Jeffrey Goldstein, Ph.D.
School of Management and Business
Adelphi University
Garden City, NY 11530

Introduction Leadership and Emergent Organizational Structures

Reflect for a moment on the term "leadership." Your impressions will include some of the following notions a position of power and authority; directing people to get jobs done; having the final say about what, who, how, where, when; "being in charge;" the capacity for imposing your mandate; and so on. Many times, of course, these notions of leadership work just fine. Yet, how appropriate are they when it comes to the enormous magnitude of those innovations that will be necessary for a health care organization to survive amidst unprecedented changes in healthcare financing, governmental regulations, technological advances, and consumer desires. "Being in charge" and "imposing your mandate" just don't seem to hit the mark in the face of the unpredictable, sudden emergence of mergers, acquisitions, and variously fashioned joint ventures.

Yet, it is precisely the unpredictable and the emergent that are so central in the new complexity sciences. Indeed, emergence in self-organizing, complex systems is one of the most fascinating areas of current research into complex systems. Specifically, emergence refers to the unanticipated arising of new higher-level systemic patterns or structures functioning according to new laws and consisting of new properties. If we can consider our institutions and businesses complex, nonlinear systems (see Goldstein, "Map-Makers, Explorers, and Tricksters..." in this Resource Guide), then it should come as no surprise that our organizations are replete with emergent phenomena. However, our understanding of what organizations are supposed to be, e.g., bureaucratic hierarchical structures, has pretty much blinded us to seeing the full extent of the emergence taking place right in front of our eyes. Moreover, if we do recognize emergent phenomena for the spontaneous and "out of control" types of system occurrences they indeed represent, our training takes over automatically and we commence suppressing them as quickly as possible. Fortunately, it seems impossible to stifle all spontaneity and creativity, so emergent phenomena in our organizations and environments are here to stay. But rather than to dismay this fact, leaders can learn to take advantage of what could prove to be an extremely powerful and constructive organizational force.

Planned Versus Emergent Leadership

It's not that emergence has traditionally had no role at all in leadership. We find emergence, for example, in what has been termed informal as opposed to formal leadership. Whereas formal leadership refers to an officially-sanctioned, imposed role in a bureaucratic hierarchy, informal leadership occurs or emerges spontaneously outside of the sanctioned chain of command. Thus, in a project team, one or more persons may informally take-on leadership roles, others in the group then choosing or not to follow these informal leaders although to do so is not officially mandated. Whereas formal leadership is the result of planning, the emergence of informal leaders is a spontaneous event and thereby represents an unanticipated innovation in an organization. And, to the extent informal leadership is emergent and innovative, it parallels self- organizing processes in complex systems.

Yet emergent, informal leadership has been given short shrift not only in management literature and research, but in the real world of businesses and institutions. It is relegated to the ranks of either "grass roots" activities (e.g., the "charismatic" type of leadership frequently seen in social reform or religious movements) or to crisis situations. Pillai (1996), e.g., found that the ratings by co-workers of leaders spontaneously emerging during crisis situations were higher in leadership ability than leaders arising in noncrisis situations. But this just proves the point that emergent leadership is generally excluded from the mainstream aspects of leadership thought to be necessary for the ongoing running of a extant organization.

Emergence and Organizational Structure

Organizations, of course, consist of both leaders and those who are led, and there are two general types of structures for connecting the leaders and led hierarchical as in the above mentioned formal leadership bureaucracy; and participative as in distributed power and authority, e.g., self-managed work groups. We can establish an Organizational Structure Grid which links the above mentioned formal and informal leadership (i.e., the sources of leadership) with these two types of organizational structure:

