Implicit, incoherent and inconsistent knowledge.(News & Views)

Abstract

Human knowledge is always impaired by four characteristic deficiencies: incomplete, implicit, incoherent and inconsistent knowledge. The three features named in the title are analysed here in some detail. Specific conditions for the formation and validity of implicit knowledge are identified. In the case of incoherent and inconsistent knowledge, constructive ways of handling are possible. The chance for novel insight, emerging from a struggle with incoherent material, is demonstrated by a historical example. The material presented here is relevant to everyday human reasoning, but particularly to reasoning under aggravating conditions, e.g., in issues, which are controversial from the beginning, or where a closed theory or a generally-accepted decision procedure does not (yet) exist.

1. The True Character of our Knowledge: Implicit, Incomplete, Incoherent, Inconsistent

All our knowledge is patchwork. Nevertheless, we can--and should--try to find out whether there is a chance to classify the main troublemaking properties. The ensuing discussion will be based on the following four characteristic deficiencies:

1. Incomplete knowledge.

2. Implicit knowledge

3. Incoherent knowledge

4. Inconsistent knowledge

In view of the abundant literature on incomplete knowledge, the latter three categories will be discussed in detail--for the moment a natural understanding will do.

However dubious our knowledge may be, we are inevitably forced to act and to make decisions, both in our professions and in everyday life. It is plausible that, even since prehistoric times, humans are somehow prepared to cope with these unwelcome characteristics of information, and at first glance one might be inclined to think that there is no necessity of a deeper scientific analysis. But anything that works easily in "normal" situations may bring about complications under aggravating conditions, such as:

1. An unusual load caused by enormous masses of data or their properties (heterogeneous provenience, disparate credibility, varying styles of representation, etc.).

2. Extraordinary importance of an evaluation or decision.

3. Adverse side-conditions of a decision process (e.g., decision under stress, time pressure or distraction).

Therefore, a deeper investigation is justified, e.g., under the aspects of decision-making, cognitive science and pedagogy; maybe even an unexpected ramification into psychopathology will emerge.

Some specific techniques have been developed in special fields of application:

1. In business data processing, modern methods of "data warehousing" and "data mining" make use of supercomputers in order to manage enormous masses of data, to retrieve relevant entries, and to derive conclusions.

2. In artificial intelligence, specific techniques have been designed for handling incomplete and inconsistent information, as well as for the use of episodic material in "case-based reasoning."

3. In any kind of management tasks (in the broadest sense of the word), we are facing material with disparate reliability, precision, and validity, from internal and external sources, from historic or recent origin, with short-term and long-term relevance, and often with an insufficient form of aggregation and representation.

4. A similar statement holds for "governmental intelligence," which is defined in one of the rare publications as "the product resulting…