Upgrade Document for “Designing Process Improvement into Management Systems”
Table of Contents
1Introduction
2Theory
2.1Two types of knowledge
2.2Knowledge Management in an Actor-Network Theory perspective
2.3Knowledge Management in a Total Quality Management perspective
2.3.1Knowledge generation
2.3.2Knowledge codification
2.3.3Knowledge transfer
2.4Organizational theory
2.5Cultural aspects
3Method
3.1Interviews and observations
3.2Document analysis
4Case description and analysis
4.1Organizational analysis
4.2The NTAX software lifecycle model
4.2.1Knowledge generation
4.2.2Knowledge codification
4.2.3Mechanisms of knowledge transfer
4.3Main ANT-concepts in a KM perspective applied within the organisation
4.3.1Heterogeneity and negotiations
4.3.2Inscriptions
4.3.3Translation
5Discussion
5.1Two ways of understanding
5.2Knowledge about the current system
5.3Knowledge beyond the current system
6Conclusion
References
1Introduction
This document has been written as an early draft for an “upgrade document” required for all PhD students at the Information Systems group at the Institute for Informatics at the University of Oslo.
2Theory
Our intent with this chapter is to describe knowledge management within a wider philosophical context that includes a perspective on technical and social knowledge representing two distinct types of knowledge, suited for solving two different types of problems. We continue by elaborating on what we see to be the characteristics of knowledge management within each of these two domains, and then conclude the theory chapter with one section on organizational aspects and one section on cultural aspects.
2.1Two types of knowledge
Kuhn (1962) argues that scientific knowledge evolves in two directions. The normal way of accumulating knowledge is by solving puzzles according to rules defined by the current paradigm of doing science within a given discipline. Every now and then, however, this paradigm may be challenged, and new ways of understanding the world may evolve.
The way Kuhn describes the process of doing scientific research is strikingly similar to how Argyris and Schön (1978) suggests how to create organizational learning.
Figure 1 – Double loop learning (Argyris & Schön, 1978)
An organization may learn through the methods of quality management, i.e. identifying errors and opportunities for improvement, and work out ways to improve the system based on such insights, as would correspond to “single loop learning”. However, from time to time it may strike the organization that the whole system should have been designed in a completely different manner, so by challenging the current assumptions and beliefs, the variables governing the single loop learning may be adjusted. This external perspective on the system is referred to as “double loop learning”.
Jashapara (2004: 135) suggests total quality management (TQM) and business process engineering (BPR) as a possible way for understanding or implementing double loop learning. In the appendix B of ISO 9004:2000, it is suggested that double loop learning should be a natural way of operating any ISO 9001:2000 certifiable quality management system, where the inner loop learning is handled by methods for creating continuous improvements (“kaizen”; Imai, 1986) while the outer loop is handled by “breakthrough management” (Juran, 1964).
What seems to us to be not all that clearly stated in the quality management literature, and knowledge management literature, is the way the inner loop and outer loop of double loop learning seem to correspond to two different ways of understanding the world. In our understanding, the difference seems to correspond to the “two cultures” of natural science and social science (Snow, 1964), meaning that the two cultures deal with two different concepts of knowledge, corresponding to whether the purpose of the research is to predict and control nature or whether it is to “understand” a culture from an anthropological point of view, i.e. to understand the language of the tribe in terms of observing what they tend to do.
Kuhn explains his gradual understanding of scientific discovery to start with the understanding of hermeneutics (Lee, 1991). A similar approach may be used for explaining Deming’s image of the organization as a learning system. In fact, Scherkenbach (1986: 35) uses a model similar to figure one to illustrate the Deming philosophy of organizational learning by illustrating process improvement within the current system as the inner loop feedback mechanism between supplier and producer, following the usual methods of statistical process control, while the outer loop feedback mechanism corresponds to consumer research.
In this presentation we consequently try to distinguish between knowledge processes related to the social science of the outer loop, using Actor-Network Theory (Monteiro, 2000; Latour, 1987) as a possible framework, while applying the traditional methods of statistical quality control (Deming, 1992) for analyzing the knowledge processes within the inner loop.
2.2Knowledge Management in an Actor-Network Theory perspective
Within the science and technology field it is of importance to be able to formulate and understand information, innovation and knowledge management processes, and hence ANT was created; in order to better understand not only what kind of knowledge that exists and gets distributed in an organization, but also how, and by that exploring how a knowledge network is created (Monteiro, 2000). ANT is, although it is called a theory, more of a material/knowledge-semiotic method that gives us a powerful linguistic tool to describe KM processes and relations.
Rather than looking at knowledge as something that is contained by someone or something, it is possible or even feasible to recognize knowledge as a network, or network of contextualized data and information, and that this network may consist of both knowledgeable and may be not so knowledgeable humans, non-humans like i.e. information processing and storing kinds of software, a web of experience, an urge to tell and a yearning to know. In all, a suitable network for the creation and transformation of knowledge from tacit to explicit and back, in addition to tacit to tacit and explicit to explicit. In an environment of learning in multiple directions and contexts, we form networks of knowledge, where we teach and learn, show and tell, exchanging a mix of tacit and explicit knowledge in a both structured and unstructured manner, with all the influencing factors that are comprised of earlier experience, education, familiarity with the tools needed to understand and manage a task, relations to others and so on. We might say that the tools for facilitating KM to a certain extent resembles Information Infrastructures as it is described by Hanseth and Braa (2000) and Monteiro (2000) as open networks, linked to others networks indefinitely, and as such, not necessarily easy to control.
