Beyond “Toxic” Relationships: Attachment, Temperament, and Difficult Dynamics

The language of popular psychology, words like "narcissist," "toxic," or "gaslighter", has become a kind of shorthand for relational pain. These labels can be validating for people who have felt confused or dismissed, but they can also flatten complexity in ways that ultimately don't serve people or their relationships. In Beyond Difficult, we look beneath the surface of these dynamics and tend to find a constellation of interacting processes: a sensitive or reactive nervous system, early attachment hurt, and parts of the personality that developed as survival strategies in childhood but cause real harm in adult relationships. Take what popular culture calls a "narcissist". In schema therapy terms, what we often see is someone with a strong Self-Aggrandising part and an Overcompensating part sitting on top of a deeply buried Vulnerable or Lonely Child part. The grandiosity, the entitlement, and the lack of empathy are coping strategies that developed because, at some point, vulnerability felt unsafe. That doesn't make the behaviour acceptable, but it can help us make sense of it.

Similarly, what people call "toxic" relationships often involve two people whose attachment scripts are activating each other's maladaptive coping strategies. An anxiously attached person whose abandonment fear is chronically triggered may pursue, criticise, or cling in ways that feel suffocating to an avoidantly attached partner, who in turn withdraws or stonewalls, confirming the anxious person's fear of abandonment. Neither person is "the problem", but together, without awareness, they create a destabilising dynamic. The guiding principle we come back to throughout the book is that all difficult behaviour is communication. Underneath every challenging behaviour is a feeling of threat, an unmet need, or a painful past visiting the present.

Temperament and attachment interact with each other, and that interaction is where the story of who we are lives. The concept of differential susceptibility is key here. We know from decades of research, including Stephen Suomi's remarkable cross-fostering studies with rhesus macaques, that highly sensitive or highly reactive individuals are more strongly shaped by their caregiving environment — for better and for worse. In supportive environments, sensitive people can flourish in extraordinary ways. In unsupportive environments, the same sensitivity makes a person significantly more susceptible to harm and negative outcomes. 

What this means clinically is that a highly sensitive child raised by a caregiver who was inconsistent, rejecting, or frightened is at particular risk of developing both an insecure attachment and a set of maladaptive personality parts that complicate adult relationships. They may be exquisitely attuned to threat, read ambiguous social cues as hostile, and have a nervous system that stays in a state of chronic low-level vigilance. From the outside, this person might look reactive, "too sensitive", or demanding. From the inside, they are doing exactly what their nervous system and early experience taught them to do to stay safe. The clinical takeaway is that temperament traits alone don't make a person difficult — it is the interaction between a person's temperament and their early environment, particularly the quality of their attachment relationships, that shapes the patterns we find most challenging in each other.

The internal working models—what we call "scripts" in the book—that people develop in early relationships follow us into every relational context, including the workplace. In fact, the workplace is one of the most fertile environments for attachment patterns to be activated, precisely because it involves hierarchy, evaluation, and belonging. Someone with an anxious attachment style may find performance feedback catastrophic, not because the feedback itself is harsh, but because their script predicts dismissal and rejection. Constructive feedback can activate their vulnerable child side, triggering shame, over-apologising, or bids for reassurance. Someone with an avoidant attachment style, by contrast, may resist collaboration, struggle with authority figures they perceive as controlling, or dismiss feedback as a way of protecting themselves from the vulnerability of being seen to need help. Their colleagues may experience them as cold, uncooperative, or arrogant.

For individuals with more disorganised or fearful-avoidant patterns—often those with more significant trauma histories—the workplace can be genuinely destabilising, with the simultaneous wanting and fearing of connection manifesting as explosive reactions or patterns of idealising and then devaluing colleagues. Research on adolescent attachment shows that insecurely attached teenagers were less likely to seek support from close friends and later entered adult relationships characterised by greater negativity and hostility—a kind of self-fulfilling prophecy in which they enter relationships expecting the worst, and either fail to call for support or select partners who don't provide it. The same mechanism operates in workplace relationships.

What makes exploring a client's own relational patterns so important is that our early relationship experiences don't simply stay in the past. They become the template through which we interpret every significant relationship that follows. As Bowlby observed, attachment influences human behaviour from the cradle to the grave. The internal working model we develop in childhood—our expectations of whether others will be available, whether we are worthy of love, whether the world is safe or threatening—operates largely outside conscious awareness, shaping how we perceive, feel, and behave in our closest relationships throughout life.

