“How do you like those boundaries?”: Metaphor and ontology of therapeutic boundaries

Abstract

Across the mental health professions, the metaphor of “boundaries” serves to designate a metaethics of therapist–patient conduct that ranges from the clinical setting to the therapeutic relationship. Though once rare, by the 1990s, boundary nomenclature had attained a notable presence, designating tangible limits to therapists’ conduct. Like fences at a precipice, clinical boundaries were introduced to safeguard therapists from perilous missteps. Mapping a variety of potential dangers, the metaphor and associated lexicon evolved most obviously from growing wariness of therapists’ sexual involvement with patients, though was informed too by the rising challenges of litigation, managed care, and the vagueness of professional codes. Being a neologism without language specific to any therapy system, the metaphor transcended many professional disagreements. Intrigued by its endurance as a ubiquitous instrument without philosophical, legal, or empirical grounding, our study explores discursive devices that bolstered its success. Early proponents capitalized on a metaphor's capacity to make visible some forms of therapy and hide others while obscuring an authorial purchase. Terminology like “slippery slope” and “boundary flexibility” highlighted therapists’ vulnerabilities to transgression yet also sustained ambivalence about the parameters of safety. Diagnoses of therapist boundary violators as psychologically impaired shifted attention toward risks facing “normal” as well as disordered practitioners, and case study vignettes enacted boundaries just as they dramatized therapists’ missteps. In its enduring success, clinical “boundaries” joined company with other professional neologisms which were motivated by then-current problems and engineered by professionals in ways that reshaped their work.

Responding to burgeoning interest in “clinical boundaries” as a code for psychotherapists’ self-regulation of ethical conduct in therapy, psychiatrists Thomas Gutheil and Glen Gabbard furnished a definition. While admitting that like other psychotherapeutic concepts, the loosely defined term boundaries “proves slippery on closer observation” (1993, p. 188), the authors nevertheless proceeded to delineate the boundaries as spatially circling the entirety of therapeutic practice: the term encompasses clinical activities of time, place, roles, space, fees, gifts, services, clothing, language, self-disclosure, and physical contact. They capitalized metaphor's capacity to describe an object in ways that are not literally true and that draw attention to a part of the object, summoning a novel view of it. Taking the word's common definition—a line marking the limits of an area—the authors stated that boundaries are sometimes crossable and even flexible. Acknowledging the lines’ mutable yet real limits, they distinguished occasional “boundary crossing,” an acceptable overstep, from “boundary violation,” “a harmful crossing, a transgression” (p. 190). Most serious of the border breaches are “sexual boundary violations,” behaviors that extend from touching to sexual intercourse with patients. They asserted, however, that all transgressions are precarious because sexual misconduct almost always begins with minor crossings—even moving from last-name to first-name salutation or sharing personal information with a patient can progress to a grave violation. The hazardous “slippery slope” from “benign” crossings to sexual boundary violations calls for constant vigilance over “even mild forms of self-disclosure” (p. 194). The boundary lines’ dangerous yet at times flexible edges can precipitate a steep descent and underscore the need for “heightened awareness” (p. 195).

Just 2 years later, it seemed “that everywhere you look, the psychotherapy world is buzzing with talk about boundaries” (Greenspan, 1995, p. 51). Gutheil and Gabbard's (1993) conceptualization flourished and now is regarded by many as foundational, deemed a “landmark” study (Adames et al., 2023, p. 162). The successful introduction of this neologism to a manifestly heterogeneous professional community was an impressive accomplishment. Along with other proponents of the nascent entity, Gutheil and Gabbard were championing an ethics-related terminology that was not in the lexicon, legal regulations, or ethics codes of the mental health professions. ¹ And unlike the physical boundaries of soccer fields, park grounds, and private property, a therapeutic boundary is something that could not be seen; it is neither visible nor measurable. Yet clinical boundaries became an enduring terminology utilized by many therapists as directives on treatment parameters. Despite having neither scientific evidence nor codification in professional ethics, boundaries came to be taken as real, attaining ontological status as an accepted professional disciplining instrument.

The subsequent surge of literature on boundaries was believed to reflect success and a “progressive raising of consciousness about the issue” (Gutheil & Gabbard, 1998, p. 409). ² That quickly rising consciousness is evidenced as well in how-to guides for maintaining boundaries (e.g., Adames et al., 2023; Koocher & Keith-Spiegel, 1998; Pope & Vasquez, 2010). In 1994, a boundary maintenance book, Keeping Boundaries, asserted that therapeutic boundary management “is the foundation for treatment technique” (Epstein, 1994, p. 9). Some held that ethics codes promote establishing reliable boundaries as “one of the most fundamental responsibilities of the therapist” (Pope, as cited in Williams, 1997, p. 70; see also Hundert & Appelbaum, 1995). Addiction counseling students learned that boundaries is “a term governing the relationship between you and your client” (White, 2004, p. 536). Every edition of Oxford University's Psychiatric Ethics since 1999 contains a chapter on boundaries, and books on boundaries were and continue to be published by the American Psychological Association (Adames et al., 2023; Knapp & Fingerhut, 2023; Steinberg et al., 2021). Boundaries slid into professionals’ quotidian vocabulary, including therapists’ talk of being “boundaried” (Heslin, 2017).

Boundaries rhetoric thus came to provide a useful disciplinary apparatus marking what should—and mostly what should not—occur in therapy, though the implications of boundary language go beyond novel terminology and a metaethics for navigating a presumably dangerous landscape. The discourse at least partially came to substitute trained judgement for boundary maintenance behaviors. Whereas one of the earliest mentions of boundaries states that sound clinical work requires “empirically derived, validated, sound technical principles” of psychoanalytic psychotherapy (Langs, 1982, p. 548), subsequent proposals by others show considerably less confidence in trained judgement, instead urging therapists to scrupulously maintain boundaries and monitor their own vulnerabilities to violations. As the boundary idea gained prominence, presumptions of the efficacy of trained judgement gave way to a pedagogy of managing risks. The mature metaphor's uptake informed a conceptual shift that muted if not translated some of the psychoanalytic principles that undergirded its earliest uses.

Realizing boundaries’ ontological status as a metatherapeutic tool was a substantial achievement, yet the question of how therapy got its boundaries has received scrutiny. While many welcomed its remedies for an unhealthy atmosphere of inadequate or ill-defined clinical ethics and heralded the definitiveness of borders (Epstein, 1994; Gabbard & Lester, 1995; Hundert & Appelbaum, 1995; D. Smith & Fitzpatrick, 1995, p. 500), not all appraisals were celebratory. While Lazarus’s (1998) tongue-in-cheek “How Do You Like Those Boundaries” summarizes the therapeutic detriments of keeping strict boundaries, other critics called out the metaphor's psychodynamic bias (Fay, 2002, p. 153) and its manufacture by an elite group of clinicians advocating constraints on therapeutic practices (Zur, 2002). These analysts tracked the adverse consequences of establishing “a restrictive notion of care,” charging clinical boundaries as a backlash against client-centered and self-actualizing therapies developed when “tearing down boundaries was encouraged” (Williams, 1997, p. 242). Yet others uncovered its sexist and hierarchical lineage that grants therapists inordinate power over clients (Anderson, 1999; Dineen, 2002). These concerns had little effect on the adoption of the concept of boundaries throughout the psychotherapy world, and its resilience and widespread success owed in part to its proponents’ visible positions in psychoanalysis, psychiatry, and psychology. ³

