A colleague and I recently proposed a model to guide the description and reform of mental health services and to clarify service evaluation (Reference Tansella and ThornicroftTansella & Thornicroft, 1998). The model, which we call the ‘mental health matrix’, consists of two dimensions, one temporal and one geographical. The temporal dimension comprises three phases: input (phase A), process (phase B) and outcome (phase C). The geographical dimension has three levels: regional/national (level 1), local or catchment area (level 2) and individual, meaning a patient or a group of patients (level 3). Nine cells are created by the intersection of these two dimensions (Table 1). The matrix can be used not only to deal with problems in the description of mental health services, but also to interpret accurately treatment outcomes. For example, to look for the possible causes of an episode of violence committed by a patient (which would be located in cell 3C, at the intersection of phase C, outcome, and level 3, the individual), one would refer not only to the process and input variables relevant to that level, the patient level (i.e. what was done before the episode of violence and what resources were available for the treatment of the patient), but also to the process and outcome on the two higher levels (how well the service responsible for the patient functions and what resources – inputs – it has at its disposal). In other words, to understand what has happened in cell 3C, we have to analyse the data and relevant facts in cells 2C and 1C and any relevant information in the other six cells of the matrix. More information on this matrix model and its possible applications is available elsewhere (Reference Tansella and ThornicroftTansella & Thornicroft, 1998; Reference Thornicroft and TansellaThornicroft & Tansella, 1999).

Table 1 Overview of the matrix model (Thornicroft & Tansella, 1999)

Temporal dimension: phase
Geographical dimension: level	(A) Input	(B) Process	(C) Outcome
(1) Regional/national	1A	1B	1C
(2) Local (catchment area)	2A	2B	2C
(3) Patient	3A	3B	3C

We can also use this matrix as a framework for studying the professional characteristics and attitudes of psychiatrists. Although it is difficult to classify members of a professional discipline on the basis of their attitudes, preferences and the choices they make in the practice of their work, one cannot deny that these exist into various clusters common to many professionals. Psychiatrists, for example, could be classified in many subtypes, but the ones most relevant to this model are, in my opinion, these two – the psychiatrist as archaeologist and the psychiatrist as architect.

The first, the psychiatrist–archaeologist, shows a predominant interest in the single patient (the individual level of our matrix). Members of this group devote most of their efforts to understanding the deepest origins of the symptoms and behaviour of their patients. Before deciding on an intervention, they take on the arduous task of understanding the dynamics that underlie the symptoms, uncovering these dynamics and using them to shed light on current problems. When these characteristics predominate in psychiatrists, and when their training and practice follow rigidly applied principles, we see assessments and treatments in which the construction of an aetiological theory (the ‘foundations’) absorbs most of the time and energy. Or, more extreme, we see treatment that excavates the foundations, analysing the various strata that come to light, so that any ‘superficial’ opportunities for solving the patient's practical problems or for improving his or her individual well-being are constantly overlooked or left to the patient's own initiative, while excavations continue at a deeper level.

The second type, the psychiatrist–architect, is particularly interested in the other two levels of the geographic dimension of our matrix (the local and the regional/national). Even when working with a single patient, this psychiatrist shows more pronounced constructive tendencies. These are manifested, for example, not only in his or her initiative and practice in organising mental health services (with respect to the two higher levels of the matrix), but also in clinical activities intended to lead to a better integration of patients into their social group, to solving problems of everyday living and social functioning and to understanding and alleviating their symptoms and suffering. Therefore this type of psychiatrist excavates in order to build on a known solid foundation, rather than to study and analyse the strata underneath. Sometimes, however, psychiatrists–architects are in such a hurry to build that they run the risk of doing so without having first established a firm base.

I have overstated the difference between these two types to make the distinction more clear and comprehensible. Clearly, good clinical practice requires that psychiatrists, as well as other mental health professionals, always possess both sets of attitudes in some degree, and that they be able to build treatment plans for the patient and to organise and coordinate mental health services that are founded on a solid base. This requires that the most significant material that comes to light during the ‘excavation’ be studied carefully, to understand in some depth the characteristics of the terrain (and the subsoil), which will influence the choice of structure. Thus, in every case, at each of the three levels of the geographic dimension of our matrix, one needs to strike a balance between the phases of preparing the foundation and of constructing the building, in order to develop a high-quality overall plan that allocates adequate resources (input) and establishes a sound approach to the work (process) to achieve the best outcome that the situation allows.

I cannot deny my own cultural interest in some aspects of the ‘archaeological’ approach in psychiatry and I am well aware of the fascination that this approach still holds (especially in South America and certain countries in mainland Europe) for many young colleagues and for those mental health professionals who devoted long years, thought and considerable personal economic resources to it, in becoming psychoanalysts. But against this fascination I can set the definition of architecture by a famous architect, Lina Bo Bardi, who disappeared in Brazil many years ago: “Architecture is a real and continuous human adventure.” Another great Italian architect, Renzo Piano, in the speech that he gave after receiving the 1998 Pritzker prize (considered the Nobel prize of architecture), said:

The duty of the psychiatrist as architect, then, is to help patients, particularly those with limited resources, to live their real human adventure, creating opportunities and services for them and with them, adopting a gentle but firm attitude and light touch, without imposing one's own agenda, but without failing to express one's own opinions. It is also a duty to help them, if possible, to overcome the pathology and handicap that, at the individual and social levels, constitute the causes of their suffering and their poverty. It is our duty to confront reality, not just to acknowledge it but to change it.