Conclusion
AVOIDING THE STRUCTURAL "QUICK FIX"
In the early stages of the quality movement in Japan in the 1950s, quality control experts applied statistical tools to more reliably check the quality of products produced. This sparked Japan's quality ascent. But more significant breakthroughs came in the early 1960s, when a few companies, led by Toyota and Komatsu, began to break with tradition by getting rid of quality control checkers, teaching the tools and methods directly to front line workers, and giving them authority to analyze their own processes. This linking of new tools with a new level of authority ignited an engine for continuous improvement.
Today, in the arena of "reengineering," a similar synergy is needed between infrastructure innovation and theory, tools and methods. Organizations are attempting to reorganize more around "horizontal" processes that cut across traditional vertical functions. But such ''horizontal process" organizations are much more interdependent than traditional functional organizations. This places a particular burden on people to learn together and practice systemic thinking.
For example, a common form of reengineering is to "co-locate" a the engineers of a large product development effort into one site, to alleviate traditional organizational "stovepipes" separating engineering subspecialties. But in itself, this "co-location'' often fails to get at the real barriers to cross-functional problem solving which are in people's heads, not within the organizational structure. Solutions to cross-functional barriers tend to emerge only when the newly "co-located" engineers can develop openness and systems thinking skills, and discover how their individual ways of working might unintentionally sabotage the development of the product as a whole.
Without a well-articulated theory and set of tools, most reengineering efforts are driven instead by vague concerns to eliminate redundancy or reduce costs. Even if such early efforts are successful, they do not build an organization's capability to continually reengineer itself. Often, the organizations become dependent on expert reengineering consultants.
Already, critiques have begun to surface about the arbitrariness and unreliability of reengineering when it is not guided by clear theory. British management historian John Thackray has written, "Re-engineering is not exactly a tool box-more of a direction, a cause, a faith in the possibilities of top-down revolutions." And when McKinsey partner John Hagel recently offered a list of common causes of reengineering failures, every item on the list was a symptom of the absence of appropriate theory: "failure to understand the processes that are being demolished before the re-engineering is implemented; attacking too many processes- there are usually only about five or six that are truly significant; exclusion of some parts of the corporation from any impact or consequences- i.e., sacred cows; and excessive speed- most successful re-engineering programmes take three to four years."
PREPARING THE SOIL AND DEVELOPING THE SEEDS
Many of the methods and tools of learning organizations will be impossible to implement widely without changes in traditional guiding ideas in management. In turn, new guiding ideas will prove impossible to instill widely without a corresponding commitment to appropriate methods and tools.
In the late 1960s, a major system dynamics study of a highly successful capital-goods manufacturer revealed that the firm had been losing market share because of its production policies. Whenever incoming orders declined, production schedules were cut hack so aggressively that delivery times increased. The product was actually less available during recessions than during boom times. Disgruntled customers would turn to competitors, who would then retain their business once industry orders began to expand again.
Inspired by the insights of the study, the company's top management instituted a new production policy during the 1970 recession to maintain production rates. Market share expanded executives estimated a net profit gain of several million dollars. Unfortunately, four years later, when the major 1974 recession hit, the firm reverted to its traditional production policies, delivery times increased, and the decline in market share resumed.
The successful production policies failed to "stick" because three generations of CEOs had invested their reputations in developing aggressive inventory control policies. Inventory control had, in effect, become one of the company's preeminent guiding ideas. If you were a production manager, there was no more surefire way to ruin your career than to be responsible for overbuilding inventory. This fear could have been changed only through concerted effort by top management to articulate new guiding ideas that could gradually supplant it. But to champion such a change would require acknowledging that the old ideas were no longer appropriate something top management was unwilling to do.
Despite its unhappy ending, the above case was more successful than many systems studies which never result in any changes in policies and practices, even temporary ones. The reason, again and again, is that the systemic insights are inconsistent with traditional guiding ideas. The precious seeds of new insight fall on barren soil.
