Take Care, but Not Too Much
The Invention of Codependency
Lily Meyersohn
Illustration by Susan Wilkinson
Do you try to please others?
Do you feel responsible for other people’s happiness?
Do you anticipate other people’s needs?
Do you wonder why other people don’t anticipate your needs?
Do you feel harried?
Do you feel angry?
Are you afraid of being alone?[1]
Since the 1970s, the term “codependent” has been used in both psychotherapeutic and self-help spaces to label one member of an unhealthy relationship, most often when the other member of the relationship is struggling with substance use, alcoholism, or another “dependency.”[2] Like other mental health idioms and diagnoses, “codependency” has been emptied of its original clinical context. Half a century on, codependency continues to refer to and pathologize a feminized identity based in caretaking and responsibility for others.
It would be easy to reject the language of codependency wholesale. Indeed, in recent years it has become fashionable to dismiss codependency as yet another harmful pop psychology label. In 2022, The New York Times ran an opinion piece by Maia Szalavitz arguing that the overuse of the concept of codependency has been toxic to people touched by addiction. The label, Szalavitz wrote, stymies efforts at recovery and drives loved ones apart. But an untold number of Americans—some publicly, many in private—have identified with the term since its debut. Co-Dependents Anonymous, or CoDA, still has over one thousand meetings in the US and convenes in more than 70 countries.
Today, “codependency” is a quasi-diagnosis and a dirty word. It is an American word that has traveled abroad. It describes a self participating in relationships in a specific way: a “bad” way. A self whose boundaries are too fuzzy, a self who is too giving, too controlling, too invested. That self has been created by a national culture that implicates its subjects in psychopathology. That self is a key to the culture’s pathology.
The history of codependency reveals something about the difficulty of sustaining and cultivating intimate relationships; about what relationships must look like to be deemed “healthy” by outsiders; about how feminized care became a site of diagnosis; about the fragile link between everyday suffering and clinical categories; and about how psychoanalytic ideas get taken up in the world of pop psychology, where they go on to lead lives of their own.
PREHISTORIES
The germ of the codependency model dates to the 1830s, when mesmerism had just arrived in the United States from Germany and France and the therapeutic ethos was taking off. In his expansive cultural history of psychotherapy, Constructing the Self, Constructing America, Philip Cushman describes how mesmerism, a psychological healing technology based on “animal magnetism,” used hypnotism to access the mind and help its recipients liberate their interior selves from the yoke of other people’s self-interests.
Through the 19th and 20th centuries, theorists increasingly conceived of sociopolitical problems in the psychological terms introduced in mesmerism’s wake. The theories that found purchase in the US were those that successfully mirrored the dominant culture and contributed to an understanding of the self-contained individual. Object relations, for instance, which put forth a vision of interiority while maintaining an individualist frame, took deep root in the US. Through her emphasis on unconscious processes like introjection and projection, the consumption and expulsion of objects, and the gratification the infant receives from nursing, Melanie Klein located the social world within the inherent psychic structures of the individual. Klein envisioned an empty self that feels, in Cushman’s words, “naturally and irresistibly driven to consume.” Interpersonal consumption is the way to create meaning. This view was not singular to Klein, notes Cushman; rather, it can be attributed to the overarching consumer society. It constitutes a throughline from mesmerism all the way to codependency.
Like object relations, ego psychology posited that children develop by taking things in. Child analyst Margaret Mahler, for example, offered a clear-cut schema of how a child grows out of symbiotic unity with the mother and blossoms into an autonomous individual.[3] Mahler’s separation-individuation theory was, according to Jessica Benjamin, a “landmark in the theory of the self,” which stressed a uniquely American vision of individuality and autonomy. The value of individualism lingers in other psychodynamic schools of thought; even family therapy, ostensibly more systems-oriented, has struggled to move outside the nuclear family.
The seventies ushered in Heinz Kohut’s self psychology. Kohut theorized that the self is formed in infancy and is the product of parent-child interactions. Kohut saw infants as “psychological consumers” that develop by using up their parents, writes Cushman. The self longs for what Kohut termed a selfobject, someone or something outside but integral to the self and necessary for normal functioning, to fill its emptiness. Though adults continue to need healthy selfobject relations, deficits first experienced during childhood can lead to all kinds of psychological symptoms: emptiness, anxiety, shame, perfectionism, rage, the uncontrollable urge to merge with others—all, tellingly, symptoms of the disease of codependency that would grip the nation come the 1980s.
