π Lecture 41 β Knowledge Sharing Structure and Services, Culture and Knowledge Communities
π Overview: This lecture explores the framework for analyzing and comparing knowledge management tools through the Knowledge Structure and Services (KSS) Matrix and Checklist. It explains how different levels of knowledge formalization and types of knowledge services can be used to classify, evaluate, and select appropriate electronic tools for managing organizational knowledge, which is critical for turning information into actionable knowledge in the Information Age.
ποΈ Topics Covered
The lecture covers electronic tools for knowledge management, the two key dimensions of the KSS framework (knowledge structure and knowledge services), the hierarchy of knowledge forms from implicit to formal knowledge, the three service layers (infrastructure, core, and packaged services), the KSS Matrix and Checklist for tool evaluation, and a comparative analysis of five leading knowledge management tools including Documentum 4i, OpenText LiveLink, Autonomy Knowledge Server, Lotus R5, and PeopleSoft CRM.
π Lecture Summary
Electronic Tools for Knowledge Management
Identifying, nurturing, and harvesting knowledge is a principal concern in the Information Age. Effective use of knowledge-facilitating tools and techniques is critical, and a number of computational tools have been developed. However, analyzing or comparing specific tools remains difficult because knowledge management is a young discipline evolving rapidly as more people encounter new problems. New technologies support previously impossible applications, and the multidisciplinary character of knowledge management combines business, management, computer science, and cybernetics. The field is frequently defined so broadly that anything can be incorporated.
KSS: Knowledge Structure and Services
Two dimensions are central to analyzing and comparing knowledge management tools: knowledge structure and knowledge service. These two dimensions can be used to form a matrix in which specific knowledge management tools are positioned.
Knowledge Structure
There is a wide range of levels of formalization or structure in the ways knowledge is represented in knowledge management systems. The knowledge forms are not discrete or exhaustive, and other levels could be added. Examples include:
- Creative knowledge is intrinsically nonformalizable and may not be representable in any formalization.
- Audio and video contain multiple "streams" of knowledge such as music, voices, faces, and objects. Humans recognize these features but creating machine recognition is extremely complex.
- A raw text document is the formal equivalent of an audio track and is comparable to natural language that is also difficult for machines to understand.
From top to bottom we increase the formalization and precision of knowledge, while from bottom to top we accommodate more informality and ambiguity. Knowledge forms toward the bottom increasingly demand knowledge engineering and incremental analysis.
- In contrast, an HTML document with markup tags can display the text's structure. Irregularities in the structure can aid in interpreting the content. For example, "wrappers" convert structural marks into semantic descriptions.
- Structured documents using formats like XML or its ancestor, SGML, explicate the semantics implicit in HTML markups. For example, instead of deducing that tag
<H1>USA</H1>indicates "USA" is a country name, an XML document could contain<country name>USA</country name>that makes the text an explicit country name. - XML documents are linear representations of "tuples" of data, the essence of information stored in databases. For example, a sequence of tags can contain a
<population>tag inside a<country>tag to indicate a relationship, facilitating efficient storage and retrieval. - Categorized information is at roughly the same level as structured information in databases. Taxonomies such as those used in biology are examples, used extensively by directory sites like Yahoo! to provide taxonomies of concepts.
- Formal knowledge is used in the mathematical sense. Logical statements such as theorems and equations are used rigorously to make sure all semantics are explicit and rules are followed, making it easy for machines to interpret.
The level of structure in the knowledge directly affects the amount of automated processing that can be performed because more structured knowledge employs powerful semantics. Managing highly unstructured knowledge requires more structured descriptions of the content, just as video indexing employs close-captioned text and HTML pages are indexed by metatags. Most knowledge found on the Web falls near the top of the scale, and most knowledge management tools concentrate on this range. Semantics and interpretation of less structured forms of knowledge depend on contextual knowledge. Raw text files representing a speech eliminate many possible ambiguities in speech recognition, using contextual knowledge about the subject, person, and voice to "reduce" less structured forms to more structured forms.
π Definition β Taxonomy: A system for categorizing and classifying concepts, ideas, or subjects, similar to biological classification systems, used extensively by directory sites.
Knowledge Services
Another useful dimension is the range of services knowledge management tools provide. By services, we mean tasks or activities in handling knowledge that can be at least partially automated. Knowledge services are divided into three main types: infrastructure services, core services, and packaged services that build on one another such that packaged services make use of core services, which employ infrastructure services.