The lower left quadrant, "Command and Control Bureaucracy," is the typical, traditional way that organizational members and leaders are connected. In spite of all the pleas to contrary on the part of management gurus and innovative leaders, it can still be said that most organizations function according to this mode of structure. The top left quadrant, "Informal Leadership," is, as stated above, a spontaneous, self-organizing process, yet, as can be seen in the grid, it retains a hierarchical structure in the sense that, e.g., during a crisis, the informal leader now "commands the troops." The past twenty years have witnesses the advent of the lower right quadrant, "Imposed Teams." First, there were Quality Circles, then Quality Teams in TQM, Project Teams in Reengineering, and so forth. What these latter organizing methods have in common is that they are imposed structures with distributed authority. It is important to note that lower right quadrant structures are not emergent but are, instead, imposed by the hierarchy. This "imposed" character needs to be underscored because it is sometimes thought they represent some kind of radical departure from the old "Command and Control" Hierarchy. Instead, they represent a somewhat contradictory message from top management that is similar to the one in the title, believe it or not, of a popular book on relaxation "You Must Relax!". Ralph Stayer, past CEO of Johnsonville Foods, experienced the futility of trying to impose participative decision-making teams after several years of mandating that his managers make decisions without his direction (see Goldstein, 1994). He finally realized he was imposing these teams just as he imposed all his other past management directions and that such imposition was not the way to proceed to generate a more effective organization. That is the why the research on imposed teams is so mixed the team structure by itself is simply not the key to success, it is rather how much real creative authority is actually allowed. In other words, neither motivation nor innovation can be created by fiat!

This brings us to the last quadrant, the upper right one, "Emergent Networks," which are neither imposed nor hierarchical. The source of "Emergent Networks" is self-organizing processes, and their power and authority is distributed. Certainly, "Emergent Networks" have been around probably as long as there have been organizations, but they are the least studied and generate the most apprehension on the part of the traditional "Command and Control" hierarchy. In a sense this apprehension is warranted because such "Emergent Networks" represent a threat to the traditional way leaders have thought businesses and institutions should be run - indeed they are "out of control" (see Kelley, Annotated Bibliography in this Resource Guide)..

With the arrival, however, of complex systems research and the concomitant interest in self-organization and emergence, "Emergent Networks" can now be better understood and can be seen to offer new possibilities for more adaptive organizational structures. Indeed, with the rapid rise of mergers and joint ventures, the real world has exceeded theory in regard to emergence. But, now the theory and research concerning emergence and innovation are there just waiting to be appropriated for our businesses and institutions. It is precisely spontaneously emerging Informal Leadership and Emergent Networks that contain the capacity for introducing those innovative structures and processes into system that are more adaptive to changing environments. So let's explore further what exactly is involved with emergence in complex systems, and, then, how leaders can, instead of suppressing them, learn to "ride these waves of emergence" toward more innovative and adaptive organizational structures.

Emergence in Complex Systems

In contemporary complex systems research, emergence refers to the arising of global, higher level patterns (i.e., structures, order, and qualities) out of local interactions of lower-level system components. Emergence is what takes place during the process of self-organization. An example from an actual physical system is the emergent structure of hexagonally shaped convection cells arising in the Benard liquid when it reaches a critical temperature, a process of self-organization studied extensively in the Prigogine School (see Nicolis, 1989). We can imagine looking down from above on the container holding the liquid when it is in this emergent state:

These emergent hexagonal cells are a startling and unexpected occurrence if they are compared with the homogeneity observed in the system before the critical temperature threshold has been reached. They express an across-the-system coherence, a collective order decidedly not present among the lower level components before their emergence.

An example of rather simple emergent patterns in an electronic simulation on a computer can be seen in the Game of Life, a cellular automaton originally developed by the phenomenal mathematician John Conway. As the Game of Life runs according to different rules linking the behavior of each cell in the array (lower level components and rules connecting components), emergence can be seen in the appearance of such higher level structures as "gliders," "space ships," and "flotillas" (Poundstone, 1985) (here the grid represents individual cells and the emergent patterns are the shapes made up by filled cells):

