warfield, john n - twenty laws of complexity - science applicable in organizations

& Research Paper

Twenty Laws of Complexity: ScienceApplicable in Organizations

John N. Warfield*

George Mason University, Institute for Advanced Study in the Integrative Sciences, Fairfax, Virginia, USA

The `LTI Set', consisting of 20 Laws of Complexity, a Taxonomy of the Laws ofComplexity, and five Indexes of Complexity, is proposed as the core of a developingscience of complexity that is applicable to resolving complexity in organizations. TheLTI Set links to these included topics:

. Alternative Science-Free Organizational Practices

. Educational Practices Appropriate to Complexity

. Quality Control of Science

. Applications of the Science of Complexity in Organizations

. Enabling Conditions for Effective Organizational Practice

A critical condition for significant advances in resolving complexity is that theorganization recognize the strong, even dominant, behavioral aspects of complexity, asreflected in the Laws; and take account of these in redefining the main role of topmanagement. That role is to set up and administer a responsive corporate infrastructureto meet the demands of complexity, along the lines set forth here. Further advances inbehavior can be made through new educational programs that reflect older scientificvalues applied to the challenges of today, in contrast to reliance on unwarrantedassumptions that undermine organizations. Appropriately remodeled to reflect therelentless demands of complexity, the university can become a model for otherinstitutions in society. Copyright # 1999 John Wiley & Sons, Ltd.

Keywords applications; complexity; management; organization; science of complexity; structure;university

INTRODUCTION

Three related contributions (a platform triad for ascience of complexity) are designated here in setform as {Twenty Laws of Complexity, a Taxonomy of

the Laws, Five Indexes of Complexity}. In abbre-viated form, the triad is referred to as the `LTI Set'(L for Laws, T for Taxonomy, I for Indexes).While the Briefs of the Laws represent the focalcontribution, they are aggregated in theAppendix to best use available space. The LTISet is augmented by a discussion aiming tointegrate these contributions in an organizational

CCC 1092±7026/99/010003±38 $17.50 Received 4 May 1997Copyright # 1999 John Wiley & Sons, Ltd. Accepted 24 April 1998

Systems Research and Behavioral ScienceSyst. Res. 16, 3–40 (1999)

*Correspondence to: John N. Warfield, George Mason University,Institute for Advanced Study in the Integrative Sciences, Mail Stop1B2, Fairfax, VA 22030-4444, USA.

context in society. The body of the paper isdedicated to the integrative discussion.

The Taxonomy aggregates the Laws into threenon-mutually exclusive categories, one of whichis further divided into three subcategories. Thesebreakouts help interpret what is occurring inapplications, and make possible discussion of thesignificance of the Laws in the body of the paper,without going overly into details. The Indexes ofComplexity provide mutually complementarynumerical measures of complexity in any part-icular application, frequently taking the value 1as the quantitative base of reference that separ-ates the ordinary (a number less than or equal to1) from the complex (a value exceeding 1, usuallyvery substantially in applications). Even thoughthe integrative goal dominates the body of thepaper, each set and subset of the LTI Set isidentified in full in the paper.

Linkages to the LTI Set

The LTI Set is intended to be the core of a scienceof complexity. But when standing alone it is notsufficient to drive the task begun here to its firstplateau. Even if a science of complexity werefully developed, much more than the sciencewould be needed in order to translate thatscience into a valuable social asset. Since thescience offered here is only partly developed it isnecessary, in order to stimulate independentassessment and enhancement of this work, tolink the LTI Set to these topics relevant to ascience of complexity:

. Alternative (Science-Free) Organizational Prac-tices: Practices carried out in organizationswithout reference to a science of complexity

. Educational Practices: Educational practices toequip people to become involved effectively indrawing value from the science

. Quality Control: Quality control in a science ofcomplexity

. Applications: Application of the science inresolving complexity

. Organizational Enabling Conditions: Organiza-tional conditions required to get maximumbenefit from applying the science

ALTERNATIVE (SCIENCE-FREE)ORGANIZATIONAL PRACTICES

It is well known that many practices arepresently advocated to promote constructiveorganizational change. They can be comparedand contrasted neither to belittle nor to enshrinethem. Instead, one begins with the firm beliefthat `one size does not fit all', and the heart of acomparative discussion ought to be mainlydevoted to discovering the best role for anygiven practice to fill, within the universe oforganizational change activities. Because thereare so many practices [end note 1, involving areference by Ackoff to `panacea overload'], astrategy of process role discovery is needed towhittle down the numbers to a manageable set.One component of such a strategy is to consideronly those practices that are offered by their advocatesas being relevant to resolving complexity. Many ofthe current practices do not involve that claim, sothey can be excluded from this discussion. Asecond component of the strategy is to discussonly those practices that, at least superficially,seem to be connected (or are said to be connected) toscience. With this two-component strategy as afilter, only five schools of thought about com-plexity appear to merit discussion here, and oneof those is distinguished by what on the surfaceseems to be an indifference to complexity. Thisdiscussion makes it possible to say why the LTISet is so critical in choosing a strategy forworking with complexity in organizations.When this strategy is used, attention automatic-ally focuses on top management of the organiz-ation, which is most likely to be concerned aboutresolving raging complexity, yet, paradoxically,least likely to be inclined to inject science into itsresolution.

The Changing Role of Top Management

Historically the complexity that faces corpor-ations and governments has been dealt with bytop management, as possibly its primary respon-sibility. A typical way to try to cope with thecomplexity is through strategic planning,possibly consisting of components that focus on

RESEARCH PAPER Syst. Res.

Copyright *c 1999 John Wiley & Sons, Ltd. Syst. Res. 16, 3–40 (1999)

4 John N. Warfield

particular time intervals. For example, theremight be a package of interrelated plans thatlook out ten years, five years, and/or two years;connected to and affecting a short-term oper-ational plan that focuses on the year ahead. Asplanning theory has matured, it has becomeacceptable to say that, for the most part, thepurpose of planning is not so much to say whatwill be done in advance, but rather to prepare themind of the executive for decisions that must bemade as time passes. Experience shows thatattempts to predict reliably what will happen inthe future and when it will happen does notgenerally produce commendable predeterminedactions. By having had some vicarious exposureto possible future events during the planningwork, the executive is presumed to become readyto tackle whatever comes along in the light ofsome previously nurtured insight derived, inpart, from that planning. When complexity is thefocus, and science is omitted from the realm ofdiscussion, assumptions take root as the replace-ment for science (often overtly, or even sub-consciously) [end note 2]. Some of theseassumptions can best be described as `killerassumptions'.

Killer Assumptions

Certain assumptions seem to be locked into thecorpus of the practice just described. In order forthat practice to be successful, those assumptionsmust be satisfied. Here are some of the killerassumptions that can undermine that practice:

. Adequate Executive Capacity. The executive hasthe intellectual capacity to comprehend:

(a) how the various interactions that take placeas situations change are interrelated;

(b) at what time actions should be initiated;(c) what alternatives are relevant when the

time comes to take those actions.

. Reliable Organizational Information Flow. Theflow of information in organizations is suffi-ciently coherent to permit the executive'sdecisions to be factored into day-to-day oper-ations reliably.

Keeping in mind that the focus here is oncomplexity, there is little knowledge to support theseassumptions in a scientifically respectable way.Considerable evidence exists that none of theassumptions mentioned is valid. Perhaps that iswhy large organizations are undertaking moreand more òff-sites' or ìnterventions' from out-side, in which small teams work together to tryto resolve situations that the older practicescannot cope with. If there is confusion at thehighest levels of management concerning howbest to proceed, perhaps a beginning can bemade by defining the requirements that acomplexity-resolving system for managingmight be asked to meet.

Defining Requirements for a Complexity-Resolving System

If present approaches to managing complexity(e.g., the US Medicare system, which is variouslysaid to involve many billions of dollars in fraudevery year) are unsound, one would hope thatresearchers would come forth with novel andtested alternatives. In searching for such altern-atives, one must constantly keep in mind thequestion: Ìs this alternative satisfactory forworking with complexity, or is it rather some-thing that is best suited to work with lessermatters?' Today, it appears that there is only onealternative (Interactive Management (IM) [endnote 3] whose advocates claim that it satisfies allof these Defining Requirements:

. Full Disclosure. Its explanation is widely avail-able in the literature.

. Replicability of Activity. The explanation is insignificant depth, so that replication ispossible.

. Specializing in Resolving Complexity. It is in-tended only for resolving complexity inorganizations.

. Sizeable Record of Value-Adding Application. Ithas been applied in many organizations andhas added significant value.

. Founded in Science. It is founded in science.

In its early stages, IM was buttressed by a`science of generic design', because it was felt

Syst. Res. RESEARCH PAPER


Twenty Laws of Complexity 5

that complexity is best resolved by well-thought-out, comprehensive designs. But it was alwaysfelt, as well, that design itself needed to befounded in a science of complexity; and thatthere would be considerable overlap between ascience of complexity and a science of genericdesign. Given that idea, it seems relevant toconsider the present state of thought aboutcomplexity, in the light of a study of variousschools of thought about it.

Schools of Thought about Complexity

The issue of whether other approaches focusspecifically upon complexity remains open. Butin an effort to close in on the issue, a study wascarried out to define `schools of thought' aboutcomplexity [end note 4]. The school that under-pins Interactive Management is called the `Struc-ture-Based School', because it proceeds from theassumption that it is the organization of com-plexity that must be the foundational way to startto resolve it. The Structure-Based School is notintended to compete with other schools. Rather,because it is more fundamental, it is intended tomake it possible for the other schools of thoughtto accomplish what is required in order for anyof those other schools to add value to a situationthat is rife with complexity. The Laws given hereare consistent with and underpin the `Structure-Based School'. The four other schools of thoughtto be identified here generally have not shownawareness of these laws, and might find theminconsistent with their views. Nevertheless, byknowing that these schools exist, the reader willbe better equipped to see what is presented herefrom multiple perspectives.

These five identified schools of thought con-cerning complexity appear to represent thevariety that is active at the present time:

. The Indifference School. Two (non-mutuallyexclusive) academic subgroups can bedescribed as examples of members of whatconstitutes a very large Indifference School (byno means confined to academia). These sub-groups are characterized either (a) by whatmay be called Interdisciplinary àpproaches' or`methods' (e.g. fostered by the Association for

Integrative Studies, a predominantly liberal-arts-faculty activity) or (b) as the `postmodern-ists' who appear to challenge all organizedknowledge. Neither subgroup recognizes com-plexity overtly in its philosophy or in itspractice.

Three schools may be characterized as today'sversion of the Comte-engendered `positivists' oras today's technical representatives of `scientism'as described in great detail [end note 5]. Theseschools prefer not to deal overtly with the logicfoundations of their models, but prefer rather toìnvent' methodologies that are not required to besupported by any underlying science. Theyconsist of the `Systems Dynamics' school, the`Chaos Theory' school, and the Àdaptive Sys-tems School'.

All actively promote models that essentiallyshut out the human being as an active com-ponent of a system, and draw much of theirreasoning directly by analogy with mathematicalsystems of varying types, mostly consisting ofsome form of differential equations. In thisrespect, they neglect to recognize many contri-butions that are reflected in Vicker's insightfulwork [end note 6], as well as in the Lawsthemselves.

. Systems dynamics (fostered by Jay Forrester,Dennis Meadows, Peter Senge, and othersoften associated with MIT, i.e., the Massachu-setts Institute of Technology, located in Cam-bridge, Massachusetts).

. Chaos theory (arising in small groups in manylocations).

. Adaptive Systems Theory (predominantly associ-ated with the Santa Fe Institute, but nowstarting to be associated with many schools ofbusiness or management).

The fifth school is the one reflected here:

. The Structure-Based School (developed by theauthor, his colleagues and associates, empha-sizing the collaborative, computer-assistedconstruction of structure of a problematicsituation as the key step in beginning toresolve complexity in that situation [endnote 7].



6 John N. Warfield

Table 1 summarizes the views of the authorabout these schools of thought. A key factor incomparing these schools of thought is theconcept of `formalism', i.e., an integrated systemof signs having the property of being uninter-preted (i.e., situation-independent) until associ-ations are made with this system [end note 8].

The vast majority of the proponents of SystemsDynamics, Chaos Theory, and Adaptive SystemsTheory hold a common point of view aboutcomplexity: that complexity is an aspect of thesystems which they explore. The differences amongthe three lie (a) in the particular formalismswhich underlie their thinking and (b) in theextent to which metaphors (e.g. `chaos', or`adaptive systems') are substituted for specificresults flowing from applications of the formal-isms.

In contrast, the Structure-Based School holdsthat complexity finds its locus in the humanmind, rather than in some corpus that the humanis striving to comprehend [end note 9]. Further,this School strives always to find adequate basesfor its assumptions in science and, especially inthe thinking of a collection of identified ThoughtLeaders, who are identified later in this paper.

Interactive Management

If it should prove true that the practice of Inter-active Management is consistent with a science ofcomplexity that is set forth, it would be mucheasier to display a comprehensive system, where

the science supports the practice, and resultsfrom the practice can be fed back to improve thescience. Moreover all of the empirical data fromthe use of Interactive Management could bebrought to bear as an evaluatory resource, andall of the learning that has gone on could poten-tially be applied to help validate the science.