Infrastructure Services
Infrastructure services are usually needed to implement any knowledge management solution. Five basic types include:
- Communication services enable electronic communication between users through e-mail, file transfer, chats, and similar vehicles.
- Collaboration services allow groups of people to communicate through online meetings, shared whiteboards, and discussion groups, as well as directory services. Built upon communication services, these tools are also known as groupware, and the best-known example is Lotus Notes.
- Translation services transform knowledge from one file format to another or from one language to another.
- Workflow management services define workflows and support online execution and control of workflows, allowing users to execute and enter results of subtasks and view status of other subtasks.
- Intranets and extranets include other infrastructure services. Intranets are Web-based applications restricted to specific organizations while extranets connect several organizations by providing access from one organization to another's content and services.
- Intelligent agents are software components capable of accomplishing tasks on behalf of a user, going beyond "information on demand" to make selected decisions based on predetermined environmental scanning methods. They can also summarize relevant data by aggregating and performing synthesizing functions before presenting to executive decision makers.
π Definition β Groupware: Software tools that enable groups of people to communicate and collaborate through online meetings, shared whiteboards, and discussion groups, with Lotus Notes being the best-known example.
Core Services
Core services define knowledge management solutions because they explicitly and directly access knowledge repositories. They are built around core processes of creating, organizing, and using a knowledge repository. Different core processes involve people or systems with different roles, including knowledge producer (creates knowledge), knowledge holder (learns from other sources), knowledge organizer (works like a librarian to allow orderly addition for retrieval), and knowledge user (consumes knowledge to execute tasks).
Key features of the five core services include:
- Knowledge generation services produce knowledge in forms that can be stored in the repository. Used by knowledge producers, these tools distill, refine, or create new knowledge, typically involving automated learning including data mining techniques and pattern recognition.
- Knowledge capture services facilitate addition to repositories, allowing users to enter new documents and may employ meta-information for indexing purposes. A simple example is the "document properties" mechanism of Microsoft Word containing author, revision number, subject, and date.
- Knowledge organization (indexing) services help knowledge managers arrange items in a repository to facilitate retrieval and use, adding to or modifying knowledge about repository indexes, taxonomies, and directories.
- Access management services determine who can access elements of the repository based on a directory of users, restricting access by permission levels.
- Retrieval services include searching and navigating functions as well as translation, visualization, and integration, creating value by making knowledge available for specific uses and providing personalization and configuration services.
Packaged Services
Packaged services aggregate lower-level services to solve specific types of problems. Much knowledge management literature concentrates on these because they are clearly connected to end-user needs. Three classes are:
- Customer Relationship Management (CRM) services provide information about a company's clients in an integrated way, allowing internal channels to share and add to the same central knowledge base. Siebel and PeopleSoft are leading providers.
- Business Intelligence services manage knowledge about competitors and partners, usually aggregating and providing unified interfaces to information from news agencies, public and private databases, economics and social information, and the World Wide Web, while filtering and classifying information into categories.
- Enterprise Information Portals (EIP) are specialized gateways providing access to internal and external sources of knowledge, offering one-stop access with typical examples including search engines and My Yahoo!.
The Knowledge Structure and Services (KSS) Matrix and the KSS Checklist
Different tools provide distinct arrays of services and manage specific types of knowledge. Two diagrams display relationships between knowledge management tools based on types of knowledge they handle and services they offer: the KSS Matrix and the KSS Checklist.
The KSS Matrix ensures that the types of knowledge handled are intimately connected with the core services provided. The horizontal axis recognizes the five core knowledge services while the vertical axis displays the eight basic levels of knowledge structure dimension. One KSS Matrix is used for each tool analyzed. Cells are filled with squaresβa large square denotes a major offering with comprehensive features, while a small square marks a service offered in restricted scope or functionality.
The KSS Checklist recognizes services beyond core services. It lists the five infrastructure services and the three packaged services, with squares indicating whether a service is provided. As with the matrix, square size represents scope of service offered.
The KSS Matrix and Checklist provide quick assessments of each tool, can be used to compare tools quickly, and more importantly, to evaluate knowledge management tools. Filling in the diagrams forces users to explore and analyze tools in detail. The KSS framework can also select tools for specific uses, with specifications represented as "target" diagrams for ideal offerings.