Our paper offers another perspective on the neologism. It tracks the play of cultural perceptions, compelling qualities of the metaphor, discursive strategies, and connections among psychotherapy schools that propelled the colloquial adoption and proliferation of boundary language across the mental health domains. Analysis of the foundational literature appearing in the 1970s and continuing into this century finds the neologism's ubiquitous acceptance to be bolstered by three devices: elucidation of a precarious situation in the profession, engagement of metaphor, and use of a psychologizing drawn from the profession's own technical language. These framed a disciplining discourse of therapists’ risks and new responsibilities for self-care, serving as foundation for the emergence and acceptance of boundaries. The paper's first section reviews the mental health profession after World War II (WWII), noting how the massive expansion of mental health care was accompanied by a multiplying of theories and techniques as well as legal and economic challenges. Highlighted is one condition that predominated in the promotion of boundaries: growing apprehensions about therapists’ sexual misconduct. And the neologism's rapid, widescale adoption was crucially facilitated by use of metaphor. Informed by metaphor studies, the second section describes how boundaries became an entity that affirms certain modes of therapy, disallows others, and conveys cultural affects and attitudes. Boundaries’ status was supported too by reflexive logics: the inward-turning use of clinical expertise to diagnose boundary-violating therapists and those at risk. ⁴ Together, rehearsals of therapy's hazards, reflexive diagnostics, and narrative displays made way for imagining a normative disciplined other: the nonpathological therapist who understands their vulnerability and consequent need to self-monitor their behavior. Despite enduring success, the metaphor encountered criticism, largely from therapists aligned with either feminist or relational psychotherapies; their exposures of its occlusions and troubling implications are examined in the final section.

Given their endurance, recognition, and continuing practical utility, clinical boundaries have become more than a figure of speech. Using Ian Hacking's model of dynamic nominalism, boundaries are understandable as an ontological entity, the outcome of a style of reasoning that not only brings things into view but also “can bring them into being” (Sugarman, 2009, p. 5; see also Hacking, 2002; Mol, 1999; Pickersgill, 2019). Much like how another spatial metaphor, “glass ceilings,” came to be recognized as a phenomenon, clinical boundaries were established through an assembly of materials and networks of professionals (Morgan, 2012, p. 161). In following boundaries’ emergence, our analysis joins company with studies of other professional classifications in which the classifiers are made a referent. Like medical professionals’ classification of “telepsychiatry” and “precision medicine,” the nomenclature of clinical boundaries was “deliberately and actively fostered by agents who reflexively reconfigure their (e.g., professional) identities in the process” (Pickersgill, 2019, p. 17; see also Galasso et al., 2024). Similar “performative nominalism” ensued in the case of boundaries, where therapists’ new nomenclature presented a new intelligibility for therapeutic practice, ultimately delivering a comprehensible, communicable reconfiguration of experts’ responsibilities.

Expansion, innovation, crises

From its 19th-century origins, psychotherapy expanded to treat ordinary maladies and serve people's growing intrigue with all things psychological (Caplan, 1998; Shamdasani, 2017). In the United States, the second half of the 20th century saw further expansion of consumers, sites of care, practitioners, and psychotherapy systems (Benjamin, 2005; Herman, 1995; Lunbeck, 1994). The growth in practitioners occurred across the mental health fields. By 1960, for instance, the number of trained and practicing psychotherapists had risen to 20,000, increasing steadily to 30,000 in 1970 and 65,000 in 1980 (DeLeon et al., 2011, pp. 40–46; Stapp et al., 1985). Between 1948 and 1976, the number of psychiatrists grew from 4,700 to 27,000, half of whom specialized in psychoanalysis in 1973 (Hale, 1995). Social workers likewise multiplied in numbers, with Master of Social Work graduates increasing from 2,162 to 5,638 between 1960 and 1970, reaching over 15,000 by 2000 (Schilling et al., 2008, pp. 103–105). Human services professionals grew 242% between 1960 and 1980, following the establishment of family and educational specialists (Jimenez, 1990, p. 9). The growth required training nonpsychiatric mental health experts, from clinically trained psychologists and social workers to public health professionals, which demanded “translating” of therapeutic practices (Devonis, 2021; Koch, 2022).

Accompanying these advances were reforms of credentials, practices, and research styles, exemplified by the creation of doctor in psychology (PsyD) programs in the late 1960s (Benjamin, 2005), ongoing revision of various professional ethics codes, incursion of managed care into the mental health professions (Boyle & Callahan, 1995; Goldstein, 2007; Hoge et al., 2000; Kochunas, 1997), and heightened awareness of litigation (Wright, 1981). Across professional schools arose disputes over the scientific bases of their techniques, notably with the calls for evidence-based practice and introduction of randomized controlled trials (RCT) to assess the efficacy of therapy (Benjamin, 2005; Goldstein, 2007; Rosner, 2018).

Psychotherapy theories, techniques, and research protocols grew almost as remarkably as did the number of practitioners and clients, sometimes leading to controversy. The detached Freudian analyst and distressed if resistant patient caricatured in films and cartoons were not representative of real-life therapeutic scenes by the postwar era when other psychotherapies were introduced. The new therapeutic approaches arising in the 1950s and 1960s ranged from Albert Ellis’s rational–emotive therapy to Carl Rogers’s client-centered therapy. “Relational” and “multimodal” therapy models challenged presumptions of therapist neutrality and abstinence, instead championing emancipatory encounters and collaborations between therapist and client (Staub, 2015). Evidence-based, nonpsychoanalytic techniques were introduced, including behavioral therapies developed in the 1960s and cognitive behavioral therapy (CBT) gaining adherents during the 1980s (Kendall et al., 1981; Newman et al., 2017; Rosner, 2018). ⁵ The 1980s saw psychologists’ recommendations for combining psychodynamic and behavioral approaches (Goldfried et al., 2011, p. 281) and clinical social work scholars debating the use of psychodynamic perspectives and evidence-driven, behavioral models (Goldstein, 2007). Alongside these developments came increasing psycho-pharmacological interventions that embraced biological explanations of mental illness in lieu of psychodynamic accounts (Braslow & Marder, 2019, p. 37). Additionally, activist movements forming in the 1960s challenged conventional notions of the authoritative therapist and submissive patient (Herzog, 2016; Richert, 2019). Feminist psychologists demonstrated psychology's sexism and patriarchal power, Black psychologists contested White-centered therapies, and gay scholars and activists rejected the pathologization of homosexuality. The epistemics of traditional psychoanalysis was scrutinized, particularly assumptions of analyst neutrality, transference, and countertransference (Eissler, 1953; Orr, 1954; Sharaf & Levinson, 1964; see also Koch, 2017; Zaretsky, 2004). Paralleling the younger nonpsychoanalytic therapies, some revisions of orthodox psychoanalysis promoted more democratic-like models and relaxed conceptions of the therapeutic relationship (Bordin, 1979). Despite ongoing controversies over technique and theory, the author of a 1959 catalogue of 36 therapeutic systems lauded the profession's “eclecticism” and flexibility over “rigid adherence to a single system of psychotherapy” (Harper, 1959, p. 149).

The psychotherapy explosion of the 1960s and 1970s paralleled North American growing cultural preoccupations with personal well-being and mental health. For many, the self became a central object of concern: self-esteem, self-actualization, and nearly everything about the self circulated through popular culture and was incorporated in many therapies (Cushman, 1990; Hofman, 2022) and self-help literature (Greenberg, 1994; Kaminer, 1992). Some observers, however, claimed that the emergent therapeutic culture marked an epidemic of narcissism and unwellness, not human improvement (Aubry & Travis, 2015).

Barely 4 decades of expansion brought new therapies, diversified services, consumers’ escalated demands for care, and consequent revisions of regulations, laws, and ethics codes. Mental health experts thus faced continuing choices, opportunities, and challenges. Historian Eli Zaretsky's chronicle of psychoanalysis’ fate during this tumultuous period is relevant to the entire mental health profession. He observed that by the end of the 1970s, the revered psychoanalytic category of hysteria “had disappeared as a clinical entity, replaced by identity movements and a new vocabulary of mental disorders.” Zaretsky concluded that “as new churches emerged, the psychoanalytic hegemony faded into history” (2004, p. 330). Directed by no single theory or technique, clerics of both the new and old therapy churches had to navigate the evolving regulations, codes, fiduciary systems, and cultural perceptions along with consumers’ growing desires and needs. Although the concept of clinical boundaries was generated in this messy environment, its propagative literature rarely acknowledges the pressures, conflicts, and movements within the field; clinical boundaries’ ecumenical perspective obscured many professional contentions just as it dampened financial and regulatory worries.

