What if Organizations Need Therapy?
Organizations sometimes go through tough times. A major change is handled badly. Trust breaks between groups that still have to work together. A conflict spreads beyond the people who were originally involved. Decisions are made that leave people feeling betrayed, frightened, angry, or unheard. People respond to those feelings, other people respond to those responses, and before long it becomes difficult to identify exactly where the original problem ends and everything that happened afterward begins.
Eventually, someone says the organization needs to heal. Usually what follows is language about rebuilding trust, moving forward, repairing relationships, or coming back together. Those are reasonable goals. They are also not particularly useful instructions.
A workplace can become civil again without repairing anything. People can attend the same meetings, answer one another's emails, complete shared work, and behave impeccably while carefully avoiding dependence, vulnerability, disagreement, or anything else that requires trust. Professionalism can make a relationally fractured system look surprisingly functional. If trust and collaboration have actually been damaged, healing has to mean something more than getting everyone to behave nicely again. So what would it mean to take the word healing seriously? What if organizations sometimes need something that looks an awful lot like therapy?
What Happens When the Caregiving System Is Hurt?
The Attachment, Regulation, and Competency framework, or ARC, was developed by Margaret Blaustein and Kristine Kinniburgh as a flexible framework for supporting children and adolescents who have experienced complex trauma. ARC attends to developmental capacities affected by trauma and to the caregiving systems that support those capacities. Blaustein and Kinniburgh describe ARC as a flexible framework, built on principles that translate across settings rather than a fixed protocol, and they extend its principles across caregivers, providers, groups, and systems of care.
Following that developmental logic beyond childhood raises questions about how ARC might function in other caregiving relationships. A supervisor can perform many of the same caregiving functions ARC asks of other caregiving systems. Managing their own affect well enough to remain useful, attuning to another person's needs, and responding effectively are as much a part of the role as supporting regulation and strengthening competency, all within a relationship that allows development to continue. But that raises another question: what happens when the caregiving system is the one that gets hurt?
Organizations are made of people in relationship with one another. When something significantly disrupts those relationships, the people responsible for providing stability and care do not magically remain outside the disruption. Supervisors may be hurt while their staff need support from them. Directors may be trying to hold programs together while needing care from the organizational layers above them. Leaders may carry more responsibility for repairing the system while simultaneously having their own legitimate experiences of what happened.
ARC is recursive. Caregiving does not move in only one direction, and no person permanently occupies the role of caregiver. Who can provide care depends on the relationship, the cohort, the moment, and the capacity available. That becomes particularly important when an organization is trying to repair.
People Will Heal Anyway
After relational rupture, people do not wait for an official healing initiative before they begin making sense of what happened. They talk, comparing experiences, getting angry, making jokes, asking who knew what and when. They decide whose interpretation makes sense to them and start seeking accountability, growing more careful around particular people while sharing less information in some places and more in others. New alliances form, and old ones get reconsidered. In other words, the system is already adapting.
That means the choice is not really between healing and not healing. People naturally try to recover from things that hurt them. The question is what shape that healing takes.
People can heal in relationship with one another, with opportunities to understand what happened and repair what can be repaired. Or they can heal around one another, developing whatever protections allow them to continue functioning. Enough of those individual adaptations can eventually become an organizational culture built around avoidance and guardedness, factions and workarounds, resentment or exquisitely polite mistrust.
There is also a timing problem here. Repair cannot necessarily begin while people are still trying to survive an immediate threat. Sometimes people need information, predictability, practical support, or simple stabilization before they can turn toward the relational injury itself.
But waiting until everyone is calm is waiting too long. Anger is part of the rupture. Grief is part of the rupture. Defensiveness, mistrust, and competing understandings of what happened may all be part of the rupture. Waiting for those things to disappear before beginning repair means waiting for the injury to heal into whatever shape it happens to take and then potentially asking people to tear it open again.
