and initial entry training. The US Army (my background) does indeed have a third wave, industrial approach to basic/AIT. I think we actually do a fair job of turning civilians into soldiers, but it is pretty much the same approach used during the two global wars: A small number of trainers with relatively limited resources use the assembly line approach to produce soldiers with a specified skill set. These are then fowarded to units where they have to have additional training before they are fully qualified. The training involves three major areas:

Socialization
Physical Fitness
Military skill (over 240 discrete skills for a Cavalry Scout, for instance: map reading, marksmanship, hand grenades, etc)

What is the problem with this logical, ordered approach to basic training? It is three-fold:

1. Trainer to trained ratio is too high. In some cases this is not a big problem. One man can train 100 men in physical fitness about as well as he can train 10. He can't do the same in, say, rifle marksmanship. As a result, a huge amount of the recruit's time is spent sitting around, waiting his turn on the range, or with the radio, or for his turn in the simulator. It is also harder for the trainers to detect and correct individuals having problems.
2. Training to the lowest possible standard. There is virtually no scope for allowing talented soldiers to learn at a faster pace. Budding SGT Rocks are kept back and forced to keep pace with all the Sad Sacks in their group. About the only area where they can get ahead is physical training. Again, this is partly due to a lack of resources dedicated to basic training, and also to the industrial training model, which values mass production over nurturing.
3. The system is ossified. When I left the basic training arena in 2002, we were still training recruits on React to Nuclear Explosion. We were not training them on many of the skills that would become important in Iraq and Afghanistan. Deleting or adding a task in the basic skill set took a minimum of two years to work its way through the various TRADOC hoops. Oh, we did what we could; we cobbled together a 'booby trap' lane, for instance, but on our own time and with resources we could scrape together.

The solution is not hard to imagine. We have more than enough motivated, talented NCOs to train recruits; they know how to train and what to train; they know what you need to train; they know how to turn lardies into studs; they know how to motivate and socialize. Imagine Gunny Highway back from Iraq for a two year tour as a basic trainer. Every four months he is handed 12 recruits and told to produce a ten-man infantry squad he would be willing to take into combat. He has access to all the equipment they would need and sufficient land and range time. He has sole authority to drop soldiers from training or recycle them. At graduation he determines what rank the soldier deserves, up to E-4. Something like that would work.

Here are the problems:

1. This would require a huge increase in the number of high-quality soldiers dedicated to training. The industrial model was chosen, after all, because it allowed for the smallest possible investment in trainers and training resources - efficiency, not effectiveness.
2. There would be some spectacular failures. When you decentralize the operation as depicted, there will be a small percentage of trainers who will turn out to be lazy, unproductive, sadistic, or simply crap. Some of these failures will involve dead, injured, or abused soldiers, along with varying degrees of property damage. Some, inevitably, will be highly publicized.

Bottom line is we will not improve basic training until we invest a great deal more resources in it and are willing to accept risk and the inevitable failures. I don't think it is a question of some new approach to training per se.