Brian Kennedy, LMFT
The partner who appears most distressed is not necessarily the only partner struggling to regulate. The partner who speaks quietly is not necessarily available for contact. A request for reassurance may be an adaptive attachment bid, a narrowed attempt to obtain relief, or a response to actual unreliability. Withdrawal may be restorative, defensive, or necessary for safety. Clinical usefulness depends on distinguishing these possibilities rather than sorting partners into the emotional one and the rational one.
This companion develops an integrative formulation of attachment, regulation, differentiation, and parts work. Its central proposal is to attend to safety, regulatory capacity, and reciprocal responsiveness before asking a couple to sustain complex exploration of difference. That proposal is a clinical heuristic, not an empirically established ordering of complete treatment models. Sometimes a differentiating act, such as naming a limit or acknowledging a real incompatibility, is the intervention that first permits regulation.
Although the clinical illustrations here focus on couples, the Attachment Patterns pathway can also be used by an individual reflecting on a friendship, family relationship, or partnership. Another person need not participate for the learner to examine a prediction, make a request, reconsider an explanation, or recognize that dependable support is missing. In individual therapy, the task includes helping the client distinguish what they can choose from what the relationship actually offers. We should not make one person’s increased flexibility responsible for repairing another person’s unreliability.
My client companion, When Distance Feels Like Danger, and the five modules translate these questions into ordinary encounters. This clinical companion helps therapists consider what clients may discover, what might be difficult, and how the learning can be brought into the room without becoming another demand to perform attachment security.
Attachment is a relational formulation rather than a personality verdict
Bowlby's secure-base formulation, Ainsworth's observations of infant-caregiver behavior, and Main, Kaplan, and Cassidy's work on attachment representation address related but distinct levels of organization. These traditions should inform assessment without being collapsed into a single casual style label (Bowlby, 1988; Ainsworth et al., 1978; Main et al., 1985).
In particular, infant Strange Situation classifications, Adult Attachment Interview states of mind, and adult romantic self-report constructs are not interchangeable. Main's use of secure-autonomous does not mean that an adult requires no attachment support. An AAI classification requires the appropriate interview, coding method, and expertise; it cannot be inferred from a person's composure in couple therapy. Research comparing interview and questionnaire approaches underscores the importance of distinguishing assessment methods (de Haas et al., 1994).
Anxious and avoidant tendencies can be useful descriptive language when tied to specific contexts. “What happens when your partner is unavailable?” is more clinically informative than “Which style are you?” Ask about exceptions, transitions, partner behavior, and the capacity to recover. Do not infer disorganized attachment from an adult's alternating pursuit and withdrawal alone.
The BKL Attachment Patterns Inventory belongs in this discussion as an educational reflection tool. It does not establish an AAI classification, diagnose attachment pathology, or verify a developmental explanation. The therapist should investigate current conditions independently of the learner's pathway result.
When a prediction changes, the feeling may still need company. Learning that a friend never received the message can revise an account of rejection without undoing the loneliness of the evening spent waiting. Before moving toward a more accurate thought, ask what the person hoped would happen and what remains unmet. A correction should help the client understand the encounter more fully; it should not become an instruction to stop feeling disappointed.
Separate activation from dysregulation and expression from capacity
Activation is not itself a treatment failure. A person may experience intense grief, anger, or fear while remaining oriented, able to hear another perspective, and able to inhibit harmful action. Conversely, low outward expressiveness may coexist with cognitive narrowing, suppression, detachment, or a strong escape impulse.
Dozier and Kobak monitored skin conductance during attachment interviews with 50 college students. Participants using deactivating strategies showed increased conductance in response to questions about threatening attachment experiences (1992). Sroufe and Waters also reported discrepancies between avoidant infants' behavior and heart-rate responses (1977). These observations challenge a simple equation between minimal expression and minimal activation; they do not establish the physiology of a particular adult sitting in the consulting room.
Covert dysregulation is therefore best used as a tentative clinical description supported by inquiry, not a hidden condition the therapist claims privileged access to. “You seem quieter, and I wonder whether you are settling, holding things in, or having trouble staying with us” allows correction. “You are actually dysregulated” can replicate the experience of being defined by someone else.
Observe functional capacities: can the client follow the exchange, describe a present sensation, distinguish an assumption from a fact, make a request, receive a response, and tolerate a limited difference? Ask whether apparent agreement is genuine or a way to end pressure. Account for language, culture, disability, medication, neurodivergence, fatigue, and preferred communication style.