Object relations and self psychology may have been less threatening to the psychotherapeutic profession than the theories of interpersonal-relational psychoanalysis. The interpersonalists saw the self as a relational process or system, not an empty container. Theorists whose ideas were more grounded in the social and cultural worlds—Harry Stack Sullivan, Erich Fromm, and Karen Horney—posed more of a challenge to the profession’s status quo. Horney’s cultural critique in particular did not find favor in this climate.[4]
“A self whose boundaries are too fuzzy, a self who is too giving, too controlling, too invested. That self has been created by a national culture that implicates its subjects in psychopathology. That self is a key to the culture’s pathology.”
Horney, one of the first women to be trained as a Freudian analyst, has everything to do with the origins of the codependency model: many scholars tie the concept back to her. Horney believed that environmental and social factors led to the neuroses she saw in patients; early anxiety-generating experiences with caregivers were especially significant. In her 1942 Self-Analysis, Horney aimed to bring psychoanalytic self-examination to a wider public. Psychoanalysis was growing popular not just for people suffering severe mental disturbance, but also as an “aid to general character development.” As part of this project, Horney outlined a theory of neurotic needs, including the need for affection and approval, the need for a “partner” who will take over one’s life, and the need to restrict one’s life within narrow borders.
In Horney’s descriptions, we hear intimations of the phrases that would one day be incorporated into the language of codependency: the neurotic’s “[i]ndiscriminate need to please others,” the “[d]read of being alone,” the “[c]enter of gravity entirely in [a] ‘partner,’ who is to fulfill all expectations of life.” They seek, as Horney put it in Neurosis and Human Growth, to “solve their early conflicts with people by ‘moving toward’ them.” They compulsively seek fulfillment from others and struggle to develop a strong sense of self.
Horney’s theory originated during a proliferation of functional diagnoses and psychopathologies. In a recent conversation, Lisa Mendelman, an expert in the medical humanities who focuses on gender, race, and affect in twentieth-century America, plainly reminded me that “being codependent doesn’t land you in a mental hospital.” Instead of dwelling on the insane/well or psychotic/neurotic distinctions, codependency attempts to diagnose people who seem healthy and functional but exhibit, as Mendelman says, a “covert unhealth.” Unlike other psychopathologies historically associated with women (e.g., hysteria, borderline personality disorder), codependency was, even in its earliest precursors in Horney’s work, a hidden psychological impairment. Horney wrote Self-Analysis in good faith, to bring psychoanalysis to a wider public. Still, its afterlife may have contributed to a culture of psychological self-surveillance.
When I reached out to the Karen Horney Clinic in New York to ask about the relationship between Horney’s legacy and codependency, the clinic’s director had a firm reply: “Horney had nothing to do with codependency. It originated in addiction world.” I posed the same question to other analysts and received similar replies; all were quick to disown a possible relationship. The analytic world might feel anxious not to be associated with the concept of codependency, since many assume that the concept has been used to pathologize people, especially women. “There is a certain liberal ethos that claims all diagnosis, all pathologizing, is bad,” notes Mendelman, “but the place where pathology goes awry is when the problem is located in the individual rather than in the superstructure.”
Nonetheless, a strong historical link exists between psychoanalysis and early theories of codependency, even if the concept has gone through its own process of separation and individuation in the last 40 years. By taking a closer look at codependency, at the diagnosis itself, we may be able to understand more about the culture that pathologizes caretaking while simultaneously demanding that the caretaking be done—quietly, in the background and, so often, for free.
WE ADMITTED WE WERE POWERLESS OVER OTHERS
The creation of the concept of codependency as we know it would have been impossible without the central involvement of the twelve-step movement. By 1951, Alcoholics Anonymous was 16 years old, and Al-Anon had just formed to support loved ones of alcoholics. Al-Anon groups followed the same rituals and steps that AA followers did; unlike AA followers, who struggled to control their alcohol use, Al-Anon members struggled with the false belief that they could influence and control another person’s addiction. Twelve-step membership grew throughout the 1970s and 1980s as more women joined. Offshoot programs, like Alateen and Adult Children of Alcoholics, sprouted up.
