Such models include values like a willingness to collaborate and an indication of where to go or what to achieve. They also depend on the people using the model—when being modelled. This means that they introduce the possibility of self-naming: the name given to the collective can be used by individual members and what members contribute (‘names’) can be recognised as required by the collective. One may differ about how to refer to this type of quality—self-naming or self-representing, or more popularly, self-organisation. The term self-naming will be used in the rest of the paper.
The transition to the models of the fifth level from those of the first four levels seems important in the history of studying organisational change. It implies something of a watershed. The first four types, for example the input–output models, are intended to ‘represent’ patterns in the behaviour of individuals, i.e. name them and make them recognisable. Representational models of this kind tend to be used to describe what may be defined as passive behaviour, i.e. behaviour that is not proactive (does not deal with the future) and not interactive (does not react to the naming). Such models identify what input to provide if one wishes to realise a preferred output, and hence may be used effectively in the control paradigm (if the models are of sufficient quality, of course). The models of the fifth level have been developed to deal with the alternative, that is, situations where behaviour is proactive and/or interactive. A different structure is needed, therefore: changes in behaviour have to be part of the models. This means the models are not representational and not part of the control paradigm. If they are to be used, such use must be part of the behaviour being modelled (as data) as well as serve as guide (name) to that behaviour. The obvious example is some collective where the model ‘of’ its achievements is also the model ‘for’ the behaviour of the individuals that collectively achieve that behaviour. This type of modelling appears to be quite powerful. It is analogous to members getting together to mirror the activities of those who developed the fifth model, and hence attempting to become a collective researcher themselves. To achieve research quality in the sense of Comte (1854), the collective has to develop its own language as a way to co-ordinate the behaviour of its members. It can be expected that there will be more than one language and that they differ from the language of variables.
Some of the aspects of the process of developing collectives (as part of the use of the fifth model) seem worth emphasising. One is that the more the collectives distinguish themselves from their environment, the more relations between members become mutual. Members no longer will be able to act as single individuals as this would imply that they aim to achieve personal objectives other than those of the collective and hence will not achieve the required type of quality. For the same reason, one cannot expect that some members aim to dominate others, i.e. impose the objective of domination over the collective objective. A direct corollary is that members have to respect and trust each other (or behave such that respect can be inferred). They have to respond to each other (i.e. act upon their observations of each other), make themselves addressable, insist on the transparency of each other’s objectives (or if one wishes, each other’s subject position) and make themselves, their objectives as well as their feelings of commitment mutually identifiable. These activities may be facilitated by the development of and adherence to rules, as an expression of members taking responsibility to accept and help maintain each other’s contributions. Adherence to the rules may be delegated to a referee, i.e. a special kind of observer. Additionally, remaining a member must be attractive—and hence collaboration must be assumed to help achieve something that each member prefers in some sense (not necessarily the same), including that it does not detract from a member’s well being, psychologically as well as physically.