Diagnosing therapy: Sex as symptom

Whether psychotherapists shared many tenets of technique, by the early 1970s almost all had apprehended a serious clinical dysfunction: sexual encounters between therapist and patient. Intimacy in therapy had been recognized from Freud onward, but sexual relationships and potential for their occurrence attained increasing visibility in the 1970s, becoming an arch symptom to be observed, measured, monitored, sanctioned, and even submitted to psychological diagnosis. ⁶

The public had long pictured psychoanalysis as the sexual science, but in the 1970s, another portrait of sex and therapy gained prominence. Attention shifted away from sex represented in the ideas of infantile and repressed sexuality and toward sexual dramas enacted between therapist and patient. Midcentury conceptions of transference and countertransference indicated lenient attitudes regarding a degree of nonsexual intimacy in the psychoanalytic relationship: in the face of expert scrutiny and self-analysis, some held that unconscious libidinal occurrences were clinically valuable (Boyer, 1961, p. 392; Eissler, 1953; Ross & Kapp, 1962, p. 646; Weigert, 1954). Such conceptions began to wane with critiques pointing that focus on countertransference was burdensome for the clinician and irrelevant for the patient (Reich, 1960; see also Jacobs, 1999, pp. 581–582). For the public and professionals alike, the “sex” associated with psychoanalysis and related therapies became less about the unconscious and more about conscious acts of misconduct. Some attributed the apparent increase in these acts to liberalized moral attitudes about sex or to novel sex therapies that fed “notions that sex and therapy could become allied as routes to personal liberation” (Kim & Rutherford, 2015, p. 286). Vocal defenders of sexual intimacy in therapy were rare, though until the 1970s, so were admonishers. ⁷ In 1970, a professional journal published a condemning report of patients’ “first-hand” accounts of therapists’ sex or attempted sex with them (Dahlberg, 1970). The popular media offered accounts of therapist–patient sexual encounters. Writing in New York Magazine, Phyllis Chesler exposed therapists’ sexual misconduct as the rampant sexism attending a patriarchal power structure (Chesler, 1972a). Several months later, Cosmopolitan magazine featured an article allegedly written to replace moralistic “gossip” with a “more candid” account of the nuances and occasional appropriateness of therapist–patient sexual relations (Gaines, 1972). Soon such defenses practically disappeared to be replaced by condemnations and legal actions, starting with the nationally publicized success of a 1975 lawsuit against a therapist who had a sexual relationship with his patient (Freeman & Roy, 1976).

Arriving during a period when “sex filled the public sphere” (Herzog, 2016, p. 72), the problem of therapist–patient sex garnered several different explanations. According to one apparently ambivalent psychiatrist, the problem arose from psychiatry's “rude awakening” by the “the Women's Movement” along with Masters and Johnson's report on the frequency of therapist–patient sex (A. A. Stone, 1984, p. 191; Masters & Johnson, 1970). Early acknowledgement of the problem came from feminist activists, scholars, and therapists who attributed sexual misconduct to psychology's patriarchal foundation. Accordingly, Chesler declared sexual relations in therapy as neither more nor less common than sex between male boss and female secretary or housekeeper, adding that “sexually seductive (or assaultive) therapists are quite ordinary in their ethical failure” (Chesler, 1972b, p. 152). Feminist psychologists’ demands prompted task forces, sponsored research, and ethics code reforms (Hare-Mustin, 1974; Kim & Rutherford, 2015; Lerman & Porter, 1990). The entire mental health industry eventually turned attention to this worrisome symptom, making it the subject of commentaries and empirical studies (e.g., Folman, 1991; Schoener et al., 1989). Responding to “anecdotal reports” of sexual misconduct, Shelden Kardener and colleagues surveyed physicians’ attitudes (Kardener, 1974, p. 1134; Kardener et al., 1973). After reporting small yet disturbing evidence of problems, Kardener sketched a psychodynamic portrait of errant (psychiatrist) physicians, though also faulted “a burgeoning number of nonprofessional modes of quasi-psychotherapy, offering a bewildering variety of techniques and philosophies” that promised self-improvement but eschewed professional responsibilities (Kardener, 1974, p. 1136). While some therapists continued to attribute purported increases in sexual misconduct to new-fangled therapies, lax sexual mores, or the Women's Movement, by the mid-1980s, therapists largely concurred that sexual encounters between therapist and patient constituted “abuse” or “exploitation.”

Sustained attention to the symptom of therapist–patient sexual conduct along with lawsuits prompted legal changes. One historical account of psychiatric ethics maintains that “judicial pressure” prompted “psychiatry to put its house in order” to clarify its position on transgressions (Bloch & Pargiter, 2002, p. 521). In the aftermath of lawsuits, legal reforms, professional debate, and research, psychiatry, social work, and psychology adopted ethics regulations prohibiting therapist–patient sexual contact (Lerman & Porter, 1990). The problem also supported a reframing of the therapeutic relationship using the business language of risk management. Risk management discourse essentially reconfigures the roles of therapist and patient as those of provider and consumer. This role shift is exemplified in psychologist Robert Woody's Protecting Your Mental Health Practice; the author devoted an entire chapter to “avoiding sexual impropriety” and a foundational claim that managing risks marked the profession's “shift from altruism to commercialism” (and a move toward “exchange relationships”). Therapy is “a business,” indeed, a “risky business” that ever faces a myriad legal and financial liabilities (Woody, 1988, p. 13). Woody concluded that risk management should be applied to all aspects of therapy: “As hard as it may be to swallow, the professional must accept the fact that mental health services must take the same precautions as any other high-risk business, only more so” (p. 189; see also Haas & Malouf, 1989; Kapp 1983). Soon, clinical boundaries appeared in risk management publications for therapists (Menninger, 1991). ⁸

The problem of sex in therapy converged with legal and financial concerns about malpractice and the consequent introduction of risk management techniques. Professional ethics and training were coming to be seen as insufficient protections against lawsuits. Therapists were warned that “By training, experience, and life-style,” they “are ill-equipped to defend themselves against formal allegations of malpractice, and many well-intentioned actions may actually increase the psychologist's vulnerability in the courtroom” (Kapp, 1983; Wright, 1981, p. 1535).

Sensationalized and submitted to intense scrutiny, the symptom of sex attained prominence just as the heterogeneous profession faced growing pains that spanned from controversies over technique to ongoing regulatory reforms. In this changeful atmosphere, the idea of clinical boundaries tendered a visibly uncomplicated and apparently theory-neutral prophylaxis against the now-visible breaches of sexual misconduct. The unacceptability of sexual intimacy in therapy had consensus across professional schools; what was to become the means of preventing its occurrence appeared to overshadow ongoing technical controversies. It also subtly shifted the ideal therapeutic arrangement toward a more transactional relationship, one more in keeping with economic as well as legal models of professional services. The acceptance and endurance of clinical boundaries owe much to the use of a capacious metaphor that delineates therapeutic limits even though it tacitly affirms certain practices while obscuring others.