Through the 1950s, Al-Anon members were encouraged to change their lives through pioneering selfhelp approaches like The Power of Positive Thinking. But in the 1960s, twelve-steppers of all fellowships allied themselves with the professional mental health field, including psychoanalysis. Al-Anon took cues from Horney and other analysts, viewing the behaviors of loved ones who “moved toward” the addict as neurotic symptoms. These “moving toward” behaviors, as Horney described them, were motivated by a “drive for total surrender,” “the longing to find unity through merging with a partner,” and the “drive to lose oneself.” Such behaviors were seen as responsible for perpetuating addiction and as typical of spouses. When family members ostensibly interfered with recovery by over-helping and “enabling” their alcoholic loved ones, they came to be known as “co-alcoholics.” The medical profession increasingly considered alcoholism a physical disease rather than a moral sin; medicalization led advocates and researchers to label alcoholics as “chemically dependent,” a term that was seen as more precise and less stigmatizing. “Co-alcoholic” was thus replaced by “co-chemically dependent”—which was soon superseded by “co-dependent” (less of a mouthful).[5]
Guided by psychoanalytic perspectives and new family therapies, the twelve-step movement considered codependency a bona fide psychological disorder. Attachment theorists, addiction theorists, and family therapists agreed with psychoanalysts that early interactions with family shape adult relational patterns and psychopathologies. Researchers also began focusing more and more on early-life trauma with the explosion in diagnosis and theorizing related to PTSD.
The focus on family systems—a concept developed by psychiatrist Murray Bowen and, by the late 1960s, central to family therapy theory and practice—meant that “codependency” cropped up regularly in mental health literature in the 1970s. Systems theory posited that addiction was a family disease. Codependency originated in “dysfunctional families,” a concept which encompassed a range of behaviors and dynamics. The addict was no longer seen as the sole instigator of family dysfunction. Instead, the entire family system was dysfunctional.
“By taking a closer look at codependency, at the diagnosis itself, we may be able to understand more about the culture that pathologizes caretaking while simultaneously demanding that the caretaking be done—quietly, in the background and, so often, for free.”
Co-Dependents Anonymous (CoDA), founded in 1986 in Phoenix, brought the twelve-step philosophy and its disease model to a wide range of people dealing with problems in relationships: dependency, intimacy problems, dysfunctional communication, boundaries, controlling behaviors, people-pleasing, self-doubt, mistrust, perfectionism, high reactivity, enabling, and obsessiveness. To the program, codependency was a treatable illness. Rather than waiting for the medical establishment to diagnose patients, meetings encouraged self-diagnosis and self-identification. CoDA still operates this way.
While twelve-steppers were self-diagnosing, researchers in the clinical sphere were coming up with their own models of codependency. Timmen Cermak and a handful of other psychologists argued that codependency should have operationally defined diagnostic criteria. Cermak, a psychiatrist and self-described codependent, came to believe that codependency was a personality disorder after working with the family members of alcoholics at the Stanford Alcohol Clinic. In 1986, Cermak wrote Diagnosing and Treating Co-Dependence and proposed that codependency should be considered an Axis II personality disorder in the DSM-III-R, to be published in 1987.
Cermak’s push for inclusion in the DSM-III-R reflected the psychiatric profession’s long interest in developing a public language. The publication of the DSM-III in 1980 had helped the public recognize itself in the updated manual’s checklists. Pia Mellody, like other codependency movement figures, was relieved to see psychologists like Cermak finally taking up codependency as a serious issue at a time when the field was still on “a primitive frontier with regard to the investigation of [this] serious personality disorder.” It was a hopeful moment for researchers and activists.