Emergent Patterns in the Game of Life

The gliders, space ships, shuttles, and flotillas represent fairly enduring, stable structures emerging at a "higher" level out of "lower" level components and their rule of interaction - the individual cells, i.e., the blocks in the grid, are either on or off (filled or empty) according to the status of neighboring cells. Similar emergent structures in electronic arrays can be seen in Kauffman's (1995) Boolean Networks or in other forms of Artificial Life (see Langton, 1986). For example, Kauffman discusses "frozen homogeneity clusters" and "walls of constancy" moving through his Boolean networks (for a discussion of Kauffman's way of understanding emergence, see Goldstein, 1993). At first sight these gliders and space ships may not appear to be particularly remarkable, complex, or even interesting. But what is amazing about them is that they have a dynamics of their own which is neither deducible from nor reducible to lower level components and rules. Thus, gliders seem to follow their own way, joining with other gliders, going around obstacles, turning direction, even stopping in one place. The antics of these emergent structures just doesn't seem understandable merely from the point of view of the lower level components and their rules of interaction. It seems that emergent phenomena have to be understood from some other vantage point than local interactions.

Emergent Behavior Follows Innovative Emergent Laws

Emergent structures have emergent laws that determine their behavior, laws which can not be deduced from lower level components alone. Indeed, one of the most intriguing things about emergent phenomena in complex systems is that they don't seem to follow from what has already been going on in the system. In the above pictured Benard System, the direction of rotation of the hexagonal convection cells are not predictable from the homogeneous nature of the liquid before it reaches the critical temperature (Nicolis, 1989). Indeed it is this quality of having a dynamic all their own that has led emergent phenomena to be called Artificial Life (Langton, 1986). They seem to have a "life" of their own, so to speak, which was not expected from the mechanical set- up of the electronic lattice alone nor the rules connecting the cells that were programmed into the cellular automata.

The arising of emergent patterns is similar to the way a 3-D image suddenly appears when staring at stereoscopic artwork (made popular, e.g., in Magic Eye Artwork - see Thing, 1993). The abrupt appearance of the 3-D image is almost like magic ( those who are unable to see the 3- D pattern think you are crazy when you do!) and seems to betray the laws of causality (Goldstein, 1996). Yet, of course, they are not magic and they are indeed embedded within a causal nexus. On the one hand, the 3-D images, like emergent phenomena, do not seem consistent with what came before, on the other hand, they must be congruent with the pre-existing conditions in the system - e.g., even in the Magic Eye Artwork, one can still discern the previous patterns along with the newly appearing 3-D form. For example, in the self-organizing Benard system, the hexagonal convection cells are certainly unexpected, yet they nevertheless remain constituted by currents of the liquid, and the patterns seen in Artificial Life are, after all, still made up of the lower level, local off and on individual cells, but now combining and recombining into global level patterns. We shall be returning to this curious mixture of lower level, "past," components and new unexpected patterns later on. For now, what needs greater exploration is how emergent structures represent radical innovations in a system, innovations which promise far greater adaptability of the system to a changing environment.

Emergent Structures as Radical Innovations

In another paper (Goldstein, 1996), I have argued that, rather than the unpredictability associated with sensitive dependence on initial conditions found in chaotic systems, it is emergence per se which presents the real challenges concerning unpredictability in complex systems. For whereas chaos can be captured in constrained regions of phase space by means of phase portraits of chaotic attractors (see Goldstein, "Map-makers, Explorers, and Tricksters..." in this volume), emergence represents a vastly greater set of possibilities of system behavior (albeit one that can still be understood in part using the construct of attractors). Thus, it seems to me that while chaos does bring with it certain conundrums to the idea of causality (e.g., see Stephen Kellert, 1993), emergence presents a far greater test to causal inference because the emergent patterns have such a drastically different internal dynamic than can be ascertained from the components themselves. That is, emergence threatens our causal intuitions since it is so radically innovative.

