Back to RMA Research Page

The Emergence of a Command Network

John W. Bodnar, CAPT, USNR (Ret) and Rebecca Dengler, 2LT, USMC

U.S. Naval Academy, Annapolis, MD 21402

Naval War College Review. 49, Autumn, 93-107 (1996)

In today's dynamic world the effectiveness of any organization at every level is dependent on how well the individuals in the unit can cooperate. Whether the organization is as small as a platoon composed of individual Marines or as large as a joint military composed of individual services, coordinated actions of any unit depend on the flow of information or commands among its individual parts. Therefore, the effectiveness of any organizational structure to respond to a rapidly changing environment depends directly on the ability of the individuals composing the unit to communicate among themselves.

We suggest that in the current Revolution in Military Affairs (RMA) huge changes in organizational doctrine are emerging, and specifically that among the most noteworthy features of the "RMA" is the military's ability to communicate on a global scale, for this ability offers an exceptional opportunity to enhance effectiveness. Organizational structures, though diagrammed as they were fifty years ago, actually operate very differently in the Information Age. In fact, procedures (or practices) and doctrine that were effective in the past for very small units can now be effective for an entire joint military.

The emerging national military organization, which we term a "command network," has the strengths of both a network and a hierarchy. It has, in fact, always been used in such settings as infantry companies, shipboard departments, or submarines - or even small primitive societies. Recently, however, command networks have arisen in much larger organizations; one unexpected result has been that the responsibilities of their members and leaders have changed dramatically.

We argue in this article that the biggest challenge in the growth of command networks - even at the national level - is that leadership and organizational doctrine must be redefined. To examine that challenge, we must ask, what are command nteworks? Why have they exploded in size in the last century? What have been their effects upon the concept of responsibility within organizations?

Informed Inaction vs. Uninformed Action

Organizational structures are based primarily on lines of communication. The strengths and weaknesses of any organization, therefore, depend on how the individuals in a given unit communicate among themselves - either by passing data (observations about the environment), information (assessments of the meaning of data), commands (orders for action), or doctrine (rules for action). In the same way that chemists use theoretical substances such as ideal gases and perfect crystals to study the structure and dynamics of molecular organizations, we will use two theoretical units - a pure network and a pure hierarchy - to study the structure and function of military organizations (Figure 1).

Figure 1. Types of Theoretical Organizations

In a pure network all the individuals are equal and automous and all possible lines of communications can be used: with no leader all organizational decisions, if any, are made by consensus, and any individual can interact directly with any other individual in the network. On the other other hand in a pure hierarchy the individuals have tightly defined lines of communication: all communications are within the chain of command where data and information only go up the chain and command and doctrine come down the chain. Of course "pure" networks or hierarchies don't exist; a real network like a commune, a car pool, or the Internet does have hierarchy in it, and a real hierarchy like Nazi Germany or the current Iraq does have network in it. Nevertheless, our theoretical organizations allow us to examine the strengths and weaknesses of each.

A network provides rapid and reliable communication which emphasizes assessment. In a network each individual is only one communication step from the source of any information. New data collected by any individual in the network are accessible to all members of the network, and all individuals in the network can become involved directly in assessment of all available data through a single communication step. Each individual must independently filter all the data to determine the proper course of action. On the other hand, in a hierarchy individuals have access to information only indirectly throught defined communications channels. Communications in a hierachy are slow because multiple middlemen may be required, and assessment may be less reliable because information is filtered at each step.

A hierarchy provides reproducible long distance communication which emphasizes command and doctrine. In a hierarchy each individual has an indirect line of communication defined to every other individual in the hierarchy. New data can be routed to commanders who assess them and command the organizational response; in this way all individuals in the organization are coordinated to a single command even if they were too distant to receive the initial data. A hierarchy can act in a coordinated manner even beyond direct communication distance because individuals can store commands as doctrine and act by doctrine even when cut off from the rest of the organization. Individuals in a hierarchy act on incomplete information because they often have only the command and not the data or reasons for it. On the other hand, in a network there is no reproducible method for integrating assessment of information to courses of action, so any coordinated action often must be negotiated on each separate occasion. Additionally, in a network there is no method for passing information beyond a single step so a pure network cannot even exist as a unit unless every individual can communicate with every other individual.

