Peter M. Senge-The Fifth Discipline：A Shift of Mind-p451-459

1、43The Fifth Discipline: A Shift of MindPeter M. SengeSEEING THE WORLD ANEWThere is something in all of us that loves toput together a puzzle, that loves to see theimage of the whole emerge. The beauty ofa person, or a flower, or a poem lies in see-ing all of it. It is interesting that the words"

2、;whole" and "health" come from the sameroot (the Old English hal, as in "hale andhearty"). So it should come as no surprisethat the unhealthiness of our world today isin direct proportion to our inability to see itas a whole. Systems thinking is a discipline for seeingwholes

3、. It is a framework for seeing inter-relationships rather than things, for seeingpatterns of change rather than static "snap-shots." It is a set of general principlesdistilled over the course of the twentiethcentury, spanning fields as diverse as thephysical and social sciences, engineerin

4、g,and management. It is also a set of specifictools and techniques, originating in twothreads: in "feedback" concepts of cyber-netics and in "servo-mechanism" engineer-ing theory dating back to the nineteenthcentury. During the last thirty years, thesetools have been applied to u

5、nderstand awide range of corporate, urban, regional,economic, political, ecological, and evenphysiological systems.1 And systems think-ing is a sensibilityfor the subtle intercon-nectedness that gives living systems theirunique character.Today, systems thinking is needed morethan ever because we are

6、 becoming over-whelmed by complexity. Perhaps for the firsttime in history, humankind has the capac-ity to create far more information than any-one can absorb, to foster far greater interde-pendency than anyone can manage, and toaccelerate change far faster than anyone's ability to keep pace. Ce

7、rtainly the scale ofcomplexity is without precedent. All aroundus are examples of "systemic breakdowns"problems such as global warming, ozonedepletion, the international drug trade,and the U.S. trade and budget deficitsproblems that have no simple local cause.Similarly, organizations break

8、 down, despiteindividual brilliance and innovative prod-ucts, because they are unable to pull theirdiverse functions and talents into a produc-tive whole.Complexity can easily undermine con-fidence and responsibilityas in the fre-quent refrain, "It's all too complex for me,"or "Th

9、ere's nothing I can do. It's the sys-tem." Systems thinking is the antidote tothis sense of helplessness that many feel aswe enter the "age of interdependence." Sys-tems thinking is a discipline for seeing the"structures" that underlie complex situa-tions, and for di

10、scerning high from low le-verage change. That is, by seeing wholes welearn how to foster health. To do so, systemsthinking offers a language that begins by re-structuring how we think.I call systems thinking the fifth disciplinebecause it is the conceptual cornerstonethat underlies all of the five l

11、earning disci-plines. All are concerned with a shift ofmind from seeing parts to seeing wholes,from seeing people as helpless reactors toseeing them as active participants in shap-ing their reality, from reacting to the pres-ent to creating the future. Without systemsthinking, there is neither the i

12、ncentive norSource: From The Fifth Discipline by Peter M. Senge. Copyright © 1990 by Peter M. Senge. Used by permission ofDoubleday, a division of Bantam Doubleday Dell Publishing Group, Inc.451452Organizational Culture Reform Movementsthe means to integrate the learning disci-plines once they

13、have come into practice.As the fifth discipline, systems thinking isthe cornerstone of how learning organiza-tions think about their worldSophisticated tools of forecasting andbusiness analysis, as well as elegant strate-gic plans, usually fail to produce dramaticbreakthroughs in managing a business

14、. Theyare all designed to handle the sort of com-plexity in which there are many variables:detail complexity. But there are two types ofcomplexity. The second type is dynamic com-plexity, situations where cause and effect aresubtle, and where the effects over time ofinterventions are not obvious. Co

15、nven-tional forecasting, planning, and analysismethods are not equipped to deal with dy-namic complexity. Mixing many ingredi-ents in a stew involves detail complexity,as does following a complex set of instruc-tions to assemble a machine, or taking in-ventory in a discount retail store. But noneof

16、these situations is especially complexdynamically.When the same action has dramaticallydifferent effects in the short run and thelong, there is dynamic complexity. When anaction has one set of consequences locallyand a very different set of consequences inanother part of the system, there is dynamic

17、complexity. When obvious interventionsproduce nonobvious consequences, thereis dynamic complexity. A gyroscope is adynamically complex machine: If you pushdownward on one edge, it moves to the left;if you push another edge to the left, it movesupward. Yet, how trivially simple is a gyro-scope when c

18、ompared with the complex dy-namics of an enterprise, where it takes daysto produce something, weeks to develop anew marketing promotion, months to hireand train new people, and years to developnew products, nurture management talent,and build a reputation for qualityand allof these processes interac

19、t continually.The real leverage in most management situ-ations lies in understanding dynamic complex-ity, not detail complexity. Balancing marketgrowth and capacity expansion is a dynamicproblem. Developing a profitable mix ofprice, product (or service) quality, design,and availability that make a s

