------ # Understanding Command and Control **Written By:** - David S. Alberts **See:** Alberts_UC2.pdf --------- ## Introduction Perhaps the most pernicious issues in understanding C2 are (1) the lack of a useful analytical definition and (2) the implicit assumption that C2 is how traditional military organisations perform the functions of Command and Control. The definitions used by the U.S. Department of Defence and NATO may be valuable in the legal, institutional, and operational settings for which they were developed, but they are not useful for analysis or research. We and others have previously pointed out the problems with relying on these “institutional” definitions when engaged in research or looking at the future of Command and Control functions.43 This is because these institutional definitions are most often the products of a committee or a coordination process and as such are politically rather than scientifically correct. They also almost always reflect current thinking and processes. These problems include the obvious fact that Command and Control are different—but closely linked—functions, and that the existing definitions focus on the legal distribution of authority, as well as the failure of these definitions to highlight the unique contributions that we associate with leaders and commanders. Working with a group of senior professionals from NATO and other nations, we have recently concluded that there is a better approach to defining Command and Control. The argument is that there are three key factors that define the essence of C2 and two important ways that those factors vary within the structures and processes of a given enterprise (service, nation, coalition, or force). These three key factors, which can be thought of as the dimensions of a C2 Approach, are the: - allocation of decision rights; - patterns of interaction among the actors; and - distribution of information. ![[UC2-Fig-11.png]] We are interested in the actual place or region in this space where an organisation operates, not where they think they are or where they formally place themselves. We are all familiar with the informal organisations that exist, often operating in very different ways than the official doctrine would dictate. For example, the formal allocation of decision rights may not correspond to the effective (actual or in fact) distribution because of a variety of factors, including traditions, culture, or level of training within a force. Similarly, the formal patterns of interaction or distribution of information desired by the leadership may differ substantially from the way information flows and is distributed in the real world because of informal relationships, linkages, sources, and the real-time requirements of a situation. We must also allow for the fact that enterprises are not homogenous in any of these three dimensions. In fact, an organization’s location in the C2 Approach space usually ranges across both function and time. For example, with respect to function, intelligence operations may operate in a different part of the C2 Approach space from logistics, and the conduct of a humanitarian assistance operation may operate with a C2 Approach that is quite different from the C2 Approach used for combat operations. Similarly, with respect to time, the crisis management phase of an operation may operate in a way that would not be appropriate if a war broke out. Hence, the C2 Approach of a given service, nation, coalition, or force may well be best understood as a region or collection of regions within the three-dimensional space rather than, as it is usually thought of, as a point within that space. ## Problem Space Clearly, some types of C2 Approaches will be better suited for certain types of problems. Mapping C2 Approaches to the types of problems for which they are relatively well-suited requires an understanding of the important ways that problems differ. We posit a problem space defined by the following three dimensions (which may not be orthogonal or totally independent): - rate of change (static versus dynamic) - degree of familiarity (known versus unknown) - strength of information position (informed versus uninformed) ![[UC2-Fig-12.png]] ### Rate of Change Static problems are those for which the situation itself does not change rapidly. For example, the trench warfare that evolved in Europe during World War I proved to be static — the front lines, the political, social, and economic operating environments, and the methods of warfare employed changed only slowly. Dynamic problems involve rapid change across all of these features—the location of critical times and places during the struggle change quickly, the operating environment is unstable, and the parties to the conflict innovate frequently and rapidly. Clearly, more static problems are amenable to more centralised decision making in which efforts can be optimised, allow preplanning of the flow of information, and can be controlled by regularised patterns of interaction that involve known groups of specialists. However, in dynamic situations, these classic Industrial Age practices will become impediments to successful Command and Control. ### Degree of Familiarity When the nature of the problem is well-known (and familiar), everything is relatively simple. First, the information requirements and who needs what information are reasonably well understood and can be supported efficiently. Second, the patterns of interactions needed are clear. This enables clear decisions about the appropriate allocation of decision rights within the enterprise. In other words, Industrial Age C2 Approaches can be used efficiently and effectively. Note that familiarity with a situation is not necessarily a correlate of the degree of situational dynamism present. For example, both NATO and Warsaw Pact military contingency planning for World War III in Europe foresaw a very dynamic battlespace. However, after decades of planning, preparation, and intelligence collection on both sides, each side believed they had a good understanding of how such a conflict would unfold. As a consequence, they developed highly specialised forces and detailed plans for the conduct of this “familiar” conflict. Of course, confidence that a situation is well-understood does not always translate into effective force development and planning. The French believed they understood the dynamics of an attack from Germany in the later 1930s and committed themselves to a strong linear defence, epitomised by the Maginot Line. However, they found themselves in a very unfamiliar conflict when the Germans unleashed their Blitzkrieg. A more knowledgeable organisation, one in which the situation is familiar to a large number of individuals, can distribute decision rights further than one in which less knowledge is present or knowledge is concentrated. ### Strength of Information Position Finally, situations can also differ in terms of the extent to which the decision making is informed or uninformed. Regardless of the degree of dynamism (though very possibly influenced by that factor) and the degree of knowledge available about it (though, again, very possibly influenced by it), the strength of the information position has an important impact on the applicability of a particular C2 Approach. As is explained in Understanding Information Age Warfare, 47 