More on Relativized Systemic

RS is a formal framework developed after MRC, for the construction of technical-scientific methods, explicitly relative to the conditions of concept emergence. RS regards any conceptual-physical entity as the result of a constructive process based on an immanent principle underneath any cognitive action, any projet : the distinction between the entity we focus on and the intersubjectively known way of specifying "what to act upon", "how to proceed" and "what are the results" to generate, describe or imagine this entity. This distinction entails that both knowledge and projects construction processes rely on anticipations, that is models, which configure any descriptive cell.

The formalization of concept genesis through an algebra physically meaningful, makes possible to figure the emergence of, up to know purely qualitative or mathematic notions, such as « physical continuity », « physical state », « persistent entity », « system »… This formalization has paved the way for RS based methods tooling, such as RSE.

New relativized and physical definitions of probability, entropy and complexity, to qualify constructed models and their connections to the domain of reality they focus on, spring from the whole construction.

The genesis of RS formal framework out of MRC is exposed in a thesis by H. Boulouet [Valenciennes, 10-03-2014].

Constructivist features

At the very root, we postulate some basically common way of living, which makes possible to build consensus about basic situation through primitive communication (pointing fingers …) : primordial description.

Any further consensual knowledge begins with the public registering of observable marks, regarded as the outward signs of interactions between some domain of reality we focus on, and the biological or artifactitious devices we use to describe, such as they have been beforehand, intersubjectively conceptualized. These conceptual-physical devices may be our own body or any apparatus aiming at standardizing qualifications.

A distinct entity then emerges as a concept from the physical substratum out of the stability of marks localized in space and time, when repeating the same experience. Let’s call contact handlers the particular places we are supposed to get in touch with, relatively to the adopted space-time referential, when conducting these experiences. Then, we understand, from our classical and spontaneous psychic standing, that we may eventually come to conceptualize the existence of an entity we simultaneously endow with a spatial and temporal shape and specific qualifications, out of these handlers, just like a blind person may come to recognize some material object using his white stick.

When a lack of physiological perception makes it difficult to localize these handlers, they may be infer from the qualifications achieved in connection with the conceptual molding set up by the prior built knowledge of the devices used and of the substratum they are supposed to interact with (such is the situation in quantum physics).

On the opposite, straightforward physiological perceptions are bound to make us absolutize the perceived contour. We then tend to identify the existence of the described entity with the resulting shape. Such a contour, however, may not be relevant with regard to other ways of qualifying, using registering devices from example.

Such a conceptual standing entails prominent methodological consequences :

• Any (no primordial) knowledge building requires the prior consensual conceptualization of the means used to capture some new piece of reality and of the substratum they interact with.
• A stable physical entity is a concept built out of statistically converging data associated to space-time localized contact handlers. They are the operational means we use to generate, to handle and to qualify the entity to be described, relatively to a finite set of aspect-views and qualification grids.
• Such a conceptualization is fundamentally relative to the physic-conceptual way of acting and to the aspect-views involved. How relevant it is, may be contextually assessed through the successful expectations it makes possible.
• The inner and the outer, such as a localization of a physical entity, are concepts built out of public contact handlers. We infer the existence of a physical entity out of these consensual localizations, but the concept of a physical entity transcends, on its own, these localizations it stems from.

Reality Conceptualization Framework

RS formal construction culminates with the definition of a physical probability law that puts forth the basic form of conceptualization of any physical support that emerges together with the assertion of a scientific knowledge or of a minimal control over some field of reality.

RS considers that asserting the factual existence of a probability law amounts to postulating that the physical support of such a law might be described by a non-linear, non-periodic but stable deterministic model : a kind of "chaotic system". ). Such a description would make emerge new entity concepts, whose description would be deterministic and genetics interdependent. This emerging organization corresponds to a <ùark>more precise level of analysis. For that reason, RS regards the description of a physical entity through a probability law as symptomatic as a level of transition between two levels of conceptualization of deterministic character.

This hypothesis reflects a psychological standing relative to the relationship between the accuracy of the measuring referential (the qualification grid) and the process producing values (initial conditions and modus operandi), such as they are conceptualized, based on preexisting knowledge: concept of epistemic referential. Periodic, not to speak of linear forms of determinisms, are regarded as exceptions.

A deterministic description may be out of reach for reasons either theoretical (limit imposed by Planck’s constant) or factual (too many dependencies, too much sensitivity to operating conditions). Conversely, in RS, any deterministic description of a persistent and stable, non-linear system, that amounts to a “physical Turing machine” -like description (defined in RS), may be used to compute a probability law relative to the chosen epistemic referential, reducing appropriately the relative level of accuracy of the terms involved. In RS, there may be as many such descriptions of the same physical support as they are viewpoints associated to specific event algebra defined on the transfer or basic description level, so, as many probability laws endowed with specific Kolmogorov spaces defined on a unique Universe of elementary events.

Such a framework may be relevant to supervise distant systems or to assess whether some usage or safety feature of a new product does conform to expectations - computation of a level of confidence: statistic convergence towards the probability law computed out of models. The same approach is applicable to scientific hypothesis in a different perspective: there, the hypothesis is at stakes relatively to facts, not the conformity of a product relatively to the project that anticipates and drives its realization.

This deep concern for the factual verifiability of our virtual constructions, at any abstraction level and at any stage of a project, has driven the application of RS to engineering activities, called Relativized System Engineering. It also paved the way for a new approach to Artificial Intelligence, stressing the parameter-types that drive a constructive conceptualization process in RS.

For an inner view on the way RS tackles identified stakes, you may download hereafter document (french written): RS approach