Still, even assuming the best results in linkinga science of complexity to the practice of Inter-active Management, one cannot feel thoroughlycomfortable because there are certain `not-said'components that leave the entire conceptualcomplexity package open to criticism; suchopenings being found at both ends of thespectrum.

Overviewing the Spectrum of Complexity fromFoundations to Results

At one end, one must acknowledge that thequality of what arises from a science is depen-dent upon an infrastructure that supports thedevelopment of the science. If that infrastructureis not adequate to enable the science to beproperly conceived, and adequately supportedwith empirical evidence, that science is rightlyopen to extensive criticism. At the other end ofthe spectrum, one must acknowledge that it isnot sufficient for an organization merely to applyInteractive Management in an effort to resolveproblematic situations. Instead, the organizationmust provide the local infrastructure to supportboth (a) the conduct of the Interactive

Table 1. Schools of thought about complexity

Name of School Underlying formalism Where complexity lies

Indifference None Unspeci®ed

Systems Dynamics Ordinary differential equations In the system

Chaos Theory Ordinary non-linear differentialequations

In the system

Adaptive Systems Theory Partial differential equations In the system

Structure-Based Formal western logic; including settheory, theory of relations, digraphtheory, lattice theory, boolean methods,and the algebra of partitions

In the mind(Much of the foundational work isrepresented by the works of C.S. Peirce,J. Piaget, M. Polanyi, and G. Vickers)




Management work, and (b) the deployment ofthe products of that work in such a way that apattern of vertical linguistic coherence evolves inthe organization, allowing the products of the IMwork to be understood all the way up and downthe corporate or government ladder.

Without such an understanding, value can belost as quickly as it was gained. Moreover, afterthe IM work has been completed, infrastructuremust be present to support the conduct ofwhatever other methods or practices are requiredin order to build on and hopefully maximise thereturns from developing the products of IM.

A long conceptual distance is traversed inmoving from the infrastructure of science to thelocal infrastructure of an organization that isworking systematically to resolve its problematicsituations. Moreover, in order to sustain organ-izational effectiveness, a dynamic process of roleredefinition must take place and must becomeembedded in the organization, so that top-levelmanagers no longer see themselves as complex-ity resolvers, but rather as strategic enablers andoverseers of the quality of information flow inthe organization in a way that is commensuratewith the nature of complexity.

Unless the killer assumptions (some of whichare discussed above) that prevent the demands ofcomplexity from being met can be dissolved forthe using organization, the entire struggle maybe for naught in that organization.

EDUCATIONAL PRACTICES APPROPRIATETO COMPLEXITY: THE CURRICULUM OFCOMPLEXITY

The first of five highly condensed patterns to beintroduced (Figure 1) is called The Curriculum ofComplexity. The reader may interpret this patternas a broad hypothesis, set forth to provide acontext for a variety of topics relevant tocomplexity.

This pattern involves attempts to aggregateearlier research products, then organize themcarefully, to try to make the results more easilyunderstood, while emphasizing complexity asthe principle theme. The biggest difficulty indoing this is trying to achieve Argyris-type

`framebreaking' [end note 10] (i.e., to dissolvepatterns of belief that are not consistent with thelaws). If that can be done, then `remodeling' is alesser task. Because of the urgency in making thiswork understood, The Curriculum of Complexityhas been prepared to serve as an overview forlearning this material. The Curriculum incorpor-ates four `Courses':

. The Infrastructure of Science

. A Science of Complexity

. The Work Program of Complexity

. Implementation: Organizations and Complex-ity

The Infrastructure of Science is a necessary firstCourse, because without an understanding of theconditions that support the ongoing develop-ment and refinement of science, it is almostimpossible to explain the Science of Complexity.Without going back to foundations, such ascience cannot be seen in perspective. Alsowithout that understanding, one cannot reallycomprehend why the Work Program is laid outthe way it is, and why organizations have tobecome aware of the demands of complexity (asinferred from the Laws) in order to provide awork environment that is adequate for workingwith complexity.

The Science of Complexity falls into place if TheInfrastructure of Science is understood. The WorkProgram of Complexity is readily illustrated bymany applications. If the first three Courses arewell understood, the organization may take thesteps needed to make it possible to carry out TheWork Program of Complexity repeatedly, on a largescale, for many issues.

A Practitioner Learning Strategy

It is possible for practitioners to learn what is inCourses 3 and 4 without taking any interest inCourses 1 and 2. That is the route that mostpractitioners have taken in the past. Probably thegreatest benefit for practitioners from studyingCourses 1 and 2 would be that they gain theunderstanding to explain why the InteractiveManagement system is unique; and why othermethods that have been and continue to be used



8 John N. Warfield

cannot substitute for the powerful, computer-assisted structuring process called InterpretiveStructural Modeling (ISM) [end note 11]. Thisshould help them explain to upper managementwhy IM is needed in the organization. This isbelieved to be quite difficult, because uppermanagement historically has been called on tomanage the complexity, without having therequisite process support. On the one hand,they feel that it is their domain while, on theother hand, they do not know (a) that suchsupport is possible and (b) how to reconceivetheir role when the support becomes readilyavailable.

The pattern in Figure 1 has provided briefdescriptions of these four courses. Now we turnto each of the four areas in turn.

QUALITY CONTROL OF SCIENCE 1: THEINFRASTRUCTURE OF SCIENCE

As was seen in Figure 1, the Infrastructure ofScience focuses upon language, thought (includ-ing behavior), and structure. The inadequacy ofprose alone to represent complexity has beendiscussed elsewhere [end note 12]. The chosenlanguage of the Structure-Based School is acombined prose±graphics language. Thislanguage is defined in A Glossary of Complexity[end note 13].

Thought Leaders

In the Structure-Based School, thought isoften seen to involve what are called `Killer

Figure 1. The curriculum of complexity




Assumptions' which disable progress. Thoughtleaders [end note 14] have identified the philo-sophical basis for destroying those Killer As-sumptions. The latter have become prominentbecause of the lack of recognition of the relevanceof certain milestones in the history of thoughtwhich are shown in Figure 2. This condensed

pattern reflects the point of view that the mostintense form of scholarship lies in philosophy.The word `philosophy' is derived as a conjunc-tion of two Greek words meaning first `love' andsecond `wisdom', so that the term translates into`love of wisdom'. Only a small percentage of theworld's philosophers exemplify that term to the

Figure 2. Chronology: seven milestones in the history of thought



10 John N. Warfield

fullest. Boxes numbered 1±5 in Figure 2represent key philosophers who have contribu-ted to the thought component of the Infrastruc-ture of Science. Thought leaders who havecontributed heavily to understanding of theinfrastructure of science include Peter AbeÂlard,Aristotle, W. Ross Ashby, I.M. Bochenski, GeorgeBoole, Kenneth Boulding, Augustus De Morgan,Michel Foucault, J. Willard Gibbs, F.A. Hayek,David Hilbert, Antoine Lavoisier, Harold Lass-well, Gottfried Leibniz, F.S.C. Northrop, CharlesSanders Peirce, Alexander Pope, Geoffrey Vick-ers, and A.N. Whitehead [end note 14].

Evolution of Thought about Thought

What is seen in Figure 2 is the evolution throughtime of thought about thought. What is portrayedin this evolution is a slow drift toward formalismin expression. Formalism refers to the use of well-defined language as a means of expression. Theseven milestones shown here, when carefullyexamined, reveal a linked pattern of support andenhancement of this slow drift, beginning wellover 2000 years ago, and reaching a state ofpragmatic utility in the last decade of thetwentieth century. In this seventh state, computersupport is provided to enable aggregated andintegrated human belief to be developed, reveal-ing interpretive patterns to help human beingsresolve complexity.

Extensive research supports the pattern shownhere, not only in terms of relevant fundamentalliterature from appropriate disciplines, but alsoin terms of empirical evidence, found in a widevariety of applications, where today's interpret-ation of this pattern has been tested and found tobe highly effective in serving the needs found inapplications.

Behavioral Pathologies

Partly because of the lack of pervasive recog-nition of the milestones in the history of thought,various behavioral pathologies have taken strongroot in organizational cultures. These are ident-ified in Figure 3.

Four milestones in the study of behavior areshown there. Personal (individual) behavior,group behavior, and organizational behavior allsuffer from various pathologies that involveunwise assumptions. Individual pathologies pro-pagate into behavior of groups; and grouppathologies propagate into the organization.These group pathologies, as the subject ofstudy, foster the definition of laws of complexitythat can be empirically discovered within organ-izations, by observing the performance of thethree behavioral categories: personal, group, andorganizational. Interactive Management, in-formed by the 20 Laws of Complexity, offers acomprehensive response to the behavioral path-ologies of individuals, groups and organizations.The macro response consists of a rigorous,structured process. The micro response designsinto the process openness and flexibility. Thecarefully planned integration of rigidity andflexibility allows the 20 Laws to be interpretedin process terms, while maintaining consistencywith the mini-infrastructure required for creativeproductivity.

QUALITY CONTROL OF SCIENCE 2:THE SCIENCE OF COMPLEXITY

The science of complexity benefits substantiallyfrom The Infrastructure of Science. By overtlyconstructing the language, drawing on the in-sights from the evolution seen in the chronologyof the history of thought, and taking into accountthe views of the thought leaders that have beenidentified in the foregoing, it is possible toconstruct a science of complexity.

Clearly a whole science cannot be presented inthis paper, but two aspects of it can be high-lighted. First, the Laws themselves, presented inthe Appendix, present a significant component ofthe science. In addition to that, it is now possibleto present five Indexes of Complexity that can bequantified, and which have been computed invarious applications of Interactive Managementto problematic situations [end note 15].

Figure 4 is a condensed pattern that shows thefive computable Indexes of Complexity. Thesestand in a role similar to that of the physical




standards that lend integrity to the physicalsciences. With the ready possibility of computingand comparing values of these indexes, andreferring them back to the seven milestones inthe history of thought, and also to the milestonesin the study of human behavior, a type of closure

is obtained. This type of closure should beviewed as compelling in terms of framebreakingand remodeling of human enterprises; and in thereconstruction of human knowledge, as recom-mended in the works of Michel Foucault andothers [end notes 10 and 16].

Figure 3. Milestones in the study of behavioural pathologies



12 John N. Warfield

APPLICATIONS 1: THE WORK PROGRAMOF COMPLEXITY

The Work Program of Complexity has been under-going empirical testing in many applications formany years. Its application has provided avariety of discoveries that have gone into theformation of the Laws. The present status of theWork Program can be indicated by noting thevariety of organizations that are now active incarrying it out, applying Interactive Managementas the sole or as a major component of what isbeing done [end note 17].

The previously shown condensed patterns,and the knowledge stemming from the laws,allows the development of the condensed patternshown in Figure 5, which represents the WorkProgram of Complexity. This is the master action

plan, which begins with the use of InteractiveManagement, primarily to carry out the first andthird components of the Work Program: namelyDescription and Design. The second component,Diagnosis, is based on the Description, andnormally does not require the kind of groupactivity that is essential to the first and thirdcomponents. The fourth and final component,Implementation, can benefit greatly if the organ-ization establishes the necessary communicationsinfrastructure that is recommended: the Corpor-ate Observatorium [end note 18].

The Behavior-Outcomes Matrix

The connections between the Work Program ofComplexity and the Laws of Complexity are of

Figure 4. Five indexes of complexity




primary importance in assessing the applicabilityof the science itself as well as the implementingsystem called Interactive Management. Unlessclear and useful connections exist and areunderstood, once again the possibility of irrele-vance arises. An appropriate means for showingthese connections could be to construct a matrixthat would show, for each component of theWork Program of Complexity, which Laws areparticularly relevant, and in what way.

Such a matrix is presented in the Appendix(see Figure 6). It is called the Behavior-OutcomesMatrix. Not only does it show which Laws areparticularly relevant to which component of theWork Program of Complexity, but it also focusesthe relevance upon the particular behavioral

component involved (e.g., individual, group,organization, process).

APPLICATIONS 2: TAXONOMY OF THE LAWSOF COMPLEXITY

Because of the growing number of Laws ofComplexity, given in detail in the Appendix, ithas become timely to develop a taxonomy of theLaws, so that they can be discussed in theaggregate in order to help interpret their impact.Heading this taxonomy is the `Law TaxonomyTriad': {Behavior, Media, Mathematical}. The spe-cifics of the Taxonomy appear in the Appendix.(It is shown there that some of the Laws can

Figure 5. The work program of complexity



14 John N. Warfield

reasonably be placed in more than one category.)The following categories and subcategories havebeen chosen.

Type A: Behavior (70% of the Laws in ThreeSubcategories Involving Human Behavior)

. Habitual Behavior: Constraints that humanbeings have imposed upon themselves, almostwithout thinking, evolved through prolongedactivity.

. Physiological Behavior: Constraints on humanbehavior imposed by their physiology.

. Organizational Behavior: Aspects of humanbehavior that inhere from their participationin organizations.

Type B: Media (10% of the Laws in OneCategory Involving Communication Media)

. Linguistics of Communication: Matters that affectmedia of human communication, including

various aspects of human written or oralinteractions.

Type C: Mathematical (30% of the Laws in OneCategory Involving Mathematical Operations)

. Mathematical Operations: Laws arrived at bymathematical operations, whether theoreticalor empirically based, or both.