A network provides rapid independent action by individuals. In reacting to the environment, an organization must collect data and assess it so decisions are made for command action. In a network each individual is equal, can operate independently, and can act rapidly at any time on the basis of all the information available to that individual directly or from other members of the network. On the other hand, in a hierarchy individuals only act through command or doctrine. A decision on a course of action often comes by passing information up the chain of command and commands back down the chain - a process which must inevitably be slower than response in a network.

A hierarchy provides rapid but inflexible doctrinal action by individuals. Rapid response can be programmed into a hierarchy through doctrine. A command decision can be made quickly when a certain situation arises if there is already a doctrine to apply. This approach allows rapid action in the hierarchy to be smarter but less flexible than in a network. It is often smarter because the doctrine is refined by integrating the ideas of many individuals on reaction to a particular type of situation based on many historical examples. Should the current situation not fit the doctrine exactly, the result may be disastrous.

A network provides flexible local control. When individuals are close enough to communicate directly, a network is most efficient in allowing them to coordinate their actions because any individual can interact equally with any other individuals. Conversely, in a hierarchy physically contiguous individuals may report through distantly related chains of command, and coordinated action may require relaying assessments far up the chains with decisions to coordinate them made far from the scene of action.

A hierarchy provides long term control. Hierarchies can coordinate action even beyond the lifetime of the individuals in the organization. Whenever a response to a particular situation works, it can be passed to all individuals in the hierarchy and further stored as doctrine for future use - even if all the individuals in the hierarchy change over the course of time. Doctrine is, therefore, the collective wisdom of all members of the hierarchy that ever lived. Should a new situation arise that does not fit the old doctrine, only the commander has the authority to change the doctrine. On the other hand, any time an individual leaves a network, the wisdom of that individual leaves also, so a network may be constantly "reinventing the wheel". Should the situation change, there is no old doctrine to unlearn for correct action.

In short, a network can assess the current situation very well and react very rapidly in local situations but has no intrinsic mechanism for trading or storing command decisions - leading to informed inaction. On the other hand, a hierarchy can coordinate action very rapidly on the basis of command decisions or doctrine often made on incomplete or out-of-date information - leading to uninformed action.

Informed Action: Command Networks

Clearly, an organization that can survive and coordinate many individuals for a long time in a rapidly changing environment must be some combination of a network and a hierarchy. Such organizations have evolved in many times and places which have characteristics we will call a command network. While the following discussion could apply equally well to a tribe, infantry company, shipboard department, small factory, a business "team", or even a small town, we will use the command organization on a submarine as an example of how a command network can integrate the strengths of a network and a hierarchy.

Figure 2. The Command Network

In a command network the organizational structure changes on a minute-to-minute basis dependent on: a) the individual tasks at hand; and b) the overall mission. It works on three levels. The reflex level includes the workers who perform the task, collect data on the environment, and pass along information up the chain of command. The autonomous level includes team leadership for specific tasks. The command level is the overall comander who sets policy for accomplishing the mission. In this structure command and doctrine follow a downward flow, but specific individuals in the chain can change as new tasks arise; however, data and information can flow freely in any direction horizontally or vertically - even jumping from top to bottom if necessary.

The operation of a command network is best illustrated by a single example of one task in a complex mission - the setting of the correct search depth for a Mark 48 torpedo by a submarine battlestations organization. When in a sub vs. sub attack scenario, the Commanding Officer (CO) is in the Control Room with task leaders such as Fire Control Coordinator (FCC), Weapons Control Coordinator (WCC), Engineering Officer of the Watch and others reporting to him. The WCC then has the weapons control team and torpedo room reporting to him. When the ship is closing to attack, the CO will insure a torpedo is ready by ordering "Make tube one ready in all respects for a submarine contact." The WCC then passes that order on to the fire control team and torpedo room, and a few minutes later with the task accomplished reports back "Tube one is ready in all respects." This procedure dates back to World War II torpedoes which ran straight until they hit something; even though a Mark 48 torpedo can have up to forty separate commands (including such things as search speed, ping interval, search pattern, and search depth), the command network insures that all forty are set properly each time with only the single order.