The same year that Cermak pushed for DSM inclusion, the concept of codependency burst out to a wider audience with the publication of Melody Beattie’s Codependent No More, a self-help bestseller. Beattie, who wrote twenty books on codependency and related issues, pinned numerous psychological problems on codependency. Beattie, and many who followed in her footsteps, saw alcoholism as a family illness that caused codependency, “similar to catching pneumonia.” It doesn’t matter who gave it to you. Once you’ve got it, Beattie wrote, “your codependency becomes your problem.”
Everyone seemed to want a piece of this healing technology. Codependency became, as two psychologists wrote by 1991, a “chic neurosis.” Clinicians tried to treat it. Support groups spread its teachings. Researchers studied it, crafted scales to measure it, and tried to develop a unified definition to describe it. Robert Burney, a self-described “counselor for wounded souls,” went further. “A Transformational Healing Process has begun on planet Earth,” Burney wrote. “Due to a profound change that has taken place in the energy field of Collective Human Emotional Consciousness, resources are now available to us to do healing that has never before been possible in recorded human history.”
The uses of the term were stretched to include families and relationships untouched by substances. Beattie’s counterparts placed the onus of codependency on dysfunctional or abusive family systems. Pia Mellody saw codependency as generationally transmitted and developed a therapy for it at The Meadows, an addiction treatment center in Arizona. She believed codependents were locked in a “joint sickness” with the addicts they loved. All across America, Mellody wrote in Facing Codependence, families suffered from this “painful and crippling disease.”
To others, like Anne Wilson Schaef, codependency was an insidious, pervasive disease in its own right. Schaef felt that codependency, like any other physical disease, came with predictable outcomes and complications: gastrointestinal problems, high blood pressure, cancer. But Schaef’s 1986 Co-Dependence: Misunderstood, Mistreated argued that codependency was a disease within a larger addictive process inherent in American culture. She tied codependency to the women’s movement: “The nonliberated woman and the co-dependent,” Schaef argued, “are the same person.” In 1990, codirectors of the National SelfHelp Clearinghouse echoed in TheNew York Times that “[t]he codependency movement may well be the psychological arm of the women’s movement.”
The women’s movement, which encouraged women to detach from demanding caregiving roles, aligned well with the construct of codependency. Both demanded that women take responsibility for their relationships. On a sociohistorical level, the construct was a “compromise formation,” Janice Haaken, a psychologist and critic, told me. On one hand, the construct broke away from some of the outdated psychiatric diagnoses that bound women to debilitating pathological categories, like hysteria and histrionic disorder. Thinkers like Schaef saw codependency as arising from the grassroots twelve-step movement, which embraced self-identification and operated outside the expert-authority model. For many women, the term offered them the power to decide whether or not they had a disease or a problem. In this sense, the self-help movement and pop psychology literature on codependency, as evidenced in the work of Beattie, Mellody, and Schaef, among others, ran parallel with a new generation of women clinicians who were flowing into the mental health professions and into psychoanalytic institutes at the time and hewing more closely to interpersonal models. For these clinicians, codependency looked like a tool of, and a partner to, women’s liberation. The construct was intensely attractive, offering both critique and cure.
Codependency discourse operated as feminism’s epistemic twin. The concept rightly recognized that many women suffered because they were embedded in harmful systems. Proponents of the term attempted to extend agency to these women by helping them detach from relationships in which they had previously felt trapped. On the other hand, codependency diagnosed care itself as a site of potential pathology, reinforcing the belief that relationality and dependency are failures of development. Codependency preached a near-unilateral detachment from struggling family members and largely failed to identify root causes outside of the nuclear family. Haaken has written that the literature offered “narrow psychological explanations of feminine pathology consistent with the conservatism of the 1980s.” Even systems theory relied on framing family problems through the lens of family deficiency rather than through a more political or structural lens. Families, Haaken writes, were seen as “governed by internal, pathological ‘rules,’” somehow separate from wider historical or political structures.
The dysfunctional family is given permission to feel that the codependent is just as much of the problem as the addict in the family. According to the logic of the codependency literature, Haaken writes, the “person who depends upon drugs for a sense of well-being is the same as the one who depends upon people for the same feelings.” It was the concept’s over-reliance on family systems theory, she goes on, that inadvertently created a “noblaming world of moral equivalents.” Though all of these issues are highly dynamic and relational, the literature on codependency flattens them. In the literature, Haaken observes, “women, in the role of wives and mothers, are in the ambivalent position of being both emotionally protecting and potentially ‘castrating’ or overpowering”—thereby perpetuating a fantasy of the omnipotent mother.