the information position of an organisation is the degree to which it is able to fulfil its information requirements. Hence, a force with very simple information requirements (for example, a terrorist organisation) may have a strong information position although it possesses relatively little information. At the same time, a coalition seeking to employ sophisticated weapons and conduct successful counter-insurgency operations may have a great deal of information, but still have a relatively weak information position because of the massive amount of information it requires. At the same time, the extent to which the information available is of high quality (correct, current, accurate, precise enough to support its use, etc.) also influences the information position of a force or an organisation. Clearly, a well informed force can distribute decision rights differently than a weakly informed one. Moreover, a well-informed force should distribute its information more broadly and encourage timely collaboration about what that information means and how to act on it successfully. All this having been said, identifying the crucial elements of the problem space and matching regions in this space to regions in the C2 Approach is a high priority. Research should include historical analyses, review of lessons learned in recent and ongoing conflicts, and experimentation. We know, for example, that the level of training and education required for some C2 Approaches are greater than those required for others. However, we do not know whether current constructs of knowledge and information contained in the NATO C2 Conceptual Reference Model are adequate to capture the essence of these issues. We also know, for example, that the ability to shift between C2 Approaches is a method for dealing with dynamic operating environments, but we do not know (1) how great a range of C2 Approaches a force or organisation can muster or (2) whether there are alternatives such as hybrid C2 Approaches that might make it possible to operate across some relatively broad range of situations. The discussion that follows starts by examining specific regions within the C2 space in order to keep the material easier to understand. However, later discussions evolve to deal with a range on each dimension and the regions they define. In addition, we begin with a general appreciation of C2, and then move to independent analysis of the command approach and the control approach. While the distinction between these two is critical for analytical purposes, they really must be understood as parts of a unified whole. While military operations can be discussed in terms of their functions of intelligence, operations, logistics, and other functional areas, they ultimately cannot be understood unless the interactions and interdependencies among them are properly appreciated. Similarly, the concept of a C2 Approach can only be understood in terms of its elements if these discussions are ultimately brought together. One of the valuable results of such an analysis is the ability to assess the compatibility (or lack thereof) between a particular approach to command and a particular approach to control. In practice, the three key dimensions across which C2 Approaches differ are not really independent, so showing them as three axes of a cube is something of a distortion. The relationships between them have been illustrated in Figure 13 as a “waterfall chart” because the most fundamental dimension is allocation of decision rights, which impacts the other two and, together with patterns of interaction, goes a long way toward determining the distribution of information. Moreover, the resulting distribution of decisions (the “real” distribution that emerges within the dynamics of a situation) may result in a change in the basic allocation of decision rights. This is an example of the type of adaptation (change in work processes and/or organisations) that occurs as part of the agility needed for effective military operations. For example, when a military situation becomes urgent (e.g., an ambush at the tactical level, the realisation that an adversary has executed an effective deception plan at the operational level and therefore friendly forces are incorrectly positioned), commanders at lower levels will not (under the doctrine of most modern forces) consult with higher headquarters about deviating from the plan or wait for a new plan, but rather take the initiative by making decisions about how their forces will immediately react. They then inform higher headquarters of what has occurred and the actions they are taking and request support so that they can deal with the ongoing challenge. (Of course, in an ideal world, the other parts of the force, including their higher headquarters, would be able to monitor the situation and would know that they had begun to take initiatives.) If these actions take them outside the existing plans or guidance, they will have altered the distribution of decision rights. ![[UC2-Fig-13.png]] As is discussed in more detail below, the allocation of decision rights either establishes or enables the establishment of the mechanisms by which those within the enterprise share information and collaborate. These structures, in turn, are major factors in determining the patterns of interaction within the social domain and the information domain. Those patterns, played out over time, have an important impact on the “real” distribution of information within the enterprise. ### Allocation of Decision Rights Decisions are choices among alternatives. Decision rights belong to the individuals or organisations accepted (whether by law, regulation, practice, role, merit, or force of personality) as authoritative sources on the choices related to a particular topic under some specific set of circumstances or conditions. The allocation of decision rights is their distribution within the international community, a society, an enterprise, or an organisation. In this context, the organisation of interest is a military, a coalition, an interagency effort, or an international effort including military elements. There can be different distributions of those rights across functions, echelons, time, or circumstances. In theory, the allocation of decision rights is a linear dimension with two logical endpoints. At one end of the spectrum is total centralisation (all the rights held by a single actor). The totalitarian regimes of Stalin and Hitler were understood to take this form, though it is unlikely that they had the ability to personally control all aspects of their societies on a continuing basis. At the other end of the logical spectrum is total decentralisation (every entity having equal rights in every decision, or a uniform distribution). While both total concentration or centralisation and perfect equality or complete decentralisation are sometimes found in very small groups, neither of these extremes is likely to be found in any enterprise that would be important to our analyses. Even the ideal “real world” political democracy uses representatives elected by individuals (who possess equal voting rights) in order to make authoritative decisions.49 Hence, for the purposes of understanding the allocation of decision rights in military organisations or civil military enterprises, the interesting range of values lies