Using the KSS Matrix and Checklist to Compare Current Knowledge Management Tools
Five leading knowledge management tools are analyzed: Documentum 4i, OpenText LiveLink, Autonomy Knowledge Server, Lotus R5, and PeopleSoft Customer Relationship Management.
Documentum 4i is an integrated software suite centered on document management, with core strengths in dealing with documents. It supports audio/video and taxonomy, with some coverage of categorized information. It does not support knowledge generation services or formal knowledge, structured information, or implicit knowledge. It is intended to be an Enterprise Information Portal tool and supports workflow management and collaboration.
OpenText LiveLink is an integrated software suite focused on document management, providing a core set of services to handle document management and structured information from databases. It is an Enterprise Information Portal tool that supports translation and collaboration, including discussion groups and group scheduling.
Autonomy Knowledge Server is a software suite for content management providing sophisticated services for classifying material based on document content. It is unique in covering formal knowledge, and its use of learning algorithms facilitates some knowledge generation services. It has some translation and collaboration services but does not incorporate workflow management.
Lotus Notes R5 is virtually synonymous with groupware, handling only unstructured types of knowledge. It supports implicit knowledge through detailed descriptions of people's information and skills. Lotus R5 does not attempt to be a packaged service as defined here because it focuses exclusively on collaboration.
PeopleSoft Customer Relationship Management is typical vertical solution software specializing in CRM. It supports structured information but only handles information about customers, making it a good choice for CRM but a poor choice for general knowledge management problems.
Conclusion
Modern-day alchemy is about turning information into knowledge. The combination of knowledge management tools with databases and knowledge in the minds of employees is fostering knowledge groups, knowledge enterprises, and knowledge industries. The KSS framework provides a convenient way to characterize knowledge management tools by defining the types of knowledge they can handle and services they provide. The KSS Matrix and Checklist help visualize coverage of specific tools, providing a convenient way to quickly compare different tool offerings. They can evaluate specific needs and match them to services provided by available tools. Services are key elements in understanding knowledge management tools, but complete evaluations should include other aspects such as hardware, software, and budget constraints. Electronic tools provide necessary "horsepower" and number-crunching ability to deal with real-world complexity, assessing difficult problems and giving us simple answers, but we must exercise caution in demanding answers that are simple rather than simplistic.
β Key Takeaways
The KSS framework provides two critical dimensions for analyzing knowledge management toolsβknowledge structure (ranging from implicit knowledge to formal knowledge) and knowledge services (infrastructure, core, and packaged services). Knowledge structure directly determines automated processing capability; more structured knowledge with explicit semantics is easier for machines to interpret, while unstructured knowledge requires more contextual knowledge and knowledge engineering. Infrastructure services (communication, collaboration, translation, workflow management, intranets/extranets) form the foundation, core services (generation, capture, organization, access management, retrieval) directly access knowledge repositories, and packaged services (CRM, business intelligence, EIP) solve specific end-user problems. The KSS Matrix and Checklist enable systematic evaluation and comparison of tools by examining which knowledge types each tool handles and which services it offers. Understanding that different tools have different strengthsβDocumentum for document management, LiveLink for enterprise portals, Autonomy for content classification, Lotus for collaboration, and PeopleSoft for CRMβhelps organizations select appropriate tools for specific knowledge management needs.
π§ Quick Revision Questions
- What are the two key dimensions in the KSS framework for analyzing knowledge management tools, and how are they used to position specific tools?
- List the eight levels of knowledge structure from least to most formalized, and explain why more structured knowledge enables greater automated processing.
- What are the three main types of knowledge services, and how do they build upon one another in the service hierarchy?
- How does the KSS Matrix differ from the KSS Checklist, and what does the size of the square in each cell represent?
- For each of the five analyzed tools (Documentum 4i, OpenText LiveLink, Autonomy Knowledge Server, Lotus R5, PeopleSoft CRM), identify their primary strength and one key limitation according to the KSS framework.
π Lecture 42 β Knowledge Application Systems and CRM
π Overview: This lecture focuses on designing knowledge application systems to support business needs, using case-based reasoning as a primary methodology. It covers the system architecture, development processes, and real-world examples from industry leaders like Daimler Chrysler and General Electric, while also introducing the concept of knowledge communities as social learning systems.