The strength in the command network is in efficient information dissemination and unspoken assessment in a tactical situation. Long before battlestations was called the Weapons Officer and Fire Control technician had a long session with tech manuals, sound velocity profiles, and weather data for the oparea and came up with a recommended list of all torpedo settings for as many conceivable attack scenarios as possible (surface or subsurface, above or below the layer, etc.); then the Weapons Officer cleared them with the CO prior to ever reaching the operating area. Thus when the order is given to make the weapon ready, both the fire control technician and the torpedo room have a plan of action; only the target and own ship depth need be compared with the latest sound velocity profile to ascertain the best search depth (so that the weapon will be on the same side of the layer as the target). The WCC obtains that data by direct observation of ship control panels or asking the FCC directly the best guess on target depth. The WCC then passes the appropriate information by sound powered phones (on a party line with the fire control technician at the Control Room firing panel, a second fire control technician at the torpedo room monitoring panel, and the torpedoman chief in the torpedo room). Then with no further order the fire control tech enters all the appropriate settings into the torpedo computer, announcing each step on the phones such as "Search depth set at 400 feet." Without any further comment the WCC in the Control Room, and the fire control technician and torpedoman chief in the torpedo room all listen and monitor the indicator lights on the appropriate panels verifying for themselves that each indicator light comes on correctly and that the actual value is consistent with doctrine and the predetermined plan of action. If all goes according to plan nothing more is said until the report "Tube one is ready in all respects." Even then, if the CO has a question or if he believes conditions have changed from the plan, he can look over the shoulder of the WCC to verify the settings. Thus every setting of the forty on the Mark 48 is discussed by at least three people, and verified by four prior to launching the weapon - even using only the age-old commands.

A command network can act as fast as a hierarchy because the traditional chain of command is used to pass command and doctrine; on the other hand a command network can assess information as well as a network because open communication channels among all individuals allow everyone who has the need and ability be a part of the decision process.

Communications and the Size of the Command Network. Command networks have been around for a very long time, but they were only used in small organizations prior to World War II. However, they have begun to emerge in much larger organizations. Now in carrier battle groups (CVBGs), the individual ships simultaneously report to an Antisubmarine Warfare (ASW) Coordinator, an Anti-Surface Warfare (ASUW) Coordinator, and an Anti-Air Warfare (AAW) Coordinator. If a ship contacts a sub and missile simultaneously it may take independent (reflex) action to address the most pressing threat while coordinating with nearby ships and the ASW, AAW, and ASUW Coordinators on what next. Thus command networks now prevail in CVBGs and are beginning to be seen in "joint" operations among all the services. But why has the explosion in size of command networks been so recent?

The reason seems to be communications capability, which may be the limiting factor in the size of any network. For a network to even exist each individual must be able to communicate on a day-to-day basis with every other individual in the network (even if they choose not to do so). Therefore, a network can be only as large as a single communications link. It follows then that to make a command network effective the commander must be able to interact with all members of the command on a day-to-day basis (even if that communication is merely monitoring communications between individuals within the command). Until the advent of radio, video and computers, large organizations could work only by implementing hierarchies, and command networks were limited by how far one could communicate (usually as far as one could drive in a day) or by how many people one could communicate with daily (about a hundred or so). Larger organizations would ultimately have to turn to hierarchies to allow long range communication or encompass more personnel (using indirect communication through multiple middlemen).

Unquestioning - or Unquestionable - Obedience? It is almost inevitable that the more hierarchical an organization, the more centralized are authority and responsibility within it. In the past large hierarchies could operate very effectively in this manner. Nonetheless, the strictly defined chains for both information and command meant that someone at the bottom of the chain would virtually always be acting on very incomplete information. Consequently it was extremely difficult to evaluate the efficacy, legality, and even sometimes the morality of an order.