well inside the two extreme points of the spectrum. ### Allocation of Decision Rights in the Industrial Age The allocation of decision rights in Industrial Age militaries reflects the basic management and knowledge structuring principles of that era. First, the problem of military decision making is decomposed into specialised functional roles using the “Napoleonic” system of command, intelligence, operations, logistics, plans, and so forth. Second, a system of echelons is employed to ensure both appropriate span of control (that is, middle managers that are available to translate guidance from above into actionable directives, to monitor and report activities in the operating environment), as well as what happens to and within the elements of the force, and to act as control agents to ensure that guidance is understood and followed. Hence, decision rights in Industrial Age militaries are centralised in those responsible for the command function, but also allocated to functional specialists (logisticians, for example) who work in echelons provided that the choices they make are consistent with the overall guidance provided by commanders. One reflection of the centralisation of these militaries is the practice of issuing orders in the name of the commanding officer, even when they may deal with some specialised functional area. Staff members and subordinate commanders have a serious responsibility for ensuring that they understand the guidance from above, which is one reason that a great deal of stress is placed on the Industrial Age concept of *commander’s intent* (as opposed to the Information Age variant *command intent*). We should note that not all Industrial Age militaries have (or have had) highly centralised decision rights. In earlier work, we demonstrated at least six successful 20th century military C2 Approaches located at some distance from the endpoints of the centralised-decentralised spectrum. Some of these have or currently employ mission type orders that devolve (or explicitly delegate) meaningful choices to lower echelons. However, none of the doctrines examined in those analyses go far enough to qualify as either self-synchronising or Edge approaches. ### Allocation of decision rights in the Information Age Information Age organisations, including militaries, are expected to have minimally centralised distributions of decision rights. An extreme case can be imagined in which there are only emergent distributions and no formal or rule-based distributions of these rights. The tenets of Network Centric Warfare indicate that self-synchronisation will be enabled when there is a “critical mass” of shared awareness (in the presence of suitable doctrine). The success of NCW and Edge C2 Approaches depend on a broader distribution of information and different patterns of interaction (the other key dimensions of a C2 Approach), but they also assume that individual entities have the capacity, information, and means to make effective decisions. Edge organisations not only assume a widespread sharing of information, but (unlike NCW) also require broad distribution of decision rights. Hence, Information Age militaries will require not only better information, better mechanisms for sharing information and collaboration, more knowledgeable personnel, and better trained personnel than their Industrial Age counterparts, but also a different C2 Approach. Leadership will also need to be different in Information Age militaries. First, it is the responsibility of leadership to ensure the competence of the elements of the force before it is deployed. At the level of creating and maintaining the enterprise, the leadership role or function includes recruiting, equipping, and training a force that is competent to perform the missions it is assigned. General Marshall was credited with playing this role successfully for the United States during World War II. Second, when the enterprise is to be employed, the leadership function includes making the mission and broad course of action or approach to accomplishing that mission (together in the form of intent) clear to all those involved, as well as ensuring that they are highly motivated. Third, when employing the force, those with leadership roles must also recognise that not all of the leaders involved in a mission will be in the same hierarchical chain of command, nor will their chains of command necessarily come together at a single point. Whether the focus is on a military coalition (in which elements of the force are ultimately controlled by the sovereign nations that deploy them) or a civil-military force that includes nongovernmental organisations and/or international organisations, there may ultimately be no specific leader with authority over all of the elements. Moreover, while coalition forces often appear to be organised into hierarchies, the forces from each nation often have somewhat different goals, rules of engagement, or constraints imposed by their national leaderships. In a similar way, different agencies in an interagency effort may pursue somewhat different agendas within the overall national guidance or may have differing legal or budgetary constraints than the military. More obviously, the charters and agendas of international organisations, private voluntary organisations, or host governments may not always be fully consistent with those of the U.S. military. Hence, more than a decade ago we began to discuss command arrangements to emphasise the fact that decision rights may be distributed quite differently in 21st century operations. We have also coined the term command intent to replace the singular (and we think incorrect) term commander’s intent, because there is no longer (if there ever was except in specific tactical situations) a single commander present in any reasonably large mission space. Rather, there are sets of leaders and commanders distributed across the functions, organisations, and echelons involved. A critical challenge for Information Age enterprises including militaries (e.g., organisations that adopt Information Age C2 Approaches) is creating *consistent* command intent across the relevant set of organisations and functions involved. Note the use of the term consistent as opposed to the stronger term congruent (which we and others have identified as ideal), which would imply that intent across the enterprise is identical. Given the variety of charters, agendas, and cultures (organisational and national) involved in any reasonably sized mission, the best that can be achieved in practice is consistency of intent. This point has also been made by those who seek to replace the classic principle of war “unity of command” with the principle “unity of effort.” Even unity of effort is, from what we have seen in Bosnia, Rwanda, Kosovo, Afghanistan, Iraq, and humanitarian assistance efforts around the world, an ideal that is seldom achieved in practice. We have concluded that what is achievable in coalitions, humanitarian, reconstruction, and peace operations is *unity of purpose*. Indeed, it is often a sign of progress if agreement on purpose allows the different organisations dealing with a situation to de-conflict their efforts (e.g., not all deliver water or food to the same locations) and very valuable if they are able to develop synergistic relationships. In situations where