ποΈ Topics Covered
The lecture begins by outlining a case-based knowledge application system design for software support teams, describing the Case Method Cycle and case development sub-processes. It then covers system architecture design with database, web-based, search, and reasoning technologies. Three industry examples are presented from Daimler Chrysler, Shuttleworth Business Systems, and General Electric. The lecture concludes with an examination of knowledge communities, their defining elements, and modes of belonging as described by Etienne Wenger.
π Lecture Summary
1. Design a knowledge application system to support your business needs. Describe the type of system and the foundation technologies that you would use to develop such system.
A knowledge-application system designed for business needs would be a case-based knowledge application for software support teams (up to 4 tiers) to address platform issues. Many organizations currently use tracking tools to control and track issues, storing ticket history with reporting features, but these tools lack mechanisms to aid in solution finding. The objective is to move beyond mere issue tracking to active solution support.
To design this case-based application, one applies the Case Method Cycle comprising six processes:
- System development process: Develop an application to store cases and retrieve them based on similar historic cases, showing resolution processes. The system must integrate with current tracking and business request tools.
- Case library development: Collect case library information from current case history repositories and develop maintenance mechanisms.
- System operation process: Define installation, implementation, and support following standard development processes.
- Database mining: Analyze collected case information to infer relationships between cases and define resolution processes.
- Management process: Ensure organizational support for the project.
- Knowledge transfer: Ensure users maintain and add cases to the case library.
Case development follows three sub-processes:
- Case collection: Determine cases for each supported application, aided by support personnel experience and historical information.
- Attribute-value extraction: Organize the support case library by identifying attributes or characteristics of each case. Determine relationships between cases for similar case searching. Consider interdependencies between different applications and map hierarchy into the database.
- Feedback: Ensure personnel maintaining the system receive necessary feedback to guarantee information quality, minimizing ticket resolution time.
π‘ Why this matters: The quality of cases and resolution processes directly impacts how quickly support tickets are closed, making feedback loops essential.
The foundation technologies include databases, Web-based technologies, search engines, and a case-based reasoning engine.
2. Design the system architecture for the system described above.
The system architecture integrates the case-based reasoning engine with a database containing the case library, web-based interfaces for user access, and search engines for retrieval. The architecture connects to existing tracking tools and business request systems to allow proper retrieval of updated information. This integrated design ensures that historical case data flows between the tracking system and the knowledge application system seamlessly.
3. Identify three recent examples in the literature of knowledge application systems.
Example 1: Daimler Chrysler Freightliner The vehicle manufacturing division chose a case-based approach for a diagnosing system. Diagnostic problems like trucks running hot or school buses vibrating were difficult to connect to particular solutions because vehicles are customized with different histories. Carlo Nardini, director of technical support, stated: "We determined that case-based reasoning gave us the chance to deal with things that were not so concrete." A model-based reasoning system might have worked for discrete problems like electrical wiring, but the multitude of indeterminate variables made case-based reasoning the most flexible option. Problems without finite descriptions or definitions demand this technology. Using a web-based interface, Freightliner offers its system to dealers, franchisees, and third-party operators like FedEx Corp., accepting legacy information from call centers and engineering groups.
Example 2: Shuttleworth Business Systems Facing the loss of two key support staff, joint Managing Director Andy King implemented a knowledge application system solution. His vision was to give customers the ability to answer their own questions by accessing a technical knowledge base. Shuttleworth implemented a help desk system first, gaining in-depth understanding of support team operations and workload. This allowed identification of problem causes originating in other departments like sales, development, or training, enabling elimination at the source.
Example 3: General Electric ICARUS System GE developed ICARUS, a Case-Based Reasoning System for Locomotive Diagnostics. The application analyzed fault logs from locomotives to predict and preempt serious faults. Previously, a rule-based system had been tried but was too hard to maintain, taking four person-years to reach a stage where it was deemed unworkable. The CBR system used a simple approach, costing approximately $250,000 and 14 person-months to build. The system solves 75% of known faults across 600 locomotives, with estimated savings of $5 million per year.
Knowledge Communities
Etienne Wenger (1998) defines communities of practice as social units of learning within larger systems forming constellations of interrelated communities of practice. The subtitle of his book β Learning, meaning and identity β shows his interest in community dimensions comprising identity, belonging, and boundaries. Wenger (2000: 229) explicitly states that communities of practice are the social containers of the competences that make up a social learning system.