Distribution of the Laws Among the Categories

As is seen in the Appendix, 70% of the Laws fallinto the Behavioral category; while 30% fall intoMathematical and 10% fall into Media. (The readeris reminded that a few fall into more than onecategory, which is why the total percentagesexceed 100.) Since none of the Laws is based inthe so-called `hard sciences', it is reasonably clearwhy technologists are unlikely to appreciatethese Laws, and are likely to make arbitraryjudgments (often efficiency-based) that encroach

Figure 6. Behaviour±outcomes matrix




on the domain of these Laws. This should alsomake clear why, in discussing the five Schools ofThought about Complexity, the three that relateto differential equations have relatively little incommon with the Structure-Based School. Thelatter school, drawing its origins largely from theInfrastructure of Science, recognizes that humanbeings create structure, whether they are scien-tists or not, that they do so with the aid oflanguage, that language is of behavioral origin,and that behavior is a social phenomenon notreadily dealt with in terms of calculus orcalculus-based formalisms.

ENABLING CONDITIONS FOR EFFECTIVEORGANIZATIONAL PRACTICE: THEORGANIZATION'S INFRASTRUCTURE

The Laws of Complexity offer science that isapplicable in organizations. This may best beseen by noting that every single Law can beconsidered as an ÌF±THEN' construction. If thisseems a bit novel, one might momentarily retreatto the laws of physics. Though these laws areoften formulated as algorithms, essentially all ofthem embrace certain preconditions which, if notmet, render the algorithm irrelevant. It shouldnot be surprising that a similar situation holdsfor the Laws of Complexity.

The Positive Construction: Orienting the LawsToward Success in Organizations

Taking any one of the Laws, the ÌF' and the`THEN' can be extracted. Consider the first-listedLaw, The Law of Triadic Compatibility, fromAppendix A as an example. The Law speaksabout compatibility of the human mind toexplore interactions among a set of elements. IFthe number involved is three or less, or IF thetotal number of elements and interactions isseven or less, THEN the mind is compatible. ThePositive Construction from this Law is clear: ifthe individual can manage to keep the numbersdown, success in reasoning is a likely outcome.

The Negative Construction: Orienting the LawsToward Failure in Organizations

As mentioned before, the ÌF' and the `THEN'can be extracted from each Law. Consideringagain the first-listed Law, the negative construc-tion tell us that attempts to work simultaneouslywith more than the specified numbers are verylikely to fail. The Negative Construction is clear:avoid exceeding the threshold of mental effec-tiveness.

The Yin and the Yang

The dual nature of all of the laws, i.e., thepositive and negative constructions, tells us onthe one hand the conditions under which theorganization need not fear the impact of the Law;while on the other hand telling us that failure isvery, very likely if the organization does notcontrol the conditions under which the Law canoperate.

Keeping in mind that all the Laws can beoperating concurrently within a given context, itshould be clear that what is required is a masterprocess that has designed into its operations theconditions required to prevent any of theNegative Constructions from occurring. This isdone by controlling the ÌF'. In other words, thedesign disables the ÌF' component of the Lawand, thereby, prevents the `THEN' componentfrom having any negative impact. That is, inessence, the reason that Interactive Managementis successful in organizations when applied towork with complexity.

The Corporate Observatorium

Interactive Management, in the absence of anadequate corporate infrastructure, can only bemarginally valuable. No matter how high inquality the products of the application of Inter-active Management may be, if they are notdisseminated to, understood by, and put intopractice by those organizational employees whohave the talent and know-how to make theresults effective, the organization cannot reap thebenefits of the work. That is why the Corporate



16 John N. Warfield

Observatorium has been conceived as the meansof making those good things happen.

CONCLUSIONS

Reductionism and systems thinking can coexist,but not in harmony. Wherever reductionism issuccessful, it will be found that prior systemsthinking enabled a problematic situation to besorted into relatively independent dimensionalcomponents. Today such components are ap-plied in organizations both in management andin technology development. But they cannot besubstituted for an effective process for workingwith complexity.

The Laws of Complexity stand not as a set ofseparate ideas to be invoked piecemeal; ratherthey are deployed like a collection of highlyarmed armadas attacking organizations indivi-dually and collectively, as a warlike body. Unlessthe variety in this armada is countered dimen-sion-by-dimension by a process founded inintegrated science, the armada will carry the day.

When applied in organizations, the Lawsshould be dually recognized: the `IF' portion ofeach Law should be suppressed with respect toits negative implications; and enhanced withrespect to its positive implications. In the hurly-burly of practice, such an approach cannot bespontaneous, but must be worked out ahead oftime. Interactive Management has been con-structed to fill that description, being foundedwithin the budding science of complexity offeredhere. Moreover the quality of that science hasbeen referred both to a long history of scholarlydevelopment, as well as immediate means ofmeasuring complexity in particular problematicsituations. Finally a growing body of empiricalevidence in practice supports the views pre-sented here.

FUTURE RESEARCH

It is inherent in science that it requires adedicated body of scholars striving constantlyto detect its flaws, while simultaneously workingto enhance its strengths. This is the general tone

that future research should take in the domain ofcomplexity. More specifically, every idea offeredhere can be challenged and put to the test. Atpresent weaknesses are found in infrastructure atboth ends of the spectrum of human activity: theinfrastructure of science, and the infrastructure ofpractice. Especially now, it is feasible to under-take infrastructural studies in organizations thatstrive to benefit from an understanding of theLaws of Complexity. These studies should focuson large-scale learning, and on constant updatingof large structures of information. Finally, theuniversity, as a primary beacon for learning,should take advantage of its unique position insociety to champion what is presented here in itseducational programs.

END NOTES

[1] Ackoff, R.L. (1995). `Whole-ing' the partsand righting the wrongs. Systems Research12(1), 43±46. In this paper, Ackoff callsattention to a large number of methods ortools that have been advanced to assistorganizations, and suggests that they arenot broadly effective. This may well bebecause organizations try to apply them toresolve complexity, in spite of the fact thatthere is no adequate logic platform showingthat they are minimally adequate for thatpurpose.

[2] Ketner, K.L. (1973). An Emendation of R.G.Collingwood's Doctrine of Absolute Presupposi-tions, Graduate Studies Number 4, TexasTech Press, Lubbock, TX. This work in-cludes several other valuable references tothe foundations of presuppositions, supple-menting C.S. Peirce's well-known taxonomyof how belief is fixed.

[3] Warfield, J.N., and CaÂrdenas, A.R. (1994). AHandbook of Interactive Management, IowaState University Press, Ames, IA. Thisbook is aimed specifically at resolvingcomplexity. The book describes in greatvariety a system of management intendedto be applied intermittently in organizationsjust to resolve complexity in those organiza-tions.




[4] Warfield, J.N. (1996). Five schools ofthought about complexity: implications fordesign and process science. In Tanik, M.M.et al., (eds), Integrated Design and ProcessTechnology (IDPT, Vol. 2), Proc. Society forDesign and Process Science, Austin, TX,pp. 389±394. This paper identifies and de-scribes the five schools of thought, andemphasizes the weak foundations on whichfour of them rest.

[5] Hayek, F.A. (1955). The Counter-Revolution ofScience: Studies on the Abuse of Reason, FreePress, New York. In this masterpiece, Hayekdocuments in great detail the onset ofpositivism in France, and its baleful appli-cation in spawning totalitarian regimes. Hedescribes the rise of `scientism', an imposterin the clothes of science, and unveils theemptiness of the underlying assumptionsthat support positivism. Some readers mightmistakenly think that the Structure-BasedSchool of Thought is a positivistic school.Nothing could be more wrong. To under-stand fully the key distinctions, it would bewell to read several books that collectivelyoffer a thorough philosophical basis for theStructure-Based School. First, one coulddevelop an understanding of Charles San-ders Peirce's views on formal logic. Peirce,in the top rank of the world's logicians, sawthe application of formal logic as a means oflearning, i.e., constructing hypotheses forevaluation, as opposed to a formula formaking final choices. He felt that formallogic was one of three `normative sciences',the other two being aesthetics and ethics. Inhis frame of thought, at the deepest level ofindividual self-governance lies aesthetics,the final arbiter, operating ineffably throughthe senses. At the next higher level ofindividual self-governance lies ethics, oper-ating ineffably through the senses under thepowerful influence of aesthetics. Just abovethe level of ethics lies formal logic; theuniquely powerful system of reasoning thatinforms the senses, having the status of anormative science, because it embodies thelove of truth in its operating purposes andconsequences. Recent views on Peirce's

normative sciences come from the latePotter, V. (1996). Peirce's Philosophical Per-spectives, Fordham University Press, NewYork.

The importance of integrating the powersof the senses with the powers of reasoninghas also been given a modern treatment inSalk, J. (1985). Anatomy of Reality: Merging ofIntuition and Reason, Praeger, New York(originally published by Columbia Univer-sity Press). To bring these lines of mutuallyreinforcing thought into one context, thereader can turn to Vickers, G. (1983). HumanSystems are Different, Harper & Row,London.

[6] A thorough airing of responsibility and itsconnection to organizations is found byreading several books by Vickers, G.(1983). Human Systems are Different, Harper& Row, London; (1965, 1983). The Art ofJudgment: A Study of Policy Making, Harper& Row, London; (original 1965, Chapman &Hall); (1980). Responsibility: Its Sources andLimits, Intersystems, Seaside, CA.

[7] The Structure-Based School has revealedthree central findings about problematicsituations that arise, typically in organiza-tions:

. There exists a plethora of component pro-blems in the problematic situation (hence,expressions such as `begin by defining theproblem' and `problem solution' are bothempty and misleading).

. There exists widespread difference of beliefamong well-informed individuals aboutwhat is most important in a situation (aphenomenon identified as `Spreadthink').

. A large number of problem interdependenciesare seen among the (usually large) num-ber of problems that arise from the mindsof the observers of a situation.

[8] Flew, A. (1984). A Dictionary of Philosophy(2nd edn), St Martin's Press, New York,pp. 123±124. From this widely acknowl-edged work, Anthony Flew has definedthe important term `formalism', which isa central concept of science, its relativeobscurity notwithstanding:



18 John N. Warfield

. Formalism: 1. (mathematics) A view pio-neered by D. Hilbert (1862±1943) and hisfollowers, in which it was claimed that theonly foundation necessary for mathemat-ics is its formalization and the proof thatthe system produced is consistent. Num-bers (and formulae and proofs) wereregarded merely as sequences of strokes,not as objects denoted by such strokes.

[9] The Nature of Complexity. In his famouspaper published in 1878, entitled `How toMake our Ideas Clear', Charles SandersPeirce talked about false distinctions thatare sometimes made in discussing beliefs.He wrote the following:

One singular deception of this sort, whichoften occurs, is to mistake the sensationproduced by our own unclearness ofthought for a character of the object weare thinking. Instead of perceiving thatthe obscurity is purely subjective, wefancy that we contemplate a quality ofthe object which is essentially mysteri-ous . . .

. . . So long as this deception lasts, itobviously puts an impassable barrier inthe way of perspicuous thinking; so that itequally interests the opponents of rationalthought to perpetuate it, and its adherentsto guard against it.

We see here Peirce's fundamental view thathuman behavior strives to focus complexityaway from the person and onto whateverthe person is unable to comprehend, be ittangible or not. Furthermore he alerts us tothe `impassable barrier' that comes whenthat mode of behavior dominates what goeson.

[10] Argyris, C. (1982). Reasoning, Learning, andAction: Individual and Organizational, Jossey-Bass, San Francisco. C. Argyris, in his studiesof behavior in organizations, discusses `fra-mebreaking' and `remodeling' as two keyactions involved in organizational reform.These terms incorporate ideas of falseassumptions that have been integrated intoframed patterns for decision-making; and

the necessity to break these frames and re-model by building new framed patterns thatincorporate higher-quality assumptions, orbetter still established scientific knowledge.

Framebreaking is implicit in M. Foucault'sinsistence on upgrading knowledge: `Wemust renounce all those themes whosefunction is to ensure the infinite continuityof discourse and its secret presence to itselfin the interplay of a constantly recurringabsence [the ``not-said''].'

Remodeling is implicit in M. Foucault's:`The problem . . . is no longer one of lastingfoundations, but one of transformations thatserve as new foundations, the rebuilding offoundations.'

The idea embodied here was stated by I.Kant in his definition: `Enlightenment isman's release from his self-imposed tute-lage.'

[11] Interpretive Structural Modeling (ISM) isunique because it instruments (i.e., providesan enabling mechanism for) the formal logicthat has evolved over more than 2000 years,finally attaining a state of pragmatic utilitywith the availability of software that sup-ports this process. This process is describedin Warfield, J.N. (1976). Social Systems:Planning, Policy, and Complexity, Wiley,New York. A version of the ISM softwarecan be downloaded without charge, alongwith a user's guide, from the Web siteidentified in end note 13.

[12] Warfield, J.N. (1993). Procrustes is Alive andWell and Teaching Composition in the EnglishDepartment, IASIS, Fairfax, VA. This report,available from the author, shows clearly thestructural constraints involved in the use ofprose, and explains why these constraintsare not compatible with the representationalrequirements of complexity.

[13] Recent IASIS publications are available thatadd considerable depth to these ideas. Theyare typically available at cost of productionand shipping. Titles can be seen at this Website: http://www.gmu.edu/departments/t-iasis. One of those publications by J.N.Warfield is entitled A Prose±GraphicsGlossary of Complexity.




[14] Thought Leaders identified here, and theconcepts that they have made available tothis work, can be seen in Appendix 1 of TheWandwaver Solution, which can be seen atthe Web site identified in end note 13.