20、trong marketposition is a dynamic problem. Improvingquality, lowering total costs, and satisfyingcustomers in a sustainable manner is a dy-namic problem.Unfortunately, most "systems analyses"focus on detail complexity not dynamiccomplexity. Simulations with thousand ofvariables and complex

21、 arrays of details canactually distract us from seeing patternsand major interrelationships. In fact, sadlyfor most people "systems thinking" means"fighting complexity with complexity," de-vising increasingly "complex" (we shouldreally say "detailed") solution

22、s to increas-ingly "complex" problems. In fact, this is theantithesis of real systems thinking.The essence of the discipline of systemsthinking lies in a shift of mind: seeing interrelationships rather thanlinear cause-effect chains, and seeing processes of change rather thansnapshotsThe p

23、ractice of systems thinking startswith understanding a simple concept called"feedback" that shows how actions can rein-force or counteract (balance) each other. Itbuilds to learning to recognize types of "struc-tures” that recur again and again: the arms raceis a generic or archetypal

24、 pattern of escala-tion, at its heart no different from turf warfare between two street gangs, the demiseof a marriage, or the advertising battles oftwo consumer goods companies fighting formarket share. Eventually, systems thinkingforms a rich language for describing a vastarray of interrelationshi

25、ps and patterns ofchange. Ultimately, it simplifies life by helping us see the deeper patterns lying behindthe events and the details.SEEING CIRCLES OF CAUSALITYReality is made up of circles but we seestraight lines. Herein lie the beginnings ofour limitation as systems thinkers.The Fifth Discipline

26、: A Shift of Mind453One of the reasons for this fragmentationin our thinking stems from our language,Language shapes perception. What we seedepends on what we are prepared to see.Western languages, with their subject-verb-object structure, are biased toward a linearview.3 If we want to see systemwid

27、e interre-lationships, we need a language of interrela-tionships, a language made up of circles.Without such a language, our habitual waysof seeing the world produce fragmentedviews and counterproductive actionsasit has done for decision makers in the armsrace. Such a language is important in facing

28、dynamically complex issues and strategicchoices, especially when individuals, teams,and organizations need to see beyond eventsand into the forces that shape change.To illustrate the rudiments of the newlanguage, consider a very simple systemfilling a glass of water. You might think,"That's

29、 not a systemit's too simple." Butthink again.From the linear viewpoint, we say, "I amfilling a glass of water."But, in fact, as we fill the glass, we arewatching the water level rise. We monitorthe "gap" between the level and our goal,the "desired water level."

30、; As the water ap-preoches the desired level, we adjust the fau-cet position to slow the flow of water, untilit is turned off when the glass is full. In fact,when we fill a glass of water we operate ina "water-regulation" system involving fivevariables: our desired water level, the glass&#

31、39;scurrent water level, the gap between thetwo, the faucet position, and the water flow.These variables are organized in a circle orloop of cause-effect relationships which iscalled a "feedback process." The process op-erates continuously to bring the water levelto its desired level.Peopl

32、e get confused about "feedback"because we often use the word in a some-what different wayto gather opinionsabout an act we have undertaken. "Give mesome feedback on the brewery decision,"you might say. "What did you think of theway I handled it?" In that context, "

33、positivefeedback" means encouraging remarks and"negative feedback" means bad news. Butin systems thinking, feedback is a broaderconcept. It means any reciprocal flow ofinfluence. In systems thinking it is an ax-iom that every influence is both cause andeffect. Nothing is ever influenc

34、ed in just onedirection.Though simple in concept, the feedbackloop overturns deeply ingrained ideassuch as causality. In everyday English wesay, "I am filling the glass of water" withoutthinking very deeply about the real meaningof the statement. It implies a one-way cau-sality"I am c

35、ausing the water level torise." More precisely, "My hand on the fau-cet is controlling the rate of flow of waterinto the glass." Clearly, this statement de-scribes only half of the feedback process:the linkages from "faucet position" to "flowof water" to "wate

36、r level."But it would be just as true to describeonly the other "half" of the process: "Thelevel of water in the glass is controllingmy hand."Both statements are equally incomplete.The more complete statement of causalityis that my intent to fill a glass of water cre-ates a

37、system that causes water to flow inwhen the level is low, then shuts the flow offwhen the glass is full. In other words, thestructure causes the behavior. This distinc-tion is important because seeing only indi-vidual actions and missing the structure un-derlying the actions lies at the root ofour p

38、owerlessness in complex situations.In fact, all causal attributions made ineveryday English are highly suspect! Mostare embedded in linear ways of seeing. Theyare at best partially accurate, inherently bi-ased toward describing portions of recipro-cal processes, not the entire processes.Another idea

39、 overturned by the feedbackperspective is anthropocentrismor see-ing ourselves as the center of activities. Thesimple description, "I am filling the glass ofwater," suggests a world of human actorsstanding at the center of activity, operatingon an inanimate reality. From the systemsperspec

40、tive, the human actor is part of the feed-back process, not standing apart from it. Thisrepresents a profound shift in awareness. It al-lows us to see how we are continually both454454Organizational Culture Reform Movementsinfluenced by and influencing our reality. Itis the shift in awareness so ard