Indeed, the characteristic of radical novelty, i.e., their surprising, unexpected, and irreducible quality, makes emergence one of the most fascinating, far-reaching, and perplexing findings in complex systems research. It is this radical novelty which, in part, renders emergents so "out of control" since what we previously knew about the system no longer seems to apply to the emergents and their unique dynamics. One way to understand this kind of novelty is to think of it as neither new copies of what was previously possible in the system nor even as new recombinations of what was previously going on. Instead, emergent novelty represents a radical departure from past practices, structures, and processes, and yet uses these past practices, structures, and processes in drastically different and unanticipated ways. (See Appendix A for a more technical treatment of a mathematical analogy for depicting the kind of radical novelty found in emergence). In an organization, for example, we saw above that Imposed Teams (from the lower right quadrant of the Organizational Structure Grid) do represent a new type of organizational structure in that their internal participative framework represents a departure from a strictly hierarchical set-up, yet, they are still the result of an imposition, a characteristic associated with the old Command and Control mindset. Therefore, imposed teams are not innovative in the sense that emergent structures are, and, accordingly, do not offer the kind of adaptability that emergent structures do They don't have this adaptive capacity because they do not represent modifications that avoid being trapped on low adaptive peaks in fitness landscapes, to use Kauffman's terminology (see, Goldstein, "Map-makers, Explorers, and Tricksters..." in this Resource Guide.)

The Black Box of Emergence

The idea of emergence is not all that new and it pays to take a brief excursion to how it has been used in the past in order to better appreciate what new insights are being brought-forth from current research. The term "emergence" seems to have been first coined in its more technical sense in the second half of the the Nineteenth Century by the philosopher G. H. Lewes (1875) to describe the arising of unexpected new qualities in a process, e.g., the gaseous nature of oxygen and hydrogen leads one to expect that their combination in a chemical reaction would eventuate in another gas, so the liquid quality of water (H2O comes as a surprise - Lewes called this new liquid property an emergent property. This concept of emergence became one of the foundations of Emergent Evolutionism, a scientific and philosophical movement in the 1920's and 1930's which believed the idea of emergence could explain not only the discontinuities found between species that the theory of evolution could supposedly not account for as well as for the radical distinction between such dichotomies as inorganic/organic, insentient/sentient, and so forth [e.g., see Morgan, 1923; Broad, 1925; and Alexander, 1966). Emergent Evolutionism investigated emergent phenomena in terms of novelty, unpredictability, nondeducibility from lower level components in the system, as well its seemingly noncausal connection with pre-existing and lower level conditions.

After the demise of Emergent Evolutionism in the nineteen thirties, the idea of emergence was perpetuated particularly among philosophers of biology (see, e.g., Polanyi, 1966) who employed the concept to buttress biological sciences against the attempt, on the part of extreme reductionists, to show that biological structures and processes could ultimately be reduced to the laws of physics alone. Similarly, the concept of emergence was utilized by the split-brain researcher Roger Sperry to speculate how mental states could be understood as having a "causal" or "determinative" influence on the physical substrate of the brain (see Sperry quoted in Stephan, 1992). For Sperry, instead of thinking that all mental states could be ultimately understood as the result of "micro-determination" from brain states, mental states as emergent phenomena could also have a role in "macro-determining" other mental states as well as brain states themselves.

But although the idea of emergence was used in these previous contexts, what was lacking in previous approaches was an ability to look deeply into which systemic processes could bring about the radical novelty and unpredictability of emergent phenomena. In effect, in earlier theorizing the processes leading to emergence were hidden inside a "black box" which took inputs (lower level system components) and produced outputs (emergent phenomena) but what exactly went on inside the "black box" was opaque:

The early Emergent Evolutionists as well as the later anti-reductionistic biologists simply did not have access to what went on inside the black box. They could only surmise what kind of things must take place inside by observing in what ways the the input could be so radically altered to partake of the properties characterizing the output. Of course, Roger Sperry and others did start peering into the black box of the emergence of mental states through their split-brain and other research, but that research was very specialized in the neurosciences and it, thereby, hardly suggests the kind of generalizations that would be needed for organizational practitioners to gain insight into the black box of organizational emergence.

