the patient feel secure enough to talk about his most private and upsetting thoughts” (Levin & Gunther, 2004, p. 58). Thus, safety is recognized as a platform that enables and promotes the disclosure, acceptance, and exploration of clients’ emotions (Greenberg, 2004).

However, unlike other approaches, the humanistic approaches rely on a non-directional concept of invitational mode, which postulates that the therapist “always invites and never insists that the client explores certain aspects of the experience” (Schneider et al., 2014, p. 526). In other words, the invitational mode invites (does not require) psychological intimacy because it values and respects the client’s need for safety (including his or her resistance to exploring vulnerabilities). The client’s acceptance (or refusal) to accept the invitation to explore can be seen as an important milestone indicating that the relationship is safe enough (or not yet safe enough) to explore certain vulnerable material. For this reason, humanistic therapists prefer to attend to or respect clients’ defenses and resistances until they develop a sufficient sense of safety (Leitner & Celentana, 1997). Other types of therapies may use different techniques, such as confronting the client and attempting to deactivate their defenses. However, humanistic techniques seem to share the same goal of making the clients feel safe enough (in one way or the other) to promote exploration or play and facilitate the development of their true selves. Concerning exploration and safety in psychotherapy, humanistic literature also uses the concept of safe emergency (Perls & Andreas, 1969), positing that psychotherapy is effective if it employs and utilizes both exploration and challenges for growth (danger) on one side and guidance and support (safety) on the other (Cozolino, 2002). The concept of safe emergency is similar to the psychodynamic concept of necessary danger (Carr & Sandmeyer, 2018; see also the systemic idea of safe uncertainty; Mason, 2015). It posits that during the co-construction of the relationship, therapists must not only build a client’s sense of safety but also use the client’s safety in a dialectic relationship with danger so that the client may eventually achieve changes and progress (Cozolino, 2002). In other words, while too much arousal and stress may activate defenses and inhibit optimal cortical processing involved in exploration, too little stress and arousal may lead to insufficient stimulation or focus required to take in new information (Cozolino, 2002). In gestalt approaches, both empathy and support (safety) as well as frustrations and confrontations (danger) are valued positively as they are deemed necessary for optimal development and growth. However, Perls cautioned against excessive focus on safety for emotional convenience, which he believed could impede clients from taking risks and developing effective coping mechanisms for life’s challenges. (Perls & Andreas, 1969). According to Gestalt theories, after the client has been dosed or fueled with sufficient safety, his or her free and creative self is expected to emerge and begin to choose how to cope with life’s surprises, insecurities, or dangers. In this respect, Rogers (1995) similarly emphasized that psychological safety allows the client to move toward self-actualization, creativity, and self-development. Moreover, the promotion of clients’ external and internal sources of safety helps clients to integrate their selves through a sense of identity, aliveness, autonomy, and boundaries (Birtchnell, 2002).

In summary, humanistic therapies identify safety as a need that organizes all human behavior and as a belief in the continuity and predictability of the world that arises from relationships, routines, and experiences. A person feels safe to the extent he or she can existentially trust significant others and the continuity of the inner and outer world. The most important aspects of personal safety – identity, autonomy, and firm boundaries – are created through present meetings of minds (such as the me-you dialogue at the boundaries of the self) within a relationship that is characterized by unconditional acceptance, empathy, presence, congruence, and the invitational mode. Similarly, like psychodynamic and cognitive-behavioral approaches, humanistic therapies also posit that safety is enhanced and maintained through the assimilation of new experiences arising from bearable safety-danger encounters (i.e., coping with tolerable risks and frustrations through a warm, accepting, and supportive relationship).

Safety and Psychotherapy: Basic Developmental Functions across Psychotherapeutic Schools

Our analysis reveals that safety is an articulated and complex concept, which has as many facets and functions as we consider different schools of psychotherapy. In ontogenesis and psychotherapy, safety is being described differently through various concepts and related functions that are both non-defensive (e.g., relational co-regulation, the window of affect tolerance, safeness behavior, internalized secure base) and defensive (safe zone, psychic retreat, safety behavior, safeguarding tendency, defense mechanisms). In the previous sections, we have seen that different psychotherapeutic approaches tend to define and use safety differently. At the same time, we believe it is possible to identify some common functions that safety plays both in ontogenesis and in clinical practice and that cut across different therapeutic orientations. These school-independent functions are described in table 1.

Table 1.

