9. The Architecture of Knowledge

It is a thesis of this work that the advancement of information technology renders choice of analytic method, not only in philosophy but wherever analysis might prove useful, dependent upon the software available to support the method, and that the architectural design of such software should take place in the context of an explicit (if generic or pluralistic) conception of analytic method.

This is an interdependency which may usefully be considered at the very earliest stages and at the highest and most abstract levels in the development of method and of information or knowledge architecture. The interdependency is such that we may be tempted to identify a certain kind of architectural design with a certain kind of fundamental philosophy or meta-philosophy.

In this chapter I undertake an analysis of knowledge architectures based the ideas about knowledge which have been so far presented. It is not desirable that an architectural discussion in a book of philosophy enter into much specific detail, so the aim here will be the analysis of certain ideas about the structure of knowledge as represented in information systems and the interaction between such conceptions of structure with the kinds of functionality which the information systems might then support, the methods which they facilitate, and the directions of future development to which they are sympathetic.

Rather than attempting wholly to effect the kind of integration between architectural design and constructive philosophical analysis, I will present these as two different perspectives upon a single enterprise, in this chapter the architectural design, in the next the philosophical perspective.

In this chapter the discussion will fall into two parts. An architecture is an abstract high level description. The analysis or evaluation of an architecture, must be undertaken against some prior conception of the aims which the architecture is intended to realize. In engineering terms these are high-level requirements. In philosophical terms, these requirements correspond to a delineation of the problem domain.

The first level at which positive philosophy departs from being purely analytic is in the choice of subject matter. A substantive statement about some practical matter, perhaps in politics or economics, may be clothed in a pure analysis, implicit in the choice of system to be studied. To promulgate ideas about how society might be organized, it would suffice to proceed by analysis, considering a class of realizations of the ideas and examining their relative merits.

This is the manner in which I proceed here. My interest is in certain approaches to the development of knowledge as a collaborative enterprise (which usually is) making effective use of our developing information and network capabilities. I begin the architectural discussion by setting out the domain of enquiry as a statement of requirements, and then proceed to consider and compare some of the ways in which those requirements might be met.

These two stages, statement of requirements, response to requirements, will not be monolithic. The requirements will be stated little by little. To each stage architectural responses are considered, and the requirements may then be augmented in the light of the analysis.

9.1 Requirements from Leibniz and Carnap

I have already identified the projects of Leibniz and Carnap as points of departure, so I begin with some key features of those projects.

The principal elements of Leibniz's project were:

Carnap's project was more narrowly scoped, but shared the first two items recast in pluralistic terms.

Here we adopt all three, adjusting the statement in the direction of pluralism, and thinking of information technology.

Thus we are interested here in:

A system for the representation of propositional knowledge.
A substantial online knowledge base represented in that system.
Software supporting the further development and application of the knowledge.

Carnap attached considerable importance to the distinction between analytic and synthetic propositions. Though he did not acknowledge the influence of Hume, he agrees with Hume in characterizing essentially the same dichotomy, Hume's fork, in three distinct ways. Carnap sought to adapt methods similar to those of Frege and Russell in the formal derivation of analytic propositions to languages in which synthetic propositions could be expressed and used in formal derivations.

Carnap's approach to the meta-theory of such empirical languages is not from our point of view entirely satisfactory. The approach envisaged here to the connection of our languages with the empirical world is entirely different. Whereas


The central idea of interest concerns the role of deduction, in the acquisition and application of knowledge, and the possible contribution of information technology in support of deduction.

Deduction is relevant in two general ways. The separation of deductive from other kinds of reason permits the deductions themselves to be checked more rigorously, and exposes the premises (which might otherwise have been obscured) on which the deductions depend. We are here concerned with a machine supported injection of formality into reasoning the intended effect of which is to improve precision, rigour and reliability.

Formality and deduction also potentially enable new kinds of functionality to be realized. This is because the search processes involved in finding proofs of logical conjectures can serve to discover witnesses for existential claims, and hence solutions to design problems. The ability to demonstrate compliance of such a solution underpins and sanitizes the application of exotic and possibly unreliable methods during the search for a solution, hence allowing solutions to be discovered which might never be found by less exotic algorithms.

9.2 Epistemic Retreat

The positivistic idea of epistemic retreat influence the requirement in various ways.

Firstly we distinguish between analytic and synthetic judgements, and between formal and informal claims. The system is primarily concerned with the formal side.

As in Hume, we accept analytic propositions as exhausting those which can be known with certainty. In fact we go one further, taking empirical claims to be at best approximations, best thought of and formally represented as models of aspects of reality. As such they should be assessed or affirmed not as true or false but in more complex and informative terms. Thus, what we affirm of a theory is not its truth but its applicability to certain aspects of reality and the accuracy and reliability with which it models those aspects under various circumstances.

Formally, we do not assert an empirical claim in connection with such a model, we instead formulate the theory as an abstract model, and on the basis of this definition we can then undertake theoretical elaboration of the theory, draw consequences in relation to its application in hypothetical situations, and formally evaluate the theory against experimental data presented in the terms of the abstract model.

Roger Bishop Jones 2012-09-23