For example, implementing systemic insights may require that diverse organizational interests cooperate in pursuing policies that might be suboptimal for individual functional areas. But such behavior can seemingly contradict traditions of functional excellence. Unless commitment to the mission and vision of the larger organization is greater than commitment to individual functional goals, functional goals will predominate.
Today, many executives are articulating a new philosophy revolving around "empowering people." But few organizations are working hard to introduce tools and methods to actually help people to make more intelligent decisions, especially decisions that improve systemwide performance. The result will likely be organizations which decentralize authority for a while, find that many poor and uncoordinated decisions result, and then abandon the "empowerment" fad and recentralize. The "empowered" soil will lie fallow, with no seeds to grow. This, of course is precisely what many of the newly "empowered" workers, cynical from past management fads, fear.
MAKING MEANING OF NEW STRUCTURES
In both political and corporate arenas, senior managers are often eager to make changes in infrastructure, believing that the more dramatic and quick the changes they make, the more long-lasting and positive the effects may be. Yet, there is abundant evidence that changes in infrastructure, like reorganizations and changes in reward systems, often have far less impact than expected. One reason is that they conflict with established guiding ideas.
Despite the eagerness and political payoff that often come from changes in infrastructure, when we first work to articulate guiding ideas, and then design the infrastructure reform in harmony with those ideas, the results seem to be far more sustainable. Links to guiding ideas allow an infrastructure reform effort to move from a reactive to a creative orientation to shift from a point of view which says (for example), "We've got to get rid of the structural barriers which are holding us back," to a point of view which says, "In the organization we really want to build, what structures (policies, reward systems, and resource-allocation mechanisms) would support our vision?"
For example, in 1990 the operations managers of Hill's Pet Nutrition Inc. distributed a list of "guiding principles," including this statement about teamwork: "People will work as a team and cooperate when they share common goals, receive proper information, have the skills to recognize, utilize and balance others' strengths and weaknesses, value teamwork, are rewarded for doing so, [and] are recognized as a team for doing a good job." Having articulated that principle, they then instituted several infrastructural reforms which resonated with it. At a new "greenfield site," they began training all their employees before the equipment arrived in the plant. They insisted that the building's architect consider team learning in the design of the building. Union-management relationships, reward and appraisals, and all the other conventional mechanisms of "infrastructure" changed to match the growing understanding of guiding ideas by people throughout the organization. Most impressively, having a set of guiding principles allowed Hill's to develop infrastructural links between their four very different manufacturing facilities, allowing the management of all four sites to act together as members of a common team.
PUTTING IT ALL TOGETHER
The power of the above ideas comes when we put the pieces together. An image emerges of the full scope of the work of building learning organizations; an image that is both more complete and more richly textured than can be seen from "the five disciplines" alone.
The triangle of organizational architecture represents the most tangible form of efforts. (Indeed, that is precisely why the triangle symbol is used: all physical structures start with the triangle. In three-dimensional construction, the most elementary physical structure is the triangle's cousin, the tetrahedron.) By contrast, the circle represents the more subtle underlying discipline-based learning cycle. (As a form, the circle is inherently abstract and intangible with no edges or vertices, with no beginning and no end, an ancient symbol of ongoing movement.) The key focus for activity is in the triangle. The central causality of change is in the circle. Both continuously affect and influence one another. Together they represent the tangible and subtle changes involved in building learning organizations.
We tend to assume that which is most tangible is most substantial, and that which is intangible is insubstantial. In fact, the opposite is true; A set of guiding ideas articulated by one generation of management can be changed by another. And infrastructure developed and implemented today can be redesigned tomorrow. A current set of tools and methods can be supplanted by a new set of tools and methods. The very reasons why we focus on the triangle because here is where we can make changes-also means that those changes can be short-lived.