The work begins when there is enough stability to attend to the injury while it is still an injury. Enough stability is different from an absence of strong feeling. Regulation does not require everyone to become calm before difficult material can be approached. It requires enough capacity to move through activation without being completely organized by it: to become angry and remain present, to step away when needed and come back, to move between hurt and curiosity without requiring either state to disappear. ARC describes trauma experience integration as fluid and state-dependent, with people moving between survival and distress tolerance, curiosity and reflection, and engaged action, often moving back again rather than progressing in a straight line.
At a group level, that capacity will fluctuate too. Modulation grows alongside repair, building as the process unfolds rather than needing to be mastered before repair can begin. Attunement matters here. Different parts of an organization may reach that point at different times. One team may already be talking, questioning, and trying to hold one another accountable while another remains focused on immediate safety and uncertainty. A recursive model does not require an entire organization to move through the same process at the same time.
ARC's own emphasis on engagement also matters. Blaustein and Kinniburgh describe engagement as investment, participation, collaboration, and felt stake in the process, well beyond simple attendance. People need to experience the goal as relevant and the process as valid and workable for them. Hesitation, withdrawal, or refusal may say as much about the fit or safety of the process as it does about anyone's interest in repair. Engagement is information, not something that can be assigned.
Someone Has to Be Able to Hold It
ARC begins its caregiving system work with caregiver affect management for good reason. People cannot reliably provide regulation and care to others when they have no capacity to manage what is happening inside themselves. Organizational repair therefore requires people who have enough capacity to help hold what emerges.
Capacity alone is not enough. This is skilled relational work. Facilitation requires substantial practice holding strong affect and tracking multiple people and relationships at once, recognizing shifts in regulation and power, and tolerating competing accounts without prematurely resolving them or intervening in a way that takes ownership of the group's process. Those capacities are central to therapeutic practice, although they can also be developed through sustained practice in other intensive relational settings. Familiarity with trauma-informed language or facilitation techniques is not the same thing as having practiced holding a room when the relationships inside it are actually under strain.
That does not necessarily mean bringing in someone from completely outside the system. An outsider may bring useful distance, but an insider may bring something equally valuable: a working knowledge of the system's history, language, and relationships, along with the tiny contextual details that can make the difference between hearing someone's words and understanding what they mean. Inside or outside matters far less than whether someone has enough capacity and depth of practice to hold this particular relational complexity. Even then, no facilitator should have to become an unfeeling vessel for everyone else's experience.
A care-team model may make more sense. Facilitation can be a function held by several people who have enough skill and capacity to move between supporting the process and participating in it. When someone who has been helping to hold the group reaches a part of the story that belongs to them, somebody else can hold the process while they become a participant.
That shift may itself be useful modeling. A person can be competent, responsible for others, and still have been hurt. They can also be hurt and still have hurt someone else. They can receive care in one moment and provide it in another. The organization is responsible for making that possible.
Telling a supervisor, leader, or facilitator to privately process their organizational hurt elsewhere so they can return ready to absorb everyone else's feelings is not caregiver affect management. It is suppression dressed up as professionalism. Individual therapy or consultation may help someone manage the experience of providing care, but it cannot substitute for having their own organizational injuries recognized within the organizational system where they occurred. The people providing care also need somewhere in the system where they can receive care.
Imagine the Hurt as a Beast
It may help to imagine every hurt that has become part of an organizational rupture as a physical object. At an ARC level, this is identification made collective: the system needs enough shared language for people to name what hurt them, connect it to what happened, and recognize it when it shows up in themselves and others. Some are enormous. Some are small. Some are old objects that became attached to the current conflict. Some came directly from the event everyone currently identifies as the rupture. Others were created by how somebody responded to it, or by how somebody else responded to that response.
Objects like this do not stay separate for long. Hurt changes how people move through a system: someone becomes more guarded or reactive, more avoidant or controlling, more vigilant or protective, and other people respond to those changes. The objects become connected through those interactions, piling up until the pattern starts to take on a shape and behavior of its own. What is left is not a heap anymore but something that moves and reacts like a creature: a beast with a tail stretching much farther back than the event everyone currently identifies as the problem. The dramatic rupture may be one of its sharpest teeth, the bite that finally makes some part of the system unable to keep organizing itself around the beast without acknowledging that it is there. Other people may have been getting bitten for a long time. Repairing only the most visible bite can leave most of the beast sitting in the room.
