April 1971

Criteria for an Adequate Meta-model

Preliminary list

- / -

(Extract from a paper entitled: Preliminary Notes on Social Relativity: with implications for conceptualization of the social process, communication, value formation and the function of non-territorial organization.)
Ordering of a dimension
Dynamic and evolutionary
Progressive unification
Justification of superseded perspectives
Integrative power
Problem relevance
Extra degree of freedom
Implementation criteria

1. Complexity

The meta-model must possess a high degree of complexity. If the model is to be used to relate observers, then it must be at least as complex as the most complex existing model relating observers. Perhaps this should rather be expressed such that the model should possess sufficient ordered complexity to be able to contain complexity without imposing upon it a degree of orders lower than that which it warrants and which is essential:

(a) for the model to be credible

(b) to avoid creating the impression of a rigid framework which would inhibit individual creativity, spontaneity and unpredetermined fulfillment.

If it is not more complex than the most complex existing model then the processes that the latter describe or rather the perspective on variety from which the model originated will not be effectively contained: given that the social process in which we are embedded is the most complex phenomena of which we know, one must make use of the most complex model that man has been able to develop to contain it adequately.

2. Simplicity

"If a model approaches or exceeds in complexity the behaviour it purports to account for, it becomes a fair question whether one has a theory of that behaviour or merely a copy of it." (John C. Loehlin. Computer Models of Personality: Random House, 1968, p. 150) . A key feature of the model is that U should have structural properties which make it simple to understand without preventing incorporation of the amount of complexity necessary. A good example of this is the simple macro-organization of crystals which does not in any way deny the existence of many more complex levels of order such as are detected by X-ray analyses.

3. Ordering of a dimension

The model should make very apparent the degree of order of each of the various phenomena it contains. Order should be a very distinct dimension. The reason for this is that there is a strong tendency to collapse hierarchies of order and to categorize the most general in the same manner as detail encompassed by it. It is as though a multi-level ordering, possibly in a number of dimensions was projected onto a two-dimensional surface.

The direct consequence of such projection is that people cease to perceive the significance of coordinative power - except as a threat or an ambition. The continuum local through national to transnational is not seen as a continuum but as projected onto the level local {from which transnational is abstract) or transnational (where the relevance to local is lost) . Similarly with the continuum specialized to transdisciplinary - from a given specialization the significance of transdisciplinary is lost because it projects poorly onto that level.

4. Open-ended

The model should be constructed in such a way that it makes no pretence to being complete. Completion precludes further growth and therefore stultifies creativity. The model should

(a) outline a perspective which suggests the existence of a multitude of other perspectives compatible with it and with each other;
(b) indicate a direction of evolution of the model and compatible perspectives namely a direction of development arising from creativity and perception of new variety.

5. Dynamic and evolutionary

The model should not be static or structural but dynamic, evolutionary and process oriented. As an evolutionary model it should suggest a direction or a manner in which psycho-social evolution is taking place. How is new variety introduced into the psycho-social system? This developmental dimension should be made very apparent.

Again there is a tendency to assume that psycho-social evolution does not take place because there is no comprehensible framework onto which to hand the concept. The non-physical growth of a child, a group or even a concept is not treated as a continuous process because we cannot handle process oriented concepts but only discontinuous concepts or slices through the growth dimension. The work of Jean Piaget has not penetrated outside his own discipline. We are faced with a myth that people once adult do not develop - or else as an exception change suddenly.

6. Progressive unification

Combining criteria 3 and 5 the model should make clear that society, if only because of limitations to living space and the increase in population is being funnelled into progressive unification as one aspect of psycho- social evolution.

The explosive fragmentation of society has been noted but this occurs within a framework of increased interdependence. Autonomy and conceptual distance are stressed precisely because geographical and temporal isolation are no longer feasible means of establishing unique territory.

7 . Comprehensiveness

The model should be comprehensive. It should not "freeze" particular parts of the social system as being irrelevant or unamenable to treatment. It should rather suggest areas of experience which have not already been contained in the model.

The model should specifically suggest the complementarity within the psycho-social process of normally dissociated phenomena or fields of experience.

8. Justification of superseded perspectives

Following on from the previous criteria, the model should not ignore out of date or non-current viewpoints but should rather justify their existence within their particular framework.

Theoretical advances in a discipline are neither instantaneous in their effect on workers in that discipline nor do they supersede all preceding perspectives. An advance in any area takes time to penetrate through the psycho-social system. At any one moment the existing perspectives in different sections of society on a particular phenomena may represent the whole intellectual history of a discipline. Primitive views may still coexist temporarily with the most advanced. Lines of development long abandoned by the main stream as representing a cul-de-sac, may continue to appeal to certain individuals as being the most significant.

No individual is up to the minute in all of the fields which are his special concern. His "out-of-dateness" should be modelled in unprejudiced terms. "Backward" evolutionary forms have often overtaken the main stream when the latter has over-specialized in response to what later proved to be environmental conditions.