Three elements define competence:
- The sense of joint enterprise: To be competent is to understand the enterprise well enough to contribute to it.
- Mutuality: To be competent is to be able to engage with the community and be trusted as a partner in interactions.
- A shared repertoire of communal resources: Language, routines, sensibilities, tools, stories, etc. To be competent is to have access to this repertoire and use it appropriately.
In a community of practice, knowing involves two components: the competence the community has established over time, and the subjective experience of the world as a member. Wenger distinguishes three modes of belonging according to different forms of participation:
- Engagement: Doing things together β the way we engage shapes our experience of who we are.
- Imagination: Constructing an image of ourselves and our communities β these images are essential for our sense of self and interpretation of participation in the social world.
- Alignment: Making sure local activities are sufficiently aligned with other processes to be effective beyond our own engagement. Alignment coordinates perspectives, interpretations, and actions.
π Definition β Communities of Practice: Social units of learning that serve as containers of competences within a social learning system, defined by joint enterprise, mutuality, and a shared repertoire of resources.
π Definition β Case-Based Reasoning (CBR): A problem-solving methodology that uses past cases to find solutions to new problems, particularly useful when problems have no finite descriptions or definitions.
π Example β Daimler Chrysler CBR System: A customized truck with a rough-ride problem involving a similar but non-matching transmission and engine type is connected to a previous instance using case-based reasoning, allowing technicians to find solutions without exact symptom-to-solution matches.
β Key Takeaways
The Case Method Cycle is the essential framework for building knowledge application systems, consisting of system development, case library development, system operation, database mining, management, and knowledge transfer processes. Case-based reasoning is most valuable when problems lack finite descriptions or definitions, as demonstrated by Daimler Chrysler's success with customized vehicles and GE's locomotive diagnostics system saving $5 million annually. Foundation technologies must include databases, web-based interfaces, search engines, and a reasoning engine, all integrated with existing tracking tools. Knowledge communities are social learning systems defined by joint enterprise, mutuality, and shared resources, with participation occurring through engagement, imagination, and alignment. The success of knowledge application systems depends critically on feedback mechanisms to maintain case quality and on knowledge transfer processes to ensure ongoing user contributions.
π§ Quick Revision Questions
- What are the six processes of the Case Method Cycle for designing a case-based knowledge application system?
- Why did Daimler Chrysler choose case-based reasoning over model-based reasoning for their Freightliner diagnostic system?
- What three elements define competence in a community of practice according to Wenger?
- What were the cost and savings associated with General Electric's ICARUS CBR system for locomotive diagnostics?
- What are the three modes of belonging in communities of practice, and how does each contribute to participation?
Here is the summary of Lecture 43, following the exact format provided.
π Lecture 43 β Role of Communities in Learning and Product Development
π Overview: This lecture explores the role of knowledge communities, particularly Communities of Practice (CoPs), in organizational learning and product development. It contrasts these communities with other group structures like teams and networks, explains how they are sustained, and details the critical role of virtual moderation in fostering knowledge exchange. The lecture underscores the shift from a management-driven approach to a supportive, ecology-based organizational environment for community success.
ποΈ Topics Covered
This lecture begins by defining and contrasting knowledge communities with teams and networks, highlighting their self-selecting membership and focus on applied knowledge. It then details the stages of community development and the critical role of virtual moderators in stimulating discussion and guiding knowledge development. The lecture concludes by outlining strategies for sustaining communities, including providing supportive facilities and a healthy organizational culture, and illustrates these concepts with the case study of the Knowledge Ecology Fair.
π Lecture Summary
Knowledge Communities
Knowledge communities are described as social structures that connect "islands of knowledge" into self-organizing, knowledge-sharing networks. They transcend disciplines, bringing different perspectives to exchange and apply knowledge. The lecture uses a table to distinguish communities from other groups. The main differences are that knowledge community membership is self-selecting rather than recruited, and the focus is on applied knowledge, with goals that are evolving and purposeful.
