Vol. 15. No. 2. CENTER FOR CREATIVE LEADERSHIP I995

A Practical Model for Organizational Learning Nancy Dixon

It is increasingly recognized today that, as organizations face continuous change, they require the use of their collective intelligence. At the same time, the basics of how collectives learn are not generally well understood. In what follows I will discuss some of these basics and outline a model for how organiza- tional learning can be encouraged.

To begin with, it is important to understand that learning is more than the acquisition of existing knowledge. Most importantly, it is an approach to human functioning that emphasizes the intention to “make sense” of our world and to act responsibly upon the understanding we derive from that sense-making.

not simply the sum of what all the members of an organization know; rather, it is the collective use of the capability to make sense of the world. Learning is a verb, not a noun; it is not the knowledge that is stored but the process that creates that knowledge. As Argyris and Schon (1978) explain, “There is no organiza- tional learning without individual learn- ing, and that individual learning is a necessary but insufficient condition for organizational learning” (p. 20).

Thus, organizations must do more than simply encourage their individual members to learn. They must also estab- lish processes for system-level learning. The necessary processes can be repre- sented by a four-step model: (1) Informa- tion is generated as individuals in the organization interact with the external environment or experiment to create new information internally; (2) the new information is integrated into the organi-

Similarly, organizational learning is

zation; (3) all the available information that is related to an issue is collectively interpreted; and (4) action is taken based on the interpretation. The action generates feedback, and that new information brings us back to the first step to begin the cycle again.

interpreting, and acting on information are not new to organizations. Typically, however, different parts of an organiza- tion focus on only one step of the cycle. For example, marketing often collects external information, research-and- development generates new ideas and products, information services distributes information internally, and top manage- ment interprets the information; action then occurs at lower levels, far removed from the people who made the interpreta- tion. When the steps of the cycle are dis- connected, organizational learning is lost.

In my view, if organizational learning is to occur, then every step of the cycle must take the collective into account.

The steps of generating, integrating,

Generate Information Information can be generated through external collection or through the internal development of ideas. Generating external information-for instance, about custom- ers, suppliers, new technology, or economic conditions-requires that organizational members cross the organization’s boundaries to interact with the world. By contrast, generating internal information occurs within the boundaries of the organization. It includes analyzing successes and mistakes, creating.experi- ments designed to provide new informa- tion, and building checkpoints into an

activity so that it can become self- correcting.

tion is the responsibility of all members of the organization, not just marketing or customer service. Although everyone should be engaged in collecting external information, a massive, repetitive effort is not needed. Rather, information should be collected by the primary users of the data-that is, by those people who can themselves act on the information. For example, at a food-processing company, complaint letters about a product are directed to the line workers who produce it, and an innovative hospital pays its employees to visit hospitals in other parts of the country when they are on vacation.

In order to make sense of complex situations, the information collected must provide a diversity of viewpoints. This can be encouraged by collecting data from multiple sources. Many organizations limit their data collection to standard practices such as focus groups or cus- tomer surveys, but informal sources may be equally productive. For example, one constructjm firm asks employees to note any new construction sites they see on their way to and from work.

information be widespread, it must be continuous. By continually seeking information, organizational members can discover new problems and opportunities.

The generation of external informa-

Not only must the generation of

AZso in this issue. .. Going International with Assessment for Development Maxine Dalton ................................. 5

IN THE CENTER Bob Lee ........................................... 7

INKLINGS David P. Campbell ........................... 8

INFORMATION ........................ 9

Center for Creative Leadership 2 Issues & Observations: Vol. 15, No. 2,1995

Integrate Information The continuous collection of information by primary users and through multiple sources is only workable if the second step of the organizational learning cycle is in place: integration into the organiza- tional context.

tion has no access to what other parts know, meaning that parts of an organiza- tion cannot learn from each other. An equally detrimental phenomenon is the inability of each part to understand its own information because it lacks the context of the whole picture-like someone closely examining a single piece of a jigsaw puzzle without access to the other pieces and without knowing what the whole picture is. To accomplish the organization’s task we must act in concert with each other. For that, we must share some understanding about what we are trying to accomplish and how we are going about it; each individual sees the whole task and the relationship of each subtask to the whole.