As communications became more sophisticated, however, it became possible to obtain information by means other than the chain of commnad. When it did, the concept of responsibility within an organization began to chnage. Moreover, judges at the trials of Nazi war criminals at Nuremberg had rejected every defense argument that "obedience to orders" could excuse grossly immoral acts. Every leader - indeed, every person in uniform - is held to be a moral being repsonsible for each action, including orders issued or carried out. The evolution of an Army training manual reflects this obligation. Its 1941 edition addressed the issue of authority in these terms: "Success in battle, which is the ultimate aim of all training, requires the co-operation of every individual to the common end. Cooperation requires centralization of authority. Someone must make the decision; when once it is made, everyone must carry out the decision energetically and unquestioningly."<1> In the 1953 version of the manual<2> a single suffix was changed in that paragraph - "unquestioningly" became "unquestionably"

This tiny change signalled a major shift in the concept of individual responsibility in an organization, toward one that combined what is demanded in a hierarchy with what is required in a network. In a network responsibility for any individual's actions rest with each individual since there are no commands or doctrine in a pure network - only data and information. On the other hand in a hierarchy responsibility rests with the commander since each individual receives only commands and doctrine - not the data and information behind the decision. In a command network data and information is shared along with the commands and doctrine so responsibility for any action lies with both - the commander because he had all the data and information to conceive the order, and the individual because he had enough data and information to assess whether the order was effective or legal.

In the command network responsibility flows both from the top down and the bottom up. However, turning this concept into doctrine has been problematic. The first codification of the concept came through W. Edwards Deming's Total Quality Management (TQM) which was modified to Total Quality Leadership (TQL) by the military. Two major new additions to leadership concepts espoused by Deming were those indicated above which could only be used by an information- and communication-rich organization: 1) effectiveness of an organization depends on collecting the most data possible and analyzing it most effectively, and 2) the best use in information comes at the lowest possible level. In essence TQM and TQL suggest that the best organization is one that comes closest to a network yet still can act effectively - a command network.

The Leadership Challenge of the Command Network

The implementation of TQL from TQM was controversial, and it sometimes became more a battle of buzzwords than a new definition of organizational doctrine for the Information Age. We suggest that this stems from the fact that leadership in a command network is significantly different from any prior military organization larger than a company or submarine. Going beyond TQL to a new organizational doctrine for a "joint" military requires reassessing and redefining organizational dynamics - especially the role of senior leadership.

The need for a new kind of leadership for the Information Age has led to a de facto emergence of a leadership model often termed "leadership by negation". Some characteristics of organizational dynamics and leadership in a comand network are counter to the way that we have run large units within our military - but which have always been around in an infantry company, a shipboard department, or on a submarine.

The organization must be close as possible to a network yet retain the ability to act effectively. The "smartest" organization is a network which can best assess data and information. However, the chaos of multiple information channels minimizes a network's ability to act rapidly and effectively. An organization which maximizes the communication channels of a network yet can still act in a coordinated manner is one always "on the verge of chaos".

It must be clearly decided who will decide. During submarine battlestations several people from the CO down to the torpedo room phone talker could set the inputs for the Mk48 torpedo. However, the organization only works effectively (and silently) if everyone in the chain knows who has responsibility for each decision.

Decisions must be made at the lowest possible level. Where in a unit any decision is made is a tradeoff: the individual lowest in the hierarchy usually is closest to the action and has the most pertinent information; but the commander usually has the most comprehensive overview of the situation. Therefore, different types of decisions are best made at differing levels. Reflex actions such as immediate actions taken for flooding on a submarine must be done as fast as possible by the first person on the scene; the "smartness" of those decisions is increased by preformed command doctrine for those actions, but even then, the person on the scene is responsible to interpret doctrine for the conditions at hand. Similarly, autonomous group actions, such as ASW defense in a CVBG, need to be coordinated, but do not require direct minute-to-minute attention of the CVBG commander - only clear doctrine of how ASW affects the unit mission. Making any decision at the lowest level provides the most rapid response, the most immediate access to primary data, and the most "brainpower" from multiple individuals directly involved, while making the decision at the highest level provides a comprehensive overview of how it affects the entire unit. Since the high level input can be handled as doctrine learned and practiced prior to a tactical situation, decisions in a command network are best handled at the lowest possible level.