Making a metaphor work

Metaphor is a crucial tool in knowledge making: a rhetorical achievement that not only commands a novel view of a thing but also transmits cultural knowledge and authorial powers (Bono, 2001; Fyke, 2024). Metaphor can be world making by “projecting our metaphorically constructed models onto matter which we shape and use to our ends as instruments of thought and action” (Bono, 2001, p. 228), and is a flexible tool that can be extended, modified, conjoined with other metaphors, and transported to reimagine other things. It is a cognitive instrument whose definitive essence “is understanding and experiencing one kind of thing in terms of another” (Lakoff & Johnson, 1980, p. 5; original emphasis). In conveying similarities with experience, metaphor invites users to select aspects of reality that seem consonant with the metaphor's targeted use (Stepan, 1986, p. 272). Though metaphors can be world making, they are grounded in experience such that “basic experiences of human spatial orientation give rise to orientational metaphors” (Lakoff & Johnson, 1980, p. 25). Lakoff and Johnson's cognitive perspective finds metaphor to be based in physical experience and then transferred to a conceptual realm; the transferred abstract concept is then taken as concrete. Regarding “boundaries,” they observed that “We experience many things … as having distinct boundaries, and, when things have no distinct boundaries, we often project boundaries upon them” (Lakoff & Johnson, 1980, p. 58). The metaphor can be imposed when no physical boundary exists (1980, pp. 29–30).

By facilitating “understanding and experiencing one kind of thing in terms of another,” metaphors give partial understandings that obscure some aspects of the referent while beckoning attention to other aspects. Which features are hidden and which illuminated are culturally contingent, corresponding with historical and collective experiences. Via metaphor, cultural meanings and affects are transported to the signified thing and, therefore, invite examining “what the metaphor drags with it” (Cowan & Rault, 2022, p. 2). The transference of cultural meanings is illustrated by the late-19th-century metaphor of “mental contagion” developed to explain crowd behavior, which associated crowds with dangers of biological contagion. The transferred negative meanings buttressed rationale for control of harrowing crowds (Richards, 2010, p. 185). A more recent metaphor, “glass ceilings,” attained social scientific recognition through transformations in experiential knowledge, cultural values, and advocacy work (Morgan, 2012). As these cases show, metaphor thus connects the linguistic, cultural, and material through an abstract authorial purchase to persuade others that the metaphor represents reality (Greenberg, 1994; Lakoff & Johnson, 1980; Stepan, 1986). However, also relevant to understanding its persuasive capacities is a materialist perspective on spatial metaphors that attends to the “interconnectedness of the material and metaphorical” (N. Smith & Katz, 1993, p. 67), revealing how uncritical embrace of spatial metaphors elides the political history of spatial logics, enabling its language to ground social meanings. Adoption of the clinical boundaries metaphor depended on the assembly of cultural awareness, personal experiences, and professional knowledge. Yet this adoption also depended on overlooking the substantial ways the metaphor's presentation of appropriate social relations conceals both certain economics as well as politics and power relations of the profession.

Making boundaries

When boundaries and accompanying lexicon were introduced to signify appropriate therapeutic practices and inappropriate transgressions, they were not the first spatial representation of therapy, which was sometimes described in terms of therapeutic “frame” and “parameters.” ⁹ Nor was the term boundaries entirely new to psychological discourse. During the postwar era, it appeared in social psychological, object relations, and family system theories. In these cases, users typically underscored the term's theoretical usefulness, not its literal representation of mental processes. According to Salvator Minuchin, the term is used “to draw an ‘artificial’ boundary around the family and the therapist. It is impossible to do otherwise” (Minuchin, 1974, p. 256). ¹⁰ Gestalt theorists similarly held that ego boundary has no structure and thus conceptualized boundary “as a verb rather than a noun” (Lewis & Schilling, 1978, p. 274). However, not all believed the metaphor to be appropriate for psychological theorizing. In his 1967 book Boundaries, Robert J. Lifton rebuked theorizing boundaries of the individual and, instead, promoted the “blurring of perceptions of where the self begins and ends” (Lifton, 1967, p. 39).

The metaphor's first distinctive, credited use in reference to therapeutic practice was by a psychiatrist concerned about deviations from conventional technique. Robert Langs called therapeutic “ground rules and boundaries” that encompass

set fee, hours, and length of sessions; the fundamental rule of free association with communication occurring while the patient is in his chair or on the couch; the absence of physical contact and other extra therapeutic gratifications; the therapist's relative anonymity, physicianly concern, and use of neutral interventions geared primarily toward interpretations; and the exclusive one-to-one relationship with total confidentiality. (Langs, 1975, p. 107) ¹¹

Deviations from psychoanalytically informed technique, Langs augured, can bring therapeutic failure and, therefore, should be undertaken cautiously and only in “clear cut emergencies” (Langs, 1975, p. 118). Though recognizing that deviations in technique can “stem from anxieties in the therapist” (pp. 134–135), he expressed little concern about that potential problem and, instead, reiterated trust in the analyst's sound training, judgement, and “considerable self-scrutiny” as insurance against problematic deviations (p. 137; see also Langs, 1982, 1984).

Soon thereafter came articles that prescriptively described therapeutic boundaries. These articulations diverge from Langs’s focal interest in good therapeutic outcome and assuredness of masterful and (psychoanalytically) well-trained providers. In contrast, they foreground transgressions of the boundary and locate therapists’ susceptibility to transgress. Although retaining Langs’s list of the specific professional activities that reside within the boundary lines, subsequent proposals emphasize border breaches, especially those involving sexual contact between therapist and patient. Any form of such contact is taken to constitute a serious boundary “violation.” Analyzing such violations, Michael Stone recalled the age-old “temptation” for male physicians to sexually engage with female patients (M. H. Stone, 1976, p. 670). He profiled female patients’ transference inclinations (“the thrust of transference”) toward emotionally and even sexually connecting with their therapists. He cautioned that the therapist can be tempted by characteristics of dangerous women patients: attractiveness, ego fragility, seductiveness, and a history of incest. Further, Stone posited, female patients who “had incestuous relationships with their fathers” tend to “play for high stakes.” Stone gave therapists the “difficult assignment” of adhering to boundaries and seeking alternatives whenever he “feels his self-control slipping” (1976, p. 675). Another psychiatrist credited with early use of the metaphor, Gerald Schamess, also detected multiple vectors of threat, and therefore exhorted practitioners to closely observe not only the “characterologically disturbed patient” (Schamess, 1981, p. 244) but also their own countertransference that demands control via boundaries of “thought, affect, and action” (p. 256). He also identified the risks of a public and professional climate that had grown wary of inappropriate therapeutic relationships and urged therapists to exercise caution, especially because that climate only appeared to be an “era of ethical relativism” (M. H. Stone, 1976, p. 675).

Subsequent boundary proposals share core premises of, first, the not-always-visible perils of trespassing borders and, second, the crucial need for well-defined, protected limit lines, yet over time, they become less reliant on therapists’ interpretation and judgment. Although boundaries were defined to encompass the entirety of therapeutic practice, from setting and payment to dual relationships, most presentations highlight sexual transgressions. Sexually inappropriate relationships and the undetectable risks of them are frequently explained in psychological terms of pathological or characterological deficits. Repeatedly emphasized are the perils of “boundary violations” that are initiated not only by certain kinds of patient groups (Gabbard, 1991; Simon, 1989) but also by vulnerable or self-deceiving therapists (e.g., Goisman & Gutheil, 1992; Gutheil, 1989; Plaut, 1997; Simon, 1992; Strasburger et al., 1992). Even the nonpathological practitioner is at often-unrecognized personal risk of slipping into inappropriate sexual contact. A therapist can be psychologically blinded, for instance, by holding “politically correct” beliefs that suppress the ubiquity of therapists’ “sexual ideation” (Gutheil & Gabbard, 1992). One preventative solution to therapists’ imperceptions and misperceptions involves taking a self-administered psychological test to assess their personal danger of committing “sexploitation” (Epstein & Simon, 1990).