Opening the Black Box of Emergence

Of course, if the process of emergence remained a black box, emergent phenomena would not present the opportunity they are presenting leaders. Fortunately, contemporary research into emergence in self-organizing, complex systems is opening-up the black box and leaders can learn to utilize what is being found inside. The box is being opened by intensive study of emergence in cellular automata and boolean networks by way of computer simulations, as well as by investigations into the mathematical dynamics of nonlinear systems helped by graphic displays of attractors by observing on a computer screen what changes in rules on lower levels leads to emergent patterns on higher levels. Like modern alchemists, Kauffman and other researchers in the area of complex, nonlinear systems can study the conditions and processes of transformation whereby emergence takes place. Indeed, a great deal is being revealed inside the black box of emergence:

Indeed, we see inside the box many new constructs that are aiding researchers in understanding how the interaction of lower level components can yield the fantastic array of emergent phenomena. For example there are the various kinds of attractors that underlie the nonlinear dynamics of emergence; there is the far-from-equilibrium (FFE) conditions leading to emergence; there are the constants (e.g., Feigenbaum's) suggesting the universal nature of some aspects of the behavior of emergent processes, there is the sensitive dependence on initial conditions (S.I.C); there are the ways emergent patterns change as the result of changes in the adaptive rules governing interaction among lower-level components; there are the fitness landscapes in which Kauffman (1995) and others are finding general principles of the adaptive value of emergent structures; there are the power law distributions found in self-organized critical states (see Bak, 1996 in Bibliography to Glossary in the Resource Guide); and finally there are two constructs that may seem new to those who already know something about complex systems, anacoluthian processes and containment, which I would like to focus on because I think they can be of help to leaders working with emergent structures in organizations. I will be calling those processes in general that facilitate emergence in complex systems "anacoluthian" processes and the way these processes need to be contained and channeled "boundaries."

Anacoluthian Processes Consistent Inconsistency

"Anacoluthian" comes from the term "anacoluthon" (Greek for inconsistency in logic") that is used in grammar to refer to a sentence that starts out in one grammatical form and then ends in another, i.e., is inconsistent. An example of an anacoluthon is the sentence, "The sun looks so strong today are you going out swimming later?" The first phrase, "The sun looks so strong "today" is a declarative assertion, whereas the second phrase "are you going out swimming later?" is a question. Hence, the sentence is anacoluthian by both following one course of logic (of grammatical construction) but then switching to another one. "Anacoluthian," then, is a term I am using to refer to any process which is consistently inconsistent (inconsistently consistent).

Now what does this have to do with emergence? First in the less important sense, as stated above, the novelty of emergent phenomena is both consistent and inconsistent in relation to the previous patterns in the system. Thus, the Benard hexagonal convection cells are consistent with the previous and lower level components of the system since they are indeed currents of liquids, but they are inconsistent in by-passing in a sudden manner from the linear spread of heat found in gradual conduction of heat to the nonlinear dispersion of heat found in the convection currents of the hexagonal cells. Their consistent inconsistency results from the shift to a new attractor that occurs at the critical temperature (technically, a "bifurcation"). Indeed, we can say that the new attractor(s) is consistently inconsistent with the previous attractor(s) governing the system. Similarly, the flotillas that show-up in the Game of Life, are consistent with lower level filling of cells but inconsistent in the way they cut across local interactions in producing a global structure that persists. Similarly, Kauffman attributes the stability of the emergent patterns in his Boolean networks to how redundant ( i.e., same or identical) patterns percolate through the system. The anacoluthian nature of this percolation is that the redundant patterns start operating according to dynamics that are not deducible from the lower level rules connecting the nodes (see Goldstein, 1994). Thus, emergent phenomena are anacoluthian in being both consistent and inconsistent with the previous patterns of the system.