The shared concern in overt and covert dysregulation is diminished flexibility or availability. The states need not have the same mechanism, severity, or consequences. Clinical fairness is not behavioral equivalence. Escalation that includes intimidation demands a different intervention from quiet overload, even if both partners report fear.
Use neurobiology to deepen observation rather than replace it
Schore's developmental account emphasizes affective communication within attachment relationships and the contribution of right-hemisphere processes to regulation. It helps direct clinical attention toward timing, prosody, expression, and embodied responsiveness (Schore, 2000). It should not be translated into a claim that one partner's right brain has taken over, or that a particular expression establishes a vagal state.
Damasio and Carvalho describe feelings in relation to bodily and life-regulatory processes (2013). An application here is to ask about the lived signal accompanying the argument: heat, constriction, nausea, trembling, numbness, or an urge to move. These descriptions are clinical data, not direct measurements of neurophysiology.
Siegel's interpersonal neurobiology framework places integration in the relationship between differentiated elements (2020). His account explicitly includes respect for individual autonomy alongside empathic connection. As a conceptual bridge, this helps avoid treating linkage and differentiation as rivals. It does not demonstrate that the treatment sequence proposed here has been experimentally validated.
These frameworks are complementary contributions, not one unified neurological explanation of couple conflict. Keep the observation, the client's report, the relational hypothesis, and the biological theory distinguishable in formulation.
Clarify what kind of regulation is occurring
In Tatkin's terminology, autoregulation is internally directed and does not require a partner; self regulation supports maintaining social engagement; interactive regulation occurs between people (Tatkin, 2013). This usage is narrower than the broader use of self regulation elsewhere in psychology. Explain the vocabulary before using it evaluatively.
Autonomous regulation is used here descriptively for managing without another person's immediate participation. It is not autonomic regulation, and it is not the secure-autonomous AAI category. Solitude may be healthy, necessary, and effective. Its clinical significance depends partly on flexibility and the ability to return to relational engagement.
Tatkin's earlier psychobiological paper distinguishes interactive regulation from both distancing strategies and nonreciprocal external regulation (2009). The distinction can guide a functional question: is the interaction expanding room for both people, or organizing another person primarily around one person's relief?
For this formulation, track three outcomes separately. First, did subjective distress become more tolerable? Second, did capacity for attention and choice improve? Third, did the exchange preserve or restore recognition of both people? A partner can feel relieved after the other capitulates while the relationship becomes less secure. A partner can remain upset after a clear boundary while the exchange becomes more honest and sustainable.
Co regulation does not require synchronized feelings, agreement, or equal caregiving in every exchange. It can be asymmetrical in a difficult moment and still be reciprocal across time. Nor should physiological synchrony, if measured, be equated automatically with therapeutic benefit: people can become more activated together.
Anxious bids and one-sided attempts at external regulation
An anxious client's urgency may make immediate reassurance a reasonable intervention. Before construing the request as excessive, assess whether the partner has been inconsistent, deceptive, contemptuous, or unavailable. Attachment anxiety can reflect both sensitized expectations and a credible current problem.
The concern is not dependence itself but narrowing mutual recognition. A request becomes clinically problematic when the other person's answer, limits, fatigue, or competing needs are persistently excluded; reassurance has to be repeated without being metabolized; or the only acceptable response is compliance with a prescribed solution.
A useful intervention joins recognition with a specific limit: “The fear is real, and we can help you ask for contact. We also need a request your partner can answer, rather than a test they cannot finish.” Invite a bounded bid, a realistic offer, and a check of impact. Avoid treating a client's imperfect capacity for reciprocity during acute distress as evidence of bad character.
Tatkin's secure-functioning emphasis contributes the principle of attending to self and partner together (2018). The clinical elaboration proposed here is to assess whether reciprocity remains available or becomes available again with support. Mutuality can include one partner receiving care first while the other's limit remains respected.
Apply the same scrutiny to the withdrawing partner. “I regulate myself” can describe useful skill or rationalize unilateral exit. Explore whether the person communicates a pause, returns as agreed, addresses impact, and can receive support. Neither attachment anxiety nor avoidance grants an exemption from responsibility.
Making room for the experience of receiving
A request can be clear and the response caring, yet receiving may remain difficult. The person may feel grateful and still want more, or become sad when an ordinary kindness makes an earlier absence more noticeable. Stay interested in what the offer actually evokes. The client does not need to produce relief to demonstrate that they received something, and discomfort does not establish that they rejected it. An honest mixed response gives the therapist more usable information than an appreciative performance.
