There is a diagram you’ve probably seen. A circle. A child moving out to explore. A child returning for comfort.
At the top: “Watch over me, delight in me.” At the bottom: “Protect me, comfort me.”
It’s elegant. It’s intuitive. It feels true.
This is the Circle of Security (COS) - an attachment-based parenting program designed to help caregivers better understand and respond to their child’s emotional needs.
And in many ways, it offers something deeply needed - A reminder that children don’t just need behaviour managed. They need relationship.
But like all frameworks that become institutionalised, scaled, and professionalised…the circle carries more than it first appears to.
What Circle of Security is trying to do: At its core, COS is built on attachment theory. It teaches that children need two things-
• a secure base to explore the world
• a safe haven to return to when distressed
The program is typically delivered as an 8-session intervention, using videos, guided reflection, and facilitated discussion to help parents-
• interpret their child’s emotional signals
• reflect on their own responses
• build more attuned relationships
It positions itself as-
• relationship-focused
• reflective
• grounded in emotional connection
And importantly, it does not present itself as behaviour modification.
COS feels like a shift toward something more human compared to compliance-driven models.
Where COS aligns with neuro-affirming values and naming what COS gets right:
• It reframes behaviour as meaning.
Rather than asking “How do we stop this?”, COS asks “What is the child needing right now?” That is a profound shift away from behaviourist logic.
• It centres co-regulation.
COS recognises that children regulate through relationship, not isolation. This aligns strongly with neuro-affirming understandings of sensory overwhelm, emotional dysregulation, and relational safety.
• It invites caregiver reflection.
Parents are not positioned as enforcers, but as participants in the relationship. It asks -
What is coming up for you? What blocks connection? What patterns are repeating? This is rare in many intervention models.
• It challenges punitive discipline.
COS implicitly pushes against time-out as default, punishment as primary tool, and behaviour suppression.
And these are important to the nervous systems of everyone in the circle.
But here’s where the circle tightens: Because even frameworks built on connection can carry quiet assumptions. And those assumptions shape practice.
1. The weight of attachment theory itself: COS is not neutral. It is grounded in decades of attachment research, which has historically -
• centred normative developmental trajectories
• prioritised caregiver-child dyads (often mother-child)
• been applied unevenly across cultures and communities
Attachment theory tells a compelling story -secure attachment = better outcomes across life, Including - emotional regulation, relationships, reduced risk of later difficulties. But this narrative can become totalising. It risks implying -If a child is struggling then attachment is the issue. If outcomes are poor, then caregiving is the cause. And suddenly, structural factors fade into the background.
2. The quiet shift toward caregiver responsibility: COS focuses intensely on caregiver sensitivity and responsiveness.
It teaches that change happens primarily through -the caregiver’s insight, the caregiver’s emotional availability, and the caregiver’s behaviour. And while this can be empowering…It can also become heavy. Because it subtly relocates responsibility from systems to families.
It does not always ask -What is the environment doing to this child? What pressures is this family under? What systemic barriers are shaping this relationship?
Instead, the intervention lens often zooms in closer...closer...closer...Until everything becomes about the dyad.
3. The evidence gap no one talks about loudly enough: COS is widely described as “evidence-based”. But the actual research landscape is… thinner than the language suggests.
Studies show - some improvements in caregiver attachment patterns, reductions in parenting stress in certain groups. But… effects are often small or inconsistent, some RCTs find limited differences compared to controls, overall evidence is limited, with small sample sizes and mixed findings. Even government summaries note - “requires further research”.
So we have a program with -high uptake, strong branding, intuitive appeal…but still developing empirical support. That gap is critical to individuals, especially when programs are scaled across systems.
4. When “understanding behaviour” still centres normalisation: COS reframes behaviour as communication. But it still operates within a developmental framework that assumes -certain emotional responses are expected, certain relational patterns are optimal. This raises a neuro-affirming tension. What happens when autistic ways of relating don’t fit the model?
For example - less eye contact, different proximity needs, non-typical expressions of attachment, parallel play over interactive play. Within COS, these can be interpreted as - signals to decode, patterns to respond to. But sometimes they are not signals. They are simply different ways of being.
5. The risk of subtle pathologisation. COS is often delivered in contexts like - child protection, early intervention, clinical services, mandated parenting programs. Which means it frequently intersects with families already labelled as: “at risk”, “struggling”, “disorganised”. In these contexts, attachment language can become - diagnostic, evaluative, tied to intervention or funding decisions. And this is where a relational model can quietly shift into a surveillance tool
Not by design, but by system. The distinction is paper thin.
6. The professionalisation of “being with”: One of COS’s central ideas is “being with” a child in their emotional experience. It sounds simple. But the program packages this into -training• manuals, certification, delivery models.
And here is the tension -
Something fundamentally human, being present with a child, becomes - structured, standardised, billable.
This is not unique to COS. But it is part of a broader pattern, where relational knowledge is captured by systems and redistributed as expertise.
So where does this leave Circle of Security in our bodies, minds, and nervous systems?
Not in the bin. But not on a pedestal either.
COS represents a shift away from behaviourism, but not a full departure from developmental normativity or system logic.
It sits in an in-between space -
• more relational than compliance models
• more reflective than prescriptive programs
• but still shaped by institutional frameworks
A neuro-affirming reframe: A neuro-affirming approach would take the useful parts of COS - behaviour as communication, co-regulation, relational safety, caregiver reflection, and extend them further:
• honour neurodivergent ways of relating
• centre autonomy, not just attachment
• widen the lens beyond the caregiver-child dyad
• explicitly address systems, not just relationships
Because children do not exist in circles alone.
They exist in - classrooms, sensory environments, policy systems, communities, cultures.
And those matter just as much as attachment. A child’s ecosystem is expansive, complex, nuanced, and must be respected.
We are not interested in rejecting relational frameworks.
We are interested in asking -
• Who do they centre?
• What do they assume?
• Where do they stop?
Because a model can be - gentle, reflective, relational…and still -
• reproduce narrow developmental norms
• place disproportionate responsibility on families
• overlook structural harm
The Circle of Security gives us a beautiful image - A child moving outward, and coming back in.
But the real question is not just “Can the child return?”
It is “What kind of world are they returning to?”
Because safety is not just built in relationships.
It is built in systems.
And if those systems remain unchanged, no circle, no matter how well drawn, can hold everything a child needs.