π Definition β Knowledge Community: A cohesive, self-selecting cluster of people within a diffuse knowledge network that focuses on exchanging, developing, and applying knowledge, often around a customer or problem focus. π Key Differences: The table shows that compared to a work group (focused on tasks) or a team (focused on output), a knowledge community is characterized by self-selection, fluid boundaries, and an evolving, purposeful focus on applied knowledge. π Example: The lecture contrasts a team, which is "recruited for team fit" and has explicit goals, with an in-house knowledge community, which is best managed "hands-off" by providing a supportive climate rather than direct orders. Etienne Wenger lists eight stages of community development: Latent, Coalescent, Active, Legitimized, Strategic, Transformational, in Diaspora, and Member able.
Virtual moderation for knowledge development
The role of a conference moderator is to stimulate virtual discussion and guide the community's knowledge development. A good moderator has enthusiasm and likes networking. They activate knowledge development through key activities like setting up conferences, defining the scope and ground rules, keeping conversations developing, and cross-linking different conversational threads.
π Definition β Conference Moderator: A person who stimulates virtual discussion and guides a knowledge community forward in its thinking, often performing this role as an important added daily activity. π Example: A good moderator's tasks include "summarizing" contributions, "managing inappropriate contributions" (often privately), and "engaging people in conversation," such as by acknowledging good contributions or seeking input from members with specific expertise. π‘ Why this matters: The moderatorβs role is not just administrative; it is central to sparking new ideas and momentum through cross-fertilization of different conversational threads.
Sustaining communities
Organizations encourage communities as an integral part of corporate knowledge programs to gain benefits like thought leadership, new product ideas, and increased knowledge capital. A potential threat is that formalization could stifle them. The watchword is endorsement, not enforcement, requiring a supportive organizational environment.
π Definition β Knowledge Ecology: An approach to knowledge management that emphasizes a supportive, natural, and self-organizing environment for communities, as opposed to a structured, top-down management emphasis. π Example: To minimize the risk of stifling communities, organizations should provide facilities for meeting (web space, physical places), offer facilitation to improve processes, provide connection information to help new members join, and encourage note-taking to share "knowledge nuggets." π‘ Why this matters: The whole ethos of a successful community is based on knowledge ecology rather than a rigid knowledge management emphasis. The Knowledge Ecology Fair (1998) is a key example, featuring virtual keynote presentations, workshops, discussion groups, and an "Open Space Circle" for participant-generated discussions. This event led to new initiatives like KEN (Knowledge Ecology Network).
β Key Takeaways
The most critical distinction is that knowledge communities are self-selecting, social structures focused on applied knowledge, and are best managed indirectly through endorsement rather than direct control. Their success depends on a high flow of communication and passionate, effective leaders who act as virtual moderators to stimulate and guide discussions. Organizations must actively foster a supportive "knowledge ecology" by providing facilities and a culture that encourages participation, while being careful not to over-formalize and stifle the community. As illustrated by the Knowledge Ecology Fair, these communities can be powerful drivers of innovation and thought leadership within an organization.
π§ Quick Revision Questions
- What are the key differences between a knowledge team and a knowledge community in terms of membership, focus, and goals?
- List four specific tasks a conference moderator performs to activate knowledge development within a virtual community.
- According to Etienne Wenger, what is the difference between a "Latent" community and a "Strategic" community?
- What is the recommended management style for an in-house knowledge community, and what is the single "watchword" for managing them?
- How does the concept of "Knowledge Ecology" differ from traditional "Knowledge Management" in the context of sustaining communities?
π Lecture 44 β Learning Organization
π Overview: This lecture explores the concept of learning organizations (LO) and organizational learning (OL), challenging traditional views of knowledge as a commodity to be transferred. It examines how the field of OL has been socially constructed and institutionalized, and critically analyzes the underlying assumptions and power dynamics within OL and knowledge management discourses.
ποΈ Topics Covered
The lecture begins by critiquing the conventional view of learning as knowledge delivery from a source to a passive recipient. It then introduces a social and situated view of learning where knowledge is created through participation and negotiation. The birth and institutionalization of the learning organization as a field is discussed, including the social process of manufacturing knowledge and creating new organizational identities. The distinction between descriptive (OL) and prescriptive (LO) research is examined. Finally, the lecture analyzes OL as a disciplinary discourse, critiquing its assumptions, and concludes with a discussion of knowledge management's reification of knowledge.
π Lecture Summary
Conventional wisdom
Our society is dominated by a view of learning as "knowledge delivery" where information flows from a knowledgeable source (teacher or textbook) to a learner lacking that information. Learning is seen as the acquisition of accumulated data, facts, and practical wisdom stored in books. This view equates learning to eating or banking: knowledge is food for the mind that the learner ingests. Teaching becomes the transfer of "gold" to pupils' heads. Learning is believed to occur mainly during early development through schooling, instruction, and training, with occasional supplementary training for job changes or updating knowledge.