competent organisations with very different organisational perspectives come together to deal with complicated and difficult problems, the minimal Information Age approach might best be described as “constructive interdependence.” In this concept, military organisations might be providing security, lift, and local communications services while international organisations take responsibility for collecting information about violations of human rights and properly identifying and registering refugees; non-governmental organisations worry about emergency supplies of food, water, medicine, and shelter; and the host government focuses on delivering local police services and ensuring that schools and medical facilities are made available to the population. Such efforts require unity of purpose, but they also make it possible for competent authorities to cooperate and collaborate intelligently and use their mutual dependencies as assets. In Industrial Age conceptions, those dependencies are bad—they reduce the control of the organisations involved. That is why they tend to seek deconfliction so that they will have all the assets needed to complete a task or mission and no other organisation will interfere with their efforts. This leads to optimisation at the task level, but actually prevents synergy. Hence, it limits overall performance. This discussion underscores the fact that decision rights are, *de facto*, distributed differently today in the vast number of operations involving the military than they have been in the past. Even in the “high stakes” efforts of the U.S. military, at this writing, in Afghanistan and Iraq, the U.S. counts heavily on coalition partners, works closely with international organisations, depends on private voluntary organisations for specialised services, and is present at the invitation of and works with, rather than controls, a host government. Indeed, the long-term measure of success (a measure of policy effectiveness) in these situations is the ability of the military to (1) engage only in those functions necessary for success, (2) turn over more and more responsibility to other competent and legitimate actors—to further distribute decision rights, and (3) eliminate the need for continued U.S. involvement (or reduce the need to a minimal level). Even with this in mind, however, critics have noted that there are two positions in which some central authority needs to play a role. The first has already been mentioned: the need to develop consistent command intent across the communities involved. In a very real sense, this is a role that can and should be broadly distributed. The “consent of the governed” is enshrined in the theory of democracy and is a major factor in self-synchronisation and in Edge organisations. In military situations, leadership plays an important role in deciding what needs attention, framing the issues, developing approaches that are both feasible and consistent with the values of the society (or societies) involved, and ensuring that the goals and fundamental approach are understood by all those involved. Even in specific engagements where self-synchronisation is intended, leadership has a crucial role in ensuring that the elements of the force are well-prepared. The role of Admiral Nelson in the Battle of Trafalgar illustrates this well. The second crucial role for leadership (or military command) is establishing the capability to communicate (enabling patterns of interaction). At the level of creating and maintaining the enterprise, this can be accomplished by building the now proverbial “system of systems” or by establishing standards or protocols that enable the appropriate level and quality of communication, information exchange, and collaboration required for success. Note that this requires either a central decision (this is the set of interaction mechanisms we will use) or development of a consensus about the appropriate mechanisms. Hence, the same goal (efficient and effective patterns of interaction and distributions of information) can be achieved with either centralised or distributed decision making. However, that goal must be achieved if self-synchronisation is to be enabled in other functional areas. For Information Age military establishments, this will occur through a combination of decisions involving both broad principles and standards. These principles and standards should be developed and implemented within the commercial marketplace. That practice will make it much easier for Information Age military organisations to share information and collaborate with non-military actors around the globe. The SPAWAR concept of composable systems61 is a recognition of this requirement. These principles and standards will also need to provide information assurance in all of their key dimensions. Note that when the time comes to employ the enterprise (or part of the enterprise), the military may be only part of the effort. In these cases, a new more temporary enterprise is being created (e.g., those responding to the 2005 Asian Tsunami or those bringing international relief to Rwanda) and some coherent decision must emerge about the mechanisms that will be used for information exchange and collaboration. Information Age militaries must also be able to avoid chaos. Truly uniform decision rights, in which every individual is involved directly in every decision and has an equal voice in each one, is unlikely to work for most military missions. The obvious exceptions are very small, highly competent, and very capable units such as Special Forces. In their classic roles, they develop rich information about each mission, tailor the team and its equipment to maximise the probability of success, plan the effort in great detail (using peer review to test the concept of operations and the plan), rehearse the mission multiple times, and distribute responsibility for tactical decisions throughout the team. Karl Weick and his colleagues also note that larger professional military (and other) organisations with complicated, high risk tasks (such as those manning the flight deck of an aircraft carrier or fighting forest fires) also delegate considerable responsibility to their junior members.63 However, in all these cases, the people in the force understand: - command intent; - relevant courses of action; - how they map into different possible circumstances; - that there is a high level of trust present in the organisation; and - that those involved are highly competent. These (and some other factors related to information distribution and patterns of interaction) appear to be the minimum conditions necessary for both the broad distribution of decision rights and effective performance. Finally, the distribution of decision rights is crucial to defining the communities of interest involved in complicated efforts like war fighting, national reconstruction, peace operations, or humanitarian operations. A community of interest (COI) is composed of all of the actors who care about and can influence the decisions made on a particular subject. For example, the logisticians supporting a force actually form a community of interest, as do those responsible for targeting or for air defence. Often the distribution of decision rights is best assessed by looking at how they are distributed within the COIs that perform the effort. There is