To gain that whole picture, the distri- bution of accurate and complete informa- tion is critical. Unfortunately, organiza- tions often obstruct information. Daft and Huber (1 987) have identified four methods of organizational obstruction: message routing (the selective distribution of information), message summarizing (the reduction of the size of the message -for example, reducing large sets of numbers to averages), message delay, and message modification. These four meth- ods affect the availability, form, and accu- racy of information in the organization.

Upper management sometimes uses these methods to control the information employees receive, but just as often. employees use them to control the information going to upper management or another department. For example, employees might modify negative information that is being reported upward; management might delay a message to employees until a more favorable time; a manufacturing department might bench- mark a product but give little thought about who else in the organization could benefit from the benchmark information.

The elements needed for the second step of the organizational learning cycle

All too often, one part of an organiza-

are the converse of message routing, summarizing, delay, and modification. All parts of the organization need to provide all other parts with accurate, timely, and complete information. There are compa- nies where all information (excepting personnel data) is stored in a single integrated database that is open to every employee regardless of position.

Tushman and Scanlan (198 1 ) note that specialization within organizations may interfere with information distribu- tion. Specialization increases the effi- ciency of information-processing within a unit but often blocks information from moving across unit boundaries. Idiosyn- cratic language and local conceptual frameworks also work against the distribution of information. How to deal with this problem? One way is the use of boundary-spanning individuals, people who can understand and translate the information to facilitate shared under- standing. Springfield Remanufacturing

Corporation engages in a different kind of translation. It has a continuous scrolling display on the shop floor to show the organization’s performance on numerous indicators. However, knowing that the numbers themselves may seem a foreign language to some, Springfield provides training in how to read financial information.

Many organizations have found unique ways of integrating information. Johnsonville Foods has a policy of giving each employee one day a year to spend with an employee in another department. At Toyota Canada, departments invite employees from other departments into their monthly meetings. Honda requires each manager to exchange jobs for a two- week period with a counterpart in another function. General Electric has imple- mented a best-practices seminar in which internal practices are explained in detail by those who have been engaged in them, and teams, attending from other units,

ORGANIZATIONAL LEARNING CYCLE

WIDESPREAD GENERATION

/ AUTHORITY TO TAKE

RESPONSIBLE ACTION ON THE INTERPRETED

MEANING

INTEGRATE NEW/LOCAL INFORMATION INTO THE

ORGANIZATIONAL CONTEXT

INFORMATION

~ ~~

Issues & Observations: VoL 15, No. 2,1995 3 Center for Creative Leadership

have time to plan together about how they might implement the practice in their unit.

To resolve the problems, organiza- tional members must not only create better processes for distribution; they must also question the organizational assumptions that mediate the full and accurate distribution of information. That questioning is related to the third step of the organizational learning cycle, collec- tively interpreting information.

Collectively Interpret Information

An easy mistake to make is to equate disseminating information with learning. Leaders often think that if they have distributed information widely, organiza- tional members will “know it.” However, individual learning theory suggests that there is a gap between receiving informa- tion and making meaning from it. When someone is given information, he or she selectively attends to it. The selected parts are examined for patterns, which are compared to the meaning the individual has stored in memory. Only when the individual has formed new relationships through this process and again stored those new relationships in memory can we say the individual has learned.

the process is yet more complex. Not only must each individual engage in the sequence just described, each must do so while interacting with other organization- al members (who are of course engaged in the same sequence themselves); and out of their interaction the organizational members must form an interpretation of the information. They need not necessar- ily agree upon an interpretation; the goal of collective interpretation is more the making of meaning than the reaching of consensus. When each member under- stands the reasoning and data that others are using to arrive at their meaning, they then understand others’ meaning more fully and, by comparison, understand their own. Collective interpretation may not develop a definitive answer but it allows them to understand the parameters of the problem more clearly.