It may help, though, to look at infrastructure as Dahlbom does, as a “regulating skeleton, providing framework and guidelines for the activity”, and even if the immaterial nature of knowledge makes it seemingly incompatible in relation to a physical infrastructure, the more important aspects of an infrastructure is in fact immaterial; agreements, standards and metrics. An important element in Information Infrastructures is the idea of gateways, which is a method for combining two, or more, initially incompatible actors (standards). In an ANT perspective this could be regarded as translation, and in KM as a tool for knowledge distribution. According to Dahlbom, the stable infrastructures of the information society would be its educational institutions, research organisations, legal systems, habits and so on, as a shared resource. (Dahlbom, 2000:217-220), and by sharing, knowledge is created and distributed.
2.3Knowledge Management in a Total Quality Management perspective
Although the best definition of Total Quality Management (TQM), in a European context, may be the evaluation criteria for the annual quality awards (EFQM, 2006), when discussion the knowledge management principles underlying the ideas of quality management, we chose to focus on the ideas put forward by Shewhart and Deming.
A good starting point for discussing KM within the context of TQM is perhaps by looking at the ISO 9000 model for designing quality management systems (figure 2). Although the figure appears to consists of a single loop, the system model can be thought of to consist of two loops or two types of logic as illustrated by the box indicating “measurements, analysis and improvement” having input from processes and products on one hand (single loop feedback) and input from customers on the other hand (double loop feedback).
In fact, if we identify the customer box on the left hand side of the diagram with the customer box on the right hand side of the diagram, the structure would topologically be possible to be described as a torus, i.e. a topology of double loop learning that is identical to the Argyris model in figure one or the Kuhn model of how science evolves.
Figure 2 – The NS-ISO 9000:2000 quality management framework
As we have already presented Statistical Process Control (SPC) as a tool for knowledge management as a previous INF5190 presentation (Øgland, Bakke, Murad & Bjørnseth, 2006), in this presentation we will just summarize some of the main insights from the presentation.
2.3.1Knowledge generation
According to Davenport and Prusak (1998: 52), knowledge generation refers to: “…, the specific activities and initiatives firms undertake to increase their stock of corporate knowledge”. They characterize the importance of knowledge generation in this way: “…since it is axiomatic that a firm’s greatest asset is its knowledge, then the firm that fails to generate new knowledge will probably cease to exist.”(ibid.: 67).
Acquisitionof knowledge, as used by Davenport and Prusak (ibid), refers to the dimension of knowledge generation where the knowledge is acquired by the organizations as well as that developed within it. In chapter six of ISO 9001:2000 there are requirements related to people having sufficient competence for performing tasks, and that the competence shall be documented.
The main process for generating knowledge with the ISO 9001:2000 model is by recording problems and areas for improvement. Although there are requirements all over the model aiding to define a consistent model that will make learning and improvement possible, the main issues are related to chapter eight in the ISO model where there focus is on applying (knowledge) metrics.
2.3.2Knowledge codification
Knowledge codification involves codified material such as texts and computer systems that organize and contain knowledge of an organization. The purpose of knowledge codification is: “…to put organizational knowledge into a form that makes it accessible to those who need it” (Davenport and Prusak 1998: 68).
Making a quality system in compliance with the ISO 9000 requirements and recommendations would be a typical example of a system for codifying, storing and managing knowledge for the purpose of improving organizational performance. As the focus of the system is on quality (conformance to standards), the process of codifications runs mostly through the processes of quality control.
In order to perform quality control of processes and products, the standard way of doing this would be by identifying metrics for all critical aspects of the organization and use statistical process control (SPC) for monitoring whether the processes are stable and under control (Deming, 1992).
In other words, codification starts by identifying what may go wrong or what we would like to improve, and then select metrics that are useful from this perspective. In chapter four of ISO 9001:2000 there are also requirements dealing with how to create a map to illustrate how the processes connect.
This may be useful for understanding how processes are related when we want to understand the root causes for something going wrong or how to predict the effects of improving related processes. What seems to correspond the most directly with the concept of knowledge maps (or topic maps) in knowledge management, however, is the quality management idea of using Ishikawa fishbone diagrams for mapping out the ontology of how things can go wrong.
The idea behind the Ishikawa fishbone diagram is that of indicating the five different “topics” that have proved useful for analyzing problems in industry; man, machine, materials, methods and metrics, catalogue problems as they occur (“occurrence”) and draw lines in order to make ontological “association”. In other words, the Ishikawa diagram, developed in the 1960s, should probably be seen as a prototype for what is now described as “topic maps”.
The main difference between the Ishikawa diagrams and the more general topic maps, on the other hand, is that the Ishikawa diagrams relates to the ontology of problems and errors while general topic maps can be used for organizing explicit knowledge in general.
