Chapter 2 – Theoretical framework of Organization learning and Literature Review
Organizational level learning – While there is broad agreement of the need for organizations, not just individuals, to learn , there is much less agreement on what organizational learning is and how and why it take place. The term “organizational learning” covers a wide range of models and concepts that are diverging rather than converging. Some of the more popular and familiar are:
– single and double loop learning (Argyris and Schon, 1978)
– experiential learning of individuals or collectives (Sata, 1989)
– a system view of learning which is constructed in Senge’s treatise (The Fifth Discipline, 1990)
Looking at twenty-five years of literature that has grown up around organizational learning, tow of the original thinkers in the field, Aryris and Schon, concluded that two different branches have emerged: one being the “prescriptive, practice-oriented, value-committed, sometimes messianic and largely uncritical” and the other is more scholarly and tends to be distant from practice, non prescriptive and open to the view that learning may be good or bad, linked or not linked, to effective action or desirable outcomes.
Organization learning literature is not more consistent and congruent which makes it difficult for practitioner to use, change or build on.
A learning organization theory by Lahteenmaki et al conceptualize organization learning in 2 ways 1) learning is either individual or organizational and 2) learning is either a process that is enacted or a set of attributes or characteristics that are held, which is often referred to as a learning organization. The learning organization theory depicts behaviors and conditions that if present would indicate a capacity to learn or expectation of an ability to learn, at either the organization or individual level (Serge 1990, Pedler, 1991).
From the above literature review, we can create a simple table to show the various segments of organizational learning literature covered (Table 1).
Segments of Organizational Learning
Organizational Learning Process Learning Organization Attributes
Individual Level Ways individual learn in organizational context End state of personal learning
Organizational Level How organizations learn – Organizational Level Learning OLL What an organization that learns look like
Our study will focus on the how the organization learns and the process of learning at the organizational level – OLL. OLL can have 3 components – type of learning to be done, method and outcome. Therefore our research work seeks to find the threads or pattern that create a typology of various learning modes connecting what is learned, how it is learned and how this learning manifests. It essentially constructs ‘learning method” as a choice in the context of the type of learning required and the type of outcome that is desired in a given situation. Therefore looking from this view of OLL, I can further define OLL for the purpose of this study.
OLL is the means by which the organization comes to hold new ideas, beliefs or knowledge and operates in a new way and behaviors and is the vehicle which the organization uses to create a change in paradigms either in degree or type.
It is these learning concepts of the extant literature, knowledge management and process innovation literature that form the framework of this study.
Knowledge Management – A study of organizational level learning will not be complete without considering the developments in knowledge management. Knowledge creation is roughly equivalent to organizational learning as it has developed in the allied field of information systems or knowledge management. As such a conceptual understanding of its role and development is essential here for this research. Knowledge as a strategy and as a resource to be managed has gained popularity in business management practice and theory (Earl 2001, Zack 1999, Nonaka & Takeuchi, 1995). It is increasingly seen as a bottom-line, results oriented solution with increasing returns in a variety of industries and functional areas.
The recent emphasis on knowledge management is the result of the convergence of tow forces – new technology and the increasing mobile and turnover rate of employee base. We have all seen the proliferation of personal computer and telecommunication technology lowered the boundaries between users. As a result, organizations became capable of producing and delivering, at low cost, more information and data than is usable. At the same time, with the telecommunications development, these data can be deployed globally enabling the organization members to integrate into their daily work. With this new insight, user of information who had traditionally demanded more data had to re-think their demands, they began to ask for less not more, They sought instead information that is more useful and that add value to their work. The attention of the actors rather than the information itself had become the scarce resources in the organization (Davenport et al, 2001). For example, a sales manager no longer wanted to see just a list (data collection) of sales metrics, she wanted to know where to focus her efforts, helps in making decisions and directions in taking actions. Collecting, codifying, storing and retrieving data, typical mechanistic information management functions, were no longer sufficient (Davenport, 2001).