When Johnsonville Foods faced a difficult question about whether to take on additional work that would stretch its

For organizational learning to occur

production capacity, Ralph Stayer, the CEO, put three questions to the members of the organization: What will it take to make it work? Is it possible to reduce the downside? Do we want to do it? The entire organization, meeting in groups over several days, struggled with these questions. Even with all the available data in the room, clear-cut answers were not obvious. The decision reached by Johnsonville organizational members was not completely unanimous, yet they agreed to act upon it. All of the members fully understood the reasoning behind the proposed action and moreover influenced that reasoning.

In comparison to the relatively formal approach to collective interpretation seen at Johnsonville Foods, at Chaparral Steel interpretation often occurs spontaneously when a group of employees come together on the factory floor to understand and resolve a problem that has just arisen. The level of formality is less critical than are the conditions that allow collective interpretation to occur.

The opportunity to interact is not in itself enough; there must be at least four basic conditions in place that support collective interpretation. First, informa- tion and expertise must be distributed. If only one or very few organizational members have all of the information related to a subject, collective interpreta- tion is not useful. The initial step of the organizational learning cycle, in which every individual engages in the generation of information, is one way of ensuring that information is distributed. Chaparral sends teams all over the world to look at new technology, talk with competitors, and interact with university researchers. These organizational members have firsthand information to bring to organiza- tional interactions.

Having organizational members continually engaged in internal and external and learning opportunities is another important way to achieve distrib- uted expertise.

Second, organizational members must hold egalitarian values. There are three core values that enhance collective interpretation: freedom to speak openly without fear of punishment or coercion; equality, which must exist for freedom to

exist; and respect, which must be present for equality to exist.

heart of organizational learning a f f i i s the power of the individual mind and equally the power of the collective mind. It acknowledges the capability of indi- viduals to think for themselves, to manage themselves, and to govern themselves. Perhaps one of the reasons so little organizational learning occurs is that the conditions of freedom, equality, and respect so rarely exist in organizations. Organizations will not be able to learn effectively until these are manifest.

Third, the organization’s size and physical arrangement must support frequent interaction between people and departments. Collective interpretation is assisted when organizational members have the opportunity for frequent face-to- face interaction. Obviously, such interac- tion is easier in a smaller organization or those large organizations which are fractionalizing into smaller business units.

Fourth, members must engage in processes and,skills that facilitate organi- zational dialogue. To carry on organiza- tional dialogue the conditions already described for collective interpretation need to be in place and in addition participants need the skills to: provide others with accurate and complete information that bears upon the issue; c o n f i others’ personal competence when disagreeing with their ideas; make the reasoning that supports their position explicit; say how they got from the data to the conclusion; voice the perspective of others; change position when others offer convincing data and rationale; regard assertions, their own and others, as hypotheses-to-be-tested; and challenge errors in others’ reasoning or data.

Although these skills might seem self-evident, their actual use is rare. It is much more common, for example, for people to withhold critical information for fear that it might embarrass themselves or others, to offer their conclusions but not the data on which they were based, and to regard their own position as ‘‘truth” and others’ positions as in error.

If the skills appear self-evident yet are in little use, we have to wonder whether there is a skill issue or if it is just

Placing collective interpretation at the

Center for Creative Leadership 4 Issues & Observations: VoL 15, No. 2,1995

that the organizational condrtions in which most of us function have made the use of these skills hazardous. In part the answer may be that the skills we use in everyday interaction are tacit; we have so often withheld our reasoning, refrained from saying what we know others do not want to hear, and held onto our position long after the evidence has proved us wrong, that it is those skills that have become automatic. If we want to employ the skills of dialogue, we may first have to “unlearn” the tacit skills that are prevent- ing effective dialogue. Although many people have written about the need for dialogue, Argyris (1986) is probably the most articulate in offering specific skills and processes both to achieve organiza- tional dialogue and to uncover the tacit skills that are preventing their use.