2.3.3Knowledge transfer
Davenport and Prusak (1998:101) state that knowledge transfer involves two actions; transmission and absorption. The goal of knowledge transfer is to improve an organization’s ability to work, so these two actions only provide value if they lead to change in behavior. However, as TQM in the context of ISO 9000 relates to the systematic aspects of knowledge transmission and absorption, we have added sections later in this theoretical section on organization theory and cultural theory.
Nonaka (1995:19) provides a model with four different modes of knowledge conversion with the basis in the terms tacit and explicit knowledge as the two main types of human knowledge.
Table3 – Nonaka’s modelof knowledge conversion
As one of the fundamental ideas in “The Knowledge-Creating Company” (Nonaka & Tackeuchi, 1995) is to explain innovation and develop processes in Japanese companies as compared to those of the West, focusing on the “tacit” culture of the East, his framework seem suitable for explaining why Quality Circles has been and still is an important aspect in Japan while it only lasted as a short-term fad in the 1980s in the West.
From the point of view of ISO 9001:2000, the idea is to have all management knowledge explicit in terms of probability distributions for making decisions, and by the process of doing so, knowledge can be codified as information to be transferred according to the traditional way of mathematical communication theory (Weaver & Shannon, 1949).
Table4 –Mathematical theory of communication (Weaver & Shannon, 1949)
The significance of going from knowledge management theory to information management theory when dealing with knowledge transfer of quality related issues is that quality management can be dealt with in objectivist ontology. By this approach, many of the problems of knowledge management are simplified to be problems of information management.
Nevertheless, in order to deal with issues that can’t be explained by numbers, tables and graphs, we need additional theory, such as the organizational theory and the theory of cultures.
2.4Organizational theory
Mintzberg (1983) definesnine organizational design parameters, categorized in four main groups. We have found three of these useful for the purpose of understanding the aspects of knowledge management needs in the type of organization we are looking at. The groups are:
- Design of individual positions – the design parameters in this group covers job specializing, formalizing or work and training and indoctrination.
- Design of “super-structure” – this covers the grouping of and the size of each unit.
- Design of communication between units covers fields like system for planning and controlling, and also coordinating communication.
- The group called “system of decisions” covers two parameters called horizontal and vertical decentralization. This has to do with where, or at what level decisions are made. It is about power structures.
These four groups are then used for analyzing the what Mintzberg see as the five different configurations:
- Simple structure – suitable for a “one-man firm” and such.
- Machine bureaucracy – typically for the industrial workplace (e.g. a factory).
- Professional bureaucracy – organizations based on professional skills.
- Organizations in divisions – an incompletely configuration but with clear terms on single fields, suitable for large and complex companies.
- Adhocracies – the innovative organization.
In order to determine a configuration, many factors have to be considered, such as situational factors. Mintzberg split these factors into categories such as; age and size, technical system, environment and power. In our KM analysis, however, we only plan to use some of the most basic aspects of Mintzberg ideas.
2.5Cultural aspects
By identifying the structural configuration of a given company, one would perhaps expect that much of organization culture is also explained as one would expect culture to evolve and adapt to fit the structure. However, in a constantly globalizing society additional cultural factors may be taken into account for understanding knowledge flow and development. In particular, we would like to see how the organization culture manifests itself or evolves in order to meet the hetrogenity among “customer” (tax payers).
Although globalization and hetrogenity may play a vital role in international business, for the processes we investigate within the tax administration, we do not expect it to play a vital part for understanding and improving the software lifecycle processes, but we would nevertheless like to add a perspective on the cultural issues as managing culture appears to be one of the most striking features of KM vis a vis TQM, e.g. through the emphasis on the interplay between tacit and explicit knowledge.
3Method
Our approach for collecting knowledge and understanding NTAX has consisted of interviews, observationsand document analysis.
3.1Interviews and observations
As one of the project members had been working as a quality manager within the IT function of NTAX and was still a part of the organization, we considered his knowledge from previous interviews and observations to be sufficient for the level of understanding that was needed for performing the analysis that was needed within the scope of the project. His previous interviews, observations and document controls had been following the framework of Action Research (Coghlan & Brannick, 2001), meaning that many of the interviews, observations and actions were done from the perspective of a change agent and thus biased in order to motivate organizational learning and improvement.
During the course of the project, however, he interviewed the current IT quality manager and made some questions to the head of the IT department and the some of the people working in software development and IT security issues.
3.2Document analysis
Although there are some differences between how work is carried out within the IT function of NTAX and how the documents explain how work is supposed to be carried out, especially if we look at various plans, procedures and strategy documents, there are also documents identifying some of these differences in order to aid the organization in reducing the gaps.
In our analysis, we have partly chosen to look at the organization from the ideal perspective of what it says it is doing and partly from our understanding of what is actually going on, i.e. by looking at quality management documents. We believe this dual perspective should correspond with our perspective on dual knowledge in section 2.1, i.e. how one type of knowledge deals with direction and narrative while another type of knowledge deals with decisions and control.