Riding this wave of anxiety-laced warnings and scrutinized limit lines, Gutheil and Gabbard (1993) crafted a definition of therapeutic boundaries that eventually became a primary citation source. Like Langs’s concept and risk management rhetoric, their definition encompasses the entirety of therapeutic practice, reveals the risky “slippery slope” and dangers of “boundary crossings,” and highlights the perilousness of “extreme boundary violations” of sexual misconduct (p. 188). With that lexicon, what some once depicted as sexual acting out or eroticized transference was redefined as a violation (Gabbard, 1995; Gabbard & Lester, 1995). ¹² The 1993 definition, however, was not essential to the metaphor's flourishing. By the time of its publication, boundary terms were already becoming common in technical articles, ethics writings, the burgeoning risk management discourse, and pedagogy (e.g., Atkins & Stein, 1993; Borys, 1994; Epstein, 1994; Folman, 1991; Waldinger, 1994). Through frequent rehearsals in psychiatry, psychology, and social work literature, boundaries came to be promoted as real or at least constituting a legitimate map for therapy. In allusions to its physicality, therapeutic boundaries were likened to modern safety improvements like the highway (Bennett et al., 1994). From such physical analogies came proverbial warnings, for instance, that “the road to sexual misconduct is strewn with boundary violations” (Simon, 1995, p. 92), and comparisons to the Grand Canyon's fences that prevent catastrophic falls over a deadly precipice (Gabbard & Lester, 1995, p. 42).

By 2000, the boundary vocabulary had achieved a place in professional literature and in colloquial usage among therapists (Anderson, 1999; Heslin, 2017). As spatial metaphors can do, that of clinical boundaries beckoned a sense of its ontological status while delivering a guide to regulate practitioners’ conduct. And as metaphors do, it harnessed cultural expectations and affective senses just as it obscured or displaced others.

Discourse supporting the metaphor

Metaphors invite seeing something in terms of something else, and the boundary metaphor evokes an understanding of therapy as bordered with definitive lines demarcating an inside and outside of appropriate practice. This understanding was realized by incorporating professional expectations and conveying associated emotions and cultural meanings of proper therapeutic conduct. Fundamental to its realization was metaphor's abstract point of view: an authorial standpoint situated above the various controversies over techniques, regulations, and pedagogies, and above the suspicions of an increasingly litigious society. With its authorial purchase, therapeutic boundaries visualized a zone of security, a space within which therapists could practice safely though also remaining vigilant. Intimating safe guardrails of practice, the metaphor dragged with it cultural associations with physical boundaries and the hazards of trespassing them. Its zone of protection reverberated with dual messages of danger and safety. With these messages, clinical boundaries resemble recently erected national border walls which, however physically sturdy, fail to contain most modern dangers to a nation yet nevertheless “constitute a spectacular screen for fantasies of restored sovereign potency and national purity” (W. Brown, 2010, p. 9). Like the nation-state’s wall functions as a theater of danger and symbol of anxious surveillance, the clinical boundary metaphor indicates reassuring protection while simultaneously sounding warning alarms. The hazards of boundary breaches are rehearsed in plentiful descriptions of the purportedly often-unrecognized perils of stepping over or, what is more worrisome, slipping across the protective line. The dangers of transgressing are demonstrated through three compelling techniques: metaphor-driven terminology, diagnoses of vulnerable therapists, and case study enactments.

Language of precarity

The 1993 definition showcases terminology that promptly entered the professional vocabulary, albeit bringing along ambiguity as well as exactitude. Parts of the boundary lexicon were adopted from other sources, whereas other parts were developed specifically for the metaphor. The term “boundary violations” most clearly conveys the wrongdoing of transgressing the limit line. It first appeared in 1976 (M. H. Stone, 1976), and by the 1990s, it came to be regularly used, for instance, in risk-management discourse (Menninger, 1991). “Sexual boundary violations” likewise attained usage in the early 1990s (Epstein & Simon, 1990; Gabbard, 1991; Gutheil & Gabbard, 1992). That freighted expression was previously linked with “sexual exploitation,” an expression introduced in the 1980s to explicitly mark intimacy in therapy as an egregious moral as well as ethical and legal breach (Schoener et al., 1989; A. A. Stone, 1984).

While “violation” is an unambiguous and affectively weighted word, other boundary terms were less precise and thus presented uncertainties. “Boundary crossing” was introduced to name the acceptability of line-crossing treatment situations. Gutheil and Gabbard claimed that not all boundary crossings lead to sexual violations, adding that different therapeutic approaches draw the lines differently. They therefore introduced “boundary crossings” as “a descriptive term, neither laudatory nor pejorative” and recommended “case-by-case” assessment of whether a crossing constitutes a violation (Gutheil & Gabbard, 1993, p. 190). Similarly uneasy about overly prescriptive advice, Robert Waldinger observed how psychiatrists’ reactive responses to public scrutiny of misconduct were leading some of them to “therapeutic orthodoxy” and adherence to the “rules” of clinical conduct at the expense of attention to individual patients’ needs. Waldinger nevertheless supported the boundary metaphor, yet to avoid rigid line obedience, he encouraged therapists’ interpretive work and more nuanced ways of distinguishing boundaries to prevent “getting stuck on the surface of the doctor–patient relationship” (Waldinger, 1994, pp. 225–226). The ambiguity of crossings was compounded by “slippery slope,” a term introduced in 1991 to denote how even ostensibly beneficial deviations (crossings) can result in sexual violation. Even a therapist who extends the clinical hour or reveals personal information can “progress down a ‘slippery slope’ to hugging the patient, kissing the patient, meeting the patient for lunch, and finally having sexual relations with the patient” (Gabbard, 1991, p. 654). The slide can begin simply “in the form of small, relatively inconsequential actions by the therapist” (Strasburger et al., 1992, p. 547). Even before slippery slope's naming, one concerned author reported how sexual exploitation sometimes “begins before it happens” and cautioned that even though treatment deviations are occasionally safe, exploitation can occur without sex (Simon, 1989). These and similar caveats about overly “rigid” and too “flexible” boundaries intensified ambiguity about the lines.

The metaphor's affect-tinged terminology demarcates zones of safety yet also conveys ambiguities that did not go unnoticed. In addition to Waldinger's call for a nuanced interpretation came recommendations for literal redrawing of boundary lines. One proposal calls for replacing the “gray areas of boundary crossings and violations” with an elaborately revised boundary map that mark spaces of “pseudoboundary violations.” It recommends replacing “boundary violations” with neutral language of “actions at the boundary” (Glass, 2003, p. 442). Another proposes a “graded-risk” model that divide crossings into four categories (Martinez, 2000). However, most allies of clinical boundaries adjudged that the uncertainties did not need redefined borders or a return to therapists’ deep introspection of unconscious processes. What was needed was therapists’ heightened responsibility to diligently monitor their behavior in therapy.

