Abstract of the original article called – 
‘Psychiatric diagnosis as a pseudo-specialist language’.
Agnew, Joyce & Bannister, Don (1973). 
'Psychiatric diagnosis as a pseudo-specialist language'. 
The British Journal of Medical Psychology, Vol. 46, Part 1, March, pp.69 - 74.

The diagnostic system in use in psychiatry is presented as a general analogue of diagnosis in physical medicine. Thereby, it is proffered as a specialist language – a construct subsystem. A specialist language may be regarded as a well-defined and publicly agreed network of terms uncontaminated by lay language with clear implications for work in a professional field. The degree of structure, independence, stability and interjudge agreement to be expected in a specialist language is well illustrated in fields such as civil engineering, law, pharmacy, meteorology, banking, marine navigation and so forth.

The idea that psychiatric diagnosis systems are too poorly structured to be looked on as technical or specialist languages has been frequently mooted and experimentally investigated. Szasz (1962) and Sharma (1970) have viewed psychiatric diagnosis as a socio-political judgement; it has frequently been shown to have low interjudge agreement in terms of decisions made about particular patients (Foulds, 1955; Kreitman, 1961); and other investigations have shown the degree to which non-technical factors appear to affect diagnostic judgements (Arnhoff, 1954).

Evaluating Diagnostic Systems

Studies to date seem not to have tested the semantic quality of the diagnostic system in use in psychiatry in the following four respects:

1. Traditional studies have checked interjudge agreement only in terms of agreement between psychiatrists as to the applicability of a particular label to a particular patient. They have not checked whether psychiatrists are agreed as to the meaning of the particular label. In construct theory terms, they have checked agreement about elements but not agreement about the relationship between constructs, and clearly such a check is necessary. We may all concur that Bloggs is a good or a depressed man without necessarily concurring about what constitutes goodness or depression. There is evidence of a possible failure of agreement on the part of the psychiatrists as to the implications of diagnostic labels in Bannister et al. (1964), where it was shown that there was little consensus as to what treatments were implied by different diagnostic labels.
2. Traditional studies have checked aspects of interjudge agreement between individual psychiatrists but they have not checked the consistency of the individual psychiatrist in making diagnostic judgements from one group of patients to another and over time.
3. Traditional studies have not provided standards of comparison with other language subsystems whereby we can estimate the relative structural qualities of diagnostic psychiatry as a category system.
4. Traditional studies have not checked the degree to which the language of diagnostic psychiatry is contaminated by (and thereby confused with) lay language.

The present experiment is designed to remedy these shortcomings, using a form of repertory grid (Kelly, 1955; Bannister & Mair, 1968) as the method of investigating language structure.

Certainly, the experiment suggests an answer to the last-line defence of users of the traditional diagnostic system in psychiatry who tend to argue that they must have some way of categorising patients and what alternative is there. The answer would seem to be that the everyday lay language would offer at least as structured, as reliable and as public a set of terms for describing human behaviour and psychological characteristics.

Summary

Eight consultant psychiatrists completed grids using their patients as elements and both formal diagnostic categories and lay descriptive terms as constructs. Results indicate that the psychiatrists are no more stable and have no greater interjudge agreement in using diagnostic terms than they achieve with everyday language. Additionally, the two languages appear to ‘mix’. It is concluded that psychiatric nosology is not a true specialist language.