good theoretical and empirical evidence that broad participation in decision making increases the quality of the decisions made. At the same time, all organisations large enough to deal with the complicated and dynamic problems involving today’s military forces are actually made up of overlapping communities of interest. Here again, the Industrial Age model of isolated communities (e.g., logistics, intelligence, plans) is both unrealistic and also enormously inefficient and, particularly in unfamiliar situations, ineffective. It is precisely the interconnections among COIs enabled by Information Age technologies and processes that make synergy and effects-based operations possible. Far more emphasis needs to be placed on these interconnections so that semantic interoperability can be achieved, which in turn facilitates shared awareness and ultimately shared understanding and synergistic actions. However, there may be a price for this broad distribution. All other things being equal, broad participation can mean slower decision making under some circumstances, but this does not have to be true. “Hard” groups, those who have worked together before on similar problems, create organisational artefacts (specialised language, work processes, a repertoire of previous decisions they can reference in their dialogue), and operate much faster than newly formed groups. This, of course, is why professional military organisations stress training and exercises. If done well, they serve to build trust, develop a common language, and create common or compatible processes. Hence, COIs that have worked together before are better equipped for Information Age C2. Moreover, for simple problems or situations (those that are familiar, have a finite number of well-understood alternative actions, and for which the correct action can be selected by rule or algorithm), military organisations develop doctrine or simple tactics, techniques, and procedures (TTPs) that guide effective execution of particular tasks. Doctrine and TTPs enable rapid, high quality decision making in those situations for which they were developed and also bring consistency to the decision making, making the behaviour of one part of the force predictable and understandable by the other parts. Such doctrine and TTPs are, of course, an example of “pre-realtime thinking” in which an organisation identifies situations that it expects to confront (e.g., crossing a river or a minefield) and develops the equipment and processes needed for success, and then builds them into its training and planning. Here again, the performance of the English fleet under Nelson at Trafalgar is instructive. He and his captains were intimately familiar with the capabilities of their own ships and crews as well as those of the adversary. They were familiar with the tactics of the enemy and the ones they wanted to use against them. They also met several times to discuss the battle plan. ### Patterns of Interaction: Social Domain The basic work on patterns of interaction is discussed in some detail in *Understanding Information Age Warfare* (2001). Three key elements are specified for Information Age networks: - Reach (the number and variety of participants), - Richness (the quality of the contents), and - Quality of interactions enabled. The first and third of these are crucial for the patterns of interaction in any future C2 Approach. Together they provide important insights into the C2 Approach of an enterprise: who is “on the net,” what is the quality of their information, and how well can they collaborate? Those who have looked at patterns of interaction often inappropriately limit their view to focus only on connectedness: who is linked to whom within a military force, in a coalition, in an interagency process, or in a broad international effort. While this is certainly part of the idea, understanding patterns of interaction requires focusing on more than just connectivity needs. First, the level of interoperability achieved cannot be ignored. Interoperability means not only technical interoperability, but also semantic interoperability (the capacity to fully comprehend one another) and “co-operability” or willingness to interact and desire to communicate clearly. Second, the range of media across which these interactions occur is also crucial. Voice connectivity over an unreliable system such as HF radio does not enable nearly as rich a set of interactions as video conferencing, the ability to exchange email, or shared whiteboards that can be used to discuss a rapidly evolving situation and the relevant courses of action available. Third, the most desirable patterns of interaction are collaborations (working together toward a common purpose), as opposed to interactions that only involve the exchange of data or information. Collaboration provides the opportunity for the parties to exchange views about the clarity of the data and information, as well as what it means or implies, not just to receive them. Fourth, digital connectivity is qualitatively better than pure voice. Voice alone means that the speaker must be clear, the transmission understandable, the listener must be paying attention and able to hear everything, and the message captured correctly. Even small errors in any of these processes (transpositions of coordinates, missing key words, etc.) in communication can lead to meaningful differences in shared information, shared awareness, shared understanding, decision making, or action synchronisation. The mechanisms by which information is exchanged may also vary across Command and Control Approaches. Traditional C2 Approaches relied heavily on information push: the originator of the information was responsible for deciding what to share, how to organise or format it, to whom to send it, and how often to update the information. This required elaborate planning, detailed standard operating procedures, and specific protocols. In many cases, it also required that the recipient acknowledged receipt. It always meant that the recipient had to ask questions if what was received was unclear or garbled, and to take the initiative if expected information (e.g., a status report or an intelligence update) was not received or if some non-standard information was needed or desired. As satellite and other systems that provided increased bandwidth became available, some items of general interest (for example, weather in an area of operations) moved to *broadcast* media. Here again, the originator was responsible for deciding content and when to update the information. When the mechanism for updating information is broadcast, the user must be synchronized in time to receive the information (unless the broadcast is recorded for playback at a convenient time). Subscription services are more selective as information is not merely broadcast into a “black hole” but is sent to a collection of subscribers who have a need for or express interest in some specific information. Thus when an individual obtains a certain piece of information, he knows who needs or wants it, rather than attempting to assess who needs what. With digital technologies and networks available, the originator of data can take a different approach to making information available. He is free to post the information in a space that is available to those consumers who have a legitimate