Psychologizing

Although diagnoses of errant or deficient psychotherapists have a longer history, those diagnoses increased in the 1970s, especially regarding therapists who had sexual relationships with patients. Sexual intimacies were sometimes attributed to patients’ pathologies (Gutheil, 1989) but were increasingly ascribed to offending therapists’ disorders or vulnerabilities. Diagnoses of these therapists involved reflexive logic in which authors drew upon their extant professional psychological knowledge to make sense of the very practitioners of that knowledge—therapists themselves. Heightened apprehension about therapist–patient sex prompted closer scrutiny of the delinquent practitioner (Dahlberg, 1970) and “seductive psychotherapist” (Marmor, 1972). The kinds of offending therapists’ psychopathologies multiplied enough to enable one psychiatrist to compile a typology divided into six major characteristics (A. A. Stone, 1984). Close analyses of the “sexually abusive practitioner” (Woody, 1988) and “at risk” therapist (Folman, 1991) resulted in discovery of an array of psychological infirmities that were incorporated in boundary discussions. Among the disorders named are depression, sociopathy, perverse sexuality, “bad character,” neurosis, psychosis, sociopathy, dysregulated ego boundaries, and defective super ego (e.g., Folman, 1991; Garfinkel et al., 1997; Gonsiorek, 1987; A. A. Stone, 1984; M. H. Stone, 1976; Woody, 1988). Of the diagnosed psychopathologies, narcissism received sizable scrutiny, and grandiosity was identified as a common trait of boundary violators. It was reasoned, for instance, that sexual activity might result from the narcissist therapist's sense of “specialness” or be “triggered by a narcissistic injury suffered by the therapist, such as marital dissatisfaction or separation” (Epstein & Simon, 1990, p. 454). These diagnostic efforts even included identification of an apparently new form of narcissism: the “lovesick” therapist who projects both his grandiose self-image and injured self onto the patient (Gabbard, 1991; D. Smith & Fitzpatrick, 1995). Although psychopathologies dominate the space and affective tone of these appraisals, also interrogated were “situational factors” and detrimental life experiences of vulnerable and at fault therapists. These situational factors include relatively common life experiences, as exemplified in the hypothetical case of “a male therapist in his mid-forties who lacks affection and intimacy in his personal life, reports feeling lonely, needy and vulnerable, is separated, divorced or unhappily married, and at the time of sexual involvement feels isolated” (Folman, 1991, p. 169).

The metaphor's mapping along with the delineating terminology mark conditions of uncertainty as well as possibilities of transgression. And when the psychological explanations—pathology, experience, and situation—are taken together, everyone is vulnerable to transgressing.

Clinical vignettes

Case study has long constituted a reliable form of evidence in the medical and social sciences, though its use diminished with the advent of evidence-based, quantitative research (Morgan, 2012). Consistent with that evidentiary practice, writings on therapeutic boundaries frequently incorporate case study vignettes, both actual and fictional. Analysis of vignettes located through a systematic sampling of articles on boundaries finds them serving to ground clinical boundaries and reveal dangerous practices. ¹³ The vignettes animate therapeutic boundaries by way of the central characters and narrative form. Whereas the boundary metaphor comprises spatial language suggestive of its existence as it visualizes a particular understanding of therapy, clinical vignettes constitute a counterpart: they at once provide evidence of and perform boundaries’ existence. The vignettes report on and illuminate problems created by certain kinds of characters: boundary violators, vulnerable therapists, and problematic clients. In the case studies, boundaries are constructed in relation to the problematic forms of therapy they would prevent or resolve, effectively casting a patient–therapist–boundary triad. Through such enactments, the vignettes exhibit the characteristics of triadic dramas.

The narrative qualities of boundary vignettes imbue the metaphor with evidential gravity and consequence, employing drama to show violations as a series of therapeutic acts. Numerous cases in the sample narrate boundary crises that involve dissolute male therapists preying on female patients (Gabbard, 1991; Simon, 1989). Several victimize therapists by reporting on problematic female patients who challenge the boundaries of therapy (Gutheil, 1989; Stewart, 1985, p. 173). On rare occasions, a vignette is set to embolden therapists’ deeper and reflective thinking. For example, Plaut (1997) provided dilemmas meant to motivate clinicians’ reflection and action by elucidating the flexibility of boundaries and consequent need for clinicians’ decision-making (see also Margolies, 1990). Plaut hypothesized the decision-making challenges with a seemingly simple case where “you leave the clinic at the end of the day and find your last client alone in the parking lot with a car that won’t start” (Plaut, 1997, p. 84); he thereby narrated an open-ended scenario in which “you,” the clinician, must determine what boundary to establish. A similar dilemma structures a case where a clinician is asked out to dinner by their recently divorced patient, to show how “boundary dilemmas present opportunities to uncover and articulate values essential to professionalism” (Martinez, 2000, p. 45).

In many vignettes, therapeutic boundaries exist in relation to problematic patients and therapists. Boundaries’ existence is indicated in demonstrations of the clinician's need for the existence of boundaries. For example, vignettes exhibit patients with borderline personalities as threatening both the ethics and technique of clinical distance and illuminating the necessity of steadfast boundary maintenance (Briggs, 1979; Simon, 1989; M. H. Stone, 1976). The borderline patient is framed as pathologically seductive in the treatment relationship and as potentially endangering boundaries. Cases starring these patients vividly rehearse their sexual threats, as illustrated in a vignette involving a “stunningly attractive” borderline personality patient and a therapist who could not “resist her charms” (M. H. Stone, 1976, p. 10). This characterization persists in another vignette portraying a “bright and attractive” borderline patient who falls into a sexual relationship with her therapist (Simon, 1995, p. 90). Some vignettes signal dangers of other patient psychopathologies such as the disordered dependent personality and its consequential neediness (Simon, 1989, p. 106) or insatiable sexual desires that closely resemble nymphomania (M. H. Stone, 1976, p. 673). Regardless of the pathology named, the case dramas enact dangers by resituating the patient's pathology, including depression and even schizophrenia, in service of materializing boundary dangers.

The vignettes also contribute to the psychologization of therapists and commonly do so by representing the connection between pathological therapists and boundary problems as a clinical fact. Gabbard (1991), for instance, reported a vignette in which a therapist with psychopathy requests his patient to perform fellatio and, in a humiliating motion, do the “duck-walk” (p. 652). Psychologizing therapists is not limited to diagnosing their pathology: some cases depict nonpathological or normal therapists as being vulnerable, especially due to adverse life experiences. With this claim, therapists suffering from the effects of divorce are identified as potential boundary jeopardizers, as seen in one case, “a Familiar Scenario,” that depicts sexual involvement between a lovesick, divorced therapist and his patient (Simon, 1989).

Situated in boundary texts, the vignettes themselves rarely feature the word “boundary”; rather, the existence of therapeutic boundaries is called forth by problematic patient–client dynamics, the potent temptations of sex, and subsequent discussion in the text. While such calling forth continues over the years, over time, the case narratives shift from describing unconscious to mostly conscious processes, from the intrapsychic domain to the behavioral one. This shift marks divergence from Langs’s (1975) original conceptualization of boundaries. He related a case involving a seductive cotransference between a male patient and a male therapist in which boundaries are repeatedly deviated from and reestablished by both the clinician and patient, as the narrative indicates the unconscious dynamics of transference and countertransference. In Langs’s cases, the adept therapist expertly handles “seduction” and “anger” along with the patient's unconscious induction of countertransference. In contrast, subsequent vignettes presented by other authors feature therapists who recognize the sexual nature of a therapeutic encounter as a conscious dynamic. Mention of the unconscious dwindles in or disappears from later vignettes, as illustrated in a 1989 case of a patient presented as consciously attempting to prey on the sexual needs of the therapist. She tries to guilt her therapist into giving her a ride home, attempts to masturbate in front of him, and eventually thrusts her pelvis against his body (Gutheil, 1989, pp. 600–601). Similar emphasis on conscious erotic acts structures a case example where the psychiatrist simply “falls in love” with her patient (Plaut, 1997, p. 78). In effect, the boundary literature eventually transforms Langs’s therapeutic dyad of patient and therapist separated by a mutable unconscious distance to a triad that summons prophylactic boundaries between patient and therapist to protect them from the risks of sex. Clinical vignettes realized this change through a dimming of the unconscious and psychological characterizations of the participants. Clinical boundaries appear as more clinically appropriate, if not essential, in the face of conscious sexual transgression when compared to nuanced, unconscious dynamics of cotransference.