A shared understanding of what happened has to be large enough to hold people's different experiences of it. That does not require everyone to agree about every fact, intention, interpretation, or decision. It does require people's experiences to become part of the story. No one gets to define another person's hurt.
That does not mean every hurt is equal. Some people may have been harmed far more severely than others. Harms differ in magnitude and consequence, in duration and preventability, and in their relationship to power. Taking someone's hurt seriously does not require pretending a paper cut and an amputation require the same response. It means one person's hurt does not invalidate another's. And neither does someone's behavior.
The legitimacy of someone's hurt does not determine the amount of responsibility they carry, and the amount of responsibility they carry does not determine the legitimacy of their hurt. Someone's hurt can be real even when their understanding of what caused it is incomplete, their interpretation of another person's behavior is wrong, or their own actions caused substantial harm. Making room for multiple experiences does not require treating every explanation of what happened as equally accurate.
"Hurt people hurt people" can become a cliché, but relational rupture gives the phrase some teeth. Someone is hurt and responds defensively. Their response hurts someone else, who withdraws. The withdrawal hurts the first person, who becomes angrier. Someone else intervenes to protect one of them and causes another injury in the process. Leadership misses what is happening, and that failure becomes another hurt. Someone reacts to that failure. Eventually there is no useful division between the people who were hurt and the people who caused harm. There is just a beast.
Understanding why someone caused harm does not excuse the harm. It simply allows two things to remain true at once: I was hurt. I hurt you. Neither statement cancels the other.
The beast is also where trauma experience integration begins to become especially visible. Blaustein and Kinniburgh describe integration as a repeated effort throughout treatment, one that connects current reactions and fragmented states with the experiences and themes that shaped them, and builds broader, more coherent narratives of self, other, and what happened, rather than a discrete final processing phase. At an organizational level, the beast serves a similar conceptual function: previously disconnected experiences begin to form a larger collective account.
Taking the Beast Apart
If the beast represents the interconnected history of hurt, repair requires something different from simply understanding how it got there. The beast is built from the hurts that became connected within the system. Some pieces come apart as people recognize and take responsibility for harms they contributed to. Others require changes to the structures that keep producing them. Still others may represent losses or painful realities that nobody caused and nobody can undo. Those pieces can still be witnessed, held, and incorporated into what the system becomes next.
Taking it apart asks something more relationally complex than simply hearing another person's experience. People have to remain in connection long enough to hold their own experience alongside someone else's, communicate across different understandings, recognize their impact, and take responsibility for the parts that belong to them. At an ARC level, that depends on relational connection as a competency: the capacity to participate effectively in relationship when the interaction becomes difficult, more than simply the presence of relationship itself.
Naming a hurt does not automatically assign blame for it. A person may discover that some part of their own hurt came from a misunderstanding or interpretation that belongs to them. Another person may have made a necessary decision that still caused harm because its rationale was never communicated, important information was withheld, or people were left alone with its impact.
Intent belongs in the story. It just cannot overrule impact. Responsibility also does not distribute evenly. Organizational power creates additional responsibility. Leadership may not have personally caused every injury in a system, but authority carries responsibility for the conditions in which those injuries occurred: what leadership knew, and what it could reasonably have known; what structures existed, and what was tolerated within them; and, once harm became visible, how the system responded.
Some of those conditions may be older than the people currently responsible for them. Organizations inherit histories, relationships, structures, and patterns of harm. A leader can inherit part of the beast without having built it. That changes culpability, but it does not eliminate responsibility. People with enough authority to change harmful conditions become responsible for what they do with them once those conditions are theirs to influence.
That does not make leadership's own hurt less real. Their hurt belongs on the beast too, without receiving preferential treatment because of their position and without being dismissed because of it. Power changes responsibility. It does not grant greater authority over the meaning of other people's experiences, and it does not revoke someone's right to have an experience of their own.
