Inside Entanglement: Mapping the Architecture Beneath the Armor
Abstract
Our most recent study established entanglement and engagement as reliable, replicable dimensions of emotional well-being. The next analysis of that same dataset of 224 adults with item-level dimension tagging asked what each dimension is made of. We found that Protective Armor and Aggression belong predominantly to entanglement, while Forgiveness, Unique Identity, and Personal Power belong predominantly to engagement. All five sub-patterns held together as internally reliable scales. Sub-patterns within the same dimension correlated with one another more strongly (.59 to .68) than sub-patterns across dimensions did (.42 to .50), and that cross-dimension range closely matches the correlation already established between entanglement and engagement themselves (r = .57) as shared in our last article, replicating the same relationship at a finer grain. Within entanglement, Protective Armor organized into three strategies resembling the fight, flight, and fawn threat responses, with anger and aggression concentrated specifically in the fight-oriented strategy. A person's composed self-presentation also proved distinct from whether that composure held under heightened emotional stress, tracking with the entanglement-side Aggression measure rather than the engagement-side measures. Together, these findings describe entanglement and engagement not as single undifferentiated stances but as two structured architectures, each built from measurable, non-redundant parts.
Keywords
Emotional safety; entanglement; engagement; emotional intelligence; emotional well-being; forgiveness; stress; identity
Revised August 2, 2026
Table of Contents Show
Introduction
At the Center for Emotional Well-being, we continue to collect data using the most up-to-date version of the Emotional Intelligence 3.0® assessment to inform model development. Our objective is a scalable model of emotional well-being. In furtherance of that objective, this article is the second in a series that analyzes a 224-person dataset collected from a non-profit organization.
This article summarizes evidence that entanglement and engagement, the two dimensions identified in our previous article, are not single undifferentiated patterns. Each is made up of distinguishable sub-patterns, measurable in their own right, that together form a coherent architecture on both sides of the emotional well-being model we are developing.
In this framework, “entanglement” refers to everyday protective patterns in emotional and relational life, not to clinical trauma responses, even though it draws on the same nervous-system dynamics that trauma research has helped illuminate.
Our first analysis established that entanglement and engagement are real, measurable, moderately related dimensions (r = .57). This analysis of the data asks the next question: what is each dimension made of?
Using data from 224 adults in the same organization and the item-level dimension tags applied directly to the assessment, we tested which sub-patterns belong to which dimension and whether they hold together as coherent, non-redundant measures.
What We Found
Each item in the assessment carries a direct tag identifying it as an entanglement item or an engagement item. When we sorted our five candidate sub-patterns, Forgiveness, Unique Identity, Personal Power, Protective Armor, and Aggression, against those tags, a clean split emerged. Protective Armor (21 of 24 items) and Aggression (6 of 6 items) belong predominantly to entanglement. Forgiveness (10 of 13 items), Unique Identity (13 of 15 items), and Personal Power (7 of 9 items) belong predominantly to engagement.
All five sub-patterns held together internally as reliable, coherent scales. The more important test was how they related to one another once correctly sorted by dimension.
How the Two Dimensions Talk to Each Other
Forgiveness, Unique Identity, and Personal Power, all three engagement sub-patterns, correlated with one another in a tight, consistent range: .59 to .68. Each of these three, in turn, correlated with Protective Armor, the entanglement sub-pattern, in a lower, looser range: .42 to .50.
That pattern is exactly what a well-behaved two-dimensional model should produce. Sub-patterns belonging to the same dimension should relate to one another more closely than sub-patterns belonging to different dimensions, and that is precisely what appeared. More strikingly, the cross-dimension range (.42 to .50) sits just below the correlation already established between the two dimensions themselves in our first level anaysis of the data (r = .57). Engagement's internal sub-patterns are not simply correlated with entanglement's sub-pattern by coincidence; they are reproducing, at a finer grain, the same relationship the two dimensions have with one another at the top level.
Within Entanglement: The Shape of Protection
One of the most revealing sub-patterns was Protective Armor, which did not behave as a single, uniform stance.
Instead, it organized into three distinguishable emotional safety strategies:
A withdrawing posture marked by helplessness and hopelessness.
Summarized as “Life is having its way with me.”
A defending posture marked by status and blame.
Summarized as “It’s all someone else’s fault because it could never be mine.”
A helping posture marked by over-functioning and rescue.
Summarized as “I have to help or else.”
These three strategies map closely—though not perfectly—onto patterns long recognized in the threat-response literature as fight, flight, and fawn. They also map closely, and again not perfectly, to Dr. Stephen Karpman’s Drama Triangle roles of Victim, Persecutor, and Rescuer.
This framework is not meant to replace those models, but to give them a measurable, assessment-based counterpart inside emotional well-being work. The strategies we describe are grounded in the same nervous-system realities those frameworks point to, yet they are defined by how they show up in everyday organizational life and by how they cluster empirically in this assessment. In that sense, they sit alongside fight–flight–fawn and the Drama Triangle as close relatives: recognizably similar, but tuned to the specific architecture of entanglement and to the practical demands of measuring and shifting these patterns at scale.
Each of these three strategies also varies in degree rather than simply being present or absent. A person can be Active in a strategy (currently relying on it under pressure), Influencing (still shaped by it, though less acutely), or Transcended (showing no trace of it at all). Fully transcending all three strategies at once proved rare (well under 1% in this sample), consistent with it being a deliberately high bar, while partial engagement with at least one strategy was common.