MOVING TOWARD
Perhaps it is true that good-enough mothering goes hardly noticed, hardly remembered. But imagine an adult in an intimate relationship still wanting care to go unnoticed. Who wants to take care in without it making too much of an impression. Only when the care makes itself too known, perhaps, does its provider become pathological. When the love is enmeshed—or enabling, controlling, loud, hysterical, too much, wrong in some other way—that is when it is codependent love.
The label of codependency says something is wrong with someone for straying too far outside herself. By contrast, when we “act masterful and bounded,” writes Cushman, when we feel “self-contained [and] autonomous [. . .] we experience an emotional affirmation of our way of being.” The underlying message is powerful: the desire for intimate, dependent relationships comes to demonstrate a lack of self-development, writes Marion Solomon in her 1994 Lean on Me: The Power of Positive Dependency in Intimate Relationships. Adults who harbor a great need for others must work hard to eradicate that feeling. This need for others is assumed to be a lingering feeling caused by the early internalization of an inadequate parental object. The child is helpless against such passive consumption. Inadequate parenting creates the possibility that we will be fated to pick bad partners over and over.
These assumptions are accepted in psychodynamic thought as well as in the twelve-step and addiction theories that were so influential to the development of the concept of codependency. In addiction theory in particular, Cushman notes, addicts are seemingly transformed via their “continual ingestion of a drug, the bad commodity.” To regain control, they must abandon the commodity in its entirety. The same holds true for the codependent.
Most literature advises the codependent to pull back from loved ones, to resist their rescuing tendencies, and to work on themselves. Al-Anon told codependents that they were just as sick as the addicts they loved. Beattie told codependents to step back from caregiving. The goals of codependency treatment, notes Robert Weiss, who coined the concept of “prodependence,” revolve around detachment, self-actualization, and the disavowal of the human need to connect.[6]
“When the love is enmeshed—or enabling, controlling, loud, hysterical, too much, wrong in some other way—that is when it is codependent love.”
This language still distorts our ideas about how relationships should look and at what point they can no longer be deemed healthy. Codependency theory presupposes that there is such a thing as a healthy, non-dysfunctional family—or a healthy, non-dysfunctional relationship—and assumes that relationships should be, above all else, healthy. That it is possible to become overly invested in someone else’s life or behavior. That if someone is overhelping, then she is controlling and enabling and ought to step away. That there is more to be gained through separation and individuation than there is to be gained from staying close. That we discover who we are alone, rather than in the context of our relationships.
If two people in a relationship are experiencing a problem, the logic goes, then the problem must exist within it or within one of its members. That logic doesn’t leave us room to encounter problems or patterns in the spaces between or across relationships. If one member of a relationship is struggling, then on one level, the codependency model does something radical by saying Look at the other person, too; they might be part of the problem.
That is a fundamentally interpersonal, rather than intrapsychic, move. But the framework stops short of real change because its focus never expands beyond the codependent’s interiority. The language of codependency thus remains limiting, though its limits tell us about the limits the dominant society imposes on ways of caring, relating, and loving.
It is widely understood that people with similar early-life challenges bond as adults. Weiss accepts the pervasiveness of what Freud saw as the repetition compulsion. At times, he writes, we will all recreate the dynamics of our childhoods. Codependency theory likewise presumes that people stay in “bad” relationships, or else repeatedly pair off with “bad” partners, to recreate those dynamics—that is, as a form of trauma repetition. Codependency theory describes bonding as little more than a reflection of unresolved early-life trauma.
Yet the human personality matures in the context of adult relationships. Even traits like autonomy, self-worth, self-actualization, and self-esteem—highly valued in an American context—develop through relationship, notes Solomon. Weiss’s model of prodependence agrees that it is this network of relationships that offers an “opportunity to thrive,” even in challenging times. Forging an adult relationship with someone whose early experiences mirror one’s own offers a chance to grow. By nurturing intimate relationships as adults, we may give ourselves a chance to heal early-life wounds. The codependency model doesn’t reflect these possibilities. In a world where children so rarely exit the first years of life even relatively unscathed, we can still continue moving toward one another as adults.