Critics also employed vignettes but did so as evidence of the metaphor's problematic epistemic foundation. Greenspan (1986) provided clinical cases that showcase the effectiveness of a self-disclosing therapist who transgresses boundaries that are understood to be sexist and patriarchal. One vignette demonstrates a feminist technique whereby the female therapist and her client shared their respective experiences with depression, a sharing that resulted in beneficial outcomes for the patient. The vignette constitutes a counterexample to restrictive techniques of conventional therapy that rely on a masculine myth of “father knows best”—the therapist who “possesses a body of wisdom about transference which is the key to unlocking the door to the patient's unconscious” (Greenspan, 1986, p. 6). Along with feminist responses to the metaphor's gendered epistemics were those advocating relational therapeutic models. In defiance of clinical boundaries, Lazarus offered instances of the therapeutic efficacy of his own boundary violations, including sharing toast and tea with an especially resistant patient (Lazarus, 1994, p. 257). Likewise, Fay (2002) recounts taking a patient with procrastination issues to the library as a means of motivating him. These vignettes by Greenspan, Lazarus, and Fay showcase the “good effect” (Lazarus, 1994, p. 259) of crossing boundaries and the “dehumanizing” (p. 256) effects of holding rigid ones. Such dehumanization is demonstrated in another personal vignette in which Lazarus called a patient to inquire about a stressful surgery that the patient had mentioned in a session, only to be criticized by the patient's wife and the psychotherapeutic community for violating professional boundaries. He remarked how “a simple act of human decency and concern had been transformed into clinical assault” (1994, p. 256).

By summoning the metaphor, the cases furnished by writers from both the endorsing and critical camps shape and affirm the ontology of boundaries, pointing to their existence in therapy. Only on rare occasions were the interpretive implications of the metaphor questioned. Palombo (1987), for one, vehemently argued that therapeutic boundaries constitute a conceptual, not practical condition, and questioned how they alter interpretations of therapy. To expose the interpretive consequences of boundaries, he reconstructed a case presented by Schamess (1986) and reinterpreted that clinical episode in terms of the psychic interplay between the interactants. Schamess’s original case centers on an impoverished woman with borderline personality disorder, Mrs. Jackson, who requests free sessions. The patient's inability to afford the treatment fee presented Schamess with a dilemma between a concerned response and “boundary maintenance.” In Palombo's reinterpretation, the case does not revolve around therapeutic boundaries or the frame but, rather, the unconscious meaning of the patient's request that indicated an unconscious desire for an omnipotent parent (Palombo, 1987, p. 292). With this reinterpretation, readers are entreated to abandon boundaries as an effective understanding of the patient and, instead, to closely reflect on and engage “therapeutic intimacy.” These diverging assessments of the same case illustrate the perceived and interpretive as well as practical implications of the metaphor's adoption.

Troubling the metaphor

Metaphor gives partial understandings of a thing: just as it exposes certain features, it hides or obscures others, including aspects of the world that are inconsistent with the metaphor. As intimated in Palombo's critique, metaphor often conceals the fact that it is a metaphor and, importantly, masks the ontological or epistemic vision that grounds it. Whereas the clinical boundary metaphor's elaborations and animations bring to light border transgressions and manifest the participants’ (psychological) status, the logic they idealize remains backgrounded. And whereas the literature's language, cultural excesses, and dramas highlight the lines demarcating realms of safety and danger, what generally remains obscured are the ways those lines sanction certain psychotherapy systems and not others. Although proffered as a metaperspective on therapeutic practice, one supposedly transcending ethics codes and therapeutic technique, the boundary metaphor poses an idealized model of therapy that effectively eschews or proscribes others. Wide-scale adoption and endurance of boundary discourse suggest tolerance if not endorsement of this idealization. However, what ideals the metaphor ultimately manifested and what horizons it darkened did not go unnoticed by everyone. A few exposed the metaphor's authoritative point of view, showing its root connections to psychoanalytic premises of neutrality, abstinence, transference, and countertransference. Most critical interrogations target one implication of maintaining boundaries: the prohibition of self-disclosure and dual relationships. Taken together, the appraisals discern much of what the metaphor masks.

Nearly from the start, the metaphor was charged with censoring, even prohibiting, some forms of psychotherapy. For some, like Palombo, its exclusionary perspective along with illogic constitute grounds for the metaphor's dismissal. He found the metaphor to confuse boundaries as set limits of the treatment situation and as mandating a “psychological separation” between therapist and patient (Palombo, 1987, p. 286). Further, it conflates ego and self-boundaries and vacillates on the matter of the boundary's qualities (rigid, porous, flexible, etc.). In place of the metaphor, Palombo advocated a “therapeutic milieu in which the therapist can achieve empathetic attunement with a patient's experiences,” a milieu fostering “intimacy within the context of absolute safety and predictability” (p. 290).

Most critical assessments, however, center on specific features of the therapeutic relationship that the metaphor naturalizes. These condemn strict disallowances in the therapist–patient relationship, such as proscription of therapists sharing personal information, altering fees, or modifying practically any established treatment parameter. Such inflexible practices of therapist–patient distance were deemed counter-productive, potential causes of therapeutic failure (Stewart, 1985) and reflections of a therapist's “defensive inflexibility” that inhibit beneficial interventions (Glass, 2003). Considerable criticism addresses the banning of “dual relationships” that prohibits any contact between therapist and patient outside the treatment parameters. ¹⁴ With related concerns, critics cited the metaphor's theoretical exclusiveness, claiming how “Non Psychodynamic approaches to boundary issues are marginalized if not delegitimized” (Fay, 2002, p. 164). Apparently frustrated by these strictures, one quoted a kindred critic calling out the “boundary police” who rigidify the therapeutic relationship (Greenspan, 1995, as cited in Lazarus, 1998, p. 144). Charging that dual relationships had become demonized, Ofer Zur analogized the metaphor's unsound premise: “To assert that self disclosure, a hug, a home visit, or accepting a gift is likely to lead to sex is like saying doctors’ visits cause death because most people see a doctor before they die” (Zur, 2000, p. 2). ¹⁵ A critique occasionally sparked publicized debate between therapists with divergent perspectives. Such was the effect of a paper decrying “anxious conformists[']” transformation of ethical guidelines into rigid, artificial boundaries (Lazarus, 1994, p. 255). Impassioned responses to the paper came from endorsers and critics alike (Bennett et al., 1994; Borys, 1994; Gabbard, 1994; Gottlieb, 1994).

Largely sidelined from the mainstream were feminist analyses of therapeutic boundaries that honed in on the metaphor's ideological premise and misunderstanding of personal boundaries. Feminist rebukes of canonical therapy's patriarchal and hierarchical power structure were accompanied by demands for respecting the personal boundaries of therapists and clients alike (L. S. Brown & Walker, 1990; Greenspan, 1986). As it was tersely put, a women-oriented approach says “good-by to the Expert–Patient model of therapy in favor of a more equalized relationship” (Greenspan, 1993, p. 234). Alternatively endorsed were egalitarian and collaborative therapeutic relationships that unavoidable breach the dual relationship prohibitions of clinical boundaries (Heyward, 1993; Lerman & Rigby, 1990; Ragsdale, 1996). Importantly, the boundary framework was charged with completely failing to challenge the “undeniable power imbalance when one person, the therapist, sets the rules and collects a fee, while the other person, the client, must adhere to the rules and pay the fee” (Margolies, 1990, p. 20). The boundaries metaphor also objectifies the client, elides her lived experience, and truncates collaborative therapeutic relations. As the metaphor gained professional usage, feminist therapists registered stronger criticisms. Asserting that boundaries reveal the profession's hierarchical power, Greenspan quoted Habermas’s depiction of professionals as “social engineers and inmates of closed systems” (Habermas, 1973, as cited in Greenspan, 1993, p. 203). Clinical boundaries “do not exist in reality,” and boundary language “camouflages the political dimension of violence against women” (Greenspan, 1995, pp. 130–131). Admitting to promoting boundaries before discerning their alliance with the dominant cultural order, Laura Brown proposed a centering on therapy's “relational nature” to rely not on rules but, rather, on the therapist's continuous “thinking, feeling, and making sense of situations,” coupled with a feminist ethic of empowering the client (L. S. Brown, 1994, p. 37). Not all shared this viewpoint. Looking back on the previous 2 decades, Tanya Dineen observed that many feminists inadvertently participated in a sexualization of boundaries, fueling a moral panic that conveyed an unfounded perception of vulnerable patients and knowledgeable therapists; in so doing, boundaries “invited an ideology of power” (Dineen, 2002, pp. 115–139).