Fosha’s work on meta-therapeutic processes invites us to explore the experience of having been helped, rather than move immediately to the next problem (2000). In AEDP, this is called metaprocessing: attending together to what a useful encounter feels like and means to the person receiving it. We might ask, “What was it like to say what you needed and find me still interested?” Let the answer guide the next moment. A welcome experience can bring tenderness or mourning as well as relief; the client’s report matters more than the affect we hoped to see.
Differentiation and attachment are distinct clinical emphases
Bowenian differentiation concerns maintaining a more thoughtful, self-defined position within an emotional relationship system, including under pressure to conform or react. Its description of realistic dependence makes clear that differentiation is not synonymous with independence (Bowen Center, n.d.).
Attachment-focused work emphasizes availability, security, access to comfort, and the expectations carried into dependence. Differentiation-focused work emphasizes a person's ability to maintain convictions and responsible action amid relational anxiety and difference. Neither description captures every intervention in either tradition.
The tensions are clinically useful. Attachment language can sometimes make a partner responsible for eliminating another's anxiety; differentiation language can sometimes dignify emotional cutoff as maturity. A therapist needs to notice when the model becomes a justification for a client's familiar strategy.
A client who says, “I cannot hear your no without feeling abandoned,” may need both an experience of continued regard and practice tolerating a limit. A client who says, “Your feelings are yours,” may need both permission to retain a position and recognition that their conduct affects someone they love.
Do not promise that sufficient attachment security will resolve incompatible values, or that differentiation will make responsiveness unnecessary. A more regulated couple may discover a genuine disagreement that cannot be negotiated away.
What is generally called for first
When partners cannot hear one another, remain oriented, or refrain from escalation, begin by changing the conditions of the exchange. This may involve interruption, reduced stimulation, more space, therapist-led pacing, or a clearly bounded pause. The aim is enough capacity for meaningful participation, not emotional flatness.
Next, if safe and welcome, facilitate a small experience of responsiveness. Help one partner make an understandable bid and the other provide an honest response. For example: “I am frightened that a break means you will not return,” and “I will return at eight; I am not able to continue productively right now.” Verify that the offer is feasible and later review follow-through.
Then explore the difference the alarm has obscured. What does each person actually want? Which need can be supported, which demand cannot be met, and which responsibility belongs to whom? This is where differentiation becomes more available without requiring that all attachment distress has disappeared.
The sequence is recursive. A clear differentiating statement can lower threat. A repair can restore curiosity. A new piece of information can reactivate alarm. Move according to demonstrated capacity and risk rather than a predetermined stage.
There are important exceptions. In coercive control, fear of retaliation, violence, or ongoing intimidation, protection and confidential assessment precede conjoint vulnerability work. Where a partner has caused a concrete injury, acknowledgment and changed conduct may matter before exercises in mutual soothing. A pause must not become a routine way to evade accountability.
The proposed order is consequently safety and usable capacity first, followed by responsive contact and exploration as appropriate. It is not attachment therapy first and differentiation therapy later.
Parts work can make attachment strategies more negotiable
An integrative parts formulation asks which experiences and action tendencies become dominant under attachment stress. A Manager may scan for rejection; another may organize self-sufficiency. A Firefighter may urgently protest or disconnect. An Exile may carry vulnerability. These are IFS roles, not attachment classifications, and a person's experience should take precedence over the map (Schwartz & Sweezy, 2019).
The following integration is a clinical proposal. Invite the client to notice a reaction without forcing an identity: “A part of me wants to make you answer right now.” Ask what that response anticipates and what it needs, while leaving its origin open. Support access to the curiosity and perspective called Self in IFS without demanding a performance of calm.
Then translate the internal understanding into an accountable relational act. “I can understand the urge to ask repeatedly, and I will make one clear request.” “I notice the urge to disappear, and I will explain my limit.” Review the partner's actual response rather than assuming internal insight has repaired the relationship.
Parts work may be particularly useful when a client shifts between contradictory positions. One part wants contact; another distrusts comfort. One part promises closeness; another experiences it as engulfment. Making room for both may reduce shame and rigid identification. It does not establish a trauma history or justify retrieving presumed memories.
Do not turn parts into additional combatants: “Your Firefighter attacked my Exile” can become a sophisticated accusation. Nor should an injured partner be asked to soothe the other's vulnerable part before their own injury has been addressed. Internal compassion and interpersonal responsibility need to remain connected.