Authority to Take Responsible Action Based on the

Interpreted Meaning If an organization involves its members in the generation of information, positions that information in the organizational context, and encourages its members to collectively interpret that information but then stops short of authorizing organiza- tional members to act on the knowledge they have derived, the learning is lost. There is in all of us a propensity to do what needs doing. To understand what needs doing but be prevented from acting on that knowledge leads to anger or despair or, in some situations, subversion.

responsibly, they must have enough discretion in their actions to make changes when and where they are needed. Rather than pre-designing as much as possible, organizations need to pre-design as little as possible. It is said that the designers of a new Honda were given only two instructions: “Come up with a product concept fundamentally different from anything the company had ever done before . . . and make a car that [is] . . . inexpensive but not cheap” (Nonaka, 1991, p. 100).

The reduction of risk is as necessary to taking responsible action as is having the authority to do so. Finally, in regard to taking responsible action, i t is important to note the need for a more equitable way

If organizational members are to act

to share in the financial gain of the organization. When organizational members have full information about the organization’s profit and loss, when they are apprised of the amount of dollar savings their actions have caused, or the increase in earnings their teams’ new process has made, when they are asked to be responsible for these actions, they will eventually expect to have an equitable share of the rewards.

A Powerful Tool Organizational learning is a powerful survival tool. It can enable an organiza- tion to transform itself, both internally and in its interaction with the environment. As a self-transformation rather than an imposed change, however, it requires the active involvement of its members in establishing the direction of the change and in inventing the means to achieve it. Organizational learning is about intention- ality-the organization transforming itself in a direction that is more satisfying to its stakeholders than its current state.

Organizational learning creates continuous change rather than resulting in a one-time change. Learning propa- gates itself.

Learning is, by its nature, unpredict- able at both the individual and organiza- tional level. When we say to individuals and teams, “challenge the assumptions you see, invent new and better ways to function, take action when it is needed, freely share the ideas you have with others,” the outcome is unpredictable. Happily so! If learning were predictable we would be limited to what is already known. It is the unpredictability of human learning that allows us to think in new ways, to create and innovate. But the unpredictability of learning also means that the change that is created, and the assumptions that are challenged, may not be in the direction that the instigators or the change anticipated. Organizational learning is the foe of those who would manage and the ally of those who would lead.

References Argyris, C . (1986, September-October).

Skilled incompetence. Harvurd Business Review, pp. 74-79.

Argyris, C., & Schon, D. C. (1978). Organiza- tional learning: A theory of action perspec- tive. Reading, MA: Addison-Wesley.

Daft, R. L., & Huber, G. P. (1987). How organizations learn: A communication framework. Research in the Sociology of Organizations, 5, 1-36.

Nonaka, I. (1991, November-December). The knowledge creating company. Harvard Business Review, pp. 96-104.

Tushman, M. L., & Scanlan, T. J. (1981). Boundary spanning individuals: Their role in information transfer and their anteced- ents. Academy of Management Journal, 24(2), 289-305.

I--

Nancy Dixon recently completed a sabbatical at the Center. She is an Associate Professor of Administrative Sciences at The George Washington University in Washington, D.C. This article is adaptedfrom The Organiza- tional Learning Cycle: How We Can Learn Collectively (London: McCraw- Hill, 1994).

Reprints If you would like a reprint of a feature in this number of Issues & Observations, please send $1.00 each for up to 10 copies, $.75 each for 11 -50 copies, or $.a each for more than 50 copies. Shipping and handling is 6% of subtotal ($1.00 minimum). Send to: hblica- tion, Center for Creative Leadership, P.O. Box 26300, Greensboro, NC 27438-6300; call 910-545-2805; or fax 910-545-3221.

1 is printed on recycled paper. 1