"Any fool can obey orders!" - the words of Sir John ("Jackie") Fisher, the great reforming First sea Lord of the Royal Navy before and during World War I. Though made, of course, in a different context, his remark has special weight in a command network, wherein an order is in essence an agreement between holders of shared responsibility. Here, plainly, is a stumbling block; but the issue arises unavoidably. In a command network, no order should be executed without examination by its recipients, for two reasons. First, the composite "brainpower" afforded by the network is mobilized only if every individual involved in an action is always thinking about its consequences. Secondly, if both the commander and the individual are responsible for an action, then both must be propietors of it. There are, of course, situations where urgency, safety, or other overriding considerations require simple and instant obedience; doctrine can identify these, and the hierarchical dimension of a command network admits of them. But they must be exceptional. In general, a commander who requires that his orders and decisions must be followed without question will soon find that he is the only one in the organization who is thinking.

Continual training is critical to command network operation. If each order is to be examined, then constant training under realistic operational conditions is required. Foreseeable or fundamental ambiguities must have been resolved in advance. In exercises, commands can afford, for educational purposes, to let unintended consequences of orders run their, course or to let subordinates question them ("But, sir . . . !") and take the time to examine the matter or at least discuss it soon afterward. As an actual current example, when Trident submarines on patrol go through launch-sequence practice, only a few officers on board know that it will not culminate in the firing of all twenty-four nuclear-armed missiles. Everone on board thus has ample opportunity to question the details and to ponder any moral reservations they may have about nuclear weapons; should the "real thing" ever occur, their crews will be ready, both technically and personally.

Commanders set guidelines for unit action. In a command network where decisions are made at the lowest possible level and practiced often, the commander gets multiple opportunities to evaluate his subordinates in action. Thus command input becomes guidance of how particular decisions affect unit mission. In a command network, the best commander will have his unit trained so that in tactical situations he will have virtually nothing to do but watch his unit perform. In a tactical situation with his team coordinators making recommendations on unit action or taking appropriate autonomous or reflex actions, his decision making input will almost exclusively be resolving disagreements among his subordinates.

The commander will let pass without comment any decision within his guidelines. Within a command network so many different action and decisions are happening simultaneously with multiple individuals monitoring, questioning, and assessing each one, most network communications must necessarily be received silently. If everyone on the submarine fire control team acknowledged each correct Mk48 setting made by the fire control technician, there would be chaos. Similarly, any individual in an E-mail network receives many times as many messages as are actually read which are many times as many messages as are actually answered. Only actions that are incorrect or not understood are overtly acknowledged. This concept is often called "command by negation" - a term full of pejorative connotations. In fact, the useful redundancy afforded by silent monitoring up and down the chain of command is a major strength of command network leadership and should earn a more positive name: perhaps "network leadership" or "command through confidence".

There is no doubt that a new leadership doctrine is required for the new emerging organizations of the Information Age. TQL and leadership by negation are a beginning. Those models can and should be transcended, however, by reassessing the fundamentals of responsibility and communication.

Toward a National Command Network.

A major advantaage of a command network is its flexibility not only to changing missions but also to changes in the size of an organization. This suggests that it may be possible to form a common doctrine within the military to form a "virtual organization" based on a command network that can rapidly regroup and reassemble in response to a variety of threats.

But what is the maximum possible size of a command network? While a worldwide hierarchy has already been achieved, a worldwide network requires that every individual can communicate with every other individual anywhere in the world and have access to all data, information, doctrine, and commands stored anywhere on the network - on a real time basis. We suggest that this will be technologically possible within the next decade; thus the ultimate size of a command network will be limited by people - not technology. Therefore, rather than belaboring technology, new doctrine should assume a worldwide network with characteristics of the World Wide Web and begin to apply lessons of the face-to-face command networks to a web-based national command network. While we can presently only identify a framework for such organizational doctrine, we suggest a "virtual" command network is possible that can rapidly form and reform effective units at the CINC level.