[15] Staley, S.M. (1995). Complexity measure-ment of systems designs. In Ertas, A.,Ramamoorthy, C., Tanik, M., Esat, I.,Veniali, F. and Bendiab, T. (eds), IntegratedDesign and Process Technology, in ConferenceProceedings, Society for Design and ProcessScience, December, pp. 153±161. This pub-lication summarizes nicely several of theIndexes of Complexity, and provides tabulardata from a series of Interactive Manage-ment Workshops carried out at the FordMotor Company in Dearborn, MI, and inthe UK.

[16] Foucault, M. (1993). (A.M. Sheridan Smith,translator from the French) The Archaeologyof Knowledge and the Discourse on Language,Barnes & Noble, New York (earlier public-ation in 1969 and 1971). This work must beread for its highly articulate expression ofwhere thought goes astray today, because itis based in foundations that are unsustain-able, and which need to be reconstructed, asdiscussed in end note 9.

[17] The empirically sensitive reader can gofurther in exploring what goes on in thepractice of Interactive Management (IM).Table 2 identifies individuals and locationsthat are presently active in this area. (Theauthor will supply more details to ques-tioners, if requested.)

[18] Warfield, J.N. (1996b). The corporate obser-vatorium: sustaining management com-munication and continuity in an age ofcomplexity. In Tanik, M.M. et al., (eds),Integrated Design and Process Technology(IDPT, Vol. 2), Proc. Society for Design andProcess Science, Austin, TX, pp. 169±172.

This paper discusses the infrastructuralrequirements of large organizations thathope to have vertical linguistic coherencein their internal communications. By con-structing an observatorium built from theproducts of applications of Interactive Man-agement, both overview and details of

organizational future activities can bemade visible to everyone, benefiting fromthe creative use of real estate.

APPENDIX: BRIEFS OF THE LAWS OFCOMPLEXITY

This Appendix presents Briefs for each of theLaws of Complexity. The principal connectionsbetween the Laws of Complexity and the WorkProgram of Complexity are shown in Figure 6, theBehavior±Outcomes Matrix (BOM). The contentsof a matrix cell are the names of the Laws that areparticularly relevant to the intersection of acomponent from the Behavioral Menu and acomponent of the Work Program of Complexity.

In addition to the study of which Laws relateto which activities, shown in Figure 6, it is alsoappropriate to recognize that an understandingof some of the Laws will make it easier tounderstand other of the Laws. This idea plays arole in constructing a learning sequence. Inpresenting the Briefs for the Laws, a particularlearning sequence has been developed which, itis hoped, will make it easier to develop anintegrated perspective on them. This sequenceappears in Figure 7. Because this document isprepared for persons who are unfamiliar withthe Laws, no attempt will be made at this pointto justify the learning pattern shown.

Taxonomy of the Laws of Complexity

As mentioned earlier, the 20 Laws have beenplaced into categories. Table 3 shows the categ-ories to which the various Laws have beenassigned.

The Revolving Door to Enlightenment

Omar Khayyam (mathematician and astronomer,1048±1122), speaking through the voice ofEdward Fitzgerald (1809±1883), vented his frus-tration stemming from his attempts to getunderstanding, as follows:

Myself when young did eagerly frequentDoctor and Saint, and heard great argument



20 John N. Warfield

Table 2. Practitioner organizations

Organization and contact(s) Origins and/or notes on activities

Centre for InteractiveManagement India (CIMI)Dr Surinder K. Batra

Dr Batra visited George Mason University while employed by Tata Ltd, and spentsix weeks observing IM in action. After returning to India, he completed hisdoctoral degree, completed his employment with Tata Consulting Services, andstarted his own business as listed above. He has worked with a wide variety ofclients on several continents

Complexity Solutions Limited(CSL)Bill Rodger, North AmericaMichael McMaster, Europe

Complexity Solutions Limited (CSL) was set up in 1997. It provides managementconsulting services to clients in government and industry, and licenses of itsprocesses to consultants and corporations. CSL provides software technologydeveloped by Desyma Decision Technologies Inc. to consultants and corporationsinterested in acquiring capability in the application of IM to the resolution ofcomplex issues. CSL licenses the software, trains facilitators in its use and providescoaching on the use of IM to speci®c situations and to building a consultingpractice. The company has licensed consultants specializing in the petrochemicalindustry, medical service provider networks and major organization change effortsin the UK and USA

CWA LtdDr Alexander Christakis,Dr Alex N. Pattakos

CWA Ltd has developed the CogniScope2 system. It requires ®ve ingredients for itswork: (1) Community of Stakeholders; (2) CogniScope2 Team; (3) CogniScope#

software for recording ideas and meanings; (4) Consensus Methods, selected fromthe universe of available problem-solving and design methods, on the basis oftechnical and behavioral criteria for productive dialog leading to action; and (5) ACollaborative Facility which promotes the comfort and contributions of theparticipants as stakeholders, and has the capability to display visually the observa-tions constructed through the dialog. CWA has had a large variety of clients, manyfrom the pharmaceutical industries. It played a major role in assisting the US Foodand Drug Administration in receiving the `Innovations in Government Award', aprized award that is comparable to the Deming Award in terms of its signi®cance.CWA boasts a long history of IM work, much of which took place before CWAstarted as a business. This includes work with American Indian tribes, US ForestService, United Fund, and others. At present CWA is working in Greece to trainhigh-level administrators in the government of Greece

Desyma Decision TechnologiesInc.Bill Rodger

Desyma was set up in 1989 to develop software used in IM. The product developed,SYNERGISTIC SOLUTIONS, has been used by Desyma since 1990 to providemanagement consulting services to clients in Canada, the USA and Europe. Thecurrent version being used is the second-generation product and is based on theexperience acquired by Desyma in applying IM in a variety of industry, govern-ment, academic and non-pro®t organizations. Desyma is currently completing thedevelopment of a new, third-generation product which will add additionalfunctionality needed by both the process facilitators and process participants.New capabilities will include multiple site support for electronic conferencing.Desyma has a VAR agreement with Complexity Solutions Limited for the marketingof SYNERGISTIC SOLUTIONS worldwide

ITESM Campus MonterreyProfessor Roxana CaÂrdenasDept. Systems Engineering,Instituto Tecnologico deMonterrey

Teaching and consulting services in IM have been provided at ITESM CampusMonterrey for about a decade. This includes consulting with industrial ®rms,carrying out projects related to government, teaching satellite classes in genericdesign science, doing short courses for other branch campuses of ITESM, andcollaborating with Professor John War®eld of George Mason University on manyprojects. These included co-author of The Handbook of Interactive Management,facilitation at Ford Motor Company, facilitation at the CSIR in Accra, Ghana; andfacilitation at George Mason University in short courses offered there. Roxana has anumber of experienced collaborators at ITESM Monterrey, including CarmenMoreÂno, who has assisted in conducting defense systems management workshopsin the USA, as well as facilitating a variety of workshops in Mexico

The Jeffrey GroupCarol Jeffrey

The Jeffrey Group has conducted Interactive Management work related to reorgan-ization of the science establishment in Ghana, and to demobilization and recon-struction in Liberia, as well as other projects of considerable variety

Table 2 continued over page




About it and about: but evermoreCame out by that same door where in I went.

The late contemporary French philosopher andchairman of the history of systems of thought atthe ColleÁge de France, Michel Foucault (1926±

1984), in his masterpiece on the `archaeology ofknowledge', believed (as described by D.W.Harding)

that our own current intellectual life andsystems of thought are built on assumptions

Table 2. Continued

Organization and contact(s) Origins and/or notes on activities

ITESM Campus LeoÂn ReynaldoTrevinÄ oCentro de Estudios Strategicos

In 1997, this group facilitated IM with people of different sectors and withfunctionaries of the municipal administrations of four Mexican counties: San Diegode la Union, Purisima del Rincon, Romita and Tierra Blanca. We examined thegeneral problematique of those counties, including their rural communities, and gotdeep into the analysis of their economic sectors, which are different in each case,and also into the analysis of their critical factors, like education, health, jobs,communitarian services, public security, public administration, water resources±management±distribution, ecology, etc. The diagnoses resulting from those sessionswere used for designing collaborative actions in each case, and these were latersystematized under general strategies which were also prioritized to give them anorder in the Integral Municipal Development Plan for each one of the counties. Themain products were: (1) a general and speci®c diagnosis, which was complementedwith statistical data and its interpretation for each one of the counties; and (2) theirMunicipal Development Plan, 1998±2002, according to ®ve big categories: EconomicDevelopment, Social Development, Educational Transformation, Conditions of Lawsand Regulations, and Good Governmental Practices. These Plans also containGeneral Objectives to be achieved for each big category, and suggested performancemeasures for each single action to be implemented

Now this group is working in planning the Socioeconomic Development for themost important county in Guanajuato in regard to economic activities and numberof inhabitants, i.e., LeoÂn, and soon will begin the same kind of task as describedpreviously, for Comonfort, Penjamo, Moroleon, Allende, and perhaps also forYuriria, Cortazar, Tarandacuao and Huanimaro. This will be a huge work, probablyrequiring more than 80 IM workshops, just to provide the basis for elaborating theirplans, 1998±2003. This link of ITESM to the communities is dictated by the ITESMMission

Professor Benjamin Broome,George Mason University

Professor Broome can be contacted through the George Mason University homepage at http://www.gmu.edu/departments

He has about a decade of experience with IM applications. In recent years he hasfacilitated IM workshops at Ford Motor Company, and in Cyprus, where he hasworked for over two years on a Fulbright Award to help bring peace to the warringfactions on Cyprus. As a result of that work he is now being asked to provide moreand varied services related to IM, both in the USA and in Europe

Phrontis LimitedTony Gill

Phrontis was set up in April 1994. It has systems thinking, in the widest sense, atthe core of its services in consultancy, training, facilitation and managementresources. Clients come from a number of industries, notably the Oil & Gas andTelecoms industries. A recent assignment involved risk assessment and transpar-ency in the Swedish Nuclear Industry using a combination of systems thinkingapproaches. Since Tony participated in a workshop on IM led by John War®eldduring 1995, Phrontis has been involved in the planning and delivery of IMworkshops for NatWest Bank and British Telecom. We see IM as an importantmanagement tool to help clients structure intractable problems involving multiplestakeholders. We have done an IM Workshop with NatWest Bank as part of aresearch project in July 1996 and another in November 1996 with Bill Rodger'sinvolvement for British Telecom European Marketing Group in Brussels. Phrontishas a web page devoted to IM: http://www.phrontis.com/facilim.htm



22 John N. Warfield

profoundly taken for granted and not nor-mally exposed to conscious inspection, andyet likely in time . . . to be discarded.

In amplifying that view, Foucault states that:

The manifest discourse, therefore, is really nomore than the representative presence of whatit does not say; and this `not-said' is a hollowthat undermines from within all that is said.

It is none too early to try to begin to correct theshortcomings in our knowledge that stem frombad (`not-said') assumptions, uncritically ac-cepted and propagated, especially in academia.Too much is at stake. Yet, such acceptanceseemingly continues its relentless advance, com-pounded by technologies that typically showlittle friendliness to their users.

The 20 Laws to be described are intended to beresponsive to Omar's complaint.

In this document, the Laws of Complexity thathave been discovered to date are presented in theform of `Briefs'. A Brief is given for each, showingthe name of the Law, its origins, references torelevant literature, the statement of the Law, andits interpretation. The list of references for theBriefs appears at the end of the set of Briefs.

There is a dual logic involved in this arrange-ment. From one point of view, the brief of theLaw should be given before it is applied toparticular situations. From another point of view,application of the Law to a specific situationsheds light on how to perceive it when it isformally presented.

As presented here, the Laws appear in aspecific learning sequence that is correlatedwith Figure 7. This sequence is believed tosupport efforts to learn the set of Laws. BecauseFigure 7 is not a linear graphic, a linear learning

Figure 7. Interactions in interpreting laws of complexity




sequence in which each Law is studied once inturn is not appropriate for learning the Laws.

Brief 1A: The Law of Triadic Compatibility

Origin(s) of Law: Empirical, (Miller, Simon);Mathematical [Applied Lattice Theory],(Warfield).

References: Miller (1956), Simon (1974), Warfield(1988).

Statement of law: The human mind is compatiblewith the demand to explore interactions among a setof three elements, because it can recall and operatewith seven concepts, these being the three elementsand their four combinations; but compatibility cannotbe presumed for a set that both has four members andfor which those members interact.

Interpretation of law: This Law expresses both ahuman limitation and a human capability. Thelimitation suggests that human beings cannotprocess interrelationships among sets of factors,issues, objects, or ideas in general, if more thanthree components are involved. The reason setforth is that the mind is incapable of recalling

into its short-term memory more than aboutseven items. (A set comprised of three elementsand the four interacting combinations of themwill consist of seven members.)

A way to show respect for this limitation is todetermine that whenever decisions are to bemade that can benefit from awareness of inter-actions, it will be advisable to choose and apply astrategy that recognizes the impact of thislimitation.

This limitation should also persuade individ-uals that intuitive decision-making, carried outwithout careful analysis, is likely to produce baddecisions and bad outcomes when complexity ispresent.

The capability that allows us to processinterrelationships among three factors Ð issues,objects, or ideas Ð should encourage us to beginto develop a facility for working with sets ofthree items. More specifically, it would beadvisable to build up a repertoire of sets ofthree items that are representative of decision-making situations, and develop skill at workingwith these sets.