This means that what a client is experiencing in a difficult adult relationship is rarely only about that relationship. The partner who feels perpetually unloved despite evidence of care, the employee flooded with shame at the mildest piece of feedback, the friend who cannot tolerate any distance without panicking—these responses are usually older than the current relationship. They are echoes of earlier experiences of loss, rejection, inconsistency, or emotional unavailability. And critically, they are connected not just to relational difficulties but to the mental health challenges that so often bring people to therapy: the anxiety that won't settle, the depression that returns, the chronic sense of unworthiness or loneliness that persists even when life looks fine from the outside.

There is a meaningful difference between two people whose insecure attachment patterns are activating each other—where curiosity and mutual understanding can shift the dynamic—and a situation where one person is consistently, strategically exploiting or harming the other. In the book we discuss the Predator, the Bully, and Attack Mode in the context of more entrenched personality pathology, particularly antisocial presentations. These modes involve an intentional use of others as instruments, a capacity to harm without genuine remorse, and a pattern that does not respond to the kinds of interventions that work in more typical relational difficulties. In these situations, the clinical priority shifts: we are no longer working primarily toward mutual understanding or repair but helping the client accurately perceive what is happening—which can be surprisingly difficult when attachment bonds are involved—and supporting them to establish and maintain limits that protect their safety and wellbeing.

The schema therapy concept of empathic confrontation remains relevant, but the target changes. Rather than helping the client understand the other person's Vulnerable Child Mode with compassion, we are helping them distinguish between compassion and self-erasure, between understanding someone's history and tolerating ongoing harm. The guiding principle becomes safety first, both physical and emotional, and sometimes the most therapeutic thing a client can do is leave. Even in exploitative dynamics, the client's attachment history is often part of the picture—not as cause for blame, but as a focus for healing. Many clients in these relationships discover a People Pleaser Mode, a difficulty with limits, or a schema around subjugation that made them more vulnerable to this dynamic, and addressing those patterns is essential both for healing from this relationship and for future relational health.

The first thing I think is important is to validate the complexity of the bind. When there is genuine attachment involved—a parent, a long-term partner, a sibling—the desire to repair the relationship is not weakness or naivety. It is entirely human, rooted in the biology of attachment itself. As Bowlby observed, attachment behaviours are activated precisely when relationships feel threatened, which means that the more painful the relationship becomes, the more intensely the client may be pulled toward it. Understanding this neurobiologically can be a significant relief for clients who have been shaming themselves for not "just leaving."

The therapeutic work in these situations often involves several strands: genuine mourning of the relationship the client needed or hoped for, which may never have existed in the way they imagined; meaning-making about what the relationship has revealed about their own modes, scripts, and patterns; and careful work with limits, not just as behavioural strategies, but as acts of self-respect and self-preservation. For many clients, particularly those with a history of subjugation or People Pleaser Modes, establishing a genuine internal limit is the real therapeutic achievement, not just the external boundary that follows from it. We also hold open the possibility that some relationships can change. When we can recognise the lens of attachment and understand that someone's difficult behaviour is coming from their own insecurities and past wounds, we sometimes gain the capacity to respond differently and that different response can, over time, invite something different in return.

Throughout the book we offer a range of practical tools, but I'd highlight a few that we find most consistently useful. The first is curiosity before reactivity: when a client encounters difficult behaviour, the most transformative first move is often to pause and ask what the behaviour is communicating. This doesn't mean excusing the behaviour or abandoning one's own perspective — it means expanding the frame before responding, which short-circuits the defensive escalation that makes difficult interactions worse. The second is mode awareness—helping clients recognise which part of themselves has been activated. When a client can notice "I've just flipped into my Angry Child Mode" or "my People Pleaser is about to agree to something I really don't want to do," they create a moment of space between stimulus and response, and that space is where their Healthy Adult Mode can intervene. The third is safety, both internal and interpersonal: a flooded nervous system cannot access the prefrontal cortex effectively, so helping clients recognise when they or the other person are too activated for productive dialogue and having strategies for de-escalation before attempting repair—is essential.

The fourth is working with limits, not just as a tactical skill but as a values-based practice. Limits that come from a client's Healthy Adult Mode—grounded in self-respect rather than reactive anger—land very differently, both for the client and for the person they're setting them with. And finally, we emphasise the importance of what we call "solving the problem, not the person." The goal in confronting difficult behaviour is not to fix or change the other person, but to address the specific behaviour and its impact. Reframing the difficult behaviour as the problem, rather than the person as the problem, tends to create far more productive conversations and, not incidentally, is much more consistent with what the research on effective feedback actually shows.