User demands had morphed from “timely, accurate information”, the mantra of the 1980’s to usability and value. There was a demand for information that is developed and presented in a way that contemplates the use of it, meeting practical needs, applying information to a specific purpose in a specific situation. As the old IBM advertising tagline said “not just data …. Reality”. What users were now seeking is information that is embedded in a context, characterized by Davenport – “knowledge has the highest value, the most human contribution, the greatest relevance to decisions and actions, and the greatest dependence on a specific situation or contest.
The second force as mentioned that drove the need for managing organizational knowledge was the increasing rate of dislocation or turnover of employees base. The length of time that an employee would stay with a firm decreasing rapidly since 1990s. Rather than lifetime employment within one firm (like Japan). The turnover was motivated by various conditions: a) they were no longer needed in a restructured and downsized organization. b) their skills and/or interests were no longer a match for the task and c) they found better alternatives elsewhere. Rather than the historical dislocation through attrition, in some cases whole functions were being terminated en masse (out-sourced). As these people left, it became apparent and problematic that the firm was also losing their experience, their knowledge and their “situated” learnings (Nidumolu, Subramani & Aldrich, 2001).
This acceleration in the loss of intellectual capital drove organizations to seek ways in which they could capture and retain vital learnings, particularly tacit knowledge, in a systematic and proactive way (Von Krogh, Ichigo & Nonaka, 2000). Tacit knowledge is defined as knowledge which is not spoken, but is implied by action or statements. This differs from explicit knowledge which is articulated and therefore able to be communicated. Tacit knowledge may be so embedded in the individual that it is not even conscious. For example, the factory worker who roles rubber onto a cylinder over and over again until it is a consistency that makes it ready for the next step, performs this task everyday, but may not be able to describe it, or make this knowledge explicit. The same is true for the engineer who reads functional specifications for an integrated circuit and then, seemingly automatically, uses computer tools to produce the designs. With the increased probability of losing these people and this knowledge, knowledge that is being used in everyday performance of the work, organizations were motivated to capture and preserve it. Information systems and data bases needed to be reconsidered in light of the need to uncover, discern and preserve critical knowledge both tacit and explicit.
The convergence of the 2 factors of user demand knowledge rather than information and the increasing rate of loss of intellectual capital drives the current emphasis on the management of knowledge. Like organization learning, knowledge management has become home to a variety of concepts and definitions. In fact, it includes views that suggest knowledge cannot be managed and can only be stimulated and channeled. The knowledge management literature addresses knowledge as 1) noun or an object 2) a verb or a process and 3) a combination of these when it is considered as a resource of the organization.
Gold, Malhotra and Segars have distilled a myriad of definitions suggested by many researchers into a comprehensive and concise set of four sub-processes of knowledge process (2001). The four sub-processes are:
1) Acquisition – accumulating existing and creating new knowledge (innovation), both of which require sharing and dissemination of personal experience (Inkpen & Dinur, 1998).
2) Conversion – making existing knowledge useful, having standards and representations that enable and support communication and dialogue across various boundaries (i.e. Personal, functional, organizational).
3) Application – using knowledge, having process for storing, retrieving and sharing which can be found for example in software development that employs open architecture and
4) Protection – discouraging “illegal or in appropriate use or theft” of knowledge.
And to further delineate the knowledge management domain upon which this research relies, 2 of the sub processes are most relevant to learning: acquisition and conversion. First, acquiring knowledge can be thought of specifically as individual or organizational learning that is based on personal experience (experiential learning). And second, converting knowledge recognizes the need to make tacit knowledge explicit and make explicit knowledge available for other’s learning and acquisition, and transfer to others, to make it common, usable to many at both the individual and organizational level (Wenger, 1998).
