In order to clarify the implications of the previous sections for some integrated approach to human and social development, it is appropriate to consider the current status of cognitive systemization. This has been the concern of Nicholas Rescher who explores the reason for systematization in the cognitive domain and shows how this is one of the crucial features of the development of knowledge (91). It is to be expected that the pattern of insights and conclusions would be relevant to development in general.

Rescher identifies eleven definitive characteristics of systematicity: wholeness, completeness, self-sufficiency, cohesiveness, consonance, architectonic structure, functional unity, functional regularity, functional simplicity, mutual supportinveness, and functional efficacy (91, p.10)). He points out, citing C S Peirce, that the need for understanding through a unified view of things is as real as any of man's physical cravings, and more powerful than many of them. The above characteristics "are constitutive components of that systemacity through which alone understanding can be achieved". (91, p.29) The point of cognitive systematization in reational terms is that (a) it is the prime vehicle for understanding by making claims intelligible, (b) it authenticates the adequacy of the organization of knowledge, (c) it is a vehicle of cognitive quality control, providing a test of acceptability, and (d) it provides the definitive constituting criterion of knowledge (91, p.29-38). Similar points could be usefully made about the integration of development.

The alternative modes of cognitive systematization are distinguished by Rescher. These are foundationalism, based on a "Euclidean model of a linear, deductive exfoliation from basic axiones "and coherentism, a network model of cyclic systematization of interrelated theses (91, p.39). The Euclidean model is typical of the logic governing formalized (intergovernmental) development programmes based on a set of principles. The network model is typical of the logic of "grass-roots development movements. From the network perspective, the Euclidean model imposes a drastic limitation by "inflating what is at most a local feature of derivation from the underived (i.e. locally underived) into a global feature that endows the whole system with an axiomatic structure". Thus although "a network system gives up Euclideanism at the global level of its over-all structure, it may still exhibit a locally Euclidean aspect, having local neighbourhoods whose systematic structure is deductive/axiomatic" (91, p.44-45)

The network model shifts the perspective, as Maruyama also notes, from unidirectional dependency to reciprocal interconnection, abandoning the concept of priority or fundamentality in its arrangement of these. "It replaces such fundamentally by a conception of enmeshment in a unifying web" (91, p.46-47), whereas the Euclidean approach gives priority to derivation from what is better understood or more fundamental.

Rescher notes (91, p.58-59) basic weakness in the latter approach was however demonstrated by Kurt Goedel (92), who showed both that the consistency of any formal axionatic system can never be proved, and that the deductive axiomatization of any such system was inherently incomplete. There are therefore always "true" statements in a given domain that cannot be derived from the chosen axioms. It would seem that this too has important implications for the limitations of development programmes elaborated on the basis of pre-determined sets of principles in some "declaration" or "world plan of action", especially since Rescher indicates the possibility of a breakdown of deductivism in the factual sciences as well (91, p.176).

Rescher also provides a valuable analysis of the limits to cognitive systematization. He identifies three possibilities: incompletability, inconsequence (or disconnectedness, compartrnentalization), and inconsistency (or incoherence). With regard to the first, he notes that it is unrealistic to expect either attainment of a completed and final state of factual knowledge, or a condition in which all questions are answered. "Accordingly, we have little alternative but to take the humbling view that the incompleteness of our information entails its incorrectness, as well" (91, p.152-3). In a more highly developed future, fundamental errors will be perceived in present formulations and programmes - as we can already detect in the development strategies of past decades.

With regard to disconnectedness, the second possibility, Rescher argues that this cannot characterize the body of our factual knowledge as a whole which can always be joined by mediating connections of common relevancy (91, p.164). The problem is rather that despite such causal linkage, there could well fail to be connections in meaning between two domains. The fundamental causal matrix in which all natural occurrences are bound together "might merely be a purely formal unity, lacking any sufficient substantive basis of functional connectedness." Nature might come to be shown as operating in an essentially compartmentalized manner. Furthermore, Rescher notes, gaps in the knowledge attainable at any time might in practice block realization of any underlying interconnectedness. This issue of compartrnentalization is of course of crucial importance in the design of interdisciplinary development programmes, for which no adequate methodology has yet emerged, partly because of separative behaviour characteristic of disciplines.

With regard to the third possibility, Rescher sees inconsistency as lying at the root of the urge to systematicity. It is the very drive toward completeness that enjoins the toleration of inconsistency upon us. But rather than implying no system at all, any inconsistency-embracing world picture involves the toleration of ungainly systems of deficient systemacity (91, p.176-7). It is a question of degree.