Crucially, this sub-pattern was not built into the model in advance. It emerged from how the items related to one another in the data. That gives the pattern added weight as a genuine finding rather than a designed outcome.
Where Anger Lives
Aggression, which incorporates anger and is called the emotional undertone, the second entanglement sub-pattern, did not spread evenly across the three protective strategies. It concentrated specifically in the defending strategy, correlating with it at .59, compared to .40 and .43 with the other two. This is consistent with a pattern noted in other research: being blamed can register in the body as something close to a physical threat. Anger, physiologically, is the body's fight response. It makes sense that it shows up most strongly in the one protective strategy that is itself organized around fighting back.
Within Engagement: Forgiveness, Identity, and Power
A closer look at the Forgiveness items revealed an internal structure of their own. Roughly half of the items track the resolution of specific grievances: letting go of grudges, finding peace with a particular past hurt, no longer needing to withhold an apology. The other half track something related but distinct: an acceptance orientation toward life in general, the sense that each moment unfolds as it should and that difficulty carries its own purpose.
These two facets are correlated (r = .27), enough to confirm they share something in common, but clearly not identical. Practically, this distinction matters most for individual insight: someone still working through a specific old hurt is in a different place than someone who intellectually accepts that life happens for a reason but has not yet released a particular grievance. For scoring purposes, however, the combined measure across both facets predicted Unique Identity and Personal Power more strongly than either facet alone, a reminder that a longer, combined scale often carries more signal than its parts measured separately.
Unique Identity and Personal Power round out the engagement side of this architecture. Unique Identity, drawn from 15 items, showed the strongest internal consistency of any sub-pattern in this study, indicating that the items reliably measure a single, coherent idea: clear values, identified strengths and gifts, known purpose, and healthy boundaries. Personal Power, drawn from 9 items, showed more modest but still acceptable consistency. A closer look at what specifically drives each of these two constructs, in the way this article examines Forgiveness and Protective Armor, is a natural next step.
When the Composed Self Cracks
Perhaps the most practically useful finding involved what we call the Managed Self, a broad measure of how resolved and put-together a person appears across the full assessment. On its own, this measure aligned closely with Forgiveness, Identity, and Power, which makes sense, since it draws on much of the same engagement-side material.
But we also tested something the broad measure alone cannot show: whether that composed presentation holds up when a person is asked a small number of pointed, specific questions designed to surface a crack under pressure. The degree to which someone's classification dropped under this closer questioning was essentially unrelated to their Identity (r = -.04), Forgiveness (r = -.16), or Power (r = .02) scores, but it was clearly related to Aggression (r = -.48). Given what we now know about which sub-patterns belong to which dimension, this is exactly the pattern a well-behaved model should produce: a stress-crack mechanism built from entanglement-flavored items relates to another entanglement measure (Aggression), not to the engagement measures. Whether the managed self holds together under acute pressure is a distinct, entanglement-side capacity, separate from how resolved a person's engagement-side self-concept appears.
Put simply: how composed someone looks is one measurement, largely an engagement-side signal. Whether that composure survives a specific kind of pressure is another, entanglement-side signal. Both are worth tracking, and they are not the same thing.
Why This Matters
For a framework built on the idea that emotional well-being has structure on both sides, these findings matter in a specific way. It is one thing to show that entanglement and engagement exist as dimensions. It is another to show that each dimension has internal architecture that behaves the way the underlying theory predicts, and that the relationship between the two dimensions reappears consistently, even at the level of their individual parts. That is a stronger, more testable claim than two overall dimension scores could support on their own, and it held up under direct testing rather than assumption.
A Note on What This Evidence Shows, and Does Not Yet Show
These findings come from a single 224-person sample. That is enough to establish that the sub-pattern structure, and its correct assignment to entanglement or engagement, is plausible and internally consistent, but it is not yet enough to call it settled. Our analysis in the last article became meaningful because the same two-dimensional structure reappeared in a second, independent sample. The natural next step for this internal architecture is the same test: does this same structure of sub-patterns, correctly sorted by dimension, reappear in a new group of people?
Where This Leads
The current evidence supports a clear picture: entanglement and engagement are each structured architectures of related, distinguishable parts, not single undifferentiated stances, and the relationship between the two dimensions holds even down at the level of their individual sub-patterns. The next steps are to replicate this structure in an independent sample, and to examine how these sub-patterns, and the distinction between a resolved self-concept and a self that holds up under acute pressure, relate to outcomes such as burnout, leadership effectiveness, and relationship quality. Even at this stage, the architecture on both sides of the model is coming into focus.
Conclusion
Entanglement and engagement are each made of distinguishable, non-redundant parts. Protective Armor and Aggression sit on the entanglement side; Forgiveness, Identity, and Power sit predominantly on the engagement side. Sub-patterns within a dimension relate to each other more closely than sub-patterns across dimensions do, and the size of that cross-dimension relationship closely echoes the correlation already established between entanglement and engagement themselves. Within that architecture, Forgiveness separates into resolving old grievances and accepting life as it unfolds; Protective Armor separates into three recognizable postures echoing fight, flight, and fawn; anger concentrates specifically in the fight posture; and a person's composed self-presentation is measurably distinct from whether that composure survives acute, targeted pressure, tracking with the entanglement side rather than the engagement side. Taken together, this is the internal map the framework has proposed, now visible in real data, and it sets a clear, testable foundation for the replication and outcome studies that come next.