These are significant features of the social process which need to be adequately represented without over-favouring the modern.

9. Self-prediction

The model should not only recognize the existence of other models but should incorporate its own existence.

A model which does not acknowledge its own existence is external to the processes which it models: if its own existence comes about as part of these processes then clearly the model is not adequately modelling the processes in question.

This is the peculiar problem of psycho-social process modelling. We cannot adopt the traditional approach of isolating a group of phenomena and producing models for those phenomena whilst ignoring the phenomena associated with the blocking out process.

10. Integrative power

The model should perform a very powerful integrative function in bringing together disparate social phenomena. This is essential to counteract the "big-bang" explosive disintegration of society whereby it is increasingly more difficult to comprehend or act in anything but a restricted domain.

Clearly the model cannot block the "big-bang" type development but it should suggest means of maintaining relationships between the most distant parts of the social system.

11. Personal

The model should not be an impersonal conceptual tool only. There should be features of it which have important implications for the problem of alienation and depersonalization. Basically the model should have "participative" features. Just as one can now justifiably speak of the need for the participation of the man-in-the-street in decision-making processes which affect him, one should be able to speak of the participation of "persons" in models which contain his activities in some way.

A model is a reifying device, namely it has strong depersonalizing and unitizing properties. "Persons" are represented as units - this is taken to mean that it is scientific.

The model to be useful must break this barrier and be both scientific and significant to the person as a person - not in terms of his purely academic interests. The model must be something that a person can use to facilitate reconceptualizing or redefinition of his relationship to his environment.

This should not be focused on general philosophical considerations but should clarify - without necessarily threatening - moment by moment relationships of the person with his environment. This does not mean that the model should in any way impose a new manner of looking at the environment it should rather clarify the manner - whatever it is - in which the person looks at the environment.

12. Problem relevance

The existence of the model must accomplish more than offer a means of logically grouping and understanding certain currently dissociated phenomena It should not be an academic talk-piece without immediate and direct consequences for the problem which it is designed to resolve - namely it should not be a "scientific" model. The model should act as a catalyst for social change.

13. Extra degree of freedom

The model should decrease the social tension by identifying an extra degree of freedom in social processes. This should be associated with the rigidification and derigidification of social structures.

The extra degree of freedom is provided because the model should offer a technique for the redefinition, when appropriate, of relatively invariant phenomena detected in the environment. The environment is then no longer need be perceived as cluttered with permanent social entities which oppose change and necessitate their total physical destruction. The destruction, of required, is accomplished through the modification of concepts.

This is at present extremely difficult because the identity of many social entities is directly bound to a rigid definition of the entity conceived as a permanent phenomena. The function of the model should be to dissociate identity from a single sot of such phenomena.

14. Non-threatening

The very existence of some models threatens certain social entities and excites their opposition. This is because some entities are highlighted as "good" others as "bad".

The model should be non-threatening in that it must justify the existence and functioning of existing social entities on their own terms. The object is not to pinpoint the bad guys but to show in what ways each entity contributes to the total social process.

But whilst being non-threatening to the identity of social entities the model should demonstrate that the characteristics of inter-entity relationships if considered significant by the entity in question do have implications for the manner in which it functions. The recognition of the significance should be facilitated by the model but without it the model should not threaten.

15. Unoriginality

The model should not be so original or presumptuous as to suggest that nothing relevant had preceded it. In a special sense it should be totally lacking in originality and should merely suggest a manner of detecting many perspectives on how existing frames of thought and activity are related without in any way suggesting the need for any modifications in those frames of reference.

16. Implementation criteria

A further set of criteria is necessary to cope with the problems of making use of some models whilst providing perfectly satisfactory descriptions may not be suitable in practical terms because they cannot be widely used and understood.

  • Simplifiability: The model should readily permit a number of levels of approximation to it - namely it should be simplifiable to different degrees of explanatory purposes and for use without necessitating loss of its main features. This is clearly essential if the model is to have any wide impact. An over- complex model does not "travel" through the different layers of the educational, public information or decision-making systems by which it must be absorbed before it can be widely used.
  • Illustratability: The model should lend itself readily to explanation by illustration rather than by text. Again, for wide impact only illustrations can cross barriers of functional illiteracy whether in the case of the educated or the uneducated. It is important to recognize how rapidly a concept becomes inaccessible when pages of text are required to explain it. The cliché "A picture is worth a thousand words" suggests the need for a model readily explainable by gestalt perception.
  • Data-oriented: The model should not offer a purely verbal explanation but should permit activation by data at a fairly low level of detail which would provide results of some value for decision-making.
  • Understanding is not enough. It must be possible to make practical use of this understanding.
  • Delivery-oriented: The model should be such as to suggest designs for adequate delivery systems to ensure its use.

creative commons license
this work is licenced under a creative commons licence.