As discussed earlier, knowledge management is not just a mechanistic process of collecting, storing and retrieving data (data processing), though that is necessary step along the way. Rather, knowledge comes about by connecting the content or the subject matter, to the context or the setting in which the event takes place (Grover & Davenport, 2001). Connecting content to content, involves relationships, both the relationship between and among people (Cross, et al, 2001) and the relationship of the data to the person, which includes both relevance and interpretation of the data. With regard to the relationship of people, when knowledge creation depends on information being shared between and among people, its very availability depends on communication, interpretation and meaning. In order for this information to transform into knowledge, there must be openness and disclosure both within the cognitive system of the source and in the relationship of the workers (Davenport at al, 2001). This sharing is a matter of trust and of personal boundaries (Prusak & Cohen, 2001) as well as a willingness to disclose. The extent to which this “context” of sharing can be made available depends on the relationships of people. Being able to access certain constructs of a person’s life (context) requires a different kind of relationship than is used in the mechanistic managing of information, namely personal and subjective rather than distant and objective. Personal and subjective business or economic relationships are complex in that they at once offer the possibility of opportunism and trust, which are antithetical in nature.
Looking at learning through a knowledge processing perspective suggests that in order for learning to be transferred they must be made explicit, so that they can be communicated, a concept which is also present in the learning theory perspective. But perhaps even more importantly, the knowledge that gets created must be capable of being unpacked so that it can be looked at and assessed for other uses and applications, separately from its environment. Only in this deconstructed form can one assess its usefulness and appropriateness for another situation or context. With these learning factors, communication and deconstruction, I see a convergence between the traditional learning theorist’s and the knowledge processing views.
Organizational learning and knowledge management have much in common as they address the phenomenon of increasing the organization’s capacity to innovate and change (Grover & Davenport, 2001). Both literature has common ground, both has:
1) developed “rich theoretical perspectives” and must now address “the How” questions of knowledge management” through field research (Grover & Davenport, 2001)
2) taxonomized various schools of knowledge management practice (Earl, 2001) and now needs to systematically, rather than fragmentally, study knowledge as a field, recognizing the differences from its “field of origin”, information systems and management, particularly the situated nature of knowledge and the socially constructed attributes of “knowledge”, the role of the human in converting data into knowledge and making it available for transfer (Nidumolu, Subramani & Aldrich, 2001)
3) identified that both organizational and individual knowledge exist and that there is a link between them.
Research needs based on current literature
Both literatures leads to a need for primary research, that builds on existing theory and frameworks and that explores in-depth the relationships between type of learning or knowledge creation, method of learning and knowledge creation and manifestation of learning what is getting unpacked and used in other areas and processes in the organization (see Figure 1).
When taken together, organization learning and knowledge management, this offers integrated perspectives from which to view OLL, rather than simply more drilling down on either the behavioral or cognitive frameworks. Following is an expanded explanation of each of the components of the frameworks (see Figure 2).
Organization Behavior view – In a literature review of types of learning, Fiol and Lyles, identify 15 different models that through various terminology and concepts are essentially about behavioral (lower-level) and/or cognitive (higher-level) type learning. In summarizing the characteristics they conclude that lower level learning addresses learning relative to routine tasks, learned through repetition, in a well understood context, typically by all or many levels in the organization; higher level learning is considered in conjunction with non routine tasks, learned through heuristics and insights, in an ambiguous context, by upper levels of management. Argyris and Schon described the type of learning relative to how much change occurred. Their view holds that learning may be constrained by existing paradigms. For example in learning how to pull the customer first, the organization may be constrained by a need to implement this within the existing organizational values. This would be considered single loop. And this is contrasted with learning that is unconstrained. Here through an understanding of what it means to be customer focused, the organization learns that existing values have to change, breaking the frame of an existing paradigm. This is termed, double loop.