It might be well to remember that manysituations in life have been approached as though

Table 3. Categories of the laws of complexity

Law of Behaviorally based (70%) Media-Based(10%)

Mathematicallybased(30%)Habitual Physiological Organizational

1A Triadic Compatibility d d

1B Requisite Parsimony d

2 Structural Underconceptualization d d d

3A Organizational Linguistics d

3B Vertical Incoherence d

4 Validation d

5 Diverse Beliefs d

6 Gradation d

7 Universal Priors d

8A Inherent Con¯ict d

8B Limits d

8C Requisite Saliency d d

8D Success and Failure d

8E Uncorrelated Extremes d

8F Induced Groupthink d d

9 Requisite Variety d

10A Forced Substitution d

10B Precluded Resolution d

11 Triadic Necessity and Suf®ciency d

12 Small Displays d d



24 John N. Warfield

there were a dichotomy involved. Instead ofallowing our thinking to be limited to dichoto-mies, we should be encouraged to move totrichotomies as a way of becoming more flexiblein thought and action, wherever appropriate.

We may also be persuaded that documentationis much more valuable than might be thought,especially if the documentation takes the form ofrepresenting systems of interactions involvingmore than three interacting members.

When we have developed patterns of inter-relationships as documentation, we may work todevelop the skill of reviewing and amendingsuch patterns. Moreover, we may begin to seemerit in group development of interrelationshippatterns, since there is little in the capability towork with three items that suggests an over-whelming power of a single individual toconstruct patterns of interrelationship that arerepresentative of actual conditions or systems, orof contemplated conditions or systems.

The limitation to interactions among threeitems suggests a very serious limitation oncreative ability as might be reflected in designof complex systems. Ad hoc designs, arrived at inordinary conversational modes (as, for example,in governmental bodies or committees) might belooked upon as unlikely to be of high quality,and likely to produce bad outcomes.

Brief 1B: The Law of Requisite Parsimony

Origin(s) of law: This law is based on thedynamics of interpreting and learning impliedby the Law of Triadic Compatibility. The Law isprescriptive, with the aim of allowing enoughtime for sequentially presented information to beinterpreted in terms of the interactions, and toallow enough listening time to help ensure thatthe information is remembered.

References: Miller (1956), Simon (1974), Warfield(1988).

Statement of law: Every individual's short-termbrain activity lends itself to dealing simultaneouslywith approximately seven items (a number that isreached with three basic items and four of their jointinteractions) [see the Law of Triadic Compatibility].

Attempts to go beyond this scope of reasoning are metwith physiological and psychological Limits thatprelude sound reasoning. For a given designer, thereis some number Kd that is characteristic of thatdesigner which typically is chosen from the set [5, 6,7, 8, 9] that represents the Limit of that designer'sshort-term idea-processing capability. If a designmethodology requires a designer to cope intellectuallyat any one time with some number of concepts Kc ,then

(i) If Kc5Kd , the designer is underburdened,being uninfluenced by the Law of RequisiteParsimony, since the designer is operating in aSituation that exhibits the Requisite Parsimony,through regulation of the rate of flow ofinformation to the designer as the designerengages in the design process.

(ii) If Kc� Kd , the designer is operating at thelimit of reasoning capability.

(iii) If Kc4Kd , the designer is overburdened andno reliance can be placed on the designer'sdecisions.

Interpretation of law: If the Law is not violated,it has no impact. If it is violated, it can be con-fidently predicted that the design Target (i.e.,whatever description or product the individualor group seeks to achieve) will embody badoutcomes that are beyond the control of thedesigner, because the design process did notexhibit the Requisite Parsimony, but insteadallowed the rate of flow of information to thedesigner to exceed processing capacity.

It may be questioned why designs havesucceeded in the past without overt adherenceto this requirement. Design Targets vary signifi-cantly in their scope. If the Law of RequisiteParsimony is being unknowingly violated, onewould expect that the impact would be revealedin the failure of large system designs. This isprecisely what is being observed all around theworld.

Those who deny the validity of and those whodoubt this Law must accept the burden ofproviding other explanations for failures. Theoften-rendered explanation `operator error' mayoften, itself, reflect the same fundamental causeto which this Law responds in terms of thedesign process.




Brief 2: The Law of StructuralUnderconceptualization

Origin(s) of law: Empirical, Mathematical logic.

References: Warfield (1979, 1991a, 1994).

Statement of law: No matter what the complexissue, and no matter what the group involved in itsstudy, the outcomes of ordinary group process (i.e.,process in which computer support for developing theformal logic structure of the issue is lacking) will bestructurally underconceptualized (as evidenced, forexample, by the lack of delineation of the cycles and ofany structural connections among them).

Interpretation of law: A proper interpretation ofthis Law requires an understanding of thefundamental nature of structure. The term`structure' is widely used by economists in avery loose way, virtually as a semi-metaphor.This widely practiced usage simply serves to puta veneer on top of what can be very preciselydefined. A proper interpretation of structurerefers to how individual substantive componentsof information or knowledge are related.

To understand the foundations of relation-ships, one needs to know something about thehistory of the development of what is called `thetheory of relations' or `the logic of relatives'. Themost foundational work done in this area wascarried out by a British professor, Augustus DeMorgan, who published his treatise in the year1847. Listen to how his work was described(sometime around the year 1867) by America'sgreatest philosopher, Charles Sanders Peirce:

a brilliant and astonishing illumination ofevery corner and every vista of logic

But the direct connection of De Morgan's workto the structure of information or knowledge didnot become crystal clear until the publication of abook by Professor Frank Harary and two juniorcolleagues at the University of Michigan in 1965[Frank Harary, R.F. Norman, and D. Cartwright,Structural Models: An Introduction to the Theory ofDirected Graphs, Wiley, New York].

In this book, Harary showed that any givenrelation corresponds directly to a particulargraphical structure; and that every relation

corresponds to some form of directed graph(`digraph') (although the utility of such acorrespondence seems limited to the so-called`transitive relations', a very large class).

Taking that information, Warfield showed inhis 1976 book [Societal Systems: Planning, Policy,and Complexity, Wiley, New York] that the mostgeneral form of digraph representing a relationexhibited several attributes:

. A hybrid structure.

. Exactly two distinctive prototypical substruc-tural types in a hybrid structure identifiable aseither (a) hierarchy or (b) cycle.

. A numerical measure of the length of anyhierarchical substructure.

. A numerical measure of the length of any cyclesubstructure.

. A numerical measure of the `width' of ahierarchy, giving a numerical meaning to`linear structure' as a hierarchy of width 1;and giving a numerical interpretation to theidea `breadth of relationship'.

. A numerical measure of the structural com-plexity of a relationship based upon thestructural features of the hybrid structure(and this measure of complexity has sincebeen joined by several new ones discovered inrecent years).

In that same work, Warfield presented a varietyof algorithms for developing such structureswith computer assistance, providing a method-ology that allowed the structures of complexissues to be created by groups of knowledgeablepeople and, thereby, opening the way for thestructure of knowledge to take its rightful placeamong the analytical and synthetic conceptsavailable to people to analyze complex issuesor systems, and to design such complex systemsin a way that would make clear how the designsrelate to the issues themselves, thereby eliminat-ing the need for vagueness in the design of suchsystems as health care systems, systems fordealing with other public policy matters, aswell as providing a similar benefit in the designof physical systems such as automobiles.

But in spite of these developments and thebroad-ranging nature of the benefits that could



26 John N. Warfield

be attained by taking advantage of them, only arelatively small number of people have learnedabout this area and have begun to take advant-age of what is known.

In the laboratory work done by Warfield andhis colleagues, a considerable amount of datawas taken based on work done by numerousgroups with a variety of complex issues. Thesedata showed typical attributes of the structuresdeveloped. It was found that over 97% of allstructures were hybrid structures (i.e., theycontained at least one cycle).

The most evident proof of structural under-conceptualization in dealing with complex issuesand complex systems is the failure even toidentify the cycles that are present in the struc-tures. A lesser evidential point is that frequentlythe hierarchical substructures are not identified.The set of hybrid structures required to show theunderlying structure of information is virtuallynever constructed. If and when the set is con-structed, it may be superficially rendered (as inSenge's `archetypes') in non-operational form,leaving the actor to invent, without recom-mended processes, the more substantive andparticular instances relevant to a given situation;which seemingly asks the actor to become aprocess inventor.

Hence the wide-ranging scope of the Law ofStructural Underconceptualization. But it mustbe realized that structural underconceptualiza-tion always implies underconceptualization. Thesituation is as though a human body werepresented without any skeleton. We would seea limp corpse with no definition of human shape.It is only the structural feature of the body thatallows it to be erect or elongated, and providesthe basis for its overall appearance.

How frightening can it be that in virtuallyevery major public issue or virtually every largesystem design seen in our society, the structuraldescriptions are not even comprehended,and not made available for view and interpreta-tion?

How disappointing and how demoralizing canit be that the latter is not being done, eventhough it is perfectly feasible to do it, to do itefficiently, and to do it in a responsible, high-quality way?

Brief 3A: The Law of OrganizationalLinguistics

Origin(s) of law: Empirical.

References: None.

Statement of law: As an organization grows,linguistic separation grows both laterally and verti-cally in the organization. At the higher vertical levels,metaphors and categories become progressively dis-connected from the relevant components at lowerlevels, leading to decisions based on perceivedrelations between categories that are not borne outby relations between the members of those categories.

Interpretation of law: Imagine that a group ofpeople is formed by selecting several individualsfrom the human race at random. Then supposethat exhaustive information is obtained whichwill reveal what language components areshared by every single member of that group. Itmay be that one of the members came from aremote tribe in Australia, and another came froma similarly remote tribe in the mountains of Peru.It may be that there is virtually no languagecomponent that is shared by the group.

The term `linguistic domain' was applied bythe Chilean scientist Maturana to describe thelanguage commonality among a group of people.

It is easy to see that if a certain groupeffectively defines their linguistic domain, theextent of that domain will normally shrink asnew members become attached to the group,unless some specific actions are taken to restoreor enlarge the linguistic domain. Enlargementwould require that every single member of thegroup take on the addition, whatever it might be,in order that an enlarged linguistic domain couldbe said to exist.

Now imagine a large organization which ishierarchically organized so that conversationsmostly occur within rather than across organiz-ational layers. Each one of these layers corre-sponds to a certain linguistic domain that holdswithin that layer. Yet as human turnover occursin a layer the linguistic domain of that layerchanges.

The maintenance of a linguistic domain relieson (a) usage of the existing domain to keep it at




the forefront of each individual's usable lang-uage, and (b) upgrading of the domain everytime a new member enters, in order to prevent itsdeterioration.

Since big organizations do virtually nothing tomaintain even a single linguistic domain, it isinevitable that over time people will only be ableto talk to one another knowingly in a given layerand then, only in the fractional terms that remainafter the natural progressive deterioration thatgoes on in these domains.

But deterioration of linguistic domains is oftenless affected by change in the human make-up ofthe relevant groups than it is by changing tech-nology. In many industries, technological changecauses significant demands to be made to incor-porate new terminology in a linguistic domain,yet the technological terminology is often sopoorly defined or so foreign that assimilation ofit into a given domain can only be done if theorganization pays the price. The price that has tobe paid is that time of the affected individualsmust be dedicated to human interaction aimed atrenewing and strengthening the linguisticdomain.

But even if linguistic domains are strengthenedin a few layers of a large organization, stillanother phenomenon becomes critical. What isbeing talked about in the lower levels of theorganization are often highly detailed subjects,these subjects never being discussed at that levelof detail in the higher levels of the organization.If there is going to be linguistic connectionbetween levels, one must recognize that themetaphors and categories (the high-level organ-izational language) have to have a strongcorrelation to the detailed information (the low-level organizational language), and that thiscorrelation has to be sustained and renewed con-stantly in order to preserve meaningful com-munication across organizational boundaries.

Empirical observation of groups who work atdifferent levels in organizations has shown thatthe relationships that high-level people constructand apply among metaphors and categoriessimply do not correlate with the lower-levelideas that high-level people assume are encom-passed within those metaphors and categories.The result is that the decisions and actions taken

at high levels in organizations often amaze theoperating levels because they make so little senseand vice versa. The reason for the lack of sensib-ility is that both levels are operating in whatmight be called incongruous linguistic domains.

Until organizations understand the necessityfor maintenance, renewal and cross-organiz-ational development of linguistic domains,those ever-present trends that work againsteffective communication will continue to beresponsible for what appears to the externalobserver as organizational incompetence.

Brief 3B: The Law of Vertical Incoherence inOrganizations


References: Alberts (1995), CaÂrdenas and Rivas(1995), Warfield (1995).

Statement of law: For any large organization (thatis unaware of this Law), there are invisible-but-potentially-discoverable patterns of vertical coherenceawaiting discovery; which when discovered, will showhow key features of that organization are (a) many innumber, (b) can be structured into categories that aremuch fewer in number, and (c) whose categories canbe structured into areas that are again much fewer innumber. The features include problems, availablesmall-scale options for resolving the problems, andother element types yet to be discovered. Thisstructure will be an `inclusion structure' from theclass of Application Structural Types described in AScience of Generic Design.

Interpretation of law: This Law relates to what Ihave called the `Alberts Pattern', based on workdone first by Henry Alberts and later by RoxanaCaÂrdenas and Jose Rivas.

