When the Tools Define the Problem

Better tools can lead to better care. But psychiatry also faces a risk that the tools themselves will define the problem. What can be measured receives attention and authority, while other aspects of a person’s life become harder to see. The understanding underlying the measurements may, meanwhile, become invisible and be taken for granted.

Two articles published in Nature on 28 October 2026 illuminate this tension. David Adam’s feature, The grand plan to transform psychiatry — by personalizing mental-health care, describes efforts to make treatment more effective through biological tests, digital data and artificial intelligence. The editorial How personalization and psychedelics are revitalizing psychiatry considers this development alongside research into psychedelic treatment.

These ambitions matter. People seeking help should not have to endure lengthy courses of treatment that do not work. But what kind of precision do we need? Measuring symptoms, predicting treatment responses and understanding a person’s life are different tasks. Progress in one area does not automatically improve our knowledge in the others.

New Technology, Familiar Questions

It is well established that people with the same diagnosis may have different needs. The importance of relationships, finances and personal history is equally familiar. These have long been central concerns in social work and social psychiatry.

What is newer is the capacity to bring together large amounts of information. Sleep, activity, genetic factors and previous treatment outcomes can be analysed to identify patterns that might improve treatment decisions.

Adam’s feature also shows how difficult this is. Many of these methods have yet to demonstrate reliable benefits in routine care. A statistical difference between groups does not necessarily tell us what help the person in front of us needs.

Technology expands what we can investigate. Yet the choice of information already rests on an understanding of what the problem is.

Tools Belong to a Context

Martin Heidegger’s analysis in Being and Time, particularly §§ 15–18, offers a way into this question. His term Zeug refers to equipment situated within practical contexts of use and purpose.

A hammer is understood as a hammer through its connections with nails, materials, the work and what is being built. It belongs to a network of references: something is used for something, which in turn serves a larger task. Investigating the hammer’s measurable properties may be useful, but it does not exhaust our understanding of it as a tool.

Heidegger refers to the Greek pragmata: what we deal with in our practical engagement with the world. He also argues that the Greeks did not fully clarify this practical character. Their language provides a clue for challenging the idea that we first encounter isolated objects and then assign meaning to them.

These references ultimately lead back to human possibilities of being and living. Heidegger calls this the Worumwillen, usually translated as the “for-the-sake-of-which”. It concerns more than the conscious goals of an individual action. Human beings participate in practical contexts, but their way of being differs from that of the equipment they use.

Applied to psychiatry, this directs attention to the tasks within which tools operate. A sleep sensor acquires significance through both what it can actually record and what the recording is intended to help us do. What makes that help relevant depends, in turn, on what is at stake in the person’s life.

This is our application of Heidegger’s analysis. It allows us to examine the assumptions underlying the use of tools rather than take them for granted.

What Do the Measurements Tell Us?

This raises a question in the philosophy of science: what kind of knowledge does a measurement provide, and what conclusions can we draw from it?

When sleep, movement or speech patterns are used as indicators of mental health, we assume that they reveal something relevant about the condition. Such assumptions need to be examined. A measure can be reliable, producing consistent results, without being valid for the purpose for which it is used.

A system might identify an unusual speech pattern with considerable certainty. Its clinical meaning may nevertheless remain unclear. Does the pattern reflect illness, cultural background, a distinctive way of communicating or a response to the situation? Accurate recording does not, by itself, establish what the observation means.

We must also distinguish prediction from explanation. An algorithm might predict which treatment someone will benefit from without explaining why it works. That could still be valuable. Conversely, a plausible explanation of a condition does not guarantee that we can predict its course or select an effective treatment.

A third distinction concerns levels of explanation. Mental distress can be investigated through biological mechanisms, personal experience and social circumstances. The fact that fear involves mechanisms in the brain and body does not establish that the cause of persistent fear is best understood at that level, or that the most effective intervention must target biology.

Problems arise when knowledge of one aspect of distress is treated as a sufficient understanding of the whole.

When Distress Has a Reason

Consider a hypothetical example.

A woman seeks help because she sleeps poorly, startles at small sounds and struggles to concentrate. She is exhausted and describes persistent anxiety. Sleep monitoring confirms frequent awakenings. A questionnaire indicates significant anxiety symptoms.

As she begins to feel safe in the conversation, she explains that her partner checks her phone, monitors her spending and responds with threats when she tries to make her own decisions. She lies awake listening for footsteps.

The measurements may be accurate. A biological explanation of her body’s state of alertness may also be correct. Yet something essential is missing if the threat at home remains unknown.