Knowledge Management View – Another perspective of the learning type comes from the field of knowledge management. In my model, we can think of learning applying to or the subject of learning being either 1) factual or declarative knowledge or 2) processual knowledge, which relates to the underlying process that affects the facts. For example, an organization may learn that they can improve their customer service up 5% by reducing the number of customers they serve by 2%. This would be declarative or learning about declarative information. On the other hand, they may learn that their process for understanding the key performance parameters must include a cross functional team and this is creation of processual knowledge.
There are three basic methods or processes that have been theorized by which organizations learn. These are 1) shared or saturated, 2) distributed, linked or networked and 3) chaotic, divergence and convergence. These methods are applied to a type of learning to produce an outcome. In reified form, they would be considered the input to a process, that process being a certain form or method of learning. These three forms that have been theorized in the literature are described below.
1) Shared/Saturated – The first method is termed “shared” or “saturated”. Here all individuals in the organization have come to some common understanding of phenomenon, such that it can be said that the organization operates or thinks in a certain way because each of the members, individually operate or think this way. Learning therefore occurs when the shared meaning of the organization members changes. An underlying assumption of this method is that it is possible for all people to have the same mental map. And that through conversation and collaboration among its members, through sufficient discussion, construction and reconstruction, all or just most of an organization’s members and therefore the organization will come to a new perspective or cognition. Often though not always, this method is characterized by use of large group events quality circles, and other programs designed to reach a broad population, which together then shapes the vision, goals, expectations and new behavioral norms. The shared model seeks to extend theories of individual learning to the organizational level. And it is these cognitive systems that must develop a shared worldview and this memory bank that must be recalled uniformly across the entire group or organization as it would for a single individual. In this method, the “sharedness” of insights and behaviors is the gate to organizational learning. For example, take the case of implementing lean manufacturing in a work group or cell. Here we have individuals who have traditionally been responsible for and measured on performing one function, at one station in a mass production process. In a lean mode, groups of workers change their focus from being good at their individual function, to one of collectively being good at high quality, low cost and total throughput. In order to succeed in this environment much more is expected of the individuals. Not only must new skills be acquired by individuals, but the organizational goals, expectations and view of priorities must change. The shared method of learning would say that in order for it to be said that learning occurred; all of the members would share the new beliefs and behaviors and have the memory of the old. Specifically, the change that would be reported would be something like – we all used to do just our won work and not care about the total throughput and now we all care about the cell first and do whatever it takes.
2) Distributed/Linked/Networked as learning process – The second method is termed “distributed: and is characterized by a network structure. Here the organization is said to have learned when changes in thought and behavior come from the linkage of nodes of knowledge. Unlike the shared or saturated method, “distributed” holds that few people with different expertise, roles or functions are linked together in such a way that if they collectively change their beliefs and behaviors, it can be said that the organization has learned. It places much more emphasis on the situated nature of learning, a learning perspective that views knowledge as embedded in the individuals, in connections between individuals and in artifacts. Recent research creates a distributed, rather than shared view of organizational cognition using social network theory. The map of participants in this learning method is much less dense and redundant than that in the shared method. Here an organization is said to know something or have learned something based on a few people who jointly hold discrete knowledge, that through connection to one another, collectively have learned. It is in fact unlikely that those with whom one typically communicates will be a part of the learning process, the new knowledge formation. This type of learning is catalyzed through non-redundant linkages and it is the links that hold the key to learning and the transfer of learnings. For example, the implementation of work cells typically results in a different flow of material through the plant, perhaps requiring a different set of tools and equipments, as well as worker skills and beliefs. As the workers focus on throughput and meeting delivery commitments, the mechanism to appraise the investment in appropriate resources to support these workers, such as these tools and equipment, may need to change. In order for the investment decision making process to change, multiple experts or functions must be involved in understanding the connections among throughput, customer satisfaction, accounting, manufacturing technology, etc. If organizational members from each of the functions collectively changed their perspective and behavior around this issue, it could be said that the organization has learned. What we would hear is that – a few of us shared information that collectively created new knowledge of what the organization needed to do and believe; we changed not based on a shared view of what was (memory) or is (shared view or common understanding of the business) but on what must be. Specifically, there would be great numbers of people who know nothing of the change in utilization accounting and a few others who describe communication and the need for understanding and linking the critical pockets of expertise and experience in the firm. This will often happen not only across functions, but across line (on the ground) and staff personnel. This view relies in part on a knowledge management perspective, specifically knowledge creation. A separate section on knowledge management is included below in this literature review.