However, this familiar conception is a highly reductive account of how people learn for at least two reasons. First, it suggests learning is separate from other activities, restricted to specific occasions like taking a class. In reality, learning is deeply rooted in everyday activities and experiences; most know-how distinguishing an expert from a novice is acquired day-to-day by acting and reflectingβthinking about what we do and why, and talking about it with others (SchΓΆn, 1983).
Second, viewing learning as a totally individual activity (like ingesting food) is misleading. People and groups create knowledge by negotiating the meaning of words, actions, situations, and material artefacts. They participate in a world that is socially and culturally structured and constantly reconstituted by all who belong to it. Knowledge is not what resides in a person's head, books, or data banks. "To know is to be capable of participating with the requisite competence in the complex web of relationships among people, material artefacts and activities" (Gherardi, 2001b). Learning is always a practical accomplishmentβ"knowledge is something people do together"βdone in every mundane activity.
The institutionalization of the field: the birth of the learning organization
The field of OL has developed as "problem driven"βproducing instrumental knowledge. However, this knowledge sets conditions for research to shift to "mystery-driven" learning. A paradigmatic episode illustrates this: at a conference, a colleague from a developing country tested whether successful firms were learning organizations but found low correlation between economic success and being a LO, leaving her puzzled.
This episode demonstrates the social process of "manufacturing knowledge": a heuristic concept (OL) acquires legitimacy in the scientific community, spreads through consultants and practitioners, produces distinguishing "characteristics," and coins the label "learning organization." This label travels through time and space, appropriated by organizations that incorporate it into their identities. A realist assumption replaces a heuristic device: learning becomes a "real" phenomenon "out there" that can be measured, compared, and validated. The touchstone for learning is the concept of changeβorganizational change is the outcome of a rational procedure of knowledge production and application.
This process of mobilizing credibility creates cultural artefacts (books, conferences, university courses) and new identities: learning organizations (LO) are born. Companies like Shell, Mercedes Benz, and Isvor Fiat baptize themselves LOs, devoting resources to creating recognizable corporate identities. The existence of LO has become naturalized through institutionalization. Mary Douglas (1986) argues that institutions are founded on an analogy with nature; naturalization of social classifications protects the institution when it is still fragile convention.
We may distinguish between LO and OL based on the dichotomy between prescriptive and descriptive research (Tsang, 1997). However, the issue is not about academics failing to generate useful implications versus practitioners lacking rigorous methodologies. The contrast is between a realist ontology (assuming learning as an empirical phenomenon) and a constructionist one. If OL is a "live metaphor" (Tzoukas, 1991)βa means to represent the organization as if it were a system that learnsβthen the problem is not what constitutes "effective learning" but determining the amount of further knowledge yielded by the metaphor proposed.
π Definition β Organizational Learning (OL) as heuristic metaphor: A means to represent the organization as if it were a system that learns, used to see new things and see existing knowledge differently.
π Definition β Learning Organization (LO): A realistic, prescriptive concept describing an ideal type of organization in which learning is maximized, often subject to empirical verification.
π‘ Why this matters: The interest of knowledge shifts from "how does an organization learn or should learn?" to "if we depict an organization as a system which learns, are we able to see something new and to see something that we already know differently?" The former question concerns explanation of OL; the latter relates to understanding (Verstehen) of it (Weber, 1922).
Organizational learning as a disciplinary discourse
The literature on LO has been suspected of colluding with the "ruling courts" governing organizations (Coopey, 1995) and employing a discourse of democracy and liberation ideologically (Snell and Chak, 1998). Both OL (heuristic) and LO (realistic) converge on the same social practice legitimizing managerial techniques based on claims of scientific knowledge. Both contribute to institutionalization of the field as a disciplinary discourse and its superseding through "institutional reflexivity."
From a Foucauldian perspective, discourses are systems of thought contingent upon material practices and informed by those practices through particular power techniques. Much of nineteenth-century social science was shaped by the "disciplinary gaze" of surveillance. In organization studies, the personnel function under "human relations" had a tutelary role, and "organizational learning" is now following suit.