need for it. While some information needs to be secure and the rights to alter information need to be controlled (as well as providing pedigrees and the other elements of information assurance), digital networks provide the opportunity for a large number of users—and users that the originator did not know needed the information—to access it once it is posted. This approach also enables information to be made available in a more timely manner. Posting information makes it necessary for consumers to know what has been made available and how to acquire it. This requires both education and tools. Moreover, when information is posted, the issues of information quality and authenticity become extremely important. Indeed, this type of information sharing allows the users to *pull* information, which is a genuinely new method of interaction. Increasingly, we are seeing references to the concept of “smart pull,” which assumes an empowered user with both the capacity for intelligent search and discovery, as well as the capability to bundle the volumes of relevant information. Information consumers must also be able to accomplish this “pull” within the time available. The richest possible pattern of interaction would be facilitated if there was a broadband system that digitally linked every entity to every other entity, had full interoperability (technical, semantic, and co-operability), and provided a cyber environment that supported continuous collaboration. At the other end of the spectrum would be a system where each actor can only talk with one, or perhaps two others (a superior and a subordinate), their interactions are very constrained (little bandwidth), and interoperability is low among some or all of the constituent elements. The types of transactions permitted and encouraged (push, post, pull, or subscribe) must also be understood if patterns of interaction are to be understood. ### Industrial Age Patterns of Interaction Within Industrial Age organisations, the patterns of interaction are designed to ensure control from the centre. Hence, the flow of information follows the “chain of command” or the management structure of the enterprise. We can see this pattern in the rule for exchanging information within the U.S. military. All official correspondence is addressed to the commanding officer of the unit, with the specific person who needs the information identified in an “attention” line. This practice reinforces the long-established tradition that information must flow along command lines and that all information within a command is the business of and belongs to the commanding officer. Hence, Industrial Age patterns of interaction mimic the hierarchical structure of the organisation. In many cases, limited bandwidth for communications and a high volume of traffic has led to not one but several sets of hierarchies in Industrial Age militaries. For example, different radio nets are used for command, artillery fire control, control of air assets, and logistics. Another factor that affects the flow of information is that each of the military services has unique communications systems. In addition, the intelligence community has specialised communications systems that are more secure and limit access to those with appropriate clearances and a need to know. The problem is that these specialised hierarchies involve very few horizontal linkages, which makes them weak sources of collaboration. These information flow patterns lie behind the ubiquitous criticism of “stove pipes” in Industrial Age militaries. They both reduce the flow of information and slow it by forcing it to pass through a number of intermediate points, the layers of middle management. Enabling the linkage of “everyone to everyone” does not imply that a direct connection exists between all pairs of entities, or that everyone talks with everyone else. It does mean that, if needed and appropriate, anyone can exchange information and interact with anyone as necessary. The Industrial Age instantiation of such a system was the telephone system, which enabled people to talk around the world, but did not provide (and lacked the capacity to provide) open lines between every pair of telephones around the globe. ### Information Age patterns of interaction Patterns of interaction are actually networks. In the case of NCW and Edge organisations, these are social networks that will be enabled by whatever mechanisms are available: courier, telephone, videoconference, local area networks, wide area networks, the World Wide Web, etc. Social networks also depend on co-operability: the willingness to work together and collaborate when appropriate. The study of interaction networks is an important area of science, research, and development. As the importance of networks in complex structures and complex adaptive systems has become recognised and tools such as graph theory have emerged to help mathematicians and scientists understand their properties, some significant insights for Information Age patterns of interaction have emerged. In many cases, these insights arise from observation and analysis of networks in the real world, ranging from physical interactions in natural systems to social networks involving humans. In other cases, they have arisen from academic efforts to understand networks and their important characteristics. For purposes of examining approaches to command and to control in Information Age militaries, we need to distinguish four different types of networks, none of which is a hierarchy: - Fully connected networks, - Random networks, - Scale-free networks, and - Small world networks. *Fully connected* networks, in which every entity directly interacts with (or is connected to) every other entity, have been explored in detail in the small group literature. In small group research (largely laboratory experiments and field studies), fully connected networks have proven effective at solving complex problems, but are often slower than more streamlined structures such as hierarchies or spokes of a wheel around a central person, particularly when the groups have no prior experience working together. For military systems, however, the important fact is that fully connected social networks do not scale well. For a fully connected network that consists of N nodes, every node that joins adds N-1 linkages. This is costly to support, quickly overloads any available bandwidth, and means that the number of interactions possible for any node rapidly becomes overwhelming. Hence, even if such a system is constructed, the nodes within it must make an enormous number of decisions about when they will interact, with whom they will interact, and how much attention they will pay to any interaction or offer for an interaction. Hence, the endpoint on the spectrum of patterns of interaction—everyone with everyone else, all the time, and using the full range of media—is simply impractical as an approach for a large-scale military organisation or an international effort. That does not mean that fully connected social networks cannot be useful in small operations. In fact, the “clusters” that are valuable in Edge organisations within other types of networks may well be fully connected themselves. Indeed some meaningful research into teams and collaboration indicates that rich connectivity can be very beneficial and may be necessary