A brief clinical illustration
In a fictional session, Rowan repeatedly asks whether Alex intends to leave the relationship. Alex responds in a controlled voice and then becomes increasingly monosyllabic. The therapist initially feels drawn to help Rowan calm down and recruit Alex as the stable partner.
Inquiry changes the picture. Alex reports a tight jaw, shallow breathing, and difficulty taking in speech. Rowan can describe the fear but is treating every limit as further evidence of abandonment. The therapist slows the exchange and invites each person to name one present experience, without asking either to interpret the other.
Rowan identifies a part that believes unanswered questions mean impending loss. Alex identifies a part that expects emotional contact to become an endless obligation. These hypotheses help organize the present experience; neither is treated as proven childhood causation.
The therapist helps Rowan request a short, specific reassurance and helps Alex make a truthful offer. Alex can say they want to continue the relationship and need a pause. Rowan practices acknowledging the offered reassurance without immediately expanding the test. They agree to revisit the substantive disagreement later.
In a later exchange, the therapist explores whether Alex actually returns and whether Rowan allows a return without an interrogation. If not, the work is not declared successful because both seemed calmer in session. Greater security requires dependable behavior outside the facilitated moment.
The therapist participates in the attachment field
The therapist’s pull in the fictional example deserves its own attention. We may feel an urgency to reassure the visibly frightened person or relief beside the partner who appears composed. We may identify with the person who feels crowded and hurry to defend their space, or with the person waiting for a response and press for more closeness. These reactions can provide questions for inquiry, but they do not establish either partner’s experience. Can we notice our own investment before it becomes the direction everyone else must follow?
Aponte and Kissil’s Person-of-the-Therapist approach emphasizes consciously and purposefully using the self we bring to the present encounter (2014). Here, that includes recognizing the shame of not helping quickly, the discomfort of another person’s disappointment, or the wish to secure agreement. Personal therapy and supervision can help us examine these responses. We can then ask more accurately whether the client needs comfort, room, practical accountability, or a correction to our understanding.
The relationship with the therapist also supplies present information. A client may tell us that a question felt intrusive, that our reassurance missed the concern, or that a pause felt like withdrawal. Before interpreting that response as an attachment pattern, consider what we actually did. A specific acknowledgment and an adjustment may be needed. My article When Another Person’s Feelings Stir Our Own develops this inquiry into the therapist’s participation and capacity for repair.
Evaluate flexibility and follow-through rather than calm alone
Useful indicators include more direct bids, less threatening pursuit, clearer pauses, reliable returns, increased ability to receive care, and better tolerance of difference. Ask whether each person can revise a mistaken interpretation and recognize the impact of their conduct.
A reduction in conflict may also reflect resignation, concealment, or fear. Follow up privately when clinically appropriate and do not assume apparent harmony establishes safety. Similarly, continued emotion does not mean that the intervention failed if choice, honesty, and responsiveness improved.
The clinical task is to help partners become more available to themselves and each other while retaining the capacity to say no, take responsibility, and make difficult decisions. Attachment, differentiation, and parts work can contribute different forms of leverage. Their integration is strongest when the therapist remains curious about the function and effect of an intervention rather than loyal to its vocabulary.
Following the five modules in clinical work
Each module uses a complete learning cycle: one manageable encounter, a possible explanation, the learner’s correction, a chosen experiment, the actual result, and a revised account. These are not five stages distributed across five lessons. When a client brings work into therapy, begin with what happened rather than how much they completed. Private rehearsal, fiction, a skipped practice, and an unchanged or difficult result can each provide something useful to discuss.
Module 1: Separate the cue from the conclusion. Help the client distinguish an observed event from its predicted meaning while keeping the wish underneath in view. Ask what information supported or changed the account, what remains unknown, and whether the disappointment still needs attention. A more accurate interpretation should leave the client more understood, not chastened for having felt something.
Module 2: Find room to choose, with support if needed. Review whether an anchor, pause, or welcome company made attention and choice more available. Did it help the client voice a need, or did it become another way to silence that need? Practical action or protection may come before a regulation exercise. Solitary calm is not a condition for deserving support.
Module 3: Ask and receive in tolerable doses. Explore the actual request, the other person’s reply, and the client’s experience of receiving or declining. A five-minute offer may be valuable and insufficient. Ask what was welcome, what remained unmet, and whether both people retained a genuine choice. The exercise should not turn disappointment into an attachment failure.
Module 4: Stay connected through a rupture. Identify what the rupture affected and what repair would need to address. Review acknowledgment, feasible return, changed conduct, and the affected person’s correction of the account. A kept return may matter while a neglected task or a painful remark remains unfinished. Repair need not produce immediate forgiveness, closeness, or equal responsibility.