A person can hold hurt in one hand while picking up responsibility with the other. Some pieces, inevitably, will remain: people may decline to participate, or may not yet be able to recognize their contribution. People leave organizations, and sometimes the accountability another person deserves simply will not come.
A model of healing that requires every injury to be resolved would leave all of us trapped by our first unresolved hurt. Community matters here too. A group can witness hurt it cannot repair. It can recognize what remains unresolved, support the person carrying it, and figure out how to move forward without pretending that accountability occurred when it did not. Repair does not require manufacturing closure.
Safety Has to Mean Something
Being heard matters. Accountability matters. Apologies matter. None of them, by themselves, make someone safe. After harm has been understood and responsibility taken where possible, another question emerges: why should anyone expect this to go differently next time? That is a competency question.
Sometimes the answer involves communication: information needs to move differently or sooner, or people need a shared understanding that reduces the likelihood that the same action will be interpreted in the same harmful way. Other times, responsibilities have been allocated according to hierarchy or habit when they should belong to the person with the strongest skill in that area, or decision-making, supervision, authority, or access to information itself needs to change.
Some of that safety also lives in routines and rhythms. Small hurts need somewhere to go: a predictable place to raise a concern, a regular moment to notice strain building, a communication pathway people can actually rely on. Repair cannot require a crisis every time. Part of developing a safer system is building enough relational infrastructure that hurt can be noticed and responded to while it is still one object, rather than waiting until enough objects have accumulated to become a beast.
This is where executive functions, self-development, and identity become organizational questions. The system has to be able to examine what happened, anticipate future problems, make decisions, change course, and develop a more coherent understanding of who it is after the rupture. ARC treats agency, self-understanding, and the capacity to act purposefully on the present and future as central to integration and competency.
And sometimes the harm itself cannot be prevented. People will misunderstand each other again, someone will miss a cue, a decision will hurt, or someone will say the wrong thing on a terrible day. Safety rarely means an organization in which nobody ever produces another hurt object. It means knowing what happens when one appears: who someone can turn to, and how quickly it will be understood, addressed, and repaired. It also means knowing the likely consequences if someone chooses to hold a hurt privately instead. A healthy system does not necessarily prevent every hurt from appearing. It becomes better at noticing one before an entire beast grows around it.
Trust Is Built After the Apology
Organizations often talk about rebuilding trust as though trust itself were an intervention. Trust behaves more like an outcome, something that accumulates once a system starts behaving differently rather than something a workshop or a memo can install. People begin to trust a repaired system by experiencing it behaving differently.
Someone names a small hurt and receives an effective response, or asks instead of assuming. A leader follows through on a change the group agreed was necessary, and information arrives when it was promised. Someone notices a misunderstanding before it becomes a grievance, and a rupture occurs and gets repaired while it is still small.
The safety plan becomes behavior. Those experiences accumulate, giving people evidence that particular forms of dependence on one another are becoming less dangerous than they had been. Trust does not necessarily return everywhere at once. Someone may come to trust a supervisor's transparency while remaining wary of senior leadership, or trust a new process without yet trusting the people responsible for carrying it out. The system becomes more demonstrably trustworthy in the places where its behavior has actually changed.
This, too, is part of integration. The rupture has been getting integrated all along: as people name what happened, connect present reactions to past events, encounter one another's experiences, take responsibility, and build different ways of responding. What changes over time is the degree to which the rupture continues to organize the present. ARC describes the goal of integration in terms of increasing capacity to use internal and external resources to engage purposefully in present life, gradually reducing the degree to which past experience overrides the present. At an organizational level, something similar may happen when the rupture becomes part of the group's collective history without remaining the primary organizer of its current relationships.
This happened to us. We understand more about how we got there. We repaired what we could. We changed because of it. Eventually, people can orient toward something else.
The Cycle Continues
Of course, none of this stays finished. ARC is developmental, recursive, and cyclical: new stressors appear, people and leadership change, resources disappear and new relationships form, and old vulnerabilities get activated under different conditions.