for success. A *random* network develops when each node has an equal probability of interacting with any other node. The distribution of interaction in random networks can be expressed as a normal or “bell” curve, though it is in fact a Poisson distribution when the interactions are relatively rare. Because of the underlying random properties, these networks are sometimes referred to as “egalitarian” networks. However, these networks are not very efficient—it may well take a very large number of steps or linkages to move from one node to another. This property changes as the size of the network (number of nodes) and the relative density of the interactions between nodes change, but it will consistently form a bell curve if the network is truly random. Random networks will also form relatively few clusters. In other words, while they have low average path length, they also have a low clustering coefficient. Hence, while a great deal of mathematical and theoretical work has been done on random networks, and some on random social networks, they have relatively little practical utility in C2 system design or implementation. Except when they are very dense and approach a fully connected network, random social networks lack resilience, the capability to persevere despite obstacles or setbacks. Because only a modest percentage of their interactions occur across any one area of the space in which they are located, removing a modest percentage of the nodes or linkages in a random network will result in it splintering into a number of unconnected structures. Hence, they are quite vulnerable to attacks and may degrade quickly if linkages are sparse. *Scale-free* networks are characterised by an extreme distribution of interactions among their nodes. A few nodes have a very large number of interactions. Most nodes have very few interactions. The term *scale-free* comes from the fact that this distribution is so extreme that it approaches an exponential distribution. (Actually, exponential distributions go to zero while scale-free networks go to a very low number, but continue their tails well beyond where an exponential distribution would reach zero.) More correctly, scale-free networks have a power law distribution.78 Rather than forming the classic bell shape, this distribution is sharply skewed toward the origin, with a long flat tail. Scale-free networks are found throughout nature, wherever complex adaptive systems develop. Examples include the distribution of branches in a river system and the distribution of nodes in the Internet. These networks are efficient (only a few steps are needed to move from any one node to another). They are also resilient, much more so than random networks. However, they can be vulnerable if an adversary knows how to find the key nodes and the linkages between and among them. The fundamental difference between random networks and scale-free networks is the existence of very long linkages that reduce the number of steps required for an interaction to move from one part of the space to another and a set of naturally occurring clusters throughout the system. These clusters are often described as “hubs” because they will tend to form around one node that is involved in one or more long-haul linkages. Even a few such linkages, provided that they link nodes that serve as hubs for clusters within each region of the network, create a remarkably strong and capable social network. Most stable, naturally evolved complex adaptive systems have been shown to operate as scale-free networks. Because of the importance of the key nodes involved in these long linkages, they are sometimes referred to as “aristocratic” networks in contrast to their random, egalitarian cousins. ![[UC2-Fig-14.png]] The richest, most efficient class of network currently under study is the *small world* network. These systems got their name from the discovery that moving information from one part of the network to another requires only a small number of steps. The distinguishing feature of small world networks, however, is a very large clustering coefficient. As a result, a link to any one node is readily tied to a number of other nodes. This is the actual structure of most effective groups of experts. They are, in one sense, random social networks growing toward scale-free social networks but not yet characterized by the number of long-haul linkages that would transform them to meet the definition of scale-free. In other words, these small world networks link together clusters that can be thought of as communities of interest or communities of practice. Each such community forms its own “small world” which may be linked to other small worlds by some individual nodes that act as connectors. These small world networks can evolve from or evolve into scale-free networks. However, if they are temporary (for example the “hastily formed networks” that arise to deal with the aftermath of a natural disaster or to carry out a specific, temporary military mission), they will function relatively efficiently and robustly while needed, then gradually disappear as the interactions become less frequent and the linkages atrophy or are abandoned. The networks required for self-synchronisation or Edge organisations are small world networks in which those with the relevant knowledge and capabilities form richly linked and frequently interacting clusters that permit them to exchange information, develop shared situation awareness, and collaborate in order to synchronise their plans (explicit or implicit) and undertake synergistic actions. ### Richest network structure The richest and most resilient network structure (the pole for the distribution) appears to be a hybrid that looks at the global level like a scale-free network, but at the intermediate level is composed of small world networks, and at the local level fully connected social networks. This combination appears to provide the blend of efficiency, effectiveness, and resilience needed for large-scale enterprises operating in multi-dimensional, dynamic environments. These patterns of interaction are capable of becoming complex adaptive systems. Of course, like other ideal types, this hybrid form does not exist today ### Distribution of Information The term *information* here covers a range of phenomena including data, information, understanding, knowledge, and wisdom. We have defined these terms elsewhere, as have other researchers. Recent work of the SAS-050 NATO Working Group has stressed that data, when placed in context such that it reduces uncertainty, becomes information, while information becomes awareness when it passes from information systems into the cognitive domain (a human brain). Humans, as individuals, actually hold awareness of situational information and combine it with their prior knowledge and mental models (which include perceptual filters that may prevent full awareness of some information) to generate situation understanding, which includes some perceptions of the cause and effect relationships at work and their temporal dynamics. As was discussed earlier, these elements of the sensemaking process also drive decision making. We speak of shared information as that directly available to more than one actor in a social network, and shared awareness and understanding as those elements that are common across more than