The language of codependency assumes that relationships aren’t supposed to suck up too much of us. Other people shouldn’t consume us—we’re supposed to consume others! Our senses of self should be stable and protected, and if one’s sense of self is too bound up in another person, the only option is to leave. If one person is asking too much and the other person is giving too much, then we must detach. If one person in the relationship is struggling, then we should come to the same conclusion: detachment. I don’t think we can let all relationships—the so-called healthy, the messy, the complicated, the abusive—be permeated by this thinking.
Certainly, one should feel able to leave a relationship if one wants that. Too many relationships are harmful and violent. Too many women are pulled to devote their time, energy, even life force, to male partners struggling with addiction or mental illness. These women should surely be able to leave their partners and receive support in doing so. It is not my role—not as a writer, nor as a psychotherapist—to determine which relationships should be cultivated and which should be abandoned. What matters is how the people in a relationship experience what goes on inside it. I’m not interested in suggesting how much one person should do for another, how much to give and how much to take: neither in telling people who identify as codependents to stop trying to help, nor in telling people who identify as addicts to stop depending. Codependency discourse does just that. Some may find that message empowering—but for others, it creates only more suffering.
In 2025, however, the term has not lost its mass appeal. This September, beloved “blackout codependent” Elizabeth Gilbert published All the Way to the River, which chronicles her journey through love addiction. Gilbert’s story represents the extremity of codependency, the worst places it might go. The memoir will surely shepherd many readers toward the term, which remains enormously influential even without Gilbert’s endorsement. Online, therapists list codependency as a specialty on their Psychology Today pages. Couples of all stripes carry the label for themselves.
What should be troubled is whether the language of codependency helps people in a wide range of circumstances make these kinds of necessary and difficult decisions about their lives. Indeed, there are other ways we might think about staying with or leaving one another—particularly in relationships where dependency is present but perhaps more equally distributed over time, where the experiences of struggling and of caring flow back and forth depending on circumstance and mutual agreement. Here, we might assume that each partner will struggle, that each already was and will again be damaged, first as children and now as adults. Perhaps we should assume, too, that even if we pick bad partners over and over, there still might be ways to love, to stay connected, that aren’t pathological. From the very beginning, this interdependence, the growing mutual recognition of self and other, is our birthright.
There is a reason that people misuse the word. It is a slip. We say, “they’re codependent,” as if codependency is something that exists between people. But the term was never constructed to work that way; it was always constructed to refer to one person alone. Thinking psychoanalytically, we might say that our modern usage belies a secret wish: to be, in fact, more dependent, even enmeshed. The wish that the world does not ask or force us to detach or abandon or individuate or disconnect. That we do give more of ourselves over to each other, perhaps even all of ourselves.
[1] Adapted from online codependency checklists.
[2] In codependency literature, the member of the relationship who is struggling with substance use or addiction is by and large referred to as an “addict” or “the addict.” In this essay, I at times use “addict” or “the addict,” especially when the original text I am quoting does so. I trust that the reader can keep in mind that this is both the accepted term in much of the literature and also a problematic one. It should be noted as well that when “codependency” originated, alcoholism and other addictions were viewed as “dependencies.”
[3] Some see Mahler as part of the object relations school, but she thought of herself as an ego psychologist.
[4] In 1941, Horney left the New York Psychoanalytic Institute following conflict over her ideas about feminine psychosexual development, which diverged sharply from classical drive theory. One version of the story has it that she was expelled; another, per Bernard Paris, is that she was demoted from instructor to lecturer and restricted to only teaching electives in 1941, which led her to resign.
[5] For the purposes of this piece, I use codependent without the dash. The dash went out of style after self-help literature on the topic became popular, but some groups like Co-Dependents Anonymous still stylize it the original way.
[6] Weiss coined the term “to describe attachment relationships that are healthfully interdependent.” He uses prodependence to specifically focus on addicts and their families, as he feels that codependency “continues to be actively promoted” in that arena in particular.