The lecture illustrates a set of premises implicit in OL and LO theorization that sustain a disciplinary discourse which disciplines concrete behaviours:
1. OL is always ameliorative and disinterested. Learning is regarded as always positiveβ"the more, the better." OL assumes an ameliorative vision where learning is incremental and knowledge cumulative. The alleged universality, neutrality, and transparency of knowledge neglect the role of power in structuring organizational knowledge. Only those in power learn the right things.
2. OL is intentional. If learning resembles appropriation of something external, the ways it is appropriated can be specified and normatively sustained. OL may be embodied in SOPs (standard operating procedures), periodically overhauled and updated (Kieser, Beck and Taino, 2001), as "the one best way of learning."
3. OL is an extorted result. The LO requires work groups to "learn" and transfer knowledge to organizational structures, with learning leading to improved performance. The use of power in transferring knowledge is silenced, and OL is conceived as grounded on free transfer, transparency, voluntariness, and chain of authorityβnot micro-conflictuality, micro-negotiation, and systematic distortion/extortion of knowledge.
4. OL presumes change but not its understanding. Learning proposes change in behavior (actual or potential) of individuals or groups, or cognitive change. It does not require individuals to understand the logic behind changes in SOPs (Child and Markoczy, 1993). If change is manifest, a learning process has taken place, but change does not require any learning.
Rorty (1989) writes that learning is part of a final vocabularyβa value in itself that cannot be further questioned. It is associated with improvement, error correction, and fast reaction to environmental changes. The positive connotation induces a priori assumption of what needs to be empirically demonstrated. Learning, as the founding myth of the OL scientific community, obscures the myopia of learning from experience (Levinthal and March, 1993).
π Definition β Disciplinary discourse: A system of thought that is contingent upon material practices and informs those practices through particular power techniques, sustaining normative behavior and providing resources for normalization.
The reification of knowledge in the knowledge management literature
A quantitative survey (Scarbrough, Swan and Preston, 1998) shows that since 1997 the term "knowledge management" (KM) has supplanted "organizational learning." Interest has switched from questions about appropriation of knowledge by individuals and organizations to techniques and technologies of knowledge management. Academic disciplines now predominant are not psychological but economic, with a new alliance between economics of knowledge and information technology.
The concepts of "knowledge work" and "knowledge worker" were first introduced by Peter Drucker (1939), set in contrast to manual work and service work. Epochal changes between the eighteenth and twentieth centuries reveal: first, knowledge was applied to artefacts, processes, and products through technologies, patents, and tacit knowledge; second, knowledge was applied to human labour through scientific analysis of work; third, knowledge was applied to knowledge itself, constituting knowledge work. The endeavour to manage knowledge as if it were a resource reflects this trend, reifying what is fundamentally a social, situated, and practical accomplishment.
π Definition β Knowledge Management (KM): The techniques and technologies for managing knowledge within organizations, which since 1997 has supplanted the term "organizational learning" and is dominated by economics of knowledge and information technology.
π Definition β Reification of knowledge: Treating knowledge as a tangible, manageable resource or commodity that can be captured, stored, and transferred, rather than as a social, situated, and practical accomplishment.
β Key Takeaways
The lecture fundamentally challenges the conventional view of learning as knowledge delivery and acquisition, instead presenting learning as a practical, social accomplishment done together through participation and negotiation. The learning organization is not a naturally occurring phenomenon but a socially constructed concept that was institutionalized through processes of legitimation, identity creation, and naturalization of social classifications. OL and LO theorization sustain a disciplinary discourse with problematic assumptions: that learning is always positive and disinterested, intentional, extorted through power, and assumes change without requiring understanding of its logic. Students must recognize that the shift from OL to knowledge management represents a move toward treating knowledge as a reified, manageable resource driven by economics and information technology, which obscures the social, situated, and practical nature of knowing.
π§ Quick Revision Questions
- According to the lecture, what are the two main reasons the conventional view of learning as "knowledge delivery" is considered a highly reductive account?
- How did the concept of "learning organization" become institutionalized and naturalized as a "real" phenomenon?
- What is the key difference between viewing OL as a realist phenomenon versus as a "live metaphor" (heuristic device)?
- What are the four premises implicit in OL and LO theorization that sustain a disciplinary discourse?
- How and why has the term "knowledge management" supplanted "organizational learning" since 1997, according to the lecture?