Different parts of the same organization may also be in very different places at the same time. A program may be ready to examine its own relational injuries while the larger division is still stabilizing. A supervisor may need to provide caregiving to their staff while simultaneously needing caregiving from the organizational system above them. Someone may be capable of helping hold one group's process while needing substantially more support in another.
Caregiver is therefore not a permanent identity in an organizational ARC model. It is a function people and systems perform for one another according to relationship, role, capacity, and moment. That has consequences for organizational leadership.
An organization cannot continually demand regulation downstream while withholding the conditions that allow the people providing that regulation to remain regulated themselves. Supervisors cannot indefinitely absorb staff distress without somewhere for their own organizational experience to be held. Leaders cannot be expected to repair programs by foregoing their membership in the very relational system that was injured.
Care has to move through the system if the system is expected to keep providing care. The goal, then, becomes building enough distributed capacity that the organization can keep doing this work, noticing and responding, regulating and connecting, developing new competencies and integrating what happens, cycling through those needs again each time a new stressor arrives. That is consistent with ARC's own refusal of a rigid sequence. The framework treats integration as ongoing and state-dependent, and its foundational strategies, engagement, education, and routines and rhythms, operate across the core targets rather than belonging to a single phase.
What if Organizations Need Therapy?
Organizations routinely describe relational injuries using words that belong to healing: rupture, hurt, mistrust, repair, recovery. Then they often reach for managerial solutions instead: communication plans, new procedures, expectations about professionalism, directives to move forward.
Those things may be useful. They are not necessarily the same thing as healing. ARC offers another way to understand the problem. If an organization is a caregiving system, then there may be moments when that caregiving system itself needs the developmental conditions it is ordinarily responsible for providing to others. People need enough support to stay present with what happened, enough shared language to understand it together, and enough relational capacity to remain connected while responsibility is taken. The system has to learn from what happened and change in ways people can actually experience. Over time, the rupture can become part of the organization's history without continuing to organize its relationships.
Structural change and relational repair depend on each other. Structural change without relational repair leaves people using new procedures inside relationships they still do not trust, adapting to updated rules the way they adapted to the original rupture: carefully, protectively, without depending on anyone more than the new system requires. Relational repair without structural change leaves people fully understood and still exposed to the same conditions that produced the original harm, trusting each other again just in time to be hurt by the same unaddressed structure. Safety requires both: enough executive and identity capacity to change what actually needs to change, and enough relational connection to make people believe in that change rather than merely comply with it.
Work like this needs skilled holding. What matters is whether someone has substantial practice holding this degree of relational complexity, regardless of the label of therapist or non-therapist, insider or outsider. Therapeutic practice is one place those capacities are deliberately developed: attending simultaneously to affect and regulation, tracking multiple relationships and competing realities, noticing power, and staying useful through rupture and repair. Some people develop comparable depth through sustained practice in other intensive relational settings. Either way, the work requires more than good intentions, managerial authority, facilitation experience, or familiarity with the language of trauma and repair.
And none of it can be done to a group. Leadership can resource it. Skilled facilitators can help hold it. People with more power can take greater responsibility for the conditions that made harm possible and for ensuring that necessary changes actually occur. But healing remains participatory. People cannot be required to reconcile, forgive, disclose, or participate in a repair process they do not want. Neither can one person's absence or refusal make everyone else's repair impossible. The people who choose to participate can build something different together while making room for unresolved hurt, absent voices, and accountability that may never come.
Maybe that is what it means for an organization to heal: to develop enough capacity to hold what happened, change what needs changing, and build relationships that are increasingly able to survive the next hurt without having to build another beast around it.
A note of thanks: This article builds on the work of Margaret Blaustein and Kristine Kinniburgh, whose ARC framework has shaped how I understand trauma, development, and the conditions that support growth. It extends that developmental lens beyond the contexts in which ARC was published while remaining grounded in the framework they created.
Blaustein, M. E., & Kinniburgh, K. M. (2019). Treating traumatic stress in children and adolescents: How to foster resilience through attachment, self-regulation, and competency (2nd ed.). The Guilford Press.