Further elaboration of this Law indicates thatthe three levels in the Pattern can typically becorrelated strongly with what have historicallybeen called the Operational Level, the TacticalLevel, and the Strategic Level; and those in turncan typically be correlated with Front-LineManagement/Labor, Middle Management, andTop Management.

Still further elaboration comes from evi-dence that these three stratified levels are not



28 John N. Warfield

adequately understood in organizations, and thatmany bad organizational decisions at each levelcome about because of this lack of understand-ing.

Brief 4: The Law of Validation

Origin(s) of law: Philosophy of science, asoriginated by Charles Sanders Peirce.

References: Churchland (1986), Deely (1991),Goudge (1969), Warfield (1994).

Statement of law: The validity of a science dependsupon substantial agreement within the scientificcommunity of meaning at its highest grade, i.e.,meaning attained through Definition by Relationship.

Interpretation of law: Many philosophersbelieve that they understand the concept ofvalid knowledge. The people that they like torefer to include Auguste Comte, Thomas Kuhnand Karl Popper. Others, who hear these peoplebeing named as the origins of the appropriateviews about what constitutes valid knowledge,are likely to accept their views without question,based on the assumption that the philosophershave adequately explored the presuppositionsunderlying the views of people such as Comte,Kuhn, and Popper.

In contrast, the mature philosophy of CharlesSanders Peirce presents a philosophy of sciencethat is not consistent with any of the foregoing.Moreover, John Deely has clarified what iswrong with the popular view of scientificvalidity, and has made clear why the popularview that there exists observer-independent`objective knowledge', which has a higherquality than ordinary knowledge, is misbegotten.

To get a hearing, one must proceed as follows:

. Explain why the prevailing views are wrong(that takes quite a bit of time and argument).

. Explain what the appropriate views are (thatrequires quite a bit of background from thelistener, which most of them who are `college-educated' lack).

. Explain why the latter views are appropriate.

Since that can't be done in a short space, we haveto appeal to the reader and make a promise to

the reader. The appeal is to suspend belief in thecommonly accepted ideas and take an interest inexploring another point of view. The promise isthat the reader who will spend enough timestudying the matter can get virtually all of theimportant ideas from the references given here.

Brief 5: The Law of Diverse Beliefs


References: Warfield (1991b, 1993, 1994, 1995).

Statement of law: Whatever the group, whatever thecomplex issue being considered by the group, at theoutset of group consideration of the issue, the indi-vidual members of the group will have quite diversebeliefs about the issue; and the probability is high thatthis situation will remain undiscovered and uncor-rected, in the absence of a group learning experienceusing a methodology whose power to produce thenecessary learning has been scientifically validated.

Interpretation of law: In order for people toshare a common point of view about a complexissue, several conditions must be met. The age-old philosophical Doctrine of Necessity under-pins this idea:

. People must all share the same linguisticdomain, in order that they can even conceiveand express jointly and sincerely the samepoint of view.

But in complex areas, empirical evidence showsthat people do not share an adequate linguisticdomain, and frequently cannot even understandthe initially expressed points of view of othersbecause they lack the substantive backgroundknowledge or experience to do so. If they do noteven share a common meaning of a critical wordor phrase, they will not be able to express anyshared point of view that they might hold, oreven test whether they hold a shared point ofview.

Even in the instances where they do share thesame linguistic domain (which our researchshows to be a rare situation, much rarer thanalmost anyone would likely believe until anopportunity is made available to observe appro-priate human interactions), some believe that




dramatic differences of opinion are a con-sequence of very different value orientations ofindividual people. But consider this. If people donot share a linguistic domain that is broadenough to enable them to express and share acommon point of view, it will never be possibleeven to determine the existence or relevance ofthe influence of presupposed different valueorientations. Therefore, even if the proponentsof the differing values theory are correct, there isno way for them to establish their correctness inthe absence of prior conditions.

Since the evidence of lack of sharing anadequate linguistic domain is compelling, onemust give credence to the former. But even if thislack is discounted, the empirical evidence showsvery clearly the absence of shared belief amonggroups of people who are supposedly knowl-edgeable in areas. So whatever the reason for thisabsence, the Law stands as an empirical fact.

This Law should compel a certain kind ofbehavior on the part of leaders who see value indeveloping a shared point of view. The kind ofbehavior that is required is to create conditionswhereby the linguistic domain of that group ofindividuals whose diversity of views creates un-manageable or ineffective organizational condi-tions is enlarged to the point where it becomesfeasible to enunciate and share a point of view.

If the leadership is unable or unwilling to dothis, at least the leadership should recognize thevalue in knowing that virtually all individuals inthe pertinent group have quite diverse beliefsabout any complex issue. The potential benefitthat may be seen by a leader is to enter the policyor action vacuum and promote one's own pointof view based on the held authority. Thepotential dysbenefit that might be seen by aleader is that the leader is usually no differentfrom any other of the relevant individuals. Theleader's views are just as likely to be unsatisfac-tory as those of any of the others.

Brief 6: The Law of Gradation

Origin(s) of law: Theory of Relations: InclusionRelation.

References: None.

Statement of law: Any conceptual body of knowl-edge can be graded in stages, such that there is onesimplest stage, one most comprehensive stage (reflect-ing the total state of relevant knowledge), andintermediate stages whose content lies between thetwo extremes.

. The Corollary of Congruence. The first Corollaryto this Law asserts that the class of situations towhich a conceptual body of knowledge may apply,in whole or in part, likewise may be gradedaccording to the demands that individual situationscan reasonably make upon the body of knowledge.This is called the Corollary of Congruence,because it relates to the congruence between theDesign Situation and Target with a restricted gradeof the Generic Design Science that is called intoplay in the specific case. Clearly the designer is notrequired to uncover every detail of relevance, nomatter what the cost. When in doubt, a conserva-tive posture will call for erring on the side of thehigher grade.

. The Corollary of Diminishing Returns. The secondCorollary to this Law is the existing economicLaw of Diminishing Returns, which states thatthe application of a body of knowledge to a DesignSituation should be made through that stage atwhich the point of diminishing returns to theSituation (as opposed to only the user) is reached.This is called the Corollary of DiminishingReturns, and it highlights a major responsibilityof the designer to make judgments about when thispoint is reached. Once again, a conservative posturewill call for erring on the side of the higher grade.

. The Corollary of Restricted Virtual Worlds. Thethird Corollary to this Law states that theidentification of the stage at which diminishingreturns to the situation is reached normallyrequires the integration of the Virtual Worlds ofthe affected parties in the situation in relation to thedimensions of the situation. This is called theCorollary of Restricted Virtual Worlds, and itreflects the need for a global point of view inmaking the kinds of judgments that are required toachieve the appropriate congruence of gradation.

Interpretation of law: The importance of thisLaw to the Science of Generic Design lies inthe guidance that it provides to the designerconcerning how to perceive any particular



30 John N. Warfield

Design Situation with respect to the Science. Inthis respect, one notes that design Targets mayrange from very small, limited-scope Targets tovery large, broad-scope Targets.

It is not reasonable to take as a criterion forGeneric Design Science that all of its Theory andall of its Methodology should be demonstrablyrequired for all design activity. On the contrary,such a Science would be too brittle for use. TheLaw of Gradation overtly recognizes that DesignSituations and Design Targets are themselvesgraded according to a variety of descriptions, notall of which can be foreseen. Accordingly, theScience of Generic Design should be appliedjudiciously, extracting from it one of its stagesthat is most appropriate for the particular DesignSituations and Design Target.

The word `generic' does not mean `alwaysrequired'. What it does mean is `covering the setof gradations of Design Situations and Targets asa whole, without overlapping the applicableSpecific Design Science; but subject to judiciousrestriction commensurate with the grade of theDesign Situation or Design Target in any part-cular instance'.

It is not the function of a Science of GenericDesign to provide a recipe appropriate to everyDesign Situation. It is the function of such aScience to actuate the designer's professionalresponsibility to assess and correlate the grada-tion in the Situation and Target against the totalsweep of the Generic Design Science, and tochoose that restricted version of the Sciencewhich will be used openly, rather than to acceptsubliminally a restricted version that leads tounderconceptualization of the Design Situationand the Design Target. It is the further functionof the Generic Design Science to provide themeans of documentation consistent with whatthe Design Situation requires.

Brief 7: The Law of Universal Priors


References: Warfield (1994).

Statement of law: The human being, language,reasoning through relationships, and archival

representations are universal priors to science(i.e., there can be no science without each of them).

Interpretation of law: The Doctrine of Necessity.The validity of this Law can be established usingwhat is called the Doctrine of Necessity. ThisDoctrine holds that, independent of the particu-lar attributes of B, if A is necessary in order for Bto exist, then A is a prior of B. (The word `prior',used as a noun, fills a need that no other wordquite satisfies.) The test of the necessity of each ofthe four factors mentioned is to imagine that theyare withdrawn, and then inquire as to whether intheir absence a science is possible.

. The Human Being. Imagine first that there wereno human beings. Accepting the commonevidence that human beings are the producersand the only producers of science, then it mustbe that the human being is a Universal Prior toScience.

. Language. Imagine next that no language wereavailable. Since all of science consists oflanguage, and nothing other than language,there can be no science without language.

. Reasoning Through Relationships. Suppose nowthat there is no reasoning through relation-ships. Since all organization of information isthrough relationships arrived at by reasoning,there can be no organization of knowledgewithout it. But science is organized knowl-edge, hence both language and reasoningthrough relationships are Universal Priors toScience.

. Archival Representation. The human being,language, and reasoning through relationshipsall can exist and persist without any archivalrepresentation, the organization being in themind. It might, therefore, be argued that thesethree are sufficient, and that archival repres-entation is not required in order for organizedknowledge and, therefore, science to exist. Butscience depends upon widespread consensus,and library after library attests to the criticalimportance of archival representation in gain-ing the necessary widespread understandingand consensus upon which acceptance asscience depends.

. Absence of Foundations. Overt recognition of thestatus of the Universal Priors to Science should




bury the modest movement to assert that thereare no foundations to (at least some) sciences.On the contrary, what is seen here not onlystates that there are some, but there are somethat are foundations to all science. If one is todistinguish one science from another, it may bethrough finding unique foundations for aparticular science that can and must beintegrated with the Universal Priors to estab-lish the decision-making basis for the particu-lar science.

. Diminution of Universal Priors. One obvious,but misguided, way to try to provide distinc-tiveness to the foundations of a science is to laythe Universal Priors on the operating table,and to diminish them to shadows of theiridentity, while retaining slices of them. Thusthe human being may be fractionated into aneconomic entity, a social entity, or other one-dimensional entity such as political, athletic,biological, etc., or through a role such asobserver of nature. Language may be dilutedby failure to establish and enforce the defi-nitions of its components, and reasoningthrough relationships may be diluted both byblurring the definitions of the relationshipterms and by disguising patterns of relation-ship. The latter can occur naturally because ofthe linear sequential nature of prose, whichdoes not lend itself to portraying patterns.Archival representations may themselves be sodiluted by the emaciation of the other threeUniversal Priors as to be helpless to offer anyassistance in searching for Referential Trans-parency.

Brief 8A: The Law of Inherent Conflict


References: Warfield (1991b, 1993, 1994, 1995).

Statement of law: No matter what the complexissue, and no matter what the group involved in itsstudy, there will be significant inherent conflictwithin the group stemming from different perceptionsof the relative significance of the factors involved inthe complex issue.

Interpretation of law: The interpretation of thisLaw is essentially the same as that for the Law ofDiverse Beliefs. The latter offers an explanationfor the Law of Inherent Conflict. Because thebeliefs are diverse, there is inherent conflictwithin the group. The two Laws mentionedhere are complementary and can often be seen asa composite that could be called the Law ofDiverse Beliefs and Inherent Conflict. Never-theless it is believed that the modest redundancyinvolved is not adequate justification to repealthe decision to present the two Laws separately.Each Law offers its own unique point of view.

In order for people to share a common point ofview about a complex issue, and thereby avoidconflict on that issue, several conditions must bemet. The Doctrine of Necessity underpins theidea that one condition is:

. People must all share the same linguistic domain,in order that they can even conceive and express thesame point of view.

But in complex areas, empirical evidence showsthat people do not share an adequate linguisticdomain, and frequently cannot even understandthe initially expressed points of view of othersbecause they lack the substantive backgroundknowledge or experience to do so. If they do noteven share a common meaning of a critical wordor phrase, they will not be able to express anyshared point of view that they might hold, oreven test whether they hold a shared point ofview.

Even in the instances where they do share thesame linguistic domain (which our researchshows to be a rare situation, much rarer thanalmost anyone would likely believe until anopportunity is made available to observe appro-priate human interactions), some believe thatdramatic differences of opinion are a con-sequence of very different value orientations ofindividual people. But consider this. If people donot share a linguistic domain that is broadenough to enable them to express and share acommon point of view, it will never be possibleeven to determine the existence or relevance ofthe influence of presupposed different valueorientations. Therefore, even if the proponentsof the differing values theory are correct, there is



32 John N. Warfield

no way for them to establish their correctness inthe absence of prior conditions.

Since the evidence of lack of sharing anadequate linguistic domain is compelling, onemust give credence to the former. But even if thislack is discounted, the empirical evidence showsvery clearly the absence of shared belief amonggroups of people who are supposedly knowl-edgeable in areas. So whatever the reason for thisabsence, the Law stands as an empirical fact.

