A controversy appears to be brewing over the U.S. Army's plan to move away from the division as its basic unit and toward the brigade, or Brigade Combat Team as the Army buzzwords it. On the surface, it appears there should be little to argue about. Most other armies abandoned the division or downsized it long ago, recognizing that it is simply too big to be commanded effectively on dispersed modern battlefields.

The controversy, it seems, is less over the move to brigades than over the question of how many maneuver battalions the Army will have left once the reorganization is complete. Here, the answer is the usual answer where numbers are concerned: it depends on what you count. An IDA study says maneuver battalions are cut by 20%, which if true, is certainly a bad move. The Army's leadership responds that IDA is not counting the recon battalion in each BCT, which is also a maneuver battalion. That may or may not be true, depending on the military situation. Like combat engineer battalions, reconnaissance battalions are sometimes used just like other maneuver battalions, because the situation demands that everyone be thrown into the fight. When the demand for cannon fodder is less intense, however, commanders usually want to avoid using units with special skills as infantry, because soldiers with special skills are harder to replace.

Far more serious than the question of whether recon battalions are or are not maneuver battalions is the matter of creeping headquarters' growth. The IDA study found that with the new BCT organization, brigade headquarters grew by about 11%. I met with the Army's "transformation task force" on force structure twice, and my strong impression from those meetings was that headquarters grow both in number and in size.

Why is this a problem? Because more headquarters and larger headquarters inevitably mean more centralization. Centralization is one of the key characteristics of Second Generation militaries, just as decentralization is a defining quality of the Third Generation. Decentralization permits outward focus and encourages initiative, which in turn together speed up Boyd's OODA Loop and improve accuracy of orientation. Centralization, in contrast, slows the OODA Loop down and blurs orientation because the picture that is the basis for decisions is many layers removed from the actual observation.

One of the reasons none of America 's armed services has yet transitioned from the Second to the Third Generation is the vast number and size of their headquarters. All those headquarters' officers are continually looking for something to do, and for some scrap of information that will give them 30 seconds of face time in the endless Powerpoint briefings that are American headquarters' main business. The result is that they impose endless demands on the time and energy of subordinate units. One Army battalion last year told me they had to submit 64 reports to their division every day.

Here we come to the central question, not only about the Army's move from divisions to brigades, but about its whole "transformation" program: is it reform, or is it just reorganization? To count as real reform, it needs to move the Army out of the Second Generation and into the Third. If all it amounts to is reorganization within a Second Generation framework, then, frankly, it's not worth the umpteen-thousand Powerpoint slides it's printed on.

If the Army's senior leadership wants reform and not mere reorganization, here's a suggestion to move the "transformation" process in that direction. Order that at the end of the day, when the new BCT structure is in place, the Army may have only half as many officers in headquarters (at all levels) as it did under the previous structure. And no, the officers cut may not be replaced by contractors. That would at least encourage decentralization, without which no reform is possible. It might also give however many maneuver battalions the Army ends up with a little room to breathe.