School-independent developmental functions of safety in ontogenesis and clinical practice

Securing survival	Facilitating restoration	Promoting exploration	Sustaining risk-taking	Enabling integration
Ontogenesis. Humans seek homeostasis, survival, and safety through evolutionary selected neural mechanisms, behaviors, and relationships with primary caregivers (however functional or dysfunctional).	Ontogenesis. The human resilience and ability to gain or restore inner safety evolve through the quality of the relationship with primary caregivers (however adequate or inadequate).	Ontogenesis. The human capacity to explore is promote by safety gained or restored through proximal and responsive (internalized) relationships with caregivers who concurrently actively promoted authenticity and exploration (however large or small).	Ontogenesis. Human ability to take risks and face danger is sustained by having constructively experienced safety in a dialectic with danger. This succeeds in renewing inner organismic balance in times of disequilibrium (e.g., challenging and conflicting experiences).	Ontogenesis. The human ability to assimilate new experiences and reorganize old ones, reform boundaries, and update identity in an integrated and coherent way requires safety provided by dreaming, mirroring, narrating, identifying and recognizing, and differentiating self from non-self.
Clinical practice. The therapist’s primary focus is to establish a safe environment and safe haven by setting therapeutic boundaries, attuning to the client, providing acceptance, holding, and containment (mainly through supportive interventions) during experiences of danger. This establishes trust, cooperation, and confidentiality enabling the client to better tolerate those experiences and feel less defensive within the therapeutic relationship. Reduced defenses promote mentalization and self- and co- regulation, thus fostering a good enough relationship (alliance building).	Clinical practice. The therapist further provides a safe environment and functions increasingly as a safe haven. Mainly through supportive interventions the therapist creates a safe enough relatiohsip marked by empathy and mentalization, which soothes and calms the client during experiences of danger. This process repairs alliance ruptures and stabilizes the client’s sense of safety, allowing them to internalize therapy as a secure base. As a consequence of this, the clients can heal and develop resilience, mentalizing abilities, and the ability to face danger without excessive reliance on defenses.	Clinical practice. The therapist takes advantage of having become a secure base to increasingly serve as a beacon of orientation and actively invite and promote the client’s exploration. This is done at a behavioral, emotional, and/ or cognitive level mainly through change-oriented interventions but still against the background of a safe- enough and supportive relationship. This further promotes the client’s curiosity, creativity, mentalization, and self- and interpersonal exploration and regulation.	Clinical practice. The therapist takes advantage of having become a safe haven and secure base to create and manage the conditions for transformative safety. This is done through an adequate and dialectic balance between supportive and change-oriented interventions against the background of a safe-enough relationship. This enables the client (i) to dialectically experience and cope with danger within an overall safe relationship and (ii) to co-regulate and assimilate new and conflicting experiences (corrective emotional experiences). As a consequence, it enhances the client’s mentalization, self-efficacy, self- and co-regulation, affect tolerance, resilience, sense of reality, and the ability to tolerate danger, cope with stress and challenges, and value risk-taking.	Clinical practice. The therapist creates conditions of integrative safety amidst internal and external stressors. This is achieved mainly through supportive interventions (dreamwork, meditation, self- narration), recognition of the client’s individuality (boundaries, personal continuity and needs, set of core values), adequate mirroring of the client’s self, mentalization, and mindsight. As a result, this enhances the client’s self-integration, self-leadership, self-support, self-compassion, self-determination, ability to make decisions based on free will, as well as the ability to recognize and determine what resonates with the client’s self and what does not (the true self dominates, defenses are deactivated).

Overall, we suggest that the discourse on client change (i.e., successful psychotherapy) may be compared to the discourse on successful child development (i.e., adaptive ontogenesis) (see Beebe & Lachmann, 2005). More specifically, we postulate that safety plays a fundamental role in the development and adaptation of both children and clients. In the beginning stages of life or therapy, it is essential for each child or client to have a caregiver who can be responsive enough to provide a sufficient amount of non-defensive (functional) experiences of safety. This creates a foundation for the child/client to form basic trust and a secure attachment, which in turn facilitates the development of a secure self and identity, self-esteem, a sense of safeness and the development of adequate mentalizing abilities. These abilities allow for the adoption of flexible strategies of emotional regulation, resulting in higher degrees of biopsychosocial adaptation (Bowlby, 1988; Cortina & Liotti, 2010; Schore, 2003; see also Palmieri et al., 2022). It is here important to stress the “enough” character of the caregiver’s/therapist’s responsivity and of the related safety he or she provides. Environments characterized by abuse, misuse, or misdosing of safety (e.g., overprotection [overcontrol, maximization of safety], overindulgence [overnurturance, overcare], underinvolvement [neglect, overpermissiveness, minimization of safety], or overdominance [authoritarianism, exploitation, abuse]) increase the risk of development of different psychopathologies (Capron, 2004; Sahithya, Manohari, & Vijaya, 2019). In fact, only environments characterized by enough safety provide opportunities for the child/client to experience growth-promoting pains, frustrations, risks, and threats. As long as the caregiver/therapist can timely and adequately co-regulate frustrations, disappointments, and insecurities and repair inevitable relational ruptures, such challenges become increasingly tolerable for the developing child/client and facilitate his or her growth and resiliency (Beebe & Lachmann, 2005; Safran & Kraus, 2014). This expands the range of biopsychosocial repertoire and allows the child or client to better interact with their environment, increasing their flexibility and adaptability (see Venuleo et al., 2020).