Later critics further interrogated the metaphor's epistemic foundation, often emphasizing one authority behind the metaphor: psychoanalytic psychiatry. As one critic put it, “the modern boundary movement” is “a runaway train propelled by psychodynamic ideologies” (Fay, 2002, p. 164). A critic of the apparent psychoanalytic grounds of dual relationship prohibitions (understood as a remnant of transference–countertransference) produced an alternative configuration of therapeutic boundaries (Zur, 2002, 2007). These resistances to the metaphor's authoritative purchase were usually, though not exclusively, grounded in nonpsychoanalytic perspectives. For instance, one defender of “other psychic models,” including psychoanalysis, condemned the metaphor's “entwining of technique and morality” (Goldberg, 2008, p. 873). Another drew from behaviorist and especially humanist therapies’ “tearing down of boundaries between therapist and patient” to highlight the “conservative view of boundaries” (Williams, 1997, pp. 241–242). Some appraisers contested the metaphor's erasure of cultural context (Combs & Freedman, 2002; Kroll, 2001) and its constrictive definition of nontherapy relationships. Whether or not these often-acidic commentaries were heeded, many supporters of clinical boundaries eventually condoned a modicum of boundary flexibility (Celenza, 2016; Gabbard, 2016, 2021), and the 2002 APA ethics guidelines lessened prohibitions against multiple relationships (APA, 2002; Knapp & VandeCreek, 2003). Observing the shifts in meaning, one critic observed how boundary guidelines change as “new perspectives, economic arrangements, and sociopolitical and risk-management pressures impinge on and shape” both psychotherapy practice and theory (Kroll, 2001, p. 274). Other critical excavations of the metaphor's concealments and obstructions nevertheless continue using alternative perspectives that range from queer (Beroset et al., 2024) to psychoanalytic ones (Barratt, 2015; Cooper, 2016). Some assessments employ psychological concepts and diagnostics, undertaking such reflexive analyses to make sense of boundary adherents’ motivations to engineer and vehemently guard clinical boundaries. One psychoanalyst, for instance, compared the “ideologically internalized” boundaries with the nontransgressible barrier of incest (Barratt, 2015, p. 212). Another interpreted the concept of sexual boundary violations as a psychological “folk theory”: it is “an institutionalized shared disinclination—whether we would call it unconscious or unthought-known” that “functions to preserve the group's standing as a good object but interferes with thinking about its problems and flaws” (Dimen, 2016, p. 363).

Conclusion

The phenomenon of therapeutic boundaries originated with no empirical discovery, momentous event, determinative theory, or relevant legislation. Rather, the novel metaphor-driven conception of psychotherapy was generated through a consolidation of then-current professional expectations along with hopes and fears about therapeutic practice. Constructed by psychotherapy professionals, clinical boundary terminology can be understood as a case of “performative nominalism” where expert classifiers are implicated in (or are members of) the kind being classified (Pickersgill, 2019). Arguably “derived from ethical, moral, professional and legal duties” (Simon, 1995, p. 90), the novel terminology reassembles therapists’ work in a different way. Like other identified cases of performative nominalism, the metaphor's “attribution of ‘newness’ has contributed to performing and propelling innovation, rather than solely reflecting it” (Galasso et al., 2024, p. 139). Developed with distinct recognition of an impropriety in therapy that was being noticed by the public and professionals alike, clinical boundaries constituted a novel disciplining device that gave therapists a comprehensible and usable map of therapy's limits and their own comportment. Along with acknowledging the dangers of sexual misconduct in therapy, proponents and allies alike deployed a reflexive psychologizing, drawing upon diagnostic knowledge, to identify potential perpetrators and perpetrators alike practitioners. The case studies vivify these theorized psychological predispositions. However reassuring cases of pathology might be to self-identified “normal” therapists, less reassuring are claims of the risks of ordinary life experiences. The metaphor's rhetoric thereby trades on senses of fear and safety; every practitioner is potentially at risk. ¹⁶

The innovative metaphor circumvented troubling debates within psychoanalysis and reigned in some nonpsychoanalytic and purportedly more liberal therapies. By introducing novel nomenclature, the boundaries metaphor sidesteps numerous controversies among psychotherapy schools and smooths over ideological differences, translating psychoanalytic principles of abstinence, neutrality, transference, and countertransference into nonpsychoanalytic language of boundaries, crossings, slippery slopes, and violations. ¹⁷ It constitutes a common language that could be shared among therapists with different training and technical practices; although beyond the focus of the present analysis, it was likely shared with clients, giving them a clearer understanding of the therapeutic relationship. Adhering to these clinical boundaries requires neither knowledge of nor belief in an unconscious since the metaphor's metaethics rhetoric is usable without technical training in unconscious processes. In fact, it conveys quite different expectations of the therapist than were proposed in its early appearance. Langs, a so-named pioneer of the metaphor, admixed psychoanalytic and boundary language to describe “psychoanalytic interpersonal boundaries” (1984, p. 7) as “spiraling conscious and unconscious communicative interaction between patient and therapist” (p. 22). Therapists needed “Basic psychoanalytic methodology,” not “preconceived theories or naïve, surface-oriented assessments of manifest meanings and reactions” (Langs, 1982, p. 114). His understanding was not sustained, and in place of extensive trained judgment for interpreting the deep meanings of therapeutic interactions, and in place of expectations for a therapist's own ongoing analysis, boundary expositions came to pronounce different responsibilities. ¹⁸ And these responsibilities gained further credence with the rise of risk management perspectives on therapeutic practices.

Although a putatively simple, lucid map, clinical boundaries compound therapists’ responsibilities. A therapist must know and abide by the metaphor's spatial guide and routinely self-monitor; they can be aided by self-administered tests on vulnerabilities and “mindful awareness” of vulnerabilities (Epstein & Simon, 1990; Gottlieb & Younggren, 2009; Johnston & Farber, 1996; Koocher & Keith-Spiegel, 2008; Pope & Keith-Spiegel, 2008, p. 640; D. Smith & Fitzpatrick, 1995). The proliferation of boundary guides on therapists’ self-monitoring behavior shows the shift from describing unconscious processes to conscious ones, from interpretation to behavior analysis. Even a critic of the metaphor's exclusivity appreciated the “greater courtroom utility” enabled by the elimination of alienating psychoanalytic language (Williams, 1997, p. 240).

As evidenced in studies of performative nominalism, novel terminology for professional activities can “performatively contribute to shaping these endeavors in different—albeit connected and overlapping—ways in an unfolding future that is always embroiled with the language that precedes and constitutes it” (Galasso et al., 2024, p. 152). ¹⁹ Debate about clinical boundaries’ restrictions and authoritative purchase continues, yet the metaphor’s performative language holds a place in textbooks, pedagogy, ethics guides, articles, and conversations. Just as critics elucidate just how “Something There is that Doesn’t Love a Wall” (Beroset et al., 2024), so the APA sponsors books such as Succeeding as a Therapist, with its nine safe action steps for navigating boundaries (Adames et al., 2023, p. 164; Knapp & Fingerhut, 2023). Clinical boundaries have become something of an “optional ontology” (Shamdasani, 2010) that rescripts the therapeutic encounter to (re)constitute therapy as conscious behavior acts and the therapist as vulnerable yet nevertheless responsible for self-protection.