However, emergence is anacoluthian in a more important sense the way in which emergence is precipitated by operations on or within a system which both continue and violate previous operating principles. That is, emergence seems to require a following and at the same time a transgression of the pre-existing patterns or dynamics in a system. For example, the emergence of innovative modifications during biological adaptation (of sexually reproductive organisms) is the result of the anacoluthian processes of recombination of genetic material and random mutations. These processes responsible for emergence are anacoluthian in that they both take the existing genetic material and, at the same time, change it during reproduction. This can be illustrated in the way John Holland has taken the principles of genetic recombination as the model for the manner in which his genetic algorithms are able to search for new solutions in possibility space. Here is how Holland pictures the anacoluthian process of genetic recombination, i.e., crossover of genetic material:

The unshaded box on the left (1110###) represents the genetic material from one parent while the shaded box beneath (000###1) represents the genetic material from the other parent. The crossover operation of combining the genetic material from both parents from the offspring follows the pattern from both parents up to a certain point but then "crosses over" and replaces one set of genetic material with another. As a result of this anacoluthian mixing, the offspring has an innovative set of genetic "programs" resulting in the emergent features of the offspring. From an adaptation point of view, this anacoluthian recombination offers the possibility of innovative modifications in an organism that may prove more fit to a changed environment.(See Appendix A for a mathematical type of anacoluthian process.)

Notice that for the outcome (i.e., the innovative, emergent modification) to have the possibility of being innovative there had to be something "consistently inconsistent" about the process leading to the outcome. In other words, innovative outcomes demand processes capable of generating innovation, and this is precisely what "anacoluthian" is meant to convey. Anacoluthian processes allow the introduction of something new by following a pattern and at the same time transgressing the pattern. If this transgression did not take place then something new could not enter the picture, there would be just the "same old same old." By trangression I am referring to some kind of "crossing-over" a preset rule, a kind of "mixing-up" of what was previously inviolably separate. But, again notice that it is not just transgression, but there is also a continuity of the previous patterns e.g., the child shows both continuity with its parents and emergent novelty with respect to them, nor can you say a child is just a combination or synthesis of the parent's genetic make-up. Rather, the mixing of the genetic material from parents leads to a nonlinear interaction resulting in emergent qualities in the child.

The anacoluthian nature of sexual reproduction leading to the possibility of emergent innovations in offspring is also aided by the presence of random mutations of genetic materials. Randomness is another source of novelty since two of the features of a truly innovative modification are both its unpredictable and unplanned characteristics. "Unpredictable" and "unplanned," though, are equivalent to "random." Anacoluthian processes, therefore, include both the consistent inconsistency we have been discussing plus the impact of random effects on a system. To summarize, anacoluthian processes facilitate emergence by continuing, transgressing, and allowing for the influence of random events. Thereby, anacoluthian operations lead to the possibility for innovative and adaptive emergent patterns.

Another way of saying this is to realize that a creative, innovative, and unpredictable outcome requires processes that must be characterized as including elements of creativity, innovation and unpredictability. That is, a leader cannot come up with innovative organizational structures by processes that themselves neither partake of creative departures from the norm, nor merely continue past structures, nor must be planned and anticipated at each stage. If a leader wants to facilitate emergent adaptations, they must be willing to facilitate organizational processes that are anacoluthian and include elements of the random (see the use of organization "noise" in Goldstein, "Map-makers, Explorers, Tricksters...." in the Resource Guide). As the saying goes, only the insane think they can produce a new outcome through doing the same old thing over and over again. Indeed when one takes a close look at what seems to work in supporting and engendering successful organizational change anacoluthian-like processes seem to be at work (Goldstein, 1994) mixing levels in the hierarchy; challenging currently held assumptions; looking at both content and process; amplifying differences, contrasts and incompatibilities; establishing new connections; taking the inside outside and bringing the outside inside (example Harley Davidson); following the plan and simultaneously being open to serendipitous random events.