This co-regulated dialectic between safety and danger (Segalla, 2018) allows the child/client to increasingly internalize the experience of sufficient safety through repeated renewals of the balance of biopsychosocial arousal. Consequently, the child/client learns new and more functional modalities of self- and interactive regulation and increases the possibility of biopsychosocial exploration. This progress enables them to move towards what’s referred to as the zone of proximal development (Vygotsky, 1978; for transposition of this concept to psychotherapeutic settings, see Leiman & Stiles, 2001). The more this is the case, the more the child/client will be able to increasingly assimilate new experiences leading to change over time (to this aim, see the concept of corrective emotional experiences; see Castonguay & Hill, 2012 for a comprehensive review). With specific reference to psychotherapy, what plays a primary role in this whole process is both relational and technical. On one side, the therapist’s ability to intersubjectively mentalize the client and, therefore, to attune to him or her (Siegel, 2010). On the other side, the ability to strategically use this attunement to deliver appropriate, well-dosed, and well-timed interventions which either support him or her or promote change at an emotional, cognitive, and/or behavioral level. These school-independent developmental functions of safety appear to be coherent with quantitative empirical findings in the field of psychotherapy research (both within and outside the framework of attachment theory). Firstly, from the perspective of attachment theory, it has been found that secure clients, compared to insecure ones, show a stronger therapeutic alliance (Diener & Monroe, 2011), a better commitment and compliance to treatment (e.g., Dozier, 1990), and a better treatment outcome (e.g., Levy et al., 2018). Analogously, secure therapists, compared to insecure ones, facilitate more often corrective emotional experiences (Dozier et al., 1994), are more able to facilitate a good alliance as well as treatment outcome (Degnan et al., 2016), and are more able to repair alliance ruptures through empathy (Rubino et al., 2000). Second, the therapist’s secure attachment is associated with his or her in-session attunement and engagement (Talia et al., 2020). On the other side, the client’s pre-treatment reflective functioning (RF) predicts his or her in-session RF as well as in-session autonomy and security (affect sharing, self-assertion, and autonomous reflection) (Talia et al., 2019). Third, the client’s in-session security is associated with both his or her working alliance and with alliance ruptures-repairs (Mallinckrodt & Jeong, 2015; Miller-Bottome et al., 2019); with other clinically productive in-session processes such as the client’s level of exploration (Parish & Eagle, 2003), self-disclosure (Saypol & Farber, 2010), and session depth (Romano et al., 2008); and with better outcome (e.g., Sauer et al., 2010).

Some other quantitative research conducted outside the specific framework of attachment theory also seems to be consistent with the basic developmental functions of safety we propose. First, a client’s sense of safety is predictive of the treatment outcome (Beck et al., 2006). More specifically, clients’ early feeling of safety predicts subsequent treatment improvement, with early alliance mediating this relationship (Friedlander et al., 2008). Moreover, clients’ concerns about safety are associated with low alliance values (Beck et al., 2006). Second, alliance ruptures are associated with a reduction in client safety, whereas an increase in client safety is associated with alliance repairs. Interestingly, therapist repairs are characterized by his or her ability to enhance a shared sense of purpose with the client following an emotional connection with him or her and requiring the client to feel safe (Escudero et al., 2012). Finally, client-rated session safety is associated with session positivity, smoothness, and depth, as well as with the therapeutic bond, confident collaboration, and overall alliance (Siegel & Hilsenroth, 2013).

Finally, also qualitative research has produced preliminary evidence which seems to be coherent with the above-described functions of safety. Regarding the clients’ perspective, one of their core experiences of what is helpful in therapy deals with safety, reassurance, and support (Timulak, 2007, 2010). Moreover, the therapist's authentic caring and boundary-setting allow clients to feel safe and connected with the therapist and consequently to engage in the “vulnerable work of self-exploration and discovery” (Levitt et al., 2016; p. 823). Finally, therapists experienced as protective and caring provide a safe environment that represents a platform for self-discovery (Kirsha, 2019). Analogously, studies on therapists’ perspective showed that the development of relational security allows the client to constructively deal with the in-session risks associated with change (Williams & Levitt, 2007) and to tolerate the vulnerability and uncertainty experienced during the client’s exploration required for change (Levitt & Piazza-Bonin, 2016). For this reason, therapists should particularly preserve clients’ safety by engaging their subjective world, especially when client-therapist disagreements occur (Williams & Levitt, 2007).