By contrast, the deep learning cycle, which seems so evanescent and uncertain at first glance, endures. Once we begin to assimilate systems thinking as a way of seeing the world we become, in the words of one manager, "looped for life." Once we learn to distinguish our assumptions from the "data" upon which those assumptions are based, we are forever more aware of our own thinking. Once we begin to operate with a genuine sense of vision, we have a permanent understanding of the difference between reacting and creating. Once a group has participated in true dialogue, its members do not forget. Changes produced by the deep learning cycle are often irreversible.
I have seen countless cases where people continue to pursue their dreams even though there is no organizational reward, once they have developed enough confidence and competence to make progress. They simply do it because "it is the right thing to do." It sometimes becomes impossible for senior management to uproot a shared commitment to systems thinking and openness, once it has become established. Learning teams within organizations simply outlive unsympathetic bosses.
This does not mean that, having begun to practice the learning disciplines, we will retain high levels of mastery automatically. As in any discipline, our level of expertise ultimately depends on how far along our own developmental path we travel, and on our commitment to continual practice. But we do not forget the basic principles we have learned. The first deep effect of the learning cycle is orientational we become oriented to a way of being that remains with us, as a sort of inner compass. We may not always operate in the manner of that discipline, but we tend to know when we are, and when we are not.
BALANCING ATTENTION BETWEEN TRIANGLE AND THE CIRCLE
When optical telescopes were the only form of astronomy, observers were trained to focus away from faint objects they were trying to detect, because the cones of the eye are actually more perceptive of object on periphery of our vision. Similarly, while changes in the circle are what really matters, attention is often best placed on the triangle of guiding ideas, infrastructure, and theory, methods and tools. These represent the operational changes where concentrated time and energy can produce results.
Yet, while we are focused on the triangle, we are mindful of the circle. Buckminster Fuller used to talk about the "Principle of Precession" characteristic of many significant change processes. When you spin a top, the primary mode of movement is rotation around its axis. But, after a while, a secondary mode of movement develops. The top begins to precess, as the axis itself slowly, gradually begins to move around its original position. This precession is quite mysterious to the casual observer server because it has no visible relation to the obvious rotation of the top. Unless we understand the dynamics of the top as a system, we might not even notice the precession, and we certainly wouldn't tend to connect this subtle movement to the spinning. So it is with the deep process of learning. For a long time it may appear that there is nothing going on except the surface activity of the triangle. People talk about new ideas. They practice the application of tools and methods. They design and implement changes in infrastructure. Yet, deeper changes are in the offing. When those deeper changes start to become evident, many people will not even notice them and those who do will often not connect them with the obvious activity.
Yet, the two are connected in subtle ways. The deeper changes are evoked only by sustaining the surface movement. If the rotation stops so too will the precession. If we stop working to articulate guiding ideas, to improve infrastructure, and to apply the tools and methods embodied in the learning disciplines, the deeper learning cycle will not progress.
Similarly, the deeper changes will gradually affect the work on architecture. Potential guiding ideas like "openness" and "localness" will have little conviction until enough people experience the collective intelligence of the whole that is possible when capabilities for dialogue, mental models, shared vision, and systems thinking develop. This is one reason we generally advise against writing down mission or philosophy statements too hastily. A premature articulation can "freeze" people around principles which have not yet been experienced, precluding deeper understanding and conviction.
RESULTS
Ultimately, learning whether it is learning to walk, ski, or compose symphonies is judged by results. The rationale for any strategy for building a learning organization revolves around the premise that such organizations will produce dramatically improved results, compared to more traditional organizations. Whether the results include profit, time to market, customer loyalty, or other agreed-upon measures of effectiveness, learning must ultimately be assessed in terms of "how well the game is played." None of us would think a product development team was learning if it did not improve its products, or a sales team if it did not establish more loyal customers.
The problem is knowing how and when to measure important results. There are two interrelated issues in assessing results of learning processes: patience and quantification.
