Multicontextualism

on principles for knowledge of differences in unity

Pieter Wisse

I have tended the growth of some new philosophical roots.

How else can I state such an impossibly immodest hypothesis in a modest manner? A claim to such basic novelty definitely sounds pretentious. Wasn’t the frontier of the philosophy of first principles closed long ago? Or will there always be space for fundamental discoveries? But, then, how did it come about that I was able to make some? Or is it only a rediscovery?

As branches of philosophy are concerned with, and comment upon, development of knowledge, philosophical roots must provide them with principles. Of course, such roots are essentially philosophical in nature. A clear distinction between roots and branches, therefore, is never possible. Earlier branches may lead to later roots; roots of philosophy grow from philosophy, too. Driven by an accumulation of actual problems, I arrived at what I propose here as today's new roots.

Regardless of realism or idealism, or of any intermediate world view, traditional theories of knowledge imply a singular relationship between information and what that information means to an interpreter. Its meaning is often denoted as knowledge about an object believed to exist autonomously or as an autonomous idea(l). Particular information is supposed to refer to a particular object only. Modern semiotics, with its familiar triadic foundation, does not escape the perspective of absolute objects, either real or idealized. Fair enough; it is recognized that the reference may be false. Even then, the fraudulent use of information relies on acceptance of singularity in correspondence between information and (non)object.

Much of philosophy, in many branches, has been dedicated to coming to terms with the problems thus generated. But such problems are best solved at the roots, with as wide a bearing as possible. As a principle of knowledge, the idea of mutually autonomous objects with corresponding supplies of information is just too simple. It does not support the increasing simultaneous and subsequent variety experienced during the encompassing life that a person, or, for that matter, many other objects, leads today.

Instead, I superimpose the concept of situation, or, as its equivalent from the perspective of information, context. At any time, a multitude of contexts may be assumed. Next, a particular object may be thought to appear in some, or all, of those different contexts. Depending on the situation or context, the object will show a characteristic identity, with contextual properties (or, in general, contextual behavior). Information, thus structured, carries multicontextualism. As a matter of principle, there is no absolute, only contextual meaning. Different meanings may occur for the same object, presuming equally different contexts.

This epistemological approach, with its strong ontological flavor, I have named multicontextualism. Only by making context an explicit, primary, and operational concept for information is it possible to label traditional theories as monocontextualist. Thus, those earlier theories are not supplanted but, rather, enveloped.

Multicontextualism offers a powerful synthesis, a blend of analytical philosophy and structuralism. Or, as a friend remarked, it unites Wittgenstein I and II.

Multicontextualism holds that information is still related to an object, but the scope of the object's relevant autonomy, as meant by particular information-as-reference, is now strictly limited to a particular context (Wittgenstein II: language game). At the same time, contextual identities may be joined to form a supracontextual object. In its supracontextual capacity, however, an object does not entertain any characteristics. These are only made specific within contexts. Its minimal supracontextual relationships provide for an object's fundamental unity while its behaviors are equally and fundamentally different for correspondingly different contexts.

In any structured information set, some information must by necessity be primitive—that is, its explicit meaning is not contained within the set but must be interpreted from without (or has been implicitly given, as with genetic inheritance). Other information may be included with references to such primitive information, or to references specified earlier. As such, it is a straightforward matter to recognize that information about objects, as the set evolves, constitute each other for the most part.

A succinct prescription for formalization of multicontextualism is presented in the first part of my book Metapattern: context & time in information models. The formalization's highly practical power is derived from the recursive application of fragmented information about contextual objects and relationships. Contexts, too, are formalized by such information, and resulting information models are both compact and flexible. The emphasis of design choices shifts to elaboration of contexts and supracontextual cohesion. It seems reasonable to expect that the full extra degree of freedom to (re)present knowledge supports a higher level of intersubjectivity; that is, of success in communication.

I originally developed multicontextualism as an approach to conceptual information modeling oriented at business and government processes. Traditional object orientation, I found, left too many problems unsolved, or without sufficiently elegant solutions. A much higher quality of such models, which include ease of communication, is a prerequisite for constructing superior (digital) information systems for all sorts of administrative purposes.

At an early stage, I recognized multicontextualism's potential for fundamental and applied philosophical inquiry. Other disciplines besides information science may also benefit, the essential criterion being that conceptual information models lie at their hearts. Examples that come readily to mind are cognitive and social psychology, linguistics and artificial or, rather, machine intelligence (expert systems). But, then, is not every science an attempt to model our world? Would it not help, also, to view light, behaving as either a wave or a particle, as a matter of context?

As in any mature discipline, the volume of philosophical work is overwhelming, making it impossible to ascertain multicontextualism's novelty. In fact, I learned from my readings that many thinkers over the course of centuries tried with various degrees of success to capture pluriformity. However, nowhere did I find a really comprehensive treatment and formalization which suited my own original purpose. Of course, when another author is shown to have made this contribution earlier I will gladly change my hypothesis, so boldly stated at the beginning of this short essay on multicontextualism. I will change it to the extent that the novelty is not mine, and give the other person(s) full acknowledgement for his (their) achievement. In fact, from research for my new book Semiosis & Sign Exchange, I have already learned that philosopher J. Dewey (1859-1952) places the concept of situation at the root of his logic as a general theory of inquiry. Then there is a so-called dialogical theory with M.M. Bakhtin and V.N. Voloshinov as groundbraking theorists. Recent examples of attempts at mathematical formalizations are publications by J. Barwise and K. Devlin.

Original or not, I will no doubt always strongly maintain that the approach to understanding along multiple contexts may be at the root of many fruit-bearing branches of philosophy—that is, of a great variety of applied disciplines—while providing opportunities for their advanced synthesis. As for myself, I will stay especially busy applying multicontextualism to conceptual information modeling for very practical results.

This text was included in Metapattern: context and time in information models (Addison-Wesley, 2001, see Appendix A).