This Law should compel a certain kind ofbehavior on the part of leaders who see value indeveloping a shared point of view. The kind ofbehavior that is required is to create conditionswhereby the linguistic domain of that group ofindividuals whose diversity of views createsunmanageable or ineffective organizational con-ditions is enlarged to the point where it becomesfeasible to enunciate and share a point of view.

If the leadership is unable or unwilling to dothis, at least the leadership should recognize thevalue in knowing that virtually all individuals inthe pertinent group have quite diverse beliefsabout any complex issue. The potential benefitthat may be seen by a leader is the opportunity toenter the policy or action vacuum and promoteone's own point of view based on the heldauthority. The potential dysbenefit that might beseen by a leader is that the leader is usually nodifferent from any other of the relevant individ-uals. The leader's views are just as likely to beunsatisfactory as those of any of the others.

Brief 8B: The Law of Limits


References: None.

Statement of law: To any activity in the universethere exists a corresponding set of Limits upon thatactivity, which determines the feasible extent of theactivity.

. The Corollary of Active Limits. The first Corollaryto this Law asserts that for any particular situation,the set of Limits can be partitioned into two blocks:an active block and an inactive block. ThisCorollary is called the Corollary of Active Limits.

The active block is the subset of the set of Limitsthat is determining at a given time, while theinactive block is not determining at that time. Theactive block may often consist of a single, dominat-ing member of the set of Limits. Such a membermay be so strong in its power to limit that, in effect,all other Limits are forced into hiding by thedominant one. When this occurs, it has bothadvantages and disadvantages. An advantage isthat the designer who recognizes this situation canfocus attention upon the dominant Limit and lookfor ways to modify its impact. A disadvantage isthat the non-active Limits may go unrecognized,only to make their impact felt later upon the designactivity that has focused overly on overcoming thedominant Limit.

. The Corollary of Movable Limits. The secondCorollary to this Law asserts that the set of Limitsalso can be partitioned into these two blocks:movable and fixed. A movable limit is one thatcan be altered, while a fixed limit is one that isunchanging. Clearly if there is a dominant Limitand it is fixed, the potential exists for wastingsubstantial amounts of time, effort, and resources ifone does not understand that it is fixed. On theother hand, if one mistakenly assumes that a Limitis fixed, when it really is movable, the potentialexists for missing opportunities for major improve-ments. This Corollary is called the Corollary ofMovable Limits.

. The Corollary of Discretionary Action. The thirdCorollary to this Law asserts that the movablesubset of Limits can be partitioned into these twoblocks: movable through discretionary action bypeople, and autonomously movable. Limits thatare autonomously movable change on their own,and thereby drive the system. Clearly the strategicposture for dealing with such Limits is to maintaincognizance of their status and to have somepredetermined alternatives in mind for copingwith them when they move into prominence. ThisCorollary is called the Corollary of DiscretionaryAction. Limits that are movable through discre-tionary action by people are, of course, those thatshould be clearly recognized by designers, and towhich attention should be given in the event thatthey are not overshadowed by more prominentLimits that effectively nullify the latent impact ofthose lying in the background.




. The Corollary of Shifting Limits. The fourthCorollary to this Law asserts that the membershipof the active blocks and of the inactive blocks of thepartitions changes with time. If, for example,discretionary action brings about a change insome movable Limit that previously was dominant,one or more new Limits will take the place of thepreviously dominant Limit. This is the Corollaryof Shifting Limits.

Interpretation of law: The significance of thisLaw to the Science of Generic Design is that itconveys the importance of discovering (a) whatthe Limits may be upon design in general andhow these Limits may relate to any particularDesign Situation and (b) those additional Limitsthat are at work in a particular Design Situation.This Law and its Corollaries impose uponTheory the requirement that it contain explicitidentification of generic Limits and explicitprovision for the incorporation of special Limits.

The Law of Limits itself provides no means ofidentifying the Limits or of partitioning themafter they have been identified. This capabilitymust arise from other sources.

Brief 8C: The Law of Requisite Saliency


References: Boulding (1966), Warfield (1994).

Statement of law: The situational factors thatrequire consideration in developing a design Targetand introducing it in a Design Situation are seldom ofequal saliency. Instead there is an underlying logicawaiting discovery in each Design Situation that willreveal the relative saliency of these factors.

Interpretation of law: Kenneth Boulding ident-ified three major reasons for poor intellectualproductivity. These are: spurious saliency (empha-sizing the wrong things, out of proportion towhat they deserve); unproductive emulation (be-having like those who help create rather thanresolve problems); and cultural lag (not usingestablished knowledge with dispatch). Charac-teristically individuals who become involved inthe design process exhibit great diversity in theirassessment of relative saliency (as indicated in the

data in Appendix 5 of A Science of GenericDesign). This diversity, if uninfluenced bythorough exploration of the Design Situation,will support unfocused dialog, unjustified de-cisions, and arbitrary design outcomes not likelyto be understood or even actionable.

The design process must incorporate specificprovision for uncovering the relative saliency ofthe factors in the Design Situation and the factorsthat characterize the Target, in order to achievethe kind of understanding that is needed to putthe design factors in proper perspective.

Brief 8D: The Law of Success and Failure

Origin(s) of law: Mathematics of discrete prob-ability.

References: Warfield (1958, 1965, 1968).

Statement of law: There are seven critical factors inthe SUCCESS BUNDLE for the Design Process.Inadequacy in any one of these factors may causefailure. The seven factors are: leadership, financialsupport, component availability, design environment,designer participation, documentation support, anddesign processes that converge to informed agreement.

Interpretation of law: This Law indicates that aScience of Generic Design must define the criticalfactors in sufficient depth to enable (a) theassessment of their adequacy and (b) theirapplication in the Design Situation. Success andfailure must also be elaborated and, in thiscontext, success in all stages of work, includingthe implementation and operation, is required inorder to proclaim that the design is successful;while failure in any stage is sufficient toconstitute failure of the design.

This Law furnishes the impetus for what iscalled the Sigma-N Concept, discussed in detailin Sec. 6.9 of A Science of Generic Design.

Brief 8E: The Law of Uncorrelated Extremes

Origin(s) of law: Empirical, Statistical Analysis.

References: Kapelouzos (1989), Warfield (1994),Warfield (1995).



34 John N. Warfield

Statement of law: No matter what the complexissue, and no matter what the group involved in itsstudy, the initial aggregate group opinion concerningthe logical pattern of the factors involved in the issueand the final aggregate group opinion concerning thelogical pattern of the factors involved in the issue (i.e.,the views at the two extremes of the application of theGeneric Design Science, before and after), will beuncorrelated; showing that significant learning takesplace through the application of the generic designprocesses.

Interpretation of law: Once it was discoveredthat there was very great diversity in the views ofindividual members of groups about the relativeimportance of elements that were generated andclarified using the Nominal Group Technique(NGT), a research question arose about thepersistence of such views after additional workwas done. Since such elements were typically thesubject of structuring work using InterpretiveStructural Modeling (ISM), a natural way toapproach this question involved a comparison ofthe results obtained using ISM with the productsof the voting done as part of the use of NGT.

While NGT is not regarded as a structuringtool, nevertheless a structure can be producedfrom the results of NGT voting. Here is how sucha structure can be produced. For each problemelement that gets at least one vote from aparticipant in group work using NGT, makenumerical assignments to those votes as follows:

The highest rated element coming from someindividual's voting gets a score of 5; thesecond highest rated element from that sameindividual's voting gets a score of 4; and soon, until the fifth-rated (lowest-rated) elementgets a score of 1. Assigning such scores forevery individual's votes, one can then com-pute a cumulative score for each element thatreceived a vote. A structure can then becreated using the relationship `has a higherscore than', and this relationship can beregarded as equivalent to ìs perceived bythe group as a more important problem than'.

Conducting the ISM session with the sameelements, typically a problematique structure isproduced using the relationship àggravates'. A

method of scoring was developed whereby eachproblem element in the problematique receives acertain net score. This net score reflects theposition of the problem element in the problem-atique, and takes into account the number ofother problems that are aggravated by a givenproblem, as well as the number of problems thataggravate a given problem. Typically problemslying at the left will get larger scores than thoselying at the right of the structure.

It then becomes possible to correlate the scorescoming from the NGT-derived structure with thescores coming from the ISM-derived structure.While the two structures do not share preciselythe same relation, one is justified in presumingthat those problems that aggravate many otherproblems are more important than the ones theyaggravate because of their power to sustain thoseother problems; while those that are, in effect,continued with increased power to do harm byothers are regarded as less important. Import-ance thereby takes on a priority status, andreflects the potential strategy of attack, workingfirst on those that have the greatest power toaggravate other problems. Some data exist toshow that this strategy has been very successful.

Judge I.B. Kapelouzos, who was on sabbaticalleave from his position on the Council of State ofGreece, decided that he would study 31 cases forwhich data were available to examine the before-and-after correlation for such structures. The`before' structure was the one produced fromthe NGT work, and reflects composite groupresults just before the ISM work begins. Theàfter' structure is the one that is available afterthe ISM work is finished.

People familiar with these concepts assumedthat the work would show some variation fromperfect correlation. Everyone was surprised tosee the results. The results showed no correlationbetween the before-and-after structures. Thestartling nature of this result could only beexplained in one way: the individuals in thegroup, after having gone through a rigorousexamination of the relationships among theproblems, learned a great deal about how thoseproblems interact (something which they couldnot readily do otherwise, as indicated by the Lawof Triadic Compatibility); and as a result the type




of strategy indicated by their product changeddramatically as a result of this learning.

These research results imply, among other things,that all those methodologies currently in vogue forgroup work, which do not incorporate in their tool kitthe ISM process whereby detailed examination ofinteractions among elements is carried out by thegroup, are sorely deficient and are likely to producevery misleading and dysfunctional conclusions.

Much more is learned from the process ofdetailed study of interactions among elementsthan intuition had suggested. Even though ISMwas specifically designed to be a learningprocess, it was not envisaged that it wouldhave the power which the Law of UncorrelatedExtremes attaches to it. Further research alongthese lines should be very valuable in adding toour limited knowledge concerning such matters.

Additional evidence to support the con-clusions embodied in this Law are reflected inthe Law of Structural Underconceptualization,where the data show unequivocally that individ-uals and groups do not even produce structureswithout the help of the ISM process, as discussedin the Interpretation of that Law!

Brief 8F: The Law of Induced Groupthink


References: Allison (1971), Janis (1982), Warfieldand Tiegen (1993), Warfield (1994).

Statement of law: The pathological behaviordescribed as `Groupthink' (e.g., in the work of IrvingJanis, and in Graham Allison's study of The Bay ofPigs incident), can be predictably induced in groupsby the behavior of individuals who put pressure ongroups to produce results under a time limit; wherecomplexity is paramount.

Interpretation of law: This Law, related toGroups, has its parallel in the Law of ForcedSubstitution, intended to explain certain beha-vior of a top-level executive.

The case study work done by Graham Allison,and the more elaborate and detailed case studiesreported by Warfield and Tiegen in the 1993document titled `Groupthink, Clanthink,

Spreadthink, and Linkthink', illustrates whatthis Law says.

Brief 9: The Law of Requisite Variety

Origin(s) of law: Mathematics, Systems Theory.

References: Ashby (1958), Warfield (1986), War-field and Christakis (1987), Warfield (1994).

Statement of law: A Design Situation embodies arequirement for Requisite Variety in the designspecifications. Every Design Situation S implicitlyrepresents an (initially unknown) integer dimension-ality Ks such that if the designer defines an integer Kmnumber of distinct specifications (whether qualitativeor quantitative or a mix of these), then:

(i) If Km5Ks , the Target is underspecified andthe behavior of the Target is outside the controlof the designer.

(ii) If Km4Ks , the Target is overspecified, and thebehavior of the Target cannot be compatible withthe designer's wishes.

(iii) If Km� Ks , the design specification exhibitsRequisite Variety, provided the designer hascorrectly identified and specified the dimensions;and the behavior of the design should be thatwhich the Situation can absorb and which thedesigner can control, subject to the requirementthat the dimensionality of the Situation is notmodified by the introduction of the Target intothe Situation.

If the dimensionality is changed thereby, the designprocess can apply the Law of Requisite Varietyiteratively, taking into account the dynamics of theSituation.

Interpretation of law: The Theory of Dimension-ality has been introduced, in part, to makepossible this formulation of the Law of RequisiteVariety, especially to enhance applicability of itto those situations where some dimensions arenaturally quantitative and some are naturallyqualitative, requiring that both kinds of dimen-sions be in a common space and subject tocomprehensive interpretation in order to achievea sound design result.

The question might be raised as to howdesigners have succeeded in the past in the



36 John N. Warfield

absence of overt response to this Law. Many, ifnot most, Targets of design are redesigns thatbenefit from decades of experience which havepermitted the development of intuitive knowl-edge that substitutes for overt application of thisLaw. Regrettably, it is this same cumulativeexperience that mistakenly leads designers andtheir managers to believe that somehow they canintuitively design systems much larger in scalethat have never been designed before.

Brief 10A: The Law of Forced Substitution


References: None.

Statement of law: Structural underconceptualiza-tion and inherent conflict lead to policy vacuums inan organization into which authority injects forcedsubstitution for absent and inadequate conceptualiz-ation, in order to avoid institutional paralysis and forself-protection.