3) Chaotic/Divergence/Convergence in learning process – The third and last model is probably the least understood and articulated and comes from the process innovation literature. Based on innovation theory, organizational level learning can be said to be a product of dynamic systems behavior. This stream of innovation research suggests that OLL is not linear, as would be found in a cyclical model. In empirical research using dynamical systems theory (DST) Cheng & Van de Ven found that the innovation process exhibited chaotic patterns, meaning it is a nonlinear system which is neither stable and predictable nor stochastic and random. They further suggest that DST provides hope by suggesting what a dynamic model of learning during the innovation journey might look like a nonlinear dynamic model of learning calls for an expanded definition of learning that examines not only how action-outcome relationships develop, but also how prerequisite knowledge of alternative actions, outcomes and context emerges. Learning in chaotic condition is an expanding and diverging process of discovery. The complexity in OLL lies in the nature of the interrelationships among the parts whose cause-effect relationships are highly nonlinear and distant in space and time (Kim, 1993). For example, in lean implementation, the way the work group changed may be described something like – for weeks we experimented and read and talked about how to do work cells. There was much disagreement, we didn’t see it the same way at all, we seemed to be coming from different places and then one day that changed. Though we didn’t necessarily see it the same way, we each knew what had to be done. There was nothing that could have brought to that point sooner other than more experimenting sooner.
The last component of our exploratory model (Figure 2) are the “outcomes” to learning that in fact give credence to and make manifest the output of the learning process. It is how we know that learning has occurred. It is not axiomatic that if there is change, there must have been learning or vice versa. Nor if there was training there must be learning. Nor if the manager or machine operator learned something, the organization did. Learning at the organizational level has it own sets of outcomes and changes.
These are connected to the learning methods through the action-outcome relationship. The methods act on the learning type and produce an outcome. These types fall into two broad categories: one is tactical and the other more strategic. For this research and keeping with my desire to integrate the practitioner and scholarly literature, particularly as it relates to pragmatic outcomes of learning, I choose to use the tactical/strategic framework which is common to the two branches rather than the view of behavioral versus cognitive. Moreover, it will better serve my purpose of investigating the typology of learning that connects types, methods and outcomes, because these concepts are linked, as opposed to juxtapose as with the cognitive/behavior framework. Learning outcomes can manifest in either/both forms.
Tactical Outcomes – These are changes in operating norms, behaviors and procedures. Here the output of the learning can be seen in action of the organization, the change in the routines of every day operational behaviors. This includes such element as the normative behavior of people or systems; of experts and technology specialists and machine operators; of database design or plant floor design. This is change that is visible, though it has come about as a result of a change in the knowledge or intelligence in the organization. It is often though not always explicit.
Strategic Outcomes – These are changes in expectations, goals, mental maps, viewpoints, myths and theories of action. Here the output of the learning can be found in the application of insights, new heuristics and a new organizational (collective) consciousness. We can look for evidence of learning in how the organization, again through its members, plan, consider, anticipate. This is change that may not be seen at the time that it occurs, but decisions based on these new insights are in evidence, supporting the notion that the learning itself is often, though not always, tacit in nature.
Altogether these concepts as depicted in Figure 2 constitute an area of investigation that I will undertake. This mode guiding framework of organizational level learning model bounds my research in the sense that it focuses on the learning process of the organization which includes reliance on the learning process of individuals. To be clear the model represents areas of inquiry rather than a hypothesis that is being tested. The specific research questions and the reasons for pursuing them will be followed in next Chapter.