Within Contained Fields Boundaries

Another crucial factor being discovered inside the black box of emergence is that of contained fields or boundaries (see Goldstein, 1994). Self-organizing processes take place within contained fields that keep the system intact and channel powerful nonlinear forces. In a sense, these fields act as boundaries providing a sense of closure to the emergent structures permitting these global patterns to go across the system. Moreover, the closure of boundaries is another way of talking about the coherence of the emergent structures. For example, in the Benard system in which the hexagonal convection cells are the emergent structures, the self-organizing processes occur within the boundaries of the container holding the liquid; see the circular container holding the liquid in the sketch of the Benard container above).

But, it's not just self-organization and emergence in physical systems that demand boundaries, they are also a crucial, but neglected factor in emergent processes in electronic arrays. For example, in cellular automata or in Kauffman's boolean networks, the components of the system, i.e., the "on and off" cells or nodes, are electronically "hardwired" so that rules of interaction between can connect them. By "hard-wired" here I am not referring to actual "wire" connections but to how the computer simulation is set up to connects the cells in terms of the rules of interaction of neighboring cells.

In this figure we see that the electronic connection between the cells in the Game of Life or in other cellular automata including Boolean Networks are bounded by the configurations possible for the cells interaction..

Emergent Leadership in String Quartets

A particularly clear example of emergent leadership can be seen in professional string quartets which operate as an intensively reciprocal, interdependent, nonlinear, complex system, and, therefore, are an organizational site where emergent phenomena might be expected to appear (see Goldstein, In Press). Since professional string quartets play pretty much the same limited repertoire of music, their success hinges on, besides high quality performance, the innovative nature of their interpretation of the music they are playing. Yet this innovative capability does not come out of the blue but appears to be, in part, a function of the innovative organizational structures of successful quartets since, although a bare skeleton of a leadership hierarchy undergirds them, successful quartets exhibit an innovative pattern of leadership and group dynamics that is neither imposed nor planned but emerges during the ongoing work of the group in practice and recitals (see Murnigan and Conlon, 1991).

A modicum of imposed structure does exist in the string quartet a first violinist in a quasi- leadership role; a second violinist backing up the first (the so-called "second fiddle") and who is thought to have less to do so takes on more business responsibilities; a viola player; and a cellist. Yet, in spite of the leadership role of the first violinist, members join the quartet because they expect to have a 25% stake in all decisions, business and artistic. Thus, although their hierarchical framework would tend to characterize them in the lower left quadrant in the Organizational Structure Grid above, i.e., the traditional bureaucratic structure, the democratic values of the members leans toward the right column, i.e., distributed power. The resulting clash between hierarchy and democracy prompts the unsuccessful quartets into either a rigidification of the hierarchy by placing all power into the hands of the leader or an overly participative, loose democracy where it is very difficult to reach decisions. But the successful quartets manage to avoid either extreme by generating an emergent structure which, while combining elements of both the hierarchical and the distributed, at the same time, transcends both through the emergence of a new emergent network (upper right quadrant). Indeed, it seems that it is this very capacity to function according to a novel emergent leadership/follower pattern which is a key to their success.

One feature of this emergent leadership/group structure was an acknowledgment that both sides, the hierarchical and the democratic, are equally important. Yet, rather than some kind of bland synthesis of both horns of the dilemma, the innovative structure emerged dynamically in that instead of trying to hash out the conflict in words, the group found that through practicing and playing that the leader's and the group's interactive roles would emerge in appropriate forms. The effectiveness of the innovative emergent network can also be seen in the way the successful quartets dealt with conflicts concerning how to interpret musical phrases conflicts about how to play a piece were often resolved by a consensual agreeing to play it one way in one concert and another way in another concert. However, this seldom needed to take place because the alternative interpretations were somehow incorporated into the first play (as one members stated, "When you play, what is right and what is wrong emerges"). Whereas, in the unsuccessful quartets there was much more planning and talking about how to resolve the conflicts in the groups and, thereby, much less time available for actually allowing the conflicts to work themselves out as the group went about its business.

