What if the symptoms we try so hard to eliminate are not the problem, but the solution — at least from the perspective of the nervous system? This is one of the most provocative and generative questions animating current discussions in psychotherapy research, and it was at the heart of a structured discussion I recently participated in at the Society for Psychotherapy Research (SPR) Annual Conference in Osaka, 2026.

The Attractor Metaphor

In dynamical systems theory, an attractor is a state — or set of states — toward which a system tends to evolve over time, regardless of where it starts. Think of a ball rolling into a valley: no matter where you release it on the hillside, it will settle at the bottom. The valley is the attractor.

Now imagine that the valley is depression, or chronic anxiety, or a pattern of emotional dysregulation. The person keeps returning to this state — not because they want to, not because of moral weakness or poor coping — but because the system has organised itself around this pattern. It has become stable. It has become, in a very real sense, the path of least resistance.

This is what we mean when we say that symptoms can be attractors. They are not random misfirings or mere symptoms of some deeper hidden cause. They are organised states — self-reinforcing, coherent, and remarkably resistant to simple intervention.

Why This Reframing Matters

The conventional medical model treats symptoms as signs pointing to an underlying pathology. Remove the pathology, the symptoms disappear. This logic works well for many physical illnesses. But in mental health, it has consistently underdelivered. Decades of research into the causes of depression, anxiety, psychosis, and personality disorder have not translated into dramatically better outcomes. Symptom reduction rates remain stubbornly modest. Relapse is the norm, not the exception.

Part of the problem may be ontological: we have been looking for the wrong kind of thing. If a symptom is an attractor — a stable dynamic pattern — then eliminating it is not simply a matter of removing a cause. You cannot drain a valley by attacking the ball. You need to reshape the landscape.

This is a profound shift in therapeutic logic. Instead of asking “what is causing this symptom?”, we begin asking: “What conditions are maintaining this attractor? What would it take to destabilise it? And what new attractors — new stable patterns — could emerge in its place?”

Implications for Psychotherapy Practice

If symptoms are attractors, then psychotherapy is — in dynamic terms — a process of attractor landscape modification. The therapist and client together are not trying to eliminate a defect; they are working to shift the topology of the person’s psychological terrain.

Several implications follow from this:

Stability before change. Attractor states are, by definition, stable. Attempting to force change from outside — through behavioural prescriptions or cognitive challenges alone — often meets strong resistance, not because the person is unwilling, but because the system is organised to return to its stable state. Effective intervention may require first understanding the attractor before attempting to move it.

Phase transitions matter. In dynamical systems, change rarely happens gradually. Systems often remain in one attractor for long periods before shifting — sometimes rapidly — to another state. Therapy may work by creating the conditions for such transitions: increasing variability, building new repertoires, expanding the range of possible states before a new stable pattern emerges.

Small interventions, large effects — sometimes. Near a tipping point, even small perturbations can push a system into a new attractor basin. This may explain why certain therapeutic moments feel disproportionately impactful — they arrive at just the right point in the person’s dynamic landscape.

The therapeutic relationship as context. From this view, the therapeutic relationship is not merely a vehicle for technique delivery. It is part of the dynamic environment that shapes the attractor landscape. Safety, co-regulation, and relational repair may themselves contribute to destabilising pathological attractors and enabling new ones to form.

Osaka and the Broader Conversation

The SPR Osaka structured discussion was an opportunity to bring these ideas into dialogue with a broader community of psychotherapy researchers — clinicians, process researchers, outcome specialists, and theorists. The conversation was generative precisely because it was multidisciplinary: different framings of the same clinical realities, held in productive tension.

What emerged from these discussions was not a unified theory — that would be premature — but a shared sense that the field needs richer metaphors. The language of symptoms-as-pathology, of diagnosis-as-essence, of treatment-as-elimination, may have reached the limits of its usefulness. The language of dynamics, of patterns, of attractors and bifurcations, opens new conceptual territory.

This does not mean abandoning what we know. Evidence-based treatments remain our best tools. But it does mean holding them differently — as perturbations to a dynamic system, rather than corrections of a broken one.

Looking Forward

The reimagining of psychotherapy through the lens of complexity science is still in its early stages. There are enormous theoretical and empirical challenges ahead: How do we measure attractors clinically? How do we identify tipping points? How do we design interventions that work with dynamic landscapes rather than against them?

These are hard questions. But they are the right questions. And they are beginning to receive the serious scientific attention they deserve.

The conversation continues — in Osaka, and beyond.