She may need treatment for her symptoms. She also needs someone to assess her safety and help her explore her options. If the sole aim is to reduce her anxiety, there is a risk of treating the alarm while the danger remains.

Her account changes our understanding of the task. Better sleep remains important, but improvement now also involves safety and the possibility of making decisions about her own life.

A Person’s Account Is Also Knowledge

A person’s own account is often described as subjective. It expresses experience from a particular standpoint. Precisely for that reason, it can reveal what measurements do not: what the person fears, which events matter and why certain choices seem impossible.

The account is not infallible. It requires examination and interpretation, as measurements do. Limited movement may be associated with depression, pain, a lack of transport or fear of going outside. The observation provides a starting point for further investigation.

Nor does a scientific commitment to scrutiny and verification require all relevant knowledge to be expressed numerically. The question is how information is justified, critically examined and considered alongside other knowledge.

Professional judgement involves assessing these connections in a specific situation. A decimal point does not make an assessment more valid if the understanding on which it rests is incomplete.

Data Confer Power

Digital tools can collect intimate information about where people go, whom they communicate with and how they sleep. Privacy concerns involve both secure storage and control: who may use the information, for what purposes and for how long? Can the person have a misleading assessment corrected or challenge an interpretation?

A risk assessment may acquire more authority than its evidence warrants. If it follows someone through the care system, an uncertain assumption can shape how subsequent professionals approach them. A possible future condition may begin to function as a description of who the person is.

Ideas about normality matter too. A statistical deviation is not, in itself, a disorder. Unusual speech patterns, limited social contact or a different sleep schedule can have different meanings. Assessing what requires treatment also involves suffering, functioning, circumstances and the person’s own wishes.

Here, philosophy of science meets ethics. Categories organise knowledge and influence the people described through them.

Vulnerability Requires Clear Boundaries

Psychedelic treatment brings questions of power and interpretation into particular focus. The Nature editorial emphasises the need for qualified staff during experiences that can be intense, unpredictable and prolonged. The US Food and Drug Administration also highlights increased vulnerability and suggestibility during treatment sessions.

A powerful experience can have profound personal meaning. Its intensity does not, however, establish the truth of a particular interpretation. Practitioners must take care not to use a patient’s vulnerability to lend authority to their own assumptions.

Patients must be able to trust that their boundaries will be respected and remain free to disagree about the meaning of what has happened. Oregon’s regulations for psilocybin services include specific provisions governing touch and prohibiting sexual contact between facilitators and clients. These rules do not establish how frequently boundary violations occur, but they illustrate the need for explicit safeguards.

Potential therapeutic benefits do not remove responsibility for how care is delivered. Safety requires clear boundaries, accountable follow-up and meaningful ways to raise complaints.

What Should Count as Improvement?

Less anxiety, better sleep and fewer depressive symptoms are important goals. Treatment outcomes must also be considered in relation to the life the person wants to lead.

For the woman listening for footsteps, improvement may mean sleeping more peacefully because she is actually safer. For someone else, it may mean recovering the energy for a meaningful activity or finding the confidence to participate in a community.

Here we return to the connection between the tool, the task and what is at stake. Measurements can show changes in symptoms. They cannot, on their own, determine which goals should take priority or which burdens of treatment are acceptable. That requires conversation and practical judgement.

Professionals bring knowledge of treatment. The person seeking help has particular insight into what is at stake. Personalisation must also include influence over goals and decisions.

When the Conditions Around a Person Must Change

The Nature editorial stresses that better diagnosis and treatment must be accompanied by prevention. Medical breakthroughs cannot replace a society that supports good mental health. The editors identify economic insecurity, sleep disruption, environmental pollution and social isolation among the factors that need greater attention.

This has implications for how responsibility is allocated. Clinicians have certain means of helping. Social services, employers and public authorities have others. If the entire task is assigned to the individual and the treatment service, important opportunities for change remain unused.

New tools deserve rigorous evaluation. We must examine what they measure, which decisions they can support and what understanding of the person underlies their use. Scientific rigour also involves recognising when the question itself needs to change.

When someone lies awake listening for footsteps, we need information about their sleep. We also need to ask who is walking down the hallway.

References

Adam, D. (2026, October 28). The grand plan to transform psychiatry — by personalizing mental-health care. Nature, 658, 1116–1118. https://doi.org/10.1038/d41586-026-03391-7

How personalization and psychedelics are revitalizing psychiatry. (2026, October 28). Nature, 658, 1101–1102. https://doi.org/10.1038/d41586-026-03402-7