We need patience precisely because deeper learning often does not produce tangible evidence for considerable time. "You don't pull up the radishes to see how they're growing," says Bill O'Brien. Yet, in effect, impatient managers often do just that to assess whether or not learning processes are progressing. As a Ford manager pointed out in one of our recent core courses at MIT, "If calculus were invented today, our organizations would not be able to learn it. We'd send everyone off to the three-day intensive program. Then we'd then tell everyone to try to apply what they'd learned. After three to six months we'd assess whether it was working. We'd undoubtedly then conclude that this 'calculus stuff' wasn't all it was made out to be and go off and look for something else to improve results."
O'Brien states a simple guiding principle: "Time periods for measurement must be congruent with the gestation period of the learning." Measurements that are made prematurely will lead to erroneous conclusions. This principle, while easy to state, can be very difficult for impatient managers and organizations to practice.
The second problem with assessing results is quantification. Again, there is a simple guiding principle: "Measure quantitatively that which should be quantified; measure qualitatively that which should not be quantified." In almost all organizational learning settings there will be some important quantifiable results: sales, time to market, product quality, total cost (especially including many costs which are often hidden, like life cycle costs), and profit. But many of the most important results of organizational learning are not quantifiable: intelligence, openness, innovativeness, high moral quality, courage, confidence, genuine caring for the customer, for one another, and for our shared aspirations. Despite the nonquantifiable nature of such results, they are not unknowable. There are many ways that people can come to agreement in making assessments of progress in producing such results. But there are also many dangers.
In particular, organizational "cultures that are saturated exclusively in scientific principles," says O'Brien," have an insatiable appetite for quantitative measurement even when they misrepresent truth and reality." For example, management often uses quantitative "proxies" for qualitative results, such as the proxies used with operating staffs. "Managers," suggests O'Brien, often "become obsessed with the proxies and not with what the proxies are intended to represent. This often causes destructive games playing in companies," even to the point of causing people to do things to make the proxy look good counter to the desired result. "There are times," O'Brien concludes, "when the organization would have been better off without a measurement than with a faulty one." But this can be a difficult lesson for control-oriented cultures.
THE IMPLICATE ORDER
Lastly, there is also a level still more subtle than the deep learning cycle. This most subtle level is, however, also the most difficult to talk about. In fact, we may only infer its presence, since there is no tangible evidence of its existence. But ultimately it may prove vital to a full understanding of the deep shifts in awareness and capabilities of learning organizations.
The physicist David Bohm (one of the main contributors to the theory of dialogue) points out that the Western word "measure" and the Sanskrit "maya" appear to derive from the same origins. Yet, in the West, the concept of measure has come to mean ''Comparison to some fixed external unit," while maya means "illusion."
"In the prevailing philosophy in the Orient," says Bohm, "the immeasurable is regarded as the primary reality. In this view," he adds. "the entire structure and order of forms . . . that present themselves to ordinary perception and reason are regarded as a sort of veil, covering the true reality which cannot be perceived by the senses and of which nothing can be said or thought."
Bohm proposed a "new notion of order" to describe this deeper reality, the "implicate order," where "everything is enfolded into everything." In Bohm's view, the implicate order is continually ''unfolding into what we experience as the manifest world, "the explicate order." More importantly, human beings participate in this "unfoldment," as Bohm called it.
The most subtle aspect of "thinking strategically" lies in "knowing what needs to happen." This is extraordinarily difficult to describe. But I know that I and many others often feel that all we are ever doing is "listening" purposefully to what is needed. George Bernard Shaw said, "This is the true joy in life, [to be] used for a purpose recognized by yourself as a mighty one.... [to be] a force of nature ..." Could Shaw's "being a force of nature" relate to Bohm's "participation" in the "unfolding" of the implicate order?" Is this what happens when we develop our sensibilities in the deep learning cycle?
Such questions may hold a particular power as we stand here at the outset of the journey of learning about learning organizations. Bohm's quest in life was toward understanding the roots of fragmentation in our ways of thinking and being. "It should be said that wholeness is what is real, and that fragmentation is the response of this whole to man's action." Insofar as the quest for learning organizations might reestablish "the primacy of the whole" in human affairs, perhaps the quests are more intertwined than we can at present know.