Conclusions and future directions

The concept of safety seems to permeate much of the discourse around psychotherapy, as revealed by our review of psychodynamic, cognitive-behavioral, and humanistic approaches. These different schools emphasize different aspects and functions of safety and draw different implications for clinical work. Psychodynamic approaches prioritize unconscious, developmental, and relational aspects of safety (e.g., holding, containing, mirroring, mentalizing, and reflective functioning). Cognitive-behavioral approaches emphasize neurobiological, psychoeducational, relational, and exercisable aspects of safety (e.g., attunement, safety/safeness behavior/ signals, narrating, priming, and sourcing). Finally, humanistic approaches see the importance of safety mainly in its dialectical and relational nature (e.g., trust, unconditional acceptance, presence, congruence).

At the same time, however, our analysis has shown how different therapeutic schools may converge around a set of basic issues regarding safety. At a very general level, we would say that safety is the sine qua non of both healthy human development and effective psychotherapy. In other words, safety plays a fundamental role in the (functional vs. dysfunctional) ontogenetic development of an individual and, exactly for this reason, it plays a fundamental role also in (effective vs. non-effective) psychotherapy. Yet, to be healthy and effective, human development and therapy do not need to be perfectly safe but sufficiently safe; this is because humans thrive if they experience enough safety leaving space for tolerable frustrations, disappointments, and insecurities that promote resilience, growth, and change. With this regard, we have proposed five basic school-independent and therefore integrative functions of safety (securing survival and defense, facilitating restoration, promoting exploration and play, sustaining risk-taking and coping with danger, and enabling balance and integration). These functions seem to be coherent, although to different extents and in different ways, with different school-specific theorizations about the nature and function of safety in psychotherapeutic contexts. They deal with the developmental processes involved in both ontogenesis and clinical practice. Each of them is differently intertwined with the concept of enough safety and appears to be healthy and therapeutic as long as it is well-timed, well-dosed, and well-structured. That is, each of these functions is consistent with the principle that the interaction between safety and ontogenesis/clinical practice is of dialectical and paradoxical nature. Safety and its functions are healthy and therapeutic if they are neither idealized nor devalued but rather provided in the right form, at the right circumstance and time, and with adequate dosages and duration.

However, more research is needed to further understand the role of safety in psychotherapy. First, it should be better explored how other therapeutic approaches than those considered in this paper conceive safety and its functions within the clinical process and to what extent they are coherent with the school-independent functions of safety we have proposed. Second, an effort should be made to more explicitly connect both the identified school-specific and school-independent functions of safety to interpersonal neurobiology and developmental affective neuroscience (Fosha et al., 2009; Schore, 2003). Such approaches have already provided very useful insights and therefore represent good candidates for the development of a more general theory of safety able to account for its various functions within psychotherapy. Third, empirical research should attempt to explicitly test the hypotheses regarding the developmental functions of safety in psychotherapy proposed in this paper. To this aim, the assessment of safety should be first further articulated through the systematic employment of both self-reports and observational instruments. Process studies should then explicitly explore the relationship between client and/or therapist safety and other clinically relevant aspects of in-session processes. These aspects include the therapeutic alliance (including ruptures and resolutions episodes), client and therapist mentalization as well as their relationship over time, interpersonal cycles, and therapeutic interventions (supportive vs. change-oriented). Thereafter, process-outcome studies should be conducted to assess the extent to which safety-related process dynamics are predictive of within-session, post-session, and treatment outcomes. Finally, moderator analyses should be conducted to assess the extent to which different variables (e.g., treatment orientation and setting, client diagnosis attachment style) influence the relationship between in-session safety-related dynamics and outcome.

We believe that the further development of a general theory of safety in psychotherapy may have relevant implications for clinical theory, practice, and research. First, it would allow bringing further the discourse on psychotherapy and its mechanisms of action back to a general theory of human functioning and development, thus contributing to moving beyond the preparadigmatic stage of current psychotherapy science (see Salvatore, 2011). Second, it might shed further light on the common versus specific factors debate. We expect that at least many of the common factors described in the literature might be related to the basic functions that safety shows in the context of human development and clinical practice. The different specific factors would be ascribed to how the different schools decline these basic functions in the context of their psychological and clinical theory. Third, it would enable clinicians of different orientations to better recognize the extent to which their clinical practice is driven by a series of principles. These principles refer to a general way of functioning of the human being and are also the expression of how their therapeutic approach sees and interprets these principles. This might contribute to the further development of a theory-informed common language, better treatment tailoring, and clinical efficacy. Finally, the development of such a general theory of safety in psychotherapy would offer a shared framework informing and orienting future research and clinical practice in psychotherapy.