one entity or node. This distribution of information dimension refers to a key result of the C2 processes within a military organisation, coalition, or international effort (e.g., reconstruction, peace enforcement, humanitarian assistance) involving military forces and civilian organisations, which might be other government agencies, international, or private entities. As noted earlier, the distribution of information is impacted by the distribution of decision rights (which includes who makes the choices about information distribution processes and the creation of the infrastructure by which information is shared and collaboration is carried out, as well as who is entitled to what information) and the patterns of interaction (who is able to acquire what information). However, the distribution of information also has a dynamic of its own. The concept of the distribution of information also includes the richness element in network-centric thinking. Richness focuses on the breadth, depth, and quality (correctness, completeness, currency, consistency, etc.) of the information that is available. The pattern of distribution of information within an enterprise arises partly from the allocation of decision rights, partly from the patterns of interaction, partly from the willingness to share information, partly from the willingness of individuals to acquire it, and partly from the tools and skills they have to acquire it. This includes and indeed is strongly impacted by their ability to collaborate, particularly when considering the more strongly cognitive aspects of information—understanding and knowledge—and the ability to share information, awareness, and understanding. Ultimately, the distribution of information governs the capacity for sensemaking at both the individual and collective levels. As a simple dimension, the distribution of information can be thought of as ranging from fully centralised repositories (e.g., the old mainframe computer that held everything for a company or organisation and the access of each user was predetermined and controlled by a central authority) to fully distributed (networked) approaches wherein everyone has access to everything, but storage is redundant. ### Industrial Age distribution of information The Industrial Age involved functional decomposition and role specialisation. Hence, during that era information was distributed according to the specific needs of each user. This required preplanning and the identification of the “owner” of each information element. Owners were responsible for ensuring the quality of the information and for ensuring that it was distributed according to the organisations plans. Changes in the processes for collecting, developing, or distributing information required centralised decision making, which would often be conducted by committees or groups of specialists. The engineers designing C2 systems spent a great deal of time developing information exchange requirements (IERs) that specified who needed access to what information under what circumstances. These IERs then drove which linkages were enabled with what capacity. More importantly, they also resulted in de facto decisions about who would not have access to information. In other words, they constrained interactions and limited the distribution of information under the Industrial Age assumption that they knew, with precision adequate to engineer these systems, the threats and circumstances under which military forces would operate. Industrial Age militaries followed this practice of preplanning, centralised systems, and constrained distributions. For example, intelligence information was kept in intelligence channels, logistics information in logistics channels, etc. These functions came together at their common bosses: the military commanders. Those in the command roles could, and often did, mandate distribution across functions where integration appeared useful or wise, but unless such mandates were issued, information stayed in functional channels. Military echelons also acted as constraints on the distribution of information. Here again, command decisions (often in the form of standard operating procedures) were required to move information across echelons. The impact of these Industrial Age patterns of information distribution can be seen in the sequential nature of the decision making processes. Figure 15 shows that this system of information distribution all but guarantees that all decisions are based on outdated information flowing upward and outdated guidance flowing downward. It also highlights the possibility for a lack of synchronisation or synergy across functional areas such as intelligence, operations, and logistics. ### Information Age distribution of information As the capacity for sharing information has grown, the distribution of information has tended to break away from rigid guidelines based on function and echelon. Similarly, where the capacity for collaboration has emerged, the quality of information has improved (people with different perspectives and different knowledge or expertise have an opportunity to comment on information and compare it with other information, understanding, and knowledge) and the patterns of distribution have been enriched. These processes are well underway in some aspects of today’s forces, but are still a very long way from the “robustly networked force” envisioned in NCW. In a genuinely Information Age enterprise (or an Edge organisation), all information is available to all the entities, with constraints minimised and focused on necessary aspects of information assurance (privacy, integrity, authenticity, availability, and non-repudiation). However, information assurance is not adequate in itself. Users must also have the tools to find relevant information (discovery and search capacity), a rich understanding of what information is available, and the capability to process it so they can “digest” it or use it to add to their sensemaking ![[UC2-Fig-15.png]] Information assurance is also, by itself, an inadequate characteristic for Information Age information distribution because it does not take into account the need for agility. For example, it needs to consider the distribution of storage as a defence against loss of connectivity from malicious attacks, natural disasters, or systems disasters, as well as the need to rapidly redistribute information in dynamic or unforeseen developments. Both robustness of the system and redundancy are key elements for maintaining effective information distribution over time and across function. ### Conclusion The core of understanding Command and Control lies in its definition and the three key dimensions by which alternative C2 Approaches can be characterised and differentiated: the allocation of decision rights, the patterns of interaction that occur, and the distribution of information. These apparently simple dimensions are interrelated and drive both the command function and the control function. Together, they define the C2 space. The chapters that follow put these key dimensions under the lens of a “value view” (Chapter 7) and “process view” (Chapter 8), both to highlight their implications and also to provide examples of how they can be expected to play out in forces, coalitions, inter-agencies, and public-private partnerships involving the military.