Interpretation of law: This Law reflects theempirical knowledge that executives in chargeof large organizations are essentially forced totake action in regard to problems of theorganization. The very large pressures on suchexecutives will be relieved in the short run bytaking action. A question of much importancehas to do with how effective such action will be.

It has been pointed out by Peter Senge thatwhile executives often have significant amountsof experience on short-term issues of relativelylittle complexity, such executives often have noreliable experience regarding longer-term issueswhich are complex. A simple explanation is thatby the time the consequences of the decisionsthey make are felt, those executives havechanged positions, and are not even around toexperience directly those consequences. Anotherexplanation has to do with the fact that complexissues are quite small in number compared to themany normal issues facing organizational leader-ship, so even the statistics work against gainingrelevant experience. It is unreasonable to expectthat the executive who is making decisions aboutcomplex issues is any better equipped to make

such decisions than anyone else inside or outsidethe organization.

The combination of being required to make adecision about a complex issue and the lack ofhigh-quality analysis and experience related tothat issue is perfectly calculated to produceaction that will not be effective and may makematters worse.

Because it is possible to apply methodologythat is compatible with and recognizes theimportance of the Laws of Complexity, it isreasonable to postulate that such an analysis ordesign could have been produced if the leader-ship were both aware of and willing to sponsorsuch activity. The choice of the term `ForcedSubstitution' recognizes that the decision-makeris substituting a `hip-pocket' or `wet-thumb'decision for what could have been a highlyinformed decision, informed, among otherthings, by the structure of the issue; and thatthe decision-maker is forced by circumstances tomake such a decision because to do otherwisewould convey an image of ignorance andindecision which (even though it might well bewarranted) is not what boards of directors arewilling to tolerate.

They will accept bad decisions (unwittingly orotherwise), unsupported by the kind of analysisand design that is now possible to attain takinginto account knowledge of the Laws of Complex-ity, but they will not support inaction.

Brief 10B: The Law of Precluded Resolution


References: None.

Statement of law: Forced substitution in organiz-ations is dominated by the combination of:

. structural underconceptualization;

. inherent conflict and diversity of belief;

. dysfunctional organizational linguistics;

which combine to preclude resolution of complexissues.

Interpretation of law: This Law is intendedto explain the reasons why complex issuesare seldom resolved in organizations. The




explanation is given in terms of what severalother Laws of Complexity have to say.

Recognizing that there is much diversity ofbelief, and much inherent conflict in the views ofindividuals concerning the relative importance ofvarious elements germane to a complex situ-ation, at the beginning the organization is inintellectual disarray about the complex issue.

If the organization does not have any effectivemethodology for learning about the issue, and ifit does not use the process of InterpretiveModeling to develop the structural patterns thatexplain the issue (and very few organizationspresently do this), whatever individual's particu-lar uninformed perceptions become the basis foraction will necessarily exhibit structural under-conceptualization, and will then promulgate anuninformed approach to an implementationscheme already lacking support in the organiz-ation.

In the absence of any well-designed means fordeveloping the necessary organizational linguis-tic domains, people will not even be able to sharea mutual understanding of what was wantedand therefore cannot be effective or evenmutually reinforcing in implementing bad de-cisions emanating from a perceived requirementto take some action.

In other words, there is an overwhelming setof institutional conditions that virtually guaran-tee that complex issues will not be resolved, andthe analysis that explains the reason for thepersistence cannot help but be supported by theanecdotal evidence being seen in everyday life asto the ineffectiveness and dysfunctionality ofsystems put in place with improper designs thatare unresponsive to the situations they werepurported to remedy.

Brief 11: The Law of Triadic Necessity andSufficiency

Origin(s) of law: Mathematical Logic.

References: Brent (1993), Burch (1991).

Statement of law: Relations are characterized by thenumber of distinct relational components, but nomatter how many such components a relation may

have, the (complex) relation can always be expressedby component relations having no more than threerelational components; but triadic relations exist thatcannot be expressed in terms solely of dyadic andmonadic relations.

Interpretation of law: Charles Sanders Peircestudied the logic of relations extensively. In therecent (1993) biography of Peirce by JosephBrent, the following passage appears:

Abstract forms of relation are objects of amathematical inquiry called the logic ofrelations (or relatives), which Peirce began toexamine in 1870 with his `Description of aNotation for the Logic of Relatives'. By 1885he had proposed in what Hans Herzberger[Professor of Philosophy, University of Tor-onto] has called `Peirce's remarkable theorem,'that there are only three fundamental kinds ofrelations: monadic, dyadic, and triadic; that bycombining triads, all relations of a greaternumber than three can be generated; and thatall those of a greater number than three can bereduced to triads. Since, in addition, triadscannot be reduced to dyads, nor dyads tomonads; monads, dyads, and triads constitutethe fundamental categories of relations. At thesame time, triads are made up of dyads andmonads, and dyads of monads.

According to Robert W. Burch [Professor ofPhilosophy at Texas A & M University], otherswho have examined related issues include Quine,LoÈwenheim, SchroÈder, Herzberger, and Ketner.The following passage appears in Burch's 1991book referenced above:

By extending both the algebraic ideas ofHerzberger and the graph-theoretical ideasof Ketner, this work proposes to develop analgebraic formalism in which a reductionthesis similar to and perhaps identical to thereduction thesis Peirce had in mind can beproved for the general case. This work alsoproposes to show that the reduction thesis itproves is consistent with the work of LoÈwen-heim and the result of Quine, despite thefact that these results may appear to conflictwith it.



38 John N. Warfield

The proof developed by Burch is long andthorny, but it has been examined by othermathematicians who have not detected any flaw.

If we accept the proof at face value, then weare impelled to note the interesting comparisonof this Law with the Law of Triadic Compat-ibility. Putting the two together we arrive at theresult that the number 3 not only is the maxi-mum number of elements whose interactions canreasonably be dealt with in short-term memorybecause of the limits of recall, but also it coin-cides with the maximum number of elementsthat must be dealt with modularly in order to beable to deal with complex relationships of anymagnitude.

The full significance of the foregoing is unclearat the present time, because of the limitedamount of investigation into the consequencesof accepting all of the foregoing as establishedscientific fact. But the potential significance is sogreat, and the absence of any evinced alternativeother than to continue the present disjointedincrementalism (`muddling through') advocatedby Braybrooke and Lindblom and so commonlypracticed in organizations; provide strong motiva-tion to those who are severely concerned with thedefects in present organizational practice to moveahead; on the basis of the hypothesis that what issaid in the foregoing paragraph is both trueand the appropriate guidance for changingorganizational practice and culture.

Brief 12: The Law of Small Displays


References: Martin and McClure (1985), oralmost any user guide to computer graphicssoftware.

Statement of law: Individuals, faced with aresponsibility to help illuminate complexity, willtypically fail to distinguish complexity from normal-ity in their choice of media for displaying their work,and will continue to accommodate their behavior tothe constraints imposed by small display media (e.g.,81

2� 11 inch or A4 paper size, and/or small computerscreens and standard-sized transparencies), instead of

insisting on matching the size of display space to thecomplexity of the subject matter.

Interpretation of law: If one imagines a scalerelated to complexity of issues, ranging from thevery simplest imaginable to the most difficultimaginable, one can then consider how the scaleof representation enlarges as the scale of the issueenlarges. Notably a point is reached where, asthe complexity grows, the scale of representationstops changing at the point where conventionalmedia have established benchmark dimensions.

Even the university has established a standardsize of chalk board, as though no matter what isbeing taught can be represented within thatscale. This assumption goes hand in hand withanother; which is that linear, sequential presenta-tions are the only kind that will ever be used inrepresenting subject matter. The fact that thelatter belief is readily contradictable seemsirrelevant to university administrations. Whychoose the university as a focal organization?Because it is this institution that blesses thepractice by carrying it out repeatedly, day afterday, to all of its clientele; thereby setting astandard for society to follow. It is, therefore, thesame institution that could change the practice,simply by acknowledging it and creating aninfrastructure to deny its universality.

REFERENCES FOR BRIEFS

Information in square brackets following the refer-ence identifies the Law(s) for which the reference isrelevant.

Alberts, H.C. (1995) Redesigning the United Statesdefense acquisition system. PhD Dissertation,Department of Systems Science, City University,London [3B].

Allison, G.T. (1971). Essence of Decision, Little, Brown,Boston, MA [8F]

Ashby, W.R. (1958). Requisite variety and its implica-tions for the control of complex systems. Cybernetica1(2), 1±17 [9].

Boulding, K. (1966). The Impact of the Social Sciences,Rutgers University Press, New Brunswick [8C].

Brent, J. (1993). Charles Sanders Peirce: A Life, IndianaUniversity Press, Bloomington, IN [11].

Burch, R.W. (1991). A Peircean Reduction Thesis, TexasTech University Press, Lubbock, TX [11].




CaÂrdenas, A.R. and Rivas, J.C. (1995). Teaching designand designing teaching. In Ertas, A., Ramamoorthy,C.V., Tanik, M.M., Esat, I.I., Veniali, F., and Taleb-Bendiab (eds), Integrated Design and Process Tech-nology (IDPT, Vol. 1), Proc. Society for Design andProcess Science, Austin, TX, pp. 111±116 [3B].

Churchland, P.S. (1986). Neurophilosophy: Toward aUnified Science of the Mind±Brain, MIT Press, Cam-bridge, MA [4].

Deely, J. (1991). Basics of Semiotics, Indiana UniversityPress, Bloomington, IN [4].

Goudge, T.A. (1969). The Thought of C.S. Peirce, Dover,New York [4].

Janis, I.L. (1982). Groupthink: Psychological Studies ofPolicy Decisions and Fiascos. Mifflin, Boston, MA [8F].

Kapelouzos, I.B. (1989). The impact of structuralmodeling on the creation of new perspectives inproblem-solving situations. Proceedings of the 1989European Congress on Systems Science, Lausanne,Switzerland, AFCET, October, pp. 915±932 [8E].

Martin, J., and McClure, C. (1985). Diagramming Tech-niques for Analysts and Programmers, Prentice-Hall,Englewood Cliffs, NJ [12].

Miller, G.A. (1956). The magical number seven, plus orminus two: some limitations on our capacity forprocessing information. Psychology Review 63(2),81±97 [1A, 1B].

Simon, H.A. (1974). How big is a chunk? Science 183,482±488 [1A, 1B].

Warfield, J.N. (1958). Switching circuits as topologicalmodels in the discrete probability theory. Trans-actions of the IRE PGEC 7(3) [8D].

Warfield, J.N. (1965). Synthesis of switching circuits toyield prescribed probability relations. 1965 Confer-ence Record on Switching Circuit Theory and LogicalDesign, Ann Arbor, MI, October, pp. 303±309 [8D].

Warfield, J.N. (1968). Switching networks as models ofdiscrete stochastic processes. In Tou, J. (ed.), AppliedAutomata Theory, Academic Press, New York,pp. 81±123 [8D].

Warfield, J.N. (1979). Some principles of knowledgeorganization. IEEE Transactions on Systems, Man, andCybernetics June, 317±325 [2].

Warfield, J.N. (1986). Dimensionality. Proceedings of the1986 International Conference on Systems, Man, andCybernetics, 2, IEEE, New York, pp. 1118±1121 [9].

Warfield, J.N. (1988). The magical number three Ðplus or minus zero. Cybernetics and Systems 19,339±358 [1A, 1B].

Warfield, J.N. (1991a). Underconceptualization. InGlanville, R., and de Zeeuw, G. (eds), Mutual Usesof Cybernetics and Science, Special issue of Systemica:Journal of the Dutch Systems Group, Thesis,Amsterdam, pp. 415±433 [2].

Warfield, J.N. (1991b). Complexity and cognitiveequilibrium: experimental results and their implica-tions. Human Systems Management 10(3), 195±202.Reprinted in Sandole, D.J., and van der Merwe, H.(eds), Conflict Resolution Theory and Practice: Integ-ration and Application, Manchester University Press,Manchester, 1993, Ch. 5, pp. 65±77 [5].

Warfield, J.N. (ed.) (1993). An Interactive ManagementWorkshop on the Variant Design Process: RRM JointApplication Development (JAD), Ford Motor CompanyResearch Laboratory, Dearborn, MI, 18 June (dis-tributed by Dr Scott M. Staley PE) [5, 8A].

Warfield, J.N. (1994). A Science of Generic Design:Managing Complexity Through Systems Design (2ndedn), Iowa State University press, Ames, IA (1st edn,Intersystems, Salinas, CA, 1990) [2, 4, 5, 7, 8A, 8C,8E, 8F, 9].

Warfield, J.N. (1995). SPREADTHINK: explainingineffective groups. Systems Research 12(1), 5±14 [3B,5, 8A, 8E].

Warfield, J.N., and Christakis, A.N. (1987). Dimen-sionality. Systems Research 4(2), 127±137 [9].

Warfield, J.N., and Tiegen, C. (1993). Groupthink,Clanthink, Spreadthink, and Linkthink: Decision-makingon complex issues in organizations, IASIS, Fairfax, VA[8F].



40 John N. Warfield

warfield, john n - twenty laws of complexity - science applicable in organizations

Documents

indexes of complexity

demands of complexity

ascience of complexity

laws ofcomplexity

taxonomy ofthe laws

integrative discussion

indexes ofcomplexity

integrative sciences