Certainly, a key question is what kind of "rules" of leadership and group interaction did the successful string quartets follow that were more likely to lead to an emergent resolution of discord? Traditional organizational theory would have it that conflict is resolved by the supposed capacity of group members' commitment to superordinate goals but this theory just isn't of much help because it commitment does not specify particular ways that conflict is overcome except to suggest, as in the emergent account being discussed here, that such patterns of conflict resolution emerge over time as the group continues to work together. But that is just to say that emergent patterns are the key. Using the framework developed in this paper, however, we can postulate that it was the anacoluthian mixing of hierarchical with democratic building blocks that was surely one key to string quartet effectiveness. (The specific nature of this anacoluthian mix is probably a unique matter of each individual working group, its kinds of tasks, environments, etc.). This emergent resolution can be seen as a kind of adaptation in a complex system in which differing, even opposing views, commingle within a shared framing of issues large enough to encompass those differences.

An anacoluthian perspective also offers insight into how successful working groups and leaders can fight the very strong pressures toward group conformity which scores of experiments in social psychology have confirmed over and over again. Anacoluthian processes might include the strengthening of "minority" positions in an organization - e.g., creative thinking and originality increases when a minority influence is allowed in a work group (Van Dyne and Saavedra, 1996). Contrary to expectations, conflict in these groups did not increase although there was an increase in personal stress on the part of the minorities. Furthermore, Jackson, et. al. (1995) found that group heterogeneity in terms of a mix of personalities, gender, attitudes, and background of experience was positively related to the creativity and the decision-making effectiveness of teams. And, Bantel and Jackson (1989) found that innovation in the banking industry was positively associated with a diversity in the experience of top management teams. Anacoluthian processes interrupt tendencies toward group conformity that is, changing the rules of interaction and then observing what emergent patterns emerge. Furthermore, leadership in emergent systems will need to be careful to distinguish "enslaved" conformist behavior and the coherence found in emergent behavior - indeed they may superficially appear the same. This is an area for much greater research and practice.

To get a better grasp of what is involved in emergent novelty and how it can be brought- forth, let's take a look at a prototypical type of novel pattern. Understanding novelty in this way will illustrate what leaders are up against if they want to be facilitators of organizational innovation.

Consider the following two sequences

a. 1 1 1 1 1 1 1 1 ...

b. 1 2 3 4 5 6 7 8 ...

Sequence a is simply a sequence of the number 1 repeated ad infinitum (e.g., from a computer which is programmed to simply print 1's). Each time the # 1 appears it is certainly a new # 1, but this type of newness or novelty is, of course, very uninteresting- in fact, it hardly seems appropriate to call each # 1 a new 1 since it is simply a repetition of the previous. Sequence b, however, is slightly more interesting in the kind of novelty being generated since each successive number hasn't appeared before on the list, but instead, consists of the addition of 1 to each previous number. Now, this second kind of novelty may be characterized by a greater degree of innovation, but, nevertheless the novelty is rather dull since it can easily be predicted what will come next, or further down the line.

Other forms of more radical novelty, however, are possible. For example, consider the following table which lists a set of rational numbers (numbers expressible as the ratio of two integers such as 1/2, 1/3, 1/4, 1/5, 2/3) and their decimal equivalents

1. 1/2 .500000000.....
2. 1/3 .333333333.....
3. 1/4 .250000000.....
4. 1/5 .200000000.....
5. 2/3 .666666666.....

The great nineteenth century mathematician Georg Cantor took a list like this and showed how one could generate a number so radically new that it could not be included in the list of rational numbers at all (Dauben, 1980; Robertson, 1995). Cantor constructed this radically new number by performing two operations drawing a diagonal across the list of decimal expressions and changing the resulting diagonal sequence as it goes along

1. 1/2 .500000000.....
2. 1/3 .333333333.....
3. 1/4 .250000000.....
4. 1/5 .200000000.....
5. 2/3 .666666666....