Module 5: Build a secure-base routine. Help the client retain one useful capacity, the conditions that supported it, an unresolved need, and a realistic route to further help. Review asking, receiving, separateness, and repair in the relationship as it actually is. Neither a workable plan nor a different inventory result certifies a lasting change of attachment style.
A more dependable bond is valuable partly because it can leave room for a life beyond watching the bond. The final module includes friendships, practical support, interests and meaningful participation alongside requests and repair. Ask what the client has more room to do, what help makes that possible, and what remains unavailable. A useful plan can preserve reliance and separateness without making one relationship the sole source of steadiness or treating independence as the endpoint.
References and clinical limits
The following sources support different levels of the argument: developmental attachment research, empirical work on expression and physiological response, theoretical neuroscience syntheses, and clinical models. They do not jointly validate this integrative sequence, nor establish that parts work improves attachment outcomes in the particular way proposed here.
Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. Routledge. Foundational secure-base theory. Publisher information.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Lawrence Erlbaum Associates. Observational infant attachment research; not an adult couple diagnostic system. Publisher information.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. Monographs of the Society for Research in Child Development, 50(1–2), 66–104. The representational contribution to attachment assessment. Original publication.
de Haas, M. A., Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (1994). The Adult Attachment Interview and questionnaires for attachment style, temperament, and memories of parental behavior. The Journal of Genetic Psychology, 155(4), 471–486. Relevant to distinguishing interview and questionnaire constructs. Study information.
Sroufe, L. A., & Waters, E. (1977). Attachment as an organizational construct. Child Development, 48(4), 1184–1199. Foundational work relating attachment organization and behavioral-physiological observations. Original publication.
Dozier, M., & Kobak, R. R. (1992). Psychophysiology in attachment interviews: Converging evidence for deactivating strategies. Child Development, 63(6), 1473–1480. Empirical evidence of increased physiological response during selected interview questions; not a test for covert dysregulation in couples. Study abstract.
Tatkin, S. (2009). A psychobiological approach to couple therapy. Author's prepublication manuscript, dated May 5, 2009. Provides the clinical distinction between interactive and nonreciprocal external regulation; some biological and diagnostic language in the early paper should not be adopted uncritically. Author-hosted manuscript.
Tatkin, S. (2013, September 16). Arousal regulation and mindfulness for couples. PACT Institute. Primary source for the model-specific regulatory terminology. Read the article.
Tatkin, S. (2018, June 17). Secure-functioning essentials: Taking care of yourself and your partner at the same time. PACT Institute. Clinical emphasis on simultaneous self and other care and the need for triage. Read the article.
Tatkin, S. (2024). Wired for love (2nd ed.). New Harbinger. Broader clinical framework for secure functioning in adult couples. Author's book information.
Schore, A. N. (2000). Attachment and the regulation of the right brain. Attachment & Human Development, 2(1), 23–47. A developmental theoretical synthesis, distinguished here from direct physiological observation. Article information.
Siegel, D. J. (2020). The developing mind: How relationships and the brain interact to shape who we are (3rd ed.). Guilford Press. Interpersonal neurobiology and integration. Publisher information. See also his overview of interpersonal neurobiology.
Damasio, A., & Carvalho, G. B. (2013). The nature of feelings: Evolutionary and neurobiological origins. Nature Reviews Neuroscience, 14, 143–152. Biological account of feelings; not a couple treatment trial. Article information.
The Bowen Center for the Study of the Family. (n.d.). Differentiation of self. Defines the Bowenian concept used in this comparison. Read the overview.
Schwartz, R. C., & Sweezy, M. (2019). Internal family systems therapy (2nd ed.). Guilford Press. Parts and polarization framework; the attachment integration in this article is an explicit clinical synthesis. Publisher information.
Fosha, D. (2000). Meta-therapeutic processes and the affects of transformation: Affirmation and the healing affects. Journal of Psychotherapy Integration, 10, 71–97. Clinical account of exploring the experience of having been helped, including receiving affirmation and mourning. Author manuscript. Publication record.
Aponte, H. J., & Kissil, K. (2014). “If I can grapple with this I can truly be of use in the therapy room”: Using the therapist’s own emotional struggles to facilitate effective therapy. Journal of Marital and Family Therapy, 40(2), 152–164. Person-of-the-Therapist model and purposeful use of one’s present self. Publication record. Author manuscript.