As the study process evolved between 1996 and 1999, however, our intuition began telling us things that were difficult to show in analytical terms because attrition-based analytical models mask some important effects. One of these was the criticality of being able to put the enemy at risk in the "dark places" and complex terrain the opposition uses to evade precision firepower, and the difficulty of doing this with "speed, knowledge, lethality and precision" alone. The number of infantry squads and fighting vehicle crews grew in our experimental designs while "over head" from platoon headquarters upward grew at a much lower rate. New technologies have the potential of enhancing the productivity of all Army soldiers in all functions, but the greatest productivity enhancements seem to be in many of the Army's supporting functions. These could yield large personnel savings that could be shifted to increase the combat fraction of the force.

A second critical insight was the high value of the operational and tactical art. Fighting an asymmetric enemy successfully is more dependent on art than science. Winning these contests involved more than leveraging our vast superiority in precision firepower and control of the air dimension to create lopsided "loss exchange ratios." It also involved leaving the enemy no unopposed dimension. In essence it meant challenging his sanctuaries, protecting our vulnerabilities against his asymmetric approaches while using our asymmetric advantages in terms of means and dimensional supremacy to erode his strength and will, disintegrate his organizational and functional coherence and dislocate his efforts in time and space. The requirements of the art become more difficult as one moves from the tactical to the operational and strategic realm. At the lower levels, information about current reality is extremely "actionable" and often decisive. At the higher levels, information about current realities is but one indication of what reality might be in some future time frame, more relevant to current decisions of the commander. Anticipating branches and sequels of the current enemy activity, and comparing those to the possibilities of the friendly side are more germane to decisions at these levels. More often the art is less in reacting to the enemy than it is in conceiving viable options and reliably imposing new realities on the enemy. Information technologies are extremely helpful in the scientific aspects of tactical and operational design, but creativity in this complex environment is still an art that depends on the brains of commanders and their staffs.

A third critical insight was that information technologies enhance span of control more than they do span of command, and that span of command is the limiting factor in units designed for combat. Recognizing, measuring, tracking and controlling many things and processes, are taxing tasks for the human brain, but relatively easy challenges for computers. The workload associated with controlling a given number of subordinates decreases significantly as information technologies are applied. Generating alternatives, accounting for and coordinating personalities, gauging and boosting morale, weighing unquantifiable risks and anticipating human reactions by both friend and foe require personal attention by leaders and principal staff. These are command workload factors. Information technologies can be helpful to commanders for this work, but far less so. In the early Army After Next organizations with spans of control of up to eight, it was often necessary to create temporary intermediate command echelons to focus the efforts of a smaller subset of organizations against particular tasks in complex settings.

A fourth critical insight was that span of command should vary from echelon to echelon, and among types of organizations, based on the complexity of the "command" workload. I believe that for several reasons, span of command can be greater at higher echelons than at lower. A corps can handle more divisions than a battalion can handle companies, or a company, platoons. Higher commanders and staffs are more experienced; higher staffs are more robust; the potential rate of change in the combat situation is more moderated; and all divisions are not as likely to be equally engaged. On the other hand, the more I began to understand what will be expected of companies and platoons -- the variety and complexity of their tasks, the intensity of combat when they become engaged, and the price of command inattention -- the more I have become convinced that information technologies will be vital to success at these levels, but will not permit an expansion of span of command beyond what we have in units today. In fact, I suspect that after more study we will want to reduce those at company level to achieve even higher leader to led ratios, to compensate for mission complexity with organizational simplicity.

Fifth, while the discussion of span of control and span of command only addresses the number of line elements in an organization, it ignores the issue of what specialized supporting organizations must be grouped at each echelon and how new technologies affect this aspect of organizational design. Every echelon requires three kinds of specialized capabilities. The first of these is support to the functioning of the commander and staff -- to facilitate decision making and other aspects of battle command, and to provide "housekeeping," protection and mobility. The second of these is to support the functioning of the line organizations, providing capabilities that are periodically required and distributed by mission need and to weight the main effort. The third of these specialized capabilities is required to set the conditions for the success of the weighted line organizations. This would require range and reach beyond the capability of line organizations. For instance, it would be important to prevent the reinforcement of the object of attack, to ensure the protection of line elements playing the key roles and to prevent surprise while engaged. The product of evolving technologies -- speed, knowledge, lethality, range and precision -- can revolutionize how these three kinds of supporting capabilities can be provided at each echelon. It will be important to maintain a minimum essential layer of capabilities close at hand for rapid response and robustness, but much more capability of all kinds could be echeloned at higher levels if the range and reach of that support -- its effects -- could reliably be placed at the disposal of subordinate echelons. I can visualize a system of combined arms support in depth as much as three echelons deep. In our current system reliable support is normally only available one echelon deep, and much of it has to drive from point "A" to "B," costing time and security resources. In the future the effects of such support could be available in minutes across wide arcs of distance without displacement.

Finally, by the end of 1999 it began to dawn on me that the logic for the number of echelons is more a matter of perspective on the problem and our ability to practice the art of war and less a matter of controlling things and processes or even of command work-load. This point is supported by Elliot Jaques and Steve Clement in their book Executive Leadership. Both have been consultants to the Army and industry for more than thirty years. They have helped many of our largest corporations shed unneeded echelons. They maintain that the chief value added to leadership at each echelon is the ability to apply a perspective to problem-solving that is different in time span and the composition of factors that condition decisions from that above and below it. They maintain that any layer which does not exhibit this uniqueness ought to be removed regardless of considerations of span of control. They insist that unnecessary layering creeps into organizations because of preconceptions about what is an appropriate span of control. In the commercial world, information technologies do not replace echelons in properly structured organizations. They do reconfigure processes, and they change the nature of work done at these echelons. Applying information technologies in the military realm will not lead to needing fewer echelons. They instead empower the command and control structures of the force to deal with uncertainty, react to change and recognize and exploit opportunities, which is the domain of battle and the essence of military art.

In my view, based on the evidence I have seen in our studies over the past seven years, every echelon between corps and platoon adds value and removing one incurs serious costs in the qualitative aspects of the contest against a thinking and adaptive enemy. The focus of the discussion should not be whether we eliminate echelons, but how to achieve several other more important objectives: how to re-structure the system of echelonment to make it more strategically responsive, achieving a balance between the demands of teamwork and the demands of adaptability; how to achieve a new integration of combined arms functions in depth within the Army power projection organizational construct and within a more integrated joint framework; and how to enhance the Army's competitive edge among the services -- its ability to field and support more infantry squads and fighting vehicle crews within whatever manpower and resource limitations may be imposed.