First, a command network should be a network when considering options but a hierarchy when issuing and implementing decisions. Given certain precautions against "groupthink" and other handicaps, networks generate the most options and produce the fullest information, so that in the evaluation and deliberation process a command network should structure itself as closely as possible to a pure network. When (as is usually the case in military operations) there are considerable time pressures, decisions are best reached by a well informed hierarchy. In any event, unless a consensus obtains, the need for closure requires an authoritative (i.e. hierarchical) decision. Once a decision is made, command and doctrine should flow downward through a hierarchy - silently monitored and validated throughout the surrounding network. Thus the individuals in a command network need to be comfortable switching between hierarchy and network modes depending on whether the organization is considering or implementing decisions.

Second, each individual should give and receive commands and doctrine through the command hierarchy. Individuals within a command network continue to follow the traditional chain of command whenever coordinated action is necessary. This means passing recommendations for action up the chain as well as commands and doctrine down the chain. As in the past orders are acknowledged by a repeat-back (signifying both understanding and acceptance of responsibility), and the order is not complete until the senior receives the acknowledgement.

Third, each individual should pass data and information to the network while monitoring and filtering information from it. This is the major challenge in defining the doctrine for a military command network. The more petinent information any individual has, the better informed his decisions or recommendations can be; the more silent monitoring that can be accomplished throughout the network, the more advised decisions and actions can become. Thus, in theory each individual should monitor all information available throughout the network; doing so in practice, however, could rapidly lead to act due to information overload. To maximize the network character of a national command while still retaining the ability to function, a number of basic steps are possible.

Fourth, commanders need to balance their dual roles - at the command level of their own unit and at the reflex level of the larger unit of which they units are a part. The military is already forming a "nested" command network where the commander of a single unit is the liaison of that single unit in a larger command network allowing that commander to perform much more effectively than in the past. With the distributed decision making inherent in a command network, the commander has fewer operational decisions to make than before; his major job is to define guidelines for action in accordance with unit mission. At the same time, his major job within the reflex level of the larger command network is to interact with other units and autonomous level coordinators to adjust his unit mission and operation to that of the larger organization. We suggest that upward and parallel links from the unit CO are currently the weakest in the emerging command network - because traditionally command operations were handled in a hierarchy. With the web becoming available the submarine or ship CO now has all the assets of the web to monitor and evaluate information from above and in parallel (just like the division officer always has done). Thus his jobs up and down the chain are, for once in history, compatible - but only if he knows how to act simultaneously at the lowest and highest levels of command.

Command networks have been around (if not by that name) in the American military for a very long time, and the best leaders knew how they operated when they were majors or lieutenant commanders. In the Information Age, forming a command network that encompasses the entire military is a possibility but depends on re-examining both leadership models and organizational doctrine. The challenge for leaders is to apply the lessons they learned and information channels they used as department heads or company officers to the larger organization they have (or will) become responsible for at more senior levels. This entails reassessing and going beyond the lessons of Total Quality Leadership and leadership by negation, taking advantage of the new ability to integrate and assess large amounts of information and adopting the concept of shared responsibility for every order. The key to jointness at a national level, as at any other level, is not technology itself but how it allows individuals within the organization to interact and communicate to make their efforts mutually reinforcing with respect to a common, proper goal.

<1>John McComsey and Morris O. Edwards. The Soldier and the Law. (Harrisburg, PA, The Military Service Publishing Co., 1941).

<2>Morris O. Edwards and Charles L. Decker. The Serviceman and the Law. Sixth Edition. (Harrisburg, PA, The Military Service Publishing Co., 1953).

Captain Bodnar holds a doctorate in biochemistry and is a member of the Chemistry Department Faculty at the U.S. Naval Academy. He served twenty-three years in the Navy and Naval Reserve. While on active duty, he served aboard two attack submarines. As a Naval Reservist he was associated for ten years with the submarine force and for six years as a technology analyst with the Office of Naval Research and the Naval War College. Second Lieutenant Dengler is student at the Basic School in Quantico, VA.

Content in this area of the website is created by faculty and is not official college material. Comments should be directed to the author.

Last modified on July 20, 2006 17:37