41、ently advo-cated by ecologists in their cries that we seeourselves as part of nature, not separate fromnature. It is the shift in awareness recognizedby many (but not all) of the world's greatphilosophical systemsfor example, theBhagavad Gita's chastisement:All actions are wrought by the qua

42、lities of na-ture only. 1 he self, deluded by egoism, think-eth: "I am the doer."4In addition, the feedback concept com-plicates the ethical issue of responsibility. Inthe arms race, who is responsible? Fromeach side's linear view, responsibility clearlylies with the other side: "

43、It is their aggres-sive actions, and their nationalistic intent,that are causing us to respond by buildingour arms." A linear view always suggests asimple locus of responsibility. When thingsgo wrong, this is seen as blame"he, she,it did it"or guilt"I did it." At a deeplevel

44、, there is no difference between blameand guilt, for both spring from linear per-ceptions. From the linear view, we are al-ways looking for someone or somethingthat must be responsiblethey can evenbe directed toward hidden agents withinourselves. When my son was four years old,he used to say, "

45、My stomach, wont let meeat it," when turning down his vegetables.We may chuckle, but is his assignment ofresponsibility really different from the adultwho says, "My neuroses keep me from trust-ing people."In mastering systems thinking, we give upthe assumption that there must be an in

46、di-vidual, or individual agent, responsible. Thefeedback perspective suggests that everyoneshares responsibility /or problems generated bya system. That doesn't necessarily imply thateveryone involved can exert equal leveragein changing the system. But it does implythat the search for scapegoats

47、a particu-larly alluring pastime in individualistic cul-tures such as ours in the United Statesisa blind alley.Finally, the feedback concept illuminatesthe limitations of our language. When wetry to describe in words even a very simplesystem, such as filling the water glass, it getsvery awkward: &qu

48、ot;When I fill a glass of water,there is a feedback process that causes me toadjust the faucet position, which adjusts thewater flow and feeds back to alter the waterposition. The goal of the process is to makethe water level rise to my desired level."This is precisely why a new language for de

49、-scribing systems is needed. If it is this awk-ward to describe a system as simple as fill-ing a water glass, imagine our difficulties usingeveryday English to describe the multiple feed-back processes in an organization.All this takes some getting used to. Weare steeped in a linear language for des

50、crib-ing our experience. We find simple state-ments about causality and responsibility fa-miliar and comfortable. It is not that theymust be given up, anymore than you give upEnglish to learn French. There are manysituations where simple linear descriptionssuffice and looking for feedback processesw

51、ould be a waste of time. But not when deal-ing with problems of dynamic complexity.REINFORCING AND BALANCINGFEEDBACK AND DELAYS: THEBUILDING BLOCKS OF SYSTEMSTHINKINGThere are two distinct types of feedback processes: reinforcing and balancing. Reinforc-ing (or amplifying) feedback processes arethe

52、engines of growth. Whenever you are ina situation where things are growing, youcan be sure that reinforcing feedback is atwork. Reinforcing feedback can also gener-ate accelerating declinea pattern of de-cline where small drops amplify themselvesinto larger and larger drops, such as the de-cline in

53、bank assets when there is a finan-cial panic.Balancing (or stabilizing) feedback oper-ates whenever there is a goal-oriented be-havior. If the goal is to be not moving, thenbalancing feedback will act the way thebrakes in a car do. If the goal is to be mov-ing at sixty miles per hour, then balancing

54、The Fifth Discipline: A Shift of Mind455feedback will cause you to accelerate to sixtybut no faster. The "goal" can be an explicittarget, as when a firm seeks a desired mar-ket share, or it can be implicit, such as a badhabit, which despite disavowing, we stick tonevertheless.In addition,

55、many feedback processescontain "delays," interruptions in the flowof influence which make the consequencesof actions occur gradually.All ideas in the language of systems think-ing are built up-from these elements, just asEnglish sentences are built up from nounsand verbs. Once we have lear

56、ned the build-ing blocks, we can begin constructing sto-ries: the systems archetypes.REINFORCING FEEDBACK:DISCOVERING HOW SMALLCHANGES CAN GROWIf you are in a reinforcing feedback system,you may be blind to how small actions cangrow into large consequencesfor betteror for worse. Seeing the system of

57、ten allowsyou to influence how it works.For example, managers frequently fail toappreciate the extent to which their ownexpectations influence subordinates' per-formance. If I see a person as having highpotential, I give him special attention todevelop that potential. When he flowers, Ifeel that

58、 my original assessment was correctand I help him still further. Conversely,those I regard as having lower potentiallanguish in disregard and inattention, per-form in a disinterested manner, and furtherjustify, in my mind, the lack of attention Igive them.Psychologist Robert Merton first identi-fied this phenomenon as the "self-fulfillingprophecy."5 It is also known as the "Pygma-lion effect," after the famous George nard Shaw play (later to become My FairLady)